This page intentionally left blank
Social Action Most human action is social action of some sort. Yet the analysis of...
42 downloads
4452 Views
949KB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
This page intentionally left blank
Social Action Most human action is social action of some sort. Yet the analysis of social, as opposed to individual, action has been somewhat neglected by philosophers. This book provides philosophical analyses of fundamental categories of human social action, including cooperative action, conventional action, social norm-governed action, and the actions of the occupants of organisational roles. A distinctive feature of the book is that it applies these theories of social action categories to some important moral issues that arise in social contexts, such as the collective responsibility for humanitarian intervention and the collective rights of minority groups. Avoiding both the excessively atomistic individualism of rationalchoice theorists and implausible collectivist assumptions, this important book will be widely read by philosophers of the social sciences, social philosophers, ethicists, political scientists, and sociologists. Seumas Miller is Director of the Centre for Applied Philosophy and Public Ethics, Charles Sturt University.
Social Action A Teleological Account
SEUMAS MILLER Charles Sturt University
The Pitt Building, Trumpington Street, Cambridge, United Kingdom The Edinburgh Building, Cambridge CB2 2RU, UK 40 West 20th Street, New York, NY 10011-4211, USA 477 Williamstown Road, Port Melbourne, VIC 3207, Australia Ruiz de Alarcón 13, 28014 Madrid, Spain Dock House, The Waterfront, Cape Town 8001, South Africa http://www.cambridge.org © Seumas Miller 2004 First published in printed format 2001 ISBN 0-511-03114-9 eBook (Adobe Reader) ISBN 0-521-78316-X hardback ISBN 0-521-78886-2 paperback
In Memory of My Father Crawford Miller (1913–2001) Who Introduced Me to Philosophy
Contents
page ix
Acknowledgments
1 2 3 4 5 6 7 8
Introduction Theories of Social Action Joint Action Conventions Social Norms Organisations, Agency, and Action Social Institutions and Social Groups Collective Rights Collective Responsibility
1 18 53 91 123 160 181 210 234
Notes Bibliography Index
259 289 299
vii
Acknowledgments
The following have offered helpful comments on various parts of the manuscript: Andrew Alexandra, John Blackler, Tony Coady, Dean Cocking, Tom Ernst, Richard Freadman, Allen Hazen, David Lewis, Ian Macdonald, Pekka Makela, Steve Matthews, Crawford Miller, Ewen Miller, Kaarlo Miller, James Nickel, Raimo Tuomela, Kerry Zubrinich, and various anonymous referees, including from Cambridge University Press. Thanks also to Professor John Braithwaite and the Reshaping Institutions Project for providing a visiting fellowship at the Australian National University to undertake work on parts of the manuscript; thanks likewise to Professor Henry Shue and the Centre for Ethics and Public Life at Cornell University. The following articles and books contain or constitute earlier versions of material published in this book. Thanks to the editors and publishers involved. “Lewis on Conventions.” Philosophical Papers 11, 2 (1982). “Conventions and Speech Acts” (University of Melbourne Ph.D., 1985, Ann Arbor: University Microfilms International, 1986). “Conventions, Interdependence of Action and Collective Ends.” Nous 20, 2 (1986). “Conventions, Expectations and Rationality.” Southern Journal of Philosophy 25 (1987). “Rationalising Conventions.” Synthese 84, 1 (1990). “Foucault on Discourse and Power.” Theoria 76 (1990). “On Conventions.” Australasian Journal of Philosophy 70, 4 (1992). “Joint Action.” Philosophical Papers 21, 3 (1992). ix
Acknowledgments Rethinking Theory: A Critique of Contemporary Literary Theory and an Alternative Account (co-author Richard Freadman). Cambridge: Cambridge University Press, 1992. “David Luban on the Obligation to Obey the Law.” Bulletin of the Australian Society of Legal Philosophy 18, 60 (1993). “Intentions, Ends and Joint Action.” Philosophical Papers 24, 1 (1995). “Socio-Political Action, Ethics and the Power of Literature.” Philosophy and Social Criticism 21, 3 (1995). “Corporate Crime, the Excesses of the 80’s and Collective Responsibility.” Australian Journal of Corporate Law 5, 2 (1995). “Social Norms,” in Ghita Holmstro¨m-Hintikka and Raimo Tuomela (eds), Contemporary Action Theory, Volume 2: Social Action. Dordrecht: Kluwer, 1997. Police Ethics (co-authors J. Blackler and A. Alexandra). Sydney: Allen & Unwin, 1997. “Individualism, Collective Responsibility and Corporate Crime.” Business and Professional Ethics Journal 16, 4 (1997). “Collective Responsibility, Armed Intervention and the Rwandan Genocide.” International Journal of Applied Philosophy 12, 2 (1998). “Authority, Discretion and Accountability: The Case of Policing,” in C. Sampford, N. Preston, and C. Bois (eds), Public Sector Ethics: Finding and Implementing Values. London: Routledge, 1998. “Collective Rights.” Public Affairs Quarterly 1, 4 (1999). “Social Norms and Practical Reason.” Educational Philosophy and Theory 31, 3 (1999). “Collective Rights and Minorities.” International Journal of Applied Philosophy (Fall 2000). “Collective Responsibility.” Public Affairs Quarterly 15, 1 (2001). “Collective Responsibility and Omissions.” Business and Professional Ethics Journal 20, 1 (2001). “Social Institutions,” in M. Sintonen and P. Ylikoski (eds), Realism in Action: Festschrift for Raimo Tuomela. Dordrecht: Kluwer, 2001. Special and more personal thanks go to the following: Ewen Miller for preparing the bibliography and the index and for his work in x
Acknowledgments relation to editing and proofing and my father, Crawford Miller, who first introduced me to philosophy and whose intellectual input over many years was very important in shaping my philosophical outlook. I dedicate this book to his memory.
xi
Introduction
If a hungry, thirsty, itchy person eats some food, or drinks some water, or intentionally scratches an itch, this is individual action. That is, it is the intentional behaviour of an ordinary individual human person; other human beings are not necessarily involved.1 If a person takes a walk down the road by herself for exercise, or eats an ice cream for pleasure, or takes a shower on a hot day, this is also individual action. Such action is not action in cooperation with, or necessarily directed at, other individuals. On the other hand, if an individual kicks or throws a football to a teammate in the course of a game of football, or puts a motion forward at a committee meeting, then this is social action. Most human action is in fact at some level, or to some extent, or in some sense, social action. Even these actions of eating, drinking, eating ice cream, individually walking down the road, or having a shower typically presuppose social forms or objects, such as farms, ice cream parlours, cups, roads, and shower rooms. But the de facto presupposition of social forms does not of itself vitiate the distinction between individual and social action. More generally, the sociality of most human action does not vitiate the distinction; it merely serves to illustrate the need to develop a more elaborate set of distinctions in this area. There is nothing to be gained from insisting that no actions are individual and/or that all actions are social, just because it might in fact be that all the actions of human beings connect in some way, however indirectly, with the actions of other human beings and with social forms and social objects. Of course it might be argued that all action is necessarily social, because sociality is a logical presupposition of action.2 But this strong claim is implausible. We need the concept of an action – and specifically that of an individual action – in order to make sense of the concepts of sociality, rather than concepts of sociality to make sense 1
Social Action of the concept of action. Indeed, this book is an attempt to fashion a particular individualistic concept of human action, and to use it as a building block in the construction of a variety of philosophical conceptions of social action types, including conventional, normgoverned, and institutional action. This process is in two stages. First, I help myself to an individualistic concept of action for the purpose of developing a philosophical account of a particular species of action, namely, joint action, where I understand joint action to be a species of interpersonal action, but not necessarily social action. I will say a good deal more about joint action, but, roughly speaking, a joint action comprises the actions of two or more individuals directed to a shared end. So members of a rowing team are engaged in joint action, as are two men moving a piece of furniture, or a group of thieves robbing a safe. Second, I use this notion of joint action to analyse a range of central kinds of social action, including conventional, norm-governed, and institutional action. In addition, I use the notion of joint action to illuminate some fundamental issues in collective morality, including the issues of collective rights and collective responsibility. Social actions are the actions of ordinary individual human persons. These include the actions of individuals performed in accordance with conventions, rules and norms, and the actions of individuals qua occupants of social, institutional, and professional roles. Some theorists claim that the category of social actions includes the “doings” of corporate entities such as governments or nations; for example, the United States declaring war on Iraq, or the Australian Federal Government introducing a new tax, such as the goods and services tax. I do not accept this claim. Accordingly, in this book I will, in effect, be concerned only with the actions of ordinary individual human beings.3 Granted that my concern is only with the actions of individual human beings, and that the concept of an individual action is logically prior to that of a social action, there is still a need for distinctions to be made between social action and other sorts of action of a nonsocial (or contingently social) sort and clarification of the concept (or concepts) of social action itself. In what follows I will introduce some intuitive distinctions between what I will term social actions and other sorts of action. The point of this is in large part to
2
Introduction demarcate my area of concern in this book. I cannot here offer exhaustive defences of these distinctions. Actions that are not necessarily social include individual actions (as already mentioned). Of these some are what might be termed natural actions. A natural action is one that is performed by virtue simply of needs and dispositions that the agent has through being a member of the human species as distinct from, say, some social group. Obvious examples are eating and drinking. Eating and drinking are not actions that logically presuppose, or logically imply, social forms. Another category of actions that are not necessarily social, or, if you like, that are social in a different sense of social, are what I will term interpersonal actions. An interpersonal action is an individual action that is interdependent with the action of some other single person, or is otherwise directed to a single person. Here the action is directed to the other person qua particular person, or qua member of the human species; it is not directed at the other person qua member of a social group or occupant of a social role or the like. So the contrast is with actions that are performed in accordance with a social form, or are directed to a number of other persons qua members of a social group, or actions directed to a single person qua member of some social group or occupant of a social role or the like. Typically, sexual acts or acts of intimate friendship or the behaviour of a newborn infant in relation to his or her mother are predominantly interpersonal actions in this sense, but institutional acts of conferring degrees or conforming to conventions of dress are not. Moreover, some of these natural, interpersonal actions are also moral actions. Consider actions motivated by instinctual feeling of sympathy for a fellow human being qua human being – as opposed to qua fellow member of one’s social group. Interpersonal actions presuppose the existence of that relationship that obtains when, so to speak, one mind confronts another mind. Such “confrontations” are everyday occurrences, but paradigmatic examples are situations in which one person is said to look the other in the eye. Here one person is aware of the other person, including being aware that the other person is aware of them. Such mind to mind interactions need to be distinguished from, on the one hand, mind to own mind (introspective) and mind to material
3
Social Action world (for example, perceptual) interactions, and, on the other hand, from mind to social world (social) interactions. Following C. D. Broad, I will call such mind to mind interactions “extraspective” interactions.4 In claiming the existence of extraspective interactions I am not committing myself to any particular analysis or theory of them, nor am I committing myself to the existence of controversial species of extraspective interactions, such as telepathy. Extraspective interactions are a perfectly ordinary phenomenon, albeit one that unfortunately has suffered neglect in recent times by philosophers.5 Indeed, most people are more certain of the existence of, and their interaction with, other minds than they are of any other sort of entity or interaction. And there is good reason for this. Extraspective states are not logically posterior to mind to material world interactions. Or at least they are not logically posterior in the manner required by the so-called argument from analogy. The argument from analogy assumes that my knowledge of the external material world (as well as my knowledge of the contents of my own mind) is logically prior to my knowledge of other minds. Accordingly, I must infer the existence of other minds on the basis of this prior knowledge of the external material world (taken in conjunction with my prior knowledge of own mind). Extraspective interactions are the most basic species of interpersonal interaction. However, interpersonal interactions include interactions that are only indirectly extraspective since there are interpersonal interactions between individuals who are not co-present in space and time, for example interactions by telephone or electronic mail. In these latter cases the extraspective interaction is between agents who are spatially and/or temporally distant from one another, and the interaction is mediated by, say, telephone wires. Indeed, extraspective interactions are so basic that they exist in the subhuman animal world. Dogs, tigers, pigs, and so on engage in extraspective interactions by virtue of their possession of primitive minds; these animals have mental states and a degree of consciousness, or awareness, of the external world and of one another. On the other hand, computers and the like do not have minds and therefore do not engage in extraspective interactions. (I reject the currently popular view – typically based on exclusively causal analy-
4
Introduction ses of mental states – that computers have beliefs, intentions, and the like.) I will assume that not all individual human actions directed at other persons qua entities with minds are interpersonal action. For example, if I intentionally hit you in order to harm you, and do so when you are immobilised by a drug (and therefore unable to respond), then this is individual, but not interpersonal, action. Moreover, it is individual action, even if you are aware that I have hit you. This assumption of mine is stipulative, but not unmotivated. It is motivated by the thought that we ought to reserve the term “interpersonal” for cases in which two or more agents actually perform different, but interdependent, actions; or at least for cases in which one agent performs an action with the intention that the other (or others) perform a second action by way of response. From the point of view of this book the most important species of interpersonal interaction is joint action: action that involves two or more agents performing individual action in the service of a shared end. And, as I will argue, joint actions are not necessarily social actions. The picture that has emerged thus far comprises the following categories of human action: (1) individual actions, that is actions involving mind to material world interactions, and actions involving mind to mind interactions that are not responded to, or not intended to be responded to, or both; (2) interpersonal actions, that is actions that involve mind to mind interactions, and are not simply individual actions, such as joint actions; (3) natural actions, that is individual or interpersonal actions which are performed by virtue simply of needs and dispositions that the agent (or agents) has through being a member of the human species.
What now of social actions? Roughly speaking, mind to social world interactions take place when one or more individual actors interact with, or direct their actions to, other individual actors (who might or might not be co-present), but do so qua parties to a convention or social norm, qua occupants of an institutional role, or qua mem-
5
Social Action bers of a social group or other social form. In other words, social actions are human actions performed in accordance with social forms such as conventions, social norms, institutions, social groups, and the like. The philosophical analysis of some of the central kinds of social action is the main concern of this book. My philosophical analysis of each of these social action categories will in each case be a reductive analysis of a certain kind. Social actions are reducible to various species of interpersonal action. For example, conventional actions will be analysed in terms of joint actions. However, these reductive analyses will not involve the reduction of social entities and the relationships between social entities. So the fact that irreducible social entities might figure in the intentional content of mental states – mental states that are in part constitutive of social and other actions – is in itself not a matter of concern. Indeed, on the conception presented here, it is to be expected. What is a matter of concern is the view held by some theorists, including Peter French and Margaret Gilbert, that social entities are themselves agents that possess mental states and perform actions. For such actions, if they existed, would be social actions, and irreducibly so. I will counter the arguments of these theorists. Social actions involve a certain kind of interdependence of action and of attitude between the members of sets of individuals, and this interdependence in turn creates the possibility of social groups. The existence of social groups, thus understood, in turn enables a certain new kind of social relationship that is not to be identified with interpersonal relationships (in the sense of mere one-to-one relationships between individuals), but rather presupposes them. This new kind of social relationship is a relationship between an individual and “the rest of the group.”6 This relation of an individual to the rest of the group arises in the context of interdependence of action and of attitude between all (or most of) the members of the group, including the individual in question. So each individual is separate from, but related to, the rest of the group. Moreover, this social relationship is not in any straightforward way an aggregate of oneto-one relationships obtaining between the members of a set of individuals. In the context of a social group, and of this social relationship between an individual and the rest of the group, the actions of individual persons can be powerfully influenced by social
6
Introduction attitudes, such as social approval and disapproval. So social phenomena, including social attitudes, do have causal impact on individual and interpersonal action. I have distinguished between two fundamental species of human action, namely individual and interpersonal action. And I have offered characterisations of natural and social actions. The terms “natural” and “social,” as I use them, are contrasting terms. However, they both presuppose the concept of action, interpersonal as well as individual action. So we have at least three kinds of action: natural individual action; natural interpersonal action; and social interpersonal action (social action). What of social individual action? First order natural individual action and first order natural interpersonal action initially give rise to social action as prior nonsocial action governed by social forms; this social action comes into existence as second order action. Consider convention-governed sexual interaction. There is the further interaction between different (n order) social actions which gives rise to (n⫹1 order) social actions as prior social action governed by higher order social forms. For example, conventions that govern action that is already conventiongoverned, such as conventions governing one’s linguistic communication at a formal ceremony. Since social action necessarily involves mind to mind interactions it is necessarily interpersonal action. So prior natural individual actions, such as eating, take on an interpersonal, indeed social, aspect once they are regulated by, say, a convention to eat with a knife and fork. However, it is important to note that the prior individual action still exists, so to speak, at the core of the newly existent social action. So it is individual action with a social aspect; it is, so to speak, social individual action. And the same point can be made in relation to prior interpersonal actions that are regulated by social forms; a sexual act still exists at the core of a marital sexual act. So marital sexual actions are social interpersonal actions. So we have at least four kinds of human action: natural individual action; natural interpersonal action; social interpersonal action; and social individual action. This fourfold distinction reflects the fact that all human actions are in the first instance either individual or interpersonal actions; natural and social are contrasting qualifications of individual and of interpersonal action. However, it can be
7
Social Action confusing to refer to action as social individual action or as social interpersonal action. So I will refer to the latter two kinds of human action simply as social action. Someone might want to insist that interpersonal actions per se are a species of social action. I do not have any strong objections to people arguing this, so long as they are not insisting that what I am calling interpersonal action logically presupposes or implies what I am calling social action. In particular, just as mind to material world (for example perceptual) interactions cannot be reduced to social interactions, so mind to mind (extraspective) interactions and natural interpersonal interactions cannot be reduced to social interactions. Indeed, social actions, in my sense, presuppose extraspective interactions, including natural interpersonal interactions (as well as mind to material world interactions). (It is, of course, also true that many mind to material world, and mind to mind, interactions are to an extent socially conditioned.) At any rate, henceforth I will use the term “social action” to designate those actions types that I want to contrast with natural individual actions and with natural interpersonal actions. Accordingly, I assert that not all human action is necessarily social action. With respect to actions that I am now calling social actions, a distinction is often made between actions that are (allegedly) constitutively social, and actions that are social in some other sense or senses. Roughly speaking, a constitutively social action is an action the social dimension of which makes it the action that it is; its social property or properties define it. It is a matter of controversy whether there are any actions that are constitutively social in this sense. Candidates for being constitutively social would be actions performed in highly formalised settings such as wedding ceremonies, debutante balls, law courts, and trophy presentations. I do not want to become embroiled in this controversy beyond making two points. First, I reject the proposition that there is a category of constitutively social actions in the sense of actions that are wholly – as opposed to, in large part, or essentially – constituted by social properties. Second, even if there is a category of constitutively social action, it is not nearly as important a category as the writings of many (including postmodernists and social psychologists) make out. As I have demonstrated, not all actions are social. Moreover, of those that are social, not all are constitutively social. Obviously the 8
Introduction above-mentioned nonsocial action types are not constitutively social. However, most concrete actions of these types are social in some other way. What other way might this be? Most actions of eating, drinking, and having sex, for example, are in fact social in some sense, although they are also at some level natural actions. The most important sense in which an action might be social is that it is socially conditioned. By the term “socially conditioned” I do not mean that the action is necessarily the effect of some deterministic causal process. Nor do I mean that there are supraindividual social entities that causally determine the actions of individual human beings. Rather, I simply mean that the action is performed in accordance with some pre-existing social form, such as a convention, or social norm, or ritual, or social role, or social group, or socially given purpose, or whatever. An action that is social in this sense is not necessarily constituted by its social dimension. Rather, in the case of an action that is socially conditioned (in this sense), a nonsocial action takes on a social aspect. Or, at least, a social action takes on an additional social aspect. But if we remove the social aspect, or aspects, of any given such action and thereby conceive of the action prior to its social conditioning, then ultimately we will come to a nonsocial action. Some examples might give an intuitive sense of what I have in mind. Eating with one’s mouth closed because of the convention to keep one’s mouth closed while eating is a case of social conditioning. The basic and prior action of eating is not social. However, the action of eating with one’s mouth closed is social in the sense that the way of performing it is governed by a convention, and this convention might serve the social purpose or ends of social facilitation and bonding that conventions of politeness typically serve. In a second example, two members of a particular society having sex involves social conditioning. The basic instinctual action of having sex is not in itself social; rather, it is natural. (It is also typically or often interpersonal rather than individual; it is interpersonal by virtue of the fact that it is directed at another individual person qua particular person and qua member of the male or female gender.) However, when two members of the same social group have sex their action is typically regulated and structured in various ways by the conventions and norms in force in that social group, and by observing these conventions and norms various social purposes or 9
Social Action ends will no doubt be served. Most actions are social, and of the actions that are social, most are social in the sense that they are conditioned by social forms of one kind or another. On this conception most actions are in fact social, but natural individual, and natural interpersonal, actions are logically prior to social actions. The social dimension principally consists in the regulation, but not the constitution, of prior individual, natural, and interpersonal actions. Moreover, the interpersonal actions in question include joint actions. There is a further important point. Many actions governed by social forms are, nevertheless, not fully determined by those forms. For example, the conventions of the English language dictate that strings of English words be ordered in certain ways and not in others. But these conventions do not fully determine which words will be used. Rather, individuals can choose which sentence to utter and choose from an infinity of possible sentences. The set of structures of social forms that constrain, but do not fully determine, human action constitutes a framework. Within this framework individual human beings can perform individual and interpersonal actions of their own choosing; they do so while continuing to comply with the relevant social forms. I do not claim that the category of mind to material world interactions is logically prior to the category of mind to mind interactions; that is, I do not claim that individual actions are logically prior to interpersonal actions in general. Nevertheless, I do claim that individual actions – including mind to material world and some extrospective interactions – are logically prior to joint actions in particular (joint actions being a species of interpersonal interaction). Notwithstanding the logical priority of individual actions over joint actions, individual ends and interests are not explanatorily prior to collective ends and interests. So the existence of collective ends is not necessarily to be explained (causally or rationally) by recourse to prior purely individual ends; nor is the motivation for the performance of joint actions necessarily ultimately to be given by recourse to some purely individual self-interest. Accordingly, I am not an atomist, but rather (in some sense) a holist;7 interpersonal interaction – including joint action – and interpersonal relationships exist, so to speak, at the ground level of explanation. 10
Introduction On this conception, then, social forms are constructions out of prior nonsocial (especially natural) individual and interpersonal actions; but once constructed these social forms become a regulating, though not fully determining, framework within which higher order individual and interpersonal human actions are performed. To be sure, these higher order actions have a social aspect; but they are individual and interpersonal actions nonetheless. So individual human beings perform their individual actions, and pursue their personal goals, within a social framework of conventions, social norms, and institutions. And individual human beings also perform their interpersonal actions – including joint actions and competitive actions – and also pursue their interpersonal goals within such a social framework.8 Moreover, the social regulation, adjustment, and structuring of prior nonsocial individual and interpersonal – including joint – actions enables the possibility of higher level thought and other rational activity, including moral deliberation. Individual human beings are agents who can think and act in rudimentary ways simply by virtue of their natural inclinations and intuitions, including their moral instincts and intuitions. However, it is their induction into the social world of conventions and norms – including the conventions governing language and the norms governing moral behaviour – that enables higher level thought and moral decision making. I do not mean to suggest that human beings do not at birth have the basic capacity for such higher level thought or moral decision making prior to induction into the social world; clearly humans have these and other capacities, while subhuman animals do not. My point is rather that these capacities require such induction if they are ever to be realised in practice; it is one thing to be born with a capacity to learn a human language, it is another to actually learn one. The conception I am offering is also in some sense collectivist in complexion; individual and joint activity, if they are to exist above very rudimentary levels, depend on social forms. By contrast, the individual and joint activity of subhuman animals remains at rudimentary levels because such activity continues to be driven only by natural instincts. Because subhuman animals lack the requisite capacities for higher level thought and moral decision making, enabling social forms do not, and cannot, develop. Consider a pack of 11
Social Action hyenas attacking a large zebra to kill it and eat it. No one hyena acting alone could kill the zebra. So not only are their individual actions caused by beliefs and intentions (or some primitive mental analogues of beliefs and intentions), but their actions are coordinated in the service of a shared purpose – to kill the zebra. This is both natural action and primitive joint action. But it is not social action in my sense. For it is not action regulated and structured by conventions, norms, institutions, and the like; and it is not action performed in accordance with higher level thought or moral decision making. Here again the contrast is between the social and the natural. Further, the social forms and objects that condition (in my sense) the actions of individual persons predate and post-date the actions, interactions, and indeed lives, of particular generations of individual persons. Moreover, social forms and objects undergo change by virtue of the joint participation in them over long periods of time of different particular individuals and different sets of interpersonally connected individuals, by virtue of changes to other connected social forms and objects, and by virtue of the different nonsocial conditions, including physical conditions, through which they persist. Accordingly, individual persons are not simply inducted into a social world; they are inducted into a sociohistorical world. This makes my conception historicist in some sense, but not necessarily historical determinist. So in what sense is my conception historicist? Individuals are inducted into the social forms of the past – including socially communicated theories, quasi-theories, and moral narratives. But these social forms do not constitute a monolithic structure; rather they comprise a miscellany of sometimes competing conventions, norms, institutions, and socially conditioned theories and narratives. The residue of the past consists in more than social forms; it also contains the memories and handed down skills that derive from the individual – as opposed to collective – lives of past generations. For example, a mother might have had personal moral experiences particular to herself, which she might make known to her daughter, but not to others. And there is this further point. Individuals do not confront the residue of the social forms of the past as atoms; they confront it – or rather participate in it – jointly. It does not follow from this that any 12
Introduction given generation of individuals can simply abandon these social forms; far from it. However, it does follow that these social forms are to a greater or lesser extent subject to change, and in some instances rejection; indeed, they are often changed and rejected in accordance with more or less rational processes. At any rate, these processes of change do not involve actions other than the actions of individual human beings. Also, these processes consist in large part of the joint activity of individual actors. An important corollary of the conception I am offering is that joint activity can take place over an extended period of time – including intergenerationally – and not simply over space. The building of the Great Wall of China and of cathedrals in the Middle Ages, the creation of the Roman Empire, the development of Western philosophy and of the literary form of the novel are in each case intergenerational joint projects. The jointness of the activity – including intellectual activity – exists across space and across extended periods of time. I would like to foreshadow here the teleological element in the account of social action that I will develop in later parts of this book. As I have mentioned, I start with the notion of an individual human action and the elements of intention, belief, end, and behaviour. Here the notion of an end is crucial; in my view ends are not reducible to intentions, on the one hand, or to desires, on the other hand; ends are an irreducible conative notion. These elements can be used to provide a philosophical account of a species of interpersonal action, namely, joint action, and in turn this notion of a joint action can be used to provide philosophical analyses of a variety of central types of social action. Roughly speaking, joint actions are comprised of individual actions performed to realise a shared end; so my conception is teleological. The end involved in joint actions is a shared or collective end; it is not simply an individual end. The possession of a collective end is not necessarily to be explained in terms of some prior motivating individual end; motivationally speaking, collective ends are sui generis. Neither individual action (directed to individual ends) nor interpersonal action (often directed to collective ends) is necessarily explanatorially basic relative to the other; as I have explained, human beings enter the world not only capable of performing, but also predisposed to perform, both types of action. 13
Social Action In proceeding in a general teleological vein I am to some extent following the rationalist, individualist, philosophy of action tradition which has its roots in Aristotle, Hume, and Kant and is associated with contemporary analytic philosophers such as Elizabeth Anscombe, Paul Grice, and Donald Davidson and, more recently – and in the context of discussions of social action – Michael Bratman, John Searle, and Raimo Tuomela. However, this way of proceeding also has a place outside philosophy, in sociological theory. Broadly speaking, it is the starting point for the voluntaristic theory of action associated with the likes of Max Weber and Talcott Parsons,9 and for the rational choice theory associated with the likes of James Coleman.10 For example, I would accept the following idea expressed by Parsons: “actions do not take place separately each with a separate, discrete end in relation to its situation, but in long complicated ‘chains’. . . . [and]the total complex of means–end relationships is not to be thought of as similar to a large number of parallel threads, but as a complicated web (if not a tangle).”11 On the other hand, my philosophical account of social action differs significantly from these theories. For example, Parsonian functionalism is too committed to notions of supra-individual functions of closed, self-regulating social systems, for my taste, and rational choice theory is too rationalist and atomistic. But the starting points for these theorists are similar to my own and to the philosophical tradition just mentioned. My conception of social action is informed by moral objectivism. In my view morality is not simply a social construction, though many social forms, such as social norms, have a moral dimension. Rather, the fact that social forms typically have a moral dimension is in part to be explained by the fact that there is a moral reality (in some sense) which to a greater or lesser extent informs particular social forms, just as to a greater or lesser extent it informs individual actions. That social forms have a moral dimension is also in part to be explained by the fact that human beings are beings with moral instincts and moral intuitions; human beings are born with moral instincts, even if these instincts exist in an inchoate form. Finally, that social forms have a moral dimension is in part to be explained by the fact that human beings are rational beings. For moral reality
14
Introduction is only able to be adequately grasped by beings who engage in reasoning. So there is a threefold distinction to be made between beings with minds – and agency – in a minimal sense, instrumentally rational beings and moral agents. Both humans and dogs have minds and perform actions, but only humans engage in reasoning, and only humans can grasp moral reality. So only humans are rational and moral agents. I should stress, though, that while instrumental rationality is a necessary condition for moral agency, it is not a sufficient condition. Without moral instincts or beliefs there is nothing for instrumental rationality to get a grip on; the ends to which the actions of merely instrumentally rational agents are directed are not necessarily morally valued. So I reject the claim that the concept of action, including purposive action, in and of itself, entails morality.12 The existence of rational psychopaths adds support to my view. In this book I am concerned with human social action, as opposed to social entities. And even within social action my aims are limited. I do not have the space to provide detailed analyses of all the various categories of social action; for example, I do not offer a detailed account of speech acts. However, the standpoint I would adopt, or have already adopted, in relation to the various categories of social action will be clear from my general account of social action. In relation to speech acts, for example, I have adopted the standpoint of agent-semantics associated with the work of Paul Grice.13 Agent-semantics involves the conceptual construction of linguistic (conventional) meaning from so-called speaker-meaning (what an individual speaker means by uttering something on some occasion). Note that on my account speaker-meaning is a species of interpersonal interaction and that intersubjective communication involves – or in the case of communication between speakers and hearers who are not co-present, presupposes – extraspective interaction. I am also concerned with social action per se, as distinct from, for example, the explanation – including the causal explanation – of social action or (so to speak) the epistemology of social action. My concern is with what social action consists in; it is not with the ultimate reasons why some particular kind of social action is performed under certain conditions or with whether or not there are discov-
15
Social Action erable causal laws that serve as explanations for such actions. Accordingly, I am insisting on a distinction that sometimes gets blurred in the writings of, for example, rational choice theorists. Some such theorists proceed on the assumption that the analysis of social action is simply the analysis of individual action supplemented by the provision of rational justifications for individual actions in contexts involving other individuals. Here the rational justifications of these actions are supposed to serve as the explanations for them. Nor is my concern with how social scientists are to determine what particular social actions are in fact being performed in some society under investigation. In insisting on a sharp distinction between what social action consists in and how we might go about determining what social actions have in fact been performed, I am rejecting the views of certain postmodern theorists who in effect collapse the distinction between the object or action in itself and the knower’s, or interpreter’s, beliefs about, and procedures for acquiring knowledge of, the object or action.14 I am not, however, rejecting in toto subjectivist accounts of social action of the kind espoused by Max Weber15 and, more recently, by John Searle;16 or even more explicitly hermeneutic conceptions, such as the “humans as self-interpreting animals” notion of Charles Taylor.17 Searle’s account combines what he calls ontological subjectivism with epistemic objectivism. On this view, social facts typically exist by virtue of the collective imposition of functions on preexisting objects or persons. For example, a log becomes a bench by virtue of being used as such, and a person becomes married by virtue of uttering the performative “I do” in a certain context. So social facts are ontologically subjective; they exist insofar as members of a social group believe or act as if they exist. Yet social facts are epistemically objective, since whether or not a log is a bench or whether or not someone is married are matters of truth or falsity to be settled by recourse to the (social) facts in question. The truth of the claim that a log is a bench or that someone is married is not dependent on the belief of the person making the claim. My own view is that there is a category of social facts that have the kind of ontological status that Searle ascribes to all (what he terms) institutional facts,18 but that many categories of social facts – including facts about many categories of social action – do not. For example, on my account conventional actions are simply regularities in 16
Introduction individual actions directed to the realisation of certain kinds of ends. Accordingly, the parties to conventions possess mental states, including intentions and beliefs about the actions of the other parties. But from the point of view of ontological categories, or of epistemological inquiry, nothing has changed; there are physical and mental facts, and there are whatever ways of knowing are appropriate for such facts. At any rate, here I simply reiterate that my concern in this book is not with the epistemology of social action or with the ontology of social entities. That said, I agree with John Searle that the concept of social action has a certain priority over the concept of social entity.19 However, it is not part of my brief to offer comprehensive philosophical accounts of social entities or objects, whether the accounts are reductive or otherwise. This is a philosophical study of human social action.
17
1
Theories of Social Action
In the introduction I offered a brief characterisation of an intuitive notion of human social action, contrasting it with notions of nonsocial action – especially natural (individual and interpersonal) action. In this chapter I will consider some of the prevailing theoretical – especially philosophical – approaches to social action. I do so primarily for the purpose of locating my own approach in the larger philosophical context; I do not pretend to offer a comprehensive picture of theories of social action, nor will I seek to provide a complete defence of my own favoured teleological account of social action. In addition, I will briefly introduce the various different species of social action, and related normative phenomena, that are the basic subject matter of this book. social power and social cooperation As will be evident by now, the underlying conception of social action that informs this book is that of cooperative individual actions directed to (collective) ends. The model on offer is what might be termed a teleological cooperation model. Such a model stands in contrast with ones that emphasise social conflict and social power. Let me consider a social power model before looking at the contrasting social cooperation model that I wish to advocate. There is a recent and widespread tendency among social theorists – perhaps more often sociologists than philosophers – to define, in effect, all human action as the exercise of power. Anthony Giddens seems to be a proponent of this view: “Agency refers not to the intentions people have in doing things but to the capability of doing those things in the first place (which is why agency implies power: cf. the Oxford Dictionary definition of an agent, as ‘one who exerts power or produces an effect’).”1 But agency and action should not be identified with power and 18
1. Theories of Social Action the exercise of power. Rather, agents need to be defined in terms of actions, and the notion of an action is logically prior to the notion of an exercise of power. Some actions are power plays, but many are not. Consider the intentional, but gratuitous, movement of one’s little finger; this is certainly not necessarily an exercise of power in relation to someone else; it is not interpersonal or social power. Nor is it an exercise of power in relation to the external world; no recalcitrant external object or force has necessarily been overpowered just because I gratuitiously move my little finger. As far as interpersonal or social relations are concerned, consider one person asking someone else the time of day; this is an interrogative speech act, but it is not necessarily an exercise of power. Moreover, many actions resulting from power relationships are not in themselves exercises of power. This is because many such actions are acts of submission to power. Consider an academic who decides not to criticise the vice-chancellor on the grounds that it might result in his being passed over for promotion. This act of omission is the result of power, but is not in itself an exercise of power. At any rate, I take it that exercises of power are one, but only one, kind of action; action is not the same thing as power or its exercise. In some related social and political theorising there is a tendency to become fixated with the power dimension of social action in particular. On this kind of view it is not that action per se is identified with or defined in terms of power, but rather that the occurrences of particular social actions and activity are to be explained by recourse to power relationships. On this view social action is taken to be principally action driven by sociopolitical forces, including national, racial, gender, cultural, ethnic, economic, class, and institutional forces.2 We are all members of social groups and social organisations (entities), and we are somehow supposedly wholly controlled, or at least causally determined, by social forces permeating these groups and organisations. Moreover, these groups and organisations are locked in conflict with one another. Thus males are supposedly principally engaged in attempting to dominate females, capitalists to dominate workers, whites to dominate blacks, and so on. Human action is understood only, or at least principally, in terms of sociopolitical power struggles. One such influential contemporary social power theorist is Foucault. In Foucault’s view: “Power is everywhere; not because it em19
Social Action braces everything, but because it comes from everywhere.”3 The first thing to note about this is that it attempts to articulate the operation of what might be termed impersonal social power. Impersonal power is power exercised upon individuals. However, it is not power exercised by particular individuals; rather it is power exercised by institutions, social groups, and so forth. Thus, Foucault says: “Power is not . . . one individual’s domination over others or that of one group or class over others. . . . Rather power must be analysed as something which circulates . . . which only functions in the form of a chain. It is never in anybody’s hands.”4 A characteristically Foucaudian example here is the way in which the institution of psychiatry coercively categorises certain individuals as being insane. This mode of power involves a process Foucault and others term “normalisation.” Normalisation is the process by which the individual is not just categorised but also controlled by the power vested in institutions and antecedent social practices.5 Contra Foucault, not all, or even most, conformity to social forms is the consequence of coercion. Clearly, there are all manner of social practices, conventions, and institutional arrangements that arise from voluntary, rational decision making in the service of mutually beneficial outcomes. Examples here would be driving on the left-hand side of the road in countries where this is the legal requirement; casting a vote in a country where one is not legally constrained to do so; undergoing a heart bypass operation in a hospital; conforming to the convention of telling the truth; keeping one’s promises. In such instances, what is at issue is in large part the voluntary participation of rational actors in mutually beneficial, indeed morally desirable, social arrangements; not simply domination and submission. Heart bypass operations and driving on the same side of the road save lives; democratic voting procedures help to control repressive, authoritarian forms of government; general conformity to the practice of speaking the truth is necessary for human communication; general conformity to the practice of keeping one’s promises is fundamental to the maintenance of economic systems based on contractual arrangements. So the attempted reduction of social action to the exercise of power, whether it be sociopolitical power or other kinds of power, is doomed to failure. In the first place, the concept of power presupposes the concept of action, and in the second place, human actions 20
1. Theories of Social Action are implicated in relationships between agents other than power relationships, for example actions spawned through friendship. Accordingly, the concept of power cannot be used to define the concept of action, nor can the concept of power be used to fully explain the occurrence of specific social actions and social activity. As is often the case with reductive programs, the net result of either of these attempted reductions is simply the obliteration of important distinctions, and therefore post-reduction there is a need to resurrect these distinctions, albeit by means of different terminology. At any rate, the appropriate response here is not to deny the power dimension of human action – its importance is a matter of common knowledge – but rather to reaffirm a different and contrasting, but nevertheless very fundamental, feature of such action, namely social cooperation.6 There are many kinds of human action which are cooperative. There are joint actions, such as playing in an orchestra or building a house. There are conventions, including linguistic conventions, which enable the collective end of communication. There are institutions and institutional actions of various types. Take the education system. This system has as a collective end the provision of a range of intellectual skills and the acquisition of certain kinds of knowledge. People learn to read, write, and count, and they acquire a body of knowledge. So an education system serves the collective end of education. Moreover, the education system relies on interdependence of action and hence cooperation. The pupils work at learning on condition the teachers work at teaching, and vice versa. Enough has been said by way of demonstrating what is in fact rather obvious; joint action, and cooperation in the service of collective ends more generally, are pervasive features of human life. But the recourse to social forms, such as conventions, social norms, and institutions, is designed to underpin a stronger claim, namely that social forms provide the framework within which most individual and interpersonal human action takes place. On this conception individuals entering the social world find themselves in a complex web of cooperative action none of which in the first instance is of their making. Since this complex web of interaction – this network of conventions and social norms and structure of institutions – constitutes the background and medium in which they act, it makes no sense to speak of individuals rejecting 21
Social Action or transforming the web in its entirety. This is really a logical point; change or rejection of any particular convention or institution is possible at any one time, but the change or rejection of the social fabric in its entirety is not.7 Immanuel Kant uses the image of a bird to make this kind of point: “The light dove cleaving in free flight the thin air, whose resistance it feels, might imagine that her movements would be far more free and rapid in airless space.”8 In their preoccupation, indeed obsession, with the repressive nature of conventions, institutions, and other social forms, many contemporary theorists make the mistake of Kant’s dove. They fail to understand that without, for example, the conventions of language, money, and literature, there would be no communication, no economy, and no literature. Just as air is necessary for birds to fly, so conventions are necessary for poetry. Notwithstanding appearances, my teleological cooperation model of social action accommodates the contrasting dimension of social action, namely conflict and power. The social world thus modelled is a sociopolitical world. Individuals, institutional actors, organisations, classes, and factions spend a great deal of time and energy in competition and conflict. In the course of these power struggles some individuals and groups amass considerable power, wealth, and status. Moreover they often do so at the expense of other individuals and groups and in part by ideological manipulation. However, the fact remains that these power struggles take place within a framework of cooperation between individuals and groups, the exploited and the exploiters, at many different levels, including the communicative level of language, the economic level of business enterprises, and the political level of government. The fact that a framework of conventions and institutions constitutes a set of enabling mechanisms for communication, education, material well-being, and so on, is entirely consistent with the sometimes widespread use, or rather abuse, of conventions, norms, and organisations to repress or coerce.9 In fact, repression of particular groups is often achieved more by excluding them from participation in educational and communicative institutions than by coercing them once they have gained entry to such institutions. The systematic exclusion of black South Africans from educational institutions during apartheid is a well-known case in point. At any rate, my claim 22
1. Theories of Social Action that language is fundamentally a conventional mechanism for enabling communication is not undermined by the fact that this same mechanism is also deployed at times in order to propagandise and manipulate. Again, my claim that democratic government is fundamentally an institution that enables the provision of various collective goods, such as stability and order, is not undermined by the fact that some such governments at times engage in repression of their citizens. Nor is my teleological cooperation model undermined by the admittedly widespread existence, and longstanding history, of wars between groups who belong to different societies. For such conflict is typically in large part outside a social framework. Moreover, protracted war has a disintegrative effect on societies and on conventions, social norms, and institutions; it is literally antisocial. This is exactly as one would expect on a cooperation model of social action. This teleological cooperation model is by no means inimical to fundamental social change. Rather, the point is that fundamental change cannot be brought about by one agent acting alone; it can only be brought about by agents acting collectively in the service of different ends or by using different means to the same ends. A single agent cannot change conventions or social norms, or transform institutions, all by themselves. Not even a very powerful political figure can transform the system of linguistic conventions overnight, and even if she can transform the political system reasonably quickly, she can do so only in cooperation with her political supporters. a taxonomy of social actions, social entities, and sociomoral properties In this section I introduce the basic taxonomy of social actions, and of the associated social entities and sociomoral properties that are my concern in this book. I distinguish between what I will term joint action, spheres of activity, conventions, social norms, rules, social groups, organisations, and institutions. Most of these types will receive comprehensive treatment later on. Specifically, I will offer detailed analyses of joint action (Chapter 2), conventional actions (Chapter 3), actions governed by social norms (Chapter 4), the actions of members of organisations (Chapter 5), and the actions of 23
Social Action members of social institutions and social groups (Chapter 6). I also offer analyses of the sociomoral properties of collective moral right (Chapter 7) and of collective moral responsibility (Chapter 8). I do not claim to offer an exhaustive or especially fine-grained taxonomy of types of social action and social entity. For example, I do not offer analyses of customs or traditions. Nor do I claim that my taxonomy is the only possible one, even at its own level of crudity. Indeed, there are other more exhaustive taxonomies available.10 Rather, I claim that the particular taxonomy that I provide, while to some extent prescriptive, has utility when it comes to understanding social action.11 Joint actions consist of the individual actions of a number of agents directed to the realisation of a collective end. A collective end – notwithstanding its name – is a species of individual end; it is an end possessed by each of the individuals involved in the joint action. However, it is an end that is not realised by the action of any one of the individuals; rather, the actions of all or most realise the end. Examples of joint action are two people lifting a table together and two men jointly pushing a car. So joint actions can be analysed in terms of individualist notions.12 Since it will turn out that on my account joint actions are the basic building blocks of social action, it is important to be clear about the nature of joint action. Earlier I distinguished between social actions and interpersonal actions. Social actions are constructions out of basic joint actions, and joint actions are not necessarily social actions. However, joint actions are necessarily interpersonal actions. Accordingly, individual and interpersonal actions are logically prior to social actions. Moreover, I hold that the individualist notions needed for the analysis of joint actions do not necessarily bring with them a normative dimension, above and beyond the minimal normativity of instrumental rationality. By instrumental rationality I mean the rationality of undertaking an action that is the means to one’s end; I ought to do x for the reason that x is the means to my end, e. Accordingly, I conclude that joint actions do not necessarily have a normative dimension (other than in this minimal sense).13 I also conclude that while joint action logically implies agents – specifically, instrumentally rational agents – it does not logically imply moral agents. Here my point is not that the joint actions performed 24
1. Theories of Social Action by human beings do not as a matter of contingent fact often possess a moral dimension; clearly they do. Rather my claim is that action per se, whether it be individual or joint action, does not necessarily have a moral dimension; the concept of action, including the concept of joint action, is logically prior to the concept of morality.14 We can further elaborate the notion of joint action by introducing the notion of joint activity. Roughly speaking, a joint activity is a complex of differential, interlocking, joint actions and individual actions directed to some overarching collective end. So building cathedrals and sailing ocean liners are joint activities. We can further distinguish between different, albeit connected and overlapping, generic kinds of joint activity, including communicative, economic, educative, sexual and religious activity. The repetition over time and duplication in space of any one of these different kinds of generic joint activity can give rise to a more or less connected, and more or less continuous, stretch of joint activity which I will term a sphere of activity.15 So the ongoing series of economic transactions across Australia constitutes a sphere of activity. Spheres of activity (by stipulative definition) are regulated by conventions. They thereby take on different specific forms according to the specific conventions that structure them. Often, though not necessarily, they are also regulated by explicit rules, including laws. And they are also often – but again, not necessarily – regulated by social norms. Here I distinguish between rules, conventions, and social norms.16 A rule is a (typically explicitly formulated) directive issued by some authority to those who are subject to it to undertake a certain course of action in certain circumstances. So laws are a species of rule. A convention is a regularity in action performed by a set of individual agents and directed to a collective end. More precisely, conventions are joint procedures, where a procedure is (roughly speaking) a habit – the automatic repetition of an action in a recurring situation. So a joint procedure is a procedure that is jointly followed because it realises a collective end. Moreover, in my view conventions qua conventions are normatively neutral – other than in the minimal sense involved in instrumental rationality. I maintain that conventions are normatively neutral notwithstanding the fact that agents can be held to have failed to conform to any given convention. On my account a conventional 25
Social Action regularity in action does not constitute a standard of behaviour to which agents ought to conform in any stronger sense than that implied in instrumental rationality. Naturally, an agent, by not conforming, might fail to realise the collective end that he or she (jointly) has; the end which is realised by conformity to the convention. However, that would be a mere failure of instrumental rationality. Here it might be helpful to draw attention to distinctions between four related, and quite fundamental, conative notions. First, there is individual action. Second, there is joint action. Third, there is repetitive individual action (including action performed in accordance with procedures). Fourth, there is repetitive joint action (including action performed in accordance with joint procedures). I am reserving the term “convention” for the main species of this fourth fundamental conative notion, namely joint procedures. I am doing so partly because I find it intuitively appealing. But my main reason is a theoretical one: Conventions conceived of as joint procedures have great explanatory power. And this is how it ought to be, given the historical and contemporary explanatory burden that is placed upon the notion of a convention. Consider in this connection Aristotle’s dichotomy between nature and convention and Hume’s discussion of conventions.17 So conventional actions are not necessarily the content of directives issued by authorities. Nor are they necessarily explicitly formulated anywhere. Using chopsticks is a convention in China, for example, but not in most parts of Australia. And driving on the left-hand side is a convention in Australia, though not in the United States of America; it is a regularity in action among Australians, which secures the collective end of avoiding car collisions. As it happens, this convention is also a law, though it might not have been. In this respect it is unlike the convention of using chopsticks. Social norms are regularities that are also norms; agents believe that they ought to conform. Here the “ought” is not that of mere instrumental rationality; it is not simply a matter of believing that I ought to conform because it serves a purpose. Some conventions and all rules are also norms in this strong sense. For example, the convention (and the law) to drive on the left is a norm; people feel that they ought to conform. This strong (and wide) sense of ought includes – but is not exhausted by – the so-called moral ought. 26
1. Theories of Social Action Further, the moral ought obviously applies to many actions directed to ends, including collective ends, and often does so by virtue of the moral significance of those ends. However, the moral dimension of such actions is not generated simply by instrumental rationality; for the ends in question are not mere ends, they are morally significant ends. So we have a threefold distinction in relation to social regularities in action: conventions, rules, and social norms. And it turns out that any given regularity in action could as a matter of contingent fact be simultaneously a convention and a rule and a social norm, or two of these, or, of course, only one of these. It should be noted here that I do not accept the reduction of the moral dimension of actions (or agents) to their social dimension. Rather social “morality” is typically in part constituted by objective morality, and there are objective moral principles and values that transcend the de facto attitudes and regularities in behaviour that constitute specific social groups and institutions. This objective morality is probably only capable of being adequately accessed by beings with an instinctual moral sense who are also deliberative agents. Such beings are not only touched by moral considerations. They are also beings who can imagine different possible courses of action, offer themselves reasons for some of these against others, and act on these reasons. A further point regarding my moral objectivism is that I regard individual human beings as the ultimate sources of moral value and bearers of moral responsibility; I reject the proposition that social forms and entities have intrinsic – as opposed to derived – moral value.18 Notwithstanding this, I am not an advocate of what might be termed individual moral atomism. In my view a large component of the intrinsic value that attaches to individuals attaches to them by virtue of their interpersonal relationships with other individuals, including relations of friendship, love, compassion, and so forth. Moreover – as I will argue in Chapters 7 and 8 – there are important moral rights and moral responsibilities which are jointly possessed. In general terms, I am seeking to drive a wedge between collectivism (including some species of communitarianism), on the one hand, and extreme forms of individualism on the other.19 A social group consists of a set of individuals who participate in a number of spheres of activity and therefore are party to a number 27
Social Action of networks of conventions (each network regulating the interactions in a given sphere of activity). In addition – by stipulative definition – members of a social group have been inducted into a common structure of institutions. However, members of a social group do not necessarily constitute an organisation or even belong to any organisation. I will elaborate the distinction between organisations and institutions, but first let me clarify the distinction between social groups and organisations. The defining elements of social groups are particular individuals standing in certain relations – including normative moral relations – to one another. The defining elements of organisations are (embodied) particular roles standing in relation to one another. By embodied it is meant that these roles are filled by individual persons – though not by any individuals in particular.20 So organisations consist of an (embodied) formal structure of interlocking roles.21 These roles can be defined in terms of tasks, procedures (in my above sense), and conventions. Moreover, unlike social groups, organisations are individuated by the kind of activity that they undertake and also by their characteristic functions or ends. So we have governments, universities, business corporations, armies, and so on. Perhaps governments have as an end or goal the ordering and leading of societies, universities the end of discovering and disseminating knowledge, and so on. It is important to note that on this (stipulative) definition of organisations they are, qua organisations, non-normative entities. In this respect they are analogous to conventions, as I have defined conventions. So being an organisation is not of itself something that is either morally good or bad, any more than being a convention is of itself morally good or bad. I can consistently hold this while maintaining that organisations, like conventions, are a pervasive and necessary feature of human life, being indispensable instruments for realising collective ends. Collective ends are a species of individual end; but merely being an end is of itself neither morally good nor bad, any more than being an intention or a belief is of itself morally good or bad. As in the case of conventions, my definition of organisations is stipulative, but not unmotivated. It is motivated in part by intuition. More importantly, it is motivated by theoretical considerations. For there is a fifth fundamental conative notion. We already have four: 28
1. Theories of Social Action individual action, joint action, repetitive individual action, and repetitive joint action (especially joint procedures). But individuals realise their ends not only by performing joint actions, including repetitive joint actions, but also by specialisation. Assume agent A performs task x, and agent B task y, and agent C task z. Assume also that: A cannot perform y or z (or at least cannot perform y or z without difficulty); B cannot perform x or z; and C cannot perform x or y. Assume finally that if A dies or leaves, B and C will identify some D to replace A; similarly if B or C dies or leaves, then some E or F will be found as a replacement. What we have is an organisation, albeit a primitive one. I hold that this notion of an organisation has great explanatory power. Moreover, it is a conative notion that has been characterised exclusively in terms of instrumental rationality; it is not normative in any stronger sense than this. In particular, it does not necessarily imply moral considerations. While my definition of an organisation does not include any reference to a normative dimension, most organisations do as a matter of contingent fact possess a normative dimension. As was the case with conventions, this normative dimension will be possessed (especially, though not exclusively) by virtue of the particular moral (or immoral) ends that an organisation serves, by virtue of the particular moral (or immoral) activities that it undertakes, and by virtue of the fact that the constitutive roles of most organisations are occupied by moral agents. This will be discussed further in this chapter.22 It should also be noted that the social norms governing the roles and role structures of organisations are typically enshrined in explicit rules, regulations, and laws, including laws of contract. For example, an employee not only believes that he ought to undertake certain tasks and not others, but these tasks are explicitly set forth in his contract of employment. Organisations with the above detailed normative dimension are social institutions.23 So institutions are often organisations, and many systems of organisations are also institutions. For example, capitalism is a particular kind of economic institution, and in modern times capitalism consists in large part in specific organisational forms, including multinational corporations. Nevertheless, some institutions are not organisations, and do not require organisations. For example, the English language is an institution, but not an 29
Social Action organisation. Moreover, it would be possible for a language to exist independently of any organisations specifically concerned with language. Institutions are (by stipulative definition) normative entities, defined in part in terms of normatively described ends, and in part in terms of activities governed by social norms. Moreover, institutions are necessarily inhabited by deliberative agents who are also moral agents (or immoral agents – the contrast is with nonmoral agents). So institutions are unlike organisations. Nevertheless, there are important relations between many institutions and organisations. First, as we saw above, many institutions are also organisations or have organisational elements. Second, organisations exist, or at least ought to exist, to facilitate institutional purposes and, ideally, morally desirable institutional purposes. I have demonstrated that social groups necessarily possess institutions. Moreover, there can be no institutions outside social groups, though the same institution can exist across different social groups, for example the institution of the English language. Further, institutions logically presuppose spheres of activity. An institution is, at least, a structure of conventions and social norms that regulates some form of generic joint activity in accordance with more or less discernible ends. Communicative activity is not necessarily institutional activity, but typically linguistic activity is. Spiritual experiences are evidently not necessarily institutional phenomena, but religious activity – including nonorganisational religious activity – is typically institutional in character. Extramarital sexual activity is not necessarily institutional in character, but marital sex is. Periodic bartering between explorers and members of local tribes is noninstitutional economic activity, whereas Finland’s export of paper and import of cars within the framework of European Union guidelines is institutional activity. A final point about social institutions. (This point is also relevant to social norms and conventions.) As is well known, institutions and other social forms reproduce themselves over time. As Giddens puts it: “In and through their activities agents reproduce the conditions that make their activities possible.”24 However, Giddens and others seem to ascribe the reason for this reproduction of institutions simply to continued interaction between conforming individuals over space and time; reproduction of institutions is a largely unintended 30
1. Theories of Social Action consequence of continuous interaction. But this as it stands is unsatisfactory. What is called for is an explanation of this continued conformity. I suggest that collective habit in a context of ongoing interaction is by itself an inadequate explanation.25 My claim is not that unintended consequences do not have an important social role to play here and elsewhere; clearly they do. Rather, the point is that this reproduction of institutions ultimately depends on the continuation of one or both of the following two general kinds of condition. The first kind of condition is that the institution in question continues to effect the purpose or collective ends to which it initially owed its existence – or has come to serve some other collective ends. For example, the conventions of the English language reproduce themselves in large part because adherence to them enables the collective end of communication to be achieved. Here, successful communication is not simply an unintended consequence or outcome of some invisible hand or a (possibly latent) supra-individual function; successful communication is something that individual agents aim at. The second kind of condition is that the institution in question continues to embody a felt moral sentiment. For example, the social norms underpinning apartheid in South Africa were sustained by the view on the part of most whites, and (in all probability) many blacks, that blacks were inferior. Once that view came to be seen as morally objectionable, the various social norms, and indeed institutions, that embodied it underwent a process of disintegration. As I have mentioned, the two main sociomoral properties that I am concerned with in this book are collective moral rights and collective moral responsibilities. Roughly speaking, there are individual moral rights, which attach to individuals qua individuals, and joint moral rights, which attach to individual persons who engage in joint action. For example, the members of a team of bricklayers have a joint moral right to the moneys paid to them for building a brick house. Collective moral rights are a species of joint moral rights, namely rights possessed in part by virtue of membership in a social group. So the right of the Iraqi Kurds to secede, if it exists, is a collective moral right. Note that on my view collective rights are not properties that are possessed by suprahuman collective entities, such as nation states or social groups. Collective moral responsibilities are, in essence, joint moral responsibilities. They are the moral responsibilities that individuals 31
Social Action incur as a consequence of their joint actions; and while these moral responsibilities attach to individual persons, they do so jointly. So if two men jointly beat a third man to the point where he dies, then they are jointly or collectively morally responsible for his death. Each man is morally responsible for the man’s death, and if one is responsible for his death then the other is, and vice-versa. Note that on my view collective responsibilities are not properties that can be ascribed to suprahuman collective entities, such as nation states or social groups or even two men considered as a single entity. philosophical approaches to social action While there is a great deal of highly developed philosophical work in the area of the philosophy of individual action, there is a relative paucity of detailed work in the area of philosophical theory of social action. The main general philosophical works in this area are Raimo Tuomela’s Theory of Social Action, The Importance of Us, and Cooperation, John Searle’s The Construction of Social Reality, and Margaret Gilbert’s Social Facts.26 Each of these authors offers a distinct approach to the theory of social action. Probably the most influential specialised work in the area is David Lewis’s Convention: A Philosophical Study.27 Lewis offers a systematic theory of one species of social action, namely, conventional action. Moreover, his basic approach has been taken over and applied to other species of social action, for example by Ullmann-Margalit and Pettit in relation to action in accordance with social norms.28 These works by Searle, Tuomela, Gilbert, and Lewis embody the main existing philosophical approaches to the theory of social action. Searle’s and Tuomela’s approach basically consists of taking over a particular theoretical approach in the philosophy of individual action and attempting to deploy it in respect of social actions. The approach is that of so-called intention-based action theory. Lewis, Ullmann-Margalit, and Pettit take over the main rival approach in the philosophy of individual action and attempt to deploy it in respect of social action. The approach is that of so-called belief/ desire action theory. Gilbert, by contrast, attempts to develop a non-individualist corporatist theory of social action. Her claim is, in effect, that the
32
1. Theories of Social Action existing work in the philosophy of individual action cannot simply be generalised and supplemented, and applied to the case of social action. Social action, she believes, is radically different from individual action. Gilbert develops an account of what she calls a plural subject. Plural subjects are essentially a species of social group. According to Gilbert, plural subjects can be ascribed beliefs and intentions that are not the beliefs and intentions of any of the individual members of the plural subject (social group). Let me make some brief remarks about each of these approaches. Gilbert’s approach is in my view fatally flawed in that it introduces a mysterious and unbridgeable gap between individual human action and social action. How does one and the same individual human person integrate their individual actions with the “actions” they somehow contribute to, or in part constitute, as members of a collective on this view? Matters become even more problematic when one considers the ontological status of the relevant intentions and beliefs involved in such supra-individual “actions.” At this point Gilbert opts for such things as collective beliefs, which can in principle exist even though no individual person actually possesses these beliefs. Suffice it to say that it would be far more satisfactory if an account could be given that explained social actions wholly in terms of the actions of individual human persons and that helped itself only to mental states that existed in the heads of individual persons. Lewis, Ullmann-Margalit, and Pettit have made some theoretical headway. However, their project suffers from an inadequate initial theoretical base. The attempt to understand human action, be it individual or social, purely by recourse to so-called belief/desire theory is misguided. Michael Bratman has shown this.29 What in the end Lewis and colleagues can offer is not theories of social action but at most theories of the rationality underpinning social action, and, specifically, an account of the structure of motivating preferences and desires. I discuss Lewis on conventions in Chapter 3. I favour the intention-based approach. Its strengths are: (1) it is based upon solid existing theoretical work in the philosophy of individual action; (2) individual and social action are taken to be different species of the same genus rather than radically different phenomena. However, as it stands I don’t believe it is sufficiently rich in theoretical machinery to adequately explain social action. In
33
Social Action particular, it requires the further teleological notion of a collective end and the notion of a procedure (roughly, a habit) in the service of collective ends. The Definition of Social Action To this point I have assumed that there are distinctions to be made between individual and interpersonal action on the one hand and social action on the other, but I have not attempted to define social action. There are a number of definitions of social action in the literature, and in this section I want briefly to consider the main ones, albeit for the purposes of rejecting them. David-Hillel Ruben and many of the rational choices theorists hold the permissive view that social action is interdependent action.30 Gilbert and, perhaps, Tuomela are committed to the restrictive view that social action is to be identified with joint action.31 There is a third influential view, associated with Wittgenstein, that social action is to be defined in terms of action performed in accordance with rules.32 My own view is somewhat more complicated than any of these. First, I hold with Ruben that social action is necessarily interdependent action. However, I hold that social action is only one species of interdependent action. For example, there are many interpersonal actions that involve interdependence but are not social actions. Ruben offers a specific analysis of interdependence, which he then uses to define social action. My view is that he casts his net too wide and includes actions that are properly speaking interpersonal, but not social. Second, while I hold that joint action is the basic building block for social action, I maintain – contra Tuomela, Gilbert, Searle, and others – that joint action is not necessarily social action. As I have mentioned, joint action is necessarily interpersonal action but not necessarily social action. Third, while it is obvious that action performed in accordance with (social) rules is necessarily social action, such action is only one species of social action. Accordingly, social action cannot simply be identified with such rule-governed action. Furthermore, neither joint nor interdependent action is necessarily social action. Accordingly, social action cannot be defined as joint action or as interdependent action. Roughly speaking, interdependent action is action in which one agent performs an action at least in part in virtue of his beliefs about 34
1. Theories of Social Action the actions, beliefs, or other attitudes of another agent or agents. Ruben has defined social action in terms of interdependent action.33 Let us look at his conception.34 According to Ruben the following four conditions are necessary and sufficient for an action to be a social action. First, there must be at least two agents, A and B, such that one agent has beliefs about the other agent and vice-versa. Second, these beliefs must themselves be part of the content of higher order beliefs. So, for example, A must believe that B has some belief about A. Third, A performs the (social) action in part on the basis of what A believes B believes about A, and vice-versa for B. Fourth, A and B mutually believe that one another’s actions are in part determined by their respective beliefs about what the other believes. (A and B mutually believe some proposition, p, if and only if A believes that p, B believes that p, A believes that B believes that p, and so on in the familiar way.) I will argue against Ruben’s conception by way of counterexample. My counter-example involves conflict of a very pure kind, and it is two individual persons who are in conflict.35 The individuals in question are strangers to one another and belong to quite different societies. These two such strangers confront one another in the desert away from any other people, social institutions, or social objects. There is insufficient water for both. Both seek to kill each other with their bare hands in order to survive, and both know that the other is seeking their death. They both mutually know that each is contriving strategies based on beliefs about what the other believes, and so on, in the manner required by Ruben’s definition. Their actions are interdependent. Moreover, their actions are driven by extreme thirst and are therefore in the first instance natural. Since the one individual directs his actions at the other individual, their actions are interpersonal. Why should this example be considered to be a counter-example? In the first place the interaction between the two individuals does not involve any social forms, such as conventions or social norms, or any social attitudes. Indeed, the social norm proscribing the killing of another person has been abandoned (though not as a result of a process of moral deliberation). So the onus is on Ruben to explain why the interaction should be regarded as social action. The example involves action on the basis of beliefs about the actions of the opposing party. But so might the actions of a man-eating tiger that 35
Social Action is aware it is being hunted by villagers whom it has been seeking to kill and eat. The tiger is not engaged in social action. Of course, Ruben holds that the beliefs in question be higher order beliefs, beliefs about the beliefs of others, and he might insist that he is speaking of beliefs in the sense of mental states with propositional content. Tigers do not have beliefs with propositional content (whether higher or lower order beliefs). But why should a difference in the level of intelligence – whether manifested in the propositional character of mental states or in the fact that they are higher versus lower order mental states – of those involved in conflict make a difference to the sociality of their interaction? Suppose that instead of a tiger and some villagers, we had villagers and a predatory Martian capable of higher order propositional beliefs, that was seeking to kill and eat the villagers. It is hard to see why the interaction would now need to be regarded as social interaction. It might be claimed that propositional attitudes and/or higher order mental attitudes, including beliefs about beliefs, logically presuppose, or imply, human language and hence (linguistic) conventions.36 But I disagree with this. I conclude that Ruben has offered a definition of social action that is too thin to do the job; this is not social action, but something short of social action, namely a species of interpersonal interaction. Let me now turn to joint actions. Joint actions are a species of interdependent action in which there is little or no conflict; joint actions involve a number of agents performing interdependent actions in order to realise some common goal. In Chapter 2 I propound the collective end theory of joint actions.37 Basically on this theory joint actions are actions directed to the realisation of a collective end.38 However, this notion of a collective end is a construction out of the prior notion of an individual end. Roughly speaking, a collective end is an individual end more than one agent has, and which is such that, if it is realised, it is realised by all, or most, of the actions of the agents involved, and is not realised by only one of those actions. Some theorists – including Tuomela and Gilbert – have argued that joint actions are necessarily social, indeed that they are the paradigm of social action. I suggest that joint actions are not necessarily social actions.39 Joint actions are sometimes social. The joint action of wearing a suit on formal occasions, given everyone else 36
1. Theories of Social Action does, is social. However, joint actions are not necessarily social and therefore social actions cannot simply be defined as joint actions. Joint actions can be one-off interpersonal actions performed by persons who are not members of the same society, who are not parties to any of the same conventions or social norms, or members of the same institutions. Suppose, on a single occasion only, I joined with a stranger in removing a fallen tree blocking our pathway. We did not communicate explicitly, went our separate ways thereafter, and my interest in his action was purely as a means to enable me to move the tree blocking my pathway. In this case, our joint action would not be social. To see this, consider the possibility that the “stranger” was a Martian, who shared no social forms with me, had the body shape and size of a doglike robot, and was not capable of any emotional states characteristic of humans, including social emotions such as social approval; the only things the Martian had in common with humans was beliefs, intentions, and ends. Given all this, surely we do not have an instance of social interaction. As we saw in relation to Ruben’s views, it is possible to argue that since this sort of joint action involves higher order mental states – such as beliefs about the beliefs, ends, and intentions of others – then it involves sociality; for it is claimed that the possession of higher order mental states requires the ability to use a language and hence conformity to conventions. However, this argument fails because it is by no means clear that what might be termed basic joint action does necessarily involve convention-governed language. Consider a primitive form of basic joint action, such as a pack of hyenas attacking and killing a large zebra in order to eat it. Here the hyenas have intentions, beliefs about the behaviour of the other hyenas and the zebra, and indeed awareness of the other hyenas and the zebra.40 Here the notion of awareness is not the notion of a propositional attitude; nor does the notion of intended behaviour necessarily bring with it propositional content. However, there are higher order mental states; for example, the hyenas are aware of the zebra’s awareness of them. Nevertheless, possession of the mental states in question does not require recourse to a convention-governed language. Again, consider a maneating tiger adopting a stalking stance in relation to its prey. In some sense the tiger intends to kill its prey, but this intention does not have propositional content. Moreover, the prey, a villager, is at 37
Social Action this point unaware of the tiger. The tiger continues to stalk, but then the villager sees the tiger; indeed, their eyes meet, so to speak. (Here we have a species of extraspective interaction.) The tiger abandons his stalking and pounces; the villager runs. The tiger became aware of the villager’s awareness of the tiger, and this caused the tiger to abandon his stalking action and to go in for the kill. But this higher order awareness is not propositional in character. Accordingly, it is difficult to see why a convention-governed language is necessarily implicated in this scenario. But now let us consider the basic joint action of humans, joint action that involves higher order mental states with propositional content. Does such joint action logically presuppose or imply a convention-governed human language? I take it that Paul Grice, Jonathan Bennett, and other proponents of agent-semantics have demonstrated that semantic facts are reducible to psychological facts about individuals. So the propositional character of conventiongoverned human language is to be explained by the propositional character of thought, rather than the other way around.41 Moreover, I do not accept that higher order propositional attitudes, in particular, logically presuppose or imply convention-governed language. On the contrary, as I outlined in the introduction to this book, I believe that extraspective attitudes – including awareness of the awareness of others, awareness of intentions of others, and so on – are a fundamental feature of human minds, and one that is logically prior to sociality. So, for example, I do not accept the need for an argument from analogy in order to demonstrate the existence of other minds. Nor do I accept the various claims, rightly or wrongly associated with Wittgenstein, that the criteria for the correctness of beliefs are necessarily provided by communities. (For more on this, see the following section on rule-governed action.) I conclude that joint action is not necessarily social action. Rather, it is a species of interpersonal interaction, and as such involves extraspective interaction, but not necessarily social interaction. Margaret Gilbert has argued that joint action examples – of the kind I have put forward as counter-examples to the view that joint action is necessarily social action – involve recourse to corporatist notions such as irreducibly collective intentions and beliefs, and plural subjects.42 My general response to Gilbert is to put forward my theory of joint actions in terms of wholly individualist notions. 38
1. Theories of Social Action In Chapter 2, I deal with some of her more specific examples and arguments to the conclusion that joint actions are irreducibly corporatist and therefore social. I conclude that neither interdependent nor joint action is necessarily social action. These species of action are certainly interpersonal, but not necessarily social. Accordingly, social action cannot simply be defined as interdependent or joint action. Social Action as Rule-Governed Action According to one strand of the analytic philosophical tradition – a strand associated with the later Wittgenstein, Peter Winch, and, in a somewhat different vein, John Searle – social action ought to be construed as rule-governed action.43 The claim here is that social actions are necessarily rule-governed, and, further, that being rulegoverned is sufficient to mark off social actions from nonsocial actions. On pain of circularity, rules in this sense are not by definition social in character. So, theoretically, there could be private rules, rules that only single individuals followed. However, these theorists argue that this is a mere “theoretic” possibility, that in reality there could not be private rules.44 Accordingly, the way is clear to define social action as rule-governed action. But why could there not be private rules? One familiar line of thought here is that thinking involves some sort of representational system, and representation involves a set of rules governing the meaning of the symbols by means of which things are represented. But such meaning rules are held to be necessarily social. A variety of reasons are offered for this claim. Some of the most influential emanate from the work of Saul Kripke.45 Kripke offers, but does not endorse, a variety of sceptical arguments against the possibility of private meaning rules. For example, Kripke’s sceptic suggests that there is nothing that could constitute an agent meaning addition by the plus sign, “⫹”, rather than quaddition (quaddition being the same operation as addition up to, but not beyond, a certain numerical threshold). Since an agent might never get to that threshold there is nothing in his behaviour to distinguish addition from quaddition. More importantly, there is (allegedly) nothing in his mental life to do the job either. The suggestion that the agent at some time in the past t1 intended that “⫹” mean addition supposedly depends on the agent being able to identify his current, t2, intention to mean 39
Social Action addition by “⫹” as being the same intention as the one he had at t1. But an intention does not have any qualitative feature that marks it off from other intentions; there is no feature that makes it the same intention at t2 as it was at t1. Now there is a great deal of confusion surrounding the use of the term “rule” in many of these discussions. Sometimes the notion of a rule in question is something akin to a rule as I have defined it above, namely a prescription or directive issued to someone to perform some kind of action in certain circumstances, although often it is unclear who the authority issuing the rule is, or else the authority in question is implicitly the community, or even the person who is subject to the rule. At other times, the “rule” is akin to a habit or standing intention or disposition; as such it is not obviously prescriptive or normative in character. I am not convinced that the concept of a rule in the sense of some kind of prescription can really do the work that seems to be required of it. This is because the notion of a prescription is itself in need of analysis. Given the limited space that I have here and the size and complexity of the philosophical literature on these issues, I am simply going to sketch a brief outline of the kind of position that I occupy. First, as I have indicated, I am an advocate of agent-semantics. Accordingly, I hold that semantic facts can be reduced to psychological facts and, specifically, Gricean – or Gricean-like – intentions. Second, contra Kripke’s sceptic, I do not see why intentions cannot be regarded as primitive mental states, albeit ones that are not individuated by their possession of some sort of qualitative character. Intentions are not sensations, but they are no less real for that.46 Third (and in Chapter 3 I will develop this argument), contra the anti-private-rule theorists, the notion of a private rule is quite coherent and is simply the notion of a procedure to perform an action in certain circumstances, and to do so indefinitely into the future.47 The key notion here is that of a standing intention to perform a set of actions indefinitely into the future. It is important to note here that a standing intention is not simply a set of different token intentions possessed at different times. So the existence of a standing intention does not necessarily involve the process envisaged by Kripke whereby the members of the set of discrete token intentions are compared with one another in order to determine whether or not they are the “same” intention. 40
1. Theories of Social Action Fourth, although having intentions – including Gricean intentions and my standing intentions – presupposes possession of a concept, this is essentially possession of an ability, and, as McGinn points out, there is no inherent susceptibility of abilities to Kripke’s sceptical arguments.48 Fifth, intentions do not necessarily have a normative dimension (in some strong sense). Consider in this connection the man-eating tiger already discussed. The tiger certainly intends to kill his victim, and do so as a means to his end of eating the victim; so his intentions and purposes can fail to be realised. But the tiger does not believe that he ought to eat his victim; he just wants to eat him, has eating him as a purpose, and intends to kill him in order that he might eat him. Intentions bring with them beliefs, and the tiger also has various beliefs. But, once again, these beliefs are not normative, other than in the narrow teleological sense that because they are beliefs they have intentionality; because they are beliefs, they, so to speak, “aim at” reality and fail as beliefs if reality is not as it is believed to be. Sixth, there are some sophisticated intentions that presuppose language and, therefore, conventions. But conventions themselves are not necessarily normative in any strong sense, but only normative in the sense dictated by instrumental rationality (see Chapter 3 for my detailed argument). Moreover, the Gricean program of agent-semantics has shown how language presupposes certain sorts of intentions that do not themselves presuppose conventions or other social forms. Seventh, if intentions are not necessarily normative – other than in the weak sense of instrumental rationality – then it is hard to see why they logically presuppose or imply social rules. For if private rules can be analysed as standing intentions, then private rules need not logically presuppose or imply social rules. Moreover, if private rules do not logically presuppose or imply social rules, then the notion of a rule cannot of itself be justifiably used to mark off social actions from nonsocial actions. At any rate, in the light of the above points, I take myself to be entitled to distinguish between social rules on the one hand and private, and perhaps interpersonal, rules on the other. It may be that some rules are such that, if they are followed, they are followed by more than one individual. But we cannot conclude from this that 41
Social Action these rules are social rules. For example, a man and his wife might have the rule that each picks up the schoolchildren on alternate days. This is certainly an interpersonal rule, but it is not necessarily a social rule. Perhaps the husband and wife are the only couple who follow this rule. What the proponent of social action as action in accordance with rules must hold is that social action is action performed in accordance with social rules. But now we require an account of the notion of a social rule that reveals the necessarily social character of actions performed in accordance with such rules. Constitutive Rules At this point an alleged distinction between regulative and constitutive rules tends to be invoked.49 A constitutive rule, but not a regulative rule, has the form “Agent A is required to count xing as ying (in circumstances c).”50 Clearly, regulative rules need not be social rules. A single parent might have a regulative rule governing the behaviour of their only child. This rule might not be in force elsewhere in the society. It is a personal, or perhaps interpersonal, regulative rule; but it is not a social regulative rule. A regulative rule governs a pre-existing action type, and such action types need not be social actions. However, the claim is that constitutive rules are necessarily social rules. So let us turn to these. There are no such things as constitutive rules that are necessarily social; rather, so-called constitutive rules are a species of regulative rule, and actions performed in accordance with such rules are not necessarily social in character.51 As noted, a constitutive rule has the form “Agent A is required to count xing as ying (in circumstances c).” Now it is presumably the case that A is to count xing as ying (in c) in virtue of the fact that, in some sense or other, to x is to y. Moreover, xing is evidently in itself (ultimately) a nonsocial action that exists prior to the rule.52 However ying is a social action that only comes into existence as a possible action type in virtue of the constitutive rule. This leads to the question, “In virtue of what does xing count as ying?” Before attempting to answer this, let us provide ourselves with some examples to work with. Examples that come to mind here are such things as that, in cricket, hitting the ball over the fence counts as six runs and the 42
1. Theories of Social Action umpire saying, “Out” counts as bringing it about that the batsman is out; or that a particular stone counts as the position of the guard in a prisoner’s late night planning of his future escape. But there is a second kind of example. This kind includes such examples as that providing an employee with a free house then counts as part of his salary, or that the hiring of two half-time tutors then counts as one full-time tutor. And it is well known that a stocking can count as a fanbelt and a log as a bench. In this second kind of example, in saying that x counts as y, one means that x and y are interchangeable or, at least, that x can stand in for y. Here x, though not numerically the same thing as y and not sharing all its properties with y, may nevertheless be able to perform the same function as y or serve the same purpose as y. Moreover, that x can count as y is due, or primarily due, to properties that x has that enable it to replace, or stand in for, y. In the second kind of example, x could not count as y primarily in virtue of someone’s decision that it so count. On the other hand, in this second kind of example someone’s decision might be a necessary condition for x counting as y. If so, it is important to note that such a decision could be the decision of an individual person acting on his own (albeit in accordance with a personal regulative rule); such a decision is not necessarily a collective or social decision. However, in the first kind of example the matter seems to be different. These cases seem primarily dependent on a decision to count x as y, albeit that in some cases this is a conventional or rulebased decision. Thus, saying “Out” counts as bringing it about that the batsman is out – or at least counts as bringing about a situation in which the batsman is to behave as if he were out – in virtue of a decision that this is to be so. Similarly, for hitting the ball over the fence, six runs are recorded and not four; and that this is so is a matter of conventional decision.53 On the other hand, in the case of the prisoner using the stone to count as the guard – and even doing so in accordance with his own private rule – there is no conventional decision; rather, it is an individual decision, and indeed one the prisoner very much wants to keep secret. No doubt in both these kinds of case – the individual and the collective kind – not any old x could count as y; x may have to have certain properties if it is to be a candidate for being counted as y. But what is characteristic of the first set of cases is that x counts as y primarily in virtue of a 43
Social Action decision by parties to the convention or adherents to the rule. So in these cases, the “counts as” relation between x and y has been created; and the relation remains in existence only so long as the relevant people (or person) continue to treat x as y and hence in fact continue to count x as y. The corollary to this is that examples of x counting as y that are primarily dependent on decisions to count x as y do not necessarily involve social actions. So the answer to the question, “In virtue of what does x count as y?” – in the cases of social actions governed by constitutive rules – is: “In virtue of a conventional or rule-based decision that this be so.” And an account of this notion of something counting as something else is also forthcoming. It seems that the expression “x counts as y” refers to a variety of similar cases including: x standing in for y; x being a means of enacting y; x being treated as if it were y; and x being ascribed the property of being y. But these relations, standing in for, being treated as if it were something else, and so forth, are not in themselves conventional or social in any sense. The example of the prisoner and the stone demonstrates this. Consider also the following example. Imagine a man who goes temporarily mad and decides to treat his dog as if it were a football; he proceeds to kick the poor animal around the room. Assume it is a one-off episode. The madman’s decision to so treat his dog is not in any way a matter of convention or norm or even private rule. So the fact that some object or action, x, counts as y does not logically imply that x counts as y by convention, or that x counts as y in virtue of some rule to the effect that x is to count as y. But in that case the expression “x counts as y” does not necessarily refer to convention or even rule-governed states of affairs. The upshot of this discussion is that constitutive rules are rules or conventions governing certain actions – describable in generic terms as actions of counting something as something else – and that these action types are not necessarily or intrinsically conventional, or social in some other sense. So constitutive rules are really only a special case of regulative rules. But we saw earlier that regulative rules are not necessarily social rules. So any attempt to mark off social from individual and interpersonal action simply by appealing to the notion of a regulative rule would fail. Moreover, there is nothing inherently social in the “counts as” device. That device can be embed-
44
1. Theories of Social Action ded in a social regulative rule, a personal regulative rule, or no rule at all. This failed appeal to constitutive rules was the final tactic in the general strategy of attempting to define social action in terms of rule-governed action. So we are now entitled to conclude that the general strategy of attempting to define social action in terms of rule-governed action has been unsuccessful. Philosophical Approaches and Strategies The theoretical conceptions we have thus far considered manifest the widespread tendency to assume that there is some unitary theory of social action, notwithstanding the fact that prima facie there are quite diverse kinds of social action. There is social action in conformity to conventions, rituals, and norms, and there is social action in accordance with social and institutional roles. The first point to be made is that this assumption of a unitary theory might be false. (This is consistent with the fact that there are some necessary conditions for being a social action, such as the condition that the action involve some reference to other people.) Accordingly, what is called for is the provision of a taxonomy of what are indisputably (and pre-theoretically) categories of social action, and the development of piecemeal accounts of these various categories. Accordingly, in this book I offer piecemeal accounts of some of the central categories of what are termed social actions. As is by now clear, on my account the notion of a joint action can be used to develop (at least partial) analyses of the main categories of social action; so the notion of a joint action is in my view a powerful theoretical tool, and one that provides a certain unity to my endeavours in relation to the analysis of social action (and also in relation to the analyses of the sociomoral properties, collective rights, and collective responsibility). In this sense my account is unitary. Finally, I will assume in this book that there is a distinction to be made between social action and natural (individual and interpersonal) action. My concern in this book is with social action alone. However, I need to reiterate three points made earlier in the introduction to this book. First, all human action is either individual or interpersonal action. Second, the category of natural (individual and interpersonal, including joint) action is logically prior to the general
45
Social Action category of social action; social actions consist in the regulation and structuring of prior nonsocial actions. Third, prior nonsocial actions are regulated and structured by social rules and other social forms (socially conditioned) to generate a higher level of action (social action). This second level of action – now social action – can itself be further regulated and structured by social rules and forms to generate a still higher level of social action. The crucial point to be made in relation to this process of regulation and structuring is that it is an enabling process; it enables the possibility of higher level individual and interpersonal action. Consider in this connection the phenomenon of language.54 the moral dimension of social action Having, so to speak, sketched the philosophical landscape with respect to social action, and offered a basic taxonomy of social action and social entity types, I now want briefly to discuss the moral dimension of social action. I need to do so because this book has the sociomoral character of human action as one of its central concerns. Specifically – as previously mentioned above – I offer detailed analyses of collective moral rights, including so-called minority rights (see Chapter 7), and collective moral responsibility, including collective responsibility for such activities as humanitarian intervention (see Chapter 8). These analyses of collective rights and collective responsibility heavily depend on my account of joint action. In addition, I discuss a variety of other moral issues in the context of offering analyses of various social action types. These include the obligation to obey the law (see Chapter 4) and the (alleged) moral status of organisations (see Chapter 5). A fundamental moral issue that arises in relation to social actions is the issue of moral autonomy. While I cannot offer a detailed philosophical analysis of moral autonomy – or the bearers of moral autonomy, namely moral agents – nevertheless moral autonomy is presupposed in many of my specific analyses. For this reason I now want briefly to consider the notion of moral autonomy. I do so in the context of my distinctions between (mere) agents, instrumentally rational agents, and moral agents. Agents act, but do not necessarily reason; instrumentally rational agents act and provide themselves with reasons for acting in terms of the ends that they have, 46
1. Theories of Social Action but the reasoning is not necessarily more than instrumental reasoning; moral agents act, provide themselves and others with reasons, and the reasons include moral reasons as well as purely instrumental considerations. Specifically, moral agents offer reasons for acting, and such reasons include their (believed in) moral ends and principles. And if a moral end is a reason for action for some agent then it is an end that is worth having, or at least it is an end that is worth having from the perspective of that agent; moral ends are more than mere ends. Also, moral agents reason about which ends they ought to choose to have; they reason about which ends are worth having. Further, moral agents have a degree of moral autonomy. Whether or not it makes sense to ascribe autonomy in some sense to (nonmoral) rational agents is not something I can discuss here, beyond making the following point.55 An autonomous moral agent is necessarily engaged in moral deliberation in relation to that agent’s action, including with respect to the moral worth of the principles and ends that guide their action; so the question arises as to their autonomy in relation to such morally informed action. For non-moral instrumentally rational agents, on the other hand, these issues simply do not arise. Non-moral (instrumentally rational) agents are not touched by moral considerations; they do not have moral sentiments or concerns. Non-moral instrumentally rational agents do not, and cannot, engage in specifically moral reasoning. Accordingly, they cannot possess moral autonomy. Perhaps they can possess some analogous kind of autonomy, but that is an entirely different matter.56 I have distinguished between agents, rational agents, and moral agents. The distinction between agents and rational agents can be made out as follows. Rational agents, but not mere agents, possess the following: (a) future directed intentions, beliefs, and/or ends; (b) intentions, beliefs and/or ends in relation to agents who are not copresent; and (c) memories of past events that took place outside the temporal horizon of their immediate experience.57 Accordingly, rational agents are able to engage in means–end reasoning, including in relation to their own future and in relation to interdependent action with other agents. So subhuman animals are mere agents, but not rational agents. Humans are typically rational agents and also moral agents. What are the implications of sociality for the moral agency of 47
Social Action human beings? From the fact that individual persons are inducted into a particular framework of conventions, social norms, institutions, and other social forms, and therefore exhibit the characteristic features and orientations of members of the social group in question, it does not follow that those individuals are not morally autonomous agents or that they are to any significant extent coerced. In participating in such a network of cooperative action, the individual adopts the ends that in part define those conventions, norms, and roles. Indeed it is inevitable that some of these socially given ends will end up being in part constitutive of the selfhood of the participating agents. However, such socially given ends by no means exhaust the ends constitutive of the selfhood of agents, and in any case these socially given ends typically consist in the structuring of pre-existing nonsocial (individual and interpersonal) ends in the manner described above. Moreover, the conventions, norms, and institutions definitive of some social group may function chiefly as mechanisms that enable individuals to achieve a high degree of individual moral autonomy and enable interpersonal relationships to flourish in ways that far outrun the dictates of conventions, norms, or roles. Notwithstanding this point concerning moral autonomy, there is a tendency to believe that the existence of cooperation of itself signals the presence of moral value and, conversely, the existence of conflict the absence of it. This no doubt in part explains the tendency to jettison the moral upon embracing the ubiquity of political struggle. But from the fact that social action is fundamentally cooperative action, it does not follow that social action is morally good. An enterprise, for example, bringing the Third Reich into existence, may be a joint enterprise, yet essentially evil. The value of, a convention for example, is to be determined principally not by the fact that it involves cooperation and realises a collective end, but rather by the value of the particular collective end that it realises. Perhaps elaborate conventions of politeness mask, and therefore indirectly serve, a morally repugnant end, namely deception. The general point to be made here is that social cooperation is not an intrinsic moral good, notwithstanding the fact that it is a necessary condition for a whole range of intrinsic moral goods, including moral autonomy. 48
1. Theories of Social Action Having argued for the possibility of individual moral autonomy in sociopolitical contexts, what is now required is further specification of the bearer of that autonomy in such contexts, the substantive moral agent. The set of properties that a substantive moral agent would necessarily possess include the following. First, he or she possesses a capacity for rational thought, including instrumental reasoning. This involves the capacity to offer reasons for and against particular moral choices, including the capacity for consistency in the making of moral judgements that he or she might make. So a substantive moral agent is in the first instance a rational agent, as opposed to a mere agent. Second, the agent possesses imagination. This involves, among other things, a capacity to envisage hitherto unencountered situations and ways of behaving. Third, the agent possesses freedom in the sense that he or she can make decisions on the basis of his or her rational thought processes and implement these even in the face of external resistance. Fourth, the agent experiences moral emotions, such as sympathy for other people, compassion, love, and so on. Fifth, the agent possesses an awareness of self, and this, together with his or her powers of rational thought and volition, enables the agent to conceive of his or her life as a totality and to develop that life in particular ways. Sixth, the agent possesses a sense of moral value. This includes the sense that certain things are worth doing and others are not, and that certain actions are morally right and others are not. Importantly, this sense exists and can be acted on despite contrary personal inclinations and various forms of external social prohibition and pressure. Moreover, the moral agent acts in accordance with the dictates of this moral sense, and in part defines himself or herself as a being who does so. So to fail to act on this moral sense will have a disintegrative effect on the agent’s personality.58 Seventh, by virtue of the above properties, and especially the capacity for sympathy and a sense of justice, the agent is able to establish intrinsically valuable relations with other agents. Eighth, the agent’s moral values and standards must cohere with one another and persist over some significant period of time. Otherwise his or her ethical dimension will become conflict-ridden and eventually disintegrate. Ninth, the structure of moral values internalised by the agent will be to some extent a response to, and a result of, the particular historically given circumstances in which the 49
Social Action agent finds himself or herself. Tenth, the structure of moral values internalised by the agent will be to some extent a response to, and a result of, the particular (historically given) sociopolitical circumstances in which the agent finds himself or herself. This last point is in need of further elaboration, since it is the source of persistent confusion. Given that the structure of moral values internalised by an individual agent is to some extent a response to, and a result of, particular historical, sociopolitical circumstances, we would expect to find that agents who belong to the same social group share a core set of moral values. It does not follow from this that each member agent is the passive recipient of the values of the group. The existence of shared values follows from the fact that groups of individuals do not confront the world wholly as atoms but rather in concert with their fellows. People who live together have to work out a coherent system of shared moral values, principles, rights, and responsibilities. Shared moral values, principles, rights, and responsibilities are embedded in, and reflect, complex, dynamic, structured processes of intergenerational joint activity. Among these moral phenomena are so-called collective moral rights and collective moral responsibilities. Moreover, at the level of social action – as opposed to individual and interpersonal action – these rights and responsibilities are quite fundamental (as I will seek to demonstrate in Chapter 7 on collective rights and in Chapter 8 on collective responsibility). The central theoretical notion in this book, namely that of a joint action, is particularly well suited to illuminate collective rights and collective responsibility. Now in some cases these shared moral values, principles, and so on might be imposed on the individual members of the group by the coercive action of the group as a whole or by some controlling sub-element. But this is not necessarily the case, and it is a matter for empirical investigation whether some value has been imposed on a particular individual, or individuals, or not. The second thing we would expect to find, given that the structure of moral and ethical values, principles, and so forth of an individual and/or a group is partly a response to and a result of particular historical sociopolitical circumstances, is some differences in ethical values from one historical society to another. In some instances this might be due to moral development. Presumably, contemporary attitudes to women in the workforce, while by no 50
1. Theories of Social Action means exemplary, constitute moral progress over attitudes prevailing in the nineteenth century. In other instances it is simply due to the different requirements of the material and social circumstances of the day. Physical courage is a great virtue in war but not nearly so important in times of peace. Physical strength is rightly valued in a society at a low level of technological development. And so it goes on. But it is important to stress that the inevitability and, indeed, the desirability of such differences in no way support moral theories such as cultural relativism or morality as ideology conceptions. Here I do not have in mind the claim that these differences between cultures, and over time, typically take place against a background of a commonality of moral values across cultures and times, though this claim is in fact true. Rather, my point is that the objectivity of moral and ethical values and judgements is not called into question by the obvious fact that different circumstances call for, and cause, different moral responses. Allowing weak and sickly infants to die might be morally right for a community living on the very edge of survival and morally wrong for us. But this might simply mean, as a matter of objective truth, that to allow infants to die under certain circumstances is morally right and under other circumstances is morally wrong. It is morally right, for example, if the infants are weak and sickly and would be a burden of such a kind as to threaten the survival of the community. Accordingly, whether we are Ethiopians or Eastenders, in jointly working out our shared moral values, including collective rights and collective responsibilities, the distinction between subjectively felt moral values (whether subjectively felt by a single individual or by the members of a group) and objective moral values needs to be kept in mind, and is in fact presupposed; one can always ask whether a given moral belief or attitude in relation to an action, person, or institution ought to be held and offer reasons for or against. I have distinguished between social and nonsocial actions, and distinguished also the concept of action from the concept of power, and social conflict and power from social cooperation. I have offered a taxonomy of social action types and sketched the main philosophical accounts of social action. I have now also located the moral dimension of social action in the general scheme of things. Let us then turn directly to the main tasks of this book, namely to offer (a) philosophical analyses of central social action types, 51
Social Action namely joint action, conventional action, social norm-governed action, and of the actions of social groups, social institutions, and organisations; and (b) philosophical analyses of the related normative phenomena of collective rights and collective responsibility. I begin with the most fundamental notion in all this – joint action.
52
2
Joint Action
The concept of joint action is fundamental to the understanding of social life in general, and social action in particular. In this chapter I will offer a philosophical analysis of joint action and defend it against rival accounts. In subsequent chapters I will use this analysis of joint action in the analysis of a variety of social action categories, including conventional action, social norm-governed action, and organisational and institutional actions. As we saw in Chapter 1, joint actions are actions involving a number of agents performing interdependent actions in order to realise some common goal. Examples of joint action are: two people dancing together; a number of tradespeople building a house; and a group of robbers burgling a house. Joint action is to be distinguished from individual action on the one hand, and from the “actions” of corporate bodies on the other. Thus when an individual walks down the road or shoots at a target, these are instances of individual action. When a nation declares war or a government takes legal action against a public company, these are instances of corporate action. Properly speaking, there are no such things as corporate actions (as I will argue in Chapter 5). At any rate, my concern in this chapter is simply with joint action. The concept of joint action can be construed very narrowly or more broadly. On the most narrow construal we have what I will call basic joint action. Basic joint action involves two co-present agents each of whom performs one basic individual action, and does so simultaneously with the other agent, and in relation to a collective end that is to be realised within the temporal horizon of the immediate experience of the agents. Examples of basic joint action are two people shaking hands, and two people manually lifting a refrigerator from the kerbside up onto a truck. If we construe joint action more broadly we can identify myriad other closely related examples of joint action. Many of these involve 53
Social Action intentions and ends directed to outcomes outside the temporal horizon of the immediate experience of the agents; for example, two people engaging in a two-hour conversation, or three people deciding to build a garden wall over the summer break. Others involve intentions and ends directed to outcomes that will exist outside the spatial horizon of the immediate experience of the agents. Thus two people might jointly fire a rocket into the extremities of the earth’s atmosphere. Still other joint actions involve very large numbers of agents, such as a large contingent of soldiers marching on a parade ground. If we construe joint action more broadly still we can identify further kinds of joint action. One such kind is what I will call joint enterprises. Instances of joint enterprises are: fighting a war; a marriage; and a joint business venture. Joint enterprises are projects involving more than one person performing various different tasks directed to diverse collective ends, which are in turn directed to some ultimate collective end or ends. Joint enterprises involve planning, adherence to conventions, and many joint actions. Joint enterprises are the collective analogue of an individual carrying out her plans for the future, for example by pursuing her career. One kind of entity that pursues a joint enterprise is an organisation (see Chapter 5). Basic joint actions can also be distinguished from what I will call joint procedures. An agent has a joint procedure to x, if he xs in a recurring situation and does so on condition that other agents x. (Procedures are distinct from repetitions of the same action in a single situation, such as rowing or skipping.) Thus people in Australia have a procedure to drive on the left-hand side of the road. Each person who drives in Australia drives on the left whenever he or she drives, and drives on the left on condition the other agents drive on the left. Joint procedures are in fact conventions (see Chapter 3).1 We can also distinguish between joint procedures (in this sense) and what I will call joint mechanisms.2 Examples of joint mechanisms are the device of tossing a coin to resolve a dispute, voting to elect a candidate to office, courts of law, and governments. Many joint mechanisms are institutional mechanisms (see Chapter 5). Joint mechanisms are not simply one-off joint actions, though a joint mechanism can be deployed on one occasion only. Nor are they
54
2. Joint Action simply joint actions performed regularly in a recurring situation, though they can be deployed regularly in a recurring situation. Rather they consist of a complex set of actions. The individual actions that constitute the complex set interlock, and there is at least some differentiation between them. Thus certain agents put themselves forward as candidates, and then other agents vote for the candidates. We can further distinguish between joint activity on the one hand, and joint practices on the other. Roughly speaking, a joint activity is a complex of differential, interlocking, joint actions and individual actions directed to some overarching collective end. So building a house and sailing an ocean liner are joint activities. Joint practices include games such as chess, and social practices such as courtship rituals. A joint practice consists of a series of actions that has a more or less definite beginning and endpoint. Such a series – which is to say, an exemplification of the practice – is repeatable, and in fact normally repeated. Each of the agents involved performs a number of different action types, and the actions of the various participating agents are interdependent. Joint practices are to some extent convention governed; that is, joint practices involve joint procedures. Further, we can distinguish between joint action and what I will call quasi-joint action. Quasi-joint action is action that just falls short of being joint. Certain cases of joint action involving free-riders are quasi-joint action. Suppose, for example, that agents in a community generally obey the law because it secures various collective ends that the agents have, such as peace and good order. Here the end of each agent that everyone (or most) obey the law might be a necessary, but not a sufficient, condition for each individually obeying the law. Suppose that in order to obey the law each agent requires – in addition to his end that the law be respected – some further end, say, to avoid imprisonment if he does not obey the law. In this example, each agent obeys the law on condition the other agents obey the law in order to realise their collective ends – almost. The fact that the other agents obey the law is not quite sufficient. The individual agent needs a further reason. Accordingly, we have here quasi-joint action (see Chapter 4 for a discussion of the obligation to obey the law). The point about quasi-joint action is that agents
55
Social Action would abandon the collective end if the action that they perform to realise that end did not also secure another individual end that each has. Further, we need to distinguish between joint action and various kinds of interdependent individual actions that are closely related to joint action. Suppose I have as an end to get to London from Oxford, and you have as an end to get to London from Oxford. You have a car, and just enough money to buy petrol for the Oxford to London trip. I have plenty of money for petrol, but no car. I suggest that I pay for the petrol if you give me a lift in your car. You agree. So we set forth. However, this is not joint action. For although I have an end to get to London, and achieve it on condition that you drive me, you are not similarly placed; certainly, like me you have as an end to get to London, but you will get to London whatever I do (although you are grateful for the petrol money). Since the interdependence is not two way this is not joint action. Now consider a second kind of closely related interdependence of individual action that is nevertheless not joint action. Consider the driving to London from Oxford scenario again. But this time let us assume that although you think you have enough money to buy petrol for the trip to London, you are mistaken; unbeknownst to you the money you had carefully saved for the trip was taken from your wallet the night before. Accordingly, when you take my money and use it to buy petrol, you actually need to do so, even though you are not aware that you need my money to get to London. So now there is two-way interdependence of action; I depend on you driving me to London; your driving to London depends on my petrol money. However, this is still not joint action because neither you nor I believe that your action is dependent on mine. So I intend to get to London by means of your car, believing that you will drive to London whatever I do; you intend to get to London by means of your car, and believe that you can do so whatever I do.3 collective end theory Recently the notion of joint action – sometimes referred to as shared action or group action or shared cooperative activity – has been receiving a certain amount of attention in the philosophical literature, and a number of sometimes closely related accounts have been 56
2. Joint Action propounded.4 Typically these accounts help themselves to concepts of intention, of goal, and of shared knowledge or belief.5 But there is divergence in respect of the particular notions of intention, goal, and belief, and of the precise way in which these notions can and are put to work in an overall account of joint action. Thus we have theorists using technical terms such as “we-intentions” or “collective intentions,” and constructing various competing sets of necessary and sufficient conditions for joint actionhood. I will now put forward a particular account of joint action, the Collective End Theory (CET),6 and defend this account against a variety of real and notional competing accounts. Basically, CET is the theory that joint actions are actions directed to the realisation of a collective end.7 However this notion of a collective end is a construction out of the prior notion of an individual end. Roughly speaking, a collective end is an individual end more than one agent has, and which is such that, if it is realised, it is realised by all, or most, of the actions of the agents involved; the individual action of any given agent is only part of the means by which the end is realised. (See below for more detail on collective ends.) So a joint action simply consists of at least two individual actions directed to the realisation of a collective end. Accordingly, individual actions x and y, performed by agents A and B (respectively) in situation s, constitute a joint action if and only if: (1) A intentionally performs x in s (and B intentionally performs y in s); (2) A xs in s if and only if (he believes8) B has yed, is ying or will y in s (and B ys in s if and only if (he believes) A xs or is xing or will x in s); (3) A has end, e, and A xs in s in order to realise e (and B has e, and B ys in s in order to realise e); (4) A and B each mutually truly believes that A has performed, is performing or will perform x in s and that B has performed, is performing or will perform y in s. (5) each agent mutually truly believes that (2) and (3).
In respect of this account the following points need to be noted. First, with respect to clause (2), the conditionality of the action is internal to the agent in the sense that if the agent has performed a 57
Social Action conditional action then the agent has performed the action in the belief that the condition obtains. Moreover, the conditionality of the agent’s action is relative to the collective end. A xs only if B ys relative to the collective end, e. So it is not the case that B ying is a necessary condition for A xing, tout court. For example, it might be that A has some other individual end, e1, and that A xing realises e1 (in addition to e). If so, A will x, even if B does not y. But it remains true that A xs only if B ys relative to the collective end, e. Again, A xs if B ys relative to the collective end, e. So it is not the case that B ying is a sufficient condition for A xing, tout court. If for some reason A abandons the end, e, then even if B ys, A will not x. Second, the notion of a collective end is an individualist notion. The realisation of the collective end is the bringing into existence of a state of affairs. Each agent has this state of affairs as an individual end. (It is also a state of affairs aimed at under more or less the same description by each agent.) So a collective end is a species of individual end. Thus CET is to be distinguished from the accounts of theorists such as Gilbert, who favour irreducibly corporate notions of intention.9 A collective end is the same as an ordinary individual end in that qua end it exists only in the heads of individual agents. But it is different from an ordinary individual end in a number of respects. For one thing, it is a shared end. By shared end I do not mean a set of individual ends that would be realised by qualitatively identical, but numerically different, multiple states of affairs. Rather, I mean a set of individual ends that would be realised by one single state of affairs; so the coming into existence of the one single state of affairs in question constitutes the realisation of all the individual ends that exist in the heads of each of the agents. For another thing, although collective ends are shared (individual) ends, they are ends that are necessarily shared; they are not ends that are shared only as a matter of contingent fact. Suppose you and I are soldiers being shot at by a single sniper. Suppose further that we both happen to see the sniper at the same time and both fire at the sniper in order to kill him. I have an individual end, and you have an individual end. Moreover, my individual end is one that I share with you; for the death of this one sniper will not only realise my individual end, it will simultaneously realise your individual end.10 However, this is not yet a collective end, for the fact that I have as an end the death of the sniper – 58
2. Joint Action as do you – does not generate independent acts of shooting. For neither of us needs the other to realise the individual end we share. Accordingly, I may well retain my individual end, when I know you have abandoned yours. As it happens we have shared individual ends; but they are not necessarily shared individual ends. Collective ends are “open” to those that have them; collective ends include mutual true belief. So if end e is a collective end, then the agents who possess e mutually truly believe that they each have e. Third, a collective end is not an intention. In CET, ends are distinguished from intentions. The notion of an end is coordinate with that of a means, but this is not so for intentions. To say that an action was intended is not in itself to say that the action is a means or that it is an end. Indeed some intended actions do not involve means or ends. Consider a gratuitous act of arm raising. I intentionally raise my arm but for no purpose; I just do it, albeit intentionally. I have no end in so raising my arm. Moreover, I have no means; I do not raise my arm by means of (say) a pulley. Rather my arm is under my direct control, and intending to raise it is sufficient for it to be raised. On the other hand, if some action is a means then there is some other action (or state) which is its end, and if some action is an end some other action is its means. Accordingly, CET is to be distinguished from on the one hand the theories of “corporatists” such as Searle, and on the other from the theories of “individualists” such as Bratman and Tuomela.11 All these theorists try to construct an account of joint action purely in terms of intentions and beliefs (or intentions, beliefs, and desires). They do not make use of an irreducible notion of having an end. Fourth, mutual true belief is to be understood (roughly) as follows.12 A and B mutually truly believe that p, if and only if: (1) p; (2) A believes that p, and B believes that p; (3) A believes that B believes that p, and B believes that A believes that p, and so on. (Hereafter for simplicity I will speak not of mutual true belief, but of mutual knowledge or sometimes just knowledge, if the context makes clear that it is mutual knowledge/true belief that is in question.) Fifth, according to CET there is a distinction between joint action and successful joint action. For something to be a case of joint action it is necessary that the constitutive individual actions be performed; 59
Social Action but it is not necessary that the end for which they were performed be realised. In a successful joint action, by contrast, the individual actions must be performed, but, in addition, the end for which they were performed must be realised. Of course in the case of some joint actions – namely those in which the end is simply to bring it about that the set of individual actions is performed – the performance of the constitutive individual actions entails the realisation of the collective end. So in these cases the performance of the joint action entails its successful performance. It follows from this that for an action to be joint it is not required that it be a matter of mutual knowledge that the end has been realised. For it might be unsuccessful joint action. Moreover, mutual knowledge that the end has been realised is not required for a joint action to be a successful joint action.13 On CET a joint action might be successful without it being known by the participants that it was successful. Indeed it is not even necessary that the individual actions have to be mutually known (or even mutually truly believed) to have been performed – as distinct from mutually truly believed to be likely to be performed – in order for them to constitute a joint action.14 For even in the case where the end of the joint action is constituted by the individual actions, there is a distinction to be drawn between joint action known by the participants to have been performed and joint action not known by the participants to have been performed. By contrast, there is mutual knowledge of the interdependence of the individual actions. See clauses (4) and (5) of the definition of CET. This mutual knowledge is what distinguishes joint action from interdependent action that is not joint. Interdependent action not mutually known by the participants to be interdependent is not joint action. Sixth, CET relies on a fourfold distinction between ends, intentions, beliefs, and desires, but makes use only of ends, intentions, and beliefs.15 So CET does not have as constitutive elements affective attitudes such as desires or preferences. Having a desire (or preference) is not the same thing, on this view, as having an end.16 Ends, like intentions, are conative states. But desires are not necessarily conative. What would be an example of a non-conative purely affective desire? John might have a desire to make love to Mary, which is dependent on his believing that he cannot make love to Mary. If 60
2. Joint Action John thought he could ever make love to Mary he would lose his desire to do so. Moreover, consideration of desires agents have with respect to the actions of other agents seems particularly likely to yield cases of desire that cannot generate actions. The desire that a parent might have that her adult offspring act entirely independently of the desires of the parent, including this desire itself, seems to be a case in point. And there are further examples. Suppose a man, A, intentionally shoots another man, B, having as an end B’s death; and suppose also that A desires B’s death. Suppose, further, that A succeeds. If so, A has brought about B’s death by shooting him. Moreover, A intentionally brought about B’s death; it was not just a matter of A desiring B’s death (or even A desiring that A bring about B’s death) and A intentionally shooting B. For A might desire B’s death (or desire that he bring about B’s death) without intentionally bringing it about; and A might intentionally shoot B without intentionally bringing about B’s death. Moreover, A might desire B’s death (or desire that he bring about B’s death) and intentionally shoot B, but nevertheless A might not intentionally bring about A’s death. This would be the case if, for example, A did not have the death of B as an end when he shot B; rather, he had as an end to prevent B escaping while he (A) figured out a way to kill him that would not look suspicious. So A’s intention to bring about B’s death does not consist of A’s desire for B’s death, or A’s desire that he bring about B’s death, or indeed any other desire. Rather, A’s intention to bring about B’s death consists of A having B’s death as an end when he shot A. These examples strongly suggest that having a desire is not the same thing as having an end. Accordingly, if joint action is action directed towards a desired end, then this is some further (and external) property of joint action. It is a property that joint action has above and beyond the property of being action directed toward an end.17 On CET, affective attitudes are properly part of the theory of the motivation for joint action; they have no constitutive role in the analysis of joint action per se. CET is thus to be distinguished from belief/desire theories of human action.18 On the other hand, in distinguishing between ends and affective attitudes such as desires, I am not claiming that there are ends that are neither directly nor indirectly the object of some motivating affective state such as a desire. 61
Social Action Seventh, CET enables a distinction to be drawn between joint actions and joint contingency plans to act. Agents have joint contingency plans to do things such as inform others if the end of some joint action has been, or cannot be, realised; help others to perform their contributory actions, if difficulties emerge. Indeed if agents are involved in joint actions then it will be rational for them to help one another and to have joint contingency plans. Nevertheless such joint contingency plans are not necessarily involved in joint actions. In particular, agents perform simple one-off joint actions without having joint contingency plans. On CET, joint contingency plans are not in part constitutive of joint action.19 Eighth, CET does not, at least in the first instance, have any intrinsically deliberative notions among its constitutive elements. In its analysis of what might be called basic joint action, it does not deploy notions such as that of plan or subplan or strategy. Such notions will need to be introduced at a later stage to deal with complex formations of action and activity, such as joint enterprises, but at the ground level, so to speak, these notions are best left out.20 This point has implications for the kind of agents that perform basic joint actions, as opposed to joint actions involving future directed intentions, beliefs in relation to future directed intentions, and beliefs in relation to agents who are not co-present. The latter kinds of joint action involve elaborate chains of means–end reasoning. All that is required for basic joint action are intentions, ends, and beliefs, albeit including beliefs in relation to the actions, ends, and beliefs of others. I suggest that the level of agency required to have such attitudes is something less than rational agency. I will refer to such agents as basic agents or simply agents (see Chapter 1 for the distinction between mere agency and rational agency). Ninth, CET, as outlined above, constitutes a core theory of joint action. But this core theory can be extended to cover a range of different cases of joint action in respect of which the following points can be made:21
(i) There are many types of joint action in which it is not the case that the performance of a contributory action by any of the participating agents is dependent on the performance of contributory actions by all of the other participating agents.
62
2. Joint Action (ii) The actions x, y, and so forth, performed by participating individual agents, might not be of the same type. Each agent might have to perform a different kind of action. Further the “action” might not even be a single discrete action. It might be a set of actions of the same or different types, or it might be an activity or even a whole enterprise. (iii) The collective end might not be a state of affairs but rather a process. And this end might relate to the actions performed to realise it as a consequence or a result or as a whole to its parts. (iv) The realisation of the collective end can itself be simply a means to some further end, or it can be an end in itself, or it can be both means and end. (v) The collective end can be a lower order collective end or a higher order collective, that is, a collective end with respect to other collective ends. For example, the members of the executive of a political party might have as a collective end to reinforce the existing collective end of members of the party to build solidarity among the members. (vi) Higher order collective ends can be what David Schmidtz refers to as maieutic ends: “ends that are achieved through a process of coming to have other ends.”22 For example, the members of the board of directors of a company might have as a collective end to settle on the collective long-term goals of the company. These ends to be chosen are final ends; they are not to be chosen for the sake of some other ends, but they are, nevertheless, not yet settled on. (vii) The collective end can be one the participating agents need to keep in mind while they are performing the actions that are the means to that end; or it might be an end that the agents will only realise if they concentrate wholly on the means; or it might be an end of which the agents are not fully conscious at any time. An example of a collective end that those who pursue it are not fully conscious of might be some end pursued by members of a particular social class, say to exploit workers. (viii) The actions constitutive of joint action can be performed at the same time and place, or at different places at the same time, or at the same place at different times, or at different places and times. Indeed the actions of the individuals can be separated by thousands of miles and/or hundreds of years, for example the building of the Great Wall of China. (ix) The actions constitutive of joint action can be interactive – in the sense that in the performance of each individual action
63
Social Action the agent adjusts his action to the action of the other, for example when dancing – or non-interactive (in this sense). (x) Joint action typically, though not necessarily, involves either direct communication (of some sort or other) between each individual and every other individual, or indirect communication between the individuals – A communicates with B and B with C but not A with C.23
Tenth, CET enables a distinction to be drawn between joint action and what I earlier called corporate action. Corporate actions presuppose the existence of some group which functions as a group. Thus a football team or a university committee or an army are groups – or, as I will say, corporate entities – in this sense. Typically corporate entities persist over time and consist of individuals who stand in certain stable relationships to one another and have differentiated roles.24 What needs to be noted here is that the notion of a joint action is not the same as that of a corporate action. Thus a one-off joint action by two strangers might not be a corporate action. For example, assume A is walking north along a pathway and B is walking south, and they happen to reach a fallen tree blocking the pathway. A and B might more or less spontaneously each take one end of the tree and move it sufficiently to enable them to continue on their separate ways. This is joint action, but it is not corporate action. Agents A and B do not constitute a corporate entity. We have seen that CET requires a distinction between collective and individual ends. But it also requires a distinction between first and higher order (second or third, and so on, order) intentions. Or, more precisely, it requires a distinction between lower order intentions and intentions with respect to intended bodily movements or with respect to intended actions. (In what follows I take it that an action consists at least of an intention and a bodily movement.) First order intentions are tied to one’s own bodily movements. Call these intentions type (a) intentions. Higher order intentions are intentions one has in respect of other intentions, and especially in respect of actions (consisting of at least first order intentions and bodily movements). One can have higher order intentions in respect of: one’s own actions, or states of affairs brought about by one’s own actions – type (b) intentions; states of affairs one brings
64
2. Joint Action about relying on the actions of other agents, but where the other agents will perform the actions whatever one does – for example, I intend to be home by six o’clock partly in virtue of the actions of the train driver, but the train driver will drive the train whatever I do – type (c) intentions; actions of other agents where these agents have an intention to perform whatever action (within some specifiable range) I intend them (or tell them) to perform – for example, I intend that the bank teller pay me money from my account – type (d) intentions. The collective end theory makes use of intentions of type (a) and (b), but not (c) or (d). According to the collective end theory, an agent cannot have an end without having: (a) an intention to perform some action or other – although an agent might have an end without having settled on the precise means, that is action, to that end, and; (b) a belief that performing that action is a means, or part of the means, to realising that end. Having an end, unlike having a desire, is an intrinsically conative notion. The belief involved here can be a belief that the action will realise the end, or is likely to realise the end, or has some chance of realising the end. Moreover, if one performs an action as a means, then one necessarily has an end, and a belief that the action is a means to that end. While to have an intention is not the same thing as to have an end, and vice-versa, in some circumstances it is appropriate to describe an end that an agent has brought about as intended. In such cases, the agent both had the outcome as an end (the end of his first action) and intentionally brought about that outcome (second action). Suppose an agent intentionally makes a loud noise in order to startle a hearer. Here the end is a state of the hearer, and the means is the loud noise. Since the agent can produce the noise at will, and since he can more or less guarantee that the noise will startle the hearer, the agent can be said not only to have intentionally produced the noise (first action) but also to have intentionally startled the hearer (second action). Another kind of case involves outcomes that are difficult to bring about. Hitting the bull’s eye in darts would be one such case. Here it seems reasonable to claim that the darts player not only performed the action of throwing the dart (first action) but also the action of hitting the bull’s eye (second action). But in that case it
65
Social Action seems that the darts player intentionally (at least in some weak sense of intention) hit the bull’s eye. Certainly we can say that the darts player tried to hit, and succeeded in hitting, the bull’s eye. However, there are many cases where ends ought not to be assimilated to intentions, even very weak intentions. One such kind of case involves inconsistent outcomes. Suppose I apply to go to Oxford University and apply to go to Cambridge University, and do so in the knowledge that I cannot attend both universities simultaneously. I can intend to apply to Oxford and simultaneously intend to apply to Cambridge; but I cannot intend to go to Oxford and simultaneously intend to go to Cambridge; for I know this to be impossible, and one can only intend what one believes is possible.25 Accordingly, we need a notion of the conative state that I am in when I seek to go to Oxford and to Cambridge. I suggest the notion of an end; there is nothing problematic about possessing two or more ends that are inconsistent with one another. However, someone might still want to insist that if I got to Oxford I did so intentionally; it is just that the intention in question was a very weak intention. Such a person might argue as follows. Intentions are such that the outcome intended is subject to a progressively weaker belief condition. Thus at one end of the spectrum the agent strongly believes that his intended outcome will eventuate; at the other end he only views it as a remote possibility. And when the outcomes are only possibilities then an agent can have two or more intentions that are inconsistent with one another. I would prefer to distinguish such weak intentions from intentions subject to a belief condition; and do so by referring to them as ends. I especially would like to do so since they are, after all, ends; such a weak “intention” involves a distinguishable intended action, which is performed as a means to realise the weak “intention.” Moreover, should we not reserve the term “weak intention” for weak intentions that are not ends? Suppose I intentionally, but gratuitously, raise my arm, but do so in the knowledge that it has been damaged in a manner that has considerably lessened my control over it; accordingly, when I form the intention to raise it I have no great confidence that it will in fact be raised. I have a weak intention, but I do not have any end in mind. It is an entirely gratuitous movement. Now consider a modification to this scenario. Suppose that my arm recovers; I now have full control over it. But, remem66
2. Joint Action bering the period when I did not have full control, I now raise my arm for a reason; I do not raise it gratuitously, as was the case in the earlier scenario. What is my reason? I now raise my arm for the sake of raising it; that is, as an end in itself. And this reason cannot be expressed by citing a further intention, as distinct from an end. I do not intentionally raise my arm for the reason that I intend to raise it.26 Another kind of case in which the assimilation of ends to intentions is problematic involves certain kinds of future directed “intentions.” Suppose an ageing farmer refuses to sell his farm, but instead assiduously maintains it. The farmer’s reason for pursuing this otherwise irrational course of action is that he hopes that his son, who is still at school, will eventually take the farm over. The farm has been in the family for generations, and the farmer wants this tradition to continue. However, the farmer has no control over his son’s future choice of occupation; indeed, he does not even want to try to influence the boy’s decision, believing an important decision like this is the boy’s alone to make. Further, the farmer believes it is unlikely that his son will in fact decide to take the farm over. Certainly the farmer assiduously maintains the farm having as an end that his son take it over and, indeed, having as a further end that his son’s future child takes it over from his father, and so on. But the father does not intend that his son take the farm over; and certainly the farmer does not intend that his son’s child farm the land. The future actions with respect to which the farmer has hopes are the actions of the farmer’s son and of future generations of the family; these actions are not in any sense the actions of the farmer. So the farmer cannot intend these actions. On the other hand, since his hopes for the future are ones he is prepared to try to do something about – he refuses to sell just yet, and he maintains the farm – they are not just hopes; they are purposes or ends that explain his current actions. And there are still other kinds of cases where ends cannot be assimilated to intentions, even very weak intentions; kinds of cases more germane to our concerns in this book. These kinds of cases involve joint actions. Here the problem is not that the outcome is only a possibility, but rather that the outcome cannot be fully within the scope of one agent’s intention, since it is within the scope of the conative attitudes of many agents, including future agents. Consider an individual contributor to a charity fund set up to raise one mil67
Social Action lion dollars; assume a woman, in contributing, say, ten dollars, has as her end that one million dollars be raised. Moreover, this particular charity has a great record of past success, so this individual – and all the other contributors – has every confidence that the million dollars will be raised. The individual has a strong belief that the goal will be achieved. Nevertheless, this individual did not intend, or even try, to raise the one million dollars. Again, some worker involved in building a part of the Great Wall of China in the early stages of its construction did not intend, or even try, to build the Great Wall of China, even if he or she believed that the Great Wall would eventually be built; for the individual knew that he or she was one of literally millions of workers, including yet to be born workers, and that he or she would be dead hundreds of years before its completion. In these cases the agent performs an action (first action) in order that his or her end be realised; but it is false that the agent performed some second action that consists of intentionally bringing about that end. The worker did not intentionally build the Great Wall of China and the charity contributor did not intentionally raise one million dollars. Though of course all the agents involved in the construction jointly built the Great Wall and all the agents involved in the charity jointly raised the money. Let us take a closer look at the argument being offered for the assimilation of ends to intentions. The suggestion is as follows. If (i) agent A intends to perform an action x, (ii) A believes that a condition, c, obtains, namely, that B will y, (iii) A believes that A’s performance of x in c will realise some state of affairs, s, and (iv) A has s as an end, then (v) A can be said to have intentionally brought about s.
The problem with this argument is that it makes essential use of what I have referred to as type (c) intentions. In the argument here, agent A has a belief that agent B will perform a certain action, y. But it is crucial to this argument that A believes that B will perform action y irrespective of whether A does x. It is crucial that the agent have a type (c) intention. For example, the intention a commuter has to go home relies (in part) on the train driver driving the train. However, in cases of joint action there is mutual interdependence of 68
2. Joint Action action, and there are no type (c) intentions. Thus, where two people lift a table together, A and B both believe in the following: The other agent will perform his/her contributory action only if he/she believes that I will perform mine. In cases of joint action, one agent’s belief that the second agent will perform his contributory actions is not a belief about a stable condition; that is, it is not dependent on the first agent’s action or at least not dependent on the second agent’s belief about the first agent’s likely action. But in that case no single agent can be said to have brought about, or to have intended to bring about, the state of affairs brought about by the conjunction of the actions of all the agents. In joint action each agent has the state of affairs as a (collective) end, but no agent intended to bring it about. In summary, then, I suggest that cases of joint action are cases where agents cannot intend the outcome of the joint action, but only have the outcome as an end. This is simply because the end is realised by the actions of all (or most) of the agents, and not by any one agent alone. (For more discussion of this issue see the section in this chapter on Bratman’s account of joint action.) Having provided an account of joint action, I want now to apply it to a number of problematic cases. The first two examples are not cases of joint action, though it might be difficult to see why they are not. The third example is a case of joint action, though it might be difficult to see why it is. First, there is the case of the conformists. A conformist is someone who wants to do whatever everyone else is doing irrespective (within some range of action) of what it is that they are doing. (It is possible that a whole group of people are conformists with respect to some action.) Such a person does not have a collective end, though he intentionally performs an action, if and only if, he believes the others will perform it. Such person A has the end that A will do whatever the others are doing. This is not equivalent to A having the end that everyone (including A) will do the same thing. In the latter case, person A has as a constitutive element of their end that other agents perform some action; that other agents perform a certain action is part of what is being aimed at. In the former case, by contrast, A’s action is wholly constitutive of the end; it is only A’s own action that is being aimed at, albeit under the description, “action that everyone else is doing.” 69
Social Action Second, there is the case of the indirect individual end. Here agent A wants to get agent B to do something, say, to come to believe that p. Here, that B comes to believe that p is not something under A’s control. A cannot intend that B come to believe that p, though A can have it as A’s end. B infers that p on the basis of: (1) B’s belief that A told B that p; and (2) B’s belief that A is a reliable source in respect of matters such as the truth or falsity of p. Surely, B coming to believe that p is a collective end on my account – A telling B that p and B inferring that p being the relevant individual actions? Not so. Rather, that B comes to believe that p is A’s individual end. B does not have this end; B is not aiming at the state of affairs of B coming to believe that p. Rather, B’s end is that B come to believe whatever is true. Therefore, in this example, there are two individual ends, but no collective end. Of course, as a matter of contingent fact, the two different ends are realised by the same state of affairs, namely that of B coming to believe that p. However, B was aiming at this state of affairs under the description, “B coming to believe what is true,” while A was aiming at it under the description, “B coming to believe that p.”27 Third, there is the case of the rowers and the cox. This case raises the issue of coordination in the context of joint action. The cox and the crew are involved in joint action in order to achieve the collective end of winning the race. But the cox coordinates the actions of the crew. How is his or her second order action of coordinating joint with respect to the crew’s first order actions of rowing? That the rowers row, and (more or less) row in time, is first order joint action. However, at this level the cox is not really involved. Rather, the cox is concerned to exercise control in respect of what might be called the fine-tuning of the rowers’ first order actions. The cox issues directives so as to ensure that the crew members row at precisely rate r1, rather than at, say, rate r2. The cox’s directives are, then, second order actions. Moreover the crew’s responses to these directives – their fine-tuning of their actions – are not joint actions (or even joint modifications of actions). Rather, they are simply the compliance on the part of each individual rower to the directives. The actions of the crew and cox are, however, joint, insofar as the crew accepts that the cox has the role of issuing directives, and the cox, for his or her part, accepts the role of issuing the directives. 70
2. Joint Action So there is joint action at the third order level between the cox and the rowers. This is the level at which the crew members decide to accept directives from the cox, and the cox decides to issue directives, in order to achieve the collective end of winning the boat race. the individual intentions model Having elaborated my account of joint action I now want to defend it against a range of competing accounts. The first competing account seeks to analyse we-intentions – intentions to perform joint actions – as the individual intentions of participating agents to perform their parts of the joint action (in conjunction with appropriate beliefs about the actions of other agents).28 On this view an agent has a we-intention that, say, the piano be placed upstairs, if and only if the agent intends to lift his side of the piano and believes the other agent will lift the other side.29 (There is in addition the mutual belief condition to the effect that each agent believes that each agent mutually believes that each will lift his or her side of the piano.) On this account there is no reference to ends, but only to individual intentions and beliefs. As Searle notes, the problem with this kind of account is that it does not get beyond individual action.30 Consider a case where each agent is lifting his or her side of the table purely for exercise, still believing the other agent will lift the other side. Each may believe that the other will lift the other side of the table. These beliefs and, for that matter, intentions may be a matter of mutual knowledge. But we would still not have a case of joint action. For the action of one agent is not dependent on the action of the other. Rather, each agent achieves his or her goal independently of the actions of the other agent. Now reference might be made to an agent involved in joint action performing “his part of X.”31 Here X refers to the set of actions performed by the agents, and “his part of X” refers to the particular individual action performed by that agent. There are three ways of understanding the phrase “his part of X.” First, it might mean that although agent A’s action fits the description “his part of X,” the agent himself is not performing this particular action under that description. On this way of understanding the matter the objection to the analysis as outlined still stands. 71
Social Action A second way of understanding the phrase is that the agent is performing the action under the description “his (my) part of X,” but the agent does not intend that X be performed. So the agent, A, only wants to make sure A does A’s part. As Tuomela himself notes, this does not give us joint action.32 In the third way of taking the term, the agent is performing the action under the description “my part of X” and the agent does intend that X be performed. This is the way that the phrase will have to be taken. However, this intention that X be performed cannot be explicitly built into the set of conditions for we-intentions.33 For to so build in this notion would generate a circularity in the analysis of we-intentions. For it would turn out that a we-intention – an intention that we such and such – was to be analysed in part by recourse to an intention that we such and such.34 It might seem that the account could be supplemented by recourse to some notion other than that of a we-intention. The only obvious candidate is the notion of a desire, where desires are understood as being sharply distinguished from we-intentions.35 Let us term such desires, intention-neutral desires. The new account would in effect consist of the following claims: (a) each agent intends to perform his or her part of a joint action; and (b) each agent, intention-neutrally, desires that the joint action be performed. Here the notion of each doing his or her part must not bring with it the notion of a we-intention. Otherwise the circularity mentioned above will return. Now it is plausible – as I have already argued – that agents can, intention-neutrally, desire things while knowing that they cannot achieve them. Thus a prisoner may desire freedom more than anything else, yet know they will never be free. Indeed it seems that some desires are such that it is of their essence that they not be realised. Such desires are non-conative purely affective states. But it seems that matters must be different in the case of intentions. It is difficult to see how an agent could have an intention to do something and yet it be of the essence of that intention that the agent know that the action would be performed, or fail to be performed, entirely independently of that intention. So it is plausible that intentions and desires are logically distinct. If so, there is little chance of being able to rescue the account of
72
2. Joint Action joint actions by recourse to desires, as opposed to some additional conative notion such as intention or end. At any rate, to return explicitly to the individual intention account, the question is whether a number of agents could desire that they all perform some action, intentionally do their parts, and yet not perform a joint action. I believe that they could. Consider the following example. Two women, A and B, desire that they both exercise. Moreover they have mutual true beliefs that they are both exercising. However A exercises whether or not B exercises, and similarly for B. There is no interdependence of action. So there is no joint action. Yet on the account before us this is a case of joint action. For they both desire that both exercise. Moreover, since both in fact exercise, each is doing her part of the “joint action” that consists of both exercising. Here, each could be doing her part, though not under the description, “doing my part.” Here, each does not even think of her exercising as being a part of the conjunct of both exercising. Obviously there is no joint action. Alternatively, each could be doing her part under the description, “doing my part,” where it is understood that this involves no reference to a we-intention. Here, each desires that both exercise, and each exercises with the intention of doing her part of the “joint action” of both exercising. But it still might not be the case that each does her part for the reason that she desires that both exercise. For A and B might be schoolchildren who have been informed that they must do their part in certain designated gym activities or risk disciplinary measures. Accordingly, A exercises for the reason that this is her part, and if she does not do her part she will face disciplinary measures. Similarly for B. Fortunately for A and B they both desire that they both exercise, but this desire is not the reason that each intends to do her part. The intention of each to do her part is tracking the individual desire each has that she not be punished. But each doing her part in order to avoid her own punishment is not joint action. Now the example could be transformed into joint action if each exercised in order that they both exercise; that is, if each doing her part meant each exercising intending that both exercised. But this reintroduces we-intentions, and hence the problem of circularity.
73
Social Action In fact there is a way out of the problem that confronts both the original individual intention account and its supplemented version. The way out is to have recourse to the notion of having an end – where ends are not desires or intentions. On this account of joint action each intends to do her part because each has as an end that both exercise. Thus if each exercises in order that she avoid punishment there is no joint action. For the end that each has can be realised by each agent exercising on her own, and this is so notwithstanding the fact that each desires that both exercise. So recourse to the notion of an end can solve the problems that beset the individual intentions account, but obviously it does so by simply transforming that account into CET. michael bratman’s account The second competing account of joint action was developed by Michael Bratman.36 This account does not seek to reduce weintentions. Rather, we-intentions are simply a basic species of individual intention. Accordingly, an appropriate theory of joint action ought to help itself simply to intentions. However, there are two sorts of intention differing only in respect of their content. While both sorts of intention are individualist, one is the intention of an agent that he or she act whereas the other – the we-intention – is an intention that we perform some action, J. These latter intentions have the form, “I intend that we J.” According to Bratman, joint actions are actions involving what he calls shared intentions. You and I have a shared intention with respect to an action J if and only if: “(1) (a) I intend that we J; (b) you intend that we J; (2) I intend that we J in accordance with, and because of, (1a) (1b), and meshing subplans of (1a) and (1b); you intend that we J in accordance with, and because of, (1a), (1b) and meshing subplans of (1a) and (1b); (3) (1) and (2) are common knowledge.”37 I have a subplan if I intend that we perform the activity in a particular way. Thus, to take one of Bratman’s examples, agents A and B might each intend that they paint the house and each have as a subplan that they do this by painting it red all over. Here the subplans mesh. A case in which they do not mesh would be one in which, although each intended that they paint the house, A in74
2. Joint Action tended that they paint it red all over, whereas B intended that they paint it blue all over. Here there are two important claims that are of interest. First, according to Bratman’s definition, I intend that we J because of (in part) your intention that we J. For example, I intend that we paint the house because (in part) you intend that we paint it. By contrast according to CET I have as an end that we paint the house and you have as an end that we paint the house, but I don’t have this end because you do. The most that could be said is that I have it as an end because (in part) I believe that you will do your part by, say, painting the bedroom. What is true is that on CET I would not have this as an end if you did not, since in the circumstances I could not realise the end by myself, and you would not do your part unless you also had that end. CET should be preferred to Bratman’s account in this respect since Bratman’s definition rules out certain cases of joint action, namely ones in which one agent does not care whether the other agent has the same joint intention or end, just as long as that agent performs that agent’s contributory action. Suppose I have as an end that we dig a tunnel under the English Channel. You are in France and will dig from Calais, and I am in England and will dig from Dover. The tunnels will connect in the middle of the English Channel giving me access to Europe, and you to the United Kingdom. However, as an Englishperson I don’t like the French and as a Frenchperson you don’t like the English. Accordingly, I don’t care whether you get to go to the United Kingdom or not. Likewise, you don’t care whether I have access to Europe. I don’t care whether you have as an end that we dig the tunnel, or whether you are simply digging a tunnel from Calais to the middle of the English Channel for a bet. My only concern is that you in fact dig a tunnel from Calais to the middle of the English Channel. So I don’t have as an end that we dig the tunnel (even in part) because you have as an end that we dig the tunnel, though as it happens you do have as an end that we dig the tunnel. Your sentiments mirror mine. Your only concern is that I dig a tunnel from Dover to the middle of the Channel, and you do not have the end that we dig the tunnel (even in part) because I have that as an end. Nevertheless we are cooperating to build a tunnel, and it is a paradigm case of joint action. In this kind of case there is interdependence of ends – as well as, of 75
Social Action course, interdependence of action to achieve that end. In the circumstances I would not have as an end that we build the tunnel unless you had that as an end. For if you did not have that as an end then, as it happens, you would not dig your stretch of the tunnel and I would not be able to realise my end. But nevertheless I do not have this end because you have this end. At most I have this as an end because I know that you will dig a tunnel from Calais to the middle of the Channel. I would not abandon my end if I came to believe that you did not have as an end that we dig the tunnel, but were simply digging from Calais to the middle of the Channel for a bet. Similarly, you would not abandon your end if you came to believe that I was simply digging from Dover to the middle of the Channel in order to win a bet. The second aspect of Bratman’s definition is the lack of unclarity concerning the relationship between the first agent and the individual contributory action or actions of the second agent; that is between, for example, me and your individual action of painting the bedroom. According to CET the relationship is simply that I believe that you are painting (or will paint) the bedroom. But this is not Bratman’s view. His analysis involves no reference to any such beliefs. Rather, on his account I intend that you paint the bedroom and do so in virtue of my joint intention that you and I paint the house. Bratman is aware that the idea of A intending B’s action might be thought to be problematic and he offers arguments in order to allay this worry.38 One of Bratman’s arguments makes a general appeal to the notion of future directed intentions. I accept that an agent can have future directed intentions with respect to his own actions. Moreover these intentions typically rely on beliefs about the actions of others. Thus I intend to go to London next year and this intention can in part depend on my belief that Qantas will have flights to London. (Here I believe, but do not intend, that the Qantas pilots will fly to London.) But Bratman has not shown, as he needs to, that there are relevant future directed intentions with respect to the actions of other agents, where these actions are essentially under the control of these other agents. Bratman’s most important argument turns on the use of an example in which an agent, Abe, intends to pump water and is assisted by various other agents – especially Diane – who intentionally in76
2. Joint Action crease the water pressure, and thereby enable Abe to pump water.39 The point about Diane is that her intention to increase the pressure is in response to Abe’s intention to pump water by means of pushing the pump handle. Specifically, Abe believes that Diane will increase the pressure if she knows that Abe intends to push the pump handle. This is not because Diane is engaged in a joint action with Abe; she does not have as an end that water be pumped out as a result of her doing her part of increasing the pressure and Abe doing his part of pushing the pump handle.40 Rather, Diane has as the end of her action merely to do what she knows Abe desires that she do, namely increase the water pressure. Nevertheless, Diane’s action is not under the control of Abe; she is a free agent, albeit a friendly one. Now I agree with Bratman that his example shows that Abe can intend to pump water, and do so in part because Diane will be responsive to this intention of his. But I cannot see that his example demonstrates how the freely performed actions of others can nevertheless be subject to our intentions. For in the first place Abe does not intend that Diane increase the pressure; he only believes that she will. In the second place there is no we-intention involved in the example. Abe intends that Abe will pump the water; he does not intend that Abe and Diane will pump the water. Let me now put forward some counter-examples to Bratman’s position. Consider this Superman example. A and B wish to insert a device into Superman’s brain in order to monitor his neural activity. However, the miniature device is not quite small enough to penetrate Superman’s skull. Accordingly, they devise a device that has two different parts, each part being in the form of a very small bullet. The two parts taken together constitute one single monitoring device. Thus A will fire one bullet and B the other bullet. But both bullets need to end up in Superman’s brain. On Bratman’s view it must be the case that A intended that B put a bullet in Superman’s brain. However B intended that B put a bullet in Superman’s brain. But it cannot be that A intended that B put a bullet in Superman’s brain, and that B intended this. (Or at least this cannot be so on the assumption that both intentions caused that action the performance of which constitutes the realisation of each of these intentions, namely the action of B putting a bullet in Superman’s brain.) For if both these intentions were realised, and if both were causally efficacious, then B’s action would be over77
Social Action determined. A’s intention would be a sufficient condition for B’s action, and so would B’s intention. But B’s action is not overdetermined. For B’s intention that B put a bullet into Superman’s brain was necessary for B putting that bullet into Superman’s brain. So no intention of A’s that B perform this action was sufficient. It might now be argued that, notwithstanding the above, A’s intention was causally efficacious by virtue of being a necessary condition for B firing a bullet into Superman’s brain. But this claim is also false. For B’s intention that B fire a bullet into Superman’s brain was not only necessary, it was also sufficient. So A’s intention cannot even be a necessary condition for B’s action. It might be now argued that although B’s intention that B put a bullet into Superman’s brain was caused by B’s intention to do this, and not by A’s intention that B do this, nevertheless that intention of B was itself caused by some other intention of A’s. But if so, what is this intention? A did not, for example, issue an order to B to fire the bullet at Superman. For B is not in any way under A’s control, and A does not believe that B is under A’s control. Indeed, it may be that A did not in any way even seek to persuade B to put a bullet in Superman’s brain. This is entirely consistent with A believing that in fact B will put a bullet in Superman’s brain. The only other of A’s conative mental states that is relevant here is A’s end – or, by Bratman’s lights, intention – that there be a monitoring device (comprising the two bullets) in Superman’s brain. Is this state either causally necessary or causally sufficient for B having the intention to put a bullet into Superman’s brain? No it is not. For once again it is B’s mental states – not any of A’s intentions – that cause B to have the intention to put a bullet in Superman’s brain. The mental states of B in question are B’s end (or intention) that there be a monitoring device in Superman’s brain and B’s belief that A will fire the second bullet into Superman’s brain. These mental states are singly necessary, and jointly sufficient, to cause B to fire the bullet he fired. At this stage it might be suggested that, notwithstanding the above, B’s end (or intention) that there be a monitoring device in Superman’s brain is itself caused by some further intention of A’s. It could then be concluded that this further intention of A’s indirectly caused B’s intention to fire a bullet. How is this process of indirect causation supposed to work? First, this (alleged) further intention 78
2. Joint Action of A caused B to have the end (or intention) that there be a monitoring device in Superman’s brain. Second, B’s end (or intention) that there be a monitoring device in Superman’s head was causally necessary for B to have the intention to fire a bullet into Superman’s brain. As we have seen, the second step is correct. Unfortunately, the first step is incorrect. The only candidate for being A’s further intention is A’s end (or we-intention) that there be a monitoring device in Superman’s brain. But obviously the end (or intention) that there be a monitoring device in Superman’s brain is not the same mental state as the intention that B have as an end (or intention) that there be a monitoring device in Superman’s brain. So this recourse to A’s end (or intention) that there be a monitoring device in Superman’s brain does not help Bratman. In response to this it might finally be suggested that I have misdescribed A’s end (or intention). I have characterised it as A’s end (or intention) that there be a monitoring device in Superman’s brain; but it should be characterised as A’s end (or intention) that A and B (jointly) bring it about (by means of firing the bullets) that there be a monitoring device in Superman’s brain. I have two replies to this. First, the existence of this new and different characterisation does not demonstrate that I have misdescribed A’s end (or intention). Indeed, this new characterisation presupposes my characterisation. For if A and B are to (jointly) bring it about (by means of firing the bullets) that there be a monitoring device in Superman’s brain, then A (and B) must have as an end (or intention) that there be a monitoring device in Superman’s brain. Second, this new and different characterisation is a description of an end (or intention) that A does not have in my example, and does not need to have. For it is a higher order end (or intention) to the effect that A has as an end (or intention) that A and B perform the joint action in question; so it is not, and cannot be, part of the analysis of a joint action. I conclude that the Superman scenario stands as a counterexample to Bratman’s analysis, and in particular to his requirement that agents involved in joint actions intend one another’s contributory actions, as opposed to merely believing that those actions have been, are being, or will be performed. Let me now consider an example of a joint action in which agents perform their contributory actions at different times. Suppose A and B are members of a 400 metre relay team. Each participant runs 79
Social Action 100 metres and passes a baton to the next relay member. This is joint action; but it is joint action in which the constitutive individual actions are performed sequentially rather than simultaneously. Now imagine an around the world relay. The relay takes years to complete and involves thousands of participants. Most of the participants will never meet one another. Assume further that many participants do not decide to participate, indeed have not even heard of the relay, until the relay is actually under way across their particular country. It would be absurd to claim that each participant has intentions with respect to the actions of every other participant. Assume the relay has been going for a number of years. Now consider the present participants in the relay. Many of these have no intentions whatsoever with respect to many of the actions of other participants. Specifically, these present participants have no intentions with respect to the past actions of many past participants. For at the time that these past actions were being performed these present participants did not even know that there was a relay being run. What we have here is a case of joint action taking place over time. The agents have a shared end, namely to complete the running of the relay, and they believe other participants have done, are doing or will do their parts. But the participants do not have intentions with respect to one another’s contributory actions. Such cases constitute a further category of counter-examples to Bratman’s thesis that participants in joint actions intend one another’s contributory actions. What is the way out of the problems that confront Bratman? I suggest that the notion of a we-intention needs to be abandoned in favour of the notion of an end. If this were done the Superman example would be construed as follows. Agent A intentionally put a bullet in Superman’s brain having as an end that there be a miniature monitoring device in Superman’s brain. Similarly for agent B. Here agent A did not intend that agent B put a bullet in Superman’s brain, but rather, A believed that B would. On the other hand, A did not simply desire that there be a miniature monitoring device in Superman’s brain. Agent A performed an action in order to try to realise this desire. Agent A had this outcome as an end. The same kind of construal is on offer for the relay example; each agent intentionally ran his or her segment of the relay, believed that the others did or would run their segments, and had as an end that
80
2. Joint Action the whole relay would be run. However no agent intended that a second agent run the segment run by that second agent. So these counter-examples can be accommodated, but only by transforming Bratman’s theory into CET. cohen and levesque’s account The third competing account is propounded by Cohen and Levesque.41 This account holds that joint action requires some notion of a goal.42 (I take it that the notion of a goal is pre-theoretically equivalent to the notion of an end.) In fact, on this view, joint action requires not one but two goals. For joint action involves, in addition to the first order joint goal of the action, a second order joint goal that there be mutual knowledge among the participating agents that the first order joint goal has been, or cannot be, realised.43 This second order joint goal is held to be in part constitutive of joint action.44 I disagree with this account. Cohen and Levesque’s second order goal is that there be mutual knowledge that the first order goal has been or has not been performed (or perhaps that it now cannot be performed).45 Note that this mutual knowledge is not mutual knowledge, or even mutual true belief, that the joint action is likely to be performed. To require this second order goal is to go too far; it results in an incorrect analysis of joint action. A joint action might have been performed even though the participants cannot come to know that it has been performed, and know that they cannot know this. It follows that each cannot have such mutual knowledge as a goal. For example, suppose A and B are spies who contrive and execute a joint mission to steal two sections of a military strategy located in different places. A must get one section and B the other. Once stolen the sections are individually forwarded to some friendly government official. One section is useless without the other. Having agreed to undertake the mission, each agent must make no further contact with the other or with the government official. After successfully stealing and forwarding a section, each agent is to assume a new identity. Any other course of action would be dangerous for the agents. A and B successfully complete their joint mission. However, they perform this joint action notwithstanding the fact that this did
81
Social Action not become a matter of mutual knowledge. Crucially neither A nor B have at any time a goal to inform the other regarding the success of the contributory actions or of the mission. For A and B know all along that they must not contact one another and that they will never know whether the joint action was successful. Take another example. Agents A and B might be jointly constructing a house. A might complete A’s part of the house, say the carpentry, and then die. B might then go on to complete the house. The joint action of building the house has been performed, notwithstanding the fact that there was not mutual knowledge of this completed joint action. Crucially, it might never have been a goal of B to inform A of the completed joint action. For A and B might have known that A would die after completing A’s part of the construction work, but before the house was completed. A further example would be one involving an end that was external to the agent’s contributory actions. Suppose five men are unlawfully hunting a tiger that is a member of an endangered species of tiger. Assume they are armed with spears. They know the tiger will probably survive if only one or two spears hit the tiger, but they believe four or five spears will probably kill it. Each throws his spear having as an end that the tiger be killed. All five throws are successful. However the tiger, although barely alive, limps off into thick scrub. The men dare not follow the tiger, since although it is probably dead, they now have no weapons and it just might have survived. Assume that in fact the tiger dies though this is not known to the five men. Surely this is nevertheless a case of joint action, indeed successful joint action. It is just that none of the participants know it to have been successful joint action. They do however mutually know that they have performed the joint action in the sense that they have completed their individual actions. But in this kind of case involving an external end, there is a distinction to be made between a joint action – consisting of the individual actions (already performed) and the collective end (not realised) – and a successful joint action (the end having been realised). To allay any doubts that in the tiger example there has been a successful joint action not known to have been successful, consider the fact that someone else might well know that the joint action was successful. Suppose a well-armed traveller happened to witness the spear throwing, followed the tiger into the scrub and then watched 82
2. Joint Action it die. The traveller knows that the joint action was successful. Moreover, the traveller reports the unlawful killing of the tiger to the authorities, who investigate the matter, including by recourse to DNA testing. As a result the men are tried in absentia and found guilty. Unfortunately, the men are never found. Moreover, they never come to know that they were being sought for the killing of the tiger or, indeed, that they had in fact killed the tiger. Clearly, it is a case of successful joint action even though not known by the participants to have been successful. These examples demonstrate, contra Cohen and Levesque, that in joint action there need not be mutual knowledge with respect either to the actual past performance of the individual actions or the realisation of the end to which they are directed, and that since participating agents can know this, it cannot possibly be a requirement on joint action that agents have such mutual knowledge as a goal, joint or otherwise. We should rather prefer the weaker mutual knowledge conditions of CET. The tiger example also demonstrates that Cohen and Levesque’s account has too narrow a focus. Their account focuses only on joint actions that are processes. In such cases the goal is simply the completion of the process. Accordingly, if each performs his or her part and there is mutual knowledge that each has done so, then there will of necessity be mutual knowledge that the process has been completed and the goal therefore achieved. But there are many cases of joint action, such as the tiger example, in which the goal or end is not internal but external to the actions performed by participating agents. In such cases, even if there is mutual knowledge that each has performed his or her action, and even if it is a joint goal that there be such mutual knowledge, it does not follow that there is a joint goal that there be mutual knowledge that the goal of the individual actions has been achieved. For the agents might know that they cannot have mutual knowledge that this goal has been achieved, and therefore it cannot be a joint goal that there be such mutual knowledge. Once again in distinguishing and accommodating both kinds of joint action CET is to be preferred.
83
Social Action john searle’s account A fourth competing account of joint action is the corporatist one propounded by Searle.46 Searle holds that joint action involves irreducibly corporate intentions. Corporatist theorists tend to run together joint action and what I have called corporate action. It is conceivable that the latter involves some irreducibly corporate notion of intention though I argue that this is not so in Chapter 5. But at any rate even if corporate actions involve irreducibly corporate intentions, it would not follow that joint actions do. My general response to the claim that joint action, as opposed to corporate action, is irreducibly corporatist, is simply to present CET. CET is an individualist account of joint action. Obviously CET has to be put to the test. But I believe that CET can accommodate the counterexamples advanced by corporatists. Searle’s example of would-be capitalists aiming at their own interest while having a collective intention to benefit all is easily accommodated.47 He is correct in thinking that the collective intention is not reducible to individual intentions. But he is wrong to think that it is therefore an irreducibly collective intention. Rather, it is a collective end of the kind deployed in CET. Let us look at his example. In one reading of Searle’s example, the agents’ actions are overdetermined. They have both individual and collective ends. Each has the individual end to get rich for the sake of being rich. In addition, each has the collective end of the common good, which each believes will be realised by each pursuing his or her own interest. Qua actions directed at the collective end, the agents are involved in joint action. Qua actions performed simply to secure their individual ends, the agents are involved in individual action. On the other way of reading Searle’s example each agent has one ultimate end, namely the common good. However, paradoxically, each of the agents must aim at his own interest in order to achieve the common good. Here the individual end is not an end in itself, but simply a means to the further end of the common good. However, the individual actions do constitute a joint action; for they are actions that are performed, indeed only performed, in order to secure the common good.
84
2. Joint Action margaret gilbert’s account A fifth competing account is the corporatist one of Gilbert. We will first look at Gilbert’s corporatist account on its own merits, and then consider her examples and accompanying arguments against individualist accounts. According to Gilbert, two or more agents engaged in joint action are “plural subjects” with a “shared intention” to perform a joint action. Here the shared intention is not reducible to individual intentions or ends, and the joint action is not reducible to individual actions (or individual actions in conjunction with individual intentions or ends). Indeed, Gilbert holds the view that plural subjects can be ascribed intentions, beliefs, and moral responsibilities that are not possessed by the individual human agents that (in some sense) constitute the plural subject. Gilbert has presented her views in a number of different places, but not always in a consistent or luminous manner.48 At any rate, she is committed to the following definition of the key notion of interest to us here, namely shared intention: “Persons P1 and P2 have a shared intention to do A if and only if they are jointly committed to intending as a body to do A.”49 Here the notion of a joint commitment is itself an irreducibly collectivist notion: “So it is not the case that each of them is the sole author of part of the joint commitment, let alone of the joint commitment as a whole. Rather they comprise the author of the joint commitment together.”50 Moreover, the notion of a commitment is itself somewhat unclear. At one place, she offers an apparently non-normative account: “If I have decided to do A, I am subject to a commitment to do A.” At others commitment seems essentially normative: “the pretheoretical judgment that shared intentions involve obligations with corresponding entitlements responds to the presence, given a shared intention, of a joint commitment.”51 I have two responses to the above analysis of shared intention. First, the crucial notion of a commitment, including a joint commitment, conflates the normative notion of putting oneself under an obligation with the (non-normative) notion of non-reversibility. Conative states, such as intentions and decisions, are (at least partially) non-reversible. The image here is of jumping off a cliff; at some
85
Social Action point the person jumping “commits” himself, and so there is no going back. Similarly, for intentions and decisions. However, this feature of non-reversibility has nothing whatsoever to do with normativity; in particular, it is an entirely different sense of commitment from that of putting oneself under an obligation, for example by making a promise. If commitment is taken to imply non-reversibility only, then the definition of shared intention in terms of joint commitments seems to be vacuous; shared intentions are (more or less) joint intentions. If on the other hand commitment is taken in the normative sense of putting oneself under an obligation, then the definition is false; intentions, including shared intentions, are not necessarily puttings of oneself under obligations (or a necessary result thereof). Second, the other crucial notion of intending “as a body” to do A conflates the notion of intending as the member of a group with the notion of an individual intention, albeit an individual intention attaching to an entity which has agents P1 and P2 as parts. If the former, then there is no apparent barrier to the analysis of the shared intentions in terms of the individual intentions of P1 and P2. If the latter, then we do not have an intention, shared or otherwise, of P1 and P2; rather, we have the intention of the (alleged) single entity that has P1 and P2 as its parts. Let us turn now to Gilbert’s examples and accompanying arguments against individualist accounts, including her argument that joint actions imply a normative dimension. I will seek to rebut these arguments and indicate how CET would handle alleged counterexamples. Let us consider Gilbert’s mushroom-picking example. This example is put forward in the context of Gilbert’s pursuit of an analysis of social action that helps itself only to individualist notions. The point of the example is to show that such a pursuit is hopeless; according to Gilbert, such analyses in terms of individualist notions will inevitably fail. Gilbert points out that a group of mushroom pickers who individually pick mushrooms are not engaged in social action. Moreover, they would not be engaged in social action even if they were mutually aware of one another’s mushroom-picking activities. Nor would the fact that one mushroom picker had a brief conversation with another mushroom picker turn their mushroom picking into social action. Finally, even if all the mushroom pickers 86
2. Joint Action engaged in an occasional two-person conversation, this would still not be sufficient for their mushroom picking to count as social action. I agree. More importantly, CET yields the same result. According to CET, this mushroom-picking activity is not social action, or even joint action; for the individual mushroom-picking actions are not directed to a collective end. So the mushroom-picking example does not show that an individualistic analysis cannot succeed. It merely shows that Gilbert did not have an adequate individualist account. Gilbert correctly claims that from the fact that agents A and B (successfully) perform a joint action properly describable as Xing, it does not follow that A performed some individual action properly describable as Xing. Thus A and B might be dancing together, but it does not follow that if A were doing the same movements alone A would be dancing.52 But this is not required of an individualist account, and certainly not of CET. CET, in helping itself to the notion of a collective end, enables a joint action description that goes beyond the description of any of the individual actions considered without reference to the collective end. Thus Fred can dial a number and Bill can pick up a receiver and there is the joint action of communication. But this is only the case if each had as a (collective) end that there be communication. In the dancing example each has as an end that they dance. Here it is the bodily movement of each, coupled with the relations between those individual bodily movements, that is the end that each has. So this end is not adequately described by a description of the bodily movements of one of the partners. CET is also able to accommodate Gilbert’s babysitting example.53 Ted and Carol are babysitting Jo’s baby, but Carol decides, with Ted’s permission, to go out to the movies. Notwithstanding her decision to go to the movies, Carol is still responsible (jointly with Ted) for babysitting Jo’s baby and she has not abandoned that responsibility. The point here is that two agents can have as end an outcome which will in fact be brought about by one of them. Ted and Carol have as an end that one or other of them is always present with the baby. Both Ted and Carol make sure that end will be realised before Carol goes to the movies. The test of this would be, say, Carol not going to the movies if she knew Ted would fall asleep after an hour or so. Of course Carol does not go to the movies in 87
Social Action order to realise the end of babysitting Jo’s baby. Carol does whatever is necessary in order to ensure that either she or Ted is present with the baby. It turns out that in her case that necessary action is quite minimal; it is the action of ensuring that Ted will stay with the baby. Let us turn finally to Gilbert’s walking example.54 Imagine Sue and Jack are going for a walk together. Each has as a goal that they walk together. Here Gilbert switches between two very different conceptions of having a goal. What she describes as a personal goal could have one of two forms. An agent could have as a goal that he or she go walking in your company. Or the agent could have as a goal that we go walking together. On the first conception the agent is simply concerned to orient his or her own behaviour to that of the other person. It is not part of his or her goal that the other person perform such and such an action. This is not a collective goal or – as I would say – a collective end. It is simply an individual goal or end. On the second conception, by contrast, it is part of the goal that the other agent perform such and such an action. The agent has a goal that we do such and such. It is important to make this distinction and keep firmly in mind that it is the second conception that Gilbert must undermine. Otherwise her arguments may seem to have more force than they actually do. Gilbert points out that the fact that two agents each have the same personal goal in performing their actions is not sufficient for these actions to constitute joint action. For each might not know that the other has this personal goal. According to CET, individual actions only constitute joint action if there is a collective end, and if the ends of each of the agents is a matter of mutual knowledge between them. Gilbert countenances this possibility – albeit with a certain amount of sliding between personal goals and collective ends – but argues that even mutual knowledge of a joint goal is not sufficient for individual actions to constitute joint action. Her argument for this claim is basically that the normative dimension of her example has not been adequately captured. The normative dimension in question goes beyond considerations of prudence and reasonable care for the well-being of others and yet falls short of moral obligations. This alleged normative dimension generates an obligation on the part of each to do his part, and an entitlement of each to rebuke the other if either does not do his part.55 The short answer is simply to deny that cases of joint action do 88
2. Joint Action necessarily involve a normative dimension beyond the considerations of prudence and reasonable concern for the well-being of another person. Consider the one-off situation I described earlier in this chapter involving two strangers who jointly move a fallen tree blocking their pathway. The relevant prudential considerations are simply that both participants in the joint action seek the realisation of the collective end, and yet it will not be realised unless each performs a contributory action. Accordingly, each is likely to try to get the other to do his part by appealing to the interest of the other. In addition there might be an element of concern for the other person; other things being equal, each does not want the other to be inconvenienced by having his pathway blocked. But in this example there is no additional normative dimension that generates an obligation on the strangers to move the log, and would entitle the one to rebuke the other if he refused to assist. This is not of course to deny that there are many cases of joint action in which there is a further normative dimension. Here there appear to be two kinds of example. First, there are joint actions in which there is the possibility of free-riding. These typically give rise to the moral obligation to do one’s part as a matter of fairness. I analyse these cases in Chapter 4, but my point here is simply that in such cases there is no additional normative dimension that stops short of a moral obligation. Quite simply there is a moral obligation not to free-ride. Moreover the collective end theory is able to show why this is so. The point about the free-rider, as opposed to the nonparticipant, is that the free-rider has the collective end and the collective end is a good. The free-rider would contribute if it was necessary to realise the collective end. However, the free-rider is unfair to the others in wanting the benefits of the joint action without performing a contributory action. Second, there are joint actions performed by occupants of institutional or professional roles. These typically have an institutional normative dimension. Thus a firefighter may be under an obligation to perform a part of the joint action of putting out a fire. Or a police officer might be under an obligation to walk together with a fellow officer while walking the beat; for the policy of the police service might be for officers to work and walk in twos in order to minimise risk to themselves. Moreover, Gilbert’s example of walking together may have a normative flavour in virtue of the fact that it is normally 89
Social Action friends who want to walk together, and friendship generates expectations and entitlements. Thus it may very well be upsetting if one’s friend suddenly leaves one’s company. At any rate, however things are in such cases involving friendship, the point is that Gilbert fails to show that in order for an action to be a joint action it is necessary that it have any additional normative dimension, let alone the institutional normative dimension she advocates. Moreover, my examples provide support for the proposition that the institutional normative dimension favoured by Gilbert is not necessarily a feature of joint actions. Finally, I deny that cases of joint action in which there is the additional institutional normative dimension require recourse to Gilbert’s collectivist ontology of plural subjects and collective beliefs. (The already demonstrated lack of necessity for such an ontology in many cases of joint action – those lacking an additional institutional normative dimension – already casts serious doubt on Gilbert’s account.) In Chapter 6 I propound an individualist account of the institutional normative dimension in question. We need to distinguish between institutional action and joint action per se. The analysis of joint action does not require any collectivist ontology of the sort espoused by Gilbert. Having provided a detailed account of joint action, and defended it against rival accounts, I now want to turn to a variety of central categories of social action, and put my notion of a joint action to work in the analysis of these categories of social action. I begin with conventional actions, the subject matter of the next chapter.
90
3
Conventions
In this chapter I will provide an account of one of the main categories of social action, namely actions performed in conformity to conventions. I have already suggested that while joint actions are not necessarily social actions, nevertheless the concept of a joint action is of fundamental importance in the understanding of social action. In this chapter I make good on this claim. For I will argue that conventions are in fact a species of joint action. Conventions play a large part in our collective lives. When we exchange money for goods, we do so in conformity to conventions. Our mode of dress and manner of eating is largely determined by convention. When we perform linguistically our actions are governed by conventions. The ubiquity of conventions suggests that an analysis of the concept of a convention would have an important place in the attempt to understand social phenomena. We find that philosophers, in particular, have used the idea of a convention freely and widely. Historically, the so-called social contract theorists, such as Hobbes and Locke, invoked the notion of a convention, and related notions of an agreement or contract, to provide the moral justification for obedience to government and the law.1 Rousseau put the proposition that political inequality depended on a kind of convention.2 More recently, philosophers have offered conventionalist accounts of analytic truths, for example, and rival theorists of speech acts have argued about whether speech acts are essentially conventional (or rule-constituted).3 So philosophers have employed the notion of a convention to make substantive claims, where the truth of these claims turned crucially on what exactly that notion of a convention was. It is difficult to see how these substantive claims could be satisfactorily adjudicated without an adequate account of conventions. As it happens, philosophers and others have propounded a number of accounts or theories of conventions. I will briefly consider two of these before 91
Social Action providing a detailed analysis and critique of the most elaborate and influential contemporary account, namely that of David Lewis. The first theory may be termed the explicit agreement theory. This theory is associated with the social contract theorists such as Thomas Hobbes. According to this theory, what makes a regularity in behaviour a convention is the fact that at some stage in the past people joined together and agreed to perform a certain type of action in certain circumstances. They now perform the conventional action in question because they are bound by their promises to do so. Aside from the problems of locating such a promise and of descendants of the promise makers being bound by the promises of the forefathers, there is the circularity involved in such an account. Promises or explicit agreements are made in some language, but languages, and for that matter the institution of promising, themselves involve conventions. The explicit agreement theory might be weakened to an implicit or tacit agreement theory. The invocation of the notion of a tacit agreement is associated with John Locke. On the tacit agreement theory there has been no actual promise; rather things are as if people had joined together and made promises. But more needs to be known about this tacit agreement – what constitutes the way in which it is as if promises had been made? Also what is to count as having entered into a tacit agreement? Non-resistance to doing an action could not, for example, count as having tacitly agreed to do it. In any case the weakening of the explicit agreement theory still leaves standing the intuitively implausible idea that conventions are followed, and ought to be followed, because the parties are morally obliged to do so having, in some sense, agreed or promised to do so. A second theory of convention is what we might term the rule theory of convention. Here rules are analysed in terms of commands. This theory has been propounded by Alf Ross.4 A convention is expressible as a command of the form “Do x in circumstances c,” and a command backed by sanctions (since otherwise what differentiates the command from, say, a request?). But if conventional actions are actions performed in order to obey a command, then it can legitimately be asked: Who issued the command? Perhaps society issued it. But there is no such issuer in other than a highly meta-
92
3. Conventions phorical sense. And, if these commands exist, then where are they formulated? Many conventions are not written down anywhere. Indeed, if they were written down, then they would already exist in a conventional (linguistic) mode. But in that case the rule analysis of conventions would be circular for it would presuppose the existence of conventions. Finally, it seems that conventions are not necessarily backed by sanctions, and where they are, this is not necessary to their being conventions. People do not drive on the left if everyone else does because they’ll be fined if they do not, though in fact they might be fined if they do not. Rather, people drive on the left, if others do, because they want to avoid collisions; it would be irrational not to do so. Again, if a less strict definition of rule in the rule theory of convention is what is required, what is this less strict definition? Can the theory be weakened so as to avoid these objections, but at the same time retain its essential properties? It is not obvious how this could be done. Margaret Gilbert has in effect tried to meet these objections by propounding a view of conventions as the self-imposed fiats of the members of a social group.5 In effect, she holds that conventions are tacitly jointly accepted principles concerning what one ought to do. These principles are not necessarily commands backed by sanctions; so Gilbert’s view does not obviously fall victim to the objections made to the rule theory of conventions. However, there are other problems. First, there is a great lack of clarity surrounding the key notion of joint acceptance. This notion has to be strong enough to ground the normative principle to the effect that the persons in question ought to do such and such. If identifying joint acceptance with joint agreement, or quasi-agreement, is relied on to ground the “ought,” then Gilbert’s account is a version of the agreement theory of convention with all its attendant problems. On the other hand, if joint acceptance means something like joint decision, or joint intention, then, as we saw at the end of the last chapter, it does not follow that those who made the decision or formed the intention are necessarily obligated to perform the action in question; the “ought” has not been adequately grounded.6
93
Social Action Let me now turn in detail to David Lewis’s theory of convention, a theory that owes a good deal to David Hume’s early account of conventions as regularities in action that serve a common interest.7 lewis on conventions Conventions are, according to David Lewis, the solutions to coordination problems.8 Now, coordination problems are given by Lewis a technical definition in game-theoretic terms.9 They are also given a definition in terms of various kinds of conditional preferences.10 It is the definition (or at least definitions) in terms of conditional preferences that will concern us here.11 A conditional preference is a preference an agent has for an action, given that other agents perform certain actions. If these other agents did not perform those actions, our first agent may not prefer to perform his action. Let us call coordination problems, thus defined in terms of conditional preferences, formal coordination problems. There is, however, a different sense of coordination problem made use of by Lewis. This second sense is given in an informal way, but is clear enough. Suppose you and I wish to meet. There are two possible meeting places, a and b (and probably more). We don’t much care which of these places we meet at. What is important is that we both go to the same place, and not one of us to place a and the other to place b. If we both turn up at the same place we have solved the coordination problem. I will call such coordination problems “informal coordination problems.” An informal coordination problem, then, is the problem of choosing between more or less equally viable but mutually exclusive alternatives. It is important to note that the alternatives are equally conventions. If a regularity in action were set up whereby we always met at a, that would become the convention. But, likewise, if the alternative combinations of actions – each going to b – always took place, that would be the convention. The idea of conventional alternatives, as opposed to a single convention and a non-conventional alternative, is central to Lewis’s theory. The intuitive underlying idea of his theory is the idea of coordinating action with respect to alternative possible conventions. It is the idea of conventions as the solutions to informal coordination problems.12 My basis for this 94
3. Conventions claim is that, first, all his paradigms of convention appear to be of this type, for example driving on the left rather than the right-hand side of the road, using currency a rather than currency b, and so on. Second, all the various formulations of his definition of convention include a clause to the effect that for a given regularity in action to be a convention, there must be an alternative conventional regularity.13 Third, Lewis rules out social contracts as conventions because (among other reasons) he thinks they have no conventional alternative.14 Finally, Lewis states (in effect) that the notion of a convention is the notion of a conventional regularity with a conventional alternative.15 The role of formal coordination problems is to capture, in precise terms, the nature of informal coordination problems. Lewis frames his clauses specifying conditional preferences in order to capture the preferences of agents involved in informal coordination problems. Lewis’s theory is far more complex than the foregoing might lead one to believe. It’s not simply a matter of agents with particular conditional preferences performing a given action in a recurrent situation. For each agent has expectations about, as well as knowledge of, the actions of other agents.16 However, I need only concern myself here with the conditional preferences element in Lewis’s definitions, and with his clause requiring a conventional alternative to a convention. In fact it would seem conditional preferences and conventional alternatives are connected. Agents have conditional, as opposed to unconditional, preferences for conformity to a particular regularity, r, just because of the availability of a conventional alternative, r1.17 I prefer to conform to r if the others do. But my preference is conditional since I would prefer to conform to r1, if they did.18 Moreover, it is from the existence of this conditional preference structure that the interdependence of action involved in conventions arises. objections to lewis Suppose circumstances are such that either we meet at a, or we meet at b, or we make no attempt to meet, that is, perhaps some of us stay at home, perhaps some of us visit our mums, and so on. What distinguishes the latter combination of actions – to simplify matters 95
Social Action I will refer to it as everyone going their separate ways – from the former two? The combination in which we all go our separate ways is a non-conventional alternative to a convention. The combination such that everyone goes to a, is a conventional alternative to the combination such that everyone goes to b, and vice-versa. In general for any given convention there will be a non-conventional alternative to that convention. Thus perhaps the situation in which there is chaos on the roads is the non-conventional alternative to the situation in which each drives on the left-hand side of the road. (The situation in which each drives on the right would be a conventional alternative.) (Let us remember that in the above we have been speaking of conventions in the sense of regularities that solve informal coordination problems, for example driving on the left.) In the light of this distinction between conventional and nonconventional alternatives to conventions, let us look at the clause in Lewis’s account that defines the preferences of agents who are party to a convention. The clause in question is: “everyone prefers everyone to conform to (the regularity) R, given that at least all but one conform to R.”19 This clause specifies the preferences of agents conforming to regularities, where those regularities solve formal coordination problems. The question is whether or not this clause characterises the preferences of agents conforming to regularities that solve informal coordination problems. In short, does this clause capture the preferences of agents who are parties to a given convention? Many combinations of actions that constitute non-conventional alternatives to conventions can be construed in such a way that the agents performing these actions have the preferences stipulated in Lewis’s clause. Take the meeting-place example. The combination of actions that is the non-conventional alternative to a convention in this example is the combination such that each goes his separate way, say, each stays at home. But it may well be that each prefers everyone to stay home, given that at least all but one stay home. For instance this would be true if each preferred that no single agent waste his time going to a or b, when the others were not going there. Most examples of Lewisian informal coordination problems would involve non-conventional alternatives to con-
96
3. Conventions ventions. Many such non-conventional alternatives could be construed in such a way that they fit Lewis’s clause. Take another example, the clothes paradigm. Assume all but one go to a party in whatever clothes they like, but George wears his dinner jacket, thinking everyone else would be doing likewise. Here each (including George) may prefer that George had not worn his dinner jacket. For George does not like his dinner jacket, but only wears it if others wear theirs, that is if a convention to do so is in force. But the existence of these latter preferences does not make the regularity in action describable as everyone going to a party wearing whatever they feel like wearing (and doing so quite independently of what anyone else is doing) a convention. On Lewis’s theory then, nonconventional alternatives to conventions can be wrongly characterised as conventions. Nor is this a problem that confronts only one formulation of Lewis’s clause. Take Lewis’s later clause: “almost everyone prefers any one more to conform to R, on condition that almost everyone conforms to R.”20 Non-conventional alternatives could, and often would, conform to this requirement. For example, in the meeting-place paradigm it may be that “almost everyone prefers any one more to stay home, on condition that almost everyone stays home.” But notwithstanding these preferences, the combination in which all stay at home that is go their separate ways, is not a convention, but a non-conventional alternative to a convention. Non-conventional alternatives come in a wide variety of types. Many, as demonstrated, fit Lewis’s clauses. These involved little or no conflict of interest. Others, such as chaos on the roads, may involve considerable conflict. Moreover ad hoc assumptions with respect to agents’ likes and dislikes would allow further variations to be generated. Given such a variety of preference structures among nonconventional alternatives, it seems unlikely that minor modifications to Lewis’s clauses could ever overcome the problem. Certain types of non-conventional alternatives could always be found that would fit any clause he proposed. (In the section on action-determining preferences we will see what the root cause of Lewis’s trouble in this regard is.) To summarise, the fact that a regularity may “solve” a Lewisian
97
Social Action formal coordination problem does not mean that it solves an informal coordination problem. Lewis’s definition, then, does not provide a sufficient condition for conventionhood. The next question is whether his definition provides a necessary condition. Many conventional combinations are ruled out by his earlier definition. Take the clothes example.21 Here everyone prefers everyone to wear, say, black ties, given that at least all but one do. Suppose many people dislike (or are indifferent to) someone or other. Thus A dislikes B, C dislikes D, and so on. Disliking B, A would prefer B to turn up wearing the wrong tie. A would like to see B embarrassed. Indeed, most people would like to see someone or other embarrassed. It is a large social circle; people cannot be expected to like everyone. Notice that this slight alteration is sufficient to ensure that the clothes paradigm violates Lewis’s clause. It is now not the case that everyone prefers everyone to wear a black tie, given that at least all but one wear black ties. Someone might maintain that, thus redescribed, the clothes case is no longer a convention. But this would surely be a mistake. In any given social milieu there are going to be elements of animosity and of competitiveness. Most people are going to dislike someone or other to the point where they would like to see that person embarrassed or made a fool of. It is extremely implausible that when such elements of enmity exist the regularity in question ceases to be a convention. It would have the result that many of the regularities we take to be paradigms of conventions would turn out not to be conventions after all. Surely the clothes case – and numerous other cases like it, for example modes of eating, greeting, how to behave at High Table, which word means what – is essentially a convention notwithstanding the element of conflict. Lewis’s initial formulation is thus unacceptable. It does not provide a necessary condition for conventional regularities. However, since the element of conflict is after all only an element, perhaps some reformulation of Lewis’s clause is what is called for. As demonstrated, Lewis himself provides such a reformulation, namely that “almost everyone prefers any one more to conform, on condition that almost everyone conforms.” 22 This clause is designed to cope with exceptions to the rule. Consider the case where someone fails to conform. The clause entails that with respect to the nonconformer, almost everyone prefers that the person conform. Lewis’s clause copes with this type of exception to the rule. In our 98
3. Conventions clothes example, given A does not conform, most would prefer that A did. But we can adjust our example slightly to defeat this interpretation. Suppose, as is often the case, a convention exists in a large community whereby most only know a minority in that community. Take, for example, a convention in some community to wear a suit at the office. Now assume that most prefer to wear a suit conditionally on some fairly indeterminate majority of the members of the community wearing a suit. However, it may well be that B, who has never met and knows nothing about A, may not prefer that A wear a suit. B is indifferent to whether or not a single individual (or even a minority) B knows nothing about conforms to the convention. But B is going to be in this way indifferent to most of the members of the community, since B only knows a minority of them. B’s action is dependent on the actions of the majority in the community. However, with respect to almost any individual (or minority) in that community, B does not necessarily prefer that individual (or minority) to conform, given all but that individual (or minority) conform. A regularity in which agents had such preferences would clearly be a convention. Indeed it is plausible that in any large community many conventions are in fact of this type. Here is another counter-example to the necessity of Lewis’s definition. Consider the clothes case again, but this time assume there is a particular minority, x, such that most people dislike members of x, and would like to see them embarrassed. Thus, given all but x conform, most do not prefer x to conform. Therefore the regularity in question is not a convention. Lewis sticks with his definition in the face of such cases. He takes them not to be conventions.23 But this is implausible. Assume once again that almost everyone (including members of x) conforms, on condition a majority, say threequarters, conform. (This is an arbitrary figure and, indeed, it could be replaced by a possibly fairly indeterminate range, anything, say, from a half to three-quarters, without affecting the point being made here.) Now if three-quarters conform but members of x do not, most prefer members of x not to conform. But if the number of agents conforming starts to fall below three-quarters, then most may well begin to prefer members of x to conform. That is, under certain circumstances most may want to (or may even have to, for example as in the stag hunt paradigm24) rely on the conformity of members of x, and thus disregard their dislike of x. It seems that, notwith99
Social Action standing the widespread dislike of x, members of x are parties to what is in fact a convention. (The same point could be made if it was assumed that the majority were merely indifferent to the minority, and did not actively dislike the minority.) Thus Lewis’s definitions as they stand rule out regularities in action which are in fact conventions. Why not try to make a further modification to his clauses? In particular, why not make a modification that allows there to be further exceptions to the rule? Where most conform but some individual (or minority) does not, the definitions require that all or most prefer that individual (or minority) to conform. This is the requirement that is causing the trouble. It cannot be wholly dispensed with. For it is necessarily the case that for any agent there is some individual, such that that agent would prefer that individual to conform, given that most conformed, and given that that individual is a non-conformer. In particular, it is necessarily the case that that individual prefers himself to conform, given that all but that individual (or most, but excluding that individual) conform. That is, some cases in which there are additional exceptions to the rule, that is, additional to the ones Lewis’s reformulation caters for, may very well be acceptable. But cases in which agents want to make exceptions of themselves are not acceptable. If an agent wants to make an exception of himself then, other things being equal, if the others conform, that agent will not conform. (I have included the “other things being equal” clause to rule out cases in which although agents may want to make exceptions of themselves at some level, when it comes to the crunch – because of sanctions, for example – each actually conforms. The account of such cases introduces distracting complications.) But if an agent fails to conform in every instance of the recurring situation, what grounds are there for saying that that agent is a party to the convention? Others prefer that the agent conform, but this is a ground for saying that others want the agent to be a party to the convention, not for saying that the agent actually is a party. We do not want, then, a definition of conventions that would have it that conventions consisted in part of preferences for making an exception of oneself. The question now arises as to how to reformulate Lewis’s clause in such a way so as to admit further exceptions to the rule, without allowing cases in which agents want to make
100
3. Conventions exceptions of themselves. It is not at all obvious how this could be done. But, more importantly, it is not clear what legitimate motivation could be provided by Lewis for such a reformulation – a reformulation that admitted further exceptions to the rule, but not cases in which agents made exceptions of themselves. In the next section I will suggest a distinction that provides a motivation for ruling out cases in which agents want to make exceptions of themselves, but lets in cases in which agents want to make an exception of individuals other than themselves. This is the distinction between actiondetermining and non-action-determining preferences. The preferences with respect to some minority other than ourselves are non-action-determining; the preferences for making an exception of oneself are action-determining. This motivating distinction can be used to differentiate between cases. It will turn out that, of the cases we have been looking at, only those in which agents want to make exception of agents other than themselves (as they stand at any rate) are genuine conventions. For these preferences are nonaction-determining. This motivating distinction is not, however, one that Lewis could legitimately appropriate. I have shown that Lewis’s definitions do not provide a necessary condition for conventionhood. And I have already argued that they do not provide a sufficient condition. The existence of a Lewisian formal coordination problem is thus neither necessary nor sufficient for the existence of an informal coordination problem. action-determining preferences Enough has been said to demonstrate that the preference structures involved in Lewis’s account are in need of fairly radical alteration. The preferences involved in Lewis’s various definitions (and indeed definitive of game-theoretic equilibria) include preferences that agents have with respect to the actions of other agents. (Of course, they also include preferences that agents have with respect to their own actions.) However, agents are not, in the situations under consideration, prepared to interfere with one another, for example agent A is not prepared to force agent B to perform action x, even though A may prefer B to perform x. A question arises as to whether these preferences with respect to the actions of other agents are
101
Social Action non-action determining. A preference is non-action determining if it makes no difference to anyone’s actual actions, or to what anyone’s actions would have been if others had acted differently. Obviously the actions in question are either the conventional actions themselves, or else the actions that are conventional or nonconventional alternatives to those conventional actions. Not all the actions not done in conformity to a convention, or its conventional alternative, are non-conventional alternatives to that convention. Some actions that are not performed in conformity to a convention or to its conventional alternative are not performed instead of, that is as alternatives to, that conventional action. They may, for example, be performed in addition to, or even at the same time as, the conventional action. The non-action-determining preferences may determine actions that are irrelevant to the convention, and its (non-conventional and conventional) alternatives. It seems that some preferences with respect to the actions of others are at least indirectly action determining, namely in the case where an agent prefers to act in accord with the preferences of others. But however things stand in general with the actiondetermining status of preferences for other actions, I want to suggest that many preferences for other actions may in fact be non-action determining (in the above sense) and that such non-actiondetermining preferences are the root of the troubles besetting Lewis’s definitions. Moreover, it is because these preferences are non-action determining that they do not yield a clear sense in which conventional actions are interdependent. For these and other reasons the non-action-determining preferences in question have no place in an analysis of conventions. Let us first see if we cannot locate the particular preferences cited in Lewis’s clauses that enabled the counter-examples to his definitions to be generated. Neither of the clauses provides a sufficient condition for conventionhood. The reason for this is that many nonconventional alternatives fit them. Thus on certain construals of, for example, the meeting-place paradigm, the combination of actions describable as everyone staying at home, is wrongly characterised by Lewis’s clauses as a convention. Assume that most stay home but that a few, unaware of the actions of the majority, proceed to the meeting place. This latter minority are upset at having wasted their time. There are two distinct types of preference of interest at this 102
3. Conventions point. Each memeber of the minority has a preference with respect to his or her own action. This is the first type of preference. Each of these preferences is action determining. In the normal course of events any such preference would have determined the preferring agent’s action. Had the agent known, as he or she normally would have, that the majority were not going to try to meet, the agent would not have tried either, but would also have stayed at home. Then there is a second type of preference. There are the preferences of each member of the majority, those staying at home, that the members of the minority, those proceeding to the meeting place in vain, not so proceed to that meeting place. The majority are sympathetic people and do not like to see others wasting their time. But this second type of preference is non-action determining. Preferences of this second type make no difference either to the actions of the majority (those with the preferences) or to the actions that are the objects of these preferences (the actions of the upset minority). Nor is there some other (relevant) possible combination of actions with respect to which these preferences would be action determining. Moreover, it is these non-action-determining preferences that are the root of Lewis’s troubles. Let us now assume that the agents’ majority are not particularly sympathetic to the timewasting minority. On this new assumption the combination describable as everyone (or almost everyone) staying at home ceases to be, by Lewis’s lights a convention. For it is now not the case that everyone prefers everyone to stay at home on condition that at least all but one do. It is now not the case that almost everyone prefers anyone more to conform given almost everyone conforms. It is by manipulating these preferences that agents have with respect to the actions of others that combinations of actions, which are in fact nonconventional alternatives to conventions, are characterised by Lewis’s definitions as now conventions, now non-conventions. The inclusion of these preferences with respect to the actions of others ensures that Lewis’s clauses do not provide a sufficient condition for conventionhood. The same points can be made with respect to the counterexamples to the necessity of Lewis’s definitions. The particular preferences causing the trouble are once again certain preferences agents have with respect to the actions of agents other than themselves. In one counter-example it is the preferences of the majority 103
Social Action with respect to the actions of some minority or other, in the other it is the preferences of the majority with respect to the actions of a particular minority. But in both cases if we made an ad hoc alteration to these preferences, the combination of actions would, by Lewis’ lights, become conventions. Thus they would become conventions if in both cases the majority decided that after all they liked the minority or minorities in question. In both cases the combinations are in fact conventions irrespective of whether or not these ad hoc alterations are made. Finally in both cases these preferences are non-action determining. These preferences make no difference to the actions of the majority, nor do they make any difference to the actions of the (disliked) minority. I take myself to have: (a) located the particular preferences that enable the counter-examples to Lewis’s definition to be generated; and (b) shown that they are non-action determining. I have not however shown that these preferences are in this way problematic for Lewis in virtue of being non-action determining. I will return to this question shortly. These located preferences with respect to the actions of others ought to be excluded from any definition of convention for the following reasons. First, as has been seen, they enable combinations of actions that are not conventions to be characterised as conventions, and combinations that are conventions to be characterised as non-conventions. Second, these preferences are non-action determining. I am inclined to think that this fact is in itself sufficient to warrant their exclusion from any role in the definition of conventions. How can the notion of a conventional action be taken to consist (in part) in preferences which are such that there are no circumstances in which any of those preferences determine that that action is performed, or that it is not performed? At the very least we are owed an explanation as to why such non-action-determining preferences ought to be included in the definition of conventions. One role for preferences is to assist in laying bare the structure of the interdependence of action involved in conventions. If a preference is non-action determining it clearly makes no contribution to the interdependence of action. Indeed, and this is my third objection to these non-action-determining preferences, these preferences currently under fire prevent the provision of a precise account of the structure of the interdependence of action involved in 104
3. Conventions conventions. Lewisian preference structures, as has been seen, are consistent with a range of different structures of interdependence of action. It is consistent with a Lewisian preference structure that each xs only on condition the others do, or that each xs only on condition the others do not y (where ying is one, but only one, way of not xing). The combination in which all meet at A, and conform to Lewis’s clauses in so doing, is an example of a combination involving the former structure of interdependence of action. The combination in which all stay at home, and conform to Lewis’s clauses in so doing, exemplifies the latter structure. Moreover, Lewisian preference structures are consistent with the same underlying interdependence of action. The different construals of the clothes example constitute different Lewisian preference structures. But we can assume that the underlying structure of interdependence of action is the same throughout these construals. Fourth, these preferences are not appropriately manifested. Such preferences, as we have seen, may be present in one instance of the recurring conventional situation, and absent in the next. Moreover, their presence may go completely undetected by other agents. If these preferences are genuine preferences, it might be thought that they must manifest themselves publicaly in one way or another. This may or may not be true. However, what is true and relevant is that it is not necessary that agents, insofar as they are parties to the convention, and are aware of the conventional behaviour of the other parties to that convention, must be exposed to actions which manifest the non-action-determining preferences in question. (Remember that what is meant here by a non-action-determining preference is a preference that does not determine a conventional action or an action that is an alternative to a convention.) Qua parties to the convention, who are fully aware of all the conventional behaviour going on, agents may well be blissfully unaware of these preferences that agents have with respect to the actions of others. I conclude that the non-action-determining preferences in question, since they are not in the appropriate way publicly manifested, cannot be (in part) constitutive of conventional action. That concludes the presentation of my reasons for expelling nonaction-determining preferences from any definition of conventions, and in particular for expelling certain of Lewis’s conditional preferences from his account. But this expulsion of certain preferences 105
Social Action with respect to the actions of others leaves Lewis’s theory in disarray. Appropriately adjusted, his clauses now read: “each (or most) prefers himself to conform, if (most of) the others conform.” Any preference an agent might have with respect to a non-conforming minority other than that agent has disappeared. But it is consistent with this clause that agents prefer to conform even when the others do not conform. And if that were so, the combination describable as all conforming could hardly be a convention. Each may conform irrespective of what the others do.25 Moreover, in excluding these troublesome preferences, the door has been opened to conflict between agents’ preferences. Thus each conformer may now prefer all or most others not to conform. A prefers A but not B to conform, B prefers the reverse. I would like finally to deal with a residual matter. We saw that certain Lewisian preferences with respect to the actions of others – preferences which enabled counter-examples to be generated – were non-action determining. But it was pointed out that the fact that these preferences were non-action-determining did not show that these preferences enabled counter-examples to be generated in virtue of being non-action determining preferences. Now I want to try to show that these preferences were in fact troublesome by virtue of being non-action determining. We intuitively judged that a combination of actions is or is not a convention, on the basis of the presence or absence of interdependence of action. Thus in the meeting-place paradigm the combination of actions describable as each going his own way independently of what the others did was judged a non-conventional alternative to a convention – notwithstanding Lewis’s definitions. On the other hand, the various clothes paradigms were judged to be conventions because agents’ actions were interdependent – again notwithstanding Lewis’s clauses. But interdependence of action is determined by action-determining preferences only. Thus, whether or not a combination is a convention is going to turn on the structure of actiondetermining preferences; non-action-determining preferences will be irrelevant. Lewis’s definitions failed to determine whether or not a combination of action was or was not a convention, and failed by virtue of consisting in part of certain preferences with respect to actions of others. But since these latter preferences are non-action determin106
3. Conventions ing, his failure looks as if it were inevitable. It is only actiondetermining preferences that determine conventionhood. The preferences in Lewisian structures that enabled counter-examples to his definitions to be generated were non-action determining. We see now that it was by virtue of being non-action determining that these preferences enabled those counter-examples to be generated. collective ends and conventions One way of eliminating non-action-determining preferences from conditional preference structures is simply to strip from these structures all preferences that agents have with respect to the actions of other agents. We are then left with: “each prefers himself to conform, given all but himself (most excluding himself) conform.” That is, we are left with a formula that specifies only (conditional) preferences an agent has with respect to that agent’s own actions. This formula is to be understood as providing a necessary, but not a sufficient, condition for the conformity of any agent to a convention. To strengthen this clause so that it provides both a necessary and a sufficient condition of the preferences involved in a convention we need to insert “only if.” We thus have: “each (most) prefers to conform, if, and only if, (most of) the others conform.” The following case demonstrates the necessity for this strengthening of our clause with the inclusion of “only if.” Suppose each man preferred to wear an expensive black tie, but preferred others not to, so that he could lord it over them. However, he would also want to wear an expensive black tie if they did, since otherwise they would lord it over him. Each thus wears an expensive black tie whatever the others do. This does not look like a convention since there is no interdependence of action. It is a situation in which each prefers to conform, given the others do. It is just that it is also a situation in which each prefers to conform, given the others do not conform. The strong formulation, “each prefers to conform, if and only if the others conform,” rules out cases such as the expensive black tie case, and for this reason should be preferred to the weak formulation. (Lewis’s way of ruling out such cases as the expensive black tie case involves recourse to non-action-determining preferences, given most wear expensive black ties, most prefer the nonwearers to wear them – a way we have seen to be problematic.) 107
Social Action It seems that conventions vary. Some, like Lewis’s telephone case, are such that each prefers to conform if and only if all the others conform. Some, like the meeting-place paradigm (on certain construals), are such that each prefers to conform if and only if most of the others conform. Still others, for example the clothes paradigm (on certain construals), are such that each prefers to conform if and only if some fairly indeterminate majority conform. Some sort of disjunctive definition would thus be called for. In what follows I will, for the sake of simplicity, make use only of the formula “each prefers to x, if and only if the others x.” Our new conditional preference formula yields straightforwardly a principle expressing the interdependence of action involved in conventions, or at least does so on the plausible assumption that agents act in accord with their preferences. That principle is, “each xs if, and only if, the others x.” (To this extent our conditional preferences clause is to be preferred to Lewis’s, which, as we saw, did not yield a precise sense in which parties to a convention performed interdependent actions.) Moreover since the preferences involved in our new principle are preferences with respect to an agent’s own actions, and preferences that, in the various possible circumstances definitive of convention and their conventional and non-conventional alternatives, determine his actions, they are action determining and publicly manifested in the appropriate way. (Remember that not all of the nonconventional actions an agent might perform are alternatives to that convention.) Our conditional preferences meet the requirement that preferences that are (in part) constitutive of conventional behaviour must be manifest in that behaviour. I would now like to turn back to the problems detected with respect to the necessity and the sufficiency of Lewis’s definitions. Lewis’s definitions were found not to provide a sufficient condition when it was discovered that they could not distinguish between conventional and non-conventional alternatives to conventions. They could not so distinguish because they included certain preferences with respect to the actions of others. Our new principle states that each prefers to x if, and only if, the others x. Notice that there must be an alternative state of affairs in which each prefers to y, in the case where the others are not xing. (Here, to y is one way, but not the only way, of not xing.) It may be 108
3. Conventions true, though it is not necessary, that in fact each ys if and only if, the others y. If this were the case then ying would be, according to our theory, a convention. If, in such a case, ying interacted appropriately with xing, ying and xing would be conventional alternatives. But what of the situation where it is not the case that each ys if and only if, the others y, but rather that each ys only on condition the others do not x? I suggest that the latter principle describes a situation which is a non-conventional alternative to a convention. Thus, to take an example, suppose each prefers to go to a (prefers to x) if, and only if the others do. Then it would be true that each prefers to stay at home (prefers to y) only on condition the others did not go to a (did not x). But each does not, in our example, prefer to stay at home if, and only if, the others stay at home. If one, a minority, or even a majority of others, decided to visit their mums rather than stay home, this would make no difference to anyone else’s actions or preferences. If we accept that non-conventional alternatives are defined (in part) in terms of the principles, each prefers to y only on condition the others do not x, and each ys only on condition the others do not x, while conventions are defined (in part) in terms of the principles, each prefers to x if and only if the others x, and each xs if and only if the others x, then obviously we have distinguished between conventions and non-conventional alternatives to conventions. Moreover, we can see where Lewis went wrong. His clauses specified a number of idle non-action-determining preferences, which are not necessary for a situation to be a non-conventional alternative, but which may in fact be present in the case of such alternatives, thereby leading those alternatives to be wrongly characterised as conventions. The idle preferences in question were, of course, certain of these with respect to the actions of other agents. Again, if we look at the problems Lewis had with the counterexamples to the necessity of his definitions, we see that they disappear under our new definitions. In the clothes paradigm (and numerous other cases of its type) each prefers to conform if, and only if, a majority (probably some fairly indeterminate majority) of the others conform – according to our new principle. Indeed this appears to be how things are. The fact that there are various minorities such that people may not wish them to conform does not, according to our definition, rule these cases out. Once again our definitions 109
Social Action succeed where Lewis’s failed, and just because Lewis’s, but not our own, included non-action-determining preferences. It seems we must accept our new principles as characterising the preference structure, and the interdependence of action, involved in conventions. It is important to see that these principles of themselves do not entail that conventions be the solutions to informal coordination problems. All that is required for there to exist a situation in which each prefers to x if and only if the others do is for there to be some alternative situation such that each prefers to y only on condition that the others do not x. But the alternative is, as such, only a non-conventional alternative. It is not a conventional alternative. There may in fact be a situation which is a conventional alternative, but this is not entailed by our principles. It is not something that arises out of necessity from the structure of interdependence of action involved in conventions. Thus that structure is not such as to entail that conventions solve informal coordination problems. In fact I will argue in the section on conventions and alternatives in this chapter that conventions are quite often not the solutions to informal coordination problems. (Of course, many conventions are the solutions to informal coordination problems.) I now want to argue that the notion of a collective end, needs to be built into our definition of convention (and indeed would have to have been built into Lewis’s definition, even had his definition been acceptable in other respects). Why introduce collective ends? The conditional preferences in our definitions are to be understood in terms of a means–end model. Agent A has a conditional preference for some action x, because it realises some end of agent A’s. (A does not prefer x in the sense of liking x just for its own sake.) Now, in the case of conventions, as we have seen, A prefers to perform x if, and only if, B performs some action, y. A, for example, prefers to drive on the left, if and only if, B drives on the left. The reason A prefers to perform x, and do so if and only if, B performs y, is that the action, x, in conjunction with the action, y, realises an end A has. Thus in the driving case, A’s end is the avoidance of automobile collisions. This is also B’s end. A and B perform actions (x and y), which jointly realise a collective end. This is not to say that the ultimate end of each is a collective end. A may ultimately only be interested in A’s own survival. But the 110
3. Conventions point is that in order to realise A’s ultimate and, as we might say, individual end, A must first realise a collective end. Collective ends, then, explicate the conditional preferences, and the interdependence of action, involved in conventions. Moreover, there are additional reasons for introducing collective ends. First, the principle, each xs if and only if the others x, is as it stands false. It may well be that an agent would, under certain circumstances, x, but x for entirely his own purposes, and thus independently of what other agents were doing. If this were so then it would be false that each xs if and only if the others x. What is required to overcome this difficulty is a relativisation of the principle to some collective end – relative to an agent’s concern to realise a collective end, that agent xs only if the others x. For, relative to that collective end, there would be no point in his xing, if the others did not also x. Second, collective ends need to be built into the actual definition of conventions in order to provide a context for conventions. Suppose there is an act of type x, such that each xs if, and only if, the others x. Moreover assume that there is some other action type, y, such that each would y if, and only if, the others yed. An agent cannot x and y at the same time. The regularity in which all x is a convention, as is the regularity in which all y. Are xing and ying conventional alternatives? Not necessarily. I may drive on the left if, and only if, others do. And I may row in rhythm, r, if, and only if, the others do. Moreover, I cannot drive a car and row a boat at the same time. Thus the additional requirement in Lewis’s definitions – the requirement that an agent cannot conform to a convention and its conventional alternative at the same time – is met. But nevertheless driving on the left and rowing in rhythm r are not conventional alternatives. Collective ends determine for us what, if anything, counts as a conventional alternative. Driving on the left and rowing in rhythm r do not realise the same collective end. Moreover, we need collective ends to tell us what constitutes a non-conventional alternative to a given convention. A nonconventional alternative is the combination of actions such that each ys only on condition the others do not perform the action x, where everyone xing realises a given collective end. Nor is it simply our account of the conditional preferences, and of the interdependence of action in conventions, that requires the 111
Social Action introduction of collective ends. Take Lewis’s clause to the effect that everyone prefers everyone to x, given that at least all but one x. His other clause requiring an alternative is (in effect) “everyone would prefer everyone to y, given that at least all but one yed.” But how does he determine whether or not ying is a conventional alternative to xing? Lewis includes in his definition mention of a recurring situation, s. Presumably ying is only a conventional alternative to xing if agents would y in situation s. But the notion of a situation s is as it stands inadequate. What constitutes the situation? The fact that all the agents live in 2000 in a liberal democratic state is not the situation – so what is? The notion of a situation employed here needs to be enriched so as to enable it to distinguish between regularities that are alternatives (in our sense) and those that are not. My suggestion here is that the notion of a collective end needs to be introduced in the way I have indicated. Finally, having introduced the notion of a collective end we dispense with the notion of conditional preferences; we no longer need the principle each prefers to x if and only if the others x. We can simply function with the notion of each agent having the collective end, believing that the other agents will x, and therefore also (intentionally) xing as a means to that end. This is not to say that the agents do not have preferences with respect both to various collective ends, and with respect to the actions performed to realise those ends. Agents do, of course, often prefer some means to another means, and desire one end more than another. Indeed, when we come to discuss the rationality of convention following we will need to say something about these preferences. But it is to say that they are not required for the constitutive account of joint action, in general, and of conventions, in particular. conventions and alternatives Are conventions always the solutions to informal coordination problems? It was shown earlier that the conditional preferences, and the interdependence of action in conventions, were not such as to entail that conventions be the solutions to informal coordination problems. Indeed, we saw in the section on Lewis on conventions that even on Lewis’s theory, conventions need not be the solutions to
112
3. Conventions informal coordination problems; a conventional regularity could have a non-conventional alternative that squared with Lewis’s clause requiring an alternative regularity. (Insofar as Lewis’s theory allowed this, it failed to achieve one of its most important aims.) The model of conventions developed above would allow both regularities that solve informal coordination problems and regularities that solve other sorts of problems to count as conventions. However a clause could be added that required that there be at least one conventional alternative to a convention. Such a clause would perform the function Lewis intended his clause specifying an alternative regularity, r1, to perform. When we reject the addition to our definition we are rejecting Lewis’s fundamental intuition that conventions are the solutions to informal coordination problems. It does not seem that there are good grounds for such an addition, as it does not seem that all regularities are solutions to informal coordination problems. First, the conditional preferences and the interdependence of action are in fact the same for both regularities that solve informal coordination problems and certain other regularities that do not solve this type of problem. This is presumably a motivation for viewing both types of regularity as belonging to the same basic category. (Examples of such regularities, which do not solve informal coordination problems, follow.) Second it seems we call many regularities without conventional alternatives conventions. Indeed, even Lewis calls some of these regularities conventions. Take Lewis’s stag-hunting paradigm. The combination of actions describable as everyone hunting rabbits is not a convention. It is not a convention either on Lewis’s theory or our own. It may well be that everyone does not prefer everyone to hunt rabbits, given that at least all but one hunt rabbits. It may well be that almost everyone does not prefer any one more to hunt rabbits given almost everyone hunts rabbits. If some individual or minority is not hunting rabbits then most may be indifferent to this. Indeed, it is possible agents do not care if a majority are not hunting rabbits just so long as they are not hunting stag. The combination in which all hunt rabbits is not a convention on Lewis’s theory. On our own theory the combination in which all hunt rabbits is a non-conventional alternative to a con-
113
Social Action vention. It is the combination such that each ys (hunts rabbits) only on condition the others do not x (do not hunt the stag). I believe this reading squares with intuition. Paradigms of the stag-hunting variety are agreed by all to be conventions, but they do not involve any conventional alternative; they do not solve informal coordination problems. Nor is the staghunting example an isolated case. We can generate like cases by making adjustments to some of Lewis’s other examples. Take Lewis’s oligopoly case. Each seller can either try to compete with the others by undercutting their prices – a situation in which sellers lose out – or each can agree with the others to a common (high) price, that is they can enter into an oligopoly. In agreeing to set a common price each agrees to set his individual price at the uniquely best price for an oligopoly in that market – something determined by an economist. It is not a matter of convention whether or not to sell goods at price x rather than y. That is a matter of calculation and is determined by the economist. What is a matter of convention is that each sells the goods at that price which maximises the revenue of the oligopoly. But there is not an alternative to this procedure. The alternative is a non-conventional price-cutting war. Take Lewis’s land case, but change the commodity to money. We are a gang of thieves. Either we can fight over the spoils after each robbery, or we can divide it equally. (No one would accept less than an equal share.) But having divided it equally there is no further question for the robbers about who gets which share. The spoils may have been put in a common bank account for safekeeping, and each robber simply writes out a withdrawal slip when the time for division of spoils comes about. Moreover there are other conventions, not generated from Lewis’s paradigms which do not have alternative conventions. Suppose in a department of philosophy the staff members must each on pain of defying the regulations, mark their own students papers. However, members of staff decide in the interests of greater accuracy, and thus fairness, to double-mark. Each double-marks only on condition the others do – it is a convention. But there is no conventional alternative. Single-marking is not a convention, and triplemarking is out of the question – it requires a greater expenditure of time and energy than is either available or acceptable. It seems that there just are a range of regularities in action, which 114
3. Conventions we call conventions, but which do not have conventional alternatives. Moreover, the double-marking example shows that there are cases in which there is a possible, that is logically possible, conventional alternative, such as triple-marking, but this alternative is unacceptable to the particular agents in question. It seems that there can be strong reasons for particular conventional alternatives to a point where some, like triple-marking, are ruled out. Thus, in the meeting-place paradigm it might be that places a and b are conventional alternatives, though place a is much preferred. On the other hand, while place c is a logical possibility, it may be completely unacceptable for other reasons. I suggest then, that conventions can be thought of as existing on a continuum. At one end each convention has at least one clear conventional alternative, and at the other there are conventions with no logically possible alternative convention. In between there are conventions for which the question as to whether they have any conventional alternative becomes fairly indeterminate. I conclude that, contra Lewis, conventions may be, but are not necessarily, the solutions to informal coordination problems. joint procedures The upshot of our discussion thus far is that conventions are regularities in action directed to a collective end. So conventions are a species of joint action. However I now want to claim that conventions are in fact a particular species of joint action, namely, joint procedures. For the purpose of what follows I introduce the distinction between regularities (in action) and what I have termed (in Chapter 1) procedures.26 In the case of a regularity in action an agent provides himself or herself (or could do so) with a reason for performing the action the agent regularly performs, and does so (or could do so) on each and every occasion of the agent’s performance of that action. Moreover the reason in question justifies that action considered on its own. The existence of such a reason (or perhaps reasons) means that there is (or could in principle be) a new decision to perform the action on every occasion on which it is performed. There is, if you like, one decision for one action. In the case of procedures, by contrast, there is a single decision 115
Social Action to perform many actions. The agent decides at time, t1, that he will perform an action, (of type) x, at time, t2, and also perform x at times, t3, t4, and so on. One decision then – the decision to adopt the procedure – governs a whole set of future actions. One process of reasoning will take place, namely that justifying the adoption of the procedure. Thereafter the agent will not provide himself with a reason for acting, he will merely identify the situation as being of the sort for which he has adopted the procedure in question, and then simply follow the procedure, perform the action. The basic characterisation of procedures, then, is in terms of a set of actions governed by a single decision, as opposed to a set of actions each member of which is governed by a different decision. It is important to note the property of conditionality which does to some extent qualify the sense in which a procedure is a single decision to perform a set of actions, but which does not negate it. For in the first place our definition of procedure is such that it is not the case that in every instance of a situation covered by a procedure a new decision to perform the requisite action type could be made after a process of adequate reasoning. This means that the matter of whether the conditions under which the original decision was made to adopt the procedure have been met is not something that can arise on every occasion, but rather only from time to time. In the second place even in those instances in which the matter of whether the conditions have been met does arise there is no wholly new decision to perform the action type in question. For there is a presumption in favour of the procedurally determined action, a presumption which can only be offset by the non-obtaining of the condition or conditions. Of course, an agent could decide on a wholesale review of a given procedure, and refuse to be bound by this presumption or to be restricted to these conditions. In that case the agent would indeed be embarking on a process of reasoning that would lead to a wholly new decision, and the agent’s subsequent action or actions would not in any sense be bound by the original decision to adopt the procedure in question. Insofar, then, as an agent in periodic reviews of a given procedure accepts the presumption in favour of the procedural action and merely tests the relevant condition or conditions, the agent’s original decision to perform a set of actions does not collapse into a series of decisions each of which governs a subset of the original set. 116
3. Conventions How could it be rational to adopt procedures rather than merely conform to regularities in action? The point of procedures is to free oneself from the need to made a new decision after a process of reasoning on every occasion of a recurring situation. Such decision making with its attendant reasoning is hugely time-consuming. What are the grounds for thinking that conventions are not simply regularities directed to the realisation of collective ends, but also procedures? These are twofold. First, many cases of actions directed to the realisation of a collective end are not conventions, even if they recur. Two men may regularly move furniture without it being a convention. Second, conventions are like procedures in that they are necessary. This is so for the following reason. An individual person’s interaction with the world is extraordinarily complex. (Here and in what follows I am restricting attention to that interaction which is in principle under the agent’s control.) Indeed it is so complex that an agent could not possibly on every occasion be consciously attending to and deciding with respect to every aspect of that interaction. Hence the need for procedures. The adoption of procedures effectively puts the agent’s actions on automatic; there is now no need on every occasion to attend to and decide upon every aspect of the interaction. Should the agent decide to attend to any aspect it is possible to do so; but the point is it is not possible to do so for every aspect. Now I suggest, further, that the complexity of the individual’s social interaction is likewise hugely complex, and thus calls for the adoption of procedures by individuals. However, in the case of an enormous amount of social interaction, individuals have collective ends, and therefore they have to adopt procedures conditionally on others adopting those procedures. Consider, for example, the number of linguistic conventions followed even in a brief and uncomplicated conversation. Members of large communities could not possibly communicate with one another unless each individual had a set of procedures and, in fact, the same set as everyone else; so each adopts a given set of procedures and does so on condition the others have adopted that set. Quite clearly, then, there is a need for procedures in relation to social interaction, and in particular, there is a need for structures of conventions. Accordingly, I offer the following theory of convention – the theory of conventions as joint procedures. 117
Social Action There is a convention to x in the recurring situation s among agents A, B, C, and so on, if and only if: (1) each has the procedure to x in s; (2) each has the procedure to x in s if and only if the others have the procedure to x in s; (3) each has the collective end e, and each xing in every instance of s realises e; (4) each mutually truly believes that everyone has the procedure to x in s.
The following points need to be made in respect of this account. The collective end e could be a recurring collective end; an end that is realised in each of the instances of the recurring situation. Thus, in the meeting place example, you and I have a collective end to meet today. But the collective end might also be a one-off collective end; an end that is realised only if the joint procedure is followed by everyone (or most) in all (or most) instances of the recurring situation. This is the case with motorists who not only have a recurring collective end not to collide with whoever is driving in the opposite direction on this stretch of road today, but also a collective end that there be no collisions on the roads at any time between any of the motorists (or at least that the number of such collisions be minimised). And there are various possible means–end and other relationships between these different collective ends. For example, it might be that satisfaction of a set of recurring collective ends is a means to, and indeed constitutive of, satisfaction of a one-off collective end. For many conventions there are going to be exceptions. Some parties to the convention will not have the requisite mutual beliefs, some will not have a procedure, and so on. We need to build in these exceptions to our definition. Thus it is only necessary that slightly less than everyone has a procedure, a collective end, and the requisite mutual beliefs. Moreover, each having a procedure will be in general conditional on most – and not all – others having that procedure. The number here will vary from case to case. Some conventions will have a threshold below which the collective end will not be able to be achieved, and thus each will have the procedure conditional on any number of others having it so long as that
118
3. Conventions number is at or above the threshold. For other conventions this will not be the case. In respect of clause (4), agents may mutually believe that each agent has a procedure, but mutually believe this under an incomplete, vague, or partially false description of what it is to be a procedure. They may, for example, mutually believe (truly) that agents have a habit of conforming. Similarly, if agents mutually believe that each agent has a collective end – something not required by my definition they may do so under an incomplete, vague, or partially false description of collective ends. This is not problematic insofar as their mutual beliefs concerning procedures and collective ends are under descriptions that are sufficient to mark off procedures and collective ends from other phenomena for most of those agents for most of the time. This is a theory of conventions considered individually. But to understand conventions fully an additional account is required of the way individual conventions relate to one another; an account of the nature and point of structures of conventions is called for. Here I will content myself with saying that conventions do come in structures, and that in fact they are layered vertically as well as horizontally (so to speak). An example of intermeshing on the horizontal dimension would be the convention to stop on a red light and the meshing convention to go on a green light. The combination of these two conventions secures the single collective end of avoiding collisions at intersections. As far as the vertical dimension is concerned, here too there is intermeshing. Thus, as I have argued elsewhere,27 a need for a conventional arrangement to deal with conflict in society gives rise to the convention of government. But this in turn gives rise to a host of additional problems in respect of the type of government, its membership and so forth, many of which are solved by conventions. In fact, central institutions such as language and government will typically involve structures of conventions in which there is both horizontal and vertical intermeshing. Thus in respect of language there are the meshing horizontal conventions in respect of, for example, the marks, “blue,” “yellow,” “green,” and so on, to stand for, respectively, the colours, blue, yellow, green, and so on. Then there are various vertical conventions of grammar with respect to such words, and which govern word
119
Social Action order, for example that adjectives such as “blue” are placed immediately before the nouns they qualify. Sometimes the word “convention” is used to refer to one-off conventions, for example a one-off agreement to use green circles to indicate the route to follow in a car rally. Gilbert takes this to constitute an argument against theories such as Lewis’s and mine that take conventions to be regularities in action.28 However, in the first place such one-off conventions are non-standard in that most conventions govern actions in recurring situations. Second, many, if not all, such one-off cases are instances of agreement. But agreements cannot be conventions except in some derivative sense. For agreements presuppose language and languages are conventional; therefore agreements presuppose conventions. One-off cases are like conventions in some ways, for example in that they involve interdependence of action and collective ends, but they are not standard conventions, and their analysis will be importantly different. Conventions must be such that they are publicly manifest, but what does this requirement amount to? Certainly an individual’s adoption of a convention is dependent on exposure to other individuals – conformity to a convention is not, for example, something wired into an individual. For conformity to a convention, in its role as conventional action, is not action an individual decides to undertake independently of knowing whether other agents have undertaken it. Moreover, whether or not agents are in fact conforming to a convention must be able to be determined by the other parties to that convention – since conformity by one depends on conformity by the others. These considerations suggest that actions governed by conventions must be externally or behaviourally manifest. However, a stronger requirement than this is not obviously called for, and the theory before us is consistent with this requirement; it simply needs to be stated as a constraint on actions governed by conventions. There may well be a convention to x even though most would prefer not to x. Take a convention to wear suits in an expatriates’ club in Malawi where the weather is very hot and the wearing of suits rather uncomfortable. Suppose each prefers not to wear a suit. How is it possible that most are wearing suits if they prefer not to? What people prefer to do and what they actually do differ on occasion due to inertia or weakness of the will. Or maybe their prefer120
3. Conventions ence not to wear suits is weaker than their preference that everyone in the club dress in the same way, and thus each prefers to wear a suit given the others do, notwithstanding the discomfort.29 At any rate, on my theory such a practice is a convention since the members of the club have the procedure of wearing suits, do so only on condition the others do, and have the de facto collective end that each wear a suit. There is also mutual true belief. But it is not, in this example, a belief in the value of the wearing of suits in the club. Rather, it is the mutual true belief that each has the procedure of wearing a suit. A related type of case is one in which most agents publicly proclaim a particular policy, while privately rejecting it. It is the conventional view but no one believes in it. Such cases have been taken by Gilbert to be problematic for individualist theories, requiring as they do that the conventional “action” be a matter of mutual knowledge or true belief.30 But here we need to be clear as to what the conventional “action” is. There may be cases where people have conventional views, which is to say that their actual beliefs are held by convention. But this is not such a case. There is no convention to believe in the policy in question. For a convention to x requires that each x; and no one in fact believes in the policy. Rather, in this case, there is a convention to express a particular view. (Such a convention is maintained by the fact that each believes (falsely) that the others believe in the view everyone expresses. Perhaps each even (falsely) believes that the others believe in the view everyone expresses, as a matter of convention. That is, perhaps everyone falsely believes that there is a convention to believe this view.) But this squares with my theory. Each has a procedure to express a particular belief; there is mutual true belief that each has this procedure, and so on. Individuals may be parties to a convention that in fact has a conventional alternative, and yet they may be unaware that there is any conventional alternative.31 This kind of case is quite unproblematic for my theory. This theory does not insist that parties to a convention be aware of some alternative convention; it does not even require that there be a conventional alternative. However, this possible lack of awareness of conventional alternatives has to be sharply distinguished from the possible lack of awareness of a nonconventional alternative. I suggested above that conventions always 121
Social Action have non-conventional alternatives, and I suggest here that parties to a convention are always aware of this non-conventional alternative, at least in the sense that each is aware that there is the possiblity of everyone failing to conform to the convention. The argument for this is simply that if someone is aware that he is performing an action then he is aware that there is the possibility of his not performing it; and if each is aware that everyone is performing a particular action – as is the case with conventional actions – then each is aware of the possibility of everyone not performing that action. A final issue concerns the normative character of conventions. Conventions are normative in some sense; parties to a convention at least feel that they and others ought to conform, and they upbraid agents who flout the convention. I claim that if parties to the convention have the collective end realised by the convention then on pain of irrationality they ought to conform; so conventions are normative in the minimal sense implicit in instrumental rationality. However, I reject the claim of Gilbert and others that conventions are necessarily normative in some stronger sense.32 However, the question of whether or not conventions are norms, and if so in what sense, is an issue that I will take up in Chapter 4 on social norms. Social norms are an important category of social actions in their own right. So let me now turn to their philosophical analysis.
122
4
Social Norms
In this chapter I am concerned with a second main category of social actions, namely social norms of action. In order to focus my discussion I will stipulate at the outset that social norms as I understand them have the following characteristic features.1 First, they are regularities in action, or inaction. Second, social norms are not qua social norms, explicit rules or laws. They are not, qua social norms, explicitly formulated; nor do they, qua social norms, emanate from any formal authority or have any formal sanctions attached to them. Social norms are not necessarily laws. As it happens, many social norms are in addition laws, and I will consider the relationship between laws and social norms in the penultimate section of this chapter, entitled Social Norms and Laws. Third, social norms have a normative dimension in the sense that participating agents believe or feel that they ought to or, more often, ought not to perform the actions prescribed, or proscribed, by the norm. This normative dimension is not merely that implied by instrumental rationality; it is not simply that on pain of irrationality one ought to do what realises one’s ends. Nor is this normative dimension, at least prima facie, prudential in character – agents do not conform because they believe it is in their self-interest – nor does it, prima facie, simply consist of feelings of sympathy for others.2 Fourth, the normative dimension of social norms provides agents who accept those norms with reasons for action, and I will consider this relationship in the final section of this chapter, entitled Social Norms and Practical Reason. Fifth, conformity by one member of a social group is in some sense dependent on conformity by the other members of the group. That is, if an action is performed in accordance with a social norm the agent who performed it did so, at least in part, in virtue of the attitudes or actions of the other agents who conform to that norm. Perhaps social norms include norms regulating manners, such as 123
Social Action not belching in public or not asking personal questions of strangers. However, in this chapter I am principally concerned only with what I take to be the fundamental kind of social norms, namely norms with (felt) moral or ethical content.3 Such norms include norms prohibiting murder, assault, theft, rape, and child molestation and rape – such norms are typically enshrined in the criminal law – norms governing business practices, and norms of loyalty to one’s group or nation. They also include norms of honour and norms of revenge. As H. L. A. Hart points out, norms with felt moral content are important for social life; they may in fact conflict with individual self-interest; and they are sustained in part by social pressure.4 However these norms are not in all cases beneficial to a society. For example, norms of revenge, although they involve intense social pressure to conform, are by and large socially undesirable.5 Some of these norms, such as refraining from violence, telling the truth, keeping promises, being faithful to one’s spouse, and avoiding incest, govern the actions of all or most members of a community; others govern behaviour in specific institutional and professional contexts. The latter norms include principles governing the constitutive activities of the profession or institution, such as the principle that doctors treat the most needy patients first, or that police officers do not shoot unarmed fleeing felons. They also include more general institutional norms, including norms of loyalty. One example here is the norm according to which doctors close ranks when one of their own is accused of malpractice; another is the norm according to which police officers do not inform on their corrupt colleagues. This chapter is in five sections. In the first section I will argue for a distinction between social norms and conventions. Here I will rely on the account of conventions put forward in Chapter 3. In the second section I will argue against a popular conception of social norms according to which they are regularities in action sustained by agents’ desires to be approved of by others. In the third section I will offer a different account in terms of beliefs held interdependently. In the fourth section I will consider the relationship between social norms and laws. In the fifth and final section I will discuss the relationship between social norms and practical reasoning.
124
4. Social Norms social norms and conventions Agents perform individual actions to realise their individual ends, and they also perform joint actions or cooperative actions in order to realise collective ends (see Chapter 2). Sometimes joint actions – and the realisation of collective ends – ultimately contribute to individual ends, but I doubt that this is always the case. The more plausible view is that human activity involves both individual and collective ends, and that neither kind of end is necessarily a means to the other kind of end. In Chapter 3 I argued that conventions are in essence regularities in joint action. So conventions are regularities in action that realise collective ends. For example, conventions of language enable the collective end of communication, conventions governing dress enable mutual identification, and conventions governing manners facilitate social interaction. As indicated in Chapter 1, I also hold that the analysis of actions should be sharply distinguished from the analysis of the motivation of actions. Actions should be defined in terms of conative and cognitive notions such as intentions, individual ends, and beliefs. Affective notions such as desire and emotion are only needed for the analysis of the motivation of actions. Since joint actions, including conventions, are after all actions, or regularities in action, it is plausible that they too can be analysed without recourse to motivational notions. Indeed, my analysis of joint actions, including conventions, operates wholly in terms of conative and cognitive notions, such as intentions, beliefs, and especially collective ends. Let us now consider the normative dimension of actions, including joint actions and conventions. It should be kept in mind that here, as elsewhere in this chapter, I am concerned with subjective normative force. A norm of social action has subjective normative force if agents believe or feel that the action prescribed (or proscribed) by the norm ought to be performed (or ought to be avoided). Needless to say social norms can have objective normative force, as well as subjective normative force. Indeed, if objective normative force attaches to a regularity in an obvious way, we might be entitled to assume that that objective normative force is subjectively felt by those who conform to the regularity in question. 125
Social Action Sometimes we say that an action ought not to be performed because it would be instrumentally irrational to do so; other things being equal, it may be irrational not to perform an action that realises an end that one possesses. I take it that the felt requirement on the part of agents that they be rational in this sense – instrumentally rational – is not the kind of normative force that is at issue when it is claimed that conventions, such as the convention to drive on the left, or the convention to double-mark exam scripts, have normative force. Certainly it is not the kind of normative force that is distinctive of the fundamental category of social norms at issue in this chapter, namely moral normative force. I also take it that the notion of an individual action can be analysed without recourse to the notion of a social norm. So much for individual actions. However, I further suggest that the notion of a joint action, including that of a convention, can be analysed without recourse to the notion of a social norm, or indeed, any notion of moral normative force. Naturally, if an agent fails to do his or her part in a joint action or convention, he or she might legitimately be accused of irrationality, namely instrumental irrationality. But such an agent is not necessarily blameworthy in any further sense. If this is so, then the tendency to think that cooperative actions have moral normative force simply by virtue of being cooperative is mistaken. In my view the normative force of joint actions and conventions derives from whatever intrinsic and consequential moral or ethical value particular actions, interactions, and ends have; it does not arise simply from the fact that they are joint actions, or even recurring joint actions. So agents who perform actions, including joint actions, and who conform to conventions are not necessarily moral agents; they might be merely instrumentally rational agents. On the other hand, agents who conform to social norms are necessarily moral agents (as well as being rational agents). Since by definition social norms have moral normative force, my suggestion that conventions do not necessarily have moral normative force enables me, indeed compels me, to maintain a distinction between conventions and social norms. But am I in fact entitled to hold that conventions do not necessarily have moral normative force? I will consider an argument for the proposition that conventions necessarily have moral normative force. I will then outline a number 126
4. Social Norms of ways in which conventional regularities can come to have objective moral normative force, but stop short of being social norms. I will then suggest that some conventional regularities that have objective moral normative force can easily and naturally come to be social norms. So it turns out that some, perhaps many, conventions are after all also social norms. However, the point is that being a social norm is a contingent feature of some conventions, not a necessary feature of all conventions. Note that my position is related to, but different from, David Lewis’s on this point.6 He also appears to hold that conventions are norms, though contingently so. However, in his view the normative force of conventions basically derives from the irrationality of failing to act in accordance with one’s own conditional preferences for conformity. A party to a convention ought to conform on pain of irrationality. However, this kind of argument, if it works, only demonstrates that conventions are norms with the force of instrumental rationality; not that they have the kind of moral normative force which I take to be characteristic of social norms. Moreover, in Lewis’s view the normative force of conventions can be derived from defining features of conventions. In my view, the normative force of conventions attaches to conventions in virtue of extraneous facts, such as the benefits that may contingently flow to parties to conventions. More of this later. Let me first consider the two arguments for the proposition that conventions necessarily have moral normative force. As we saw in Chapter 2, according to Margaret Gilbert joint actions, and therefore (on my account of conventions) conventions, necessarily have normative force, albeit what she terms institutional normative force. We saw reason to reject Gilbert’s argument. However, there is another argument we need to consider. This argument for the moral character of conventions is one advanced by Thomas Scanlon in his discussion of promises.7 Scanlon himself does not use this argument in relation to conventions. In his discussion of promises, Scanlon argues that moral obligations can be generated when agents intentionally cause other agents to have beliefs about the actions of those agents, and when those other agents act on those beliefs and are harmed in so doing. Scanlon invokes the principle that agents are under a moral obligation not to intentionally cause false beliefs. This is apparently a 127
Social Action species of the obligation not to deceive. However, on my account of conventions, the failure of an agent to conform to a convention he has conformed to in the past would not in general involve any such deception of the other parties to the convention, since the nonconforming agent does not intentionally cause the other agents to have a false belief that the first agent is or will conform. Rather the other agents come to so falsely believe on the basis of the evidence of the past conformity of the first agent to the convention. And there is no obvious general normative principle to the effect that if you intend to discontinue performing a certain kind of action or activity, then you should avoid behaving in such a way that others might come to take your behaviour as evidence that you will continue to perform the kind of action or activity in question. In the course of his discussion Scanlon also invokes the principle of due care – agents are under a general obligation not to harm others. Specifically, agents ought to avoid causing other agents to have false beliefs, if having such beliefs causes harm. However, on my account of conventions there is no question of agents being harmed, even in the minimal sense of not having their desires satisfied. Rather, the only “harm” brought about is that of other agents failing to have some of their collective ends realised. But not realising an end does not of itself constitute harm. Whether or not an agent’s failure to realise his or her end is harmful to that agent depends on the nature and consequences of the particular unrealised end in question. Naturally, an agent or group of agents whose ends were never or seldom realised might over time suffer harm qua agent. But that is a different matter. I conclude that Scanlon’s argument cannot be used successfully to demonstrate that conventions in themselves have moral or even quasi-moral (including institutional) normative force. Notwithstanding this discussion some conventions do in fact have moral normative force. How can this be? I suggest that there are a variety of reasons why parties to specific conventions may incur moral obligations to other parties to the conventions in question, and as a result come to believe that they have these moral obligations. (Shortly I will detail some of these reasons.) However, it does not follow from this that moral normative force is a necessary feature of conventions, or that conventions are necessarily social norms. For the existence of these moral obligations – whether felt or not – is a 128
4. Social Norms contingent feature of specific conventions. Nevertheless, if parties to a convention incur moral obligations to one another for some specific reason, and if these moral obligations are strongly felt and mutually acknowledged, then the convention may come to be a social norm. So some, perhaps many, conventions are also social norms. But it remains true that in order for a regularity in action to be a convention it is not necessary for it to be a social norm. The converse point should also be made, namely that in order for a regularity in action (or inaction) to be a social norm it is not necessary for it to be a convention. The existence of social norms proscribing such actions as murder and torture is sufficient to demonstrate this. One reason why parties to a convention might incur moral obligations is the possibility of free-riding. Consider what happens if someone fails to do his part to conform to a convention because on some occasion it is not in his interest to do so. Such free-riding will typically attract moral disapprobation in virtue of the perceived unfairness of the failure to conform. An example of this is pushing in front of others in line. Note that the requirement not to push in derives from the unfairness of receiving the benefit provided by the particular convention of lining up, not from the mere fact that there is some convention or other in force. In the film, The Life of Brian, a line was formed for the purpose of distributing crosses. The bearer of the cross was then nailed to his cross! Presumably, no one would object to someone pushing in in this situation. A second reason is the moral unacceptability of a failure to realise the collective end of the convention. Consider the example of driving on the left-hand side of the road in Australia. This is a convention – a regularity in action that realises a collective end. However, drivers are considered to be under a moral obligation to conform to this convention in virtue of the threat to their own life and the lives of others if they do not. A third reason is the belief that the act – which also happens to be governed by a convention – is intrinsically good, or at least is believed to be so. For example, there might be a convention to offer condolences to the bereaved, but the act of offering condolences might also be felt by members of a community to be intrinsically morally good. Finally, I will mention a more general reason why a given conven129
Social Action tion might have moral normative force. Individual conventions can attract normative force in virtue of the fact that they are constituent elements of structures of conventions, and the structures have normative force. Such structures typically provide a benefit to everyone, though constitutive conventional elements may only benefit some people sometimes. However, all agents may be under an obligation to conform to these conventional elements on all occasions in virtue of the benefits provided by the structure as a whole. For example, the convention in a community that cars give way to the right may not suit you because of the particular routes that you travel. However you might be very happy with all the other individual conventions that constitute the structure, including the convention that cars drive on the road and pedestrians use the footpath, that cars drive on the left, and so on. Accordingly, you are obliged to give way to the right. This point can be strengthened when we consider that in the case of certain structures of conventions – for example, those governing the linguistic system – it is not simply that everyone generally benefits, but that it is inconceivable that individual agents could function beyond a rudimentary level without being parties to this structure of conventions. Accordingly, everyone has a very fundamental interest in conforming to such a structure of conventions, even if on occasion they have an interest in flouting one or other of the constituent conventions, or one of the conventions most of the time. the social approval theory Having distinguished conventions from social norms it is now time to try to provide an account of social norms. I do so in the context of wanting to bypass the view that social norms are, so to speak, sui generis. Some sociological theorists, including Durkheim, proffered the view that social norms were part of the basic furniture of the world, and that while we could discover and ascribe various properties to social norms, there could be no analysis of them, and in particular no reductive analysis in terms of individual persons, actions, and attitudes. I find this Durkheimian perspective unduly pessimistic and prefer to consider more analytical acccounts.8 The fact that everyone in a given social group believes that a certain action morally ought to be performed (or ought not to be 130
4. Social Norms performed) in a recurring situation does not make the action a social norm. It may simply be an individual norm, which happens to be followed by everyone, but each acts independently of everyone else. Nor would the fact that everyone believed that everyone else ought to perform the action entail that it was a social norm. For, notwithstanding these beliefs, it could still be the case that the behaviour and attitudes of each were independent of anyone else’s. For example, each might act wholly out of a desire to keep faith with his or her own moral beliefs. If so, we would have a set of individual norms, but not a social norm. Accordingly, any account of social norms first has to concede that social norms involve interdependence of action. Indeed, at the outset of this chapter this feature was mentioned as being one of the defining features of social norms. Second, any adequate account of social norms has to concede that conformity to a social norm is not necessarily a matter of direct or indirect self-interest, and not a matter of self-interest whether that be conceived of as individual self-interest, or as the general interest of the group. I take the work of Jon Elster to have established this point. Elster invokes a whole category of norms which are counterexamples to all these versions of the self-interest view. His most powerful examples are norms of revenge.9 In Sicily, for example, there is, or at least used to be, enormous social pressure to engage in revenge killings. But very often conformity to a norm of revenge is not in the self-interest of the conformer, since, for example, it may well lead to his death. Nor is it in the interest of the community. For over time the community can be decimated if it conforms to such a norm. Indeed whole villages in Sicily were decimated for precisely this reason. Consider another example – the infamous Ik of Uganda.10 These people conformed to social norms – such as the norm that prescribed that one must return a favour. But the point is that many of these social norms were inconsistent with human flourishing and destructive of community life.11 Let me now consider the most salient analytical account of social norms, namely that which defines conformity to social norms by recourse to the notion of social pressure; we conform because others disapprove of us if we do not, and we desire to be approved of, or at least not to be strongly disapproved of. This kind of account is 131
Social Action implicit in many discussions of social norms, and has been explicitly advocated by a number of theorists, including Pettit.12 Let us refer to it as the social approval theory. Notice that the social approval theory is consistent with the fact that many norms are sustained by sanctions in the narrow sense of punitive measures such as physical punishment or incarceration. In addition, the theory is consistent with our intuition that social norms involve social attitudes of approval and disapproval, and that these attitudes function as a motivating factor in our conformity. So conformity to a social norm by one agent is dependent on the other members of the social group. However, this interdependence is not interdependence of action. One agent does not perform action x on condition the other agents perform action x. Nor on this account is there interdependence of attitude. One agent does not approve or disapprove of an action on condition the other agents approve or disapprove of that action. Rather, the interdependence is between the action of one agent and the attitudes of the other agents. An agent performs an action x on condition the other agents approve of him or her xing. It is important to note that this kind of interdependence of action on attitude needs to be distinguished from two other kinds, specifically: (a) an agent xs on condition others disapprove of his xing – the agent is a non-conformist motivated by a desire to do the opposite of what is expected of him; (b) an agent xs on condition everyone, including himself, approves of his xing – the agent conforms but is motivated by a desire to meet his own expectations of himself, as well as the expectations of others. I will return later in this section to the kind of interdependence of action on attitude described in (b). I will now consider four kinds of problem for the social approval theory, which is the theory that an agent conforms to a social norm on condition that other agents approve of his or her conformity, and that his or her own approval is not necessary. One problem with the social approval theory is that none of the conforming agents needs to approve of his or her own performance of the action or inaction that constitutes conformity to the norm. But it is a characteristic of social norms that failure to conform generates feelings of guilt, or, more particularly, shame – shame
132
4. Social Norms being the social analogue of guilt – even if others are unaware of one’s failure to conform. People do not simply conform because others want them to, and they want to please others. Rather, they conform because the whole community, including themselves, believes it to be right to conform. Fashions pose a second problem for the social approval theory. Fashions are not social norms, but fashions meet the defining conditions put forward by the social approval theory. In the case of a fashion each agent does what others approve of, largely because they approve of it. So each wears flared trousers this week largely because others approve of it, and drain-pipe trousers the week after because that is what others approve of at that time. Fashions serve to draw attention to a fundamental deficiency in the social approval theory. In the case of social norms, but not fashions, the prescribed action is performed in large part because it is, in itself, strongly approved of by everyone or most. The primary focus of each is on the goodness or rightness of the action. While the approval of others is necessary and important, it functions principally as a reinforcing factor in relation to the maintenance of the beliefs of each individual. This explains the stability of social norms in a community over time. In the case of fashions the primary focus of each is on what others think, rather than on the action considered in itself. Each performs the action that is approved of by the others, mainly because it is approved of by the others. So no one much cares which particular action she performs (within limits), just so long as it is the one approved of by the others. This explains the instability of fashions in a community over time. The third problem with the social approval account of social norms is that it cannot distinguish between social norms and certain kinds of socially competitive behaviour, which are not social norms. Assume that everyone wants to gain special attention by, say, wearing brightly coloured clothing. Assume also that no one wants anyone else to wear brightly coloured clothing. For if someone else wears brightly coloured clothing alone he or she will steal the limelight. Moreover, the worst outcome for each is wearing brightly coloured clothing when everyone else does. For then it will be obvious that each has tried and failed to gain special attention, as opposed to simply not trying and not gaining special attention. Let us also
133
Social Action assume that wearing brightly coloured clothing is in itself regarded as a desirable action. Other things being equal, each would rather wear bright clothing than wear dull clothing. In our example, given no one xs, everyone disapproves of anyone else xing (where to x is to wear brightly coloured clothing). Moreover, this fact helps to ensure that no one xs. Why? Because each will act on their disapproval of anyone else xing alone; each will x if someone else does, in order to prevent that someone else from stealing the limelight. However each knows this, and does not want to bring about a situation in which he or she is xing alongside a whole lot of others. Accordingly, no one xs. In the circumstance that no one is wearing brightly coloured clothing, the fact that each disapproves of anyone else wearing brightly coloured clothing helps to ensure that in fact no one wears brightly coloured clothing. Therefore, according to the social approval theory, no one wearing brightly coloured clothing must be a social norm in this community. But it is not a social norm. Rather, it is a stalemate in a competitive struggle for status. So much the worse, then, for the social approval theory. A fourth kind of problem for the social approval theory is that it does not accommodate the collective attitudinal character of social norms. On the social approval theory the attitudes of other agents motivate a given agent’s action, but these attitudes exist independently of one another, and they motivate only as an aggregate. By contrast, collective attitudes exist and motivate as an interdependent structure of attitudes. So a non-conforming agent confronts not only an aggregate of disapproving attitudes, but a mutually reinforcing structure of disapproving attitudes. Moreover, this structure of interdependent attitudes includes the disapproving attitudes of each to his or her own failure to conform. Consider the following example. All the members of some social group are faithful to their spouses, but only because each member doesn’t want to upset his or her spouse. Now this is not a social norm. Moreover, on the social approval account it is not a social norm. For the action of each is not dependent on the attitudes of all or most of the others. Rather, the action of any one agent is dependent on the attitude of one other agent, namely his or her spouse. So far so good for the social approval account. But let us change and complicate the example. 134
4. Social Norms Assume a homosexual community of married men in which although each agent desires to have sex with every other agent, each agent desires that his spouse not have sex with any other agent, and, in addition, each agent desires that he not upset any other agent’s spouse by having sex with that agent. What we have here is a generalised interpersonal regularity in behaviour. But it is not a social norm. Consider the motivation for any given failure to commit infidelity. With respect to any given putative act of infidelity each is only motivated by the attitudes of two other people. Fred wants to have sex with John and John with Fred, but Fred is married to Bill and John to Bob. The reason Fred and John do not have sex is that Fred does not want to upset Bill or Bob, and John does not want to upset Bob or Bill. But, on the social approval theory, fidelity in this community is a social norm. For fidelity in any given marriage is sustained by the attitudes of almost everyone else. The reason Fred and John do not have sex is Fred and John’s desire not to upset Bob and Bill; the reason Fred and Edward do not have sex is Fred and Edward’s desire not to upset Fred and Edward’s spouses, and so on. Now an advocate of the social approval theory might seek to escape this problem by maintaining that since each single (putative) act of infidelity is not blocked by the disapproval of all (or most) others, but only by the disapproval of two others, fidelity is not by its lights a social norm. But the adequacy of this response wholly depends on how we choose to describe the situation. If the “action,” or rather state of inactivity, governed by the norm, is described as “being faithful to one’s spouse;” then we have an “action” that is sustained by the attitudes of almost everyone else. For example, Fred’s ongoing fidelity to John – his refraining from sexual activity with Bill, Bob, Edward, Edward’s spouse, and so on – is sustained by a whole set of attitudes of disapproval, including attitudes of disapproval belonging to John, Bob, Bill, Edward’s spouse, Edward, and so on. At the very least the social approval theory owes us an explanation for rejecting this way of describing the situation in favour of its proposed alternative. The advocate of the social approval theory might respond that there is in fact an explanation for rejecting this way of describing the situation. The advocate might claim that in the case of social 135
Social Action norms the actions or inactions approved or disapproved of by each of the members of the community must be the same set of actions or inactions. Now in my example this condition is not met. For in the homosexual community, first, Fred and John having sex is disapproved of only by Bill and Bob, but not, for example, by Edward. (And even if Edward disapproves this will have no impact on Fred or John.) Second, and crucially, Fred’s being in a state of ongoing faithfulness to John is not an “action” that anyone (apart from John) approves of. (Of course, Fred desires to avoid doing what John disapproves of, namely Fred being unfaithful. But Fred does not disapprove of Fred being unfaithful.) However, this move is not one that is available to advocates of the social approval theory. For the social approval theory is based on the assumption that there need not be a set of actions or inactions that is approved or disapproved of by everyone, or even almost everyone. For the social approval theory entails the possibility that each not approve of his or her own conformity. The homosexual example just described accommodates this feature of the social approval theory. Fred and John do not disapprove of Fred and John having sex – even though everyone else does; Bill and Bob do not disapprove of Bill and Bob having sex – even though everyone else does; and so on. It follows from this that the social approval theory entails the possibility that in respect of any social norm there is no set of actions or inactions that everyone, or even almost everyone, approves or disapproves of. To see this, consider the homosexual example again. Fred disapproves of John, Bill, Bob, and Edward having sex with anyone, other than their respective spouses and Fred. John disapproves of Fred, Bill, Bob, and Edward having sex with anyone, other than their respective spouses and John. So the set of actions Fred disapproves of differs from the set of actions John disapproves of. For example, Fred, but not John, disapproves of John having sex with Bill, John having sex with Edward, and John having sex with anyone else (excepting Fred and John’s spouse, Bill). Similarly, the set of actions Bill disapproves of differs from the set of actions Edward disapproves of, and so on, for all the members of this homosexual community. The homosexual example draws attention to that deficiency in the social approval theory, being that it fails to incorporate the 136
4. Social Norms approval or disapproval of an agent with respect to that agent’s own actions. But it also, I suggest, highlights a further connected deficiency. For the underlying reason that we do not regard fidelity in this community of homosexuals as a social norm is that there is no collective, as opposed to aggregate, disapproval of infidelity. The attitudes of disapproval are not interdependent and mutually reinforcing. So each putatively unfaithful spouse does not confront a collective attitude of disapproval. To clarify this, consider a revised version of the homosexual community example. In this revised scenario each approves of his own conformity to the norm. However, no one else is aware of this. So the attitude of each to his own conformity is not something that concerns others. But it is a characteristic feature of social norms that non-conformers are upbraided and shamed into conformity. It is not simply that others believe that an agent ought to conform; rather they believe that the agent ought to believe that he or she ought to conform, and feel guilty about failing to conform. Indeed, the institutional device of confession, which is pervasive in many cultures, hinges on precisely this feature of social norms.13 I conclude that the social approval theory fails to do justice to the collective character of the attitudes of approval and disapproval constitutive of social norms. Contra the social approval theory, each agent approves of the conformity of everyone or most (including his or her own conformity) and disapproves of non-conformity (including his or her own); and these attitudes of approval and disapproval of each are dependent on the same attitudes of the others. There is interdependence at the level of attitudes, as well as at the level of action. The four arguments and counter-examples put forward here point to a number of significant deficiencies in the social approval theory of social norms. Cumulatively they constitute a sufficient reason to reject the account. In the following section on social norms I will tentatively offer a different kind of account; one that accommodates the various problems thus far identified. Notwithstanding my rejection of the social approval theory of social norms, I do not dispute the obvious point that social approval is a powerful and pervasive social force, and one that maintains a large number of regularities in action, including fashions and other forms of social conformity. 137
Social Action One of the most important aspects of the desire for social approval is the desire for status, whether it be the status of a pop star, sports star, or academic star. But I have demonstrated that the concept of a regularity in action sustained by the desire for status is not to be identified with the concept of a social norm – a regularity in action sustained by moral beliefs. On the other hand, it is conceivable that a desire for status, and a corresponding adulation of those with status, comes to assume such importance that it overrides hitherto accepted moral considerations; perhaps this was the case with monarchs in the past, and is now the case with pop stars in Western societies. That is, among the members of some social group the desire for, and adulation of, status may come to assume the role and substance of a moral belief. In that case, the regularities sustained by these beliefs in the importance of status would be social norms. Naturally, such regularities might not be objectively morally valid; indeed, their existence might be cause for moral concern. But that is a different matter. social norms This discussion suggests that an adequate account of social norms has the following features. First, a social norm is a regularity in action (or inaction) in a social group, and the existence of this regularity is a matter of common knowledge.14 Second, members of the social group disapprove of any failure to perform the action (or inaction), including a failure by themselves, and this disapproval is also a matter of common knowledge. Third, this attitude of disapproval has moral force. (This is a defining feature of the category of social norms being analysed.) Fourth, the attitude of disapproval within the social group is what I have termed a collective attitude, rather than a mere aggregate of individual attitudes. Fifth, the collective disapproval of members of the group at least in part sustains conformity to the norm because each desires to do what we (collectively) approve of, and to avoid doing what we (collectively) disapprove of. This is consistent with some norms being in part sustained by sanctions (in the narrow sense of physical punishment, incarceration, and so on). In the light of this discussion, I offer the following analysis of social norms. 138
4. Social Norms A regularity in action, r, is a social norm among a group of individuals if and only if: (1) most of the individuals conform to r ; (2) each (or at least most) believes that each (including himself or herself) morally ought to conform to r ; (3) (1) and (2) are matters of common knowledge; (4) (2) is a necessary condition for (1) in virtue of the fact that almost every individual conforms to r in part because he or she believes he or she morally ought to, and in part because almost everyone else believes he or she morally ought to;15 (5) the belief of almost any individual that each (including himself or herself) morally ought to conform to r is in part dependent on almost everyone else’s belief that each morally ought to conform to r.
Armed with this characterisation of social norms, I offer the following remarks. First, some actions prescribed by social norms are also conventional actions, some are joint actions but are not conventional actions, and some are individual actions. Driving on the lefthand side of the road is a convention in Australia and a social norm – a social norm enshrined in the law. Sex between spouses is a joint action and a social norm. Refraining from killing or assaulting another person is an individual action (inaction), a social norm, but not a convention. Second, social norms are in part constitutive of a framework of accepted moral principles, against which the members of some social group act and interact.16 The argument for this is simply that agents cannot deliberate and decide in relation to all their morally informed actions – they perform too many actions, including too many interpersonal and social actions, for this to be possible. Third, many social norms are internalised by their adherents in such a way that their commitment to the actions or inactions prescribed by these norms comes to be in part constitutive of the selfhood or identity of individual adherents. So the content of the beliefs of these moral agents is the same as the content of the social norms of the social group to which these moral agents belong. But the moral beliefs of a person determine what is most important to that person; indeed, in part they constitute the person’s identity. Accordingly, once an individual internalises and identifies with a social norm it becomes a matter of integrity for that individual to 139
Social Action act in accordance with that norm and a potential threat to his selfhood, if he does not. Consider in this connection hitherto morally upright police officers who are seduced into engaging in corrupt activities and are then investigated and exposed. Suicide is a not uncommon result.17 However, there are differences between an individual’s failure to adhere to his or her moral principles and the individual’s failure to adhere to the social norms of the group to which he or she belongs. The former failure is necessarily a reflection on the individual’s integrity. Not so the latter failure. For from the fact that some regularity in action is a social norm in the group to which an individual belongs, it does not follow that the individual has a commitment to that kind of action. Social norms and individual moral principles typically coincide, but this is not necessarily the case. Individuals will often flout objectively valid social norms for reasons of self-interest. Corrupt actions are a case in point. Some individuals flout objectively valid social norms, not for reasons of selfinterest, but because their own subjectively felt (but objectively invalid) individual norms dictate that they do so. Consider in this connection assassins or bank robbers who are entirely committed to what they regard as their profession.18 Widespread corruption is typically the result of a large number of individuals ignoring objectively valid social norms. On the other hand, some social environments are constituted by objectively invalid social norms. Consider the social norms that govern the activities of members of the Mafia. Such social environments are constitutively corrupt. Sometimes objectively valid social norms come into conflict with regularities driven only by social approval. Consider a social climber driven by a desire to have social status. Such a person might well mouth the kinds of views that would win approval within a social group, and in doing so fail to comply with the social norm of truthtelling. Indeed, in some social groups there might well be a systematic tension between the framework of social norms on the one hand, and the set of regularities driven only by social approval on the other.
140
4. Social Norms social norms and laws Given the above account of social norms, it is easy to see why citizens feel they ought to obey many of the laws of the land, and in particular criminal laws. For the criminal law is an explicit formulation (backed by penal sanction) of the most basic of a society’s social norms. Citizens believe that they ought not flout the laws against murder, theft, rape, assault, and so on, because these citizens have internalised a system of social norms that proscribes such behaviour. Putting matters simply, for the most part any given citizen does not commit murder in part because he or she believes it is wrong to murder, and in part because others believe it is wrong to murder. Unfortunately, there are some citizens who have not internalised the system of social norms, or who have not internalised that system sufficiently. Accordingly, there is a need to buttress the system of social norms by the construction of a criminal justice system. The latter system involves the detection of serious moral wrongdoing and the trial and punishment of offenders. This account of compliance with laws explains why citizens feel an obligation to obey the criminal law, but it does not provide a justification for an objective obligation to obey the criminal law. Nor does it explain the felt obligation – let alone any objective obligation – to obey those laws that are not criminal laws. Let me deal with the former problem first. Any objective obligation to obey the criminal law will be based on the objective moral merits of the specific criminal laws in question. For example, there is an objective moral obligation to obey the law against murder because the social norm proscribing murder is not only a subjectively felt obligation, it is also an objective moral obligation. The criminal law, the social norm, and the objective moral principle coincide. How do matters stand when a given criminal law does not coincide with a relevant objective moral principle? In the case where the criminal law in question infringes some central moral principle or right, there is no moral obligation to obey the law. For example, in South Africa under apartheid it was a criminal offence under the Immorality and Mixed Marriages Act for persons belonging to different race groups to have sexual relations with one another or to get married. 141
Social Action However, in other cases there might still be an obligation to obey criminal laws that do not infringe important objective moral principles, or that infringe only relatively minor objective moral principles. For example, in some states of Australia it is a minor criminal offence to possess marijuana. Arguably, such criminal laws – while they ought to be repealed because they unnecessarily restrict certain freedoms – do not violate any important rights. So perhaps such criminal laws, while unjustifiable, ought to be obeyed. I suggest that if such laws ought to be obeyed, they ought to be obeyed by virtue of a general obligation to obey the law generated by other considerations. What are these other considerations? They are considerations that generate the objective obligation to obey laws that do not embody moral (or immoral) principles, that is, much of the noncriminal law. In other words, we have come to our second problem. In the remainder of this section I will offer a detailed treatment of this problem and the various attempts to deal with it. It has proved difficult to provide the required rational underpinning for this felt moral obligation to obey the law. For example, theories in terms of consent to obey the law seem unable even to get to first base; most citizens have simply never consented to obey the law. More recently, Hart and Rawls have developed the so-called argument from fair play.19 Roughly the idea is that (many) laws should be construed as mutually beneficial cooperative schemes. And, if this is so, it is unfair to disobey the law when one has accepted its benefits; as a beneficiary one is under a moral obligation to obey the law even if it is not in one’s interest to do so. However, the argument from fair play has come under attack, including from Nozick and Simmons,20 and these attacks have in turn provoked a reworking of the argument by Luban. In this chapter I criticise Luban, but do so with an eye to improving the argument from fair play. My discussion makes use of the Collective End Theory developed in Chapter 1. Luban, in his influential book, Lawyers and Justice,21 argues that the justification for a moral obligation of citizens to obey at least some laws – laws that are not evil, unfair, or hopelessly stupid22 – lies in the fact that each of these laws constitutes an important, or at least, reasonable, cooperative scheme23 for the members of the citizenry, and that to break such a law is unfair to those who obey it. It should be noted that even if Luban’s argument demonstrates that 142
4. Social Norms there is an obligation to one’s (law-abiding) fellow citizens to obey these laws, he has not shown that these law-abiding citizens have a right to enforce compliance with these obligations. Luban mentions four conditions in his discussion of the moral obligation to obey a law:24 (1) the law is generally beneficial – the so-called generality requirement25 to the effect that the law benefits citizens and does so in a fair and non-discriminatory way; (2) most citizens comply with the law; (3) citizens accept the benefits of the law; (4) the law is an important, or at least reasonable, cooperative scheme. Luban claims that conditions (1), (2), and (4) are jointly sufficient to establish an obligation to obey a law.26 Luban also argues against Simmons’s claim that condition (3) is necessary.27 That is, Luban rejects the claim that there is an obligation to obey a law only if the person thus obligated had an opportunity to decline the benefit arising from that law, and chose to accept the benefit. Simmons thinks condition (3) is necessary because being an active participant – as opposed to merely conforming to the law – is necessary, and (3) is a necessary condition for being an active participant.28 It is not clear whether Luban accepts Simmons view that (3) is a necessary condition for being an active participant (hereafter, participant) in a law. What is clear is that Luban, in claiming that (1), (2), and (4) are jointly sufficient to generate an obligation to obey a law, commits himself to rejecting the proposition that being a participant is a necessary condition for having the obligation to obey a law.29 Luban’s initial point is that the benefits of most laws are in fact thrust upon the citizen; citizens are in general not in a position to reject the benefits laws confer. But the fact, if it is a fact, that the benefits of laws are thrust upon citizens does not in itself show that participating in laws is not necessary for being under the obligation to obey the law. In the first place, it might not be the case that a necessary condition for participation in a law is that there has been an opportunity to refuse the benefits. In the second place, it might be that: (a) participation is a necessary condition for having the obligation to obey the law; (b) there is no participation because the benefits of laws are in fact thrust upon citizens; and, therefore, (c) there is no obligation to obey the law. Luban does not think he needs to assume that there is in fact an 143
Social Action obligation to obey the law.30 He takes himself to have an argument against the claim that being in a position to decline benefits is a necessary condition for being obligated to a law. Indeed, he takes himself to have a further argument for the stronger claim that conditions (1), (2), and (4) jointly constitute a sufficient condition for having the obligation to obey the law. Let me now turn to Luban’s arguments. Luban argues that there are cases in which it is unfair to free-ride even though there is no opportunity of declining benefits.31 In such cases one is under an obligation not to free-ride. Luban’s argument, if valid, would show that being in a position to decline benefits is not a necessary condition for having the appropriate obligation. Luban puts forward a number of cases in order to try to demonstrate this point. One such case is that of failing to help clean up glass on one’s street, and then making full use of the fact that the street has been cleaned up by others. (The street will be cleaned up by others whatever one does.) This case is different, he suggests, from failing to assist others to plant and to tend flowers on the median strip of the street. (The case is similar to the street cleaning example in that the others will plant the flowers whatever one does, and one is not in a position to decline this benefit.) Luban suggests that the difference is that in the first example it is essential that the street be cleaned. That is, Luban at this point introduces condition (4). More precisely, Luban suggests that the more important or reasonable a cooperative scheme is, the less it matters if the benefit received is actively accepted. Acceptance does not matter in the glass example, but does in the gardening example. In the glass example, but not the gardening example, free-riding is unfair and disrespectful to one’s fellows.32 Luban’s examples demonstrate that being in a position to decline benefits is not a necessary condition for having the obligation to contribute to a cooperative scheme. However, I will argue that being in a position to decline benefits is not a necessary condition for being a participant in a cooperative scheme. Therefore being a participant may well be a necessary condition for being obliged to contribute to a cooperative scheme. Further, I reject Luban’s view that the conjunction of conditions (1), (2), and (4) provides a sufficient condition for having the obligation to obey the law. 144
4. Social Norms Before proceeding any further, it may be helpful to be clear about individuals’ rights and obligations in the standard cooperative scheme. In such a scheme there are two levels of “activity”; the level of contribution and the level of acceptance of benefits. Individuals perform a contributory individual action in order to realise a collective end. (For an analysis of the notion of a collective end see Chapter 2.) In cooperative schemes the collective end consists of some good or benefit. In the case of laws, typically (though not invariably) the collective end can be realised without the participation or contribution of everyone, and the collective end realised is a collective good. (A collective good is one such that if it exists it is necessarily able to be enjoyed by everyone. Clean air is an example.) Accordingly, in the case of laws there is the possibility of freeriding. In cooperative schemes, obligations and rights arise as follows. If an individual contributes to the scheme, then the individual has a right to the benefit. If an individual does not contribute, then the individual has no right to the benefit, unless by the consent of those who contributed to the provision of that benefit, for example if they offer the benefit as a gift. This gives rise to at least three possibilities. In two of these the agents fail to contribute. First, there is the agent who both contributes and accepts the benefits. This agent is a standard participant and is under an obligation to contribute. Second, there is the bona fide nonparticipating agent who refuses to contribute and refuses the benefit. Such an agent is not under an obligation to contribute. Third, there is the free-riding agent who always takes the benefits and yet fails to contribute unless it is necessary in order for the scheme to succeed. This agent is under an obligation to contribute having taken the benefits.33 A problem arises in cases in which it is not possible (or not possible without considerable difficulty or hardship) for an agent to refuse the benefit provided by some cooperative scheme. Can agents be obliged to contribute to the scheme if they cannot refuse the benefits? Such problematic cases include Luban’s flower example and his glass example. In cases where agents are not in a position to decline a benefit, what makes any given person a bona fide non-participant – without rights to benefits or obligations to contribute – rather than a free145
Social Action riding participant who wants to exercise a right to a benefit, but does not want to discharge the obligation to contribute?34 What makes an agent a free-riding participant, as opposed to a non-participant, in such cases, is what makes an agent a free-riding participant in any case, namely the fact that the agent would contribute if it were necessary in order to provide the benefit. The freeriding participant, unlike the nonparticipant, would contribute if he or she had to; for participants – whether free-riders or not – are committed to realising the purpose or end of the cooperative scheme.35 So Luban’s flower example involves a non-participant and his glass example, a free-riding participant. And this is the reason, contra Luban, why the agent in the glass example, but not the flower example, is under an obligation to contribute to the cooperative scheme. The agent in the glass example, being a participant, is under an obligation to contribute. It might be argued that in cases in which the benefit is imposed, there is no way of determining whether an agent is a free-riding participant or a bona fide non-participant. I reject this argument. There is all sorts of evidence to distinguish free-riders from nonparticipants. There can be evidence for the agent’s willingness to contribute if his or her contribution is necessary in order to realise the collective end of the scheme. For example, in the Luban glass case, if the free-rider has an expensive car with thin tyres, needs to drive to work, and swept up glass the time before when most people were out of town, then we have evidence that this agent is a participant seeking to free-ride. (I will say more about the evidence for free-riding in relation to laws, in a moment.) Another kind of evidence would be the agent’s attitude to the contributions of other agents. For example, in the Luban gardening example, if the agent said nothing to the neighbours when he or she noticed that the flowers were not being tended, and indeed were beginning to die off, then we could assume that the agent was a non-participant in the cooperative scheme. The agent is not prepared to do anything to rescue the scheme. I conclude that while Luban has shown that being in a position to decline a benefit is not a necessary condition for having the obligation to obey the law, he has not shown that being a participant is not a necessary condition for having the obligation to obey the law. 146
4. Social Norms Let me now turn to Luban’s view that, taken together, conditions (1), (2), and (4) constitute a sufficient condition for generating an obligation to obey the law. (Naturally, if (1), (2), and (4) are (jointly) sufficient, then being a participant is not necessary.) Condition (2) is that most citizens in fact comply with a law. If most people do not comply with a law then the law is failing to realise its collective end and it becomes difficult to see how under these circumstances an individual could be under an obligation to obey it. So condition (2) looks as though it must be a necessary condition for generating an obligation to obey a law. What of conditions (1) and (4)? Luban terms condition (1) the generality requirement. This condition amounts to Luban’s requirement that a law be fair and neither stupid nor evil. For presumably a cooperative scheme that is not stupid, evil, or unfair is so in virtue of the following facts. First, it provides a benefit that outweighs the cost of contributing to it, and there is no known and clearly preferable alternative way of securing the benefit. Second, it benefits everyone, and does not require anyone to contribute a greater share than anyone else in relation to the benefit that person receives. Condition (2), taken in conjunction with condition (1), is not sufficient to generate an obligation to contribute to a cooperative scheme. For, in the first place, the benefit in question may be quite trivial, and contributing to the provision of trivial benefits is not a matter of moral obligation. And in the second place, a given agent may well have some alternative course of action, which will provide the agent with some other benefit which may be of greater importance to him or her than the benefit to be derived from participating in the cooperative scheme. It is not that the benefit provided by the cooperative scheme is trivial. It is just that some other individually attainable benefit is more important to that agent. Are conditions (1) and (2), taken in conjunction with condition (4), sufficient to generate an obligation to obey the law? Condition (4) rules out trivial benefits. However, there is nothing in these three conditions to rule out the possibility of an agent who does not want to contribute to a cooperative scheme because he or she wishes to pursue some other individually attainable benefit which is of greater importance to that agent. Suppose, for example, that the roads in some neighbourhood 147
Social Action become snowed over. The members of the community regularly go out and clear the snow off the roads. But suppose there is a somewhat reclusive composer who is actually prepared to forgo driving during the relatively short winter rather than see to it that the roads are passable. The composer’s life would be made fairly difficult by impassable roads. For example, the composer would not get any fan mail, and would have to stockpile food. However the composer would rather this than have to regularly perform the somewhat arduous and time-consuming task of shovelling snow.36 Sometimes an agent or agents have an obligation to conform to a scheme that burdens that agent or agents, but which significantly benefits another agent or agents. But such an obligation has little to do with the fairness of a cooperative scheme. Rather, it concerns the importance or moral value of the collective end realised by the cooperative scheme. Such obligations arise, especially, in cases of need – as opposed to desire for a benefit – and the greater the need, the greater the disadvantage one ought to be prepared to suffer in order to help fulfil that need. The need in question may belong to a majority or a minority of the participants in the scheme. In the latter case, the collective end of the scheme does not consist in a collective good or benefit. Suppose that in the snow-clearing example it was known to the composer that some other members of the community needed access to a hospital in the city. There would now be an obligation on the members of the community, including the composer, to ensure that the roads were kept clear. But this has little to do with the fairness of contributing to a scheme from which one benefits. The composer is obliged to help the infirm, irrespective of the fact that to do so does not benefit the composer. The upshot of this discussion is that there are (at least) two sorts of basis for an obligation to contribute to a cooperative scheme. There is the obligation, if any, deriving from the moral value or importance attaching to the collective end realised by the cooperative scheme. Fairness is the other basis of the obligation to contribute. The obligation of fairness derives from the fact that having become a participant in a cooperative scheme, and therefore a beneficiary of it, one is under an obligation to do one’s part to realise that benefit. In the case of some cooperative schemes, there is no moral obli148
4. Social Norms gation to become a participant. However, in some of these, if one is a participant, fairness demands that one contribute. In other cases it is morally encumbent on agents to secure some collective end, irrrespective of whether the cooperative scheme that secures this end is a fair one. In still other cases considerations of both fairness and the moral value of the collective end of a cooperative scheme generate obligations to contribute. I have argued that Luban fails to demonstrate that: (a) being an active participant is not a necessary condition for being under an obligation to obey a law; and (b) conditions (1), (2), and (4) are sufficient for generating the obligation to obey the law. I have also argued that one source of Luban’s problems is a failure to adequately accommodate the distinction between the fairness of a cooperative scheme and the value of the collective end that the scheme might realise. I want now to argue that a further source of Luban’s problems is that his account is focused too narrowly on the individual law or cooperative scheme. If the moral obligation to obey laws is to be properly understood, whole structures of laws – whole structures of cooperative schemes – need to be considered. For if this is done it becomes clear that in many instances the apparently bona fide non-participant is actually a participant in the overall structure of cooperative schemes, but is nevertheless unfairly trying to opt out of certain individual constituent schemes. He or she is in reality a free-rider. The agent wants to opt out of those individual constituent schemes that benefit others but not the agent, while expecting others to participate in individual constituent schemes that benefit the agent but not the others. In such cases the free-rider is involved in an inconsistency, and hence unfairness, across individual cooperative schemes. The law is a good example of such a system of cooperative schemes. Many laws only benefit some individuals. However, the issue is whether the whole system of laws on balance benefits everyone and to a reasonable extent. If so, then failing to contribute in the case of a particular law may well be inconsistent and unfair to law-abiding citizens. Here, breaking the law is simply a special case of unfair free-riding in a cooperative arrangement of the sort envisaged by Luban. As such, breaking the law will constitute a failure to discharge a moral obligation. And there is a further point. It may be that the collective end 149
Social Action realised by the whole structure of laws, say the collective end of good order, may be so important that everyone is under an obligation to obey the law, even though the structure of laws is unfair. This is not to say that the importance of realising the collective end of the system of laws in some cases might not be outweighed by the unfairness of that system. How do we determine in relation to a given instance of law breaking whether it is a case of free-riding or of non-participation? In respect of law breaking we need to distinguish on the one hand, between individual law breaking and law breaking by collectives, and on the other between breakage of an individual law and rejection of the system of laws or of large fragments of a system. I have already suggested that, in relation to fairness, what is important is the whole system of laws. It is whether or not an individual or group participates in the whole system of laws that is important. I suggest that, in relation to systems of laws or large fragments of systems, there can be clear evidence that some group is essentially a non-participant being forced to participate. In South Africa under apartheid, that group consisting of black South Africans provides one clear example of this. This does not show that with respect to all groups desirous of non-participation there could be evidence. Presumably there are instances where there could not be evidence. But in any society where a group has a strong desire not to participate, and where there is some space to vent opposition to the system of laws (or some large fragment of the system) there is likely to be such evidence. What of the individual in relation to the system of laws? Let us set aside those laws that embody social norms and express moral prohibitions, such as laws against murder. I suggest that it is almost never the case that the set of the remaining laws of some legal system – or large fragments of the set of the remaining laws – are such that one person only (or even a handful of people) would rather not be a participant in those laws. Rather, it is almost always the case that it is some group of individuals that would prefer not to participate. Typically, such a group will consist of those individuals who are being burdened in various ways by the system of laws. But in that case, in most situations where there is some space for groups to express their desire not to participate in the system of laws, the individual will be able to manifest his or her desire not to participate 150
4. Social Norms in that system of laws; he or she will be able to do so in consort with other individuals who have a similar desire. In this section I have argued the following. First, contra Luban, the notion of an active participant in a mutually beneficial cooperative scheme can be used to generate the obligation to obey the law, provided there is evidence that the participants would have complied with the law and accepted its benefits if this was necessary for the realisation of the collective end of the cooperative scheme. Second, Luban fails to demonstrate that one is under an obligation to obey a law if the law is a generally beneficial and important cooperative scheme that most people comply with (but are not active participants in). Third, we need to distinguish between a moral obligation generated by the fairness of a law and one generated by the moral value of the collective end secured by a law, and accept that both kinds of moral obligation can be used to justify obedience to the law. In this chapter I have put forward a detailed account of a central category of social action, namely social norms, and distinguished them from the category of regularities in action based only on social approval, for example fashions. I have also considered a further category of regularities in action, namely laws, and used my analysis of social norms and that of joint action (developed in Chapter 2) to provide an explanation of the moral obligation to obey the law. Now I want to consider the relationship between social norms and practical reason. In our focus on the rationality of conformity to cooperative schemes (including conventions), and in particular on the sources of the moral obligation to conform to such schemes, it has turned out that conformity to such schemes may in the last analysis be motivated by commitment to moral principles. Indeed, given that these schemes involve social action directed to collective ends, it is highly likely that the moral principles in question are also social norms. So let us now turn directly to practical reasoning in relation to social norms. social norms and practical reason The rationality of agents who act in conformity to social norms is practical, as opposed to theoretical, rationality. It is practical reasoning since it has action as its point. This is not to say that the conclu151
Social Action sion of a piece of practical reasoning is necessarily an action, as opposed to a proposition. Perhaps the conclusion of a piece of practical reasoning is a proposition to the effect that the agent ought to perform such and such an action. I will not be concerned here to defend any particular view on this dispute.37 Nor will I take sides in relation to the dispute as to whether all practical reasoning is ultimately means–end reasoning (or at least means–end reasoning supplemented by minimal theoretical reasoning, for example, simple deductive reasoning); so I will not dispute the existence of other forms of reasoning, including reasoning from principles that is not reducible to means–end reasoning.38 I will simply speak in terms of a number of prima facie different sorts of reasoning, leaving aside the issue of reducibility. Finally, and relatedly, I will also avoid the issue as to whether moral principles or beliefs can in themselves have motivational force.39 In order to get an understanding of reasoning in accordance with moral principles embodied in social norms it might be helpful to remind ourselves of the structure of means–end reasoning. Roughly speaking, individual means–end reasoning proceeds in the following manner. First, identify the end that one has. Second, identify the means to that end. Third, perform the action that is the means to the end.40 Notice that here ends are not necessarily goods or benefits; they are simply ends. In the case of cooperative schemes – involving, as they do, joint actions directed to collective ends – matters are more complicated. But here is a rough summary of such reasoning. First, agent A has end e, and agent B also has end e, and so does C and so forth. Moreover, A has end e only on condition that B has end e, and so on. Second, A and B and C and so on believe that e cannot be achieved other than by A, B, C, and so on, each performing some action x.41 Third, A, B, C, and so on each has the true belief that each other has performed, is performing, or will perform x, and these beliefs are a matter of common knowledge. Four, A, B, C, and so on each performs x. Notice once again that here ends are not necessarily goods or benefits; they are simply ends. Now let us consider means–end reasoning in relation to ends that are also regarded as goods or benefits. First, agent A comes to have e as an end for the reason that A regards e as a good or benefit, g. Second, agent B has end e, and agent C also has end e, and so does 152
4. Social Norms D and so forth. A, B, C, and so on have e as an end for the reason that they regard e as a good or benefit or as a means to some other good or benefit. But they might not regard e as being the same good or benefit as does A, or indeed as being the same good or benefit as one another do; indeed, they might not even regard e as a good or benefit. For example, you and I might jointly kill a stag. But you might do it for the sake of the killing; you regard killing the stag as a good thing. I, on the other hand, might be involved in killing it purely to eat it. The good aimed at is the relief of hunger; I might be indifferent to killing as such, or even abhor it. Note that here A has end e only on condition that B has end e, and so on. So e is a collective end. 42 Second, if A and/or B and so on, have competing ends, e overrides the other ends, and does so by virtue of the greater importance that A, B, and so on, attach to the goods or benefits that the realisation of e will bring. Third, A and B and C and so on, believe that e cannot (perhaps, without difficulty) be achieved other than by A, B, C, and so on, each performing some action x. Fourth, A, B, C, and so on each has the true belief that each other has performed, is performing or will perform x, and these beliefs are a matter of common knowledge. Fifth, A, B, C, and so on, each performs x. So much for the basic structure of practical means–end reasoning in the case of both individual and joint actions, including joint actions performed to produce what are believed to be benefits or goods. It is important to note that in light of the need to reason from beliefs about goods or benefits to ends, the reasoning in this example is not pure instrumental reasoning, at least as we are using the term. So our distinction between moral reasoning, including moral reasoning involving ends, and instrumental reasoning has been preserved. Now we need to consider reasoning from principles, specifically moral principles that are also social norms. Presumably, individual reasoning from a moral principle, say the principle of truthtelling, proceeds in roughly the following manner. First, identify the relevant moral principle, for example the principle to tell the truth. Second, determine which action is prescribed by that principle in the context in question. Third, perform the action by speaking the truth in relation to the matter in question. Now let us consider reasoning in accordance with a social norm. First, identify the relevant moral principle adhered to by all of us, 153
Social Action for example truthtelling. Second, determine which action is prescribed by that principle (social norm) in the context in question. Third, perform the action prescribed by the social norm. In light of the above, I want to draw attention to three constraints on practical reasoning in accordance with social norms, at least in the elementary cases of such reasoning. (I will look at some more morally complex cases later on.) The first constraint arises from the fact that reasoning in accordance with a social norm is typically reasoning in accordance with what I termed in Chapter 2 a procedure, as opposed to a mere regularity. In the case of a regularity in action an agent provides himself or herself (or could do so) with a reason for performing the action he or she regularly performs, and does so (or could do so) on each and every occasion of his or her performance of that action. In the case of procedures, by contrast, there is a single decision to perform many actions. One decision, then – the decision to adopt the procedure – governs a whole set of future actions. One process of reasoning will take place, namely that justifying the adoption of the procedure. Thereafter the agent will not provide himself or herself with a reason for acting; he or she will merely identify the situation as being of the sort for which he or she has adopted the procedure in question, and then simply follow the procedure and perform the action. You will recall that there is a presumption in favour of the procedurally determined action, a presumption that can only be offset by the non-obtaining of the condition or conditions under which the procedure is to be performed – especially the condition that the agent (or agents) has the end that the procedure realises. The second constraint on practical reasoning in accordance with social norms – at least in the elementary cases – arises from the fact that any given agent’s conformity to social norms is by definition to some extent dependent on the moral attitudes of the others who conform to that social norm. Here it is important to see that the interdependence is in the first instance interdependence at the level of attitude, as opposed to interdependence at the level of action. It is not that I tell the truth in part because others tell the truth. Rather I tell the truth in part because I believe that I ought to tell the truth, and my belief that I ought to tell the truth is in part
154
4. Social Norms dependent on the fact that the others believe that I (and others) ought to tell the truth. This interdependence of moral attitude might seem to make action performed in accordance with social norms profoundly irrational. What is rational about conformity to principles merely because they are socially accepted? This charge misunderstands the claim being made. First, on the account of social norms that I have presented, an agent’s belief in (say) truthtelling is not completely dependent on a similar belief of others; there is only partial dependence. Second, an agent could engage in rationally justified revision of any of the moral principles he adheres to, including his belief that he and others ought to tell the truth; but he could not revise all or most of these principles simultaneously. Or at least he could not do so while simultaneously continuing to cooperatively interact with the fellow members of his “moral community.” Third, the members of the “moral community” could jointly over time engage in rationally justified revision of large fragments of the structure of moral principles that they adhere to. In short, the second constraint on practical reasoning in relation to social norms amounts to the acknowledgment of the limits on the capacity of each agent living in a “moral” community to rationally justify the whole set of moral principles to which he or she adheres. There is, so to speak, a presumption in favour of the social norms adhered to by the members of the community to which he or she belongs. This presumption in favour of any given social norm can be offset by rational considerations; but there is nevertheless a presumption to be offset. The existence of this presumption amounts to a constraint on each agent’s practical reasoning in relation to social norms that embody felt moral principles. And there is a third constraint on practical reasoning in relation to social norms; or at least reasoning in relation to social norms that derive from projects whose ends embody felt moral values. As Michael Bratman has successfully argued in relation to future directed intentions, it might be rational for an agent with such intentions not to reconsider one of those intentions, even though it might be rational to reconsider that intention, and indeed change it, from an all things considered external viewpoint.43 The general point here is that agents that are long-term planners need to build a degree of
155
Social Action stability into their future directed intentions or ends if they are to achieve them; they need to focus on the means to the ends, rather than be constantly questioning the rationality or wisdom of the longterm ends themselves, or embarking on different projects that would realise different ends. So there is a presumption in favour of maintaining, rather than abandoning, or even questioning, long-term ends and future directed intentions. In light of this, what is our third constraint on practical reasoning in relation to social norms that derive from long-term projects or enterprises? The third constraint arises from the fact that an agent living in a “moral” community typically contributes to a variety of long-term joint projects or enterprises that realise collective ends that are pursued (at least in part) because they are believed to be morally worthy ends; such ends include collective goods. Consider in this connection a school teacher, a doctor in a hospital, a police officer, or a worker in the clothes industry. And consider a taxpayer or a voter or a parent or a spouse. On pain of failing to realise any such projects, it would be irrational for all or most of the members of a community to be constantly questioning the rationality or morality of the ends of such projects or enterprises. Moreover, as was the case with social norms that embody felt moral principles, the fact that an agent believes that the ends realised by such collective enterprises or projects are morally worthy is to some extent dependent on the moral attitudes of the others who jointly pursue those ends. One citizen believes in the provision of public education by means of funds raised by taxation, but does so on condition (at least in part) that the other citizens believe in the provision of public education by that means. Accordingly, such an agent ought to have a presumption in favour of not abandoning the felt moral ends embodied in the long-term projects or enterprises pursued in the community to which he or she belongs. This presumption in favour of any such project or enterprise can be offset by rational, including moral, considerations; but there is nevertheless a presumption to be offset. The existence of this presumption amounts to a constraint on each agent’s practical reasoning in relation to the felt moral ends constitutive of joint projects or enterprises, and therefore in relation to social norms that might derive from the pursuit of those ends. So there is a presumption in favour of paying one’s taxes. 156
4. Social Norms Thus far I have only considered what I have termed elementary cases of practical reasoning in relation to social norms. Specifically, I have not considered cases of practical reasoning in situations in which the relevant social norms prescribe contradictory courses of action. Consider the case in which the Nazis are coming for a Jewish person hidden in your house. You can choose to tell the truth or tell a lie. If you lie you may save a life; if you tell the truth then the result will certainly be the death of an innocent person. Now in some of these kinds of cases there might be a socially accepted higher order moral principle – a social norm – in place; a higher order social norm according to which (say) the social norm of saving lives always takes precedence over the social norm of telling the truth. However, in many situations in which different social norms prescribe contradictory courses of action, there are no readily applicable higher order social norms. Accordingly, moral agents will need to resolve these situations on a case by case basis. These situations have a moral complexity that calls for moral deliberation, and not simply the mechanical application of pre-existing (including higher order) social norms. In these morally complex situations the three constraints on practical reasoning in relation to social norms cannot be respected, or at least cannot be fully respected. It is obvious that the first constraint cannot be respected, since these situations need to be looked at on a case by case basis. But the second and third constraints are also inappropriate. For these situations are by definition ones in which the individual moral agent (or agents) involved necessarily have to make moral judgments that outrun existing social norms, including socially accepted moral ends, and which therefore have not been socially sanctioned. Needless to say, these situations still involve practical reasoning in relation to social norms; social norms provide the starting point in this reasoning, even though they do not fully determine the outcome of it. From the perspective of this chapter, these kinds of morally complex situations are of two main kinds. The first kind involves situations calling for an individual to perform some individual action or other. The Nazi case might be an example of the first kind. The second kind involves situations in which two or more individuals find themselves, and what is called for is a joint action. A community deciding whether to go to war or suffer invasion might be an exam157
Social Action ple of the second kind. I will now outline the basic steps in practical reasoning from social norms involved in these two kinds of situations, beginning with the kind of situation calling for individual, but not joint, action. The reasoning in this kind of situation is roughly as follows. (1) Identify the relevant social norms – especially competing social norms – and any relevant individual and morally significant principles or ends. (2) Determine the relative moral weight to be given to these various social norms, and individual principles and ends, in the situation in question. (3) In the light of the above weighting – and any judgments concerning preferred means to adopted ends – determine what individual action ought to be performed. (4) Perform the action thus determined.
The reasoning in the second kind of situation – calling for joint action – is roughly as follows. (1) Jointly identify the relevant social norms – especially competing social norms – and any relevant morally significant collective ends. (2) Jointly determine the relative moral weight to be given to these various social norms and collective ends in the situation in question. (3) In the light of the above weighting – and any jointly determined judgments concerning preferred means to adopted collective ends – jointly determine which joint action ought to be performed. (4) Each performs the action that contributes to the joint action thus determined, or at least each forms the (shared) belief that each ought to perform his or her contributory action.
These outlines of practical reasoning in relation to the application of social norms in morally complex situations are outlines only; they are crude and incomplete summaries. Moreover, nothing has been said of a wide variety of still more complex situations, including ones in which individual moral principles and ends might compete with collective ends, as well as with social norms. Nevertheless, it does signal the direction in which this more detailed work needs to go. This section has sought to provide the beginnings of a philosoph158
4. Social Norms ical articulation of practical reasoning in relation to social norms, relying on a specific philosophical account of social norms. It has done so for the straightforward case in which an action is prescribed by a social norm, as well as for some of the morally complex cases in which social norms prescribe contradictory courses of action. In earlier sections of this chapter I provided an account of social norms, including their differentiation from conventions, and deployed that account to address the question of the moral obligation to obey the law. In the following chapter I turn to consider actions performed in the context of organisations.
159
5
Organisations, Agency, and Action
organisations Thus far in this book on social action we have considered in detail only single joint actions and regularities in action, especially conventional action and norm-governed action; we have not looked at structures of human action. In this chapter and the following one we will consider structures of human action, and especially organisational action and institutional action. In Chapter 1 distinctions were made between organisations, social groups and institutions. I now want to invoke these distinctions and elaborate on them. In this chapter I consider organisations. In the following chapter I offer detailed treatments of institutions and social groups. My central concern in this chapter is to argue that the “actions” of organisations are reducible to the actions of individual human persons. I do not accept that macro-entities, such as nation states and corporations, have beliefs and intentions, and consequently are either rational or moral agents. Properly speaking, all social actions are performed by individuals, not social entities. It should be noted that in advocating this form of individualism I am not committed to any of the following claims: (1) Social groups and/or organisations are reducible to the individuals who (at least in part) constitute them.1 (2) Social groups and/or organisations do not have causal powers. (3) Social groups and/or organisations cannot be blamed in some non-moral sense of blame for causing undesirable states of affairs, such as pollution of the environment. (4) It is never appropriate for individuals, courts of law, and so on, to treat organisations (in particular) as if they were rational or moral agents.
We have already distinguished between social groups on the one hand, and organisations on the other, and between organisations and institutions. But let us elaborate on these distinctions. 160
5. Organisations, Agency, and Action As we have seen, the defining elements of social groups are particular individuals standing in certain relations – including normative relations – to one another. The defining elements of organisations are embodied particular roles standing in relations to one another. That the roles are embodied means that they are filled by individual persons – though not by any individuals in particular.2 Accordingly, the University of Melbourne is the same organisation as it was fifty years ago, even though none of the present employees of the University of Melbourne was an employee fifty years ago. On the other hand, the University of Melbourne qua social group – the social group that consists of the individual human members of the University of Melbourne – is not the same social group as it was fifty years ago. This is because the membership of the earlier university is different from that of the current one. And there is a further defining characteristic of organisations. Individuals realise their ends not only by performing joint actions, including repetitive joint actions, but also by specialisation. Assume agent A performs task x, and agent B task y, and agent C task z. Assume also that: A cannot y or z (or at least cannot y or z without difficulty); B cannot x or z; and C cannot x or y. Assume finally that if A dies or leaves, B and C will identify some D to replace A; similarly, if B or C leaves or dies, then some E or F will be found as a replacement. What we have is an organisation, albeit a primitive one. So organisations consist of an (embodied) formal structure of interlocking roles.3 And these roles can be defined in terms of specialised tasks governed by procedures and conventions. Moreover, unlike social groups, organisations are individuated by the kind of tasks that their members undertake, and also by their characteristic functions or ends. So we have governments, universities, business corporations, armies, and so on. Perhaps governments have as an end or goal the ordering and leading of societies, universities the end of discovering and disseminating knowledge, and so on. Most societies at most times have made use of, and been comprised in part of, organisations. Moreover, the structure of organisations has varied enormously. Some are extremely hierarchical with an emphasis on controlling individual behaviour and attitudes. Military organisations have traditionally been of this kind. It is often claimed that Japanese organisations, including corporations and 161
Social Action government departments, are also of this sort, though with the qualification that employees are looked after and treated well as long as they conform to prevailing conventions and norms, and obey their superiors. Other organisations, such as Western universities, have been more collegial in character. It is important to note that on my (stipulative and non-systematic) definition of organisations they are, qua organisations, nonnormative entities. In this respect they are analogous to conventions, as I have defined conventions. In short, I reject the proposition that being an organisation is of itself something that is either morally good or bad, any more than being a convention is of itself morally good or bad. (I argued this point in relation to joint actions – and therefore, in effect, in relation to collective ends – in Chapter 2, and in relation to conventions in Chapter 3.) I believe that I can consistently hold this conclusion while maintaining that organisations, as well as conventions, are a pervasive and necessary feature of human life, being indispensable instruments for realising collective ends. Collective ends are a species of individual end; but an end, merely by being an end, is not in itself either morally good or morally bad, any more than an intention, by being an intention, and a belief, by being a belief, are in themselves morally good or morally bad. While my (somewhat prescriptive) definition of an organisation does not include any reference to a normative dimension, most organisations do in fact possess a normative dimension. As was the case with conventions, this normative dimension will be possessed by virtue of the particular ends that an organisation serves, as well as the particular activities that it undertakes. Further, most organisations possess a normative dimension by virtue (in part) of the norms governing the constitutive organisational roles. More specifically, most organisations consist of a hierarchical role structure in which the tasks and procedures that define the individual roles are governed by norms. It is not simply that an employee in fact undertakes a particular set of tasks or tends to comply with the directives of his employer. Rather, the employee undertakes those tasks and obeys the directives of his employer by virtue of the social norms governing the employee’s and employer’s roles. The employee believes that someone in his position ought to perform the tasks in question and believes that the employer has a 162
5. Organisations, Agency, and Action right to issue directives to his employees. Moreover, this belief is shared and acted upon by the other employees (and employers) and is held by each conditionally on its being held by the others. In short, the behaviour of the employees and employers is governed by social norms as I have defined these in Chapter 4. It should also be noted that the social norms governing the roles and role structures of organisations are typically enshrined in explicit rules, regulations, and laws, including laws of contract. For example, an employee not only believes that he ought to undertake certain tasks and not others, but these tasks are explicitly set forth in his contract of employment. Aside from an organisation’s defining (and often rule and lawgoverned) tasks and procedures, there is typically – though not necessarily – an important informal dimension of an organisation roughly describable as organisational culture. Organisational culture – as distinct from culture in other senses – is the ethos or “spirit” that pervades an organisation. Culture in this sense comprises many of the attitudes prevalent among the members of the organisation. Moreover, culture in this sense can powerfully influence the manner (at least) in which the members of the organisation undertake their various roles. So while the job description may say nothing about being “secretive” or “sticking by one’s mates come what may” or having a hostile or negative attitude to particular social groups, these attitudes and practices may in fact be pervasive; they may be part of the culture. In extreme cases, such as some modern police services, the official doctrine, including formal regulations of the organisation, may explicitly proscribe attitudes and practices that are in fact pervasive in the organisation by virtue of being constitutive elements of the organisation’s culture. Needless to say, such organisations are dysfunctional owing to the conflict between organisational goals and structure on the one hand and organisational culture on the other.4 Institutions in our sense are not the same things as organisations. An institution is (by stipulative definition) at the very least a structure of conventions and social norms that regulates some generic form of joint activity in accordance with more or less discernible ends. Examples of generic joint activity are communication, economic activity, spiritual activity, and sexual activity. Communicative activity is not necessarily institutional activity, but typically linguistic activity is. Exchanging furs for knives is economic activity, but not 163
Social Action necessarily institutional activity. Spiritual experiences are evidently not necessarily institutional phenomena, but religious activity is typically institutional in character. Sexual activity is not necessarily institutional in character, but within a marriage it normally is. Since by definition institutions, but not organisations, are normative in character, organisations are not necessarily institutions. Moreover, institutions are not necessarily organisations. Consider in this connection a human language such as English. The English language is an institution, but it is not an organisation. So much for the distinctions between organisations on the one hand and social groups and social institutions on the other. Let us now turn to the “actions” performed by organisations. The “actions” performed by organisations I will term “corporate actions.” In the next two sections of this chapter I will argue that properly speaking there are no corporate actions in the sense of actions performed by corporate agents possessed of beliefs and intentions above and beyond the beliefs and intentions of individual persons. Here it is important to note that the notion of a corporate action is not the same as that of a joint action. Joint actions are interdependent individual actions (see Chapter 2). For example, assume A is walking north along a narrow pathway and B is walking south, and they happen to see each other coming. A and B might more or less spontaneously keep to the left sufficiently to enable them to continue on their separate ways. This is joint action, but it is not corporate action. For, in this case, agents A and B do not constitute a corporate entity. In the next section I consider a particular conception of organisations and corporate actions for the purpose of rejecting it. In this conception organisations perform corporate actions in much the same way as individuals do, namely by virtue of their possession of mental states such as beliefs and especially intentions. Organisations, accordingly, are literally agents. The best-known advocate of this view is Peter French.5 While he is concerned with business corporations his position is one that, if true, would apply to all or most organisations.
164
5. Organisations, Agency, and Action corporate intentions French defines corporations in terms of their corporate internal decision (CID) structure.6 This structure consists of, first, hierarchically ordered positions with their attendant tasks, for example director, engineer, typist; second, procedures, for example the board of directors of Gulf Oil has the procedure of voting on whether to join a cartel; and, third, company policies, for example to maximise profit. I take it that there is a strong presumption against the ascription of mental states, such as intentions and beliefs, let alone full-blown moral agency, to institutions and organisations, including corporations.7 If we follow French in ascribing such sophisticated intentions to Gulf Oil as the intention to maximise profits, then we are apparently committed to ascribing to Gulf Oil a whole network of sophisticated propositional attitudes concerning economic production, the workings of markets, and so on. For, as recent work in the philosophy of mind has taught us, propositional attitudes, such as beliefs and intentions, cannot exist as stand alone entities, but only as interrelated components of networks of propositional attitudes. Thus a being that intends to maximise profits must have beliefs about what a profit is, and therefore what cost and revenue are (profits equal revenue minus costs), what a market is, what production, supply, demand, and so on, are. Moreover these propositional attitudes are not rudimentary ones directly tied to specific forms of behaviour or external sensory stimuli. They are not, for example, simple beliefs of the kind a dog comes to possess when it discovers that an object in its sensory field is in fact a bone. Rather, the relationship between these sophisticated beliefs is indirect, complex, and traversable only by reflective reasoning. Indeed, a being with such a network of propositional attitudes would be capable of high level thought, and therefore be possessed of a language in which to do this thinking. Moreover, this agent’s thought processes would include planning for its future and doing so on the basis of its past mistakes and the likely responses of other corporations. Thus, French states: “Simply to be a moral person is to be both an intentional actor and an entity with the capacity or ability to intentionally modify its behavioural patterns, habits, or modus operandi after it has learned that unto165
Social Action ward or valued events (defined in legal, moral, or even prudential terms) were caused by its past unintentional behaviour.”8 Such a corporate agent is self-reflective; it not only distinguishes its present from both its past and its future, and itself from other corporations, but it also reflects on itself for the purpose of transforming itself. Such a being has higher order propositional attitudes, including beliefs about its own beliefs and intentions, and conceives of itself as a unitary whole existing over time. It looks as though we now have a fully conscious, indeed self-conscious, being on our hands. Nor do matters rest here. For if we are prepared to grant Gulf Oil a mind, then why not all its subsidiaries, as well as all other companies and subsidiaries worldwide? Indeed, how can we stop at corporations? Surely governments, universities, schools, supermarkets, armies, banks, political parties, trade unions, English soccer team supporters’ clubs (at least) now all have minds, albeit in some cases smaller minds (so to speak). Not only do we have a self-conscious mind, but we have an ever-expanding community of self-conscious minds. It might be replied to this that I have made French’s conception seem more bizarre, or at least more exotic, than it really is by failing to draw attention to the fact that these corporate intentions – and therefore corporate minds – are dependent on individual human intentions and minds. But, according to French: “This is not an epiphenomenal theory nor is it a causal generation theory. I do not believe that corporate acts depend upon individual acts, if that is supposed to mean that the descriptions of the events in question that says [sic] they are individual human intentional acts are regarded as privileged.”9 In fact French seems to hold to a relation of interdependence between distinct but equally real “minds.” Thus: “The corporation’s only method of achieving its goals is the activation of the personnel who occupy its various positions.”10 On French’s view both corporate and individual intentions are equally real and neither is reducible to the other. Accordingly, on French’s view, we are threatened with a proliferation of fully fledged selfconscious minds. I also take it that there is a strong presumption against the ascription of moral agency, in particular, to corporations. For such ascription appears to have at least two untoward consequences. The first of these is that ascription implies a diminution in the responsibility 166
5. Organisations, Agency, and Action of individual human beings for the activities of corporations. It appears to license individual employees to hide behind company procedure and policy, rather than to take responsibility for the actions they perform in accordance with those procedures and policies. It is important to note here that holding a single entity, the corporation, morally responsible is not the same thing as holding a set of individual members of the corporation collectively (or, as I would prefer to say, jointly) morally responsible. Ascribing moral agency to the corporation introduces the practice of ascribing moral responsibility to organisations themselves, as distinct from ascribing it either individually or collectively to the human beings who constitute those organisations. It is also important to note that holding the corporation morally responsible is not the same thing as holding it legally responsible as a device to ensure internal disciplinary measures are imposed by management to curtail the wrongdoing of individual members of the corporation. Second, the ascription of moral agency to corporations has the consequence that corporations do not simply exist for the benefit of individual human beings; corporations have moral value in themselves. Accordingly, once we have determined that a particular corporation no longer serves the interest of shareholders, managers, workers, consumers, and the community of human beings in general, we cannot simply dissolve it. The corporation allegedly has interests and, in virtue of being a moral agent, moral value, above and beyond those of individual human beings. Says French: “It is important however that the converse of that popular slogan is conceptually erroneous. The converse expresses the proposition that the firm’s good is but the sum of the goods of those associated with it. The good of the firm, on such an account, is additive or cumulative, and of course dependent. The firm is supposed to exist only to further the aggregate goods of those associated with it.”11 Consider the case of a corporation that is producing goods no longer wanted, that has shareholders who have other much more lucrative investment opportunities, and managers and workers who have other better paid more rewarding jobs to go to. On French’s conception this corporation is a moral agent with interests above and beyond any of these individual interests. Accordingly, these shareholders, managers, workers, and consumers are faced with a moral dilemma. But surely they are not under even a prima facie 167
Social Action moral obligation to buy unwanted goods, work unwanted jobs, and forgo lucrative investment opportunities merely in order to preserve the corporation. I have suggested that there is a strong presumption against ascribing either rational or moral agency to corporations. But let me now turn to the details of French’s conception and look at some of the arguments adduced in favour of ascribing mental states to corporations. As will become evident, it is somewhat unclear what, precisely, French’s conception is. According to French, when a decision of, say, the company directors is made in accordance with company procedures and is consistent with company policy, then there is a corporate intention. French considers the actions of executives of Gulf Oil. He says: “The event of those executives voting [to join the cartel] is redescribable as . . . ‘Gulf Oil Corporation intentionally decided to join the cartel.’ ”12 There seem to be two different ways of taking French at this point. Perhaps every decision made by an employee in the course of performing his or her job is also an instance of a corporate intention so long as that decision is in accordance with company procedures and policies.13 On this construal when one of the financial advisers (salespeople) for Guardian Life answers the phone, saying “This is Guardian Life,” and completes his or her conversation by saying “Have a nice day,” the corporate mind has intentionally communicated with someone, or at least has intended that its instrument (the salesperson) communicate what he or she communicated. After all it is company policy and standard procedure in the sales division of the corporation to use these expressions and to mention the name of the company as often as possible. This way of construing corporate intentions has untoward consequences. For now most, or very many, of the millions of intentional actions performed by shareholders, managers, and workers in the course of their work have a kind of shadow corporate intention. In that case – given that intentions cause their actions – every individual employee’s action is causally overdetermined. But surely employees’ intentions are not redundant with respect to their own actions. And if they are redundant, in what sense are any of the employees’ actions their own? Moreover, each of these millions of intentions presumably pre168
5. Organisations, Agency, and Action supposes a network of other propositional attitudes, including other intentions, beliefs, and so on. Surely any theory that implies the existence of millions of such shadow propositional attitudes is highly implausible.14 I conclude that the consequences of accepting this construal of French’s theory are very unpalatable indeed. On a second construal the corporate mind is to be located in the vicinity of the control centre of the corporation; corporate intentions do not lie behind the actions of any or all of a corporation’s employees, but rather behind those of the managers, or at least behind the important actions of the managers. Of course the fact that an important decision is made by a manager does not entail a corporate intention. The CID structure is necessary for corporate intent, and managers can fail to follow procedures and policies. However, if an important decision is made by a manager and it is in accordance with corporate procedures and policies then there is a corporate intention. Moreover, on this second construal, it is only such decisions that involve corporate intentions. At times French seems to be advocating this view: “What matters is . . . that it has some operative decision in place that functions as an incorporator of the decisions of corporate managers.”15 This second way of taking French will not accommodate cases in which lower level employees act in accordance with company procedures, but the procedures themselves are flawed. Perhaps a design fault in a car arises because of inadequate checking procedures in the lower echelons of a corporation. Such a case involves a defect in the CID structure itself, but it does not involve any action or omission by a manager. Therefore, on the construal before us, it does not involve corporate intention or negligence.16 So, on this construal of French, in such cases we have not corporate, but individual, fault. An example of such a case might be one similar to that involving Ford’s Pinto car. Imagine that due to a defective checking procedure a particular make of car has a design fault that causes the car to explode when involved in even minor crashes. We could, of course, hold the managers morally responsible for any deficiencies in the CID structure. However, such a move is not open to French. For this move locates corporate fault in the actions and minds of individual humans, namely the managers. But it is precisely the point of French’s conception to ascribe full rational 169
Social Action and moral agency to corporations via their CID structure. If the deficiencies in a CID structure could be laid at the feet of managers, mere individual human beings, French’s conception would be undermined. French himself invokes higher order procedures and policies to deal with deficiencies in lower order procedures and policies.17 Thus, in the Mt. Erebus Air New Zealand disaster, a particular administrative procedure used in the organisation was held by the Royal Commission of Inquiry to be the main cause. Subsequent to the disaster Air New Zealand refused to change this procedure. Presumably French believes that the corporation, Air New Zealand, did not have a higher order policy of changing lower order procedures when they were revealed to be deficient. (The alternative is to hold that the particular managers of the day failed to adhere to the higher order policy of remedying deficiencies in lower order procedures. But by French’s argument this would not demonstrate corporate fault.) Therefore we have corporate, as distinct from merely managerial, fault. The problem with this move is that it simply pushes the problem back one step further. Who is responsible for the non-existence of higher order policies to correct deficient lower order procedures? It is hard to see – given the risks to life involved in air transport – how anyone but individual humans, and especially managers, could be primarily responsible for the non-existence of such a policy. Certainly corporate intentions or negligence as manifest in the CID structure cannot be invoked. For according to the view before us corporate intention or negligence must be manifest in some appropriate fragment of the corporation’s CID structure. But in the case of a corporation without any policy or procedures with respect to deficient existing procedures and policies, there simply is no existent higher order fragment of its CID structure for such intentions or negligence to be manifested in. The corporation qua embodied CID structure does not have the wherewithal to remedy deficient lower order policies and procedures, and therefore cannot be held either intentionally or negligently to have failed to remedy them. Moreover, even if there were such higher order corporate policies and procedures to remedy lower order deficient procedures, there would still arise the question of the remedying of these higher order policies and procedures, should they themselves turn out to have 170
5. Organisations, Agency, and Action deficiencies. In relation to a hierarchy of deficient procedures and policies of the sort involved in the Air New Zealand crash, it seems the buck must stop at individuals. On a third and final construal of French he is to be understood as claiming that at least in the case of certain procedures we must postulate the existence of corporate intentions. The idea is that certain procedures involving the joint action of agents give rise to corporate intentions in their very application. Here French has drawn attention to a particular species of procedure, one involving joint, as distinct from individual, action. To that extent his argument, even if sound, is relatively weak. At any rate even this argument is unacceptable. Suppose, to take one of French’s examples, the board of directors of Gulf Oil has the procedure of voting on important issues, with the majority vote to carry the day.18 At times there is not consensus. In such cases it can be said that Gulf Oil voted for a particular course of action, but it is not true that all the individuals voted for that outcome. Therefore there is a corporate action – and a corresponding corporate intention – that is not reducible to the individual intentions. In the following section of this chapter I provide an account of procedures such as voting, acting through representatives, deploying a hierarchical decision structure, jointly making specialist contributions to a complex collective good, and so on.19 I term these procedures joint mechanisms, or institutional mechanisms. The important point for our purposes here is that the notion of a joint mechanism involves no essential recourse to supra-individual minds, intentions, or beliefs. So even this final argument of French fails. I conclude that French’s arguments for the existence of corporate intentions, and therefore for corporate moral agents, fail. Let us now turn to my own positive account of corporate actions. corporate action Corporate actions are the actions of corporate entities, including organisations. Corporate entities consist in part of roles filled by individuals. These roles are related to one another in such a way as to constitute a structure of roles. Further, this embodied structure of roles exists to serve some collective end or ends. Thus the army 171
Social Action consists of officers, foot soldiers, and so on, and the whole structure of roles dictates actions performed for the purpose of winning wars. Most corporate entities consist at any given time of individuals who constitute a social group. But it is the individuals qua occupiers of the roles that constitutes the corporate entity. I suggest that a role is simply an abstraction from proceduregoverned tasks, and the actions that constitute the undertaking of those tasks. Qua abstraction, the role is simply a specification of the task or tasks, and the procedures that govern it. To say that the role is filled is just to say that some individual is undertaking that task or tasks, and following those procedures. By undertaking a task I simply mean performing a relatively complex set of differentiated actions directed at some given end. The notion of undertaking a task is admittedly somewhat vague, but this is inevitable. The main point is that to perform a task is to do more than perform a discrete action, or repeatedly perform an action. The notion of a procedure in use here is the one I defined in Chapter 3. Roughly speaking, someone has adopted a procedure if he or she regularly performs a given action type in certain specified circumstances. Accordingly, my notion of a procedure is relatively wide and informal. Procedures can be private or public, implicit or explicit and they may be enforced or freely chosen. Since procedures are in effect regularities in action, it follows that proceduregoverned tasks are regularities in the performance of a task. Accordingly, the occupant of a role is engaged in a task that he or she undertakes repeatedly. To engage in a task on one occasion only is not to occupy a role. These procedure-governed tasks are typically associated with various attitudes, such as beliefs, desires, and values. I say these attitudes are associated with the roles rather than constitutive of them since it is possible to occupy the role without adopting the attitudes. Of course, there are some beliefs and intentions that are constitutive of the role, but these are simply those beliefs and intentions that are constitutive of the actions one performs in occupying the role. On this conception of a role, chess playing is a role and so is woodchopping. However if one scratches one’s head when it is itchy one has not adopted the role of a headscratcher. For scratching, while it is repeatable action, is not sufficiently complex to count as a task in my sense. Again, someone in jail who plans and executes a 172
5. Organisations, Agency, and Action complex series of actions to escape is involved in the activity of escaping from prison but has not adopted the role of prison escapee. For the task is not procedure-governed and the task is a oneoff. On the other hand, if someone was continually escaping and being caught and developed a set of procedures for escaping, then we could come to think of that person as having the role of an escapee. On this account of corporate “action” we can distinguish between the actions constitutive of the performance of the task or tasks of the occupant of the role, on the one hand, and the “action” of the corporate entity that is constituted by the roles, on the other hand. Corporate “action” is not merely the individual action of occupants of roles. It is the action of the corporate entity per se. Now, some alleged actions of corporate entities are in reality the joint actions of the individual persons who comprise the corporate entity.20 Consider, for example, the claim that the army (corporate entity) won the battle. Many such cases are, I suggest, reducible to the joint actions of individual human beings. layered structures of joint action At this point the notion of what I will term a layered structure of joint actions needs to be introduced. Suppose a number of “actions” are performed to realise some collective end. Call these actions level two actions. Suppose, in addition, that each of these level two actions is itself a joint action whose component actions are individual actions directed to a collective end. Call these component actions, level one actions. So the realisation of a collective end of the level one actions is the performance of the level two action. One illustration of the notion of a layered structure of joint actions is an army fighting a battle. The individual members of the mortar squad jointly operate the mortar in the service of the collective end of reducing the ranks of the enemy soldiers. The set of tanks jointly move forward firing in the service of the collective end of destroying the enemy’s gun emplacements. Finally, the set of foot soldiers jointly position themselves to realise the collective end of holding the ground vacated by the retreating enemy force. The actions of each of the individual foot soldiers, mortar squad members, and individual members of the tank team are level one actions. More173
Social Action over, the actions of the individual foot soldiers taken together constitute a level one joint action; as do the actions of the individual members of the mortar squad, and the actions of the individual members of the tank crews. However, the infantry and the cavalry (the tanks) and the mortar squad jointly act to realise the collective end of defeating the enemy. So each of these joint actions constitutes a higher level “individual” action performed to realise a further collective end, namely defeat of the enemy. Each of these actions is a level two action. Moreover, the set of these level two actions, taken in conjunction with the collective end to which they are directed, constitutes a level two joint action. I suggest that the claim that the army won the battle describes a layered structure of joint actions. If so, then there is no need for recourse to “actions” other than the joint actions of individual human beings. For the level one actions are joint actions of individual human beings; and the level two actions are also the joint actions of individual human beings, albeit joint actions that are logically dependent on the level one joint actions. Let me now turn to corporate action that is not in this way reducible to the joint actions of individuals. There are two kinds of such corporate action. There are individual corporate actions and there are joint actions performed by corporations. A joint action is an action performed by two or more corporate entities in order to realise a collective end. As such it is logically dependent (so to speak) on individual corporate action. Let me then turn to individual corporate action. joint mechanisms Such individual corporate action is the result of what I will call a joint mechanism. Accordingly, I maintain, the fundamental notion required to understand (individual) corporate action is the technical notion of a joint mechanism. Let me now turn to the characterisation of the notion of a joint mechanism. In so doing I will acquit my case introduced at the end of the previous section of this chapter against French’s final argument for corporate intentions. Suppose we know that in some area there are bound to be disagreements not resolvable by recourse to rational argument to a point of consensus. In this area we might decide to adopt proce174
5. Organisations, Agency, and Action dures to resolve disputes. Thus, voting, tossing a coin, courts of law, governments, and so on are procedures in this sense. These are not in themselves procedures of the sort mentioned above. We might toss a coin to resolve a dispute on one occasion only. And there are still other kinds of cases that are not procedures. We might use a particular agreed upon signal in one instance of a non-recurring situation. It is thus application of a procedure in a different sense of a procedure from that discussed above; yet it is not an attempt to solve a disagreement or dispute. The application of these new kinds of procedures, even on a single occasion only, is action performed to realise a collective end. Let us call such procedures as these joint mechanisms. Now in some cases, that these joint mechanisms are used might be a matter of having a procedure in my earlier sense. Thus, if we decided that (within some specified range of disputes) we would always have recourse to tossing a coin, then we would have adopted a procedure in my earlier sense. Accordingly, I will call such joint mechanisms, joint procedural mechanisms. Joint mechanisms (and, therefore, joint procedural mechanisms) consist of (a) a complex of differentiated but interlocking actions; and (b) the result of the performance of those actions. Thus a given agent might vote for a candidate. He or she will do so only if others also vote. But further to this, there is the action of the candidates; namely that they present themselves as candidates. That they present themselves as candidates is (in part) constitutive of the voting mechanism. Voters vote for candidates. So there is interlocking and differentiated action. Further, there is some result (as opposed to consequence) of the joint action, the joint action consisting of the actions of putting oneself forward as a candidate and of the actions of voting. The result is that a candidate is voted in. That there is a result is (in part) constitutive of the mechanism. That to receive the most number of votes is to be voted in is (in part) constitutive of the voting mechanism. (This is not to say that the voting mechanism might fail on occasion; but that is a different matter. It also needs to be noted that this kind of result which is, in part, constitutive of a mechanism, is not as such the collective end of the mechanism; though, of course, joint mechanisms do have collective ends.) The complex of joint action and its result is, qua application of the mechanism, in principle repeatable. But the complex of actions 175
Social Action performed in any given application (and/or its result) must be able to differ from one occasion of application to the next. Joint mechanisms are not simply (complex) joint procedures. Rather, the action to be performed in the application of the mechanism is specified in such a way as to enable it to consist of different possible actions on different occasions of application. Or at least, the action is specified in such a way as to enable it to have a different result from one occasion to the next. I am speaking here of the result, which is (in part) constitutive of the mechanism. Thus one must vote for a candidate, but which particular candidate is not specified. Or one must toss a coin, but it is not known in advance whether the result will be heads or tails. Moreover, this characteristic of delivering different actions or results on different occasions of application is a characteristic that can exist in virtue of different features of joint mechanisms. We have just seen, in the case of the voting mechanism, that one such feature is the capacity for differing actions to be performed on different occasions in conformity to the mechanism. We have also just seen, in the case of the mechanism of tossing the coin, that agents performing the same action on every occasion can deliver a different result from one occasion to the next, and do so in virtue of an internal property of the mechanism. (In the case of flipping a coin, the internal property is the equal probability of a tossed coin coming down heads or tails.) However there are further kinds of cases. One of these is the capacity of the mechanism to deliver a different result in virtue, not of different action input or of an internal property of the mechanism, but rather of some feature of the situation in which the mechanism is applied. Thus, an utterance of the term “the bank” might produce a different result in one context (a conversation about money) from the one it produces in another (a conversation about the Hawkesbury River). Moreover, it may be that there are certain linguistic mechanisms especially, which are such that the context always makes a difference to the result (whether by disambiguating or in some other way). Another kind of case involves what I will call resultant actions.21 Thus, in rugby, a scrumhalf might have the procedures of tapping his left knee, if the lineout ball is to be thrown to the front of the lineout; tapping his right knee, if it is to be thrown to the back; 176
5. Organisations, Agency, and Action tapping a knee once, if it is to be lobbed high; and tapping a knee twice, if it is to be thrown fast and low. The action of tapping his left knee twice means that the ball is to be thrown fast and low, and to the front of the lineout. However, this is a resultant action, since it is derivable from the two basic procedures of tapping a knee twice and tapping the left knee. More precisely, it is the application of a certain sort of joint procedural mechanism. The action of tapping the left knee twice is not in itself a procedure, since it may only be performed once, or indeed never. So the action of tapping the left knee twice is not a resultant procedure, rather it is a particular application of the mechanism. The mechanism consists of: (a) the scrumhalf’s two interlocking and differentiated procedures – the procedure of either tapping his left knee or tapping his right knee, and the procedure of either tapping his knee once or tapping his knee twice; (b) the knowledge of the members of the lineout, and of the thrower of the ball, of these procedures; and (c) the result of both the application of a procedure, and of knowledge of that procedure, namely that the ball is thrown, say, low and to the front of the lineout, and that someone at the front of the lineout jumps in a certain way. Notice that this mechanism enables different actions to be performed (and different results to ensue) from one occasion of the application of the mechanism to the next. This possibility (or these possibilities) arises in virtue of two features of the mechanism. First, the mechanism involves disjunctive procedures. The scrumhalf can choose between two actions per procedure. This is one reason an application of the mechanism varies from one occasion to the next. Second, the procedures interlock so as to produce a single complex action, which in fact may not have been performed before, even though each of the procedures has been followed before. Thus the scrumhalf may only once tap his left knee once, although he has many times tapped his left knee, and many times tapped a knee once. This is another way in which applications of the same mechanism deliver different actions from one occasion to the next. Language appears to consist in part of joint mechanisms involving resultant actions.22 Assume that there are the following joint procedures in a community: Utter “John” when you have, as an end, reference to John; “Fred” when you have, as an end, reference to “Fred”; utter “is a man” when you have, as an end, ascription of the 177
Social Action property of being a man. Then there might be the resultant joint action to utter “Fred is a man” when you have, as an end, ascription to Fred of the property of manhood. So joint mechanisms and joint procedural mechanisms are such that their action inputs, and/or their results, may differ from one occasion of application to the next. This creates the possibility of individual choice, and indeed uniqueness, of resultant action. For example, the scrumhalf, although he is constrained in the sense that he has to follow the procedures, nevertheless he can choose between disjuncts of the procedures, and thus choose his resultant action. It also means that on occasion joint mechanisms and joint procedural mechanisms deliver outcomes other than those desired by at least some of the participants. These two characteristics of joint mechanisms have the implication that in one sense the input of joint mechanisms is not a joint action, and the output is not a collective end. In particular, under the description, means to achieving my individual end, the action is not a joint action, and the end is not a collective end. Thus if I flip a coin in order to win some money, my action, qua means to winning the money, is not joint, and my end is not collective. However, in another sense it is. Under the description, application of the mechanism, the action of each of the individuals is a joint action. And under the description, output of the application of the mechanism, the output is a collective end. Thus, if I flip a coin having as an end that someone (you or I) wins the money, then my action of flipping the coin, taken in conjunction with your action of calling “heads” (or “tails”), is joint action and my end is a collective end.23 Finally, let me say that, as with joint procedures, there are relations of interlocking and of presupposition between joint mechanisms. Perhaps the joint mechanisms of government and of the law interlock. The system of courts and of the law is a joint mechanism whereby the actions of members of a community are regulated and coordinated, the joint mechanism of the government is a joint mechanism that enables a community to be governed. But the roles of government and of the law interlock, the one being scarcely possible without the other. Voting in a democracy might be a presupposed joint procedural mechanism. The presupposing procedure is to obey laws promulgated by the elected government. Now the law is a mechanism in 178
5. Organisations, Agency, and Action our above sense. It is a device for coordinating actions and achieving an orderly society. However, if the citizens would only obey the laws passed by a democratically elected government, then the procedure of obeying the law presupposes the joint procedural mechanism of voting. quasi-joint action Thus far I have argued that the “actions” of organisations can be conceived of as various species of joint action, including layered structures of joint action and action in accordance with joint mechanisms. However, there is a different kind of problem facing my analysis of the “actions” of organisations in terms of various species of joint actions. This problem stems from the excessively individualistic character of many of the actions of individual members of organisations. Suppose, for example, that workers in a car factory only work for their pay packets. They are not interested in whether or not any cars are produced. Certainly, they perform whatever tasks they are paid to perform, but they care nothing about the outcomes of the combined efforts of themselves and their workmates. Evidently, there are no collective ends, and therefore no joint actions. Or at least there are no collective ends or joint actions in the lower echelons of the organisation; the upper echelons would presumably have collective ends and engage in joint actions. Of course it might be that no one in the organisation possesses collective ends or engages in joint action. Rather, all members of the organisation, whether in the upper or the lower echelons, pursue only individual ends. I suggest that such an organisation would come to grief very quickly indeed. Consider a car factory in which no one had as a collective end to produce cars. Could anyone have the production of cars as an individual end? Certainly not; for no one could achieve that end on his or her own. It follows that no one in the car factory has the production of cars as an end. Surely, in that case it would be unlikely, or at best fortuitous, if in general cars were produced in that car factory. At any rate, in the situation described each worker has an individual end only, namely to get his or her pay packet, and each intentionally works to achieve that individual end. However, we can imagine two alternative scenarios. In the first scenario each worker has 179
Social Action the production of cars as a collective end. However, the realisation of this collective end is itself a means to an individual end, namely getting a pay packet. In this scenario the workers are engaged in joint action, albeit as a means to securing individual ends. In the second alternative scenario each worker has the production of cars as a collective end, and in addition has, as an individual end, getting a pay packet. However, each worker knows that he or she will be paid, even if cars are not produced; so realising the collective end is not a means to realising the individual end. On the other hand, in this scenario, each would not have the collective end unless he or she each had the individual end. For no one would be prepared to work if he or she were not paid. This is a case of what we termed in Chapter 2 quasi-joint action. The collective end is not a means to an individual end, nor is the collective end sustainable without the individual end. Rather, the existence of the collective end depends in part on the existence of the individual end. I suggest that most organisations involve for the most part quasi-joint action. However, I grant that this is an empirical question. In this chapter my main purpose has been the negative one of arguing that the “doings” of macro-entities, and especially organisations, are reducible to the actions of individual persons. I have also suggested that many alleged corporate actions are in reality either reducible to layered structures of joint actions or understandable in terms of my notion of a joint mechanism. I have further suggested that the joint actions of members of organisations are probably more often than not instances of quasi-joint action, rather than fullblown joint action. In the next chapter I turn to a consideration of what is arguably the most important kind of structured human action, namely actions governed by social institutions. In doing so, the relationship between the organisational, institutional, and social group action will be displayed.
180
6
Social Institutions and Social Groups In Chapter 5 distinctions were made (especially) between social institutions, social groups, and organisations, and a detailed account of organisations, and especially the “actions” of organisations provided. I now want to look in more detail at social institutions and social groups. We begin with social institutions. In order to understand social institutions we need to first understand the notion of a sphere of activity, which in turn depends on the notion of generic joint activity. Roughly speaking, a joint activity is a complex of differential, interlocking, joint actions and individual actions directed to some overarching collective end. So building cathedrals and sailing ocean liners are joint activities. We can further distinguish between different, albeit connected and overlapping, generic kinds of joint activity, including communicative, economic, educative, sexual, and religious activity. The repetition over time, and duplication in space, of any one of these different kinds of generic joint activity can give rise to a more or less connected, and more or less continuous, stretch of joint activity, which I earlier (in Chapter 1) termed a sphere of activity. So the ongoing series of economic transactions across Australia constitutes a sphere of activity. Spheres of activity (by stipulative definition) are regulated by conventions. They thereby take on different specific forms according to the specific conventions that structure them. Often, though not necessarily, they are also regulated by explicit rules, including laws. And they are also often – but again, not necessarily – regulated by social norms. As indicated in the first section of Chapter 5, an institution is (at least) a structure of conventions and social norms that regulates some form of generic joint activity in accordance with more or less discernible ends. Communicative activity is not necessarily institutional activity, but typically linguistic activity is. Spiritual experiences 181
Social Action are evidently not necessarily institutional phenomena, but religious activity – including non-organisational religious activity – is typically institutional in character. Extra-marital sexual activity is not necessarily institutional in character, but marital sex is. Periodic bartering between explorers and members of local tribes is non-institutional economic activity, whereas Finland’s export of paper and import of cars within the framework of EU guidelines are institutional activities. Social institutions are (by stipulative definition) normative entities, defined in part in terms of normatively described ends, and in part in terms of activities governed by social norms. So social institutions are a species of social entity that is logically dependent on a certain species of social action. Further, by virtue of their inherent normative character institutions are different from organisations. An institution, but not an organisation, by definition reproduces itself (or is disposed to do so). An institution reproduces itself because its members strongly identify with its ends and social norms, and induct the next generation into the institution. Accordingly, members of an institution, but not necessarily of an organisation that is not also an institution, typically share a common history over a number of generations. Notwithstanding these differences between institutions and organisations, there are important relations between many institutions and organisations. First, as we saw above, many institutions are also organisations, or have organisational elements. Second, organisations exist, or at least ought to exist, to facilitate institutional purposes and, ideally, morally desirable institutional purposes. So much by way of offering a preliminary characterisation of social institutions. The currently most influential philosophical account of social institutions is that propounded by John Searle. In the next section, Searle’s account is examined and in large part rejected. In the section following I offer my own account of social institutions. In the third section I consider social groups. searle on institutions According to Searle, institutional facts are ultimately physical objects, or states of affairs, or events, upon which what he terms statusfunctions have been collectively imposed.1 I say “ultimately” because 182
6. Social Institutions and Social Groups although status-functions can be collectively imposed on pre-existing institutional facts, any such iterated structure of status-functions must terminate at some point in physical objects or events (or more precisely, must terminate in what he terms brute facts).2 Something has a status-function if it has, or those who use it have, deontic powers. Thus a police officer has a status-function, and therefore a set of deontic powers, including rights to stop, search, and arrest people under certain conditions. A five-dollar bill is a piece of paper (a physical object) the bearer of which has various deontic powers, including the right to exchange the bill for goods to the value of five dollars. These status-functions, and therefore deontic powers, have been collectively imposed in the sense that the relevant members of a community accept or agree to or otherwise treat the objects or persons that possess these status-powers as if they do in fact possess them. But in accepting or so treating, for example, the police officer as if he has the right to arrest people, the police officer comes to have that right. According to Searle’s argument, if no one ever paid any attention to police officers they would cease to have any deontic powers and therefore any status-function; indeed they would cease to be police officers. Similarly, if no one was prepared to exchange five-dollar bills for goods, then these bits of paper would cease to have any status-function, and the bearers of them would cease to have any deontic powers. Searle’s account of institutions makes use of four basic notions: (1) (2) (3) (4)
imposition of functions; the deontic powers of institutional persons and objects; the distinction between constitutive and regulative rules; collective intentionality.3
Let us look more closely at these notions, beginning with functions and deontic powers. Functions and Deontic Powers Searle’s notion of function concerns what it is that a group collectively imposes on a physical phenomenon. For example, if members of a community began to sit on a log then the log would in effect have become a bench. So a function – that of being used to sit on – would have been collectively imposed on a physical object. On Searle’s conception all functions ultimately depend on the existence of physical objects on which functions are imposed. How183
Social Action ever, some functions – such as the function of being a chair – depend on the specific physical properties of the object on which they are imposed. Thus a log can become a bench only if it has a certain size and shape. By contrast, the key feature of institutional facts is that they involve functions that cannot be imposed simply by virtue of the specific physical properties of the phenomena on which they are imposed. Rather, possession of the function exists by virtue of the collective character of the imposition.4 So being a chair is not an institutional fact; rather, its functionality exists by virtue of its specific physical properties. On the other hand, being a medium of exchange is an institutional fact; its functionality supposedly exists by virtue of collective imposition rather than specific physical properties. It should be noted that, according to Searle, functions can be individually imposed on physical objects by individuals. This would be the case if, for example, Robinson Crusoe alone used the log as a bench. Further, if a log had the function of a bench collectively imposed on it, then its being a bench would be a social fact.5 However, the fact that the log was a bench would not be an institutional fact; for its being a bench does not depend on the collective character of the imposition of this function. According to Searle, in the case of institutional facts the imposition of a function consists in the assignment not only of a function, but also of a status with deontic properties.6 These deontic properties – rights and duties and the like – are regarded by Searle as powers. If a police officer has a right to arrest you, then he has a power to do so; if you have a right to goods worth five dollars in exchange for your five-dollar bill, then you have a power to receive those goods. And, to reiterate, these status-functions or powers are possessed by virtue of the agreement or acceptance of the community; the powers are collectively imposed and maintained.7 The tight connection that Searle makes between function imposition and deontic properties is problematic. Searle asserts that the notion of a function necessarily brings with it values and deontology. Thus if the purpose of the heart is to pump blood then it ought to pump blood.8 No doubt purposes and ends can be assessed in the sense that they can be said to have been realised or to have failed to be realised. Moreover, if one has an end then, other things being equal, one ought to enact the means to that end. So functions, goals, 184
6. Social Institutions and Social Groups and the like bring with them normativity in the minimal sense described in Chapter 1, namely that pertaining to instrumental rationality. But the claim that functions and purposes bring with them normativity in any stronger sense has not been shown. Specifically, Searle has not shown that his deontic properties, such as rights and duties, can be derived from the notion of a function. Nor has he shown that deontic properties can be derived from his notion of function taken in conjunction with collective intentionality. A more plausible source of some deontic properties is rules – in my above defined sense of that term, which is explicitly formulated directives issued by authorities to perform certain kinds of actions under certain circumstances. However, Searle claims that “rules” in the very general sense of general policies, are capable of abuse, and hence have a normative status.9 As I made clear in Chapter 1, I do not accept the proposition that general policies or habits necessarily generate standards of conduct – as opposed to mere regularities in behaviour – and therefore deontic properties. Certainly Searle offers no demonstration of his claim. At any rate I suggest that this non sequitur – moving from individual or collective imposition of functional properties to deontic properties – infects Searle’s account of status-functions. For it is perfectly possible for a person or an object to have a set of functions – and be treated as having those functions – without any associated (noninstrumental) deontology. For example, a set of individuals might use a certain sort of relatively rare shell as a medium of exchange, and do so notwithstanding the fact that no one had any desire to possess these shells independent of the fact that they could be used as a medium of exchange. In this scenario all that is required is that each exchanges shells for goods, and goods for shells, intends to continue to do so, and believes that all the others do and intend likewise. Of course it would add greatly to the stability of this arrangement if these shells were somehow authorised as an official medium of exchange, and if a (rule-constituted) system of rights and duties in relation to the exchange of these shells was introduced and enforced. However, such a deontological structure is not a necessary feature of the system of exchange.10 Moreover, even where a deontological framework of the kind described by Searle has been adopted, the relevant deontic powers do not subsume, take the place of, provide the basis for, or go hand 185
Social Action in hand with, the substantive functions of the objects or persons in question. Here by substantive functions I mean the functions that are central and necessary to the object or person being the kind of thing that it is. Consider an incompetent surgeon, who is incapable of performing a successful operation on anybody, and who largely avoids doing so, or, when he or she absolutely has to, always ensures that he or she is part of a team comprised of other, competent, surgeons and nurses who actually do the work. By virtue of being a fully accredited surgeon this person has a set of deontic powers, including the right to perform surgery, and others have deontic powers in relation to him or her, including the right that he or she performs operations competently and with due diligence. Moreover, these deontic powers are maintained in part by (say) the Royal College of Surgeons, his or her colleagues, and the community. However, the surgeon simply does not possess the substantive functions of a surgeon. The deontology is there, but the substantive functionality is not. Accordingly, it is simply false to claim, as Searle must do, that he is a surgeon.11 If someone cannot perform, and knows nothing about, surgery he or she cannot be a surgeon, irrespective of whether he or she has the highest professional qualification there is, is treated as if he or she were a surgeon, and indeed is widely believed to be the finest surgeon in the land. It might be thought that this kind of example suits my argument by virtue of the physical character of the role of surgeon. But my point can be made in relation to non-physical roles and objects. Consider a philosophical article published in a leading journal. This is an institutional fact. Now recall the notorious Sokal hoax widely distributed on the Internet in 1999. Sokal, a physicist, decided to write a piece of gibberish and submit it to a learned journal specialising in postmodernist writings. The journal accepted his piece and published it. So the piece of writing was institutionally certified as being a very high quality philosophical article. Moreover, many who read it believed this to be the case. In reality, however, it was nothing of the kind; it was gibberish and intentionally created by Sokal as such. So the “philosophical article” possessed the relevant collectively imposed deontic properties, and therefore on Searle’s account constituted the institutional object, philosophical article. However, this piece was not a philosophical article, for it lacked the intellectual properties necessary for it to be so. Accordingly, the 186
6. Social Institutions and Social Groups collective imposition of deontic properties was not sufficient to generate the relevant institutional fact. Institutional authorisation and collective intentionality cannot turn entirely incompetent surgeons into real surgeons, nor can they turn gibberish into real philosophical articles. I conclude that in relation to institutional facts Searle’s notion of deontic power – whether collectively imposed or not – does not necessarily subsume, take the place of, provide the basis for, or go hand in hand with, substantive functionality. Substantive functionality and deontic powers can come apart. But let us turn now to Searle’s notion of collective intentionality, and the related notions of social acceptance and social agreement. Collective Intentionality, Collective Acceptance, and Constitutive Rules Searle’s apparatus of status-functions is created and sustained by what he terms collective intentionality. As we have just seen, his view that the substantive functions of objects and persons are subsumed by, or necessarily go hand in hand with, deontic powers is incorrect. Moreover, it is now clear that collective intentionality cannot generate the substantive functions of many institutional persons and objects. What of the relationship between collective intentionality and deontic powers? Let us get a little clearer on the notion of collective intentionality. As explained in Chapter 2, collective intentionality is for Searle a primitive notion expressed by locutions such as “we-intend.” A weintention is not reducible to an individual intention, nor to an individual intention in conjunction with other individual attitudes such as individual beliefs.12 As will be evident from my discussion in Chapter 2 on joint actions – especially my critique of Searle – I do not accept Searle’s conception of primitive we-intentions. I take the view that an analysis is possible using only individualistic notions, including especially the notion of a collective end. In any case Searle’s notion of primitive we-intentionality is under-theorised. He has not provided an account of it, analytical or otherwise. But, that aside, what of the role of collective intentionality – however it is to be understood – in relation to the creation and maintenance of deontic powers? Consider the deontic powers associated with the utterance of sentences, such as the right of hearers 187
Social Action that speakers would aim at the truth. Searle must hold that such a right can only exist if a group adopts some sort of we-attitude, such as we-accept or we-agree or we-belief or we-intention, to these deontic powers. And presumably the same goes for other moral rights, such as the right to life, the right to freedom, and so on. Now prima facie hearers have a moral right that speakers aim at the truth, and a moral right to life, and so on, quite independently of whether the community we-intends that this be the case, or accepts that it is, or whatever. So the source of at least some central deontic properties, namely some moral properties, is evidently not collective intentionality. Naturally, whether or not deontic properties, including rights, are respected might ultimately turn on the attitude of the community. So general community “acceptance” of deontic properties, including rights – in a weak sense of acceptance – is necessary if a specific deontological framework is to take effect in that community. So much is trivially true.13 I conclude that Searle has not established that all, or even most, deontic properties associated with institutional objects and persons are (so to speak) ontologically dependent on collective intentionality.14 However, are there other cases of deontic properties – especially those involved in authority relations – which might be ontologically dependent on collective intentionality.15 Consider Peter Sellars in the movie Being There. Sellars plays the role of a gardener who for various reasons begins to be treated by the staff of the president, and ultimately by everyone, as if he were the president of the USA. Eventually, he is in a position to run for office and be elected. Unfortunately, he has no understanding of the political system, or of relevant policies, and has no leadership qualities whatsoever. Nevertheless, it seems to be the case that the gardener can become the president by virtue of collective acceptance. Institutional authorities are vulnerable to a degree that other institutional forms, such as languages, are not. As Searle points out, the communist government of Russia turned out to have clay feet.16 Once people chose not to obey its directives, it was finished; it simply ceased to function or exist as a government. However, it is difficult to see how the English language could go out of existence in such 188
6. Social Institutions and Social Groups spectacular fashion, for it depends on millions of often-disconnected communicative interactions between millions of different people. There is a reason for the vulnerability of institutional authorities. In the special case of institutional authorities, deontic properties are, as Searle holds, ontologically dependent on collective acceptance; no collective acceptance, no deontic properties. The point here is not simply that (say) rulers cannot exercise their right to rule if their right to rule is not collectively accepted. Rather a ruler does not even possess a right to rule unless he or she is able to exercise authority over his or her subjects. This seems to be a general feature of the deontic properties of those in authority. So Searle is right about institutional authorities, but wrong about other institutional forms. Moreover, Searle is also correct to speak of the rights possessed by institutional authorities as being powers. Indeed, the actions of those in authority, including those whose authority is not morally legitimate, constitute in large part the exercise of power, and not merely the exercise of authority. But from this it does not follow that all social actions somehow necessarily involve exercises of (social) power. In particular, contra Foucault and others of his view, it does not follow that actions in conformity to conventions and social norms are necessarily exercises of power, either on the part of those who conform, or on the part of those upon whose conformity one’s own conformity is dependent. This point ought to be evident from the analysis of conventions provided in Chapter 3. If an action is not necessarily an exercise in power, then presumably a joint action is not necessarily an exercise in power. But since conventions are a species of joint action, then it is reasonable to hold that conventions are not necessarily exercises in power. I now want to focus attention on the connection between collective acceptance and substantive functionality. I will begin with the special case of institutional authorities. What does it mean to collectively accept or collectively treat someone as if he or she is a leader? Presumably it means that people obey that person’s directives because they believe that they ought to do so. But now what is the actual or realised substantive functionality of a leader? Surely it is in large part getting people to obey one’s directives because they believe that they ought to do so. So, in the case of leaders, and other institutional authorities, I suggest that Searle’s notion of collective 189
Social Action acceptance is redundant; it collapses into actual or realised substantive functionality. To possess the functional properties of a leader is (in large part) to have one’s directives obeyed by one’s followers. But to have one’s directives obeyed by one’s followers is to be collectively accepted. What of the relation of collective acceptance and substantive functionality in the case of other kinds of institutions, such as money or language? Once again I suggest that Searle’s notion of collective acceptance collapses into the actual or realised substantive functionality of the institution.17 If people exchange dollar notes for the purpose or end of receiving goods (and exchange goods for the purpose of receiving dollars) then they have “accepted” dollars in the only important sense of that term; so there is no need for an additional notion of collective intentionality or collective acceptance or treating dollars as if they were dollars. A similar point can be made for language. If speakers and hearers utter a set of specific noises for the purpose or end of referring to specific things, communicating thoughts to others, and so on, then they have “collectively accepted” the language in the only important relevant sense of that term. That is, the substantive functionality of language has been realised; the speakers and hearers are using language. And there is this further point. Speakers and hearers – and persons using a medium of exchange, and so on – typically act in conformity with conventions and, indeed, social norms. So the substantive functionality of these institutions is realised in accordance with conventions and social norms. Accordingly, there is collective acceptance in the sense of conformity to conventions and social norms. When we use language or a medium of exchange we typically do so in part by conforming to the relevant conventions and social norms. So much is obvious. A final point about Searle’s notion of collective intentionality and its relation to rules. In his later formulations it turns out that the weintentional state in question is conventional in character; he speaks of conventional powers.18 Searle nowhere offers an analysis of conventions. However, he does distinguish them from rules. For Searle, conventions, in contrast with rules, are arbitrary.19 Moreover, as we have seen, he takes the view that rules are necessarily normative phenomena. There are a number of unanswered questions here. What exactly 190
6. Social Institutions and Social Groups is the nature of rules and of conventions, and precisely how are these phenomena to be differentiated? While Searle does not explain the nature of conventions or the nature of rules in general, he does offer an account of the distinction between constitutive rules and regulative rules. Moreover, constitutive rules have a key role in his account of institutions. Unlike regulative rules, constitutive rules “create the possibility of certain activities.”20 Constitutive rules have the form “x counts as y in c” where y is the function imposed on x in conditions c.21 So constitutive rules are, according to Searle, very important in the construction of social institutions. Now it is by no means clear that some conventions could not have the form x counts as y in c. On Searle’s view of conventions as arbitrary, a convention-governed imposition of y on x (if it were possible) would necessarily be a matter of arbitrary choice. But this seems to be so in many instances. Consider a convention among military strategists to use certain shaped pebbles to stand for troops in their strategising concerning troop movements. More important, it is by no means clear that the notion of a rule – as opposed to a certain realised rule content, namely that persons count something as something else – is actually doing any work here. I argued above that function imposition does not necessarily bring with it values and deontology. Now I suggest that function “imposition,” whether collective or not, does not require rules, or even conventions. Consider that example of military strategists collectively imposing functions on physical objects. If they used toy soldiers that looked like real soldiers then the arbitrary character of their decision would diminish. So according to Searle’s argument there would not be a convention. But equally there might be no rule. It might be a one-off episode, never to be repeated. The point is that it is the notion of counting one thing as another thing that is doing the work; whether they count x as y in accordance with a convention or rule or by virtue of some other collective device is not important. This latter point is really a special case of the more general point made earlier, that collective acceptance collapses into actual or realised substantive functionality. For what is critical is that the relevant individuals perform the action, that is treat the gardener as if he were president, count the pebble as a soldier, use the paper as a medium of exchange, and so on. What is critical is not constitutive rules, but rather substantive functionality. 191
Social Action In the opening section of this chapter I located the notion of a social institution in relation to that of an organisation, and in the introduction to this chapter I offered a preliminary description of social insitutions. Thus far in this chapter I have provided a critique of Searle’s account of institutions. In the next section I will return to my own view. social institutions As earlier indicated, I agree with Searle’s general view that social actions have a certain logical priority vis a` vis social entities, including social institutions. However, I do not subscribe to Searle’s account according to which the social actions in question are essentially the collective imposition of status-functions on pre-existing social objects, and, ultimately, natural objects. Rather, I view institutions as structures of conventions and norms that regulate various species of generic joint activity, and I view institutional actions as actions performed in accordance with the conventions, norms, ends, roles, and so forth, that constitute institutions. Moreover, I hold the view that institutions, in part, should be marked off from other related phenomena, such as (noninstitutional) organisations, by virtue of their normative dimension. And I accept Searle’s view that this normative dimension includes the ascription to persons and objects of status-bearing deontic properties. However, as I have just argued, I do not accept that the source of those deontic properties is collective acceptance, other than in the special case of institutional authorities. In relation to institutional authorities, I endorse Searle’s view that the relevant deontic properties are powers, and that those in authority are engaged in the exercise of these powers. But, contra Searle, I do not accept that the exercise of power is constitutive of other social actions, such as actions in conformity to conventions or social norms. Moreover, in my view, the normative dimension of institutions, including hierarchical institutions comprised of authority relations, involves more than the deontic powers Searle gives to them. For one thing, it involves the felt ethical or moral worth of the ends or purposes of institutions, and specifically the kinds of subjectively felt moral goods that institutions produce or fail to produce. Accord192
6. Social Institutions and Social Groups ingly, the definition of an institution will typically include a description of the (alleged) moral good that it purports to produce. For example, governments should provide leadership, universities purport to produce knowledge and understanding, language enables the communication of truths, marriages facilitate the caring and moral development of children, economic systems ought to produce material well-being, and so on. Moreover, these moral goods, or at least believed moral goods, are, normatively speaking, the collective ends of institutions, and as such they conceptually condition the social norms that govern, or ought to govern, the constitutive roles and activities of members of institutions, and therefore the deontic properties that attach to these roles. To use one of Searle’s examples, a police officer has certain deontic powers of search, seizure, and arrest, but these powers are justified in terms of the moral good, law and order (say) that it is, or ought to be, the role of the police officer to maintain. Further, and again, a defining property of an institution is its substantive functionality, and so, contra Searle, a putative institutional entity with deontic properties, but stripped of its substantive functionality, typically ceases to be an institutional entity, at least of the relevant kind; would-be surgeons who cannot perform surgery are not surgeons. Here, by substantive functionality, I have in mind the specific defining ends of the institution or profession. In the case of institutions, including professions, the defining ends will be collective ends; they will not in general be ends that an individual could realise by his or her own action alone. In short, the theory of institutions, and of any given institution, is a teleological theory. Further, institutions in general, and any given institution in particular, require both a teleological descriptive theory, and a teleological normative theory. Naturally, whether or not my commitment to teleological descriptive theories of institutions is warranted depends on empirical facts. If it turned out, for example, that most or all institutions did not have collective ends that were regarded either as intrinsic moral goods, or the means to intrinsic moral goods (derived moral goods) – that is, the participating agents did not in fact seek to realise the relevant putative defining collective ends – then my teleological descriptive theory would be false; I would have to abandon it. Moreover, the falsity of the teleological descriptive theory would put pressure on the acceptability of any teleological nor193
Social Action mative theory of institutions. If it turned out that no institution at any time or place in fact involved to any extent the pursuit of the relevant kind of collective end that was an objective (intrinsic or derived) moral good, then this would make it implausible to claim that institutions nevertheless in general ought to aim at collective ends that are objective (intrinsic or derived) moral goods. But it seems quite clear that participants in many institutions do as a matter of fact seek to realise collective ends that are believed to be moral goods, and are in fact moral goods, and that these collective ends are sufficiently fundamental to be regarded as definitive. Consider the institution of the English language. If the relevant speakers and hearers ceased to have as a collective end the successful communication of their beliefs, intentions, and the like, then the English language would go out of use. Again, consider the institution of the nuclear family. A mother and a father typically (jointly) act having as an important end the material and moral well-being of their child. For example, one parent might work and the other perform a child-minding role. It is obvious that the well-being of the child is not simply an unintended consequence of their actions; it is something that they explicitly have in mind, and indeed, in all probability, they will explicitly explain or justify their actions in terms of their contribution to this collective end. Now consider educational institutions. Teachers and students typically have as a collective end the attainment by the students of intellectual skills and knowledge. That the students come to possess these skills and knowledge is an end that teachers and students not only jointly pursue, but explicitly claim that they are pursuing in school mottos, discussions with students, and so on. Indeed, often the most powerful criticism that students or teachers believe they can make of a particular class is that the students do not learn anything. In relation to this, it might be argued that, first, nevertheless there are some institutions, especially economic institutions, where there are no such collective ends and, second, that I have failed to identify the most important social consequences of the kinds of institutional activity that I have discussed. This second point is hard to rebut in the absence of an account of what these social consequences are supposed to be. Candidate accounts include those propounded by Marxists and functionalists.
194
6. Social Institutions and Social Groups According to a functionalist, the most important social consequence of child rearing might be to ensure an orderly and harmonious society. According to a Marxist, the most important social consequence of schools might be to inculcate an ideology that enables class exploitation. I accept that sometimes the most important consequences of social institutions are consequences that are not also collective ends of the kind that I have described. But this does not show that these collective ends are not definitive of social institutions, and that the social consequences in question are definitive of institutions. Air pollution and associated ill-health might turn out to be the most important social consequence of the use of taxis in New Delhi but, nevertheless, taxis are to be defined as modes of transport – people use taxis in order to get from one place to another – and not in terms of their contribution to air pollution. Even so, air pollution is a huge problem in New Delhi, and, if possible, taxis should be built so as to limit their polluting effects. Similarly, if some schools are primarily sites of ideological indoctrination, then this needs to be looked into and changed; but the claim that necessarily, and/or by definition, schools are predominantly sites of ideological indoctrination is simply false. Further, quite often macro-level social consequences are in fact a collective end of institutions. A single teacher and her students have as a collective end the education of these particular students. The realisation of this collective end contributes to the realisation of a larger collective end, namely, the education of all the students in the school; and the realisation of this larger collective end contributes to the contribution of a still larger collective end, namely the education of all the students in the whole education system. What we have here is what I described in Chapter 5 as a layered structure of joint action. What of the other complaint: that some institutions, notably the economic system, do not centrally involve collective ends? In Chapter 5 I argued that the best way to understand organisational action, including in the case of business organisations, was in terms of quasijoint action. There is a collective end, for example the production of cars, but there is also an individual end, which is securing a wage. Now while the collective end of producing cars is not simply a means
195
Social Action to the individual end of securing a wage, nevertheless the collective end would not be pursued unless workers were simultaneously able to realise the individual end. Perhaps the most influential model of economic behaviour is that put forward by rational choice theorists. However, rational choice theory is a theory that purports to supply explanations of social action. By contrast, my account is an account of what social action consists in. It is consistent with my account of business organisations as layered structures of quasi-joint action that the ultimate motivation of the actions of economic actors is some benefit to themselves; but my account does not require that this be so. The rational choice model faces problems that my account does not face. Consider the problem posed by free-riding to rational choice accounts that seek to explain the production of public goods in terms of the individual actions of rational self-interested actors. As Mancur Olsen famously demonstrated, such rational selfinterested actors would free-ride, rather than make their contribution to the production of the public good.22 By contrast, my account of organisations is not committed to self-interested actors. Indeed, all institutions, including most organisations, involve conformity to social norms, and – as I made clear in Chapter 4 – conformity to social norms is not able to be understood simply in terms of selfinterested motives. As far as the economic system itself is concerned, for example the modern capitalist system, it is inevitably a mix of competition and cooperation. However, the conventions, laws, and institutional regulations that constitute the framework within which competitive economic behaviour takes place ought to be in large part contrived in terms of collective ends that are objective moral goods, and especially the provision of material well-being to the participating economic actors. The fact that the current capitalist system is inadequate in this respect merely serves to illustrate the need for institutional reform. Arguably, its inadequacies pertain more to economic justice than to economic efficiency. For there are profound questions of economic justice, both in relation to the differential economic benefits that go to those explicitly engaged in economic activity within the existing system – consider the salaries of many chief executive officers of multinational companies versus the salaries of “sweatshop” workers in some of the factories in Asia owned 196
6. Social Institutions and Social Groups by those multinationals – and in relation to those who are largely excluded from the existing economic system. Consider, in the latter connection, many women in contemporary Western societies, and many citizens of the nation states of Africa. A different kind of objection to my teleological account of institutions is that it fails to accommodate manifestly morally unacceptable institutions, such as the institution of slavery. Assume that slavery in (say) the USA was set up as an institution for the purpose of providing cheap labour. There are two salient questions. First, what is the defining collective end or ends of this institution of slavery? Second, in what sense, if any, was this end a believed and/or actual moral good? Presumably, the ultimate collective end of this institution was the material well-being of the relevant non-slave population. This was believed to be a moral good. Obviously, insofar as it was an end that excluded the slave population, it was, at the most, a limited moral good. Moreover, this collective end involved the systematic violation of the human rights of the slaves. So whatever the slave owners might have believed, the institution was morally unacceptable by virtue of the immoral means used to realise its collective end. But was there in fact a layered structure of collective ends to which the slaves contributed by engaging in joint activity? There might be a tendency to deny this because the slaves were coerced into performing the tasks that they performed. That the slaves were coerced is obviously true, but it is not necessarily relevant. An agent who is coerced into performing an action, might nevertheless, still intentionally perform the action, and perform it in order to realise the relevant end. For example, if I threaten to shoot you if you do not shoot Fred, and you intentionally fire your gun having as an end to kill Fred, then you intentionally kill Fred, albeit under mitigating circumstances. Perhaps you are not morally responsible for killing Fred; but nevertheless you did kill Fred. Matters would be different if you intentionally fired the gun, but did not have the death of Fred as an end; for example, if you intentionally fired but nevertheless tried to miss him. Similarly, if a group of slaves individually worked in the cotton field having as an end the harvesting of the cotton in that field, then they were engaged in joint action directed to a collective end. On the other hand, if each worked individually without (say) coordinating their actions in the service of the collective end, then there is 197
Social Action no joint action. It is an empirical question whether or not slaves engaged in joint activity directed to collective ends and, if they did, to what extent they did. The answer to this empirical question is probably in the affirmative, at least in many cases; for in all probability slaves need to engage in a considerable extent of joint activity directed to collective ends, if the institution of slavery is to work. Naturally, the fact that a group of persons is forced to have collective ends that only benefit their masters – as well as being forced to intentionally perform actions that are deeply inimical to their wellbeing – is bound to be profoundly damaging to their moral personalities; more damaging than if they performed the actions, but did not pursue the ends. This is in part because the ends that one spends most of one’s life pursuing are bound to be in part constitutive of one’s moral personality. To the extent that the teleological account of institutions pays due heed to the importance of collective ends – including the collective ends of slaves – it is, if anything, better equipped than other accounts to make sense of the damaging and immoral character of the institution of slavery. At any rate, it now appears that the institution of slavery is not a counter-example to the teleological account of institutions. Thus far I have spoken in terms of the theory of institutional action where institutions have been taken to be different and separate “entities.” However, there is also a need for a theoretical account of the theoretical interrelationships between different institutions. It is clear that on my teleological account of institutions any given institution is to be understood in terms of the collective end or ends to which their activities are and/or ought to be directed. However, there still remains the question of the relationships between institutions. One issue concerns the extent or degree of any required relationship. Another concerns the nature of the required relationship. As far as the extent of the relationships is concerned, in the post-Enlightenment West this interaction between institutional organisations belonging to the same society has typically taken place in the context of a commitment to a basic separation between them. Governments must stand apart from corporations lest public and private interests are confused, and corporations must stand apart from one another in the interests of competition. In communist regimes, by contrast, the doctrine of organisational (or at least institutional) separation, including separation of powers, has not 198
6. Social Institutions and Social Groups been adhered to. Japan constitutes an interesting third model. While Japan is obviously in some sense a liberal democratic state, there has been an extent of government, corporation, and bureaucracy linkage that is at odds with the notion of institutional separation. Moreover, there is some evidence that in recent years in the Western liberal democracies the doctrine of institutional separation is under threat in the face of policies coming under the banner of so-called economic rationalism. Such policies include the privatisation of law enforcement agencies and prisons, and the outsourcing of administrative functions. As far as the nature of the relationship between institutions is concerned, this is presumably to be determined primarily on the basis of the extent to which the differential defining collective ends of institutions are complementary rather than competitive, and/or the extent to which they mesh in the service of higher order ends. Finally, the substantive functionality – now collective end or ends – of an institution has a distinctive feature, which is in itself a (further) defining property of institutions. This feature helps to explain the sense in which institutional action – as opposed to merely conventional or norm-governed action – is, or can be, creative. The institutional creativity I want to emphasise is not that made so much of by Searle, namely social construction by means of social acceptance. Rather, my concern is with the capacity of institutions to not only constrain, but to create the possibility of, novel or new or unique individual and social actions. The most obvious example of this is the oft-cited capacity of languages to enable the generation of entirely new sentences. Such institutional action is often the result of what I have called in Chapter 5, a joint mechanism.23 As we saw in that chapter, joint mechanisms are also a defining feature of organisations. Note also that joint mechanisms are in themselves normatively neutral. Joint mechanisms, like organisations, only take on a normative dimension in institutional contexts. To recap joint mechanisms consist of: (a) a complex of differentiated but interlocking actions; and (b) the result of the performance of those actions. Thus a given agent might vote for a candidate. He or she will do so only if others also vote. But further to this, there is the action of the candidates; namely, that they present themselves as candidates. That they present themselves as candidates is (in part) constitutive of the voting mechanism. 199
Social Action Voters vote for candidates. So there is interlocking and differentiated action. Further, there is some result (as opposed to consequence) of the joint action; the joint action consisting of the actions of putting oneself forward as a candidate and of the actions of voting. The result is that some candidate is voted in. That there is a result is (in part) constitutive of the mechanism. Moreover, joint mechanisms are a pervasive feature of organisations and institutions. Examples include voting mechanisms, tossing a coin, courts of law, governments, and so on. This completes my discussion of social institutions. I now turn to social groups. social groups We have already noted that a social group consists of a set of individual persons who are (at least) the current participants in some common structure or structures of conventions (including at least a structure of linguistic conventions).24 We have also noted that conventions are essentially facilitative and instrumental social forms, whereas social norms embody the moral principles and values of a social group. This is why we earlier insisted that social groups by definition also involve a common structure of social norms (in the sense of accepted – though not necessarily objectively justifiable – principles of morality). Yet such a structure of social norms is necessarily embedded in the fundamental institutions of the social group in question. Hence the further condition for being a social group, namely a common structure of fundamental institutions, including at least linguistic, kinship, legal (or quasi-legal), and economic institutions. Most English and German people share a structure of conventions, including the conventions of the English language, and share also a common structure of social norms, including those embodied in the criminal laws of both countries. They also share a fairly similar set of structures of fundamental institutions, including the institutions of the modern nuclear family, capitalism, and the liberal democratic state. But the Germans and the English do not constitute a single social group. One reason for this is the lack of a common intergenerational history. The history of the English certainly intersects with that of the Germans, but they are nevertheless relatively 200
6. Social Institutions and Social Groups distinct. Let us then add the following condition for being a social group: a common stretch of intergenerational history. The resulting definition remains inadequate. Many minorities with strong claims to being distinct social groups have shared a common stretch of intergenerational history with the larger community in which they are a minority, and have participated in common structures of conventions (including linguistic conventions), social norms, and at least some fundamental institutions. However, their point is that they have done so under duress, and that there are important “cultural” differences, which have been suppressed. Consider the Kurds in Turkey or Iraq. Here we need to distinguish two conditions: (1) The membership of the social group in question is able to be differentiated from the membership of other social groups in virtue of (a) their common ancestry, and (b) their participation over generations in some common set or sets of joint actions, activities, and/or projects, including joint, so to speak, epistemological and narrative projects (roughly, joint, intergenerational attempts to understand “the world”). In short, condition (1) expresses the proposition that a set of individual persons have shared a common life over generations. (2) Their moral rights have been violated, and have been violated because they are members of the social group to which they belong. Roughly speaking, social groups that meet condition (2) are oppressed peoples.
Clearly (1), but not (2), is a condition for being a social group. Condition (2) presupposes the existence of a social group. So condition (2) cannot be a necessary condition for being a social group. On the other hand, condition (1) ought to be sufficient to mark off minority social groups from the members of an oppressive “host” social group. Typically, the former will not share the same joint activities and projects, and especially not the same narrative projects, with the latter. Moreover, even where they participate in the same joint activities and projects, “fault lines” of social tension and dissent will indicate that joint “participation” is coercive in character. Social Groups and Morality Our definition of a social group consists of the five conditions mentioned in the previous section. A number of questions that now arise 201
Social Action concern the moral dimension of social groups. One set of questions concerns the moral worth and moral rights of social groups. Do social groups necessarily have moral value and associated moral rights, analogous to the way in which individual persons have moral value and associated rights? Specifically, are these defining conditions sufficient to generate a collective or group right to survival in the sense of self-preservation of the identity of the group and to autonomy, analogous to individual rights to life and to individual autonomy? A second set of questions concerns the moral development of social groups. What are the mechanisms available to social groups to enable them to undergo change, especially moral development, and do so on the basis of a reflective, as opposed to merely passive, process? A third set of questions pertains to the interaction between members of different social groups. How is such interaction to be understood, and in what sense, if any, does it undermine the very notion of a stable, unitary social group? In relation to the first kind of question – do social groups necessarily have collective rights to autonomy and to self-preservation – the answer appears to be in the negative. For one thing, social groups have to be viable as independent entities if their claim to rights to autonomy and cultural self-preservation are not to be empty. Thus a social group may need to have a stretch of territory it can inhabit, and to which it has a moral entitlement. For another thing, predatory social groups such as the Sicilian Mafia or the Chinese Triads may well meet all five conditions. Indeed, various socioeconomic classes, such as the eighteenth-century British aristocracy, the nineteenth-century British working class, the whites in apartheid South Africa, and the blacks in apartheid South Africa, also appear to meet them. But the claim that any of these groups has collective rights to autonomy (at least) is very implausible. And the claim that the Sicilian Mafia or the Chinese Triads has collective rights to the preservation of their culture is clearly false. Rather, the predatory criminal groups should be destroyed and their members incarcerated. As for the other groups, the power of the exploiting ruling classes ought to be diminished and the rights of the exploited classes respected. Accordingly, if we are to ground these alleged collective rights we 202
6. Social Institutions and Social Groups will have to look beyond any definition of what it is to be a social group. In short, there is, I suggest, a disanalogy between social groups and individual human beings at least in relation to autonomy and cultural self-preservation. All individual persons have a fundamental right to autonomy and to self-defence, irrespective of their moral character. I do not lose my right to autonomy or to selfdefence because I am a bad person who is leading a worthless life. This is, of course, consistent with being restrained or even locked up if I am a danger to other persons. Not so social groups. In the case of social groups there is no right to continued existence or to possess autonomy unless the social group can pass various moral tests in relation to its contribution to the moral well-being and autonomy of its members. Moreover, social groups also need to pass tests in relation to their treatment of other social groups, including at least the test of not seriously harming other groups. Indeed, in a world of increasing interaction between social groups, it is more and more reasonable to require very strict (albeit minimum) standards in respect of the behaviour of social groups in relation to other social groups. Thus, when it comes to the collective rights of autonomy and cultural self-preservation of social groups, it all depends, so to speak, on the moral character of the particular social group in question. Let us now turn to the second sort of question, concerning the possibility of social groups collectively undergoing reflective moral development. Here a key issue is the availability of mechanisms to revise the social norms of social groups. The social norms of social groups are subject to revision and application to new contexts. Revision is possible because of inconsistencies between different social norms, and because there is a gap between individual moral principles and social norms, and between individual and collective interest on the one hand, and social norms on the other. Moreover, in societies comprised of multiple social groups, or social groups in contact with other social groups, revision is possible because of differences between the social norms of the different social groups. This process of revision and application involves communication and (at least ideally) that communication is informed by a process of collective deliberation. Such communication includes what I will term socially directed action. Socially directed action 203
Social Action is a form of communicative social action. Indeed, it constitutes a further basic category of social action, though not one that I intend to elaborate in great detail in this book. Much socially directed action is ritualistic in character. Consider the repetition of popular narratives, or the varieties of traditional ceremonies. Presumably such socially directed communicative action exists largely to reinforce or at best renew prevailing social norms. However, there is one species of socially directed action that is somewhat different. This kind of socially directed action can perhaps best be captured by somewhat exaggerated examples, such as the following.25 A black man and a white woman go out together to a social function in a racially segregated society. They know that what they are doing is socially unacceptable. What they are doing involves resisting social pressure, including pressure from themselves, and constitutes a public rejection of the social norms of the society to which they belong. This action is in a certain sense individualistic. It is not an action prescribed by a convention or norm. But the action has a social aspect. The social aspect of the action lies in the relation between the two individuals on the one hand and the social group to which they belong on the other; the social group is the audience for an intentional communicative action performed by the two individuals. The woman (at least) is communicating: (a) as a member of the social group; and (b) to those other people as members of that group. In communicating in this way, she conceives of herself as being on the one hand different from, and indeed at odds with, the rest of the group. On the other hand, she also, and simultaneously, conceives of herself as being a member of the group. To perform a socially directed action of this sort is not to conceive of oneself simply as an individual standing outside the group, nor to conceive of oneself simply as a member of the group conforming to the conventions and norms of the group. Rather, it is to conceive of oneself as standing in a relation to the rest of the group for the purposes of communicating a matter of importance to the group. Such socially directed actions are part and parcel of the process of revision and application of social norms. In this connection it is important to note the nature and role of the public institutions of communication in a society, and specifically – in modern societies – the mass media. Such institutions are 204
6. Social Institutions and Social Groups comprised in part of public communicators who are engaged in socially directed actions. Naturally, such public communicators can engage in socially directed action that facilitates regressive social norms or debases morally desirable ones. In this connection I make two points. First, such public communication does not have to be explicitly concerned with social norms to have implications for those norms. For example, a diet of films glorifying violence might be taken to imply that routinely deploying violence to resolve one’s problems is acceptable, and perhaps even laudable. Given other contextual factors, such socially directed action emanating from public communicators may well have a corrosive effect on existing social norms proscribing violence. Second, public communication via the mass media by definition takes itself to be important, no matter how trivial its content actually is, and no matter how trivial it presents itself at one level as being. It takes itself to be important simply by virtue of occupying the limited space available for public communication via the media, and by virtue of the vast number of people being communicated to. Only a relatively few public communicators actually occupy this space, so those that do must be important. For example, the fact that (say) a range of game shows might present themselves as “just for fun” does not mean that they are in fact for fun, or do not in reality view themselves with great seriousness. For one thing, the comperes of these shows become rich and famous celebrities. Did they just want to have fun? For another, the contestants who fail to win the expensive prizes on offer presumably do not view their losses as simply “entertainment.” What of the audiences? Judging by the level of interest in these shows, and in the private lives of the celebrities who present them, the audiences also regard these shows as important. Even for the audiences it is apparently not just a matter of entertainment. Social Groups and Transcultural Interaction Transcultural interaction is, and always has been, a pervasive feature of social groups. How is it to be understood, and what are the implications for (so to speak) the ontology of social groups? Given this interaction, does it really make much sense to speak of social groups, as distinct from organisational structures – such as the na205
Social Action tion state – that might be composed of various collections of individuals? We need first to clarify the notion – or, as it will turn out, the multiple notions – of transcultural interaction.26 Most societies in most periods of human history have engaged in most of the various kinds of generic joint activity that comprise spheres of activity and underpin social institutions, and have done so because these activities are grounded in basic human needs. Moreover, throughout the course of history many, if not most, social groups have interacted with some other social groups communicatively, economically, sexually, and so on. So a sphere of activity is not by any means necessarily an intrasocietal or intracultural phenomenon. Since communicative and economic (and so forth) interactions – at least to the extent that they are voluntary interactions – are to some extent structured by conventions, so trans-societal and transcultural communicative and economic (and so forth) interactions will be structured by conventions, conventions to which the members of both interacting social groups will be party. Dialogue between members of different social groups, societies, and cultures presupposes a common language. This remains true notwithstanding problems of differences in interpretation. For example, French used to be the international language and now English is. Trade between societies presupposes a commonly adhered to system of exchange, whether that be a barter system or a monetary system or some other system. Let us call mere convention-governed interaction between distinct social groups, trans-societal interaction. The existence of trans-societal interaction displays the distinction between social groups and mere participants in a sphere of activity. A social group consists of a set of particular individuals who participate in a number of the same spheres of activity, and therefore are party to a number of the same networks of conventions (each network regulating the interactions in a given sphere of activity). In addition, however, members of a social group (by stipulative definition) share a common intergenerational history, and have been inducted into a common structure of social norms and social institutions. What of the interaction between organisations belonging to different social groups? Historically there has been a great deal of political and economic (cooperative) interaction between economic 206
6. Social Institutions and Social Groups or business organisations and between political organisations, such as governments. However, a good deal of it has been coercive in character. Consider nineteenth-century British gunboat diplomacy pursued for the purpose of ensuring the opium trade had access to China. On the other hand, much of it has been broadly cooperative in character. And the current mania for regional trading blocs, economic communities, and the like indicates a belief that cooperation can be mutually beneficial. But, as with individual persons, including individual members of the same organisation, cooperative interaction between different organisations, and between organisations belonging to different societies, presupposes a common set of conventions, such as linguistic conventions and conventions governing the exchange of goods. Accordingly, one species of trans-societal interaction is trans-societal organisational interaction. However, there is a different kind of trans-societal interaction involving organisations. Historically, as well as in the contemporary world, there have been many different kinds of trans-societal organisations, including empires, military organisations, religious organisations, and multinational corporations. Such trans-societal organisations presuppose not only that the two or more social groups have an overlapping sphere, or spheres, of activity and an adherence to a common set of conventions, but also that they belong to at least one of the same organisations. Insofar as such organisations have a minimal impact, or are unwanted impositions, on a pre-existing social group, they do not necessarily destroy the identities of the different social groups that belong to the trans-societal organisations in question. Accordingly, the notion of trans-societal interaction within a given organisation is quite coherent. What about trans-societal institutions? There are such things as trans-societal institutions. The English language is a case in point. However, the notion of a trans-societal institution contains an inherent tension. This is because a social group is in part defined in terms of a particular set of social institutions. So two different social groups could not have all, or even most, of their institutions in common; if they did it would cease to make sense to claim that they were in fact different social groups. That said, I do not dispute that entrants into contemporary cultural milieux from traditional societies, traders, criminals, and warriors can belong to more than one social group; such people can be 207
Social Action cultural hybrids. Such individuals belong to more than one set of institutions. They can be both members of the institution of the English language and members of the institution of the Swahili language. Or perhaps they are members of a traditional tribal society and also members of the institution of international capitalism. Note that a group of individuals need not participate in the same set of organisations in order to belong to the same social group. For example, in the case of the Indian or Chinese or Anglo-Celtic diasporas, we have the same social or cultural group belonging to different sets of organisations, including different nation states – the nation state being an organisation of organisations. Many Australians are members of the Anglo-Celtic cultural group, though they belong to different organisations, including a different nation state, from their British counterparts. Note also that the notion of an institution defined in terms of both conventions and social norms allows us to distinguish between two species of social groups, namely societies and cultures, and therefore to distinguish between trans-societal and transcultural interaction. A group of individuals who live together, are engaged in joint projects, and share a set of networks of conventions, a structure of social norms, and at least some interlocking structure of institutions and organisations, including, in particular, the institution of government, are a society. By contrast, a group of individuals who do not necessarily live together at a particular time under a common government, and who do not belong to any of the same organisations, but who nevertheless share the same set of conventions, social norms, and institutions and have at least a large stretch of their history in common, are a cultural group. So the Japanese in Japan constitute the largest part of a society and the largest part of a single cultural group. Perhaps the Japanese and the Ainu (a cultural and racially distinct minority in Japanese society) are part of the same society, but belong to two different cultural groups. Perhaps they are part of a largely monocultural, but nevertheless to some extent multicultural society. The Indians in Malaysia and the Indians in Fiji are not a society, though they are part of the same cultural group. Further, economic interaction between the Chinese in Taiwan and the Chinese in Hong Kong is not transcultural interaction, though it is transsocietal interaction. However, the economic interaction between 208
6. Social Institutions and Social Groups Australians and Japanese is both transcultural interaction and transsocietal interaction. Note also that social groups are not the same things as cultural groups; as explained, members of a social group, but not necessarily a cultural group, live together sharing the same particular networks of conventions, social norms, and institutions. On the other hand, members of a social group do not necessarily constitute a society, since there can be more than one social group within a society; members of a society, but not necessarily a social group, have the same institution of government.
209
7
Collective Rights
Thus far in this book I have put forward a teleological cooperative model of social action and offered piecemeal philosophical analyses of what I take to be the main social action types, including conventional and norm-governed action and organisational and institutional action. In all this the notion of joint action has played a pivotal role, and thereby provided a certain unity to my account. In the remaining two chapters I turn to the analysis of two fundamental moral properties that arise in the context of social action, namely collective right and collective responsibility. Once again, the central theoretical notion underpinning these analyses is the notion of joint action. I will consider collective rights in this chapter, and collective responsibility in the following one. A discussion of collective moral rights raises a number of important issues.1 One such issue is the claimed existence of supraindividual entities, such as nations and ethnic groups, that possess moral properties, including moral rights, above and beyond the moral properties possessed by the individual persons that compose them. Another issue is the possibility that fundamental human rights possessed by individuals might be justifiably overridden by collective rights. A third issue is the moral claims of minority cultural groups. Such claims are typically couched in terms of collective rights. A fourth issue is the justification for interference by one state in the affairs of another state. One form of justification in such cases is the alleged collective right of the members of a community to outside assistance. Many collective moral rights are joint rights of individual persons. My notion of a joint right presupposes the analysis of joint actions provided in Chapter 2. Given this analysis of collective rights as joint rights of individual persons, there may well be no need to postulate supra-individual entities with moral rights above and beyond those possessed by individuals.2 On the other hand, there might well be a 210
7. Collective Rights need to go beyond a conception of individuals which conceives of them as having rights – and moral value – only in virtue of properties they possess as individuals, such as the capacity for self-interested rational choice. And there might well be a need to adjudicate conflicts between (non-joint) individual rights and joint (individual) rights. In some of these cases joint (individual) rights might override (non-joint) individual rights. There are joint rights to collective goods possessed in virtue of joint production of those goods. There are joint rights not to be harmed. There are joint rights to assistance. And there are joint rights of participation. The class of collective rights cuts across all these categories, but does not exhaust all of them. For example, some joint rights are rights in relation to interpersonal, as opposed to collective, goods. Friendship is an interpersonal good. Specifically, I will argue that many collective rights are joint rights to collective goods possessed in part in virtue of membership of a social group. Here, the category of collective goods includes collective harm avoidance as well as certain sorts of assistance to collectives. Further, the notion of a social group involves at least shared social norms, shared knowledge, and mechanisms for group decision making.3 In Chapter 4, I provided an analysis of social norms, and in Chapter 5, of mechanisms for group decision making. In Chapter 6, I outlined a conception of social groups. I will further argue that there are probably no minority moral rights, at least in the sense of what Kymlicka terms “special,” as opposed to universal, rights.4 Special rights are rights that one community has that another does not. Rights claimed to be minority rights, such as the land rights and hunting rights of indigenous peoples, and the political and language rights of some minority cultures in semi-autonomous states, turn out to be either moral rights that are not minority moral rights, or else merely morally justified legal minority rights.5 The basic notion of a moral right that I will be working with is the one typically advanced in the philosophical literature. Roughly speaking, a moral right entails correlative duties on the part of some other agent, including at least the duty to refrain from preventing the bearer of the right from exercising his or her right. It is also a very weighty moral consideration, but not a moral absolute.6 211
Social Action So a collective right has these two properties. However, collective rights, if there are any, have further properties, namely ones that distinguish them from individual rights. What those properties could be is at this stage quite unclear. However, we can delimit collective rights by saying what they are not. A collective right is not a right possessed by an individual person in virtue of a property or properties possessed by that individual considered only as an individual. So the right to life is definitely not a collective right, but in all probability the right of the Finnish people to self-determination is a collective right. We need to distinguish between collective rights that are rights held against other social groups and collective rights that are rights held against individual members of the rights-bearing group itself. Collective rights of the former kind include, but are not exhausted by, rights of a social group held against the larger community to which it belongs. These are Kymlicka’s special rights. Rights of the group held against its members he terms “group rights.”7 An instance of collective rights held against the larger community might be the land rights of the Australian Aboriginals held against other Australians. An instance of a “group right” might be the right of Islamic communities to enforce the teachings of the Islamic faith to all members of their community. An instance of a collective right held against other separate social groups might be the right of one nation state not to suffer interventions in relation to their domestic policy by other states. Moreover, theoretically a collective right could either be a right that attaches to a supra-individual social entity, such as the Volk or the Sioux nation, or it could be a right that attaches to individual members of a social group.8 Proponents of the former view of collective rights include Van Dyke, Michael Macdonald, and (in a somewhat different vein – collective rights as rights of artificial persons) Re´aume: Proponents of the latter include Ian Macdonald, Hartney, and Narveson.9 Further, collective rights – whether conceived as supra-individual rights or as the rights of individual members of social groups – could be rights to individual or collective goods. Roughly speaking, a collective good is one that cannot be exclusively enjoyed by an individual member of a social group. If it is enjoyed by one it is enjoyed by all (or at least most). An example of a collective good might be an 212
7. Collective Rights educated or tolerant or well-ordered society. Moreover, the notion of a collective good here used is to be construed fairly widely so as to embrace avoidance of collective harms. An example of a collective harm to a social group is invasion of its territory and loss of selfdetermination. I believe that there is a strong presumption against accepting the existence of supra-individual entities possessing moral rights. For one thing, such entities are ontologically mysterious. For another, the acceptance of them brings with it moral and political dangers of the kind evidenced in the relationship between supra-individual collectivism and the emergence of totalitarian movements. Nor are such movements a thing of the distant past. As recently as 1997, the Co-Prime Minister of Bosnia and Herzegovina, Dr. Haris Silajdzic, had this to say: Around 250,000 innocent people were killed and another 200,000 suffered in concentration camps in the process. Two million people were forced to leave their homes in Bosnia and some of them will never return. A Bosnian resistance was made out of scratch and sometimes Bosnians stopped Serbian tanks literally barehanded. The world, with honorable exceptions, maintained its neutrality, meaning that the Bosnians were kept unarmed. For some strange reason the West did not at least in the beginning identify with the Bosnians against fascism. It took the revelation of the existence of concentration, rape and death camps in Bosnia and an embarrassing media campaign for the Western governments to act. In the meantime, Bosnia resisted and suffered beyond comprehension. . . . Could the Bosnian case be a paradigm, not only in explaining the past but also in determining our future actions? Is Bosnian multi-ethnic democracy versus mono-ethnic totalitarianism, and its [Bosnia’s] religious and cultural multi-culturalism versus unilateralism, globally relevant to the future of humankind?10
At any rate, the burden of my case here against supra-individual collective rights consists in the provision of an alternative individualist account; not only are such supra-individual rights unwanted, they are probably not needed. In my view, if there are collective rights, then they are rights that attach to individual members of a collectivity. But how do we now distinguish between individual and collective rights? The most obvious way is simply to hold with Ian Macdonald11 that 213
Social Action collective rights are rights that individuals have by virtue of membership of some community or social group, and individual rights are rights individuals possess simply by virtue of some property or properties they possess as individuals. For example, French language rights are collective rights because they are possessed by members of the French-speaking community and possessed by these individuals in virtue of their being members of that community. On the other hand, the right to life is an individual right possessed by all humans. No doubt there is some such distinction to be made along these lines. Moreover it is plausible that there are individual human rights of the latter sort. Such rights evidently include not only the right to life, but also the right to personal security, the right not to be arbitrarily detained, the right to freedom of movement and of thought. However, there seem to be counter-examples to this way of drawing the distinction between collective and individual rights. In the first place, there are some paradigmatic collective rights that do not fit this definition of collective rights. In the second place, there are some rights, typically regarded as individual rights, that are nevertheless rights individuals possess partly in virtue of membership of some social group. Let us look first at some paradigmatic collective rights that would turn out to be individual rights on this account. Consider the Australian national or community right to exclude others from its territory, or the right of the Kurds to secede from Iraq. These are not rights that any or each single individual member of the communities in question have. Joe Ocker of Balmain in Sydney, Australia, does not have a right to exclude would-be immigrants to Australia.12 An individual Kurd does not have a right to secede. So the view that collective rights are simply individual rights possessed in virtue of membership of a social group cannot be correct. But now let us look at some rights considered to be individual rights and see how they fare on this account. Consider the right to vote. (If it is thought that the right to vote is a legal but not a moral right, substitute some more general right such as the right to political participation.) This is often regarded as an individual right. Certainly, it is a right that individual persons possess. Moreover, it is (apparently) a universal right.13 In many past societies slaves and 214
7. Collective Rights women were excluded from voting; evidently there was a failure to recognise a universal moral right. Surely, the right to vote is a paradigmatic individual right, to be contrasted with, for example, special language rights claimed by minorities. However, a necessary condition for having a right to vote in, say, the United Kingdom, is membership of a social group, namely the community consisting of residents of the United Kingdom. Chinese living in China do not have a moral right to vote in the United Kingdom. So, according to the membership of a social group criterion, the right to vote must be a collective right. Nor does it help to move to a higher level of abstraction and hold that everyone has a right to vote in some social group or other. For in the first place, this underdescribes the situation. Britons do not simply have a right to vote in some social group or other; rather, they have a right to vote in the United Kingdom and not elsewhere.14 In the second place, it may well be false. Presumably lone individuals choosing to operate outside all political communities do not have a moral right to vote in any political community. Perhaps the best response to this apparent counter-example is to bite the bullet and claim that the right to vote is – notwithstanding appearances – a collective right. After all, as we have just seen, there is arguably no universal right to vote as such, but rather only a right to vote in the French elections, or a right to vote in the United Kingdom elections, or a right to vote in some other election. Of course, it is not membership of any old social group that grounds the right to vote, it is membership of certain specific kinds of social groups, and especially membership of political communities. But this consideration does not count against claiming the right to vote as a collective right; it merely compels us to be more precise in our definition of voting rights. The main consideration against counting the right to vote as a collective right is fairly weak. This is the consideration that it is not simply membership of particular kinds of social group that grounds the right; the right to vote also in part depends on the possession of an individual property or right, namely, individual autonomy (or some cognate notion). I conclude that, if the right to vote is a moral right, it is a species of collective right, albeit a pervasive one. It is pervasive because most people belong to political communities. The discussion thus far has shown that the distinction between 215
Social Action individual and collective rights is not simply a distinction between individual rights possessed in virtue of membership of a social group, and individual rights not so possessed. Matters are more complex than this. However, it does not follow from this that we need to abandon individualist accounts and return to the unwanted view of collective rights as supra-individual rights. Nor does it follow that we need to abandon as one criterion of a collective right that it be possessed in part in virtue of membership of a social group. Rather we need to focus on some more specific species of individual right. My suggestion is that we take seriously the notion of a joint right and see where this leads us. joint rights Joint rights are rights that attach to individual persons, but do so jointly. On this view Joe, together with the other members of the Australian community, have a joint right to exclude would-be immigrants to Australia. Another example is the right to secession.15 Arguably the Kurds in Iraq have a right to secede. But, if this is a right, it is not a right that some Kurdish person has as an individual, even an individual qua member of the Kurdish people. After all, an individual person cannot secede. The right of the Kurds to secede – if it exists – is a right that attaches to the individual members of the Kurdish social group, but does so jointly. If collective rights are joint rights – I do not say that joint rights are necessarily collective rights – then a number of questions present themselves. First, precisely what are joint rights? Second, what sort of goods are joint rights rights to? Are they rights to collective goods? Finally, there is the question of the relationship between joint rights and membership of a social group. For example, are collective rights joint rights possessed in virtue of membership of a social group? Let me now try to answer some of these questions, beginning with the definition of joint rights.16 Roughly speaking, two or more agents have a joint moral right to some good, if they each have an individual moral right to that good, if no one else has a moral right to that good, and if the individual right of each is dependent on the individual rights of the others. 216
7. Collective Rights Let us assume that political participation in Canada is a good. Let us further assume that each Canadian citizen has a moral right to political participation in Canada, non-Canadians do not have a right to political participation in Canada, and the right to political participation of each Canadian is dependent on the possession of the right to political participation in Canada of all the other Canadians. In that case Canadians have a joint moral right. Such joint rights need to be distinguished from universal individual rights. Take the right to life as an example of a universal individual right. Each human being has an individual right to life. However, since my possession of the right to life is wholly dependent on properties I possess as an individual, it is not the case that my possession of the right to life is dependent on your possession of that right.17 However, notice that joint rights can be based in part on properties individuals possess as individuals. The right to political participation is based in part on membership of a political community, and in part on possession of the property or right of autonomy. This also raises the question of whether or not joint rights are necessarily possessed in part on the basis of membership of a social group. I suggest that they are not, and that therefore the way is open to define collective rights as joint rights possessed in part on the basis of membership of a social group.18 But I am getting ahead of myself. So much for joint rights per se. Let us now consider collective goods.19 Here there are a number of distinctions to be made. As Raz points out, there are necessarily collective goods and ones that are merely contingently collective.20 A right of access to a water supply might only be contingently collective. This would be so if, when the water supply is cut off, everybody’s supply is cut off. But under a different system selective cutting off is possible. By contrast a tolerant society is necessarily a collective good. The tolerance of the society is not something that could be channelled to certain individuals only. Following Re´aume, in respect of necessarily collective goods, we can further distinguish between those that an agent can choose not to enjoy himself or herself, and those that he or she cannot choose not to enjoy.21 Perhaps clean streets are of the former kind and a law-abiding society of the latter. A recluse could not be prevented from enjoying clean streets or a law-abiding society, but he or she 217
Social Action could choose not to enjoy clean streets by never going out. On the other hand even by staying at home he or she cannot choose not to enjoy a law-abiding society. Let us now explore the connection between joint rights and collective goods. First we need to get a clearer picture of joint action or activity.22 Here I rely on the account of joint action provided in Chapter 2. According to that account, two or more agents perform a joint action if they each have the same end, and they each perform an individual contributory action in the belief that the others will do likewise. Accordingly, they ensure that the end is realised. Thus Fred and Bill each have as an end that the dining table be lifted, and so Fred lifts one side of the table, Bill the other side. Here each intentionally lifts his side in the belief that the other will lift the other side, and the table will be lifted. Note that such an end has a special feature. It is an end that is realised by the agents acting jointly. Let us call such ends collective ends, notwithstanding the fact that the state of having an end is a mental state existing in the mind of an individual person; it is not a mysterious supra-individual mental state. Sometimes the end realised in joint action is a collective good. If so, then a joint right may well be generated. It is easy to see why these agents, and not other agents, would have a right to such a good; they are the one’s responsible for its existence, or continued existence. It is also clear that if one participating agent has a right to the good, then – other things being equal – so do the others. That is, there is interdependence of rights with respect to the good. Let us look a little closer at this joint right. First, what is the basis of this right? Certainly, it is not the fact that I am a member of a social group. So joint rights are not necessarily collective rights. Or at least they are not necessarily collective rights on the plausible assumption that collective rights are the rights of social groups. Perhaps the basis of this joint right is simply the fact that I performed some individual action. This is not quite correct. I do not have a joint right in virtue of performing an action that is describable independently of the actions of the other agents. For example, I do not have a right to use a house we jointly built simply in virtue of the fact that I built one of the walls. Rather, I have a right in virtue of having contributed to the building of the house by building that wall. My wall building considered by itself would not give rise to a 218
7. Collective Rights right in respect of the house. For there might not be a house, but only my wall. Second, what is the relationship, if any, between joint rights and collective goods? The good consists in the first instance in a collective end in respect of which the participants possess a joint right. Accordingly, even if the good considered in itself is not a collective good, each of the individual (jointly held) rights to it is a right to it qua collective good. For example, using two boats and a single large net we could jointly catch a hundred fish. By prior agreement we could possess individual rights to fifty fish each. But this agreement is something additional to the joint right, and indeed presupposes it. Imagine that unexpectedly the good produced could not be parcelled out in the manner envisaged in the agreement. Perhaps we caught only one very large fish, or no fish, but instead a rare and valuable old ship. If so, each individual could still claim an individual (jointly held) right to the good, and therefore legitimately insist on making some different agreement (or perhaps no agreement). Thus far my discussion of jointly held rights to jointly produced goods has implicitly assumed that the good produced is not constituted by the actions that produce it. But there are joint actions the collective end of which just is those joint actions themselves.23 And of these, some are cases in which the collective end is a good. Thus two people dancing together may have as a collective end, in itself, simply that they dance together. Here the good aimed at is just the joint activity that realises the good. By contrast, a team of street sweepers may have as a collective end in itself that the streets are clean, but not have their joint activity of sweeping as an end in itself. Here the good aimed at – the clean streets – is not constituted by the joint activity that brings it about. Let us now consider goods that consist in the joint activity undertaken to produce them. Re´aume has given consideration to such goods.24 She claims that such goods are necessarily collective in the sense of goods such that an agent that contributes to their production cannot be excluded from their enjoyment, either by another agent or by the agent himself or herself. She gives, as an example, friendship. If Fred enjoys the friendship of Mary, then Mary enjoys the friendship of Fred. Other examples are political participation and cultural participation. Let us call such goods participatory goods. 219
Social Action As Re´aume points out, the examples of friendship and culture are problematic as instances of goods in respect of which persons have collective rights, since it is by no means clear that there is a right to such goods.25 For such cases involve the production of a benefit that can only be enjoyed by those who contribute to the production. There are no merely instrumental individual actions to a collective end; rather, the joint action is the end. But in that case there is no point to claiming a right to enjoy the good, having participated in the activity; having participated, a person has necessarily enjoyed the good.26 The production of participatory goods – as distinct from the enjoyment of these goods, having participated in their production – involves moral rights. Specifically, there are rights in relation to participation, and these are conceivably of two sorts. First, there are rights to participate. These can be held against those who are already participating, or against outsiders seeking to prevent an individual from participating, or indeed prevent everyone from participating. Second, there are rights to enforce participation.27 Are either of these kinds of right collective rights? Rights to participate are joint rights. I have a right not to be prevented from participating, and I have this right jointly with you. Rights to enforce participation are more problematic. There are some joint rights to enforce participation. For example, the members of the community may have the joint right to enforce participation in the war effort. If we assume that the war effort is a good then the war effort is a participatory good. Presumably the war effort is only an instrumental, and not an intrinsic, good. Perhaps parents, and even the members of a community, have a joint right to enforce the participation of children in educational programs.28 If so then this is a joint right to enforce participation in a participatory good that is also an intrinsic good. At any rate, if my joint right to participate (or to enforce participation) is in part based on membership of a social group, then I suggest the joint right is a collective right. If the joint right is based entirely on properties other than membership in a social group, then it is not a collective right. So political participation is a collective right. What of other participatory goods mentioned by Re´aume, such as friendship and culture? Certainly culture and even friendship have
220
7. Collective Rights been claimed to be collective rights. However, I suggest that they are not, since the right not to be excluded from participation in these goods is not even in part based on membership of a social group. Moreover, I am not convinced that there are rights of enforcement in relation to these participatory goods, other than in special cases, such as the education of children. Friendship, communication, intellectual inquiry, artistic expression, and spiritual expression appear to be human rights, not collective rights, in the sense that they are not possessed by groups or individuals qua members of groups. Moreover, they are rights to participate, and also rights not to be forced to participate. In denying that such rights are collective I do not mean to imply that they are necessarily individual rights, at least in the sense of rights a solitary individual can possess and enjoy simply in virtue of properties he or she possesses as an individual. Rather, many such rights are a species of joint right, which I will refer to as interpersonal rights. They are rights individuals possess in virtue of needs they have to form relations, or engage in activities, with other individuals.29 Moreover, the enjoyment of such rights is necessarily joint enjoyment. If A is enjoying his right to have a friendship with B, then B is necessarily enjoying B’s right to have a friendship with A. Further, they are not rights held against the other individuals with whom they may form relationships or engage in joint activity. Rather, they are rights held against third parties who might seek to interfere in these relations or joint activities. Sometimes these third parties are themselves participants with some of the participants. For example, John may try to prevent his friend Mary from being friends with Jane. Importantly, for our purposes here, such interpersonal (joint) rights in relation to participatory goods are not collective rights in the sense of being rights possessed in virtue of membership of a social group. They are basic human rights to form relations and engage in joint activities with other humans by mutual consent. Consider the attempt of the Eastern European communist regimes to prevent intercultural intellectual interaction between academics in Eastern Europe and academics in the West. Consider, also, the policy of the Taliban in Afghanistan not only to force women not to show their faces in public – they must wear veils – but
221
Social Action also to restrict their individual freedom to form friendships, especially friendships with males who are not members of their family, or who are not approved of by their fathers or husbands. Here we need to distinguish the exercise of basic interpersonal human rights from the organisations and institutions which might claim to embody them. In some cases such claims seem false. Consider, for example, the so-called cultural imperialism of large Western media and entertainment corporations. These corporations are able to ensure that the cultural material produced by themselves is dominant in domestic, international, and intercultural communication. But it is by no means clear that this material should be considered to be an expression of the culture of a people, as opposed to the consumer product of a few organisations. In other cases an institution might seek to impose its “culture” on individuals. Such imposition looks to be in violation of human rights. Notwithstanding these examples, there appear to be some special cases in which imposition of participatory goods on individuals might be justifiable. The most obvious of these is the imposition of education on children. Other cases might be ones involving institutions that have been overpowered by external, or indeed internal, coercion, and are in need of special protection in the short term in order to rebuild themselves. Perhaps the Russian Orthodox Church is an example of this. Such imposition is at least in principle permissible in virtue of the fact that human rights, including human rights in relation to participatory goods, are not absolute. Thus far we have discussed two kinds of joint rights in respect of goods that have been jointly produced – participatory and nonparticipatory – and I have argued that such rights are collective if they are rights possessed in part by virtue of membership of a social group, but not if they are not. However, there are at least some joint rights to goods which have not been jointly produced. These include joint rights to assistance or protection.30 The category of minority rights evidently includes rights to goods that have been jointly produced and goods that have not been jointly produced. Kymlicka evidently has both categories in mind in his discussions of minority rights.31 However, as we will see, he has a particular concern with collective rights to assistance or protection.
222
7. Collective Rights minority rights The issue of minority rights arises in relation to communities within communities, such as Native Americans or Aboriginal Australians. Sometimes it is argued that these minority groups have rights that all groups have. For example, right-wing Afrikaners have argued that each group has the right to exclude any other group from living in a certain area. At other times the claim is that minority groups have rights that the other members of the larger community do not have. For example, Native Americans, but not other Americans, might have rights to own property within traditional tribal land areas. Or Native Americans might have voting rights in local elections, even though these rights are denied to non-Native American transient workers. These alleged minority rights are paradigmatic collective rights. What is the argument for the existence of these rights? One of the most compelling arguments for the existence of minority collective rights is that propounded by Kymlicka.32 He points out that a social group’s cultural structure – its institutions, such as its system of education and language – is a necessary condition for the individual freedom and well-being of members of that group. Specifically, societal culture enables “meaningful individual choice and supports self-identity.”33 However, the cultural structures of many (but not all) minority groups, such as Native American, are under threat in a variety of ways from the majority social groups in which these minorities find themselves. Therefore, the freedom and self-identity of members of minority social groups is under threat. At least part of the solution, according to Kymlicka, is to grant minority rights, such as local voting rights, land rights, and language rights. For such minority rights protect the cultural structure of minority social groups, and thereby enable the possibility of individual members of those social groups possessing freedom and self-identity. Kymlicka’s argument relies in part on empirical claims concerning, for example, the effectiveness of minority rights in protecting cultural structures. If we grant him these then I believe the argument is plausible. Nevertheless, it is important to be clear on what the argument has demonstrated, and what it has not demonstrated. First, the argument overlooks, or at least significantly understates, 223
Social Action the extent to which past and present domination and exploitation of a cultural group undermines the self-identity of individual members of that cultural group by undermining in the first instance their self-worth. Self-worth – and not simply protection of cultural structures – is fundamental to the maintenance of self-identity. Second, the argument, if it succeeds, merely demonstrates that, under certain contingently existing conditions, it is desirable to put certain legal minority rights in place. It does not demonstrate the existence of minority moral rights as such. Here I assume that moral rights are not equivalent to legal rights, and not even equivalent to legal rights that ought to be enacted. During the apartheid era South African police routinely shot dead black South Africans who were protesting against being forcibly removed from their homes, having their freedom of movement (especially in relation to job seeking) restricted, and so on. On such occasions the moral rights of black South Africans were violated, notwithstanding the fact that the police were often acting entirely within their legal rights and the black South Africans acting illegally. Moreover, it is completely unconvincing to claim that the possession or exercise of these moral rights is equivalent to an as yet unrealised, but desirable, legal state of affairs. What the legal state of affairs is, or ought to be, is not the substantive issue. The substantive theoretical issue concerns the moral grounds – in the shape of moral rights or moral values of some other kind – for the black South Africans’ correct belief that they ought not to be forcibly removed from their homes, shot dead when they peacefully demonstrate, and so on. The substantive practical issue is how to ensure that these moral rights (or values) are not violated. That probably involves enacting legal rights, but not necessarily. Perhaps it simply calls for revolutionary intervention to protect these peoples rights, followed by the destruction of the state – including the whole system of laws – in order to establish an anarcho-socialist moral community in which moral rights are respected, but not by way of legal rights.34 On Kymlicka’s account, minority legal rights ought to be enacted under two conditions. First, doing so is the only way to maintain certain institutional structures. Second, maintaining these institutional structures is the only way to guarantee an individual moral rights, namely the moral rights to freedom and self-identity. Here we have minority legal rights in the service of an individual moral 224
7. Collective Rights right, but we do not have minority moral rights. (I do not claim that Kymlicka takes himself to have demonstrated the existence of special moral rights, as opposed to the desirability of enacting special legal rights.) Moreover, it is by no means clear, as Kymlicka appears to assume, that we can read off in any straightforward way what those legal minority rights ought to be, given the individual moral rights, and given some specific set of institutional structures that enable the possibility of individual freedom and self-identity. For one thing, those existing institutional structures may be doing a bad job; perhaps the institutional structures could be adjusted, or even substantially changed, in order to better facilitate individual freedom and self-identity. For another, there will inevitably be other moral considerations in play, considerations other than these individual moral rights of the members of the minority. For example, some legal provision for Australian Aboriginal land rights looks to be morally justified, but not to the extent of disabling the entire national economy. Third, as Taylor points out, it is only an argument for the protection of existing people, not for the protection of future generations of minority social groups.35 For presumably, future generations could possess freedom and self-identity if they were inducted into a culture other than that of the existing people. But typically the concern of minority groups is with the survival of their culture, and not simply with the protection of the existing people. Indeed, arguably Taylor’s point can be extended to the current generation of members of at least some contemporary cultural communities. For in some cases – perhaps the French community in Quebec is a case in point – it might not be the case that the erosion of their cultural structures would have significant negative impact on their individual freedom and self-identity. Given the similarity and the porous nature of contemporary French culture vis-a`-vis contemporary English culture (and vice-versa), it might be that the gradual and orderly replacement of French cultural institutions by English ones will leave the freedom and self-identity of the current members of the French-Canadian community in reasonable shape. Even if this is so, the French-Canadians may well still have a moral right to maintain their culture. What this line of reasoning demonstrates is that French-Canadian cultural rights are probably not 225
Social Action grounded, or not only grounded, or not necessarily grounded in their rights to individual freedom and self-identity. The argument for the existence of collective moral rights – as distinct from legal rights – must I think proceed on two assumptions. First, all social groups, and not simply minority groups, have these rights. For if this is not so it is hard to see how they could be moral rights, as distinct from legal rights in the service of some morally desirable goal. Second, these rights will have to be such that they have as their main point the survival of the community’s existing way of life, and not simply the well-being of individual members of the community. What kinds of collective rights could satisfy these two conditions? Two putative kinds of collective rights that suggest themselves are the right to collective self-defence, and the right to political selfdetermination. These two putative rights are salient because the whole point of the movement in favour of minority rights is to ensure that a community and its way of life survives, and survives as a self-determining community. But in relation to other social groups this fundamentally means that the community and/or its culture is not attacked, eroded, or otherwise destroyed by external forces, and as well, that it determines its own future. In respect of the first requirement we can distinguish between self-defence in the sense of defence of the lives of the members of the community, and self-defence in the sense of the defence of the social forms definitive of the community, including its fundamental institutions, such as its language, kinship system, and so on. Call this cultural self-defence. Let us look at each of these collective rights. Self-defence in the first sense is presumably based on the individual right to life of the members of the community. Each individual has a right to life, and therefore an individual right to self-defence. In addition, individuals have at least an obligation to protect the lives of their family and friends. Further, the members of a community have a joint obligation to protect the lives of members of the community from other members of the community and from external threats. Or at least they have a joint obligation to do so at some threshold level of threatened lives. Perhaps in complex societies there is a kind of moral division of labour in terms of which this joint obligation can be discharged by means of police services and the armed forces.36 226
7. Collective Rights Is this joint obligation a collective duty? It is, if the joint obligation is in part based on membership of the community. Whether and to what extent this joint obligation is based on membership of the community is a difficult issue, which I cannot settle here. Suffice it to say that there are joint obligations to protect the lives of members of communities other than one’s own. For example, there was a joint duty on the part of the leadership of the various members of the United Nations to intervene militarily in Rwanda to protect the lives of Tutsis.37 A consequence of this is that – contra Raz and Re´aume – there might well be individual rights to collective goods, and specifically individual rights to a jointly produced collective good of protection and/or assistance.38 Let us concern ourselves in what follows with the rights to selfdetermination and to cultural self-defence.39 The right to selfdetermination can be cashed as the joint right of members of a social group to political participation.40 This right is collective since it is based in part on membership of the social group. Non-members do not get to participate. Moreover, it is a right held against members and non-members alike. On this view the right to selfdetermination brings with it the individual right to political participation. So on this account there is no collective right to selfdetermination for totalitarian states. What of cultural self-defence? I have already argued that some cultural rights – namely, rights in respect of participatory goods – are joint rights, but are not collective rights. However, these joint rights might be thought to generate at least two sorts of collective rights, namely collective rights to protect the exercise of cultural rights and collective rights to support them.41 These rights are collective, for ought implies can, and individuals, and indeed external societies, are typically not able to adequately protect and support the joint right of members of a whole community. If this is correct then the collective rights of members of a community in relation to the (joint) cultural rights of their members is analogous to the right of members of communities to protect and support their individual rights, including their various individual freedoms. Indeed, these collective rights of the members of the community – both in relation to the joint cultural rights and various individual rights – might in turn justify the establishment of organisations and occupations the designated function of which was to 227
Social Action support and protect particular cultural and individual rights. Perhaps police services are an example in relation to individual rights to property, personal security, and such. Perhaps publicly funded arts bodies are an example in relation to joint rights to artistic expression. Since collective rights to support and protect cultural rights derive from these cultural rights, that is derive from the joint rights to participatory goods, then the former cannot be invoked for the purpose of justifying the infringement of the latter. What are examples of these collective rights? Suppose it turned out that the exercise of the joint right of the members of a community to, say, produce their own culturally specific films was unable to be exercised because of lack of private funding and sales revenue. If so, then the collective right of the members of the community to support these cultural rights could be discharged via public funding. Or suppose the French-speaking community of France was concerned about the standard of French in the community and/or worried about the extent to which extraneous linguistic elements were creeping into ordinary usage, as a result of, say, large numbers of foreign tourists. Perhaps in these circumstances the members of such a community have a collective moral right to launch educational programs in schools and workplaces to promote the French language. Again, suppose a large American media corporation wanted to provide pornographic films by satellite to a Muslim community. Suppose such films were deeply offensive to the religious sensibilities of members of the community, and there were concerns that children would view the films, and concerns also on the part of the adults themselves that they might end up viewing them. Presumably the members of the community have a collective right to protect the exercise of their cultural rights by disallowing the media corporation to communicate its films. Note that such collective rights might generate collective legal rights. Perhaps the French community will need to pass a law providing for a legal right to all French mother tongue speakers to have access to French-language courses. So far so good. However, we have thus far not found it necessary to postulate minority (collective) moral rights. There are other 228
7. Collective Rights claimed collective moral rights in relation to goods that are not participatory goods; these are goods that are not produced by those who have a right to them. Such rights might include communal rights in relation to curbing the procreation of children in a context of overpopulation; the use of land under communal systems of ownership; placing strictures on the joint right of two individuals to have sexual relations, given well-motivated social taboos on intercourse between, say, fathers and daughters.42 Evidently, the existence of such collective moral rights might be thought to translate into minority moral rights in the case of minority social groups. I cannot pursue here the details of each of these and other specific (alleged) minority rights. What I can do is address the kind of in-principle problem raised by minority rights. The problem arises when there is more than one cultural group within a social group, and specifically in communities with minorities. The problem can be posed in the form of a dilemma. The first possibility is that although you are a minority, you are a genuine community. You are possessed of a distinctive culture, a common ancestry, a territory to which you have a right, and so on. If so, you have a right to self-determination, and therefore a right to secede. Alternatively, you do not have a sufficiently distinctive culture and/ or common ancestry, and so on, and therefore have no right to selfdetermination. If so, you are not really a genuine community, and therefore not a community within a community, but, rather, at most a regional or class or special interest group within a single community. Surely in the setting up of this dilemma the notion of a community has been conceived of too tightly and has excluded too many groups that are, properly speaking, communities. We can have, and in fact do have, overlapping communities, such that a group of individuals exercises some of its collective rights in one community, and the remainder of its collective rights in another community (or communities). In effect, this is what many multicultural societies and federations of communities propose, and there are viable societies of this kind. These evidently include Belgium, Switzerland, and Canada. If this is the model, then perhaps there are no minority rights as such; rather, a given set of individual persons can exercise different collective rights in different communities. For example, the Chinese 229
Social Action community in Australia can exercise its collective right to support Chinese culture by paying taxes, which in turn translates into Australian government support for the government-funded multicultural television station. Again, Aboriginal people in Australia can exercise some of their political rights in their indigenous political institutions, and others in political institutions in which all Australians participate. It might be argued that this way of describing the situation will not do for some minority rights. For in some cases a minority possesses the same rights as the majority, as well as an additional set of rights not possessed by the majority. Let us consider some cases. Consider the case of a guaranteed number of seats in government for some minority.43 Apparently, the minority has a special right denied to others. Certainly this is an additional legal right. But it is hard to see how it embodies an additional moral right. Rather, it is an institutional device put in place on the basis of the recognition that there are different interest groups within societies, and that simple majority voting systems can enable interest groups with large numbers to exercise too much power relative to interest groups with small numbers, and that therefore there needs to be counterbalances to this built into the electoral system. One such counterbalance is giving a minority interest group a guaranteed number of seats, another is proportional representation, still another is an upper house numerically reflective of the number of states (in a federation) rather than population size. In short, we have a non-moral legal right in the service of consequentialist goals. Let us consider land rights. Apparently, some Australian Aboriginals have communal land rights – including rights of exclusion – in virtue of native title legislation, as well as the individual property rights enjoyed by other Australians. However, Aboriginal people have been granted these legal rights in virtue of the process of dispossession of land that Aboriginal people suffered at the hands of the colonisers of Australia.44 Aboriginals have these legal rights to land under native title legislation in virtue of being the original occupants of the land in question, and of having inhabited it continuously. So they now enjoy two sorts of land rights (legal rights) in virtue of the different systems of land “ownership” that the different cultures deploy. However, from the point of view of moral rights this is a mere legal technicality. If one social group dispossesses a second 230
7. Collective Rights social group and later gives some recognition and compensation for this dispossession by enacting land rights legislation, then we have a situation of violation of a moral right followed by recognition and compensation. We do not have the existence of some additional special moral right. To see this, assume that such recognition and compensation was accompanied by an insistence that the members of the first group had to forfeit their legal rights to whatever additional land they might have bought from those who dispossessed them subsequent to the act of dispossession and forfeit also any legal right to buying any more of this land in the future. Assume also that the justification offered for this was the moral unacceptability of a group of individuals having two sets of moral rights when other groups only had one set. I take it that such an argument would be morally outrageous. The fact that I have a moral right to recognition and compensation in virtue of some past violation of my rights is hardly equivalent to the enjoyment of a special moral right not enjoyed by others who did not have their rights violated. Consider now the example of government support for the language rights of a minority.45 According to our earlier account, all communities have a (derivative) collective duty to support their cultural rights, including language rights. Accordingly, in multicultural states there may need to be, for example, more than one official language. But I am not thinking here of states in which there are, or ought to be, two or more official languages. In such states, the minority social group (or groups) does not have a special (legal) right that the majority group does not have.46 Rather, I have in mind a state in which the minority happily receives instruction and communications in the official language of the nation state to which it belongs in the same way that everyone else does, and, in addition, its own language is supported in schools, and so on. So the minority has a benefit that no one else has. That is, the minority social group is having it both ways.47 If such a minority language is under attack this may warrant special legal rights to, so to speak, thwart the attack, but legal rights that stop short of ensuring the survival of the language. For example, suppose extremist ethnocentrists in some public schools and universities refused to allow some minority language to be an op231
Social Action tional part of the curriculum, notwithstanding the fact that a critical mass of students and their parents wanted the language to be included in the curriculum, the relevant teachers and other resources were available, and so on. Special legislation might need to be enacted to nullify such attacks. But this amounts to special legal rights to support a non-special moral right. It is frequently pointed out that attacks on a culture, including a language, are more subtle and insidious than this. For example, the personnel in key institutions might not speak the language and there may be no provision for languages other than the dominant one.48 Accordingly, a process of linguistic erosion takes place. Indeed, the threat to a culture from such processes lies at the heart of the concerns of Canadian policy makers and theorists such as Kymlicka. Here distinctions would need to be made between “natural” erosion, so to speak, and attacks in the sense of morally unacceptable erosion. Moreover, we need to keep in mind that these distinctions are now being made in the context of a prior decision – and a morally justifiable one – that the language in question not be given full official status as a language of public communication. At any rate, as I argued above, the need for provision of special legal rights to protect moral rights that are under attack is not the same thing as the existence of special moral rights. Once again I cannot see that the existence of such special legal rights brings with it a commitment to special moral rights. For one thing, such legal rights might take different forms at different times, and even go in and out of existence, depending on the unfolding nature and continuation of the threat. Accordingly, they do not look very much like moral rights. On the other hand, if a language is not under attack, but is simply dying through a lack of interest on the part of the members of the relevant linguistic group, then there is no right of intervention to preserve it. Suppose the (artificial) language, Esperanto, was dying but is now undergoing a revival due to the increase in interest on the part of some social group. If the members of this social group had not been prepared to support this language in the past, surely the government did not have a duty to ensure its survival. What of the situation now? Now that the Esperanto speakers have decided that they want to revive Esperanto – but not abandon English – do they have a moral right to government support for Esperanto as well 232
7. Collective Rights as English? Given that such support relies on the taxes of the nonEsperanto-speaking majority – who are to see no benefit themselves – it is hard to see what the basis would be for this alleged moral right. Naturally, if the minority language in question had suffered suppression at the hands of the conquerors of the social group in question then matters would be different. Consider, in this connection, the Gaelic language in the Highlands of Scotland. The Scots were conquered by the English, who pursued a policy of suppression of the Gaelic language, including making it unlawful for Gaelic to be used as the medium of instruction in schools in Scotland. However, Gaelic never died out, and recently it has undergone a revival. Moreover, it is no longer unlawful for Gaelic to be used as a medium of instruction in schools in the Highlands. Are the Gaelic-speaking Highlanders entitled to support in relation to both the English and the Gaelic languages? I suggest that they are. So Gaelic speakers are entitled to two sorts of language rights (legal rights). However, this does not show that Gaelic speakers have two sets of moral rights, any more than our earlier discussion of Aboriginal land rights showed that Aboriginal Australians had two sets of moral rights in relation to land ownership. If one social group dispossesses a second social group of its language and later gives some recognition and compensation for this dispossession by enacting language rights legislation, then we have a situation of violation of a moral right followed by recognition and compensation. We do not have the existence of some additional special moral right. In this chapter I have argued that many collective moral rights are in reality joint rights to collective goods possessed in part in virtue of membership in a social group. I have also argued that there are probably no minority moral rights. The alleged minority moral rights considered turned out to be either minority legal – but not moral – rights, or moral rights, but moral rights possessed by all members of social groups, and therefore not minority moral rights. My discussion has relied on the notion of a joint action developed in Chapter 2, and has made extensive use of the notion of a social group – a notion I defined in Chapter 6. In the following, and final, chapter of this book I will consider the correlative issue to that of collective rights, namely collective responsibility.
233
8
Collective Responsibility
In this final chapter I will offer an account of the notion of collective moral responsibility. I will argue that collective moral responsibilities are joint responsibilities of individual human persons. This conception of collective responsibility as joint responsibility presupposes, and is heavily reliant on, the notion of joint action that I presented in Chapter 2. So the unifying and pivotal role of my notion of joint action in this philosophical study of social action continues to the last. Further, I will argue that collective rights, as defined in Chapter 7, typically generate collective moral responsibilities on the part, not only of other collectives, but also of the bearers of the collective rights themselves. For example, Russian citizens have a collective right to be free of economic domination by organised crime groups. If so, this collective right generates two separate sets of collective responsibilities. It generates a collective responsibility on the part of Russians to put in place law enforcement mechanisms to curtail such domination. It also generates a collective responsibility on the part of members of law enforcement agencies outside Russia to assist the Russians in relation to this crime problem; or at least it does so, given that they are able to assist them, and given that the Russians cannot deal with their crime problem by themselves. Along with rights go responsibilities. Along with collective rights go collective responsibilities. Before providing these slogans with some content, we need first to be clearer on the notion of (moral) responsibility and then on the notion of collective (moral) responsibility. moral responsibility We hold people responsible for actions, and we also hold them responsible for omissions. I will consider responsibility for omissions 234
8. Collective Responsibility in a later section in this chapter. For the moment let us consider responsibility for actions only. We need first to distinguish some different senses of responsibility. Sometimes to say that someone is responsible for an action is to say that the person had a reason, or reasons, to perform some action, then formed an intention to perform that action (or not to perform it), and finally acted (or refrained from acting) on that intention, and did so on the basis of that reason or reasons. Note that an important category of reasons for actions is the one encompassing ends, goals, and purposes; an agent’s reason for performing an action is often that the action realises a goal the agent has. Moreover, it is assumed that in the course of all this the agent brought about or caused the action, at least in the sense that the mental state or states that constituted his or her reason for performing the action was causally efficacious (in the right way), and that his or her resulting intention was causally efficacious (in the right way).1 This sense of being responsible for an action is different from other cognate senses of responsibility. These latter cognate senses include the notion of a responsible agent – as distinct from someone being responsible for an action. An agent is a responsible agent if he or she is not insane, is not under the influence of drugs, and so on.2 I will dub the very first of these senses of being responsible for an action “natural responsibility.” It is this sense of being responsible – intentionally performing an action and doing so for a reason that I will be working with in this chapter.3 On other occasions what is meant by the term “being responsible for an action” is that the person in question occupies a certain institutional role, and that the occupant of that role is the person who has the institutionally determined duty to decide what is to be done in relation to certain matters. For example, the mechanic has the responsibility to fix the brakes on my car when it goes in for a service, irrespective of whether or not the mechanic does so, or even contemplates doing so. A third sense of “being responsible” for an action is a species of our second sense. If the matters in respect of which the occupant of an institutional role has an institutionally determined duty to decide what is to be done include ordering other agents to perform, or not 235
Social Action perform, certain actions, then the occupant of the role is responsible for those actions performed by those other agents. We say of such a person that he or she is responsible for the actions of other people in virtue of being in charge of them or of being the person in authority over them. The fourth sense of responsibility is in fact the sense that we are principally concerned to elucidate, namely moral responsibility. Roughly speaking, an agent is held to be morally responsible for an action if the agent was responsible for that action in one of our first three senses of being responsible, and that action is morally significant.4 The ways in which an action can be morally significant are too many and varied to be detailed here. They include instances of infringing or conforming to a moral principle or right, causing great good or evil, and being motivated by a moral emotion. Suffice it to say here that this notion is relative to one’s favoured moral theory or, at least, favoured moral theoretical assumptions. By my theoretical lights, the notion of moral significance is relatively narrow, and needs to be contrasted with wider notions, such as the ones expressed in phrases like, “it was a good or bad thing to do,” or “it ought or ought not to be done.” Indeed, I hold the view that most actions of most people are without moral significance, in this sense of moral significance. So I do not accept certain versions of utilitarianism, which are committed to the view that all or most of our actions are morally significant because they contribute causally to the sum of pain or pleasure, or happiness or unhappiness, or the satisfaction or dissatisfaction of desire of someone or other at some stage or other. Moreover, the notion of moral responsibility has stronger and weaker forms, and the “actions” in respect of which one might be morally responsible admit of gradations. In relation to stronger and weaker forms consider that one might be less morally responsible for an unpremeditated act of wrongdoing than for a premeditated one, or less responsible for an outcome that was a foreseen consequence of one’s action, but not intended. In relation to gradations of “action,” consider a person who fires a bullet at a business competitor with the intention of killing that person. Coincidentally the person suffers a terminal heart attack a split second before the bullet arrives.5 Given the notion of natural responsibility on which the 236
8. Collective Responsibility above definition of moral responsibility in part relies, the shooter is not morally responsible for the death of the person. This is so notwithstanding the fact that the shooter would have killed the competitor if the competitor had not suffered a heart attack. On the other hand, it is consistent with the above definition to hold that the shooter is morally responsible and blameworthy for a closely related action, namely attempted murder. It is also consistent to hold that the shooter is morally responsible for shooting the competitor. It is even consistent to hold that the shooter is almost as bad, as he or she would be if the competitor had not suffered a heart attack. To those who hold that the shooter is just as bad, I would reply that what one actually does counts for more (even if sometimes not much more) than what one would have done, but for luck. At any rate we can make our first preliminary claim concerning moral responsibility. (1) If an agent is responsible for an action in the first, second or third senses of being responsible, and the action is morally significant, then – other things being equal – the agent is morally responsible for that action, and can reasonably attract moral praise or blame and (possibly) punishment or reward for performing the action.
Here the “other things being equal” clause is intended to be cashed in terms of exculpatory conditions, such as that he was not coerced, he could not reasonably have foreseen the consequences of his action, and so on. If a person is responsible for some action in the second or third senses of being responsible, then typically that person is institutionally accountable for the action. To be institutionally accountable is to be subject to a requirement to explain and justify one’s actions to some institutional body or person. Sometimes institutional accountability brings with it institutional, including legal, liability; and an adverse judgment on the part of those to whom one is accountable can result in the infliction of punishment. Having distinguished four senses of responsibility, including moral responsibility, and distinguished each of these from accountability and liability, let me now turn directly to collective responsibility.6 237
Social Action collective moral responsibility As is the case with individual responsibility, we can distinguish four senses of collective responsibility. In the first instance I will do so in relation to joint actions. Let us remind ourselves what a joint action is. Roughly speaking, two or more individuals perform a joint action if each of them intentionally performs an individual action, but does so with the true belief that in so doing he or she will jointly realise an end that each of them has. As we saw in Chapter 2, having an end in this sense is a mental state in the head of one or more individuals, but it is neither a desire nor an intention. However, it is an end that is not realised by one individual acting alone. So we have called such ends, collective ends. For example, you and I lifting a refrigerator onto a truck is a joint action, since you lift one side and I the other, each of us lifts one side truly believing the other will lift the other, and each of us has as an end that the refrigerator be situated on the truck. Agents who perform a joint action are responsible for that action in the first sense of collective responsibility. Accordingly, to say that they are collectively responsible for the action is just to say that they performed the joint action. That is, they each had a collective end, each intentionally performed his contributory action, and each did so because each believed the other would perform his contributory action, and that therefore the collective end would be realised. Here it is important to note that each agent is individually (naturally) responsible for performing a contributory action, and responsible by virtue of the fact that the agent intentionally performed this action, and the action was not intentionally performed by anyone else. Of course the other agents (or agent) believe that the first agent is performing, or is going to perform, the contributory action in question. But mere possession of such a belief is not sufficient for the ascription of responsibility to the believer for performing the individual action in question. So what are the agents collectively (naturally) responsible for? The agents are collectively (naturally) responsible for the realisation of the (collective) end that results from their contributory actions. In our refrigerator-lifting example, each is individually (naturally) responsible for lifting one side of the
238
8. Collective Responsibility refrigerator, and the two agents are collectively (naturally) responsible for bringing it about that the refrigerator is situated on the truck. Further, on my account, to say that they are collectively (naturally) responsible for the realisation of the collective end of a joint action is to say that they are jointly responsible for the realisation of that end. They are jointly responsible because: (a) each relied on the other to bring about the state of affairs aimed at by both (the collective end); and (b) each performed his or her contributory action on condition, and only on condition, the other or others performed theirs. Here, condition (b) expresses the interdependence involved in joint action. Moreover, it needs to be noted that this interdependence is relative to a collective end. It might be that one (or more) of the agents has some other end, and that the agent also performs the action he or she performs to realise this second end. For example, assume a team of burglars steals a work of art from an art gallery, and that in so doing one burglar turns off the security system before another smashes the protective glass and seizes the work of art. Here, each has as a collective end the theft of the work of art. But perhaps the first burglar had an additional end; perhaps this burglar intended to turn off the security system having as an end to put his or her technical ability to the test. In such cases an agent would have performed the contributory action even if the other agents did not perform theirs. The presence of such additional ends existing side by side with the collective end does not remove the interdependence of action definitive (in part) of joint action. For even in these cases, each agent performs his or her contributory action (qua contributory action) if and only if the others perform theirs relative to the collective end. Again, if the occupants of an institutional role (or roles) have an institutionally determined obligation to perform some joint action, then those individuals are collectively responsible for its performance in our second sense of collective responsibility. Here, there is a joint institutional obligation to realise the collective end of the joint action in question. In addition, there is a set of derived individual obligations; each of the participating individuals has an individual obligation to perform his or her contributory action qua contrib-
239
Social Action utory action. (The derivation of these individual obligations relies on the fact that if each performs his or her contributory action then it is probable that the collective end will be realised.) Notice that typically agents involved in an institutional joint action will discharge their respective individual institutional obligations and their joint institutional obligation by the performance of one and the same set of individual actions. For example, if each of the members of a police task force performs his or her individual duties, having as an end the smashing of a drug ring, then, given favourable conditions, the task force will smash the drug ring. But one can imagine a detective who recognises his or her individual institutional obligation, but not his or her jointly held obligation to realise the collective end in question. This detective might have an overriding individual end to be promoted; but the head of the task force might be ahead of the detective in the queue of those to be promoted. So the detective does not have smashing the drug ring as a collective end. Accordingly, while the detective ensures that he or she discharges his or her individual obligation to (say) interview a particular suspect, the detective is less assiduous than he or she might otherwise be because he or she wants the task force to fail to smash the drug ring. There is a third sense of collective responsibility that might be thought to correspond to the third sense of individual responsibility. The third sense of individual responsibility concerns those in authority. Here we need to distinguish two kinds of cases. If the occupant of an institutional role has an institutionally determined right or obligation to order other agents to perform certain actions, and the actions in question are joint actions, then the occupant of the role is individually (institutionally) responsible for those joint actions performed by those other agents. This is our first kind of case; but it should be set aside, since it is not an instance of collective responsibility; the person in authority does not have collective responsibility. In the second kind of case it is of no consequence whether the actions performed by those under the direction of the person in authority were joint actions or not. Rather, the issue concerns the actions of the ones in authority. In what sense are they collective? Suppose the members of the Cabinet of country A (consisting of the Prime Minister and the Cabinet Ministers) collectively decide to 240
8. Collective Responsibility exercise their institutionally determined right to order their army to attack country B during peacetime. The army does what it was ordered to do. The Cabinet is collectively responsible for starting the war in some sense of collective responsibility. There are a couple of things to keep in mind here. First, the notion of responsibility in question here is, at least in the first instance, institutional – as opposed to moral – responsibility. Second, the “decisions” of committees, as opposed to the individual decisions of the members of committees, need to be analysed in terms of the notion of a joint institutional mechanism, introduced in Chapter 5. So the “decision” of the Cabinet can be analysed as follows. At one level each member of the Cabinet voted for or against the army attacking country B; and let us assume some voted in the affirmative, and others in the negative. But at another level each member of the Cabinet agreed to abide by the outcome of the vote; each voted, having as a collective end that the outcome with a majority of the votes in its favour would be pursued. Accordingly, the members of the Cabinet were jointly institutionally responsible for the decision to order the army to attack country B. So the Cabinet was collectively institutionally responsible for starting the war; and the sense of collective responsibility in question is joint (institutional) responsibility.7 What of the fourth sense of collective responsibility, collective moral responsibility? Collective moral responsibility is a species of joint responsibility. Accordingly, each agent is individually morally responsible, but conditionally on the others being individually morally responsible; there is interdependence in respect of moral responsibility. This account of collective moral responsibility arises naturally out of the account of joint actions. It also parallels the account given of individual moral responsibility. Thus we can make our second preliminary claim about moral responsibility. (2) If agents are collectively responsible for the realisation of an outcome, in the first or second or third senses of collective responsibility, and if the outcome is morally significant, then – other things being equal – the agents are collectively morally responsible for that outcome, and can reasonably attract moral praise or blame, and (possibly) punishment or reward for bringing about the outcome.
241
Social Action Here we need to be more precise about what agents who perform morally significant joint actions are collectively morally responsible for. Other things being equal, each agent who intentionally performs a morally significant individual action has individual moral responsibility for the action.8 So, in the case of a morally significant joint action, each agent is individually morally responsible for performing his or her contributory action, and the other agents are not morally responsible for his or her individual contributory action. But, in addition, the contributing agents are collectively morally responsible for the outcome or collective end of their various contributory actions. To say that they are collectively morally responsible for bringing about this (collective) end is just to say that they are jointly morally responsible for it. So each agent is individually morally responsible for realising this (collective) end, but conditionally on the others being individually morally responsible for realising it as well. So, in the example of the burglars, burglar A might be individually morally responsible for turning off the security system in the art gallery and burglar B individually morally responsible for smashing the glass shield protecting the work of art. However, A and B are jointly morally responsible for the theft of the work of art. Having thus elaborated what I take to be a, if not the, central notion of collective moral responsibility, I now want to turn to a number of problems that arise in relation to collective moral responsibility for actions. The first of these concerns overdetermination, and the second concerns the actions of members of organisations. In the final section of this chapter I will consider collective moral responsibility for omissions. Overdetermination The first puzzle arises in collective action situations in which the outcome of the collective action is overdetermined by the actions of the agents involved.9 Suppose, for example, that each of eleven men inflicts a single stab wound on a twelfth man, John Smith, intending to kill him. Assume the stabbings are simultaneous. Smith dies from his wounds; however, seven stab wounds would have been causally sufficient to kill him. That is, seven stab wounds are individually causally necessary and jointly causally sufficient to kill Smith. Therefore, no single stab wound (of the eleven) is either causally necessary or sufficient for Smith’s death. So while each of the men 242
8. Collective Responsibility performed an action, that is a stabbing, that was causally necessary and sufficient for wounding Smith, not one of the eleven men performed an action that was either causally necessary or sufficient for Smith’s death. So each of the men is individually morally responsible for wounding Smith, but what about the moral responsibility for killing him? It might be thought that if a person has not performed an action that was either causally necessary or sufficient for a person’s death, then that person cannot be held responsible for the person’s death. So none of the eleven men is responsible for Smith’s death. But if none of the eleven is responsible then no one is responsible. For the cause of Smith’s death was the stab wounds, and these were performed by the eleven men. Notwithstanding the above claimed lack of individual moral responsibility, it might be held that the eleven men were collectively morally responsible for Smith’s death. But even this appears to be false, since only the actions of seven were necessary for Smith’s death. So at best we are entitled to conclude that (an unspecifiable) seven of the eleven men were collectively responsible for Smith’s death, but no individual was responsible. This conclusion is very unpalatable indeed. For one thing, it sets up an unbridgeable gap between collective responsibility and individual responsibility; a collective can be morally responsible for an outcome, even though none of its members are. For another, it licenses the commission of immoral acts, so long as they are collective actions involving overdetermination; individual perpetrators are not morally responsible for heinous crimes, so long as they commit those crimes collectively and their actions overdetermine the outcome. Moreover, the same kind of argument can be used to deprive those who collectively produce goods of the benefits of their efforts. Suppose eleven men each score twenty runs in a World Cup cricket game, and this score of 220 turned out to be too difficult for the opposition to achieve. In reply, the opposition team could only manage a score of 139. So 140 of the 220 scored by the eleven was sufficient for victory in the game. Since each of the eleven men only scored 20 runs, none of the innings was either necessary or sufficient for the victorious batting total of 140. Accordingly, not one of the eleven was responsible for the victory, and so no one was responsible. But in that case presumably the eleven should not receive 243
Social Action their payments for scoring a higher total than their opponents and thereby winning the game! We first need an analysis of the kind of collective actions at issue. We have one at hand: the previously described account of joint actions. So we can conceive of such cases of collective action as actions directed to a collective end; in our first example, the collective end is the death of Smith, in our second, the production of a victorious batting total. Each of the eleven men has the collective end as an end. Moreover, each of the eleven performs the act of stabbing (or the hitting of 20 runs) as a means to the collective end he has. Further, the actions of the eleven agents are interdependent. That is, each performs his contributory action if he believes the others will perform theirs, and each does so only if he believes this. Why are the actions interdependent? They are interdependent by virtue of the existence of the collective end possessed by each of the eleven agents, and toward the realisation of which each of the individual acts is directed. Now there are a number of possible reasons why the agents came to have the collective end that they had. One such basis is mutual agreement; each forms the collective end only after each has agreed to do so. So perhaps in the stabbing example, the agents did not want some of their number to directly participate in the killing of Smith with the others being mere bystanders; rather, the policy was one in, all in. At any rate, strictly speaking, seven acts of stabbing were individually causally necessary and jointly causally sufficient to bring about the desired outcome, which was the death of Smith. So the basis for all eleven agents coming to form the collective end was not that a stabbing contribution from each was causally necessary to the outcome desired by all. Be this as it may, my point is that whatever the reason why each came to have the collective end in question, once each had come to have that collective end then there was interdependence of action. That is, each stabbed only on condition the others stabbed. So the full set of eleven acts of stabbing can be regarded as the means by which the collective end was realised; and each act of stabbing was a part of that means. Moreover, in virtue of interdependence, each act of stabbing is an integral part of the means to the collective end.
244
8. Collective Responsibility Accordingly, I conclude that all eleven agents are jointly – and therefore collectively – morally responsible for Smith’s death. For each performed an act of stabbing in the service of that (collective) end, of Smith’s death; and each of these acts of stabbing was an integral part of the means to that end. It is clear that my notion of an integral part of the means does not entail that an action that is an integral part of the means is causally necessary (or causally sufficient) for the outcome; it is enough that such an action make a causal contribution. However, it might now be asked: What is the precise meaning of “integral part of the means”? In what sense is the action integral, if it is in some sense interdependent with other contributing actions, but not causally necessary for the outcome? Such an action is an integral part of the means if: (a) the action makes a causal contribution to the outcome; (b) it is intentionally performed as part of the means to the outcome; and (c) the outcome is the collective end of the agent who performs the action. In short, an action is an integral part of the means if it is a contributory element of a joint action, and as such participates in the web of interdependence definitive of joint actions. So each of the eleven individual acts of stabbing Smith is an integral part of the means of bringing about Smith’s death, notwithstanding the fact that none of these acts was causally necessary (or causally sufficient) for this outcome. Note the following residual point. You will recall from Chapter 2 that it is not definitive of joint action that each perform his or her contributory action on the condition, and only on the condition, that all of the others perform theirs. Rather, it is sufficient that each perform his or her contributory action on the condition, and only on the condition, that most of the others perform theirs. So the interdependence involved in joint action is not necessarily complete interdependence. Nevertheless, if the action of one agent (or more than one agent) is not interdependent with any of the actions of the other agents, then the action of that first agent (or agents) is not part of the joint action. So, if one (or more) of the eleven agents involved in the stabbing example performed his act of stabbing independently of the rest, and if the rest performed their acts of stabbing independently of that one agent, then the action of that one agent would not be part of the joint action. The action of that
245
Social Action one agent would not be part of the means to the collective end; and the agent could not be said to have had the death of Smith as a collective end.10 While recourse to the notion of a joint action, and (especially) the collective end theory of joint action, might be thought to have solved this problem of overdetermination, there are other kinds of problems of overdetermination. Consider a revised version of the stabbing example, according to which there is not one joint action, but rather two joint actions. Assume that five of the eleven men belong to one gang, and the other six men to another gang. Assume further that, coincidentally, each gang attacked Smith at the same time, but that five stab wounds would have been causally sufficient to bring about his death. So the death of Smith is overdetermined by virtue of there being two joint actions, each of which was causally sufficient to bring about Smith’s death. For reasons of space, my response to this argument is necessarily a brief one. In my view, if an action is a means to some end, and if the action is sufficient for the realisation of that end, then the agent who performed the action has a (natural) responsibility for bringing about the end. So the fact that the outcome in question might be overdetermined by virtue of the existence of some second action performed by some second agent does not remove the responsibility of the first agent for the outcome in question. Consider two assassins who work entirely independently. By coincidence each assassin fires a bullet at the president of the USA, and the two bullets lodge simultaneously in the brain of the president, killing the president instantly. Assume further that either one of the bullets would have been sufficient to kill the president. I take it that each of the assassins is guilty of murder, and each is guilty by virtue of having intentionally (and individually) shot the president dead. So moral responsibility can be ascribed in cases in which an agent’s action is causally sufficient, but not causally necessary, for a morally significant outcome.11 The stabbing example in its two-gang version is analogous to the assassin example. There are two independent actions, albeit two joint actions, and each of these (joint) actions is sufficient for the death of Smith. I conclude that, just as the two assassins are both morally responsible for the murder of the president, so the members of both of the two gangs are morally responsible for the death 246
8. Collective Responsibility of Smith. The only difference is that each of the assassins is individually responsible for the death of the president, whereas the members of the first gang are jointly responsible for Smith’s death, as are the members of the second gang. Collective Responsibility and Organisations The second problem in relation to collective moral responsibility for actions arises in the context of the actions of large groups and organisations. In order to focus our discussion I will consider a wellknown and extreme example of collective wrongdoing, namely the Rwandan genocide. According to Fergal Keane, an influential journalist who undertook investigative activity in Rwanda at the time of the genocide, the genocide in Rwanda was a premeditated collective enterprise.12 From 1990 thousands of Hutus were organised into citizen militias (Interahamwe) by the Rwandan army under the directions of Rwandan Hutu government leaders, including President Habyarimana. (Rwanda was a one-party state governed by the MRND – National Revolutionary Movement for Development – led by the president.) Lists of Hutus and Tutsis were compiled on the basis of identity cards, an identity card system having been put in place in 1926 during the Belgian colonial period. At a prearranged moment the Interahamwe were to go on a genocidal rampage against Tutsis. That moment turned out to be – whether by design or opportunity – the shooting down of the plane carrying the Rwandan and Burundian presidents, Habyarimana and Ntaryamira, on April 6, 1994. According to Fergal Keane, “In the ensuing 100 days up to one million people were hacked, strangled, clubbed and burned to death.”13 Clearly each individual who committed murder is individually morally responsible for that murder. However, there is an additional dimension of collective responsibility. This dimension arises in virtue of the overall collective end of the organised murder of Tutsis, namely genocide. Apparently, individual murders were committed as a means to an overall collective end: the elimination of all Tutsis in Rwanda. So the perpetrators were collectively morally responsible for genocide. The nature and extent of the contributions of individual members of the Interahamwe and their leadership to this collective end of genocide varied. For example, some assisted by identifying indi247
Social Action viduals as Tutsis, but without actually taking anyone’s life. Perhaps those in subordinate positions have diminished moral responsibility relative to those in positions of authority. But even if none has full responsibility, none can escape partial moral responsibility for the realisation of the collective end to which they intentionally contributed. These kinds of collective action situations embody the previously described collective action problem of overdetermination. For the actions of some subset of the members of the Interahamwe (and the Rwandan army and its leadership) were causally necessary and sufficient for the genocide. Accordingly, there is the danger of not being able to ascribe moral responsibility for the genocide to any of the murderous members of the Interahamwe. Naturally, each member is individually responsible for his or her individual act of murder, or for assisting in the murder of a given person, or for planning an attack, or for whatever contributory action he or she performed. But how do we escape the conclusion that no one is morally responsible for the genocide as such? As with our stabbing example, the actions of the members of the Interahamwe are interdependent in virtue of the collective end, namely genocide. Naturally, this interdependence is far more complex, given the existence of a hierarchical organisation, the Rwandan army, and its more loosely structured extensions, the citizen militias. Moreover, the contribution of each individual to the outcome is far more various, and in general quite insignificant, given the large numbers of people involved. Let us now consider the structure of interdependence in the Rwandan genocide. As we have seen, the genocide was overdetermined by the actions of the members of the Interahamwe. However, the genocide is disanalogous to the stabbing example in various respects. For example, we have assumed in the stabbing example that each agent acted as an autonomous agent. For example, each was free to refuse to be involved. This was evidently not the case in the Rwandan genocide. Some moderate Hutus, who refused to participate, themselves became targets for the murderous Interahamwe. Accordingly, many ordinary Hutus – depending on the nature and extent of their participation – might have diminished responsibility for their actions; for example, a Hutu who in fear of his life made known the whereabouts of Tutsi neighbours might have diminished responsibility for the death of those Tutsi neigh248
8. Collective Responsibility bours. And so might youths who were coerced by soldiers into joining the citizen militias and coerced also into participating in their murderous activities. At any rate, at this point the notion of what I termed in Chapter 5 a layered structure of joint actions needs to be introduced.14 Suppose a number of “actions” are performed in order to realise some collective end. Call the resulting joint action a level two joint action. Suppose, in addition, that each of the component individual “actions” of this level two joint “action” is itself – at least in part – a joint action with a second set of component individual actions. And suppose the member actions of this second set have the performance of this level two “action” as their collective end. Call the joint action composed of the members of this second set of actions a level one joint action. An illustration of the notion of a layered structure of joint actions is in fact an army fighting a battle, or an army – together with organised militias – engaging in genocide. At level one we have a number of joint actions. The members of one of the militia groups attack a village and kill all the inhabitants. At the same time other militia groups are doing likewise in other villages. So there is a series of level one joint actions. Now, each of these (level one) joint actions is itself describable as an individual action performed (respectively) by the different militia groups, namely the action of destroying and killing all the members of such and such a village. However, each of these “individual” actions is part of a larger joint action directed to the collective end of genocide. For each of these individual attacks on villages is part of a larger plan coordinated by the Rwandan army and its Hutu leadership. So these “individual” actions constitute a level two joint action directed to the collective end of genocide. Accordingly, if all, or most, of the individual actions of the members of the Interahamwe and Rwandan army were performed in accordance with collective ends, and each of the resulting level one joint actions was itself performed in accordance with the collective end of genocide, then, at least in principle, we could ascribe joint moral responsibility for the genocide to the members of the Interahamwe and the Rwandan army. Perhaps in reality these conditions did not obtain, or did not obtain to a sufficient degree. At any rate, the upshot of our discussion is that agents involved in complex cooperative enterprises can, at least in principle, be 249
Social Action ascribed collective or joint natural responsibility for the outcomes aimed at by those enterprises, and in cases of morally significant enterprises, they can be ascribed collective or joint moral responsibility for those outcomes. This conclusion depends on the possibility of analysing these enterprises in terms of layered structures of joint action.15 collective responsibility for omissions Finally, let us turn to the collective moral responsibility for omissions, specifically, the collective moral responsibility on the part of social groups to intervene to prevent harm to others. Following Henry Shue, let us accept the proposition that basic moral rights, including the right to security, generate rights to protection and assistance.16 With the establishment of the nation state, and specifically of policing institutions, the responsibility for protecting and assisting those whose life or security is threatened from within a society has to a large extent devolved to the police. When these rights are externally threatened it is the military institutions of the state that bear the responsibility. So the state has a special responsibility to protect and assist its own citizens when there are either internal or external threats to their basic rights. So far so good. But what are we to say about cases in which the state is no longer willing or able to protect the rights to security of its citizens? Indeed, in some of these cases, the state is itself the source of the threat. The Rwandan genocide is one such case. Let us once again use the Rwandan example to focus our discussion. Henry Shue has persuasively argued that the state has obligations other than the obligation to promote the interests of its citizens.17 Specifically, the state has an obligation not to unduly harm citizens of other states. Examples of such obligations include the obligation not to attack other states purely for economic gain, the obligation not to deplete the ozone layer by destroying forests, and so on. Here I want to go further and suggest that the state not only has moral obligations not to harm citizens of other states, it also has moral obligations to assist and protect the rights of citizens of other states. Specifically, it has these obligations when three general conditions are met: (i) the rights in question are basic moral rights, such as the right to security; (ii) the domestic state is not willing or able to 250
8. Collective Responsibility protect these rights; and (iii) the external state is able to protect these rights, whether by unilateral intervention, or by collective intervention with other states and/or internal social groups, such as the Rwandan Patriotic Front. As we have seen, the basic moral rights in question ground collective rights, which in turn generate collective responsibilities on the part of citizens of external nations. However, these collective responsibilities to intervene involve a number of collective action problems. In this final section of this book I want to address a version of one such problem. The problem can be illustrated by the following redescription of an infamous incident that took place in New York in 1964: On the night of March 13–14, 1964 thirty-eight New Yorkers were awakened in the early hours of the morning by the frenzied cries for help of a young woman, Kitty Genovese, the victim of a savage physical assault perpetrated by a man who had accosted her while she was on her way home. Over a period of about forty minutes, the assailant made several separate attacks on her, while she struggled, battered and bleeding, to reach the sanctuary of her apartment. Her screams of anguish and her calls for help were heard by at least thirty-eight neighbours, who, in the privacy and anonymity of their own homes, witnessed her struggle, yet offered no assistance in any form, whether through direct intervention or through the simple expedient of telephoning the police. A neighbour finally summoned the police, after first calling a friend to seek advice as to what to do. A patrol car arrived on the scene within two minutes but this prompt assistance was too late to save the young woman, who died on the way to hospital. It is clear that Kitty Genovese was a victim not only of her assailant’s viciousness, but also of her neighbours’ inaction.18
The first question that arises in relation to this scenario is the question as to who is to be the one to intervene by calling the police. Most, if not all, of the neighbours were capable of telephoning the police, but each decided not to, or decided not to until it was too late. So each intentionally refrained from calling the police, notwithstanding the fact that there was a moral obligation to do so. Note that while there was a moral obligation to phone, it does not attach to any one person without attaching to any other person; nor does it seem to attach to everyone simultaneously, since it is only one phone call that is needed. Perhaps each person has an individual, 251
Social Action but conditional, moral obligation. Each has an obligation to phone on the condition that each of the others does not phone. So if someone rings the police, then the others have not failed to discharge their individual moral obligations. On the other hand, if no one rings the police then each has failed to discharge his or her individual moral responsibility.19 In short, we have interdependence of individual moral responsibility to intervene, though not joint moral responsibility to intervene. Perhaps the Rwandan genocide had the structure of the Kitty Genovese case. There was an obligation for some or other (reasonably powerful) nation state to intervene to prevent the genocide, but none did. On the other hand, perhaps the Rwandan genocide did not have the structure of the Kitty Genovese episode. For there are a large number of complex and ongoing international social conflicts, and perhaps no one state – even a superpower, such as the USA – has the capability to successfully intervene in all of them. If so, then there is a disanalogy between the Rwandan and the Kitty Genovese episodes. We need first to clarify the international response to the Rwandan genocide. When the slaughter began on April 6, 1994, the UN forces in Rwanda were ordered not to intervene. Indeed, shortly thereafter, the number of UN forces was reduced. The UN Security Council did issue a resolution condemning the killing, but it stopped short of using the term “genocide.” Had the word “genocide” been used, then the UN would have been legally obliged to intervene. In May 1994, the UN resolved to send 6,800 troops, but delays resulted due to a dispute between the UN and the USA over equipment costs. Finally, a French contingent was dispatched to southern Rwanda, but had little overall effect. The worst excesses of the genocide were not brought to a halt by any actions of the UN, but rather by the Rwandan Patriotic Front, which defeated the Rwandan army and the Interahamwe, and installed a new government in July 1994. We can therefore conclude the following: (i) the UN intentionally refrained from discharging its institutional duty to intervene in relation to a morally significant matter (genocide) and is therefore morally responsible for that omission; (ii) the salient “great powers,” the USA and France, intentionally refrained from discharging their (conditional) individual moral obligations to intervene to prevent 252
8. Collective Responsibility the genocide; (iii) the individual members of the UN Security Council intentionally refrained from discharging their joint institutional duty in relation to a morally significant matter (genocide), and are therefore jointly morally responsible for that omission. Thus described, the Rwandan genocide seems analogous in important respects to the murder of Kitty Genovese. It was not necessary for all the member states of the United Nations to intervene, and given the complexities of the situation, unilateral action by, say, the USA or France, might not have been desirable. However, it was desirable, indeed obligatory, that some members of the UN jointly intervene in the Rwandan genocide. Thus, given that each state intentionally refrained from intervention, all the member states of the United Nations are morally responsible for failing to prevent the genocide. Thus described, the failure on the part of (at least) the USA and France to intervene in relation to the Rwandan genocide seems analogous with the Kitty Genovese episode; in each case it was a failure to discharge an individual conditional obligation. But failure on the part of a set of agents to discharge their individual conditional moral obligations is not the same thing as collective moral responsibility for an omission. But what of the UN (and the Security Council)? Were its member states collectively morally responsible for the non-intervention in the Rwandan genocide? To assist us with this question let us consider a variation on the Kitty Genovese episode. Assume that in this second scenario the neighbours do not have access to phones; physical intervention is necessary. Let us assume that one (or even two or three neighbours) acting on their own are highly unlikely to be unsuccessful, and indeed would put at risk their own physical wellbeing. What is needed is the performance of a joint action involving all of the neighbours. Let us further assume that although this collective action is considered by the neighbours, they make a collective decision not to intervene, and do so having as an end that Kitty’s murder not be prevented from occurring. Accordingly, each neighbour performs an intentional act of refraining from acting, having as an end that Kitty’s murder not be prevented from occurring. Here we have a species of joint action, namely intentional joint inaction; although each of the neighbours refrains from acting, each does so in the service of a collective end, 253
Social Action namely the end that Kitty’s murder not be prevented from occurring, or at least, not be prevented from occurring by the neighbours. Accordingly, there is individual moral responsibility on the part of each of the neighbours for their refraining to act, and there is joint moral responsibility for failing to prevent the murder of Kitty. We can conclude that the neighbours are collectively morally responsible for failing to prevent the murder of Kitty in the manner of collective moral responsibility set forth in the first section of this chapter. Perhaps the UN’s failure to intervene was similar to this second version of the Kitty Genovese episode. For there was an intentional act of refraining from acting. Moreover, perhaps the UN decision not to act was made in the service of the (collective) end that the Rwandan genocide not be prevented from occurring or, at least, not be prevented from occurring by the member states of the UN. If so, the member states of the UN were collectively morally responsible for failing to prevent the Rwandan genocide in the manner of collective moral responsibility set forth in the first section of this chapter. Let us now consider a third version of the Kitty Genovese scenario; and perhaps the UN’s failure to intervene in Rwanda should be assimilated to this third version, rather than to the second version, just given. On this third version the agents intentionally refrain from intervening, but there is no collective decision not to intervene. Moreover, in refraining from intervening the agents do not have any collective end in mind. Rather, each individually decides not to intervene because it is too much bother to intervene. However, each is aware of the possibility of joint intervention, and indeed it is discussed and is seen to be perfectly feasible, and possibly morally obligatory. Let us further assume that one neighbour decides to intervene on his own, believing that others will join in and assist him. However, they do not, and his efforts are to no avail. Indeed, he is bashed for his troubles by Kitty’s murderer. Presumably, there is a joint moral responsibility to intervene, and this responsibility is grounded in Kitty’s moral right to protection from those able to protect, or at least on the part of those able to protect her without putting themselves at great risk. So Kitty’s right to protection generates an obligation on the part of each of the neighbours to have her protection as an end, and to pursue suitable 254
8. Collective Responsibility means to secure that end. The salient means is for all, or most, to intervene; so each should form an intention to intervene, but do so on the condition the other intervene. Naturally, there is no obligation on the part of any single neighbour to intervene if they know that the others are not going to do likewise; for in that event the end would not be realised, and the single neighbour would be putting himself or herself at great risk. Accordingly, in order to come to have the protection of Kitty as an end, each agent needs to know that the other agents have, or will have, this as an end and therefore will perform the requisite contributory means. So each is under a moral obligation to intervene on condition, and only on condition, that the others do. Here it is important to distinguish the case where each will intervene if and only if the others do, but does not have as an end to protect Kitty; rather, each has (say) the individual end to avoid bringing about a situation in which the others intervene, but he or she does not. This latter situation is a case of having the individual end of not being the odd person out; but this is not the same as having the collective end of bringing about the prevention of the murder of Kitty Genovese. At any rate, on this third version of the Kitty Genovese scenario the following ascriptions of moral responsibility seem warranted; and to the extent that the UN’s failure to intervene in Rwanda can be assimilated to this third version, then analogues of the following ascriptions of moral responsibility will hold true of the UN member states. First, a given neighbour is individually morally responsible for failing to individually intervene in order to prevent Kitty’s murder, if: (a) Kitty was murdered; (b) that neighbour intentionally refrained from individually intervening to assist in the prevention of that murder, even though the other neighbours did intervene, or that neighbour would have intentionally refrained from intervening, if the others had intervened; and (c) the intervention of that neighbour, taken in conjunction with the intervention of the other neighbours, would have been sufficient to prevent Kitty’s murder. Second, the neighbours are collectively, that is, jointly, morally responsible for failing to prevent Kitty’s murder, if: (a) Kitty was murdered; (b) the neighbours intentionally refrained from intervening; (c) each of the neighbours intervening (having as an end to prevent the murder of Kitty) would have resulted in the prevention of her murder; and (d) each of the neighbours would still have 255
Social Action intentionally refrained from intervening (and intervening with the end of prevention of the murder of Kitty) even if the others had intervened with that end. It should also be noted that if a given neighbour is not individually morally responsible for failing to individually intervene in order to prevent Kitty’s murder, then that neighbour does not share in the collective moral responsibility for failing to jointly intervene in order to prevent Kitty’s murder. So the neighbour who unilaterally intervened to assist Kitty, and was bashed for his troubles, escapes both individual responsibility and any share of collective moral responsibility. The Kitty Genovese episode and the Rwandan genocide illustrate the general moral imperative not to allow the basic rights of others to be violated. This moral imperative generates obligations to cooperate and/or put in place cooperative schemes, including institutional mechanisms to prevent or abort violations of basic rights.20 International social conflicts can be regarded as partially analogous to recurring Kitty Genovese scenarios. The relevant analogous principle here is that the UN and other formal or informal blocs of states should develop institutional mechanisms for joint armed intervention in international social conflicts and should deploy those mechanisms to prevent large-scale serious rights violations. In fact the UN has already developed some such mechanisms, including a mechanism to intervene in the case of genocide. However, as the Rwandan tragedy demonstrates, these mechanisms are not necessarily used. However, there is a disanalogy. For unlike Kitty Genovese scenarios, such intervention in international social conflicts involves layered structures of joint actions. At the second level of “action,” a given state has a moral responsibility, jointly with other states, to develop and deploy institutional mechanisms to ensure humanitarian armed intervention in international social conflicts. At the first level of action, an individual citizen of a given state has a moral responsibility jointly with the other citizens to lend his or her support to the creation and implementation of governmental policies that would develop and deploy the mechanisms for the sorts of humanitarian armed intervention in question. So the first level of joint responsibility interacts with the second level. The upshot is that citizens of a given state have joint moral 256
8. Collective Responsibility responsibilities with one another, and with citizens of other states, to support armed intervention to prevent large-scale serious rights violations in international social conflicts. The existence of these moral responsibilities implies a shift in the concept of citizenship, and a concomitant shift in the concept of the state (at least as these concepts have often been understood). Specifically, large-scale violations of basic moral rights generate moral responsibilities to intervene on the part of foreign citizens; and the state must be viewed not only as an institution that exists for the protection of the rights of its own citizens, but also as one that exists for the protection of at least some of the basic rights of foreign citizens. Clearly there is much more and better work to be done on collective moral responsibility for omissions, including the collective responsibility of members of communities and organisations. And there is further work to be done on the question of sharing collective moral responsibility for outcomes and omissions; sometimes those who are jointly responsible for an outcome or omission are each fully responsible and sometimes only partially responsible for it. However, that work cannot be done here. Instead this discussion of collective responsibility as joint responsibility concludes not only this chapter, but also this study of social action. For I have now completed the main tasks of this book: the tasks of providing philosophical analyses of central social action types – joint action, conventional action, social norm-governed action, and the actions of social groups, social institutions, and organisations – and of the related normative phenomena of collective rights and collective responsibility.
257
Notes
Introduction 1. I will not in this book be centrally concerned with so-called mental acts or with action that might be short of intentional action. Moreover, I do not want to get bogged down in disputes regarding different theories of individual action. Roughly speaking, I take it that so-called basic individual action consists of a mental state, such as an intention, a bodily movement, and a causal nexus between the intention and the bodily movement. 2. By logical presupposition, priority, and implication I do not mean to invoke logic in the strict sense of deductive logic, including notions of logical entailment. Nor do I mean to invoke notions of supervenience. Rather, I intend to invoke the familiar, if somewhat vague, notion of conceptual connectedness. 3. There is a tendency in some quarters to operate with an insufficiently differentiated notion of social action. All action is seen as social; indeed, all action tends to be viewed as constituted by social realities. One source of this tendency derives from the abandonment of substantive notions of the self and agency; the self is conceived by many poststructuralists, for example, simply as a social construction. Thus Roger Fowler asserts that “a real person can be seen, as the social psychologist sees him, as a construction of roles acquired through the process of socialisation.” Roger Fowler, Linguistics and the Novel (London: Methuen, 1977), 128. In a similar vein, Foucault claims: “The subject is a plurality of possible positions and functions.” “Order of Discourse,” in Language and Politics, ed. M. Shapiro (Oxford: Blackwell, 1984), 129. Suffice it to say here that I reject social constructivists’ accounts of the self. I also reject the Humean constructed self. 4. C. D. Broad, The Mind and Its Place in Nature (London: Kegan Paul, 1928) 328. 5. But see Edmund Husserl, Cartesian Meditations, trans. Dorion Cairns (The Hague: Martinus Nijhoff, 1960) 89–90; and David Bell, Husserl
259
Notes to pp. 4–16
6. 7.
8.
9.
10. 11. 12.
13.
14.
(London: Routledge, 1990) 218–19. See also Thomas O. Buford (ed.), Essays on Other Minds (Chicago: University of Illinois Press, 1970). For a discussion of issues in this area see Nicolas Berdyaev, Solitude and Society (London: Centenary Press, 1938), chap. 11: 106–14. Charles Taylor, “Atomism,” in his Philosophy and the Human Sciences: Philosophical Papers 2 (Cambridge: Cambridge University Press, 1985): 187–210. And individuals and groups can pursue individual and interpersonal actions outside or substantially outside, such a social framework. Consider two warring tribes, or the example of two men trying to kill one another described in Chapter 1. For an elaboration of the voluntaristic theory of action (including in relation to Max Weber), see Talcott Parsons, The Structure of Social Action: A Study of Social Theory with Special Reference to a Group of Recent European Writers) vols 1 & 2 (New York: Free Press, 1968). See also Alfred Schutz, The Phenomenology of the Social World, trans. George Walsh and Frederick Lehnert (Chicago: Northwestern University Press, 1967); and George H. Mead, Mind, Self and Society (Chicago: University of Chicago Press, 1934). J. S. Coleman, The Foundations of Social Theory (Cambridge, Mass.: Harvard University Press, 1994). Parsons, Structure of Social Action, 229. The general view that agency entails moral agency is associated with Alan Gewirth. See his Reason and Morality (Chicago: University of Chicago Press, 1981). For arguments against the proposition that the concept of action entails that the agent ascribes worth to the action, or the purpose for which the action was performed, see Virginia Held, “The Normative Import of Action,” in Gewirth: Critical Essays on Action, Rationality and Community, ed. Michael Boylan (Oxford: Rowman and Littlefield, 1999). Since Gewirth seems to assume that aims or intentions are wants or desires, and that agents can reflect on their ends (Reason and Morality 38), his notion of an agent may be a “thicker” one than the notional one that I have in mind. H. P. Grice, Studies in the Way of Words (Cambridge, Mass.: Harvard University Press, 1989). Seumas Miller, “Speaker Meaning and Assertion,” South African Journal of Philosophy 4, no. 2 (1985): 48–55; “Performatives,” Philosophical Studies 45, no. 2 (1984): 247–60; “Truthtelling and the Actual Language Relation,” Philosophical Studies 49, no. 2 (1986): 281–94. Richard Freadman and Seumas Miller, Rethinking Theory: A Critique of Contemporary Literary Theory and an Alternative Account (Cambridge: Cambridge University Press, 1992).
260
Notes to pp. 16–22 15. Max Weber, Economy and Society: An Outline of Interpretive Sociology, ed. G. Roth and C. Wittich (Berkeley: University of California Press, 1978), vol. 1: 4. 16. John R. Searle, The Construction of Social Reality (London: Penguin, 1995). 17. Charles Taylor, “Interpretation and the Sciences of Man,” in his Philosophy and the Human Sciences, 15–57. 18. Searle, Construction of Social Reality, chap. 2. 19. Ibid., chaps. 1 and 2. Searle explains social facts in terms of collective intentionality and constitutive rules.
1 Theories of Social Action 1. Anthony Giddens, The Constitution of Society: Outline of the Theory of Structuration (Cambridge: Polity Press, 1984), 9. 2. See, for example, the works of the French social philosopher, Michel Foucault. See Seumas Miller, “Foucault on Discourse and Power,” Theoria 76 (1990): 115–25. Much of the following section on Foucault is derived from this paper. 3. Michel Foucault, History of Sexuality, Volume 1: An Introduction, trans. R. Hurley (New York: Pantheon, 1978), 93. 4. Foucault, “Two Lectures,” 98. 5. In fact, Foucault insists that the power involved here is not merely coercive but constructive: It doesn’t just compel certain forms of behaviour; it actually produces a certain kind of being. Thus, Foucault states: “The individual which power has constituted is at the same time its vehicle.” Foucault, “Two Lectures,” 98. 6. Much of the following section on the cooperation model of social action is to be found in Seumas Miller, “Sociopolitical Action, Ethics and the Power of Literature,” Philosophy and Social Criticism 21, no. 3 (1995): 93–110. 7. Naturally, one could destroy the social fabric by simply destroying the society, that is by killing everyone, including oneself. The social fabric presupposes living human beings. Alternatively, an individual could reject the social fabric by (say) having a massive lobotomy, but yet remaining alive. So by individual I mean a more or less fully functioning adult human person with rational, linguistic, and moral powers. 8. Immanuel Kant, Critique of Pure Reason, trans. J. Meiklejohn (New York: Wiley, 1943) 6. 9. Searle, Construction of Social Reality, 104, offers a somewhat undifferentiated account of social power – what he calls “conventional power.” Specifically, he fails to sufficiently distinguish between the repressive deployment of conventions and norms on the one hand, and their
261
Notes to pp. 22–27
10.
11.
12.
13.
14. 15.
16.
17. 18. 19.
enabling role on the other. Moreover, much of what he has to say about the exercise of social power pertains to the exercise of power by persons in authority – especially by those occupying the top positions in large organisations – rather than to social forms in general. See, for example, Larry May, The Morality of Groups (Notre Dame: University of Notre Dame Press, 1987), chaps. 1 & 2; Raimo Tuomela, The Importance of Us (Stanford: Stanford University Press, 1995), chap. 4; and Margaret Gilbert, On Social Facts (Princeton: Princeton University Press, 1989), chap. IV. Seumas Miller, “Joint Action,” Philosophical Papers 21, no. 3 (1992): 275–97; “On Conventions,” Australasian Journal of Philosophy 70, no. 4 (1992): 435–44; “Social Norms” in Contemporary Action Theory, Volume 2: Social Action, ed. Ghita Holmstro¨m-Hintikka and Raimo Tumela (Dordrecht: Kluwer, 1997); “Individualism, Collective Responsibility and Corporate Crime,” Business and Professional Ethics Journal 16, no. 4 (1997): 19–46; “Collective Rights,” Public Affairs Quarterly 13, no. 4 (1999): 331–47; “Collective Responsibility,” Public Affairs Quarterly 15, no. 1 (2001): 65–82. See Seumas Miller, “Intentions, Ends and Joint Action,” Philosophical Papers 24, no. 1 (1995): 51–67, for a discussion of the main analyses of joint action. See also Tuomela, Importance of Us. Naturally, joint actions, being actions, involve intentions and beliefs. And intentions and beliefs have a minimal level of normativity in virtue of their intentionality or directedness. Roughly speaking, intentions and beliefs fail or succeed; intentions succeed if the actions thus intended are carried out, and beliefs succeed if the world is as they make it out to be. See note 12 to the Introduction. This notion of a sphere of activity is different from, but has resonances with, Walzer’s notion of a sphere of justice. See Michael Walzer, Spheres of Justice (New York: Basic Books, 1983). Tuomela operates with a different taxonomy, namely what he calls s-norms (expectation based) and r-norms (agreement based). See Tuomela, Importance of Us, chap. 10. Aristotle, Nichomachean Ethics (any edition), bk V, chap. 7; and David Hume, A Treatise of Human Nature (any edition), esp. bk III, pt. II, 2. I do not mean to imply that there are no ultimate sources of moral value that are nonhuman and nonsocial, for example, God. For two influential communitarian perspectives, see Alasdair Macyntyre, After Virtue (Notre Dame: University of Notre Dame Press, 1984); and Michael Sandel, Liberalism and the Limits of Justice (Cambridge: Cambridge University Press, 1982).
262
Notes to pp. 28–35 20. See Tuomela, Importance of Us (1995), chap. 8, for one account of roles. 21. See Rom Harre´, Social Being (Oxford: Blackwell, 1979), 37–43, for an account of structure. 22. Here and elsewhere in this book I do not always bother to distinguish objective from subjectively felt moral considerations. Suffice it to say that social norms involve subjectively felt moral considerations. See Miller, “Social Norms.” 23. For another account of institutions see Harre´, Social Being, 97–100. 24. See Giddens, Constitution of Society 2. 25. Collective habit is, in effect, what Giddens seems to be suggesting. See Giddens, Constitution of Society, 2, 25. 26. Searle, Construction of Social Reality; Gilbert, On Social Facts; Raimo Tuomela, A Theory of Social Action (Dordrecht: Reidel, 1984), Importance of Us ; Raimo Tuomela (with Kaarlo Miller), “We-Intentions,” Philosophical Studies 53, no.3 (1988): 367–89. 27. David Lewis, Convention: A Philosophical Study (Cambridge, Mass.: Harvard University Press, 1969). 28. Edna Ullmann-Margalit, The Emergence of Norms (Oxford: Oxford University Press, 1977); Philip Pettit The Common Mind. (Oxford: Oxford University Press, 1993), and “Virtus Normativa: Rational Choice Perspectives;” Ethics 100 (1990): 725–55. See also Russell Hardin, Collective Action (Baltimore: Johns Hopkins University Press, 1982). 29. Michael Bratman, Intentions, Plans and Practical Reason (Cambridge, Mass.: Harvard University Press, 1987). 30. David-Hillel Ruben, The Metaphysics of the Social World (London: Routledge and Kegan Paul, 1985), 105–6. 31. This view is implicit in Tuomela’s work. At the very least Tuomela is committed to the view that if an action is joint, then it is social. 32. Peter Winch, Idea of a Social Science (London: Routledge and Kegan Paul, 1958). Searle, in his Construction of Social Reality, offers a definition of a species of social actions, namely institutional actions, and does so in terms of (in effect) the concept of joint action in conjunction with the concept of constitutive rule. Roughly, an institutional action is an action performed in accordance with a socially imposed constitutive rule. Whatever the merits of this proposal as a definition of institutional action, it does not provide a definition of social action in general. 33. Ruben, Metaphysics of the Social World; and David Shwayder, The Stratification of Behaviour (London: Routledge and Kegan Paul, 1965). 34. Ruben, Metaphysics of the Social World, 105–6. 35. Naturally there are, in addition to pure conflict cases and pure cooperative cases, cases of mixed conflict. Most such mixed conflict cases
263
Notes to pp. 35–42
36. 37.
38. 39. 40.
41.
42. 43. 44.
45. 46. 47.
48. 49. 50.
will be social. See Seumas Miller, “Conventions and Social Contracts,” Philosophical Papers 16, no. 2 (1987): 85–106. Donald Davidson, “Rational Animals,” in Actions and Events, ed. Ernest Le Pore and Brian McLaughlin (Oxford: Blackwell, 1985), 473–80. Michael Bratman, “Shared Co-operative Activity,” Philosophical Review 101, no. 2 (1992): 327–41; Philip R. Cohen and Hector J. Levesque, “Teamwork,” Nous 25 (1991): 487–512; Miller, “Joint Action.” Miller, “Joint Action.” Gilbert, On Social Facts, chap. II; and Tuomela, Theory of Social Action, 12. Winch (Idea of a Social Science, 50) seems to suggest that intentions are commitments and therefore involve normative rules and hence sociality. I discuss this general issue in Chapter 3 in relation to Margaret Gilbert. Suffice it to say here that intentions are surely not necessarily commitments precisely because the latter but not the former is an intrinsically normative notion. H. P. Grice, “Utterer’s Meaning, Sentence-Meaning, and WordMeaning,” Foundations of Language 4 (1968): 225–42; and Jonathan Bennett, Linguistic Behaviour (Cambridge: Cambridge University Press, 1976). Gilbert, On Social Facts, chap. V. See Ludwig Wittgenstein, Philosophical Investigations (any edition); Searle, Construction of Social Reality; Winch, Idea of a Social Science. See, for example, Saul Kripke, Wittgenstein on Rules and Private Language (Oxford: Blackwell, 1982). See also Colin McGinn, Wittgenstein on Meaning (Oxford: Blackwell 1984). See also Donald Davidson’s related arguments concerning language and thought in “Rational Animals.” See also Pettit, The Common Mind, chap. 4. Kripke, Wittgenstein. McGinn, Wittgenstein on Meaning, 159. The basic notion of a procedure was introduced by Grice, “Utterer’s Meaning, Sentence-Meaning, Word-Meaning.” The notion was developed and deployed in relation to conventions in Seumas Miller, Conventions and Speech Acts (University of Melbourne PhD, 1985, Ann Arbor: University Microfilms International, 1986), and “Rationalising Conventions,” Synthese 84, no. 1 (1990). See also Bratman, Intentions, Plans and Practical Reason, and Joseph Raz, Practical Reason and Norms (London: Hutchinson, 1979). McGinn, Wittgenstein on Meaning, 169–70. J. R. Searle, Speech Acts: An Essay in the Philosophy of Language (Cambridge: Cambridge University Press, 1969), and Construction of Social Reality. For discussions and criticisms of the notion of a constitutive rule, see J.
264
Notes to pp. 42–54
51.
52.
53.
54.
55.
56.
57.
58.
Garcia, “Constitutive Rules,” Philosophia 17 (1987), and Raz, Practical Reason and Norms, 108. I have propounded this argument in relation to speech acts in Seumas Miller, “Speech Acts and Conventions,” Language Sciences 22, no. 2 (2000): 155–66. I say “ultimately” since x might count as y where x is itself a social action, but only if there was some w such that w counted as x and w was not a social action (or some v, which counted as w and v was not a social action. . . . or some u . . . ). The conventional decision is that in such cases six runs are to be recorded. Naturally, whether or not a particular batsman hit the ball over the fence – and is therefore entitled to have six runs recorded – is a matter for the individual umpire to decide on the day in question. Or at least language as it is understood by so-called agent semanticists, such as H. P. Grice and P. F. Strawson. Grice, Studies in the Way of Words, and P. F. Strawson Logico-Linguistic Papers (London: Methuen, 1971), essay 8. And we could define a moral agent as a being who necessarily possessed moral autonomy or only as a being who was capable of possessing it, but might not in fact possess it. I have opted for the former. The same point could be made with regard to autonomy, as opposed to moral autonomy. It might be argued that rationality does imply morality, but that I have chosen too thin a notion of rationality in opting for mere instrumental rationality. I do not reject this claim; my points pertain only to rationality narrowly conceived, namely instrumental rationality. Such agents would also need to have a (perhaps minimal) capacity for theoretical reasoning, for example simple deductive reasoning from one belief to another. See Charles Taylor, Sources of the Self (Cambridge: Cambridge University Press, 1992), chap. 2.
2 Joint Action 1. It might be suggested that the existence of a stand alone joint procedure involving only two agents does not constitute a social form, and that therefore a joint procedure is not necessarily a convention. If this is so then I would reply that only joint procedures that are part of a structure of interconnected joint procedures are, properly speaking, conventions. In all probability there is no determinate answer to this sort of dispute, and stipulation is called for. 2. As will become evident from the discussion in Chapter 5, joint procedures and joint mechanisms are not mutually exclusive.
265
Notes to pp. 56–58 3. I accept that there is an element of stipulation in ruling out such cases as instances of joint action. The stipulation is justified by the utility of the general account of joint action, which takes mutually truly believed interdependence as paradigmatic of joint action. 4. See: Tuomela, Theory of Social Action, esp. chap. 5, and Tuomela (with Miller), “We-Intentions,” and Tuomela, Importance of Us ; Gilbert, On Social Facts, esp. sec. IV; John R. Searle, “Collective Intentions and Actions,”in Intentions in Communication, ed. Philip R. Cohen, Jerry Morgan, and Martha E. Pollack (Cambridge, Mass.: MIT Press, 1990), 401– 15; Cohen and Levesque, “Teamwork”; Bratman, “Shared Co-operative Activity,” and Michael Bratman, “Shared Intention,” Ethics 104 (1993): 97–113; Miller, “Joint Action.” 5. For the most comprehensive and detailed set of analyses of collectivity concepts in the literature see Tuomela Importance of Us. 6. For an earlier elaboration of CET, see Miller, “Joint Action.” For still earlier elaborations and applications of this account in relation to conventions, speech acts and social contracts, see Miller, Conventions and Speech Acts, Seumas Miller, “Conventions, Interdependence of Action and Collective Ends,” Nous 20, no. 2 (1986): 117–40, “Conventions and Social Contracts” and “On Conventions.” 7. The notion of a collective end was introduced in Seumas Miller, “Lewis on Conventions,” Philosophical Papers 11, no. 2 (1982) p. 6. See also Miller, Conventions and Speech Acts, sec. 6, and “Conventions, Interdependence of Action and Collective Ends,” 133–4. See Searle, “Collective Intentions and Action,” and Gilbert, On Social Facts, sec. IV. 8. Something weaker than belief is sufficient here; perhaps “thinks it is likely,” or even “thinks it is quite possible.” Ditto for the other beliefs – including the mutual true beliefs – involved in the definition. However, absence of any degree of belief would not be sufficient. For if the first agent did not think there was any chance the other agent would do his part, then how could the first agent be thought to have the end in question? After all, the end can only be achieved if both agents do their parts. 9. See note 4 to this chapter. 10. It might be argued that the two individual ends are not the same in that necessarily I have as an end not only that the sniper die, but that I be the one to kill the sniper. (And likewise for your individual end.) (See John R. Searle, Intentionality: An Essay in the Philosophy of Mind (Cambridge: Cambridge University Press, 1983), esp. chap. 3.) But my example is intended to be illustrative only; it serves this purpose even if it does not correctly describe reality. On the other hand, if this
266
Notes to pp. 58–61
11. 12.
13. 14. 15.
16.
17.
18.
claim is true it will cause me to revise my account of collective ends; strictly speaking they will consist of a set of different individual ends to the effect that each has as an end not only that some state of affairs obtain, but that he or she contribute in part to the realisation of that state of affairs. However, this revision to the content of collective ends would not be fatal to my account. In any case, I am already committed to the claim that ends bring with them means in the sense that an agent performs the means in the belief that it is the means, or part of the means, to his or her end. See note 4 to this chapter. There is a considerable literature on mutual belief and related notions. See, for example, J. Heal, “Common Knowledge,” Philosophical Quarterly 28 (1978): 116–31; Lewis, Convention; S. Schiffer, Meaning (Oxford: Oxford University Press, 1972); N. Smith (ed.), Mutual Knowledge (London: Academic Press, 1982); D. Sperber and D. Wilson, Relevance (Oxford: Blackwell,1986). It is of course necessary that agents mutually know what the end is. Again, agents must mutually know what the end is. In fact, I distinguish between two sorts or “intending”: trying, and intending proper. The latter but not the former entails a belief that one can succeed in the action “intended.” This kind of distinction is a familiar one in the literature. See, for example, Bratman, Intentions, Plans and Practical Reason, 136. It is of course also true that the end of some intentional action can itself become an intended action. See Searle, Intentionality. Sometimes preferences are defined in terms of what is chosen or what an agent is disposed to choose. Suffice it to say that we need some distinction between, on the one hand, desires or wants that rationally interact to generate choices of action and, on the other hand, choices of action per se. So we need a notion of preference that reflects our comparative desires or wants, prior to their interaction with beliefs, to generate choices of action. This is the way I am understanding preferences. This is not to say that agents can have ends that are not desired, or that are the means to ends that are desired, or that they cannot have actions desired as means. See Bratman, Intentions, Plans and Practical Reason, for criticism of the belief/desire theory of action. For a belief/desire account of one species of what I take to be joint action, namely conventions, see Lewis, Convention. For the argument that conventions are joint actions, see Miller, “Joint Action” and “On Conventions.”
267
Notes to pp. 62–76 19. See Cohen and Levesque, “Teamwork,” for a contrasting account. 20. CET contrasts here with Bratman’s account: see his “Shared Intention” and “Shared Co-operative Activity.” 21. For an earlier account of these applications of CET, see Miller, “Joint Action.” 22. David Schmidtz, “Choosing Ends,” Ethics 104 (1994): 228. 23. See Cohen and Levesque, “Teamwork,” 506, for a contrasting account. 24. I discuss corporate action in Chapter 5. See also Miller, “Individualism, Collective Responsibility and Corporate Crime.” 25. This example was inspired by Michael Bratman’s video game example. See Bratman, Intentions, Plans and Practical Reason, 114. 26. This point is not vitiated by the often made distinction between prior intentions and intentions in action. Whether or not a gratuitous action is to be analysed in terms of prior intentions as well as intentions in action or simply in terms of intentions in action, the distinction between gratuitous actions and actions done for their own sake remains; and this distinction cannot be made out by recourse only to the notion of an intention. 27. This is not to say that in many conversations the speaker and the hearer do not have as a collective end that the hearer come to believe whatever is true. 28. Raimo Tuomela and Kaarlo Miller, in their “We-Intentions,” 375, appear to endorse this kind of account. (But see 376 of the same article.) In Tuomela’s later account, in Importance of Us, 125–6f and 426, Tuomela does not endorse it. 29. Tuomela (with Miller), “We-Intentions,” 375. 30. See Searle, “Collective Intentions and Actions,” 404–5. Searle ascribes this view to Tuomela, perhaps unfairly. 31. See Tuomela (with Miller), “We-Intentions,” 375. 32. Ibid., 376. 33. Ibid., 375. 34. Indeed, elsewhere Tuomela seems to accept this circularity, but argues that it is non-vicious. I do not understand how this is supposed to be non-vicious circularity. See Tuomela, Importance of Us, 143–5. 35. Something like this seems to be proposed by Tuomela in his “What Are Goals and Joint Goals?” Theory and Decision 28, no. 1 (1990): 1–20. Tuomela’s analysis of goals is in terms of wants and intentions. See, especially, 5. Cohen and Levesque, “Teamwork,” 494–5, also seem to define goals in terms of desires. 36. Bratman, “Shared Intention” and “Shared Co-operative Activity.” 37. Bratman, “Shared Intention,” 106. 38. Bratman, “Shared Co-operative Activity,” 330, and 333; “Shared Inten-
268
Notes to pp. 76–88
39. 40.
41. 42. 43.
44.
45. 46. 47.
48.
49.
50. 51. 52. 53. 54.
55.
tion,” 104 and 109; and “I Intend That We J,” in Contemporary Action Theory, Volume 2: Social Action, ed. Ghita Holmstro¨m-Hintikka and Raimo Tuomela (Dordrecht: Kluwer, 1997), 49–63. Bratman, “I Intend That We J,” 56–61. If in fact it is a joint action, as it is in danger of becoming, then Diane will have the water being pumped out as an end. But in that case I would resist the claim that Abe intended the water to be pumped out. Moreover, Bratman would have to ascribe to both Abe and Diane the intention that we pump the water out. Cohen and Levesque, “Teamwork.” Goals or ends are contrasted by Cohen and Levesque with intentions. Cohen and Levesque, “Teamwork,” 499–500. Cohen and Levesque have an argument for this (see 492). The argument turns on entirely contingent facts in some cases of joint action or at least joint processes, in which mutual knowledge that others are doing their parts might “dissipate.” But such contingent facts about some cases do not motivate requirements for a general account of joint action. Cohen and Levesque seem to hold that agents have goals (but not intentions) with respect to one another’s contributory actions. (Cohen and Levesque, “Teamwork,” 496.) This view is open to a variant of my objection to Bratman. Cohen and Levesque, “Teamwork,” 496. Searle, “Collective Intentions and Actions.” Ibid., 404. See J. R. Hobb, “Artificial Intelligence and Collective Intentionality: Comments on Searle and on Grosz and Sidner,” in Intentions in Communication, ed. Philip R. Cohen, Jerry Morgan, and Martha E. Pollack (Cambridge, Mass.: MIT Press), 445–59. For a paper that details different interpretations of Gilbert’s account of collective responsibility, see Pekka Makela, “Collective Moral Responsibility: A Collective as an Independent Moral Agent,” Australian Journal of Professional and Applied Ethics (forthcoming). Gilbert, “What Is It for Us to Intend?” in Contemporary Action Theory, Volume 2: Social Action, ed. Ghita Holmstro¨m-Hintikka and Raimo Tuomela (Dordrecht: Kluwer, 1997), 65–87. Ibid., 72. Ibid., 76. Gilbert, On Social Facts, 165. Ibid., 160. Ibid., 158–64. See also her “Walking Together,” in The Philosophy of the Human Sciences, ed. P. A. French, T. E. Euhling, Jr., and H. K. Wittstein (Notre Dame, Ind.: University of Notre Dame Press, 1990), 2–6. Gilbert, “Walking Together,” 6.
269
Notes to pp. 91–98 3 Conventions 1. Thomas Hobbes, Leviathan (any edition). John Locke, Of Civil Government: Book 2 – An Essay Concerning the True Original Extent and End of Civil Government (any edition). 2. Jean-Jacques Rousseau, The Social Contract (any edition), and Discourse on the Origins of Inequality (any edition). 3. Searle, Speech Acts. See also Miller, “Speech Acts and Conventions.” 4. Alf Ross, Directives and Norms (London: Routledge, 1968). 5. Gilbert, On Social Facts, 369. 6. Gilbert is apparently aware of the problem, but does not provide a solution. Rather, in effect, she merely announces the existence of some apparently sui generis social form, which is neither a joint intention nor a joint agreement. But what precisely is this notion? See her On Social Facts, 378–9. 7. Hume, A Treatise of Human Nature (any edition), esp. bk. III pt. II 2. 8. See Lewis, Convention, and “Language and Languages,” in Language, Mind and Knowledge: Minnesota Studies in Philosophy of Science, Volume 7, ed. K. Gunderson (Minneapolis: Minnesota University Press, 1975). See also Miller, “Lewis on Conventions,” “Conventions, Interdependence of Action and Collective Ends,” and “On Conventions.” Indeed, a good deal of the material in this chapter was published in an earlier form or forms in these articles. 9. Lewis, Convention, 14, 24. 10. Ibid, 76. On the notion of a preference, see my comments in chap. 2, note 16. 11. For detailed criticisms of Lewis see Gilbert On Social Facts, chap. 6. 12. Margaret Gilbert and others have objected to the notion of conventions being the arbitrary solutions to coordination problems. See On Social Facts, 340. Lewis does not have to be committed to arbitrariness per se, other than in the sense that there is an alternative that each would conform to if the other did. But it might be a worse alternative. 13. Lewis, Convention, 24, 42, 58, 76, 78, 79, and “Language and Languages,” 5. 14. Lewis, Convention, 95. 15. Ibid., 70. 16. Ibid., 76. 17. Ibid., 21, 76. 18. Ibid., 76. 19. Ibid. 20. Ibid., 78. 21. Ibid., 45–6.
270
Notes to pp. 98–123 22. Ibid., 78. 23. Ibid., 45–6. 24. Ibid., 7. In the stag-hunting paradigm, each can either individually hunt rabbits or collectively hunt the stag. If the stag-hunting alternative is chosen, then each has to rely on the others – stags are large animals able to be killed only by many hunters acting collectively. 25. If we build in an alternative regularity such that each conforms to it if the others do, then there will be one, but only one, set of circumstances in which each agent would not prefer to conform to the first regularity given the others did not. But at best that would make conformity to this first regularity conformity to a non-conventional alternative to a convention. 26. Miller, “Rationalising Conventions.” (Much of this section is taken from this article.) The basic notion of a procedure was introduced by Grice in “Utterer’s Meaning, Sentence Meaning, Word Meaning.” The notion was developed and deployed in relation to conventions in Miller, “Conventions and Speech Acts.” See also Seumas Miller, “Conventions, Expectations and Rationality,” Southern Journal of Philosophy 25 (1987): 357–72. Michael Bratman also developed the related notion of a policy. See his Intentions, Plans and Practical Reasoning, 87–8. Gilbert speaks of a principle of action. See her On Social Facts, 373–4. 27. Miller, “Conventions and Social Contracts,” 85–106. 28. Gilbert, On Social Facts, 364. 29. Lewis, Convention, 46, 47, 84. 30. Gilbert, On Social Facts, 375. 31. Tyler Burge, “On Knowledge of Conventions,” Philosophical Review 84 (1975): 249–55. 32. See Gilbert, On Social Facts, 374.
4 Social Norms 1. For an early and impressive account (and taxonomy) of norms in general, see Raz, Practical Reason and Norms. See also Ullmann-Margalit, The Emergence of Norms. For more recent theories of social norms in particular see Jon Elster, The Cement of Society (Cambridge: Cambridge University Press, 1989), chap. 3; Philip Pettit, “Virtus Normativa: Rational Choice Perspectives,” Ethics 100 (1990): 725–55; and Coleman, The Foundations of Social Theory. Much of the discussion in the early parts of this chapter is taken from my “Social Norms.” 2. Jon Elster, “Rationality and Social Norms,” Logic, Methodology and Philosophy of Science 8 (1989): 531–52; “Rationality, Emotions and Social Norms,” Synthese 98 (1994): 196–202; and “Norms of Revenge,” Ethics 100 (1990): 862–85.
271
Notes to pp. 124–142 3. I use the terms “moral” and “ethical” interchangeably. I accept that they can in other contexts be usefully distinguished. For example, morality tends to be used to refer to minimum standards governing interaction between agents; ethics tends to be used to embrace aspirational values, including personal values. 4. H.L. A.Hart,ConceptofLaw(Oxford:OxfordUniversityPress,1961),84–5. 5. Elster, “Norms of Revenge.” 6. Lewis, Convention, 97. 7. T. Scanlon, “Promises and Practices,” Philosophy and Public Affairs 19, no. 3 (1990): 199–226. 8. Emile Durkheim, The Rules of Sociological Method (Glencoe, Ill.: Free Press, 1965). 9. Elster, “Norms of Revenge.” 10. C. M. Turnbull, The Mountain People (London: Jonathan Cape, 1973). 11. A further version of the self-interest view invokes sanctions in the narrow sense of punitive measures such as physical punishment, but excluding social pressure; social norms are regularities backed by sanctions. But many social norms do not have sanctions in this sense and punitive measures are often inadequate if people believe conformity is inimical to their interests and/or morally unacceptable. 12. See Pettit, “Virtus Normativa.” 13. See Seumas Miller (with J. Blackler), “Restorative Justice: Retribution, Confession and Shame,” in Restorative Justice: From Philosophy to Practice, ed. H. Strang and J. Braithwaite (Aldershot: Ashgate, 2000), 77–93. 14. There is a considerable literature on common knowledge and related notions. See, for example, Lewis, Convention, 97–100; and N. Smith, Mutual Knowledge. 15. There is an important and long-standing issue concerning the possibility of beliefs – as opposed to desires and the like – providing the motivation for actions. This issue connects with another issue, namely, cognitivism in relation to moral “judgments.” Crudely, how can moral “judgments” motivate action, if they are cognitive states and actions are motivated by affective states? See Michael Smith, The Moral Problem (Cambridge: Cambridge University Press, 1994). 16. For this kind of notion, see Searle, Intentionality, chap. 5, and The Construction of Social Reality, chap. 6. 17. See Seumas Miller, “Corruption and Anti-corruption in the Profession of Policing,” Professional Ethics 6, nos. 3 and 4 (1998): 83–107. 18. See J. Katz, Seductions of Crime (New York: Basic Books, 1988). 19. H. L. A. Hart, “Are There Any Natural Rights?” in Political Philosophy, ed. A. Quinton (Oxford: Oxford University Press, 1967), 53–66; and John Rawls “Legal Obligation and the Duty of Fair Play,” in Law and
272
Notes to pp. 142–152
20.
21.
22. 23. 24. 25. 26. 27. 28.
29. 30. 31. 32. 33.
34. 35.
36.
37.
Philosophy: A Symposium, Proceedings of the 6th Annual New York University Institute of Philosophy, ed. Sidney Hook (New York: New York University Press, 1964) 3–18. R. Nozick, Anarchy, State and Utopia (Oxford: Blackwell, 1974), 90–1, and A. J. Simmons, Moral Principles and Political Obligations (Princeton: Princeton University Press, 1979), 101–2. David Luban, Lawyers and Justice: An Ethical Study (Princeton: Princeton University Press, 1988), 35–49. An earlier version of the argument in this section appeared in my article, “David Luban on the Obligation to Obey the Law,” Bulletin of the Australian Society of Legal Philosophy 18, no. 60 (1993): 10–19. Luban, Lawyers and Justice, 35. Ibid., 41. Ibid., 38, 41. Ibid., 43–4. Ibid., 38–9. Simmons, Moral Principles, 125. Luban, Lawyers and Justice, 38; and Simmons, Moral Principles, 122–5. According to Simmons, to be an active or real participant one must have at least tacitly consented or played some sort of active role in the cooperative scheme (Simmons, Moral Principles, 123.) Luban, Lawyers and Justice, 38. Ibid., 39. Ibid. Ibid., 42. Such an agent needs to be distinguished from the exploiting nonparticipant, the agent who takes benefits whenever he or she can but never contributes even if it is necessary for the scheme to succeed. Such an agent is a parasite rather than a free-rider. Simmons, Moral Principles, 122. Cooperative schemes are best understood as a species of joint actions; joint actions involve a collective end that is only achieved by the actions of the individuals, each agent only contributing if the others do. See Miller, “Joint Action,” and “On Conventions.” See also Chapter 2 of this book. This is not an example of the individual having an excuse for not contributing, as would be the case if he or she were ill or had to mind the children. See Luban, Lawyers and Justice, 45. See Robert Audi, Practical Reasoning (New York: Routledge, 1989), 24– 5. An earlier version of the argument in this section appeared in my “Social Norms and Practical Reason,” Educational Philosophy and Theory 31, no. 3 (1999): 313–27.
273
Notes to pp. 152–168 38. Ibid., chap. 3. 39. Ibid. 40. Or at least conclude that one ought to perform the action x. Here x is the best means to the end. 41. Or at least that each xing is the best means to realising e. Here the notion of a best means is fully determined by the end e for example that x is better than (say) y because x more fully realises e. 42. Here the conditionality of A having end e on B having end e results from the fact that, in order for A’s end e to be realised, B has to perform some action, x. But B only performs x if doing so will realise B’s end e. Thus, A having end e is (indirectly) dependent on B’s having end e and vice versa for B’s having end e. 43. Bratman, Intentions, Plans and Practical Reason, chap. 6. Bratman expresses a similar point, regarding what he calls policies, to the one I refer to as the first constraint.
5 Organizations, Agency, and Action 1. On the issue of reducibility of social entities, see Ruben, Metaphysics of the Social World. See also Seumas Miller, “Against Collective Agency,” Grazer Philosophische Studien (in press). 2. See Tuomela, Importance of Us, chap. 8, for one account of roles. 3. See Harre´, Social Being, 37–43, for an account of structure. 4. See Seumas Miller, John Blackler, and Andrew Alexandra, Police Ethics (Sydney: Allen & Unwin, 1997), chap. 9; Seumas Miller, “Authority, Discretion and Accountability: The Case of Policing,” in Public Sector Ethics: Finding and Implementing Values, ed. C. Sampford, N. Preston and C. Bois (London: Routledge, 1998), 37–53. 5. See also Gilbert, On Social Facts, chap. V. 6. Peter French, Collective and Corporate Responsibility (New York: Columbia University Press, 1984), 39, 41, 48–9. An earlier version of the argument in this section appeared in my “Individualism, Collective Responsibility and Corporate Crime.” 7. For criticisms, see Larry May, “Vicarious Agency and Corporate Responsibility,” Philosophical Studies 43 (1983): 69–82. 8. French, Collective and Corporate Responsibility, 165. 9. Ibid., 110. 10. Ibid., 45. 11. Ibid., 98. 12. Ibid., 44. French frequently speaks in terms of redescription of events in terms of corporate intentions and decisions, leaving it unclear as to the ontological status of these events under their new description. At
274
Notes to pp. 168–177
13.
14.
15. 16. 17.
18. 19. 20. 21.
22.
any rate, and as already mentioned, it is clear that French’s corporate intentions are real, not reducible to the intentions of individual human beings, and not merely epiphenomenal. So corporate intentions have causal efficacy. Indeed, it is probably definitive of intentions that they have causal efficacy, and, in particular, that they cause actions. For our purposes it does not matter whether the procedures in question are thought of as recognition rules or not. At times French invokes this well-known distinction of H. L. A. Hart’s, between first order rules and recognition rules (roughly, the rules for making new rules, and so on). In Collective and Corporate Responsibility, French relies on a belief/desire theory of intentions. In his later work, Corporate Ethics (Fort Worth: Harcourt Brace, 1995), 11, he accepts the irreducibility of intentions to belief/desire pairings. But this makes no difference to my argument here. A proliferation of beliefs, desires, and intentions (or simply intentions) is as problematic as a proliferation of beliefs and belief/desire pairings. French sometimes invokes his belief/desire theory to show that corporate and individual reasons can come apart. Certainly the individual needs to take the corporation’s policies (reasons) sufficiently to heart so as to act in accordance with them. But this says nothing about the individual’s ultimate reasons for so doing. These ultimate reasons may be a desire for wealth. French, Collective and Corporate Responsibility, 65. Of course, this is not to say that corporations are not vicariously liable for such actions. Or, more specifically, “CID structures that facilitate non-programmed decision-making.” See French, Collective and Corporate Responsibility, 168. It is important to see that French is here committed to higher order procedures, and not simply higher order decisions by managers. French, Collective and Corporate Responsibility, 44. Miller, “Joint Action,” 291–5. This kind of view is propounded by Tuomela, Importance of Us, chap. 5. Grice first introduced this notion: see his “Utterer’s Meaning, Sentence-Meaning, Word-Meaning,” and his Studies in the Way of Words, 129. Grice (Studies in the Way of Words, 129) developed his notion of a resultant procedure (as opposed to a resultant action) for precisely this purpose. Philip Pettit takes the view that such procedures give rise to collective intentions and beliefs, and also collective reasons (Pettit, “The Discursive Dilemma and Social Ontology,” background paper for ESSO 2000: Social Ontology and the Common Mind, Erasmus University, Rotterdam, July 2000). For criticisms, see my “Against Collective
275
Notes to pp. 177–186 Agency,” Grazer Philosophische Studien (in press). Roughly speaking, I argue, against Pettit, that all his examples involve joint mechanisms, and that joint mechanisms can be understood individualistically in the manner I have outlined. 23. Note that some joint mechanisms involve multiple joint actions and collective ends. For example, voting for one’s leader might involve the action of voting – jointly performed with all other voters – having as a collective end that the candidate with the most votes becomes leader. But some voters might also vote for, say, Jones, having as a collective end that Jones is voted in; the latter end is collective since each of Jones’s supporters acts jointly (votes) with the rest of the supporters.
6 Social Institutions and Social Groups 1. Searle, Construction of Social Reality, 41. An earlier version of the discussion of Searle in this chapter is in my “Social Institutions,” in Realism in Action: Festschift for Raimo Tuomela, ed. M. Sintonen and P. Vlikoski (Dordrecht: Kluwer, in press). 2. Ibid., 27, 34. 3. Searle ascribes a certain priority to the institution of language – language is somehow constitutive of institutional reality: Searle, Construction of Social Reality, chap. 3. Searle’s account of institutions also involves what he calls background abilities (chap. 6). 4. Searle, Construction of Social Reality, 39. 5. Ibid., 88. 6. Ibid., 100–1. 7. For a related “collective acceptance” view, see Raimo Tuomela’s “Collective Acceptance, Social Institutions and Social Reality,” unpublished paper (Helsinki: University of Helsinki Library). 8. Searle, Construction of Social Reality, 15–16, 19. 9. Ibid., 48. 10. Searle, in a later paper (“Responses to Critics of The Construction of Social Reality: Reply to Raimo Tuomela,” Philosophy and Phenomenological Research 57 (1997): 449–58, claims that the point of the deontic power is to enable the performance of the function. But the function of money, being exchange, can be performed by shells (without deontic status), and the function of surgeons by non-accredited persons with surgical skills. That is, it is false that deontic powers are necessary for the performance of these functions. 11. Searle, Construction of Social Reality, 49–50. Searle oscillates between claiming that there is a collective imposition of substantive functionality
276
Notes to pp. 186–210
12. 13.
14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.
and claiming that there is a collective imposition of additional, ancillary deontic powers. I do not dispute the latter. Ibid., 24–6. As Raimo Tuomela points out in his unpublished paper, “Searle, Collective Intentionality and Social Institutions,” (Helsinki: University of Helsinki Library), 2, the notion of collective acceptance as used by Searle is multiply ambiguous and in need of conceptual analysis. Tuomela (with W. Balzer) offers his own account in “Collective Acceptance and Collective Social Notions,” Synthese 117 (1999): 175–205. In Searle’s terms, “the Y content is imposed on the X element by collective acceptance.” Searle, Construction of Social Reality, 104. See, for example, Richard De George, The Nature and Limits of Authority (Lawrence: University of Kansas Press, 1985). Searle, Construction of Social Reality, 91. Raimo Tuomela makes a similar point in his “Searle on Social Institutions,” Philosophy and Phenomenological Research 57 (1997): 2. Searle, Construction of Social Reality, 104. Ibid., 28. Ibid. Ibid. Mancur Olson, Logic of Collective Action (Cambridge, Mass.: Harvard University Press, 1965). The following discussion is a version of what appeared in the last section of Miller, “Joint Action.” See Miller, “On Conventions.” This notion of a socially directed action is discussed in Miller, “Sociopolitical Action, Ethics and Literature.” See Seumas Miller, “Social Norms, Corruption and Transcultural Interaction,” Theoria 92 (1998): 57–77.
7 Collective Rights 1. For discussions of these issues and others, see Michael Macdonald (ed.), Canadian Journal of Law and Jurisprudence 4, no. 2 (1991) (special issue on collective rights); Judith Baker (ed.), Group Rights (Toronto: University of Toronto Press, 1994); Ian Shapiro and Will Kymlicka (eds) Ethnicity and Group Rights, Nomos 39 (New York: New York University Press, 1997). For earlier versions of the material in this chapter see my “Collective Rights” and “Collective Rights and Minority Rights,” International Journal of Applied Philosophy 14, no. 2 (2000): 241–59. 2. In this paper I will be concerned primarily with the collective rights of
277
Notes to pp. 210–211 social groups, as opposed to the (alleged) collective rights of organisations such as corporations, governments and the like. I have elsewhere argued against the ascription of collective rights to organisations, institutions, and the like. See Miller, “Joint Action” and “Corporate Crime, the Excesses of the 80’s and Collective Responsibility,” Australian Journal of Corporate Law 5, no. 2 (1995): 39–51. See also chap. 5. 3. For a discussion of external judicial adjudication in relation to internal church decision making and the implications for the autonomy of a minority within a larger society, see Denise G. Re´aume, “Common-Law Constructions of Group Autonomy: A Case Study,” in Ethnicity and Group Rights, Nomos 39, ed. Ian Shapiro and Will Kymlicka (New York: New York University Press, 1997): 257–89. 4. Will Kymlicka, “Individual and Community Rights,” in Group Rights, ed. Judith Baker (Toronto: University of Toronto Press, 1994), 19. See also his “Liberalism, Individualism and Minority Rights,” in Law and the Community, ed. A. Hutchinson and L. Green (Toronto: Carswell, 1989): 181–204; Liberalism, Community and Culture (Oxford: Oxford University Press, 1989); and Multicultural Citizenship: A Liberal Theory of Minority Rights (Oxford: Clarendon Press, 1995). Kymlicka also contrasts his special rights with group rights. The latter are rights a group has against its members’ individual rights. The former are rights a minority community has against the larger community to which it belongs. See Kymlicka, Multicultural Citizenship, chap. 3. 5. For a discussion of a version of this view, see Thomas W. Pogge, “Group Rights and Ethnicity,” in Ethnicity and Group Rights, Nomos 39, ed. Ian Shapiro and Will Kymlicka. I disagree with Pogge if he holds – as he seems to – that neither ethnicity (defined in terms of commonality of ancestry and culture) nor religion ought to be considerations in the ascription of legal rights to groups. For example, one religious group cannot be, and ought never to be, treated as more valuable than another religious group in the political domain. His position rests on the false proposition that from a political perspective every religion is as good as every other religion. For example, certain religious groups might advocate theocracy. In relation to ethnicity, it seems to me that a common ancestry and culture are in fact considerations – though not decisive ones – that would need to be taken into account in determining group legal rights. Certainly they would have to be given more weight than – to use his example – being a Porsche driver. The reason for this is quite straightforward; an individual’s identity is in part a function of his or her ancestry and culture. But the legal right to form self-determining political communities ought surely take into consid-
278
Notes to pp. 211–214
6.
7. 8.
9.
10. 11. 12.
13.
eration the identity of the individuals contemplating such an association. There are, of course, a great number of further distinctions that can be made in relation to rights, such as claim rights, liberty rights, active rights, and passive rights, and so on. For discussions of these and other distinctions in relation to collective rights, see James Nickel, “Group Agency and Group Rights,” and Jacob T. Levy, “Classifying Cultural Rights,” both in Ethnicity and Group Rights, Nomos 39, ed. Ian Shapiro and Will Kymlicka. See Kymlicka, Multicultural Citizenship, chap. 3. Michael Macdonald, “Should Communities Have Rights? Reflections on Liberal Individualism,” in Michael Macdonald (ed.), Canadian Journal of Law and Jurisprudence 4, no. 2 (1991), 218, distinguishes collective rights from what he calls “class action rights.” Thus the members of a club might have a class action right, but the club per se has a collective right. I think we can generate collective rights from the individual, but nevertheless jointly held, rights of members of groups. But such joint rights need to be distinguished from his “class action rights.” As will become evident, the latter are much narrower than the former. Michael Macdonald, “Should Communities Have Rights?”, Michael Hartney, “Some Confusions Concerning Collective Rights,” and Jan Narveson, “Collective Rights,” all in Michael Macdonald (ed.), Canadian Journal of Law and Jurisprudence; Ian Macdonald “Group Rights,” Philosophical Papers 18, no. 2 (1989): 117–36; Vernon Van Dyke, “Collective Entities and Moral Rights: Problems in Liberal Democratic Thought,” Journal of Politics 44 (1982): 21–40; Denise G. Re´aume “The Group Right to Linguistic Security: Whose Right? Whose Duties?” in Group Rights, ed. Judith Baker (Toronto: University of Toronto Press, 1994), 1. Bartels World Affairs Fellowship Lecture, delivered at Cornell University, October 16, 1997. Ian Macdonald, “Group Rights.” See also Hartney, “Some Confusions Concerning Collective Rights,” 311. For an argument that such rights are weaker than might be thought in virtue of certain rights of immigrants, see Joseph A. Carens, “The Rights of Immigrants,” in Group Rights, ed. Judith Baker (Toronto: University of Toronto Press, 1994), 142–63. The sense of the term “universal” in use in relation to moral rights can be confusing. Sometimes it means simply that it is a right possessed by all human beings. At other times it means it is a right possessed in
279
Notes to pp. 214–217
14.
15.
16. 17.
18.
19.
conformity to the so-called principle of formal equality – treat all cases the same unless there is a relevant difference. If used in this sense perhaps all moral rights are universal, since to legitimately ascribe a moral right to some individual or group, and not a second individual or second group, would seem to depend conceptually on the provision of some difference between the first and the second individual, or the first and the second group. It is sometimes claimed that individuals, but not groups, have a moral right to be treated with formal equality. But in fact neither individuals nor groups have such a (non-trivial) moral right. Rather, if an individual or group x has a moral right to, say, good g, then, according to the principle of formal equality, individual or group y also has a right to g, unless there is a relevant difference between x and y. In that case it is not a matter of having a right to be treated with formal equality. It is simply a matter of having your individual or group right to g respected. Re´aume seems to hold that both individuals and groups have a moral right to be treated with formal equality. Re´aume, “Individuals, Groups and Rights to Public Goods,” 22. Britons may in fact have voting rights in some other countries, if they are residents and there is a historical connection. However, typically they do not have two votes, one in the United Kingdom and one in, say, Australia. On secession and the right thereof, see Allen Buchanan, Secession: The Morality of Political Divorce from Fort Sumter to Lithuania and Quebec (Boulder, Colo.: Westview, 1991). See Miller, “Joint Action” and “Collective Rights.” Joint rights also need to be distinguished from conditional individual rights. By mutual consent I might have a right to fish in your river on condition that you have a right to hunt in my woods. However, neither I nor you have a joint right; rather, we each have a conditional individual right. For one thing, the content of my right brings with it no essential reference to the content of yours. For another, I can unilaterally extinguish your right, as you can mine. Thus, the distinction between joint and collective rights is important. Moreover, as Hartney (“Some Confusions Concerning Collective Rights,” 311) points out, the distinction is in keeping with established usage. There is a further distinction that is typically made in respect of collective goods, namely the distinction between collective goods that are non-rival in consumption and ones that are rival. A non-rival good is such that the consumption by one person does not reduce the consumption by another. The former are referred to as public goods. It
280
Notes to pp. 217–220
20.
21. 22.
23. 24.
25. 26.
might be held that a good that is not non-rival in consumption must be an individual good. However, I do not accept this. There is the possibility of a good that cannot be produced other than by joint effort, and is therefore reasonably referred to as a collective good. An example is the Great Wall of China. There might also be goods that cannot be enjoyed by one person without being enjoyed by others, but are nevertheless scarce. An example would be the various kinds of relational goods, such as camaraderie within a circle of friends. Perhaps, beyond a certain optimal point, the larger the circle of friends the less possible camaraderie there is within a group. If so, at points above the threshold, each new addition to the circle diminishes the enjoyment of available camaraderie. Joseph Raz, “Rights-Based Moralities,” in Theories of Rights, ed. J. Waldron (Oxford: Oxford University Press, 1984), 187. See also his The Morality of Freedom (Oxford: Clarendon Press, 1986), chap. 8. Denise Re´aume, “Individuals, Groups and Rights to Public Goods,” University of Toronto Law Journal 38 (1988): 1–27. For a more detailed account of this view and defence against other accounts, see Miller, “Joint Action” and “Intentions, Ends and Joint Action.” See Miller, “Joint Action,” for a taxonomy of joint actions and collective ends. Re´aume, “Individuals, Groups and Rights to Public Goods.” See also Leslie Green, “Two Views of Collective Rights,” in Michael Macdonald (ed.), Canadian Journal of Law and Jurisprudence 4, no. 2 (1991), 315– 28. Re´aume, “Individuals, Groups and Rights to Public Goods,” 9. Re´aume offers an interest-based account of both individual and collective rights. See her “Group Right to Linguistic Security.” More specifically, she claims that the individual has no interest qua individual in participatory goods (p. 11). As Hartney points out, this is false. See his “Some Confusions Concerning Collective Rights.” But see Re´aume’s rejoinder (in her “Group Right to Linguistic Security,” n.11, p. 139). I prefer to avoid talk of interests in this context, partly because I am unsure how wide the notion of an interest is supposed to be. I might discover that I have deer hunting rights in the Highlands of Scotland as a result of being a descendant of Lord Muck. But where does that leave me, if I am a vegetarian? At any rate, suffice it to say here that while in my view individuals can have interests, as individuals, as participants in interpersonal relationships, and as members of social groups, I am not committed to a position that grounds all moral rights in interests.
281
Notes to pp. 220–222 27. Or rights to enforce non-participation, as in the case of social taboos, such as incest. Here there is a joint right (indeed collective right) to prevent members of the group from participating in a joint action (incest) that realises a collective end – an end that is either intrinsically bad, or at least has bad consequences. 28. Following Green, “Two Views of Collective Rights,” 324, we can usefully distinguish between a rights bearer, and the one who has the right to exercise or waive the right on behalf of the rights bearer. Children may have rights, which their parents have the right to exercise or waive on their behalf. Perhaps the (alleged) right to enforce participation in educational programs is an exercise of the parents’ right to exercise the child’s right on behalf of the child. 29. It might be argued that these cultural rights are social, as opposed to merely interpersonal. I reject this. The right to communicate is – at least in the first instance – the right of an individual mind to communicate to another individual mind. Artistic expression can, of course, take a collective form, as in the case of an orchestra. But here – first point – the right of the individuals to play music together on the basis of individual mutual consent is not dependent on membership of some prior social or cultural group. Indeed, it might be that a musical band sets out to constitute itself of individuals from diverse social and cultural groups for the purposes of generating transcultural music. Second, the right of group audiences is logically dependent on the right of an individual person to be an audience. Again, the right to friendship is – at least in the first instance – a right held between two individuals. 30. Are rights to collective goods that have not been jointly produced necessarily joint rights? I think not. If a good is contingently collective – but not jointly produced – there seems to be no reason to think a priori that the right to it must be joint. This is because an individual would have a right to the good, even if there were no other individuals. After all, the existence of the good does not depend on the efforts of other individuals. (Or in the case where a good was produced by individuals, but not jointly, a single individual could produce the good, the first individual might have a right to it, but the second individual person who could produce the good might not have a right to it in virtue of not possessing whatever properties generate the right. Consider a situation in which there are two agents, A and B. A has right to a preventative herbal cure for a life-threatening disease only A has. B has the ability to easily prepare the herbal cure, but B does not have the right to the cure since B does not have the disease.) Moreover, if others do exist this does not affect the possibility of any individual’s
282
Notes to pp. 222–226
31. 32. 33. 34.
35. 36.
enjoyment of his or her right. Hence there seems no impediment to the exercise of individual rights. What of necessarily collective goods that an agent cannot be excluded from by other agents? Here the same argument holds. Again there is apparently no impediment to the exercise of individual rights. The case of necessarily collective goods that an agent cannot be excluded from either by himself, or by other agents, is more problematic. Could there be an individual right – as opposed to a joint right – to such a collective good? The point here is that everyone has a right to the good, but if I choose not to exercise my right then you cannot exercise yours. Consider the case of “lights on” in the dormitory. Either lights are on or they are off. Moreover, everyone views lights on as a good, and as indeed a right. However, some want to exercise their right at times when others do not. Here we seem forced to conclude that there are no individual rights, since if they existed they could not in principle be exercised. Are there any joint rights? Once again there seems to be no joint right, since it is difficult to see what such a joint right would be a joint right to – given there is no right that would overrule those who wanted the lights off at a given time. Perhaps the group has a right not to have the lights cut by someone outside the group, if the group can agree. That is, there is a conditional joint right in relation to the lights. The condition is that the group reach agreement. But surely this is better described as a set of individual rights conditional on agreement. After all, given the agreement, I now have a right to lights on, independently of you. See note 4 to this chapter. Ibid. Kymlicka, Multicultural Citizenship, chap. 5. For a thorough demolition of the attempt to reduce moral rights to legal rights, see Joel Feinberg, Freedom and Fulfilment (Princeton: Princeton University Press, 1992), chap. 8. Charles Taylor, Multiculturalism and the Politics of Recognition (Princeton: Princeton University Press, 1992), 40. The existence of this moral division of labour has implications for Raz’s view that there cannot be individual rights to collective goods (Raz, “Rights-Based Moralities,” 187–90). (See also Re´aume, “Individuals, Groups and Rights to Public Goods,” 5–7.) If we assume (simplistically) that, for example, the medical profession produces the public good of a disease-free environment, the police produce the public good of law and order, the teachers and academics an educated society, and so on, then we might be positioned to argue that each individual has a right to a number of collective goods, only one of which the individual
283
Notes to pp. 226–231
37.
38. 39.
40.
41.
42. 43.
44.
45. 46.
47.
contributed to the production of. I suggest, however, that in reality we have a number of sets of joint production by a minority of a particular collective good, and the joint right of all to all of the collective goods, in virtue of mutual interdependence between the various specialist producers. For an account of the Rwandan tragedy, see Fergal Keane, Season of Blood: A Rwandan Journey (London: Viking, 1995). See also Seumas Miller, “Collective Responsibility, Armed Intervention and the Rwandan Genocide,” International Journal of Applied Philosophy 12, no. 2 (1998): 223–38. Raz, “Rights-Based Moralities,” 190; and Re´aume “Individuals, Groups and Rights to Public Goods,” 3–4. As many writers have pointed out, even limited political rights in relation to self-determination are especially problematic – in comparison with cultural rights to self-defence – in virtue of their tendency to contribute to destabilisation. But see Kymlicka, Multicultural Citizenship, chap. 9. 186–214. But see James A. Graff, “Human Rights, Peoples and the Right to SelfDetermination,” in Group Rights, ed. Judith Baker (Toronto: University of Toronto Press, 1994). Perhaps these rights can be exercised by the government on behalf of the members of the community. See Green, “Two Views of Collective Rights,” 324. I argue (in unpublished material) that the concept of a joint right can handle communal land ownership systems and sexual taboos. For a detailed discussion of different forms of representation, and one consistent with what I have to say here, see Kymlicka, Multicultural Citizenship, chap. 7. Kymlicka discusses the land rights claims, as well as other claims of minority rights, of indigenous peoples and ethnic minorities under the headings of what he terms the equality argument, the role of historical agreements, and the value of cultural diversity. See his Multicultural Citizenship, chap. 6. However, he seems to understate the extent to which historically indigenous people were in essence dispossessed of their land. For a discussion of the issue of language rights in general, see Re´aume “The Group Right to Linguistic Security.” Nathan Brett, “Language Laws and Collective Rights,” in Michael Macdonald (ed.), Canadian Journal of Law and Jurisprudence 4, no. 2 (1991): 347–60. See Hartney, “Some Confusions Concerning Collective Rights,” 313, regarding the legal rights of English speakers in Quebec who were
284
Notes to pp. 231–235 educated outside Quebec to have their children educated in English, as opposed to the legal rights of those English speakers in Quebec who were educated in Quebec. The Government of Quebec defended the narrower right against the wider one – the wider one is part of the Canadian Charter of Rights and Freedoms – on the grounds that the wider one was enacted only to protect a minority language in each province. It is not, they argued, an individual right. But the Englishspeaking minority in Quebec is not threatened, even if they are denied the wider right. Therefore, they can be denied it. This argument confuses language rights which all are entitled to – in Canada, entitlements of English and French speakers – with additional legal rights needed to protect threatened minority languages. An example of the latter might be the laws to restrict business signs in English in Quebec. See Brett, “Language Laws and Collective Rights,” and Green, “Two Views of Collective Rights,” 325. 48. See Brett, “Language Laws and Collective Rights,” and Re´aume, “The Group Right to Linguistic Security.”
8 Collective Responsibility 1. See Donald Davidson, “Freedom to Act,” in Essays on Freedom of Action, ed. Ted Honderich (London: Routledge and Kegan Paul, 1973), 37– 66 and Searle, Intentionality, chap. 3. 2. Perhaps there is also being responsible for an action in the sense of freely performing the action. This notion of a freely performed action is notoriously difficult to pin down. According to one popular line of thought a freely performed action is an action such that the agent could have done otherwise. But this line of thought has been heavily criticised. I do not have an adequate analysis of the notion of a freely performed action. See John Martin Fischer (ed.), Moral Responsibility (Ithaca, N.Y.: Cornell University Press, 1986), 14–15. For attempts to analyse free action and related notions of autonomy by recourse to higher order attitudes, see Harry G. Frankfurt, “Three Concepts of Free Action,” in Fischer, Moral Responsibility, and Gerald Dworkin, The Theory and Practice of Autonomy (Cambridge: Cambridge University Press, 1988). For accounts of one’s responsibility for one’s character see Ferdinand Schoeman (ed.), Responsibility, Character and the Emotions (Cambridge: Cambridge University Press, 1987). 3. A further distinction in this area is that made between retrospective and prospective responsibility. Retrospective responsibility is responsibility for actions that have already been performed, as opposed to responsibility for future actions (prospective responsibility).
285
Notes to pp. 236–246 4. Obviously moral responsibility is sometimes retrospective and sometimes prospective. 5. This is a version of an example provided by Dean Cocking. 6. On the notions of joint and collective responsibility, see D. E. Cooper, “Collective Responsibility,” Philosophy 43 (1968): 258–68; R. S. Downie, “Collective Responsibility,” Philosophy 44 (1969): 66–9; Joel Feinberg, Doing and Deserving (Princeton: Princeton University Press, 1970), chap. 9; Virginia Held, “Moral Responsibility and Collective Action,” in Individual and Collective Responsibility, ed. Peter A. French (Cambridge, Mass.: Schenkman Publishing Company, 1972); Dennis F. Thompson, Political Ethics and Public Office (Cambridge, Mass.: Harvard University Press, 1987), chap. 2; Michael J. Zimmerman An Essay on Moral Responsibility (Totowa, N.J.: Rowman and Littlefield, 1988) and “Sharing Responsibility,” American Philosophical Quarterly 22, no. 2 (1985): 115–22; Larry May (ed.), Collective Responsibility (Totowa, N.J.: Rowman and Littlefield, 1991). For earlier versions of the arguments in the following sections see my “Collective Responsibility,” and “Collective Moral Responsibility for Omissions,” Business and Professional Ethics Journal (in press). 7. This mode of analysis is also available to handle examples in which an institutional entity has a representative who makes an individual decision, but it is an individual decision, which has the joint backing of the members of the institutional entity, for example an industrial union’s representative in relation to wage negotiations with a company. It can also handle examples such as the firing squad in which only one real bullet is used, and it is not known which member is firing the real bullet and which merely blanks. The soldier with the real bullet is (albeit unknown to him) individually responsible for shooting the person dead. However, the members of the firing squad are jointly responsible for its being the case that the person has been shot dead. 8. Here, the moral significance of the individual action derives from the moral significance of the collective end to which the action contributes. So, strictly speaking, each individual action is morally significant qua contributory action. Naturally, each individual action may also have moral significance in itself. Thanks to Pekka Makela for this point. And each contributing agent might be fully or only partially responsible for the outcome. 9. There are less developed discussions of the first three of my four puzzles in Miller, “Collective Responsibility, Armed Intervention and the Rwandan Genocide.” 10. An agent could make a causal contribution to say, Smith’s death by stabbing Smith, fail to participate in the joint action, and yet still be
286
Notes to pp. 246–250
11. 12. 13. 14. 15.
16. 17.
ascribed some form of moral responsibility, namely individual responsibility. This would be so if the other ten agents acted independently of the agent in question, yet the agent stabbed Smith – knowing the others had done likewise – having Smith’s death as an (individual) end. See Russell Hardin, Morality within the Limits of Reason (Chicago: The University of Chicago Press, 1988), 155–60, for a contrary view. Keane, Season of Blood. Ibid., 29. Miller, “Collective Responsibility, Armed Intervention and the Rwandan Genocide.” An analogue of the concept of a layered structure of joint action can be deployed to provide solutions to related collective-action problems of the sort put forward by Frank Jackson in “Group Morality,” in P. Pettit, R. Sylvan, and J. Norman (eds.), Metaphysics and Morality: Essays in Honour of J. J. C. Smart (Oxford: Blackwell, 1987), 91–110. Assume that there are a thousand bandits who each steal a thousand beans from a thousand villagers. In the first scenario, each bandit steals a thousand beans from a single given villager. (Following Jackson, call this the vertical version.) Since the villager has no beans he dies, and the individual bandit who stole all his beans is responsible for his death. Further, the overall consequence of each bandit stealing a thousand beans is the death of all one thousand villagers. In the second scenario, each bandit steals one bean from each villager. (Following Jackson, call this the horizontal version.) The overall result is the same as in the first scenario, namely the death of the thousand villagers. However, seemingly no bandit is responsible for the death of any single villager. For the action of any one bandit in respect of one villager cannot make a difference to whether that villager lives or dies; one bean is neither here nor there in the diet of a villager. These kinds of examples do not involve either individual or collective ends, but rather foreseen consequences. Moreover, in the horizontal scenario the contribution of each of the bandits to the death of each villager is a very minimal one. Nevertheless, each bandit could contribute to a joint action, the collective end of which was the avoidance of harm to the villagers; moreover, in the light of the consequences of the bandits not doing so, each bandit ought to so contribute. So, it is at least in principle possible to ascribe collective moral responsibility to the bandits. (Each bandit would not have full responsibility.) Henry Shue, Basic Rights (Princeton: Princeton University Press, 1999). Henry Shue, “Eroding Sovereignty: The Advance of Principle,” in The Morality of Nationalism, ed. R. McKim and J. McMahan (Oxford: Oxford University Press, 1997), 340–60.
287
Notes to pp. 251–256 18. Leon Shaskolsky Sehleff, The Bystander: Behaviour, Law, Ethics (Lexington, Mass.: Lexington Books, 1978), 1–2. 19. I accept the existence of positive rights, such as the right to security and subsistence. Such rights give rise to individual and collective responsibility. For the case in favour of positive rights, and an explanation of how rights generate responsibilities, see Henry Shue, Basic Rights. Thanks to Pekka Makela for a clarifying discussion of the notion of a conditional individual obligation. 20. It might be argued that in these Kitty Genovese type cases it is not rational for any of the agents to act, and that therefore there can be no ascription of individual or joint moral responsibility for the adverse outcome of the failure to act. Why is it not rational for any agent to act in these cases? Because each should only act if he or she knows the others will act, and no one can know that the others will act. Thus A should not x unless A knows that B will x. But A cannot know that B will x. For A knows that B will x only if B knows that A will x, and ex hypothesei B does not know that A will x. So there is an irresolvable boot-strapping problem. (See, for example, Hardin, Morality within the Limits of Reason, 155–60.) However, this way of describing the problem helps itself only to the mutual knowledge of the interdependence of action; it ignores the mutual knowledge of the ends or goals of the agents involved, and such mutual knowledge of ends makes a difference to at least some central categories of collective omission cases. Consider the following patterns of reasoning. If agents A and B mutually know that each has the end e, and that A xing and B xing will realise e, then A knows the following proposition: If A xs then B will x. (And B knows that if B xs then A will x.) For if A xs, B will x because by doing so B will realise B’s end e. So A xs, and so does B. Parallel rational solutions are available for the three, four, and so on cases (up to some indeterminate nperson case). In the three-person case, A knows that if A xs, then B will reason as follows: A has xed, so if I, B, x, then C will x. (C will x, because if C xs – given A and B have xed – C will realise C’s end e.) So A xs, B xs and C xs. In the four-person case A will x and ascribe to B the reasoning A deployed in the three-person case; in the five-person case A will x and ascribe to B the reasoning A deployed in the four-person case, and so on for the six- and seven-person cases, and so on.
288
Bibliography
Aristotle. Nichomachean Ethics. Any edition. Audi, Robert. Practical Reasoning. New York: Routledge, 1989. Baker, Judith (ed.). Group Rights. Toronto: University of Toronto Press, 1994. Bell, David. Husserl. London: Routledge, 1990. Bennett, Jonathan. Linguistic Behaviour. Cambridge: Cambridge University Press, 1976. Berdyaev, Nicolas. Solitude and Society. London: Centenary Press, 1938. Bratman, Michael. Intentions, Plans and Practical Reason. Cambridge, Mass.: Harvard University Press, 1987. ———. “Shared Co-operative Activity.” Philosophical Review 101, no. 2 (1992): 327–41. ———. “Shared Intention.” Ethics 104 (1993): 97–113. ———. “I Intend That We J.” In Contemporary Action Theory, Volume 2: Social Action, edited by Cihita Holmstro¨m-Hintikka and Raimo Tuometa. Dordrecht: Kluwer, 1997, 49–65. Brett, Nathan. “Language Laws and Collective Rights.” In Michael Macdonald (ed.), Canadian Journal of Law and Jurisprudence 4, no.2 (1991): 347–60. (Special issue on collective rights). Broad, C. D. The Mind and Its Place in Nature. London: Kegan Paul, 1928. Buchanan, Allen. Secession: The Morality of Political Divorce from Fort Sumter to Lithuania and Quebec. Boulder, Colo.: Westview, 1991. Buford, Thomas O. (ed.). Essays on Other Minds. Chicago: University of Illinois Press, 1970. Burge, Tyler. “On Knowledge of Conventions.” Philosophical Review 84 (1975): 249–55. Carens, Joseph A. “The Rights of Immigrants.” In Group Rights, edited by Judith Baker. Toronto: University of Toronto Press, 1994, 142–63. Cohen, Philip R., and Levesque, Hector J., “Teamwork.” Nous 25 (1991): 487– 512. Cohen, Philip R., Morgan, Jerry, and Pollack, Martha E. (eds). Intentions in Communication. Cambridge, Mass.: MIT Press, 1990.
289
Bibliography Coleman, J. S. The Foundations of Social Theory. Cambridge, Mass.: Harvard University Press, 1990. Cooper, D. E. “Collective Responsibility.” Philosophy 43 (1968): 258–68. Davidson, Donald. “Freedom to Act.” In Essays on Freedom of Action, edited by Ted Honderich. London: Routledge and Kegan Paul, 1973, 137–56. ———. “Rational Animals.” In Actions and Events, edited by Ernest Le Pore and Brian McLaughlin. Oxford: Blackwell, 1985, 473–81. De George, Richard. The Nature and Limits of Authority. Lawrence: University of Kansas Press, 1985. Downie, R. S. “Collective Responsibility.” Philosophy 44 (1969): 66–9. Durkheim, Emile. The Rules of Sociological Method. Glencoe, Ill.: Free Press, 1965. Dworkin, Gerald. The Theory and Practice of Autonomy. Cambridge: Cambridge University Press, 1988. Elster, Jon. The Cement of Society. Cambridge: Cambridge University Press, 1989. ———. “Rationality and Social Norms.” Logic, Methodology and Philosophy of Science 8 (1989): 531–52. ———. “Norms of Revenge.” Ethics 100 (1990): 862–85. ———. “Rationality, Emotions and Social Norms.” Synthese 98 (1994): 196–202. Feinberg, Joel. Doing and Deserving. Princeton: Princeton University Press, 1970. ———. Freedom and Fulfilment. Princeton: Princeton University Press, 1992. Fischer, John Martin (ed.). Moral Responsibility. Ithaca, N.Y.: Cornell University Press, 1986. Foucault, Michel. History of Sexuality, Volume 1: An Introduction. Trans. R. Hurley. New York: Pantheon, 1978. ———. “Two Lectures.” In Power/Knowledge: Selected Interviews and Other Writings 1972–77, edited by C. Gordon. Brighton: Harvester, 1980. ———. “Order of Discourse.” In Language and Politics, edited by M. Shapiro Oxford: Blackwell, 1984. Fowler, Roger. Linguistics and the Novel. London: Methuen, 1977. Frankfurt, Harry G. “Three Concepts of Free Action.” In Moral Responsibility, edited by John Martin Fischer. Ithaca, N.Y.: Cornell University Press, 113– 24. Freadman, Richard, and Miller, Seumas. Rethinking Theory: A Critique of Contemporary Literary Theory and an Alternative Account. Cambridge: Cambridge University Press, 1992. French, Peter. Collective and Corporate Responsibility. New York: Columbia University Press, 1984. ———. Corporate Ethics. Fort Worth: Harcourt Brace, 1995. Garcia, J. “Constitutive Rules.” Philosophia 17 (1987): 251–70.
290
Bibliography Gewirth, Alan. Reason and Morality. Chicago: University of Chicago Press, 1981. Giddens, Anthony. The Constitution of Society: Outline of the Theory of Structuration. Cambridge: Polity Press, 1984. Gilbert, Margaret. On Social Facts. Princeton: Princeton University Press, 1989. ———. “Walking Together: A Paradigmatic Social Phenomenon.” In The Philosophy of the Human Sciences, edited by P. A. French, T. E. Euhling, Jr, and H. K. Wettstein, Midwest Studies in Philosophy, vol. 15. Notre Dame, Ind.: University of Notre Dame Press, 1990. ———. “What Is It for Us to Intend?” In Contemporary Action Theory, Volume 2: Social Action, edited by Ghita Holmstro¨m Hintikka and Raimo Tuomela. Dordrecht: Kluwer, 1997, 65–87. Graff, James A. “Human Rights, Peoples and the Right to Self-Determination.” In Group Rights, edited by Judith Baker. Toronto: University of Toronto Press, 1994, 186–214. Green, Leslie. “Two Views of Collective Rights.” In Michael Macdonald (ed.), Canadian Journal of Law and Jurisprudence 4, no. 2 (1991) (Special issue on collective rights), 315–28. Grice, H. P., “Utterer’s Meaning, Sentence-Meaning, and Word-Meaning.” Foundations of Language 4 (1968): 225–42. ———. Studies in the Way of Words. Cambridge, Mass.: Harvard University Press, 1989. Hardin, Russell. Collective Action. Baltimore: Johns Hopkins University Press, 1982. ———. Morality within the Limits of Reason. Chicago: University of Chicago Press, 1988. Harre´, Rom. Social Being. Oxford: Blackwell, 1979. Hart, H. L. A. Concept of Law. Oxford: Oxford University Press, 1961. ———. “Are There Any Natural Rights?” In Political Philosophy, edited by A. Quinton. Oxford: Oxford University Press, 1967, 53–66. Hartney, Michael. “Some Confusions Concerning Collective Rights.” In Michael Macdonald (ed.), Canadian Journal of Law and Jurisprudence 4, no. 2 (1991) (special issue on collective rights), 293–314. Heal, J. “Common Knowledge.” Philosophical Quarterly 28 (1978): 116–31. Held, Virginia. “Moral Responsibility and Collective Action.” In Individual and Collective Responsibility, edited by Peter A. French. Cambridge, Mass.: Schenkman Publishing Company, 1972. ———. “The Normative Import of Action.” In Gewirth: Critical Essays on Action, Rationality and Community, edited by Michael Boylan. Oxford: Rowman and Littlefield, 1999, 13–28. Hobbes, Thomas. Leviathan. Any edition. Hobbs, J. R. “Artificial Intelligence and Collective Intentionality: Comments on
291
Bibliography Searle and on Grosz and Sidner.” In Intentions in Communication, edited by Philip R. Cohen, Jerry Morgan, and Martha Pollack. Cambridge, Mass.: MIT Press, 1990, 445–59. Hume, David. A Treatise on Human Nature. Any edition. Husserl, Edmund. Cartesian Meditations. Trans. Dorion Cairns. The Hague: Martinus Nijhoff, 1960. Jackson, Frank. “Group Morality.” In Metaphysics and Morality: Essays in Honour of J. J. C. Smart, edited by P. Pettit, R. Sylvan, and J. Norman. Oxford: Blackwell, 1987, 91–110. Kant, Immanuel. Critique of Pure Reason. Trans. J. Meiklejohn. New York: Wiley, 1943. Katz, J. Seductions of Crime. New York: Basic Books, 1988. Keane, Fergal. Season of Blood: A Rwandan Journey. London: Viking, 1995. Kripke, Saul. Wittgenstein on Rules and Private Language. Oxford: Blackwell, 1982. Kymlicka, Will. “Liberalism, Individualism and Minority Rights.” In Law and the Community, edited by A. Hutchinson and L. Green. Toronto: Carswell, 1989, 181–204. ———. Liberalism, Community and Culture. Oxford: Oxford University Press, 1989. ———. “Individual and Community Rights.” In Group Rights, edited by Judith Baker. Toronto: University of Toronto Press, 1994, 17–33. ———. Multicultural Citizenship: A Liberal Theory of Minority Rights. Oxford: Clarendon Press, 1995. Levy, Jacob T. “Classifying Cultural Rights.” In Ethnicity and Group Rights, edited by Ian Shapiro and Will Kymlicka. New York: New York University Press, 1997, 22–66. Lewis, David. Convention: A Philosophical Study. Cambridge, Mass.: Harvard University Press, 1969. ———. “Language and Languages.” In Language, Mind and Knowledge: Minnesota Studies in Philosophy of Science, Volume 7, edited by K. Gunderson. Minneapolis: University of Minnesota Press, 1975, 3–35. Locke, John. Of Civil Government: Book 2 – An Essay Concerning the True Original Extent and End of Civil Government. Any edition. Luban, David. Lawyers and Justice: An Ethical Study. Princeton: Princeton University Press, 1988. Macdonald, Ian. “Group Rights.” Philosophical Papers 18, no. 2 (1989): 117–36. Macdonald, Michael (ed.). Canadian Journal of Law and Jurisprudence 4, no. 2 (1991) (special issue on collective rights). ———. “Should Communities Have Rights? Reflections on Liberal Individualism.” In Michael Macdonald (ed.), Canadian Journal of Law and Jurisprudence 4, no. 2 (1991) (special issue on collective rights), 217–37.
292
Bibliography McGinn, Colin. Wittgenstein on Meaning. Oxford: Blackwell, 1984. Macyntyre, Alasdair. After Virtue. Notre Dame, Ind.: University of Notre Dame Press, 1984. Makela, Pekka. “Collective Moral Responsibility: A Collective as an Independent Moral Agent.” Australian Journal of Professional and Applied Ethics (forthcoming). May, Larry. “Vicarious Agency and Corporate Responsibility.” Philosophical Studies 43 (1983): 69–82. ———. The Morality of Groups. Notre Dame, Ind.: University of Notre Dame Press, 1987. ———. (ed.). Collective Responsibility. Totowa, N.J.: Rowman and Littlefield, 1991. Mead, George H. Mind, Self and Society. Chicago: University of Chicago Press, 1934. Miller, Seumas. “Lewis on Conventions.” Philosophical Papers 11, no. 2 (1982): 1–8. ———. “Performatives.” Philosophical Studies 45, no. 2 (1984): 247–60. ———. “Conventions and Speech Acts” (University of Melbourne PhD, 1985; Ann Arbor: University Microfilms International, 1986). ———. “Speaker Meaning and Assertion.” South African Journal of Philosophy 4, no. 2 (1985): 48–55. ———. “Truthtelling and the Actual Language Relation.” Philosophical Studies 49, no. 2 (1986): 281–94. ———. “Conventions, Interdependence of Action and Collective Ends.” Nous 20, no. 2 (1986): 117–40. ———. “Conventions, Expectations and Rationality.” Southern Journal of Philosophy 25 (1987): 357–72. ———. “Conventions and Social Contracts.” Philosophical Papers 16, no. 2 (1987): 85–106. ———. “Rationalising Conventions.” Synthese 84, no. 1 (1990): 23–41. ———. “Foucault on Discourse and Power.” Theoria 76 (1990): 115–25. ———. “On Conventions.” Australasian Journal of Philosophy 70, no. 4 (1992): 435–44. ———. “Joint Action.” Philosophical Papers 21, no. 3 (1992): 275–97. ———. “David Luban on the Obligation to Obey the Law.” Bulletin of the Australian Society of Legal Philosophy 18, no. 60 (1993): 10–19. ———. “Intentions, Ends and Joint Action.” Philosophical Papers 24, no. 1 (1995): 51–67. ———. “Sociopolitical Action, Ethics and the Power of Literature.” Philosophy and Social Criticism 21, no. 3 (1995): 93–110.
293
Bibliography ———. “Corporate Crime, the Excesses of the 80’s and Collective Responsibility.” Australian Journal of Corporate Law 5, no. 2 (1995): 39–51. ———. “Social Norms.” In Contemporary Action Theory, Volume 2: Social Action, edited by Ghita Holmstro¨m-Hintikka and Raimo Tuomela. Dordrecht: Kluwer, 1997, 211–29. ———. “Individualism, Collective Responsibility and Corporate Crime.” Business and Professional Ethics Journal 16, no. 4 (1997): 19–46. ———. “Authority, Discretion and Accountability: The Case of Policing.” In Public Sector Ethics: Finding and Implementing Values, edited by C. Sampford, N. Preston, and C. Bois. London: Routledge, 1998, 37–53. ———. “Collective Responsibility, Armed Intervention and the Rwandan Genocide.” International Journal of Applied Philosophy 12, no. 2 (1998): 223–38. ———. “Corruption and Anti-Corruption in the Profession of Policing.” Professional Ethics 6, nos. 3 & 4 (1998): 83–107. ———. “Social Norms, Corruption and Transcultural Interaction.” Theoria 92 (1998): 57–77. ———. “Collective Rights.” Public Affairs Quarterly 1, no. 4 (1999): 331–417. ———. “Social Norms and Practical Reason.” Educational Philosophy and Theory 31, no. 3 (1999): 313–27. ———. “Restorative Justice: Retribution, Confession and Shame” (with J. Blackler). In Restorative Justice: From Philosophy to Practice, edited by H. Strang and J. Braithwaite. Alsershot: Ashgate, 2000, 77–93. ———. “Speech Acts and Conventions.” Language Sciences 22, no. 2 (2000): 155–66. ———. “Collective Responsibility.” Public Affairs Quarterly 15, no. 1 (2001): 65– 82. ———. “Against Collective Agency.” Grazer Philosophische Studien (in press). ———. “Collective Moral Responsibility for Omissions.” Business and Professional Ethics Journal (in press). ———. “Collective Rights and Minority Rights.” International Journal of Applied Philosophy 14, no. 2 (2000): 241–58. ———. “Social Institutions.” In Realism in Action: Festschrift for Raimo Tuomela, edited by M. Sintonen and P. Ylikoski. Dordrecht: Kluwer (in press). Miller, Seumas, Blackler, John, and Alexandra, Andrew. Police Ethics. Sydney: Allen & Unwin, 1997, 1–240. Miller, Seumas, and Freadman, Richard. Rethinking Theory: A Critique of Contemporary Literary Theory and an Alternative Account. Cambridge: Cambridge University Press, 1992, 1–296. Narveson, Jan. “Collective Rights.” In Michael Macdonald (ed.). Canadian Journal of Law and Jurisprudence 4, no. 2 (1991) (special issue on collective rights), 329–45.
294
Bibliography Nickel, James. “Group Agency and Group Rights.” In Ethnicity and Group Rights, Nomos 39, edited by Ian Shapiro and Will Kymlicka. New York: New York University Press, 1997, 235–56. Nozick, R. Anarchy, State and Utopia. Oxford: Blackwell, 1974. Olson, Mancur. Logic of Collective Action. Cambridge, Mass.: Harvard University Press, 1965. Parfit, Derek. Reasons and Persons. Oxford: Oxford University Press, 1984. Parsons, Talcott. The Structure of Social Action: A Study of Social Theory with Special Reference to a Group of Recent European Writers, Volumes 1 and 2. New York: Free Press, 1968. Pettit, Philip. “Virtus Normativa: Rational Choice Perspectives.” Ethics 100 (1990): 725–55. ———. The Common Mind. Oxford: Oxford University Press, 1993. Pogge, Thomas W. “Group Rights and Ethnicity.” In Ethnicity and Group Rights, Nomos 39, edited by Ian Shapiro and Will Kymlicka. New York: New York University Press, 1997, 187–221. Rawls, John. “Legal Obligation and the Duty of Fair Play.” In Law and Philosophy: A Symposium, Proceedings of the 6th Annual New York University Institute of Philosophy, edited by Sidney Hook. New York: New York University Press, 1964, 3– 18. Raz, Joseph. Practical Reason and Norms. London: Hutchinson, 1979. ———. “Rights-Based Moralities.” In Theories of Rights, edited by J. Waldron. Oxford: Oxford University Press, 1984, 182–200. ———. The Morality of Freedom. Oxford: Clarendon Press, 1986. Re´aume, Denise G. “Individuals, Groups and Rights to Public Goods.” University of Toronto Law Journal 38 (1988): 1–27. ———. “The Group Right to Linguistic Security: Whose Right? Whose Duties?” In Group Rights, edited by Judith Baker. Toronto: University of Toronto Press, 1994, 118–41. ———. “Common-Law Constructions of Group Autonomy: A Case Study.” In Ethnicity and Group Rights, Nomos 39, edited by Ian Shapiro and Will Kymlicka. New York: New York University Press, 1997, 257–89. Ross, Alf. Directives and Norms. London: Routledge, 1968. Rousseau, Jean-Jacques. The Social Contract. Any edition. ———. Discourse on the Origins of Inequality. Any edition. Ruben, David-Hillel. The Metaphysics of the Social World. London: Routledge and Kegan Paul, 1985. Sandel, Michael. Liberalism and the Limits of Justice. Cambridge: Cambridge University Press, 1982. Scanlon, T. “Promises and Practices.” Philosophy and Public Affairs 19, no. 3 (1990): 199–226.
295
Bibliography Schiffer, S. Meaning. Oxford: Oxford University Press, 1972. Schmidtz, David. “Choosing Ends.” Ethics 104 (1994): 226–51. Schoeman, Ferdinand (ed.). Responsibility, Character and the Emotions. Cambridge: Cambridge University Press, 1987. Schutz, Alfred. The Phenomenology of the Social World. Trans. George Walsh and Frederick Lehnert. Chicago: Northwestern University Press, 1967. Searle, John R. Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge University Press, 1969. ———. Intentionality: An Essay in the Philosophy of Mind. Cambridge: Cambridge University Press, 1983. ———. “Collective Intentions and Actions.” In Intentions in Communication, edited by Philip R. Cohen, Jerry Morgan, and Martha E. Pollack. Cambridge, Mass.: MIT Press, 1990, 401–15. ———. The Construction of Social Reality. London: Penguin, 1995. ———. “Responses to Critics of The Construction of Social Reality: Reply to Raimo Tuomela.” Philosophy and Phenomenological Research 57 (1997): 449–58. Sehleff, Leon Shaskolsky. The Bystander: Behaviour, Law, Ethics. Lexington, Mass.: Lexington Books, 1978. Shapiro, Ian, and Kymlicka, Will (eds). Ethnicity and Group Rights, Nomos 39. New York: New York University Press, 1997. Shue, Henry. Basic Rights. Princeton: Princeton University Press, 1996. ———. “Eroding Sovereignty: The Advance of Principle.” In The Morality of Nationalism, edited by R. McKim and J. McMahan. Oxford: Oxford University Press, 1997, 340–60. Shwayder, David. The Stratification of Behaviour. London: Routledge and Kegan Paul, 1965. Simmel, George. The Sociology of George Simmel. Trans. Kurt H. Wolff. Glencoe, Ill.: The Free Press, 1950. Simmons, A. J. Moral Principles and Political Obligations. Princeton: Princeton University Press, 1979. Smith, Michael. The Moral Problem. Cambridge: Cambridge University Press, 1995. Smith, N. (ed.). Mutual Knowledge. London: Academic Press, 1982. Sperber, D., and Wilson, D. Relevance. Oxford: Blackwell, 1986. Strawson, P. F. Logico-Linguistic Papers. London: Methuen, 1971. Tannsjo, Torbjorn, “The Morality of Collective Action.” The Philosophical Quarterly 39, no. 155 (1989): 221–8. Taylor, Charles. Philosophy and the Human Sciences: Philosophical Papers 2. Cambridge: Cambridge University Press, 1985.
296
Bibliography ———. “Foucault on Freedom and Truth.” In Foucault: A Critical Reader, edited by D. C. Hoy, Oxford: Blackwell, 1986. ———. Multiculturalism and the Politics of Recognition. Princeton: Princeton University Press, 1992. ———. Sources of the Self. Cambridge: Cambridge University Press, 1992. Thompson, Dennis F. Political Ethics and Public Office. Cambridge, Mass.: Harvard University Press, 1987. Tuomela, Raimo. A Theory of Social Action. Dordrecht: Reidel, 1984. ———. “What Are Goals and Joint Goals?” Theory and Decision 28, no. 1 (1990): 1–20. ———. The Importance of Us. Stanford: Stanford University Press, 1995. ———. “Searle on Social Institutions.” Philosophy and Phenomenological Research 57 (1997): 435–41. ———. “Collective Acceptance, Social Institutions and Social Reality.” Unpublished paper. Helsinki: University of Helsinki Library. ———. “Searle, Collective Intentionality and Social Institutions.” Unpublished paper. Helsinki: University of Helsinki Library. Tuomela, Raimo (with Balzer, W.) “Collective Acceptance and Collective Social Notions.” Synthese 117 (1999): 175–205. Tuomela, Raimo (with Miller, Kaarlo). “We-intentions.” Philosophical Studies 53, no. 3 (1988): 367–89. Turnbull, C. M. The Mountain People. London: Jonathan Cape, 1973. Ullmann-Margalit, Edna. The Emergence of Norms. Oxford: Oxford University Press, 1977. Van Dyke, Vernon. “Collective Entities and Moral Rights: Problems in LiberalDemocratic Thought.” Journal of Politics 44 (1982): 21–40. Walzer, Michael. Spheres of Justice. New York: Basic Books, 1983. Weber, Max. Economy and Society: An Outline of Interpretive Sociology, edited by G. Roth and C. Wittich. Berkeley: University of California Press, 1978. Winch, Peter. Idea of a Social Science. London: Routledge and Kegan Paul, 1958. Wittgenstein, Ludwig. Philosophical Investigations. Any edition. Zimmerman, Michael J. “Sharing Responsibility.” American Philosophical Quarterly 22, no. 2 (1985): 115–22. ———. An Essay on Moral Responsibility. Totowa, N.J.: Rowman and Littlefield, 1988.
297
Index
When discussion examples are numerous, they are listed under the topic separately from other subheadings. Endnotes are listed as n. or, if more than one per page is listed, as nn. aboriginals, Australian, 212, 230–1 acceptance, collective, 277n.13; as substantive functionality, 189–90, 191, 199 Afrikaners, 223 agent preferences, 101–7 agents: interactions between, 76; responsible for realizing collective ends, 238–9. See also moral agency; rational agency agent-semantics, 15, 38, 265n.54 Ainu of Japan, 208–9 apartheid system, 22–3, 31, 141, 150, 202, 224 Aristotle, 26 artistic expression, 228, 282n.29 attention-gaining behavior, 133–4 attitudes: affective, 61; associated with roles, 172–3; collective/interdependent, 134, 137, 154–5; propositional attitudes, 36, 165 Australian aboriginals, 212, 230–1 authority: whether involving collective or individual responsibility, 240–1; institutional, 188–9, 192–3 autonomy: moral, 46–7, 48; social, 202 Being There (film), 188 beliefs, 13, 125, 266n.8, 272n.15; belief/desire theory, 267n.18,
275n.14. See also intentions; mutual knowledge Bennett, Jonathan, 38 Bosnia, 213 Bratman, Michael: belief/desire theory of action, 267n.18; ; counterexamples, 77–80; on future directed intentions, 155–6; shared intentions model of, 59, 69, 74–7, 81, 267n.15 Broad, C. D., 4 capitalism, 196 Chinese: in Australia, 229–30; Chinese Triads, 202; in Taiwan and Hong Kong, 208–9 Cohen, Philip R. and Levesque, Hector J., mutual goals and knowledge theory of, 81–3, 269nn.43, 44 collective actions, 53; Kitty Genovese (example), 251–2, 288n.20 collective ends: agents’ responsibility for realizing, 238–9, 244; characteristics of, 10, 13, 21, 57–8, 148– 50, 194; collective goods comprising, 145, 219; higher order, 63, 199; individual ends and, 69, 162, 195–6, 266–7n.10; intentions distinguished from, 65–71, 80–1; joints actions to attain, 24, 58–9, 60, 61, 152; moral value adhering
299
Index collective ends (cont. ) in, 148, 151, 156, 194. See also collective goods; intentions; outcomes collective ends (examples): auto factory, 195–6; fallen tree, 64, 89; Great Wall of China, 63, 281n.19; the rowers and the cox, 70–1; soldiers killing sniper, 58–9; stabbing of Smith, 246–7, 286–7n.10 collective end theory (CET) of joint action, 36, 56, 57–62, 142, 246; belief/desire theories distinguished from, 60–1; core theory and applications, 62–4, 266nn.6, 7; corporatist theories distinguished from, 58, 59, 64; individualist theories distinguished from, 58, 59. See also collective ends; joint action; teleological account of social action collective goods, 145, 217–19; joint actions themselves as, 219–20; joint rights to, 211, 216–17, 218–19, 233, 282–3n.30 collective intentionality, 59, 84, 187– 90; deontic properties and, 187, 276n.10; rules and, 185, 190–2; status-functions and, 184–7, 192 collective intentionality (examples): mediums of exchange, 184, 185, 189–90; truth telling, 187–8; wouldbe capitalists, 84 collective (moral) responsibility, 1, 46, 241, 243; for omissions, 250; for outcomes of joint action, 247–50; for realizing collective ends, 238–9. See also overdetermination collective rights, 2, 210–12, 279n.8; adhering to individual members, 161, 212–14, 292n.28; collective moral rights, 31–2, 46, 210, 226, 233; dependant upon collective membership, 214–15; joint rights and, 200–1, 280–1n.19; minority
rights and, 223–6; relationship to individual rights, 212, 214–16; whether a social group has, 202–3. See also joint rights collective rights (examples): Australian right of exclusion, 214, 216; Kurds’ right of secession, 214, 216 committee decisions, 241 communication: language enabling, 23; sphere of, 163, 181–2. See also language; media communist regimes, 198, 221 communities, overlapping of, 229–30 competition, 133–4, 196 conditional actions, 58 conditional preferences, 94, 105–6, 108; and collective ends, 110–12 conflict, 23 conformity, 69, 98–9, 100, 101, 107. See also norms, social; rules constitutively social actions, 8 constitutive rules, 42–5, 191; log as bench (example), 43 contingently social action, 2 contributory actions, 238, 245 conventions, 2, 22, 25–6, 38, 120–1; collective end theory of, 107, 110– 12, 129; conventional alternatives, 95, 113, 115, 121–2; interdependence of action in, 104–5, 112; joint acceptance theory of, 93–4, 121; as joint procedures, 115–19; nonconventional alternatives, 95–7, 101, 102, 108–9, 115; normative force of, 122, 126–7, 128–30; social contract theory and, 91–2, 95; structures of, 97, 119–22, 130, 206; as tacit agreements, 92. See also coordination problems and conventions; linguistic conventions conventions (examples): driving on the right side, 110, 129; motorists, 118 cooperative schemes, 152, 273n.35; laws as, 142–6, 148–9
300
Index cooperative schemes (examples): flower gardening, 144, 146; glass cleanup, 144, 146 coordination problems and conventions, 270n.12; action-determining vs. non-action-determining preferences, 101–3, 105–7, 107–10; formal coordination problems, 95, 98, 101; informal coordination problems, 94–5, 98, 101, 112–13; Lewis’s theory on, 94–5, 102, 107, 113, 114; Lewis’s theory – counter arguments to, 95–101, 102–4, 106– 7, 109; minority vs. majority preferences, 99–100, 102–4, 106 coordination problems and conventions (examples): clothes and accessories, 97–9, 107, 109–10, 120– 1; double-marking case, 114–15; gang of thieves, 114; meeting place, 102–3, 108; oligopoly case, 114; rabbit hunt, 113–14; stag hunt, 99, 113–14, 271n.24; telephone case, 108 corporate actions, 164, 171–3; joint actions distinguished from, 164; management and, 169–70 corporate actions (examples): Air New Zealand disaster, 170; design fault in a car, 169; employees, 168– 9; Gulf Oil, 165–6, 168; production of goods no longer wanted, 167 corporate entities, 2; hierarchical, 161; mental states ascribed to, 167– 71; moral agency ascribed to, 166– 8; as role structures, 171–3 corporate internal decisions (CID) theory, 165–6, 168, 169–71, 174, 274–5n.12 corporatist theories, 53, 64. See also by theory eg., collective intentionality corruption, 140 criminal justice system, 141 cultural groups: differences between, 51; self-defence or preservation of,
223–4, 226, 227–8; self-identity of, 224; social groups distinguished from, 209. See also minority cultural groups culture/society, 209, 220–1, 261n.7 deductive reasoning, 152 deontic properties: collective intentionality and, 187, 276n.10; of status-functions, 184–7, 192 desires, 60–1, 72–4; desire for approval, 124, 130; for status, 139, 140 Durkheim, Emile, 130 economic behavior, 163–4, 196 economic institutions, 181, 194; transsocietal, 208–9 economic justice, 196–7 educational institutions, 21, 181, 194, 195 Elster, Jon, 131 ends. See collective ends European Union (EU), 30, 182 ‘‘extraspective’’ interactions, 4 fair play argument, 142, 149 Foucault, Michel, on social power, 19– 21, 189, 261nn. 2, 5 Fowler, Roger, 259n.3 freedom, 49, 285n.2 free-riding, 89, 129, 144, 146, 196; non-participation and, 146, 149, 150; as quasi-joint action, 55 French, Peter: corporate internal decisions (CID) theory, 165–6, 168, 169–71, 174, 274–5n.12; on organizations as agents, 6, 164. See also corporate actions friendship, 3, 220–2 functions. See status-functions future directed intentions, 47, 76, 155–6
301
Index game-theory, 101 Genovese, Kitty murder (example), 251–2, 288n.20 Gewirth, Alan, 260n.12 Giddens, Anthony, 18, 30–1 Gilbert, Margaret, 32–3, 270nn.6, 12; on institutional normative force, 122, 127; joint acceptance theory of conventions, 93–4; on joint actions as social actions, 36–7, 38–9; objections to, 88–9, 90; plural subjects theory, 6, 32–3, 58, 85–90; plural subjects theory – objections to, 87, 88–9, 90 government, 119, 120, 198–9 Grice, H. Paul, 15, 38, 275–6n.22 Hart, H. L. A., 124, 142 Hartney, Michael, 212 health, environmental threats to, 195 hierarchical organizations, 161–2 historical conception of moral values, 12–13, 49–51 Hobbes, Thomas, 91–2, 95 humanitarian intervention, 46, 210; moral responsibility in, 252–6, 257. See also Rwandan genocide human rights: collective rights and, 210; individual, 214, 221; international, 250–1; right to life, 217, 226. See also Rwandan genocide Hume, David, 26, 94 Hutus, 248, 249. See also Rwandan genocide Immorality and Mixed Marriages Act (South Africa), 141. See also apartheid system Indians, in Malaysia and Fiji, 208 individual actions: collective actions distinguished from, 53; constituting social actions, 6–7, 160; interdependence of, 56; carpooling (example), 56; social actions distinguished from, 1–3, 5, 7
individual ends, 125; collective ends and, 69, 162, 195–6, 266–7n.10 individual intentions model, 58–9, 64, 71–4; women exercising (example), 73–4 individual (moral) responsibility, 243, 254 individual rights: to self-defence, 203. See also human rights institutional accountability, 237, 335– 6 institutional joint actions, 160, 240; detective and drug ring (example), 240; social consequences of, 194–7 institutional norms, 89–90, 124, 182; firefighter (example), 89 institutions: authority relations in, 188–9, 192–3; defined, 22, 28–31, 163–4, 181, 182; descriptive theory of, 193; hierarchical, 192; Marxist account of, 194–5; normative dimensions of, 184–7, 192, 193, 193– 4, 197–8; organizations distinguished from, 163–4; structures and interrelationships of, 28, 198– 9, 225; substantive functionality of, 189–90; teleological account of, 193–4, 197–9 instrumental reasoning, 153, 265n.56 intellectual freedom, 221–2 intentional joint actions, 33–4, 253–4 intentions, 13, 40–1, 59, 60–1, 125, 268n.26; and collective ends, 65, 69; future directed, 47, 76, 155–6; higher order, 64–5; individual intentions model, 58–9, 64, 71–4; mental states and, 77–9; ‘‘weintentions,’’ 57, 72, 74, 80, 187; weak intentions, 66, 66–7 intentions (examples): arm raising, 67; A shoots B, 60–1; college applications, 66; dart playing, 65–6; John in love with Mary, 60–1 interactive joint actions, 63–4 interdependence of actions: of indi-
302
Index vidual actions, 6, 56, 60; of joint actions, 34–5, 239, 244, 245–6; social norms involving, 131, 132 intergenerational history, 200–1 Interhamwe, 247, 248. See also Rwandan genocide international social conflicts, 253, 256 interpersonal interactions, 3–4, 5, 118 interpersonal rights, 221–2 intervention. See humanitarian intervention Jackson, Frank, 287n.15 Japan, 199; Ainu cultural group of, 208–9 joint actions, 59; basic or generic, 2, 38, 53, 69–71, 181; defined, 17, 24– 5, 53–4, 164, 238, 262n.13; directed toward collective ends, 24, 58–9, 60, 61, 152; interdependence of, 34–5, 56, 239, 244, 245–6; social actions and, 34, 36–7; structures/layered structures of, 173–4, 249–50, 256–7, 287n.15; as in themselves collective goods, 219– 20. See also collective end theory (CET) of joint action; conventions; institutional actions; organizational actions; quasi-joint actions; social actions joint actions (examples): army fighting, 249; carpooling, 56; joint mechanisms, 54–5, 174–9, 199– 200, 276n.23; joint procedural mechanisms, 175, 176, 178–9; language consisting in part of, 177–8; resultant actions, 176–7, 178 joint mechanisms (examples): coin tossing, 54, 175, 176, 178; rugby game, 176–7, 178; voting process, 175, 176, 178–9, 199–200 joint obligations, 226–7 joint practice, 55
joint procedures, 54, 265n.1, 275– 6n.22; conventions as, 115–19. See also conventions; joint action joint rights, 216, 222, 280n.17; collective rights and, 200–1, 280– 1n.19; to collective goods, 211, 216– 17, 218–19, 233, 282–3n.30. See also collective rights joint rights (examples): Canadian citizenship, 217; fishing, 219; house building, 218–19; table lifting, 218 knowledge. See mutual knowledge Kripke, Saul, 39–40 Kurds: as a minority cultural group, 201; right of secession (example), 214, 216 Kymlicka, Will, 211, 212, 222, 223–5, 278n.4, 284n.44 land rights, 212, 284n.44; of Australian aborigines, 212 language: as an institution, 23, 164, 188–9, 199; joint mechanisms in, 177–8; linguistic conventions, 37, 117; linguistic erosion, 232–3; minority languages, 231–2; social groups sharing, 130, 200, 206, 208 laws: benefits of, 142–5; as cooperative schemes, 142–6, 148–9; fair play argument for obeying, 142, 149; legal rights of minorities, 224– 5, 230, 278–9n.5; moral obligations and, 141, 142–4, 147–8; and social norms, 124, 141–2; systems of, 149– 50 Lewis, David, 32, 33; on coordination problems and conventions, 94–5, 101, 102, 107, 113, 114; counter arguments to, 95–101, 102–4, 106– 7, 109; preference structures, 101, 105, 107, 110, 112; on the normative force of conventions, 126–7, 128–30. See also coordination problems and conventions
303
Index linguistic conventions, 37, 117. See also language Locke, John, 91 Luban, David, 142–4, 147, 149, 151 Macdonald, Ian, 212, 213–14 Macdonald, Michael, 212 macro-entities, 180. See also organizations Mafia, 140, 202 Makela, Pekka, 286n.8, 288n.19 manners, 123–4 marijuana, laws on, 142 Marxist account of institutions, 194–5 means–end reasoning, 152–3, 274nn.40,41 means–end relationships, 14, 118, 254–5; action as integral part of the means, 244–5 media, 203–5, 228 mental states, 5–6, 10, 49; ascribed to corporate entities, 167–71; higher order, 37–8; and intentions, 77–9; representational thinking, 39–40; self-awareness, 49 military organizations, 161, 226 Miller, Seumas, 266nn.6, 7, 271n.26, 284n.42 minority cultural groups, 201, 210, 211; Ainu in Japan, 208–9; Australian aboriginals, 212, 230–1; Chinese in Australia, 229–30; French Canadians, 225–6, 228 minority languages, 231–2 minority rights, 46, 210, 228–9, 278– 9n.5; indigenous people’s rights, 208, 211, 212, 223, 230–1; Kymlicka on, 222, 223–5, 278n.4; legal rights, 224–5, 230, 278–9n.5 minority vs. majority preferences, 99– 100, 102–4, 106 moral agency, 11, 27, 46–7, 49–50, 165, 260n.12, 265n.55, 285n.2; ascribed to corporate entities, 166–8; social norms and, 15, 124, 126, 128–9, 139–40, 152, 263n.22. See
also agents; moral obligations; moral principles; moral responsibility ‘‘moral’’ community, 155 moral complexity, 157–8 moral complexity (examples): helping a Jew hide from the Nazis, 157; war or invasion, 157–8 moral development, 50–1 moral division of labor, 283–4n.36 moral objectivism, 14–15, 27, 129 moral obligations: laws and, 141, 142– 4, 147–8; snow clearing (example), 147–8; social norms and, 148–9. See also moral responsibility moral principles, 49, 272n.3; higher order, 157; sociohistorical influences on, 12–13, 49–51 moral reasoning, 153, 153–4, 265n.56 moral responsibility, 234, 237, 285n.3, 288n.19; causal sufficiency and, 246; to intervene in cases of violence, 252–6, 257; shooting a competitor (example), 236–7; weaker forms of, 236. See also moral obligations moral rights: basic, 211, 216–17, 257, 279–80n.13; collective, 31–2, 46, 210, 226, 233 moral value: adhering in collective ends, 148, 151, 156, 194; adhering in individual actions, 286; adhering in joint action, 242; as moral weight, 158 mutual goals and knowledge theory, 81–3 mutual goals and knowledge theory (examples): house construction, 82; military spies, 81–2; tiger hunt, 82– 3 mutual knowledge, 59, 60, 81–3, 88, 121 Narveson, Jan, 212 nation-state, 250
304
Index Native Americans, 223 natural actions, 3, 5, 7–8, 12, 45–6, 51 natural responsiblity, 236, 237–8 non-conventional alternatives, 95–7, 101, 102, 108–9, 115 non-participation, 146, 273n.33, 282n.27; free-riding vs., 146, 149, 150 nonsocial actions, 7–9. See also individual actions; interpersonal interactions; natural actions normalization, 20 normative dimension: of conventions, 122, 126–7, 128–30; of institutions (teleological account), 184–7, 192, 193, 193–4, 197–8 norms, social, 26–7, 123, 124, 189, 271n.1; changes in, 204; collective attitudinal character of, 134, 137; as conventional actions, 139; and desire for approval, 124, 130; distinguished from conventions, 124, 125–30; embedded, 200; governing institutions, 124; internalized, 139– 40; moral principles involved in, 124, 126, 128–9, 139–40, 152; social climber (example), 140; subjective vs. objective force of, 125. See also conformity; institutional norms; laws; social approval theory Nozick, R., 142 nuclear family, 194 oligopolies, 114 Olson, Mancur, 196 omissions, collective moral responsibility for, 250 one-off conventions, 120 organizational actions, 160, 164, 195; layered structures of joint action, 173–4; as quasi-joint actions, 179– 80. See also corporate actions organizational actions (examples): auto factory, 179–80; military battle, 173–4
organizational culture, 163 organizations, 28–30, 161–2, 180; as agents, 164; ; employees of (example), 162–3; hierarchical, 161–3; institutions distinguished from, 163– 4; normative dimensions of, 162–3. See also corporate actions outcomes, 65–6; Great Wall of China (example), 68; mutually interdependent, 67–9. See also collective ends overdetermination, 242, 248–9. See also collective (moral) responsibility overdetermination (examples): multiple stabbing, 242–3; World Cup cricket game, 243–4
participation: in cooperative schemes, 143–4; compulsory, 222; enforcement of, 220–1 participation (examples): compulsory education, 220; war effort, 220; political, 220; right to participate, 220. See also non-participation participatory goods, 219–20, 220–1; joint actions and, 221; moral rights adhering to production of, 220 Pettit, Philip, 32, 33 plural subjects theory, 58, 85–90; nonreversible commitment in, 85–6; objections to, 87, 88–9, 90 plural subjects theory (examples): babysitting, 87–8; dancing, 87; mushroom-picking, 86–7; walking together, 88, 89–900 police services, 226, 228 political inequality, 91 political participation, 220; enforced on individuals, 222 political repression, 22–3 political rights, 230 popular narratives, 204 postmodernism, 16 power model of social actions: Foucault on, 19–21, 189, 261nn. 2, 5;
305
Index power model of social actions (cont.) Giddens on, 18; and the teleological cooperation model, 22, 48 practical reasoning: constraints on, 154; interdependence of moral attitudes and, 154–5; in morally complex situations, 157–8; and social norms, 124, 151–2; truth-telling (example), 154, 155, 157 preferences: action-determining vs. non-action-determining, 101–3, 105–7, 107–10; conditional, 94, 105–6; preference structures, 101, 105, 107, 110, 112; of the minority vs. the majority, 99–100, 102–4, 106. See also coordination problems and conventions procedurally determined actions, 154, 156 procedural reasoning, 154 procedures, 116, 271n.26; conditional property of, 116; governing role–related tasks, 172–3; higher order, 170–1. See also joint procedures procreation, 229 prohibitions, 150 promises, 127–8 propositional attitudes, 36, 165 protection: of cultural structures, 223– 4, 226, 227–8; right to, 250, 254 quasi-joint actions, 55–6; free-riding as, 55; organizational actions as, 179–80 quasi-joint actions (examples): auto factory, 179–80; obeying the law, 55–6 rational agency, 11–12, 47–8, 265n.56; hyena hunt (example), 11–12 rational choice theory, 34, 196 Rawls, John, 142 Raz, Joseph, 217, 227, 271n.1, 283– 4n.36
Re´aume, Denise G., 212, 217–18, 219– 20, 220–1, 227, 282n.26 reductive analysis, 6 regularities of actions, 26, 94, 97, 113, 117, 154; moving furniture (example), 117 regulative rules, 191 relationships, right to form, 221–2 religious institutions, 164, 181 representational thinking, 39–40 repression, political, 22–3 revenge, norms of, 131 roles, 171–3 Ross, Alf, 92–3 Rousseau, Jean-Jacques, 91 Ruben, David-Hillel, 34, 35–6, 37 rules: and collective intentionality, 185, 190–2; constitutive rules, 42–5; defined, 25; deontic properties and, 185; exceptions to (nonconventional alternatives), 95–7, 101, 102, 108–9, 115; rule theory of conventions, 92–3. See also conformity Rwandan genocide, 227, 247–50, 250; compared with Kitty Genovese murder, 252, 253–6 Rwandan Patriotic Front, 249, 251 sanctions, 93 Scanlon, Thomas: on the principle of due care, 128; on promises, 127–8 Schmidtz, David, on higher order ends, 63 Searle, John R., 32, 261–2n.9, 263n.32, 276–7n.11; collective intentionality theory of, 59, 84, 187– 90, 276n.10; on institutional facts, 16–17, 263n.32; on social actions as rule-governed, 39; on social entities as the site of social actions, 71, 192. See also collective intentionality Security Council (UN), 252–3 security institutions, 250. See also protection self-awareness, 49
306
Index self-defence: collective, 226, 227–8; individual, 203 self-determination, 227 self-interest view, 131, 272n.11, 281n.26 Sellars, Peter (role as gardener), 188 Serbian mono-ethnic movement, 213 sexuality, 163, 181; extra-marital sexual activity, 182; sexual acts and relationships, 3, 229 shame, 132–3 shared intentions model, 59, 69, 74– 7, 81, 267n.13; collective ends distinguished from, 80–1 shared intentions model (examples): English Channel tunnel, 75–7; relay team, 79–81; Superman, 77–8; water-pumping, 76–7, 269n.40 Shue, Henry, 250 Silajdzic, Haris, 213 Simmons, A. J., 142, 143 slavery as an institution, 197–8 social actions, 18, 34–9, 125, 181, 206, 259n.3, 262n.15; defined, 1– 3, 5–6, 8, 10, 23–4; individual action distinguished from, 1–3, 5, 7; individually performed actions constitutive of, 6–7, 160; joint actions as necessarily social, 36–7, 38–9; moral dimension of, 46–51; natural actions distinguished from, 3, 5, 7– 8, 12, 45–6, 51; procedurally determined, 154, 156; as rule-governed behavior, 39–42; social entities as the site of, 71, 192; temporal dimensions of, 79–81; as voluntary actions, 14, 19, 20–1. See also joint actions; spheres of social action; teleological account of social action social actions (examples): hyena hunt, 37; man-eating tiger, 35–6, 37–8; voting, 20 social approval theory, 124, 130, 131– 2, 138; objections to, 132–8 social approval theory (examples): fashions/clothing (examples), 133–
4; sexual/homosexual fidelity, 135– 7 social change, 23, 204 social contract theory, 91–2, 95 social forms, 10–11, 15. See also conventions; institutions; organizations; social groups; structural features social groups, 6–7, 207–8, 229; cultural groups distinguished from, 209; defined, 27–8, 161, 200–1, 206; institutions distinguished from, 160–1; intergenerational history a requirement of, 200–1; moral dimension of, 201–2, 203–4; organizations distinguished from, 160–1; shared language a requirement of, 130, 200, 206, 208; whether collective rights attach to, 202–3 socially conditioned actions, 9–10 socially directed actions, 203–4 social norms. See norms, social social taboos, 282n.27 society/culture, 209, 220–1, 261n.7 sociopolitical model. See power model of social actions Sokal hoax, 186–7 South Africa, 141. See also apartheid system ‘‘special’’ rights, 211 spheres of social action, 181, 206, 262n.15. See also by sphere e. g., economic institutions; religious institutions spirituality, 164, 181–2 state obligations: regarding individuals of other states, 250–1. See also humanitarian intervention status, desire for, 139, 140 status-functions, deontic properties of, 184–7, 192 status-functions (examples): incompetent surgeon, 186; log as bench, 184; mediums of exchange, 184, 185, 189–90; Sokal hoax, 186–7
307
Index structural features: collective ends adhering in, 149–50; cultural structures, 223–4, 226, 227–8; interrelated institutional structures, 28, 198–9, 225; layered structures of joint actions, 173–4, 240–50, 249, 256–7, 287n.15; role structures, 171–3; structures of conventions, 97, 119–22, 130, 206; trans-societal structures, 205–6 subjectivist account of social action, 16–17, 263n.22 substantive functionality, 276–7n.11, 277n.13; collective acceptance as, 189–90, 191, 199; collective ends and, 199; of institutions, 189–90 suicide, 140 supraindividual entitites: collective rights adhering in, 212; whether they possess moral properties, 210, 213 Taliban of Afghanistan, 221–2 Taylor, Charles, 16, 225 teleological account of social action, 13–14, 18, 22, 23, 210; joints actions to attain collective ends, 24, 58–9, 60, 61, 152; social cooperation, 21, 196. See also collective end theory (CET) of joint action
teleological model of institutions, 193– 4, 197–8 time: future directed intentions, 47, 76, 155–6; temporal dimensions of social actions, 4, 13, 53, 79–81 totalitarian movements or states, 213, 227 traditional ceremonies, 204 trans-societal interactions, 205–7 truth telling, 154, 155, 157, 187–8, 268n.27 Tuomela, Raimo, 32, 34, 36–7, 59, 262n.16, 263n.31, 266n.4, 268nn.28, 34, 277n.13 Tutsis, 247, 248–9. See also Rwandan genocide Ullman-Margalit, Edna, 32, 33 unilateralism, 213 United Nations (UN), 227, 252–3 values. See moral principles Van Dyke, Vernon, 212 Walzer, Michael, 262n.15 wars and conflicts, 23 Weber, Max, 16 ‘‘we-intentions,’’ 57, 72, 74, 80, 187 Western governments, 198 will, weakness of, 121 Winch, Peter, 39, 264n.40 Wittgenstein, Ludwig, 39
308