Abstract
Central elements of W. V. Quine’s epistemology are widely and deeply misunderstood, including the following. He held from first to last that our evidence consists of the stimulations of our sense organs, and of our observations, and of our sensory experiences; meeting the interpretive challenge this poses is a sine qua non of understanding his epistemology. He counted both “This is blue” and “This looks blue” as observation sentences. He took introspective reports to have a high degree of certainty. He endorsed outright Hume’s “skeptical” argument concerning induction. His naturalized epistemology is simply naturalistic, or scientific, epistemology stripped of the project of rational reconstruction, and is thoroughly normative. Quine was unconditionally a scientific philosopher who took our theories to be answerable ultimately to our perceptual experiences (sensations)—and conditionally an empiricist, empiricism being a scientific theory that has no competitors worthy of the name. I attempt to make all of this clear, and conclude by offering a concise formulation of his epistemology.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
One cannot help wondering how sanguine he himself was about the matter. As late as 1986 he wrote the following about his famous thesis of the indeterminacy of translation, which had been widely discussed for many years:
Unlike so many, Gibson fully understands the difference in status that I ascribe to the indeterminacy of translation and the under-determination of natural science. I count on the clarity of his presentation for a swelling of the ranks of those who get the point.
See (Quine (1986a), p. 155).
Quine (1969). Quine had presented the paper at the Fourteenth International Congress of Philosophy in Vienna in 1968.
(Quine (1966a), p. 212). My emphases.
(Quine (1992), pp. 20–21). My emphases.
(Quine (1995), p. 76). My emphases.
(Quine (1960), pp. 2–3). My emphases.
Though Quine speaks easily here of what he takes to be some obvious facts about sense data and their epistemic limitations, in the opening section of a late chapter called “Ontic Decision,” he devotes several pages to dismissing any claim they might have to inclusion in our ontology. In my view, nothing he says there runs counter to anything I say here, but any discussion of his ontological views would take us far beyond the scope of this essay.
(Quine (1960), p. 22). My emphases.
Quine does not, of course, see the individual’s positing of physical objects as deliberate, or motivated:
[T]he positing of the external objects of common sense is an original trait of human nature. ... It would be senseless to speak of a motive for this archaic and unconscious posit, but we can significantly speak of its function and its survival value; and in these respects the hypothesis of common-sense external objects is quite like that of molecules and electrons.
See (Quine (1966a), p. 210).
(Quine (1986b), p. 533).
(Quine (1986c), p. 335).
(Quine (1981a), pp. 184–185) . Some more recent writers have offered accounts of Quine’s epistemology in which neither experience nor introspection makes any appearance, perhaps because they share the very misunderstandings we have seen Quine respond to. See, for example, Hylton (2007), Fogelin (2004), and (Orenstein (2002), pp. 173–178). One noted Quine commentator, having read a draft of this paper that included these passages, asked, “Would Quine really approve of making use of the notion of introspective access?”
(Quine (1963), pp. 43, 45). My emphasis.
This is implicit in an earlier remark of Quine’s: “A conflict with experience at the periphery occasions readjustments in the interior of the field” (my emphasis). We are not permitted to allow such conflicts to stand.
The scope of this constraint is quite narrow in two respects. First, even if I had to acknowledge a moment ago that there appeared to be brick houses on Elm Street, I am now free in principle to judge that there did not then appear to be such; no statement is immune to revision. Second, if, say, there is a red chiliagon on the wall, it will almost certainly not appear to me either that there is or that there is not. If so, then the constraint simply does not come into play, and I am insofar free to hold either that there is, or that there is not, such a figure there—or, of course, to abstain from judgment.
A word of caution. As Quine is a fallibilist through and through, he would hold that even my belief that there appear to me to be brick houses on Elm Street is fallible. However, this is one of three places I am aware of where he is plausibly read as suggesting that introspective reports enjoy an unusually high degree of certainty. (The others are Quine (1986c), and his response to a proposal of Lars Bergström’s, cited in note 35.)
(Quine (1963), p. 44). My emphasis.
(Quine (1981b), p. 20). My emphases.
Quine (1991).
(Quine (1981a), p. 180). See References for The Roots of Reference. My emphases.
Quine (1986c, pp. 239–240). Last two emphases mine.
(Quine (1992), pp. 20–21). My emphases.
Even after the appearance of this late work, in which he once again underscores one of the central roles of experience in his epistemology, he continues to be misread. Laurence BonJour has written, for example, that “[His account of these matters in ‘Two Dogmas’] is elaborated and developed in many of Quine’s later writings, except that talk of experience is replaced by behavioristically and ‘naturalistically’ more respectable talk of ‘sensory stimulations’.” See (BonJour (1998), p. 90).
A few words of clarification are in order. Clairvoyance theorists do not contend that the world is evidenced to us solely via clairvoyance, rather than via our senses, but only that it is sometimes so evidenced; thus their theory is not what one might call a “head-to-head” rival of empiricism. It and countless other theories—including the theory that Thor sometimes hurls thunderbolts—are “scientific” in the bare-bones sense of offering purported explanations of certain (alleged, though also in some cases actual) events, and Quine takes them to be moribund because they have utterly failed to satisfy the minimal requirement imposed by the scientific enterprise on theories that are to be given serious consideration: that they sometimes successfully predict the occurrence of clearly specified observable events.
Quine’s openness—and guardedness—concerning all sorts of radical theories should be noted:
Any occult phenomenon—any clear case of telepathy, teleportation, or clairvoyance, a ghost, a flying saucer—any of these would delight the scientific mind. ... The mechanisms of the occult phenomena would cry out for investigation.
... Time being in such short supply and the parascience buffs so prodigal with their purported prodigies, the scientist is bound to make short shrift of shoals of parascientific claims. None warrants examination unless it bears promise of unimpeachable documentation.
See “Anomaly” in Quine (1987).
(Quine (1973), p. 39).
For a fuller discussion of this and related matters see my “Observation”, in A Companion to W. V. O. Quine, edited by Gilbert Harman and Ernie LePore, forthcoming from Wiley-Blackwell.
(Quine (1969), p. 88).
(Quine (1981b), p. 25).
(Quine (1992), p. 43).
(Quine (1992), p. 3).
(Quine (1960), p. 41).
(Quine (1960), p. 44). Although Hylton fails to recognize the observational status of subjective observation sentences in his discussion of these matters in Quine, he does present most of what I have said here as an account of objective observation sentences (pp. 135–136). He appears to think, however, that Quine developed such an account only very late: “In two late pieces [‘In Praise of Observation Sentences’ (Journal of Philosophy, 1993) and ‘I, You, and It’ (Orenstein and Kotatko, eds., Knowledge, Language and Logic, 2000a)] ... Quine ... puts forward a criterion of observationality according to which present impingements suffice not for the truth of the sentence, but rather for immediate and unreflective assent to the sentence ....” But Quine never thought that present impingements sufficed for the truth of any objective observation sentence, and all of the central elements of his account of observation sentences appear in §10, “Observation Sentences,” of Word and Object: definition of observationality in terms of observers’ assent, recognition of both degrees of observationality and the fallibility of observational sentences, and a general description of the sorts of conditions under which a subject may wrongly assent to (or dissent from) even a maximally observational sentence.
(Quine (1973), p. 40). My emphasis.
(Quine (1969), p. 87). My emphasis.
(Quine (1981a), p. 180).
(Quine (1975b), pp. 67–81, 73). My emphases.
To avoid familiar worries about how “appears” or “seems” sentences are to be understood, we could take a cue from Quine and formulate them in terms of being in certain neuroperceptual states:
... Instead of “it seems that the stick is bent” let me say something like: “I am in a neural state commonly induced by the sight of a bent stick.” This dodge amounts to reinstating sense data and then construing them neurally .... (Quine 1986c, p. 336)
For economy’s sake, however, I shall for the most part retain the familiar terminology.
It may appear to be a bit odd to speak of “agreement” about the truth of subjective observation sentences, since each individual who responds affirmatively to, say, “Looks blue?” will be reporting on his own subjective (neuroperceptual) state. But this is no cause for concern; Quine’s aim in defining “observation sentence” was to identify those sentences that capture an individual’s sensory evidence about the world, and the fact that his definitions admit subjective, as well as objective, sentences is all to the good, as we shall see. In any case, as noted earlier, talk of agreement could be avoided simply by adapting a different Quinean definition: “S is an observation sentence for a group if it commands the same verdict from almost all linguistically competent witnesses of the occasion.”
Having long been puzzled about the matter, I wrote Quine in 1989 suggesting that he should count many sentences of the form “a looks F” as observation sentences, since they meet his test of observationality at least as well as do their corresponding sentences of the form “a is F.” He replied that he had always done so, and in at least one place he explicitly identifies one as such: “An observation sentence that is perhaps minimally theoretic is ‘This looks blue’.” See (Quine (2000a), p. 413).
Although both objective and subjective observation sentences count as such without qualification, a significant difference between them is worth noting. While we can readily understand how someone’s belief that there is a cup on the table could be mistaken—he might be hallucinating, say—understanding how his belief that there appears to be a cup on the table could be so is quite another matter. For example, the possibility that he is hallucinating that there appears to be a cup on the table—that is, that there only appears to appear to him to be one there—is at best opaque, and likely nonsensical. Even Quine offers no suggestion about how to understand the possibility of such an error, noting only that future developments in neurophysiological theory might occasionally give us grounds for concluding that one had occurred:
If with the progress of neuropsychology we were actually to isolate a neural process that lent itself pretty neatly to identification with the [experience] in question, we might thereafter impute occasional errors of introspection rather than give up the neat identification. (Quine 1986c, p. 335).
Indeed, in Quine (2000a) he strongly suggests that he takes all such assertions by fluent speakers to be true:
An observation sentence should be true, on [Bergström’s] account, on just the occasions that would prompt the subject’s assent to it. An observation sentence that perhaps meets this condition is ‘This looks blue.’ I see only such cases as fulfilling Bergström’s empiricist conception of truth. (p. 414; my emphasis)
I take this to be a slip on Quine’s part, since it is incompatible both with his universal fallibilism and with his suggestion (see note 36) that future developments in neurophysiological theory might provide us with grounds for concluding that some assertion of this sort was mistaken; I note it because it underscores the fact that he took subjective observation sentences to have a high degree of certainty.
As Quine again makes clear in the opening sentences of his last public lecture:
Language is for communicating with our fellows about matters of mutual concern. Its focus is not the private sensations that are our evidence for what is happening around us; [its] focus is rather the things and events out where we all can jointly observe them (from Quine 2000b, pp. 287–291).
(Quine (1992), p. 12).
(Quine (2008), pp. 94–95).
In Fogelin (1997), for example, Fogelin considers Quine’s responses to the possibilities that he might be a brain in a vat, or dreaming, but takes no note of his Humean skepticism.
(Quine (1975b), pp. 67–68).
Historians of philosophy seem to be quite clear about this, but “Cartesian skepticism” still looms large in epistemological writings. For some discussion of this situation, see Johnsen (2009).
Note, incidentally, that on this view the posits that are epistemologically comparable to physical objects include, in addition to Homeric gods, Evil Demons and neurophysiologists stimulating our brains via powerful computers. Quine has no interest in these classical hypotheses. So far as I know, he never mentions them unprompted, and his replies, in a taped conversation, to questions about two of them are casually dismissive: “I dismiss [the dream hypothesis] as very unlikely,” and “it would really be an implausible achievement, at this stage anyway, [for a neurophysiologist] to rig up” a brain and feed it a steady stream of experiences. The conversations are published by R. V. Fara under the title In Conversation with W. V. Quine (London: Philosophy International, 1994).
(Quine (1992), pp. 12–13). My emphasis.
(Quine (2000a), p. 412).
(Quine (1986a), p. 157).
(Quine (2008), p. 134).
(Quine (1994), p. 503) . My emphasis. Here Quine (rightly) bypasses the question of how to determine whether the course of nature has changed, given the failure of some prediction. Would the mining of a blue emerald at some time after \(t\) mean that the course of nature had changed, or would it be further evidence that emeralds are grue (green if observed before \(t\), blue otherwise)? See Goodman (1965).
(Quine (1995), p. 20). My emphases. Strictly speaking, his first sentence should have read, “[N]atural selection has endowed us with standards of perceptual similarity that have meshed pretty well with what have been natural trends ....”
In Bergström (2008), Bergström takes Quine to endorse “skepticism” only in the anodyne sense of accepting fallibilism—the view that there can be no “guarantee” that our theory of the world implies no false predictions. But experiential or observational data are the only possible source of support for a theory of the world, and Quine insists that taken in isolation they provide no support for any theory.
(Putnam (1983), p. 244).
(Quine (1986d), pp. 664–665).
(Quine (1995), pp. 49–50).
For a fuller examination of “Epistemology Naturalized,” see Johnsen (2005).
(Quine (1969), p. 75).
(Quine (1969), p. 78). My emphases.
Quine, “The Scope and Language of Science,” in (Quine (1966b), p. 216). My emphases.
(Quine (1975b), pp. 74–75).
For a fuller discussion of these matters in the context of Quine’s “Epistemology Naturalized” see Johnsen (2005).
(Quine and Ullian (1970), pp. 44–51).
With the exception, of course, of such deviants as some insane individuals.
(Quine (1969), p. 75). Emphases mine, except for the two tokens of “is”.
(Quine (1992), p. 19).
(Quine (1969), pp. 75, 76 and 78). My emphases.
(Quine (1981c), p. 474). My emphases. Strictly speaking, Quine should have recognized four sorts of predictions that are in principle derivable from our theories, and are of some interest to epistemologists: predictions of observable events, of observations of such events, of sensory stimulations and of experiences (and/or the neuroperceptual events with which they are identical). Hylton, in Hylton (2007), demurs. He suggests not only that when Quine describes himself as “speaking, oddly perhaps, of the prediction of stimulation,” he is speaking oddly “to the point of being misleading,” but that “The point is perhaps better put in terms of predictions of observation sentences” (p. 14). But observation sentences are of course not events, hence not predictable. Perhaps he had in mind utterances of such sentences, which are in principle predictable, but it is hard to see how any point Quine might have been making in speaking of predictions of stimulations could have been put better by speaking in terms of predictions of utterances of observation sentences.
(Quine (1969), pp. 82–83). My emphasis.
Quine (1993). “We are not aware of our neural intake” (p. 111).
Here the fact that we sometimes disagree about what is evidence for a given theory reemerges, and my remark should be understood as restricted to theories on which we all agree.
This tension in Quine’s writings may be responsible for one found in Hylton (2007). Having argued early (p. 14) that Quine does not think our theories yield predictions of subjects’ stimulations, Hylton later attributes to him the view that “[Stimulations of our nerve endings] provide the only empirical constraint on our system of knowledge ...” (p. 91).
(Quine (1960), pp. 32–33, 42).
(Quine (1986d), pp. 564–568).
In fairness to Quine, it should be said that a sufficiently careful reader of Word and Object would not have fallen into such a misunderstanding. On page 39 he sums up some of his remarks as follows: “We have now seen that stimulus meaning as defined falls short in various ways of one’s intuitive demands on ‘meaning’ as undefined .... Yet stimulus meaning ... may be properly looked upon still as the objective reality that the linguist has to probe when he undertakes radical translation. For the stimulus meaning of an occasion sentence is by definition the native’s totalbattery of present dispositions to be prompted to assent to or to dissent from the sentence ....”.
(Quine (1992), pp. 2–3).
In Hylton (2007), Hylton attributes the view that our evidence consists of our stimulations to Quine (pp. 88–89), and he counts observation sentences as theoretical—as “integrated into wider theory” (p. 140)—but fails to see, hence does not consider, the consequent problem that Quine here addresses: the need to find some way to formulate our evidence in sentences in order to reveal the logical relations between evidence and theory. Hylton is of course right to hold that objective observation sentences are in a sense theoretical, namely, the sense in which, for Quine, all talk of bodies is theoretical. But in that sense of “theoretical,” sentences recording the evidence for scientific theories are themselves theoretical, whereas the crucial epistemological distinction that Quine is here trying to elucidate is the one we make when working within the scientific enterprise: the distinction between what we there count as evidence (hence not as theoretical) and what as theory.
(Quine (1992), p. 2).
See note 35.
(Quine (1981c), p. 474). My emphases.
(Quine (1992), pp. 72–73). My emphasis.
References
Bergström, L. (2008). A defense of Quinean naturalism. In C. B. Wrenn (Ed.), Naturalism, reference and ontology: Essays in honor of Roger F. Gibson. New York: Peter Lang Publishing.
BonJour, L. (1998). In defense of pure reason. Cambridge: Cambridge University Press.
Fogelin, R. J. (1997). Quine’s limited naturalism. Journal of Philosophy, 94, 543–563.
Fogelin, R. J. (2004). Aspects of Quine’s naturalized epistemology. In R. Gibson (Ed.), The Cambridge companion to Quine (pp. 19–46). Cambridge: Cambridge University Press.
Goodman, N. (1965). Fact, fiction, and forecast (2nd ed.). Indianapolis: Bobbs-Merrill.
Howson, C. (2000). Hume’s problem. Oxford: Oxford University Press.
Hylton, P. (2007). Quine. New York: Routledge.
Johnsen, B. C. (2005). How to read epistemology naturalized. Journal of Philosophy, 102(2), 78–93.
Johnsen, B. C. (2009). The argument for radical skepticism concerning the external world. Journal of Philosophy, 106(12), 679–693.
Orenstein, A. (2002). W.V. Quine. Princeton: Princeton University Press.
Putnam, H. (1983). Why reason can’t be naturalized. In H. Putnam (Ed.), Realism and reason, vol. 3 of Philosophical papers (pp. 229–247). Cambridge: Cambridge University Press.
Quine, W. V. (1960). Word and object. Cambridge: MIT Press.
Quine, W. V. (1963). Two dogmas of empiricism. In W. V. Quine (Ed.), From a logical point of view (pp. 20–46). New York: Harper & Row.
Quine, W. V. (1966a). On mental entities. In W. V. Quine (Ed.), The ways of paradox (pp. 208–214). New York: Random House.
Quine, W. V. (1966b). Posits and reality. In W. V. Quine (Ed.), The ways of paradox (pp. 233–241). New York: Random House.
Quine, W. V. (1969). Epistemology naturalized. In W. V. Quine (Ed.), Ontological relativity and other essays (pp. 69–90). New York: Columbia University Press.
Quine, W. V. (1973). The roots of reference. LaSalle, IL: Open Court.
Quine, W. V. (1975a). On empirically equivalent systems of the world. Erkenntnis, 9, 313–328.
Quine, W. V. (1975b). The nature of natural knowledge. In S. Guttenplan (Ed.), Mind and language (Vol. 73, pp. 67–81). Oxford: Clarendon Press.
Quine, W. V. (1981a). Responses. In W. V. Quine (Ed.), Theories and things (pp. 173–186). Cambridge: Belknap Press.
Quine, W. V. (1981b). Theories and things. Cambridge: Harvard University Press.
Quine, W. V. (1981c). Reply to Stroud. Midwest Studies in Philosophy, 6, 473–474.
Quine, W. V. (1986a). Reply to Roger F. Gibson, Jr. In L. Hahn & P. Schilpp (Eds.), The philosophy of W.V. Quine (pp. 155–157). Carbondale: Southern Illinois University Press.
Quine, W. V. (1986b). Reply to P. F. Strawson. In L. Hahn & P. Schilpp (Eds.), The philosophy of W.V. Quine (pp. 533–535). Carbondale: Southern Illinois University Press.
Quine, W. V. (1986c). Reply to Arnold Levison. In L. Hahn & P. Schilpp (Eds.), The philosophy of W.V. Quine (pp. 335–337). Carbondale: Southern Illinois University Press.
Quine, W. V. (1986d). Reply to Morton White. In L. Hahn & P. Schilpp (Eds.), The philosophy of W.V. Quine (pp. 664–665). Carbondale: Southern Illinois University Press.
Quine, W. V. (1987). Quiddities. Cambridge: Harvard University Press.
Quine, W. V. (1991). Two dogmas in retrospect. Canadian Journal of Philosophy, 21, 265–274.
Quine, W. V. (1992). Pursuit of truth. Cambridge: Harvard University Press.
Quine, W. V. (1993). In praise of observation sentences. Journal of Philosophy, 90, 107–116.
Quine, W. V. (1994). Response to Hookway. Inquiry, 37, 502–504.
Quine, W. V. (1995). From stimulus to science. Cambridge: Harvard University Press.
Quine, W. V. (2000a). Response to Bergström. In A. Orenstein & P. Kotatko (Eds.), Knowledge, language and logic (pp. 412–414). Dordrecht: Kluwer.
Quine, W. V. (2000b). Three networks: Similarity, implication, and membership. In A. Kanamori (Ed.), Proceedings of the twentieth world congress of philosophy, vol. 6: Analytic philosophy and logic (pp. 287–291). Charlottesville: Philosophy Documentation Center. (Reprinted in Quine, W. V. (2008). In D. Føllesdal & D. B. Quine (Eds.), Confessions of a confirmed extensionalist and other essays (pp. 493–497). Cambridge: Harvard University Press.)
Quine, W. V. (2008). In D. Føllesdal & D. B. Quine (Eds.), Confessions of a confirmed extensionalist and other essays. Cambridge: Harvard University Press.
Quine, W. V., & Ullian, J. S. (1970). The web of belief. New York: Random House.
Acknowledgments
I am deeply indebted to Roger Wertheimer for his obdurate balking at sundry opacities in early drafts of this essay, and grateful to him and to Dagfinn Føllesdal, Gilbert Harman, Lars Bergström, John Perry, Joshua Brown, Edmund Gettier, Carl Feierabend, Joshua Weisberg and two of this journal’s readers for encouragement and insightful suggestions concerning later drafts.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Johnsen, B.C. Reclaiming Quine’s epistemology. Synthese 191, 961–988 (2014). https://doi.org/10.1007/s11229-013-0302-4
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11229-013-0302-4