QUINE VERSUS THE APEMAN

ED BRANDON

Access, vol. 3, no. 1, pp. 19-23 (1984)

However poignant the destruction of ancient ways of life, philosophers of education should applaud the recent assault on a moribund tradition by Evers and Walker (1982) under the banners of W.V.0. Quine. Their target is analytical philosophy of education (APE), especially as exemplified by Professor P.H. Hirst. Their ammunition is taken from the gleaming stores of formal logic and Quine's personal reserve of "Dali-esque" pragmatism (Gellner's characterization, 1981, p.64). The day after such an attack one is certainly not likely to find much of the cautious and confused London Institute of Education (I forego another acronym) verbiage in what once passed for good health.

While in general applauding the introduction of modern techniques into the philosophy of education, I must admit to sharing with Gellner a feeling that Quine's vision is somewhat too idiosyncratic to serve as a stable basis for a defensible philosophy of education or to offer a short-cut to Evers and Walker's rendition of Marx. There are other routes to good sense or the community of ends.

I shall take just one of Evers and Walker's arguments, one which they do, however, regard as of considerable importance in showing that knowledge is a "seamless web" rather than a "partitioned set" of Hirstian forms. It has a wider significance since it purports to show that certain familiar ways of classifying knowledge are indefensible, indeed unintelligible. One might well have agreed with Evers and Walker that Hirst has signally failed to justify, explain, explicate, or elucidate his own proposals, while thinking that it still made sense to talk, now of items of mathematical knowledge, now of items of historical knowledge, now of items of moral obfuscation. But no: "we should give up the task of trying to demarcate a special set of statements called ‘mathematical truths’" (1982, p.38). And a little earlier we are told that a familiar way of talking, albeit described arcanely as "quantification into opaque contexts", is "unintelligible".

For myself, I cannot find much sympathy for many philosophers' predilection for finding what other people say unintelligible or meaningless. No doubt if it were so, one would not stand any chance of having to admit that what they said was true; but 50 years after Popper's first book it is high time philosophers were prepared to take a few risks. And the considerations that have been offered for thinking that theologians, say, have been talking nonsense are very good reasons for thinking they have been speaking falsely and can be disregarded by anyone with a concern for truth. Despite the contortions the positivists got into with supposed meaninglessness, it is disheartening to see Dummett’s disciples, and now Quine’s, making the same utterly implausible accusations, even if in somewhat different directions.

But let us turn to their argument. We need a sentence forming operator on sentences, "It is an item of mathematical knowledge that… " which I shall write "Math fact that… ". We are then asked to consider the argument:

It is evident, we are told, that (3) does not follow from (1) and (2); in Quinean terms, "Math fact that… " creates an opaque context into which one cannot in general substitute co-referring terms salva veritate. Noting that (1) and (3) are instantiations of the existential generalization (4) [I am using (Ex) for the existential quantifier]:

they conclude that quantification into opaque contexts is unintelligible, at least with the standard "objectual" understanding of quantification.

Before replying directly to this argument, let us note a consideration ad homines. As Evers and Walker are no doubt aware, their argument would not only rule out Hirstian form of knowledge operators, but also our ordinary pedagogical talk of what people know or believe or fear: what Quine and others call the propositional attitudes. We are presumably to cease trying to demarcate what Jim has learnt from what he doesn't yet understand. Their argument would also, I suspect, rule out their own talk of items of ideological distortion, but I have yet to find relevant premises that Evers and Walker would have to accept. (Consider the possibly shifting reference of "religion" in this imitation of our original argument:

In the case of the propositional attitudes, Quine of course will provide alternative things for us to do that are recognizably in the same ball-park, and perhaps belief and the rest are otiose notions (cf. Kaplan, 1983); but should we concede so readily that they are unintelligible? Quine has recently said that "I see the verb 'believe' even in its de dicto use as varying in meaningfulness from sentence to sentence" (1981, p. 122), but this is clearly a consequence of holding his philosophical views rather than an "innocent" judgment.

Intelligibility in these sorts of case is notoriously hard to settle (or, as I have suggested earlier, the issue is really about something else), but we need very strong reasons for thinking that something people normally say is really unintelligible. Evers and Walker have offered us the failure of an apparently simple inference, so let us look more carefully.

Lest numbers and their genuine mysteries bemuse us, let us take another comparable argument, this time with an operator for historical facts, "Hist fact that… ". The argument is:

Aqain I think our inclination is to say that the conclusion fails to follow from the premises. Evers and Walker tell us that we therefore cannot really understand the operator; but while they spend a page on "substitutional" quantification, they fail to mention another explanation of these failures of inference that has been current for many years (it was suggested in Smullyan, 1948) and which will certainly explain at least their examples. The point is that definite descriptions can be seen as logically complex, as they are in Russell's theory. Consequently the logical analysis of (3), (7), or (10) involves two operators, the existential quantifier belonging to the definite description and the special operators we have introduced. Once you have two operators, you have the problem of deciding their order, or more to the point, their relative scopes.

Recognizing that there are two operators to contend with, we can move on to see that our conclusions, (3), (7), and (10), are possibly ambiguous, and that as they stand they are not fully analysed. One interpretation does in fact follow from the premises given, the other does not. Let us take the two readings of (10):

I claim that (11) follows from (8) and (9), while (12) doesn't. Reverting to Evers and Walker's argument, we can see that what follows from its premises (leaving out most of the Russellian complexity) is (13):

Whereas their own conclusion, (3), would be most plausibly construed as (14):

If we may rely on Quine's own remarks about parallel cases in talk of belief or necessity, one condition for validly inferring conclusions with the logical structure of (12) or (14) would be (15), where I use "D" to stand for any of the special Hirstian operators:

"The Fer" being the designation introduced in the second premise of our various examples. What this amounts to is that if it were a mathematical fact that the number of coins in my pocket at time t = 2, or a historical fact that I live in Jamaica, then the original conclusions would have gone through. This seems as it should be.

It might be noted in passing that we can see here the awkwardness of Hirst's apparent claim that every proposition belongs to one of his forms. The original argument involved one mathematical and one empirical premise; Hirst seems obliged to allocate the conclusion, (3), or rather the proposition operated on in (3) by our special operator, to one of his forms. But the rest of us are not under any such obligation. We can say it is a bastard. In querying Evers and Walker's attack on Hirst, I am of course not to be interpreted as defending Hirst's distinctive views.

Smullyan's insistence that we take note of the scope of operators has been taken further by Prior who used attention to scope as a means of denying the existence of "opaque" contexts altogether (1963). Prior's approach has recently been extended and again applied to Quinean arguments by Williams (1981, esp. ch. IX). The work of Prior and Williams will also show why we should not be overimpressed with the Evers/Walker Quinean dichotomy of objectual versus substitutional modes of understanding quantification.

But while the formal details can be handled, I think, in Prior's manner, it might also be useful to have a more informal approach to the issue raised by the Evers and Walker argument. The point then is that we can put more or less weight upon the semantic components of the designators we use. Consider the following series of virtually synonymous claims:

I do not think it is easy to say when it becomes inappropriate to prefix "Hist fact that…" to these sentences. If the change comes between (18) and (19), it is presumably because my reference to where I now happen to be obtrudes in a way in which an appeal to where we all happen to be does not. But the point is that we can use designators with or without attending to their own specific meanings; I could use "the island where I now live" to convey no more than "here" does in (16); but I could also, and perhaps more naturally, use it in the way analyzed in (12). Distinguishing scopes gives a clear division; the reality of language use is more blurred. It is not, I think, unintelligible, either in that word's ordinary meaning, or even in its philosophical transmogrification.

I do not expect that these brief remarks will do much to alter Evers and Walker’s view. As is obvious from their reference to a variety of logical issues, the argument they offered and the reply I have made raise a whole host of interconnected problems which are certainly not addressed here, nor unfortunately anywhere else. As Manser notes (1982), there are surprisingly few confrontations in philosophy and obviously I have not given Evers and Walker enough reason to switch their allegiance in philosophical logic. The point, however, is more for the uncommitted: Quine's philosophical logic is not the only one on offer; there are other ways of being rigorous which arguably take a less cavalier approach to the logical structure of the natural languages we speak. To do so is not, of course, to suppose that those languages embody truths immune to ordinary rational investigation, but perhaps it allows us to describe the world and our thought about it without the Dali-esque prohibitions Evers and Walker demand.

REFERENCES

Evers, C.W. and J.C. Walker (1982). 'The Unity of Knowledge.' Access 1, (2).

Gellner, E. (1981). 'Pragmatism and the Importance of Being Earnest.' In R.J. Mulvaney and P.M. Zeltner (eds.) Pragmatism: its Sources and Prospects. Columbia, S.C., University of South Carolina Press.

Kaplan, M. (1983).'Decision Theory as Philosophy'. Philosophy of Science, 50 (4).

Manser. A.(1982). 'Language, Lanquage-Games and the Theory of Meaning.' The Aristotelian Society, Supplementary Volume, LVI.

Prior, A. (1963). 'Is the Concept of Referential Opacity Really Necessary?' Acta Philosophica Fennica, XVI.

Quine, W.V.0. (1981). Theories and Things. Cambridge, Mass., Harvard University Press.

Smullyan, A.F. (1948). 'Modality and Description.' Journal of Symbolic Logic, 13.

Williams, C.J.F. (1981). What is Existence? Oxford, Clarendon Press.

URL http://www.uwichill.edu.bb/bnccde/epb/quinevapeman.html

HTML last revised July 7th, 2000.

Return to list of publications.