BOOKREVIEWS Rudolph H. Weingartner, The Unity of the Platonic Dialogue. Indianapolis and New York: The Bobbs-Merrill Company ("The Library of Liberal Arts"), 1973, ix + 205 pp., including Index. Paperbound: $ 2.95 Humans can be deceived from what is apparent. In his "Introduction" Weingartner argues that many commentators have ignored the fact that Plato wrote dialogues. He wishes his commentaries to be faithful to that fact. To be faithful thereto one must, he claims, refrain from supposing in the work of Plato the form of Aristotle. One must consider every question, assertion, argument and myth within its proper dramatic context. Above all one must comprehend each dialogue as a whole, for only then will its "theme" and the "purpose" of its author become apparent. In order to comprehend the whole in question one must undertake a rigorous examination of its parts. Weingartner's exegetical method expressly depends on the claim that Plato's dialogues are "unified" wholes. He intends in the book under review to provide partial evidence for this claim by examining, in turn, three dialogues: Cratylus, Protagoras, and Parmenides. His intention, however, is superior to his accomplishment. In his chapter on the Cratylus (a revision of an essay published in 1970) he obscures the order of the speeches which constitute the dialogue; while trying to show that it is a defense of dialectic he fails to exhibit the logographic necessity in virtue of which the dialogue is indeed one. Nor can it be said that he declares the Protagoras in its wholeness. He fails, for example, to indicate why the dialogue begins as it does; he does not grasp the import of references to the hound-hunt, beauty, etc. Finally, his chapter on the Parmenides is largely an examination of the conversation prior to Parmenides' exemplification of the mental gymnastic in which he has encouraged Socrates to exercise himself; it is therefore preparatory, at best, to the synoptic commentary we seek. One reason for Weingartner's failure is apparently this, that while the application of his exegetical method requires acquaintance with the Greek language, he admittedly falls short in this respect. A second reason for his failure is, unaccountably, the intermittency of his attention to dialogical context. For example, he argues that the Parmenides marks a turning point in Plato's conception of the objects-of knowledge and of the way to knowledge. But his argument depends ultimately on a neglect of the contexts in which Socrates' speeches about separate forms occur: he incautiously supposes that such isolated passages yield a Platonic "theory of forms." Again, he argues that Plato's Protagoras is in truth a spokesman for popular opinion. Although this claim is very likely true, it cannot be justified apart from an account of the apparent inconsistency between the Sophist's initial statement of what it is he teaches (318e5-319a2) and the claim which he makes in his long speech; but Weingartner overlooks this inconsistency, and moreover forgets the initial statement. A third reason for his failure is a tendency to appeal directly to modern commentaries rather than to the dialogues themselves. For example, while discussing Platonic dialectic in his chapter on the Cratylus he refers to Robinson's account of it but does not compare 390c10-12 with 398d3-8, or consider 388b10-c8 with 424b7-425a5. Weingartner is not yet sufficiently exercised in the way of interpretation which he has proposed. It does not follow that his commentaries are always without merit. His analyses of certain passages - for example, Prt 328d-333b, 351b-360a, and Prm 132a-133a - deserve close examination. These are passages about which there has been considerable discussion in our time. Weingartner is acquainted with these discussions; he regards them appreciatively and critically. Stewart Umphrey The New School for Social Research
Book Reviews
75
John Kekes. A Justification of Rationality. Albany: State University of New York Press, 1976. Pp. xiv, 260. Indexed. $ 20. The age-old war between those who think that reason is objectively grounded and those who deny such ground has never aroused much interest. Disagreements are usually settled by uncritical appeals to certain beliefs or by appeals to force. Dogmatism, relativism, and skepticism are the commonplace outcomes of demands for justification. Why attend to another skirmish in a war that has had little influence on ordinary practice? In their less theoretical moments, even relativists and skeptics agree that this is not the best of possible worlds. Improvements can arrive by three means: by the determining forces of history, by accident, or by rational plan. The desired accidents are unlikely and history cannot be trusted to determine the right ends. Reason would seem our only hope, tfreason is grounded in the r+al world, and tfhumans use reason. Thus, the importance of John Kekes' tasks in A Justification of Rationality: 1.) to defend rationality against its enemies, 2.) to show why humans ought to be rational. Kekes' expresses his basic thesis in a line: "Being human dooms one to having problems, and [rational] theories are the most useful devices for solving them." (p. 167) Kekes' divides his book into four parts: a statement of the skeptical challenge to rationality, an assessment o f pragmatic, common sense, ordinary language, and scientific counters to the skeptical challenge, Kekes' own account of rationality, and a refutation of skepticism. Reason can be attacked by the following dilemma: One reasons or he does not. If one does not reason, then he is a skeptic on the order of Cratylus (he does not even speak). If one does reason, then he finds that reason destroys itself with arguments. (See Hume's Treatise,Book I.) Kekes' reading of the Humean horn of the dilemma is that the rational acceptance of any standard requires supporting reasons, which leads either to an infinite regress of reasons or to circular reasoning. The skeptical challenge is to provide reasons for using reasons, but to avoid infinite regress and circularity. The initial pragmatic reply, that only practice can vindicate reason and that reason is useful, is replaced by Carnap's distinction between internal and external questions. The justification of rationality is an external question, and thus, either illegitimate or practical (non-theoretical). If the question is practical, then the answer is obvious. Kekes' criticism of this defense employs his thesis that rational ends, goals, and standards must be justified. "Practical" is short for "practical toward some end," and pragmatism cannot justify ends, it presupposes them. "The intelligibility of any theory presupposes common sense" is taken as a comment on the psychology of discovery and not on the nature of primary knowledge. Common sense is wrong on occasion and it rejects theoretical accounts that are successful in characterizing and dealing with the world. Kekes concludes that an appeal to common sense does not remove the skeptic's challenge. Kekes' treatment of the ordinary language defense centers on two themes: the paradigm case argument against idealism and solipsism, and Winch's contention that a standard of rationality is implicit in each form of life. The paradigm case argument is represented as follows: Philosophy is not about the facts, but about correct ways of talking. Ordinary language is all right. The meaningfulness of certain descriptive uses of language entails that these statements are frequently true. (How else could we have learnt the use?) Thus, the common sense world exists. Kekes replies that the three premises o f the paradigm case argument are only partially correct, and that no ontological conclusions can be drawn from them. Although philosophy is not about the facts, it may be concerned with more than correct ways of speaking. Ordinary language is all right as far as it goes. Typical philosophical goals, such as clarity, precision, justification, and evaluation, are not encompassed by the pragmatic constraints of ordinary language. There is work to be done by language, work that ordinary language does not do. Finally, a counterexa m p l e - "Nothing is really red" could be true and"red" could have a use, to wit, to describe what is apparent. Kekes concludes that solipsism is compatible with ordinary language. (p. 75)
76
The Journal o f Value Inquiry
Winch's thesis depends on the notion of "a form of life." Understanding Winch's proposal entails acquaintance with the criteria that license the use of "a form of life." Kekes argues that these criteria must either be internal or external to a particular form of life. If internal, then the criteria needed for sorting forms depend upon the prior sorting of forms of life, which means that the criteria cannot be internal. If external, then rational criteria exist which transcend all forms of life. Winch's thesis fails because "The notion of a form of life is incoherent when it is transformed from a convenient metaphor into the cornerstone of rationality." (p. 106) Reading Hume would discourage one from offering the success of science as a counter to skepticism. Reading Popper might tempt one to make the offering. If truth is viewed as a regulative ideal toward which scientific theories tend, then the possibility of error can coexist with the growth of "knowledge." Scientific theories become better approximations of reality as they increase in scope, as they lead to greater success in problem solving, and as the number of counterexamples to theories diminish. Kekes characterizes Popper's position as "weak fallibilism," the thesis that although error is always logically possible there are rational grounds for belief. Kekes argues that Popper fails to escape skepticism because weak fallibilism entails strong fallibilism, the thesis that there is no rational ground for belief. Science necessarily has presuppositions which are not established by science. The rationality of these presuppositions is attacked by skepticism, an attack which science is in no position to meet. Having found the preceding counters to skepticism inadequate, Kekes presents us with his theory of rationality. The presentation has three parts: the external account, the internal account, and an answer to the question, "Why should one be rational?". The success of Kekes' book rests on the truth of his external account. The internal standards of rationality are said to depend on the external standard, and it is said that the nature of the external standard entails that humans should be rational. The external standard is the existence of problems, an existence which is independent of any particular context and which is objective. Kekes notes that many problems are removable (e.g., finding a hammer), while others endure (e.g., physical handicap). Enduring problems are not solved, humans must learn to cope with them. It is argued that rational theories are our most successful means of coping with enduring problems. Four internal standards of rationality are presented: logical consistency, conceptual coherence, explanatory power, and criticizability. These standards should be accepted because "conformity to them leads to successful problem solving." (p. 133) Kekes attempts to justify each internal standard by showing how it leads to successful problem solving. Since the ultimate test of a standard is its utility, one is not surprised to read "there are theories that are both illogical and rational," or that logical consistency is a necessary condition of rationality most of the time. (p. 149) The standard of conceptual coherence prohibits self-defeating claims. An explanation is powerful to the degree to which it removes problems. Criticizabifity requires the possibility of a theory being mistaken. "Truth and falsity are timeless but unattainable ideals." (p. 160) The argument that humans ought to be rational rests on the claims that every human has problems and that it is in each human's interests to cope with his problems. Kekes argues that rational theories are the best means of coping with enduring problems, and thus that one who wishes to act in his own best interest should accept rational methods. The final part of the book is devoted to the defeat of two versions of skepticism: solipsistic and metaphysical. Solipsistic skepticism denies a rational ground for belief in an external world. What solipsismcannot distinguish, according to Kekes, is the difference between a correct and an apparently correct mental state, a failing due to the absence of the only referent that makes the distinction possible, the external world. Thus, sollpslsm deprives atself of the distinction between true and false statements and, since descriptive expressions are meaningful only if sometimes true, solipsism cannot account for something that exists, meaningful language. Metaphysical skepticism, taken to be the skeptie's last line of defense, asserts that there is no rational way of deciding which metaphysics is correct. Thus, there is no rational way of deciding
Book Reviews
77
between idealism and materalism, atomism and holism, nominalism and realism, and so on. Kekes replies that metaphysical skepticism is defeated by showing that metaphysical theories can be rational, that is, by showing that they can meet his five standards of rationality. He concludes that "rationality, metaphysics, and the refutation of skepticism thus stand or fall together." In what remains of this review I wish to raise several problems indicative of what I consider the main failing of Kekes' book, and then to give reasons for claiming that the book has failed to achieve its purpose. Hume is less of a skeptic than Kekes makes him to be. The key to Hume's Treatiseis the seven philosophical relations, the first four of which are necessary- they cannot change if their terms do not change. Hume intends "cannot," not "have not," an intention which is reaffirmed in Section 1 of Part IV when he allows that simple arithmetic relations can be known with certainty. Hume does not give an account of the possibility of certainty (he had to leave something for Kant), but it is clear that he does not think that such knowledge requires an appeal to evidence more basic than the relations themselves. Perhaps some things don't need an argument. If there are any necessary truths which are not human conventions, then they must have an unchanging ontological ground. The best candidate ground might be found in an examination of the problem of universals. Kekes quickly dismisses that problem, or at least what he calls the traditional version of the problem, in favor of Wittgenstein's discussion of family resemblance. Wittgenstein's discussion seems to be more about the problem of natural kinds than it does the problem of universals. Taken as an observation on the problem of universals, the notion of family resemblance leads either straight to nominalism (each of the members of a family is an individual, not something general) or back to the original question. (In virtue of what are all of these called cases ofproposifional knowledge, now that we have observed that propositional knowledge and carnal knowledge are related only in a family way?) Kekes' brush with the problem of universals deprived him of some ontological ground which might have terminated his endless search for reasons. The demand for reasons is one of Kekes major weapons in disposing of a variety of philosophical positions. The skeptics attack on reason is made to depend on two premises: every standard must be justified, and no standard Can justify itself. (p. 17) When G. E. Moore says (and correctly) "I can know things which I cannot prove," Kekes refers to the apparent perversity of the argument, and explains this by saying that here " p r o o f " means "logical entailment." (p. 54) We are told "the acceptance of scientific inquiry as a paradigm of rationality requires a demonstration of the presuppositions upon which scientific inquiry rests." (p. 83) The demand for reasons might be used one more time. What reasons can be given for the acceptance of Kekes' external standard for rationality? How is the reply that the existence of problems is simply a fact, no further reasons are needed, different from Moore's claim that he knows some things which he cannot prove? In Book Gamma of the Metaphysics, Aristotle considers the need for first principles. He remarks, "for not to know of what things one should demand a demonstration, and of what one should not, argues for want of education." (1006a) A response to a demand for justification can take one of three paths: an infinite regress of reasons, circular reasoning, or a presentation of true first principles which ground the justification. The first two alternatives justify nothing, a denial of the third is a denial of the possibility ofjustificafion. There are places in Kekes' book where he seems to recognize this problem, such as the move to an ontological ground for justification (the existence of problems). But one wonders about the status of this ground when he is told that truth and falsity are unattainable ideals. The major flaw in Kekes' work becomes apparent when one considers the concept of an epistemically serious argument, an argument where the premises are known at least as well as the conclusion. Kekes' argument exhibits "To be human is to have problems" and "Rational theories are the best means o f solving problems" among his first principles. These principles cannot be viewed as self-evident necessary truths. According to Kekes, the test of any theory lies
78
The Journal of Value Inquiry
in its ability to solve problems. In addition to its success, a rational theory must be logically consistent (most of the time), conceptually coherent, powerful in explanation, and criticizable. To be criticizable it cannot have true first principles. Kekes' theory of rationality is, on his own account, a rational theory. (9. 190) Kekes' theory is well-grounded to the degree to which it is successful. Since pragmatic criteria are special cases of inductive criteria, the support for Kekes' theory rests with an inductive argument. Thus, all of rationality hangs on the justification of induction, and on the particular induction demonstrating the success of Kekes' theory. Is the law of non-contradiction less wellknown than this induction? Are simple arithmetic relations less well-known than this induction? Is the necessity of ontological categories, such as quantity, quality, and relation, less well-known than this induction? Kekes' argument fails the test of epistemic seriousness. Near the end of his book Kekes says that rationality is a method, not a state of mind. (p. 258) But one can always ask, "What does it mean to say that one knows that a method is correct?". One cannot know a method by a method, or as Nietzsche remarked about Kant's first Critique: "[he answers] "by means of a means" - but unfortunately not in five words." William J. Edgar Suny at Geneseo
ERRATUM T h e J o u r n a l o f V a l u e I n q u i r y , V o l u m e X I , i s s u e 4. I n t h e a r t i c l e b y J. M a r t i n S t a f f o r d , On Distinguishing Lust, l i n e 28 o n p a g e 2 9 2 s h o u l d r e a d : without commitment
Between Love and
o r l o v e w i t h o u t r e s p e c t ' 5 (p. 119). I f t h i s is a l l o w e d ,