Bryn Mawr Classical Review 97.7.27

Jonathan Barnes, Logic and the Imperial Stoa. Leiden: Brill, 1997. Pp. xi, 165. $66.25. ISBN 90-04-10828-9.

Reviewed by Brad Inwood, Department of Classics, University of Toronto,

The topic of this compact monograph (number 75 in the Philosophia Antiqua series) defies expectations. The general view that in the Imperial period the Stoics did not do much interesting work on logic is solidly entrenched and supported by a great deal of respectable evidence; hence one might easily doubt whether there is a book to be written on the topic. However, if anyone actually could write the book with this title, it would be Jonathan Barnes, whose keen eye for logic is matched by a rhetorical skill of which few logicians can boast. Both virtues are needed, for without the aid of considerable rhetorical elaboration (i.e., auxesis) this book would have been impossibly short.

Barnes explains the genesis in his preface: it began as a seminar paper, became a lecture which turned into an article; thence a monograph. But how? He admits that there is too little material and that what there is is pretty thin stuff ("gossamer"). "Well, I took the thing up as a wager. Whether or not the bet was won, I persuaded myself that the subject was, after all, good for a hundred pages of print" (x).

Barnes doesn't say how much he had riding on the bet, but I wouldn't myself want the task of adjudicating the inevitable dispute. The reader might think that he must have won. After all, he has written a whole book on Stoic logic in the Imperial period, hasn't he? Well, yes and no. For in fact the book is about Epictetus' use of Stoic logic and not much else, and some of what is in the book might fairly be diagnosed as mere rhetorical amplification.

Barnes claims (ix) that the argument of his book is delimited by the careers of Seneca, Epictetus and Marcus Aurelius. Since other Stoics (such as Musonius, Cornutus, and Euphrates) are complete non-starters, a study of logic in the works of these three philosophers would count as a book on Imperial Stoic logic. But by p. 11, at the end of a chapter which surveys the little we know about logic in the first two centuries A.D. and carefully defines what is meant by "rejecting logic" (as so many later philosophers were alleged to do) Barnes gives up on the emperor. "I shall say no more about Marcus, who is a hopeless case". Marcus had once expressed gratitude for having avoided dealing with logic, and from Barnes' point of view he is indeed beyond salvaging. There is more on Galen and Sextus (indeed, more on Alexander of Aphrodisias) than there is on Marcus Aurelius.

The situation is better with Seneca. Barnes correctly infers from Seneca's often extravagant hostility to logic and from the discouragement he offers to Lucilius that logic was an extremely popular philosophical theme in the early empire, a point which emerges even more clearly from his analysis of Epictetus. Furthermore, he gives us a careful analysis of the very limited sense in which Seneca rejected logic: sophisms are of little use to a philosopher, and in the same way other demonstrations of self-indulgent intellectual trifling are to be discouraged. What Seneca actually thinks about logic is much milder than the stereotyped blanket rejection: logic is merely useful, an instrument for philosophy and no more. In this, Seneca's attitude toward logic is not (Barnes claims) very different from those of Alexander of Aphrodisias and Galen who were in their own way utilitarians about logic. He carefully distinguishes Seneca's attitude toward logic from his attitude toward physics (which has considerably greater independent value). But that is it for Seneca. Not a very rich haul.

With Epictetus matters are different. The chapter on Epictetus stretches for over a hundred pages and is articulated into ten sections which look rather like chapters themselves. Here, then, is the heart of the book; the two-page conclusion (126-7) is mostly about Epictetus, and the book concludes with a valuable appendix consisting of an edition, translation and commentary on Discourse 1.7 (On the Use of Changing Arguments, Hypothetical Arguments, and the like). Barnes, then, wins his bet in so far has he has managed to produced a book on Imperial Stoic logic (Epictetus is an Imperial Stoic, the meat of this book is about his logic, so the book is about Imperial Stoic logic). But he ought to lose the bet on the grounds that the book is really about Epictetus alone (much of what is valuable in the Seneca chapter could easily have been worked into the margins of the treatment of Epictetus).

The hundred pages on Epictetus are among the best written on any aspect of his philosophy, and are well worth the reading even for those who care little for logic and whose passion is (like Epictetus') for ethics. Barnes shows himself to be a wonderfully sensitive interpreter of Epictetus, especially keen in disentangling the rapid changes of speaker and outbursts of prosopopoeia which make his artfully casual diatribes such tough reading. For example, fragment I of Epictetus is usually read as his rather Cynical rejection of the utility of physics. Read properly, it is his ironic and aggressive assertion of the necessity of physics; Epictetus remains a Stoic in doctrine by courtesy of careful reading and thoughtful punctuation (pp. 25-27). Barnes also shows in section A that Epictetus was in fact moderately well schooled in logic. In two cases (Disc. 1.8 and 4.1.61) the usual interpretation of the text is rejected in favour of a more careful reading which leaves Epictetus a more competent logician than he has normally been taken for.

In section B Barnes shows again what he inferred from a consideration of Seneca, that there is every reason to believe that the rejection of excessive devotion to logic reflects a positive mania for the subject among philosophers in the first century A.D. The conventional picture of Imperial philosophy neglecting logic for ethics rests on the fact that later ancient authors have preserved for us the voice of protest rather than the mainstream. In sections C and D he elaborates Epictetus' view of the centrality of practical application in philosophy. His objections are often not to logic per se, but to anything detached from the concrete improvement of our lives. Logical theory for its own sake, like the commentary-based syllabus of the contemporary philosophical schools, does nothing to establish what is true or to improve the moral quality of our lives. For that we also need askesis (and Barnes has an interestingly pragmatic view of moral practice which he sets in illuminating contrast to that of Pierre Hadot -- p. 47 n. 101).

In sections E and F Barnes turns to Epictetus' positive views about the role of logic in philosophy; Disc. 1.7 plays a particularly important role. The key point here is that Epictetus is (as Adolph Bonhoeffer so often saw) a very conservative thinker, harking back at least to Chrysippus. For Epictetus as for Chrysippus logic has substantive moral value. As a rational animal, man cannot really fulfil his nature unless he reasons, and reasons well. "Epictetus thinks, first, that you ought to argue, and that you ought to study logic. He does not think, merely, that if you argue, then you ought to argue well" (p. 65). This is a more than instrumental view of logic; Epictetus values logic more than Seneca did.

In sections G and H Barnes assesses Epictetus' attitude and contribution to more technical questions: paradoxes and sophisms (he preserves important information and does not treat them with contempt) and the nature of logical analysis (alas, we learn little). With hypothetical arguments in section I and so-called changing arguments in section J we get much further. Hypothetical arguments, Barnes argues, are not the province of a separate kind of logic (as Peripatetic hypothetical syllogisms might be thought to be), but rather the dialectical application of perfectly normal Stoic logic, where the second premiss of a modus ponens argument is asserted hypothetically within the context of a dialectical encounter. "If P then Q; but P, therefore Q." There is nothing logically special here if one says instead "let's suppose that P". What is special lies in the dialectical context and the (once again) interestingly Chrysippean attitude to dialectic which Epictetus took.

Section H deals with an apparently obscure theme, logoi metapiptontes or changing arguments. We know little about such arguments outside the pages of Epictetus (one reference in Sextus), but are helped by the knowledge that propositions could change their truth value in conventional Stoic logic. Sextus' contribution (at PH 2.229) is discussed in a ten-page excursus. But the Epictetan payoff is worth the wait. If, as Barnes argues, a changing argument is one in which one of the premisses changes its truth value, then Epictetus' interest in them, and in the proper way to behave when dealing with them, shows again that Epictetus is pursuing the question of how one ought to respond to paradoxical or problematic arguments in a context of face-to-face dialectical encounter. Dialectic of the sort presupposed by Aristotle's Topics was evidently alive and well in the time of Epictetus (just as it was in the time of Chrysippus), and Epictetus displays pragmatic good sense at 1.7.16-19, where he claims that if our opponent's premiss somehow changes its truth value, then we ought to withdraw our assent to it. Chrysippus did something similar when faced with an obviously dangerous sorites -- he kept silent before he fell into the trap.

On matters of logic, then, Epictetus emerges as a distinctively interesting philosopher. There is no evidence that he developed the topic, but abundant evidence that he was well trained in it, and that his attitude towards the use of logic and its role in a philosophical life was not that different from what we may reasonably suspect about Chrysippus. The difference between them lies, one may conclude, in the fact that Chrysippus seems to have thought that logic was of greater independent intellectual interest than Epictetus did -- which is hardly surprising, given the enormous contribution Chrysippus made to logic himself. The image of Epictetus we are left with is one which Bonhoeffer would recognize; it is important to extend Bonhoeffer's picture of Epictetus' place in the history of the school to logic as well, and Barnes has done that with wit and literary grace.

And now the complaints. First, there is no doubt that the book is padded by interesting digressions, needless use of argument in utramque partem, and an unusually leisurely pace of argument. Second, although typographical errors are rare, there is a surprising number of type-setting errors. The commonest of these occurs very frequently indeed: a word in the middle of a line appears hyphenated as though it were at line end (e.g., "dis-tinction" and "dis-covery" in the middle of the line on p. 69). This trivial nuisance can only be the result of haste in the final stages of book production, and Brill (whose series is thriving at the moment) ought really to take the time to finish off its books properly before printing them.

The final verdict? With some hesitation I would say that Barnes should lose his bet and pay up. The book is really about Epictetus and the title really ought to announce that clearly. It is, however, an excellent book, worth reading by anyone interested in Epictetus, even those left cold by logic generally. As penance for the venial deception wrought by the title, I suggest that Barnes take on the job of producing a general book on logic in Imperial philosophy generally. Epictetus might then be seen to occupy a very interesting place alongside representatives of all the major schools: Alexander, Galen, and Sextus, among others.