The title and subtitle of Christopher Craig’s book point to its broader and narrower goals: one the one hand, a detailed study of the rhetorical figure known as dilemma; on the other, an attempt to explore non-rational methods of persuasion in Cicero’s forensic speeches. In particular, Craig considers the way in which the very presence of dilemma may impel the jury to accept arguments which the cold light of reason would reveal to be specious. In both aspects of his book, Craig has much to say that is both interesting and valuable; in the narrower study of dilemma he gives a careful and detailed analysis of the evidence, while in the broader study of persuasion he deals with issues that are crucial to the interpretation of Ciceronian oratory. He does not pursue these issues as far as he might, but his study is suggestive in the best sense of the word.
What dilemma is can be illustrated by an example given by Aristotle ( Rhet. 1399a19ff.) and cited by Craig (11): “If you say what is just, the people will hate you; but if what is unjust, the gods will.” In other words, damned if you do, damned if you don’t. Craig (25) defines dilemma as “the offering to the opponent (not necessarily in direct address) of two choices such that he must choose one or the other, and either choice hurts him.” Most frequently, as Craig’s discussion shows, it is used in an ethical context: if you do (or did) X, you are a villain; if you do (or did) not-X, you are a fool. Dilemma is an extremely powerful argument, if rightly deployed: indeed, in certain circumstances it is unanswerable. Craig’s choice of dilemma as his paradigmatic rhetorical form is an excellent one: dilemma not only is, in logic, irrefutable, it is an argument that is, as Craig says, immediately recognizable. What Craig does not say in his introductory discussion—although he recognizes it rightly and repeatedly throughout his book—is that while being formally a powerful argument it is in fact (at least as deployed by Cicero) almost invariably bogus: only if the two choices in fact exhaust the available possibilities is it logically valid. The most succinct modern example of the bogus dilemma is the famous question “Do you still beat your wife?” The trick lies in framing the choice such that it appears exhaustive while in fact is not. As Craig puts it (63): “The very use of dilemma seems to confer legitimacy on the assumptions that inform it.” Most Ciceronian dilemmas are tricks; and, as Craig’s discussion shows, therein lies the opportunity for oratorical skill and the need for critical analysis.
Craig’s book is divided into nine chapters. The first discusses the place of his work in recent Ciceronian scholarship, explains his choice of dilemma as an exemplary rhetorical figure, and provides a conspectus of ancient discussions of dilemma, concentrating in particular on Cicero’s own youthful De inventione. That is followed by chapters devoted to seven speeches: Pro Roscio Amerino, as one of Cicero’s earliest speeches, and then in chronological order the six in which the figure of dilemma is most frequently employed: the Divinatio in Caecilium preceding the Verrines, Pro Roscio Comoedo, Pro Sulla, Pro Caelio, Pro Plancio, and the Second Philippic. The last chapter is a brief, analytic summary of the uses to which Cicero puts dilemma and related forms; it is followed by an appendix giving text and translation of all dilemmas in Cicero’s speeches (coded to match the categories in chapter 9) and another giving tables of the syntactical variations of dilemma forms. The format of chapters 2-8 is similar: a discussion of the case and an outline of the speech; analysis of each instance of dilemma, showing how the dilemma contributes to the structure and argument; and a concluding summary drawing parallels between the uses in the speech under discussion and those discussed previously. Craig pays particular attention to formal questions, and he is very careful—and enlightening—in showing how Cicero extends the uses of dilemma as described in textbooks of rhetoric ( De inventione, Ad Herennium, and to a lesser extent Quintilian) to fit his own needs: it moves from being purely an attack on one’s opponent to a means of attracting sympathy for one’s own client or oneself.
The speech-by-speech format adopted by Craig has advantages and disadvantages. It is important, as he points out, to examine each instance in context, and not merely to catalogue or categorize; and he does examine them, one by one, in order. But exhaustiveness can be exhausting; and so too are the extended discussions of each case, with long and redundant bibliographies in the first footnote of each chapter and often with far more detail than is germane to his own argument. Moreover, in a number of cases he spends far too much time explaining and then modifying the idiosyncratic interpretations of W. Stroh. At the same time, however, Craig’s thoroughness in describing each case and speech is a great help if one is not altogether familiar with the orations discussed. Although Craig claims (x) to expect readers to know the speeches before reading his book, that is not really necessary; and some of the speeches which he analyzes—Pro Roscio Comoedo above all—are rarely read by anyone but a specialist. Furthermore, the discrete consideration of each speech will make it possible for those who teach Pro Caelio or the Second Philippic to read Craig’s chapters on them and profit from his lucid analyses.
What interests Craig most of all is what he calls the “presentational” aspect of dilemma: its use (cf. p. 168) as argument rather than within an argument. It is this aspect that leads to the broader considerations in his book, to which I will return shortly. But not all the speeches discussed are equally revelatory of presentational dilemma: in Pro Sulla, for instance, Craig admits defeat (103): “Dilemma forms, while remarkably plentiful in Pro Sulla, are not integral to the persuasive strategy in a unitary sense; the special qualities of the device and of its repetition do not function together to make a single definable contribution to the persuasive strategy.” An honest conclusion, but then one asks why Craig included the speech at all. In this speech, as in Pro Plancio, the use of dilemma is largely to place Cicero himself in a sympathetic light, and to defend his client through his own character. Craig reveals these techniques well, but they have little to do with the special qualities of dilemma, much more with the “rhetoric of advocacy” analyzed by George Kennedy some years ago. The special nature of dilemma does emerge in several other speeches, in which Craig’s analysis is more successful. Thus, part of the purpose of the Divinatio is to show that Cicero is a better speaker than Caecilius (and would therefore better represent the Sicilians against Verres); in order to do so, he uses dilemmas to show that Caecilius would be unable, in his prospective prosecution of Verres, to handle the dilemmas that would be thrown at him by Hortensius, Verres’ advocate (cf. esp. p.60). Similarly, in the discussion of Phil. 2, Craig again shows dilemma to be the figure used to demonstrate the speaker’s mastery of rhetoric. To some extent, this approach is also valid for Pro Roscio Comoedo and Pro Caelio, in each of which drama is part of the subject as well as part of the technique of the speech. Dilemma, as Craig’s account of the first of these shows, is a specious manipulation of language; and it is most at home in cases where the use of language or of oratory is itself an issue. As Craig says (41), the dilemma is “a symbol of memorable and irrefutable argument” and thus it is the master-figure for the ability of oratory to persuade—and to deceive.
Craig knows very well how deceptive dilemma is—indeed, how deceptive Ciceronian rhetoric is in general—but his approach to the subject is excessively trusting. Thus he seems to believe that handbooks of rhetorical theory such as De inventione are reliable guides to what one might expect Cicero’s practice to be, and the single sentence in Quintilian on the topics to be considered in a Divinatio—a sentence which offers no explicit ordering of the topics—is taken (50 and 55) to be a norm of disposition against which to measure Cicero’s Divinatio. Craig is also unclear about the norms of truthfulness in forensic oratory: he chooses to believe that Cicero was telling the truth in the Verrines when he said that Caecilius had a genuine cause of enmity with Verres, but not in the Divinatio, when he makes light of it. Both statements are motivated by context, and the truth can not be determined—certainly not enough to base an argument about Cicero’s technique on the “facts” of the case. In discussing Pro Caelio, Craig creates a Cicero who is trying to avoid an admission “in quotable form” (110) that Caelius had an affair with Clodia—as if the prosecution required a verbatim transcript in order to attribute words to the defense. So too, in discussing other cases, he seems to think that the prosecution speech to which Cicero is replying in fact contained the arguments and emphases that Cicero attributes to it. A defense lawyer is trying to get his client off, not to report the prosecutor’s case fairly: we generally have no other evidence, but we need to be content with ignorance.
Truth is not an issue in a Ciceronian forensic oration: winning is. Indeed, it is arguable that no client defended by Cicero was “innocent” in strict law of the crime of which he was accused: it is perfectly clear that Cluentius, Archias, and Sestius were guilty as charged, not to mention Murena, Flaccus and Milo. Hence, unless we have some “objective” external evidence, we might as well dispense with trying to figure out what the “truth” of a case is, and how Cicero manipulates it. We will never know, nor should it matter. Where the real problem arises is in trying to decide whether the truth of the case made any difference to the judge or jury either. Most interpreters, including Craig, assume that it did; hence Craig resorts several times to the desperate argument that Cicero intended speeches to be heard differently by different segments of the audience. In the case of Pro Roscio Comoedo, this is singularly unconvincing: we are left with a judge, Piso, who is expected to be hoodwinked by Cicero’s false description of an arbitration over which he (Piso) had himself presided. But it should be pointed out that Craig makes a brave attempt to explain a speech that is both incomplete and (to me at least) completely incomprehensible. But if a speech is filled with material and arguments bound to be recognized as specious by at least part of the jury, then Craig’s concern with the non-rational effects of rational argument is absolutely central to our understanding of Ciceronian oratory. He does not manage fully to come to terms with this issue; it is perhaps symptomatic of his difficulty that he varies between speaking of dilemma as irrefutable argument and speaking of dilemma as the symbol of irrefutable argument.
Over and over again, Craig points out the misleading logic underlying a Ciceronian dilemma; over and over again, it is obvious that Cicero is using dilemma to play games with words, to quibble with his opponent, to play fast and loose with fact, law, and language. One of Craig’s great services in this book, indeed, is to make it abundantly clear how cavalier Cicero could be with his case and with his audience. But then, one is forced to ask, are these dilemmas irrefutable arguments, or are they symbols of irrefutable argument? By using such a visible form, Cicero draws attention not to the overwhelming strength of his case, but to the overwhelming strength of his rhetoric: not to irrefutable argument, but to irrefutably specious argument. Strange as such a suggestion may sound, it is in fact in harmony with many of the other tactics used by Cicero in his courtroom speeches: his regular practice of turning the assault away from his client to himself ( Pro Sulla and Pro Plancio); his misdirection (as in Pro Roscio Comoedo and Pro Caelio) by the use of comic patterns; his general tendency—blatant in Pro Plancio—to avoid discussing the formal charge itself. All that these tactics (and others as well) do is to show clearly how little Cicero cares about the actual defense of his client; and those few cases in which Cicero actually does talk about the case (e.g. Pro Caecina) should only serve to heighten one’s sense of the peculiarity (to use no stronger word) of Cicero’s defense tactics in most of his cases. It is in this context that the use of dilemma belongs: it is, as Craig rightly says, a symbol of irrefutable argument, but one being flaunted before the jury as an indication that the argument is nonsense. One of Craig’s reasons for choosing dilemma as his paradigmatic argument is that it was well known to the audience (from the rhetorical training they had received in their youth) and well understood. Fair enough; but if the audience understood it, they understood it not as a knock-down argument, but as a trick. Anyone educated in rhetoric would view a dilemma as a red flag: in the press of the trial, they might not figure out what the dilemma was concealing: but that it was a deception could never be doubted. What won the juries over was not the validity of Cicero’s case, but the amazing boldness of his argument; not truth, but sheer, unmitigated effrontery. What Cicero makes almost explicit in the Pro Caelio is true of his other speeches as well: they were entertainment, and they were rewarded as performance.
All this can be taken one stage further. The speeches that we have are not the speeches Cicero delivered, but the texts of those speeches that he chose to publish later. Whether they are close to the original speeches is unimportant and impossible to determine; but they are representations of genuine speeches, and meant to be taken as such. But how would the contemporary reader react to the texts that we have, and which Craig has analyzed so carefully? How would Sulla or Plancius feel in studying a text which reveals on close reading (e.g. Craig’s) just how fallacious the defense was? How would Cluentius feel in reading the speech which, while getting him acquitted, makes his guilt palpably clear? And how would a jury feel after letting one of these people off, and then seeing how they were hoodwinked? Questions like this lead to the further question: if abstract virtues such as truth or innocence really mattered in these cases, how on earth could Cicero publish the speeches?
He obviously did, and we have most of those that he chose to publish. Revealing his techniques and the duplicity of his arguments clearly did not distress him or his clients or the once-and-future juries who will have read them. But that leads back to what I think is the larger truth embedded in Craig’s discussion of dilemma: that truth itself, the guilt or innocence of Cicero’s client, was rarely very important. As noted above, Craig is more concerned with the “truth” of Cicero’s cases than the evidence will allow; but the very importance of dilemma shows that the quest for truth in such matters is ill-conceived. The blatant fraud of most dilemma-arguments establishes a collusive bond between advocate and jury: a recognition on both sides of the fraud, and a simultaneous recognition that it is both entertaining and unimportant. In most cases, probably not even the prosecutor cared whether or not the alleged crime had been committed: they wanted the defendant punished, and that is a very different thing. Modern pieties about justice and trial by jury have little place in the Ciceronian court, and Craig’s examination of dilemma provides compelling evidence (if more was needed) that modern notions of proof and evidence do not belong there either.
 Craig’s book is too interesting to dwell on its minor flaws, but they should be mentioned. The bibliographical footnotes are referred to above: they are highly repetitive, and there are grammatical errors or sentences out of order in several, leading me to suspect that an editor asked for them, and failed to edit them. There are other signs of authorial or editorial sloppiness: dilemmas are supposed to be bold-face in Latin and in translation; but in some cases they are not, and in some what is in bold face in Latin does not match what is bold face in English. In terms of language, Craig consistently misuses “substantial” to mean “substantive” and refers to “synthetic” dilemmas when he means “fallacious” or “false”; he also uses “recur” for “return”, “adduction of” for “reference to”, and “affirmed” for “confirmed”; in general, I found his prose heavy and hard to read. And it is somewhat disconcerting that there are mistakes in one of Craig’s few translations of Cicero ( Planc. 4, pp. 129-30; he generally uses the Loeb), and that one finds flamen divo Iulio and (correctly) flamen divi Iulii on successive pages (163f.).