Barnes, Jonathan. Truth, etc

Edward M. Engelmann

BARNES, Jonathan. Truth, etc. Oxford: Clarendon Press, 2007. viii + 551 pp. Cloth, $65.00–“Ancient logic is not sexy,” laments Jonathan Barnes in his introduction to Truth, etc. Yet in the tradition of Lewis Carroll, Barnes has given us a book on logic that is witty, entertaining, and idiosyncratic. In a word, sexy. Not that it is an easy read, or that the arguments are not subtle, or the scholarship not impeccable. Barnes makes use of passages drawn from a large cast of characters, both well known ones such as Aristotle and Proclus, and ones not so well known such as Dionysius Thrax and Herminus. He also makes many references to modern logic. The book is based on the John Locke Lectures given at Oxford, 2004, and it sticks to the lecture format in that it contains no bibliography or scholarly references. It does, however, contain footnoted Latin or Greek texts for each passage quoted, a general Index, and an Onomasticon (one-line biographies of those quoted).

It should be pointed out that in Truth, etc. only the first chapter has anything to do with the notion of truth. The “etc.” forms the main body of the work, on ancient logical and grammatical theory. In what follows, no more than indications can be given as to the full contents.

Chapter 1, “Truth,” opens with a reference to the Stoic Chrysippus’ maxim that “every assertible is either true or false” (p. 1). This statement of the law of excluded middle sounds straightforward enough, but then, what is an assertible? Simplicius and others suggest that an assertible, as contrasted with a question or imperative, is defined as what is either true or false. Apart from the problematic nature of this definition when applied to Chrysippus’ maxim, what about assertibles that are neither true nor false (future contingents), both true and false (paradoxes), and true at one time and false at another? In considering these questions, Barnes discusses whether Aristotle and other ancients thought of the “is” in such statements as “the number 2 is even” as timeless or tenseless, as modern logicians do, or whether they thought of them as true omnitemporally, that is, true in the past, now, and in the future. This leads to a consideration of whether the ancients held to something like Tarski’s theory of truth. Barnes concludes that they do, but if an assertible is primarily a thought, as the Stoics held, then “‘snow is white’ is true iff snow is white” would be the case only if someone actually had the thought that “snow is white.” This Barnes takes to be absurd. Barnes ends the chapter with a linking of Chrysippus’ maxim, which he has interpreted to mean “every assertible is at every moment of existence either true or false,” to his Stoical metaphysical fatalism: every assertion of future events is either true or false. This contrasts with the Epicureans, who said that statements of future contingents are neither true nor false, and thus denied Chrysippus’ maxim.

In chapter 4, entitled “Forms of Argument,” Barnes discusses what makes Aristotelian (and Stoic) logic a ‘formal logic’. Aristotle himself, the inventor of formal logic, does not explicitly consider this issue, but Alexander of Aphrodisias and others do in some detail. They distinguish the matter of a syllogism–the terms–from the form–the figure and mood. By using symbols for terms, says Alexander, Aristotle shows that the terms themselves can be disregarded when considering the validity of a syllogism. Only the form of a syllogism is relevant.

But then, asks Barnes, what is the status of these symbols, or, as Barnes calls them, ‘logical letters’? Are they like modern Fregean variables? No, says Barnes: variables have no reference and sentences using them can be true or false only through the use of quantification. Aristotle’s sentential descriptions of syllogism, however, cannot be quantified, and so they have no truth value. Barnes suggests that Aristotle’s use of letters is to be understood in the same way as Euclid’s use of diagrams. According to Barnes (following Proclus), Euclid first proves a theorem for a specific figure (for example, a specific triangle), and then universalizes the result to every possible specific triangle that can be referred to. Similarly, Aristotle’s logical letters are meant to have some specific reference, but no specific reference in particular.

Chapter 5, “Is Logic a Science?” considers the Parapatetic position that logic is not a science, because it is ‘topic neutral’. Rather, logic is an art, a technique which is applied to objects of science. Barnes asserts that logic must surely be a science in the Aristotelian sense, because it is the study of a unified and determinate species of objects. Indeed, the Aristotelian syllogistic exemplifies the idea of a demonstrative science: it is an axiomatic deductive system. The axioms are perfect syllogisms such as Barbara and Celarent, while the items proved are the imperfect syllogisms. Much of the chapter is concerned with examining this Aristotelian concept of perfection. Barnes holds that perfection is (strictly, it seems) an epistemological notion concerning self-evidence of validity. He considers the rule of the infamous dici de omni et nullo in establishing perfection. Ultimately, he says, the dictum does not establish it, and that, contrary to what Aristotle claims, no syllogism is perfect. Indeed, the principle of conversion (used to prove the validity of some imperfect syllogisms) is a better candidate for perfection.

The initial sections of the final chapter return to the question broached in the initial sections of the previous two chapters: what is the philosophical status of the Aristotelian syllogistic? Barnes considers the ancient quarrel between the Stoics and Peripatetics such as Alexander of Aphrodisias as to whether logic is a part of philosophy (Stoics) or simply a tool of philosophy (Peripatetics). The Peripatetics held the latter position principally because they thought of syllogistics in terms of methodological application in scientific demonstration. Barnes finds this position to be at best pointless and at worst pernicious. Insofar as Aristotelian logic is a formal logic (chapter 4) and is a science (chapter 5), its status must certainly be greater than the merely utilitarian one assigned to it by the Peripatetics. And Barnes hints that Aristotle would agree with him. The rest of the chapter is taken up with discussion of some consequences of the Peripatetic position.

No review can do full justice to the diverse contents of Truth, etc. One can certainly take issue with many of Barnes’ assertions. Of course, that is beside the point, as any good philosophical work is bound to be provocative. One might wish that Barnes had not simply assumed modern logical theory to be superior to the ancient, and had attempted to show how deeper philosophical presuppositions of ancient logic might throw light on those of the moderns. More trivially, although one can admire Barnes’ ‘in your face’ attitude towards certain scholarly conventions, this publication of his lectures could have used at least a few explanatory footnotes. For it is not quite the case that, as he claims, the book “does not presuppose any prior acquaintance with logic, either ancient or modern” (p. vii). Nevertheless, for anyone who finds ancient logic interesting and worthwhile, Truth, etc. is an indispensable read.–Edward M. Engelmann, Bolton, Massachusetts.

Yonkers remanufacturer sees profits in old autos

Thinking of England