Abstract
It is a consequence of Quine’s confirmation holism that the logical laws are in principle revisable. Some have worried this is at odds with another dictum in Quine, viz., that any translation which construes speakers as systematically illogical is ipso facto inadequate. In this paper, I try to formulate exactly what the problem is here, and offer a solution to it by (1) disambiguating the term ‘logic,’ and (2) appealing to a Quinean understanding of ‘necessity.’ The result is that the different theses in Quine’s philosophy of logic are to be situated within different contexts of inquiry.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Berger essentially uses this conflict as a way of criticizing Quine’s “dispositional account” of linguistic meaning. However, Berger does not consider the sort of interpretation I hold on the matter, and for this reason his criticism of Quine strikes me as hasty.
Levin, on the other hand, sympathetically tries to solve the conflict in Quine, in ways different from my own solution. Levin’s idea is that, for Quine, the laws of classical logic strictly speaking are unrevisable, although the logic we use in practice may undergo variation over time (see p. 57). I am somewhat amenable to this idea if ‘classical logic’ is understood as Symbolic Deductive Logic (see Sect. 2 above). Yet the distinction between Symbolic Logic and the logic we use does not, in itself, resolve the conflict in Quine. For the conflict can be located merely in what Quine says about the logic we use in practice: In practice, our current logic might deviate from what it is, yet apparently the current logic is a logic that must be respected in the translation of a language.
An anonymous referee has reminded me that Quine eventually rejected holism in favor of “moderate holism.” The moderate holist grants that individual statements are not dis/confirmed by experience, yet she does not think experiences bear on one’s theory as a whole. Instead, she says (roughly) that experiences dis/confirm “chunks” (proper parts) of a theory (see, e.g., Quine 1991). Thus, the present discussion should be understood as just addressing the earlier holist Quine.
It is too weak just to say that experience can, as a matter of psychology, prompt someone to revise logic. The interesting claim is that logical laws can be revised as a rational response to recalcitrant experiences. An anonymous referee has asked, however, if a rational revision of logic is even possible, since respecting the laws of logic would seem critical to the rationality of a revision. The question is a good one, but since the matter is rather involved I shall not pursue an answer here.
Presumably Quine’s constraints here are also subject to revision, as part of the theory of theory-choice.
Joe Ullian has objected that talk of logic “governing” a language is not a very Quinean way of talking. He suggests instead that we talk of our linguistic behavior as conditioned by our inferential practices and logical intuitions. This is in fact what I had in mind when I spoke (shorthandedly) of logic “governing” a language. I believe one could simply replace my talk of logic “governing” a language with Ullian’s way of talking, without affecting the main thrust of this paper.
Here ‘adequate translation’ is meant to indicate a translation which is, minimally, (1) empirically adequate, and (2) in line with various Principles of Charity, including the Principle of Logical Charity. On a sympathetic reading, Quine did not intend this notion of “adequate translation” for translations of formal languages (after all, we can stipulate a formal language with a deviant logic). But I assume formal languages are not at issue here.
The contemporary analog of this idea is found in “theory-theory” accounts of language acquisition, including the Language of Thought hypothesis (see Fodor 1975). Such accounts are debatable of course, but it is not my intent to debate them here. Suffice it to say that Quine espoused such an account. Yet notably, Quine also thinks of language acquisition in behavioristic terms—language acquisition is a process of conditioning verbal responses in relation to certain stimuli (see, e.g., Chap. 1 of his 1973). Nonetheless, Quine seems to think that radical translation provides a model of what occurs when the child is thus conditioned (see especially his 1968). Of course it may seem strange for a behaviorist to adopt such a model. For on this model, radical translation would presumably occur “in the mind” of the child—and such “mental processing” seems incompatible with behaviorism. But as Quine (1968) stresses, his behaviorism is in no way committed to denying “internal mechanisms.” Indeed, one would naturally look to such mechanisms in explaining how conditioned responses take root in the child. Yet there are further questions on the issue, but I shall not address them here. For present purposes, we only need observe that Quine does deploy the radical translation model. Whether his behaviorism allows such a model is another matter.
For Quine, ‘obvious’ does not indicate some special epistemic property of logical laws; rather, the term ‘obvious’ just indicates that a statement is one which, by and large, speakers assent to unhesitantly (1970, p. 82). It would seem, however, that speakers’ hesitation to assent may come in degrees; if so, obviousness and ultimately acceptability of a translation would also come in degrees. (Thanks to Jim Baillie for drawing my attention to this point.)
However, Quine (1990) may hint at a distinction between Basic Logic* and Symbolic Deductive Logic.
Cf. Quine (1953), pp. 29–30.
Here, biconditionals, conditionals, and other apparently modal statements are to be read in the material mode rather than formally. Thus, “If p then q” is read as “Not (p and not-q)” rather than “Necessarily, not (p and not-q).
See, e.g., Quine (1963b), pp. 74–76. In this region of the text, it is easy to misconstrue Quine as genuinely committed to particular modal truths. One can also misread Quine (1960), §41, and Quine (1953) in this way. However, in each case I take it that Quine is arguing by reductio on the modal logician’s assumptions. In (1963b), Quine reasons from these assumptions to the claim that the logical/mathematical notion of necessity “has no relevance” (p. 75). In (1960) §41, the reductio is completed by the so-called “slingshot” argument (the label comes from Barwise and Perry 1981). That argument is also key in his (1953), though in that paper, it is only the first wave of attack.
Thanks to an anonymous referee for raising this question. On either answer I propose, the referee’s concern is that Quine’s view will entail that (at least in some inquiries) some contingent obvious truths will be counted as necessary, e.g., ‘some dogs are black.’ This is indeed a weighty problem, though it seems rooted in the problems with Quine’s notion of necessity, see Sect. 3 above.
References
Bar-On, D. (1992). Semantic verificationism, linguistic behaviorism, and translation. Philosophical Studies, 66, 235–259.
Barwise, K. J., & Perry, J. (1981). Semantic innocence and uncompromising situations. In P. A. French, T. Uehling, & H. K. Wettstein (Eds.), Midwest studies in philosophy, vol. 6: Foundations of analytic philosophy (pp. 387–403). Minneapolis: University of Minnesota Press.
Berger, A. (1990). A central problem for a speech-dispositional account of logic and language. In R. Barrett & R. Gibson Jr. (Eds.), Perspectives on Quine (pp. 17–35). Cambridge, MA: Blackwell.
Fodor, J. A. (1975). The language of thought. Cambridge, MA: Harvard UP.
Levin, M. (1979). Quine’s view(s) of logical truth. In R. W. Shahan & C. Swoyer (Eds.), Essays on the philosophy of W.V. Quine (pp. 45–68). Normal, OK: University of Oklahoma Press.
Quine, W. V. (1936). Truth by convention. (Reprinted in The ways of paradox and other essays, 2nd edition (pp. 77–106). Cambridge, MA: Harvard UP.).
Quine, W. V. (1953). Two dogmas of empiricism. (Reprinted in From a logical point of view, 2nd edition (pp. 20–46). Cambridge, MA: Harvard UP.).
Quine, W. V. (1960). Word and object. Cambridge, MA: MIT Press.
Quine, W. V. (1963a). Carnap and logical truth. (Reprinted in The ways of paradox and other essays, 2nd edition (pp. 107–132). Cambridge, MA: Harvard UP.).
Quine, W. V. (1963b). Necessary truth. (Reprinted in The ways of paradox and other essays, 2nd edition (pp. 68–76). Cambridge, MA: Harvard UP.).
Quine, W. V. (1968). Linguistics and philosophy. (Reprinted in The ways of paradox and other essays, 2nd edition (pp. 56–58). Cambridge, MA: Harvard UP.).
Quine, W. V. (1970). Philosophy of logic. Englewood Cliffs, NJ: Prentice-Hall.
Quine, W. V. (1973). The roots of reference. La Salle, IL: Open Court Publishing.
Quine, W. V. (1977). Intentions revisited. (Reprinted in Theories and things (pp. 113–123). Cambridge, MA: Harvard UP.).
Quine, W. V. (1990). Comment on Berger. In R. Barrett & R. Gibson Jr. (Eds.), Perspectives on Quine (pp. 36–37). Cambridge, MA: Blackwell.
Quine, W. V. (1991). Two dogmas in retrospect. Canadian Journal of Philosophy, 21, 265–274.
Acknowledgments
I wish to thank Dorit Bar-On, Ephraim Glick, William Lycan, Ram Neta, Michael Resnik, Joseph Ullian, and two anonymous reviewers for their comments on earlier drafts. I also owe thanks to audiences at the 2003 Northwest Conference in Philosophy, and at the 2004 Central Division Meeting of the American Philosophical Association.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Parent, T. Quine and Logical Truth. Erkenn 68, 103–112 (2008). https://doi.org/10.1007/s10670-007-9080-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10670-007-9080-z