Review: [Untitled] Reviewed Work(s): The Liar. An Essay in Truth and Circularity. by Jon Barwise; John Etchemendy W. D. Hart Mind, New Series, Vol. 98, No. 391. (Jul., 1989), pp. 451-453. Stable URL: http://links.jstor.org/sici?sici=0026-4423%28198907%292%3A98%3A391%3C451%3ATLAEIT%3E2.0.CO%3B2-4 Mind is currently published by Oxford University Press.
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/oup.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.
The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers, and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take advantage of advances in technology. For more information regarding JSTOR, please contact
[email protected].
http://www.jstor.org Sun Sep 2 01:13:47 2007
Book Reviews
The Liar. A n Essay in Truth and Circularity. By Jon Barwise and John Etchemendy. Oxford and New York: Oxford University Press, 1987. Pp. x i i t 185. L15.oo. This is a splendid book. Jon Barwise and John Etchemendy have striking new ideas and material. These they have thought through deftly and masterfully. Their selection from their material and their exposition measure u p to Horace's ideal that ars est artem celare. This is a book to seize the philosophical imagination. Their examination of the liar paradox falls into three parts. T h e heart of part I is a fascinating new conception of sets. Their exposition follows Peter Aczel. O n the now dominant iterative model of sets, we picture sets as 'built up from below' by iterating (through the transfinite) the taking of sets of things previously taken. I t is built into this model that a non empty set always has a member disjoint from it; this principle is called the axiom of foundation. I f we picture sets and their members as points, and membership as an arrow from members to a set, then foundation says that every set has a lineage without closed loops all the way back to 'the beginning'. Aczel lets us take any directed graph, even ones with closed loops, as a picture of membership. T a g nodes to which no arrow goes with things without members. A decoration of a tagged graph assigns an untagged mode n the set of all and only the things at nodes whose arrows go to n. Aczel replaces foundation by an axiom of anti-foundation: each tagged graph has a unique decoration. Barwise and Etchemendy explain Aczel's work shrewdly: they provide no excuse for any philosopher to lose heart; and they whet the appetites of the educated. Let a be an object, and q, a property. Think of the proposition that a has q as the ordered pair of a and q. If we construe the ordered pair of x and y as {{x),{x,y)), it is easy to draw a wellfounded lineage for this set down to x and y. Now alter this graph so that the node tagged by x is the node where the ordered pair of x and y lies. T h e result is a picture of the proposition p that p itself has q. So we have a fixed point theorem: for any property q, there is a proposition p 'saying' that p itself has q. Aczel's ideas let us model self-reference so naturally. (It is fun to look u p 'selfreference' in the index of The Liar.) I n part 11, Barwise and Etchemendy develop a conception of truth for propositions that they name, without historical commitment, after Russell. O n this view, a model of the world is, very loosely speaking, a collection of propositions; a model makes a proposition true when it is in the model, and a proposition p is true in a model when the proposition that p is true is in the model. From this conception they draw the moral that the Russellian liar proposition 'is made false by the world but its falsity cannot be construed as a fact in the world', and they call this moral 'somewhat puzzling' (p. 79). O n the Austinian view in part 111, an (accessible) proposition is, roughly, an ordered pair whose first member is a (contextually demonstrated actual) situation and whose second member is a type of situation (roughly, a Russellian proposition); the Austinian Mind, Vol. xcix
.
391.
.
July 1989
@ Oxford University Press 1989
452
W . D. Hart
proposition is true if its first member is of the type that is its second member. For any situation s and any proposition p, there is a unique proposition, F(s,p), about s that p is false. By the axiom of anti-foundation, there is a function f whose value for any situation s is F(s,f(s)), the proposition about s that f(s) is false. O n the Austinian view, what is really happening in the liar is that for any actual s, f(s) is false about s, but the fact of its falsity, while in the world of which s is part, is not in s (p. 135). Barwise and Etchemendy want a diagnosis of the liar. By this they mean more than a conversion of the original paradoxical argument into a reductio ad absurdum of some of its (probably enthymematic) premisses. They also want a reasoned choice of which premisses are so discharged and this should include a positive statement of the real lesson taught by the liar; hence the above remarks about the moral of, and what is really happening in, the liar. But on what basis can such a choice be reasoned? Their announced basis seems to be 'our vretheoretic intuitions about truth and the world'; they claim that the Austinian treatment measures u p to these better than the Russellian (p. 20). Call these intuitions semantics. One might wonder whether semantics is articulated in any established and systematic body of doctrine with a record of successful applications outside its own borders, and thus worry that semantics is not a good basis for reasoned choice. But their announcement of their basis seems inadequate to the power of their practice. For they preserve the old association between the semantic and the set theoretic paradoxes, and they charmingly assimilate their Austinian diagnosis of the liar to the treatment of Russell's paradox characteristic of the now dominant iterative conception of sets: as Russell's paradox shows that there is no set of all sets, so the liar shows that we cannot assert propositions about the universe of all facts (p. 173). Set theory is a more established and systematic body of doctrine than semantics, and has a better record of successful applications outside its own borders, so it might seem a better basis for reasoned choice. Could there be a principled correlation between various set theories and various theories about 'truth and the world'? They ask how much we give u p if we surrender the belief that propositions can be about the world as a whole, and they answer: arguably, not much (p. 174). An older answer might have been logic itself, that is, the unqualified generality Frege and Russell took to be definitive of the truths of logic.' When we now say that '(Vx) (x = x)' is valid, we mean that every member of every non-empty domain is self-identical. O n the iterative conception, that domain is in another, but not in itself; '(Vx) (x = x)' does not seem the unqualifiedly general truth simpliciter that made it a law of logic for Frege and Russell. Perhaps we cannot mean absolutely everything all at once by 'all', but is it so plain that this conception of logic isn't much? S o for all their emphasis on distinguishing in future between negation and denial (and their apparent countenancing of negative atoms), and even if it would have cost complexity to include a version of quantification in their simple illustrative language, one wishes they had included a version of it to compare with the idea that 'all' means absolutely everything. T h e universal quanitifer is how we
'
See, for example, F. B. Fitch, 'Self-Reference in Philosophy', Mind, 1946, pp. 64-73; Thomas G. Ricketts, 'Objectivity and Objecthood: Frege's Metaphysics of Judgment', in Frege Synthesized, ed. Haaparanta and Hintikka, Dordrecht, Reidel, 1986, pp. 65-95; and Peter Hylton, Russel and the Emergence of Analytic Philosophy, Oxford, Oxford University Press, forthcoming.
Book Reviews
453
express universality, and according to Barwise and Etchemendy (and ~ a r s k i ' before them) the universality of colloquial language is the issue raised by the liar.
University College, London
W. D. HART
Representation and Reality. By Hilary Putnam. Cambridge, Mass.: Bradford Books, 1987. Pp. xv + 136. L14.95. I.
Introduction
This book is an exercise in intellectual honesty. In it Hilary Putnam, a man with more claim than anyone else to be called the Father of Functionalism, argues that functionalism is false, that it won't answer the questions philosophers want to answer about the nature of intentionality. The arguments of the book show, claims Putnam, that we must drop a certain view of the mind, the view that what is real is somehow 'under', 'below' or 'more fundamental than' everyday appearances. His reasons for the conclusion are three: (i) (ii)
Intentionality is not reducible (to physical and/or computational properties or relations) Intentionality is not primitive (there is no property such that all instances of a given intentional phenomena have it in common) and yet:
(iii) Intentionality is not eliminable/mythical (unless truth is too) Putnam concentrates on establishing the first and second claims, but the argument for the third claim, to which chapter four is devoted, is ingenious and deserves careful attention. I won't discuss it here in the detail it deserves but urge readers to study it themselves, especially if they find Eliminativism attractive. Putnam's arguments are basically very few and very simple but his style sometimes makes it difficult to see this. In places the book manages to be repetitious yet confusing and Putnam has a disconcerting tendency to veer off into mathematical proofs accessible only to the mathematically sophisticated, which won't endear him to the average reader (see especially chapter four and the appendix). About a third of the book is devoted to the old arguments for environmental and social determination of meaning (for which we once again owe a debt to the younger Putnam), but this is not really a criticism, because we are shown how these arguments are thought by Putnam to work with the new arguments derived from the normativity and holism of meaning, to falsify any possible version of functionalism. It is important while reading the book to keep in mind a distinction between the reduction, in Putnam's sense, of mental states such as believing Fa, (there would be a reduction of the state of believing Fa to a physical/ computational state if we were to find a physical/computational property or relation co-extensive with, Alfred Tarski, 'The Concept of Truth in Formalized Languages', in Logic, Semantics, Metamathematics, trans. J.H. Woodger, Clarendon Press, Oxford, 1956, pp. 164-5 Russell too deserves a nod here: see 'hiathematical Logic as based on the Theory of Types', in Logrc and Knowledge, ed. Robert C. hiarsh, London, George Allen and Unwin, 1956, pp. 59-102. This is not to suggest that Russell's and Austin's views do not contrast as Barwise and Etchemendy suggest.
