Progress on Two Fronts W. V. Quine The Journal of Philosophy, Vol. 93, No. 4. (Apr., 1996), pp. 159-163. Stable URL: http://links.jstor.org/sici?sici=0022-362X%28199604%2993%3A4%3C159%3APOTF%3E2.0.CO%3B2-3 The Journal of Philosophy is currently published by Journal of Philosophy, Inc..
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/jphil.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.
JSTOR is an independent not-for-profit organization dedicated to and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact
[email protected].
http://www.jstor.org Sun Jun 17 05:08:00 2007
T H E JOURNAL OF PHILOSOPHY VOLUME XCIII, NO.
4, APRIL 1996
PROGRESS ON TWO FRONTS*
I
n these pages I update my views on two epistemological matters.
I. PREESTABLISHED HARMONY
Correct translation, we used to say, preserves meaning. In Ward and Object,' chary of mentalism and meaning, I represented correct translation of an observation sentence as preserving what I called its stimulus meaning. I explained its stimulus meaning as the range of stimulations, any one of which would prompt the subject to assent to the sentence. Each stimulation, in my sense of the term, was the near-simultaneous firing of some subset of the subject's neuroceptors. Not that I pictured the translator a s neurologizing. He would just guess his translation of the native's observation sentence in the obvious way, picturing himself in the native's stance and considering what English sentence it. would prompt of him. Eventually, he would build his translations of theoretic sentences upon those of observation sentences by what I called analytical hypotheses; still no talk of nerve endings. Talk of stimulus meaning and of analytical hypotheses was rather my business, my theory of the translator's activity. Stimulus meaning was what, theoretically speaking, correct translation of an observation sentence preserved. This is uncomfortable theory, however. It calls for sameness of stimulus meaning of the native sentence for the native and the English sentence for the translator, and hence a sharing of stimulations by native and translator. Well, they cannot share neuroceptors, so
* Thanks to discussions with Burton Dreben, this paper has much improved through several drafts. ' Cambridge: MIT, 1960. 0022-362X/96/9304/159-63
O 1996 The Jour~lalof Philosophy, Inc.
159
160
THE JOURhrAL OF PHILOSOPHY
we must settle rather for homology of receptors. Such homology is by no means to be expected, and anyway surely should not matter, as I remarked in a lecture five years later.2 I expressed the same discomfort again in 1974,' citing Charles Darwin%n the diversity of nerve nets from individual to individual. But I saw no steps to take. The matter surfaced again at a conference with Donald Davidson, Burton Dreben, and Dagfinn Follesdal at Stanford in 1986. Davidson favored treating translation purely in terms of the external objects of reference, bypassing consideration of neural receptors. Such is the familiar practice also of experimental psychologists, for whom the so-called stimulus is out there, shared by subject and experimenter. I recognized the adequacy of this object-oriented line in describing the procedures of translation, having described them thus myself, and its adequacy likewise for lexicography. There remains a problem for epistemology and neuropsychology, however, when we reflect with Darwin on the intersubjective diversity of nerve nets and receptors. How does the mere sameness of the distal cause, the jointly observed object, prevail over the diversity of the proximal segments of the causal chains, inside the two observers, and still issue in agreeing response? In short, why can the translator and the lexicographer blithely rest with the distal stimulus, as indeed they can? Let me pinpoint the problem. A rabbit appears, the native says 'Gavagai', and the translator conjectures 'Rabbit'. On a later occasion they espy another rabbit, the translator says 'Gavagai', and the native concurs. The two occasions were perceptually similar5for the native, by his subjective standards of perceptual similarity, and likewise for the translator by his independently testable subjective standards of perceptual similarity. Anatomic likeness of the native's receptors and those of the translator could have helped to account for this agreement, but that is out. What then does? If I had thus pinpointed my uneasiness back then, I would have quickly resolved it. What we have is a preestablished harmony of standards of perceptual similarity, independent of intersubjective likeness of receptors or sensations. Shades of G. W. Leibniz, thus, but without appeal to divine intervention. The harmony is explained by a yet deeper, but more faltering preestablished harmony between "Propositio~~al Objects," 1965, published in Ontological Relativity and Other Essays (New York: Columbia, 1969); see page 157. Roots of Reference (La Salle, IL: Open Court, 1974), pp. 23-24, 41. Origin ofSpecies (London: 1959), pp. 45-46. ' Roots ofReference, pp. 16-18; From Stimulus to Science (Cambridge: Harvard, 1995), pp. 17-18.
PKOGRESS ON TWO FRONTS
161
perceptual similarity and the environment. This, in turn, is accounted for by natural selection, as follows. We have, to begin with, an inductiue instinct: we tend to expect perceptually similar stimulations to have sequels that are similar to each other. This is the basis of expectation, habit formation, and learning. Successful expectation has always had survival value, notably in the elusion of predators and the capture of prey. Natural selection has accordingly favored innate standards of perceptual similarity which have tended to harmonize with trends in the environment. Hence the success, so much better than random, of our inductions and expectations. Derivatively, then, through our sharing of an ancestral gene pool, our innate standards of perceptual similarity harmonize also intersubjectively. Natural selection is Darwin's solvent of metaphysics. It dissolved Aristotle's final cause, teleology, into efficient cause, and now Leibniz's preestablished harmony as well. Harmony without interaction: that was the subtlety. We take its ubiquitous effects for granted, not thinking them through. It was only in From Stimulus to Science ( p p 20-21) that I got the matter into proper focus and explained it as here. Note that the harmony was needed not only in respect of the rabbits, in my example, but also in respect of the utterance of 'Gavagai'. The sound of it was perceptually similar for the native and likewise for the translator when it was spoken by either of them. This applies equally, of course, to the child's learning of words from the mother. The effect of the intersubjective harmony, we see, is that what the two observers agree on is the shared distal subject matter and not the unshared proximal stimulations. The latter, however, are what are related by perceptual similarity, though not intersubjectively. Might it be more straightforward to readjust the terminology and treat the external subject matter as the relata of perceptual similarity? One reason not to is that it would put the whole vital causal structure of preestablished harmony back where it had been, out of sight and out of mind. Another reason is individuation. What is distally perceived may be an ill-defined event rather than a body, or, if a body, a body in an ill-defined range of perspectives, accommodating our two observers on the occasion in question and perhaps excluding others. I cut through all this by adhering to the subject's neural intakes as the relata of that subject's relation of perceptual similarity, which then enjoys harmony, generous but probably not complete, with that of other subjects.
162
THEJOURNAL OF PHILOSOPHY
Ironically, I had described this harmony of perceptual similarity standards already in Roots of Reference, on the very page (23) where I then proceeded to bemoan my dependence on intersubjective matching of neural intakes. I even accounted for the harmony by natural selection, shared environment, and heredity, just as in the present pages and in From Stimulus to Science. My account differed only in one crucial detail. Today, I have the two subjects exposed to the same external event, whereas in Roots of Refeence I was still grudgingly according them homologous neural intakes. It made a difference of twenty years. I left myself and others hanging with this enigmatic envoi: "I shall not pause over the lesson, but there is surely one there" (op. cit., p. 23). 11. THEORY-LADEN OBSERV.4TION
In his early years, the only way a child can go wrong on an observation sentence is by affirming it in a situation that would not have commanded the adult's outright assent. Second thoughts are not yet relevant; they become so only at a later stage, when scientific theory has begun to interrelate observation sentences and generate conflict. The adult may affirm 'There's a rabbit' outright, and then rescind it when a later observation suggests that it was the neighbor's cat. In Word and Object (pp. 42-44) I consequently recognized degrees of observationality, and treated the intrusion of theory or collateral information as dilution of observationality. But in later writings, I held observationality as absolute, based on immediacy of assent, and then I accommodated the intrusion of theory by contrasting the holophrastic conditioning of the observation sentence to neural intake with the analytic relations of the component words to the rest of language. The sentence figures holophrastically both in the infant's first acquisition of it and in the scientist's immediate assent to it when testing a theory. My attention was turned anew to these matters by a recent letter from Lars Bergstrom adumbrating an empirical theory of truth. I ended up not adopting it, but it has prompted me to reflect further on the bipolarity of the holophrastic and analytic perspectives, as against the gradualism of observationality in Word and Object, or, as we might say, a gradualism of theoreticity. In conclusion, I retained the absolute notion of an observation sentence as simply an occasion sentence that commands the subject's immediate assent, however fallible and revisable. Fallibility is then accommodated in a separate dimension, theoreticity, which invests observation sentences in varying degrees.
PROGRESS ON TWO FRONTS
163
As before, all observation sentences are holophrastic in their association to stimulation. Thus it is that assent to them is immediate, unless the stimulation is off in the fringe of appropriate stimuli. Also, as before, it is in the analytic or word-by-word relations of the sentence to others that theory comes in, in varying degrees. Theoreticity is at a minimum in the observation sentence 'That's blue', and appreciable in 'That's a rabbit'. It may even be said to invest innate perceptual similarity, for theory is implicit in our innate readiness to integrate varied perspectives into an objective reality. A square on the floor, seen from various angles, projects on the retina now a square and now one and another trapezoid, but all these neural intakes are perceptually similar; we are said to "see the square." I am thus won over to Thomas Kuhn and others who have insisted that observation is inseparable from theory. But my observation sentences stay on in their old definition and their old role as clauses in the observation categoricals that are the checkpoints of science. This fits nicely with scientific practice. The experimenter settles for criteria that he can confidently adjudicate on the spot: confidently, not infallibly. He looks carefully, and if he is subsequently in doubt he repeats the experiment if he can. The finality of an experiment is historical, not logical. The evidential burden borne by a highly theoretic observation sentence such as 'There was copper in it', when it is invoked at a checkpoint, can certainly be shifted to less theoretic observation sentences by deeper analysis. The ideal of reducing all checkpoints to minimally theoretic observation sentences like 'That's blue', however, is the old phenomenalistic reductionism of which I have long since despaired. Observation categoricals, formed of observation sentences two by two, remain the checkpoints of scientific theory as before, and the experimental refutation of an implied categorical refutes the theory that implied it. But the refutation is not final, for the refutation of the categorical was not final; the experimenter may have misjudged the protasis of the categorical as fulfilled, or misjudged the apodasis as failed, such being the fallibility of observation. We just expect him to do the best he can. W. V. QUINE
Harvard University