SOURCE: "Arguments in Context: Aristotle's Defense of Rhetoric," in Aristotle's "Rhetoric," edited by David J. Fuley and Alexander Nehamas, Princeton University Press, 1990, pp. 129-66.

[In the following essay, McCabe defends the structure and the content of Rhetoric, arguing that both support Aristotle's view that rhetoric is indeed an art and that it can be practiced in a legitimate manner.]

Is the opening of Aristotle's Rhetoric a muddle, an agglomeration of two versions of the text, haphazardly assembled? Or is there a coherent strategy to be found here? It has been persuasively suggested that two different strands of argument within the Rhetoric correspond to two stages in the development in Aristotle's logic (an earlier, Topics-based stage, and a later one that uses the theory of the syllogism put forward in the Prior Analytics).¹ But I shall argue that the Rhetoric is not ill-knit after all. For the appearance of fracture derives from Aristotle's standard practice of considering the views of the many and the wise to arrive at a coherent theory of his own. In this case, I claim, Aristotle explicitly confronts the opposed theories of his predecessors: Plato, who denied that there could be an art of rhetoric, and the tradition of the rhetors and sophists which maintained that rhetoric is the overarching science because the practice of rhetoric is all-pervasive. This debate about the status of rhetoric turns on two issues: The first is theoretical: is the art of rhetoric properly systematic—is rhetoric a techne-? The second is ethical (and political): is the practice of rhetoric legitimate or disgraceful? Aristotle, by showing how the first question can be answered in the affirmative, gives an account of how rhetoric may legitimately be practiced. In the process he avoids both Plato's absolute condemnation of the rhetors and Isocrates' wholehearted endorsement of their art. The result is a subtle theory of rhetoric which is complex but not incoherent.²

Plato's Challenge

Notoriously Plato had argued that the art of rhetoric is either no art form at all (because the practice of rhetoric is merely flattery, like cooking) or turns out to be philosophy. He makes the case for the first in the Gorgias and for the second in the Phaedrus; but the grounds for both are the same and threefold. He has an epistemological argument, an ethical argument, and a political one; the three are closely tied together.

Because it aims to persuade, ordinary rhetoric is fake and insincere. It puts on the clothing of something else³ and pretends to be a real skill, concerned to confer real benefit. In fact, however, it is mere hedonist flattery, an irrational knack, not a proper art (technē, Grg. 464b ff.). Why is rhetoric not a techne? Because the rhetor is the jack of all trades (as Gorgias boasts); he is able to persuade anyone about anything, but he lacks a subject matter of his own (Grg. 458e ff.). He has no real understanding (Grg. 465a); instead his objective is to make himself plausible (pithanoteros, Grg. 459a; cf. Phdr. 272d) in order to convince the ignorant, not to teach the truth (Grg. 459a, cf. Phdr. 272e) nor to improve the souls of his audience, as the politician should (Grg. 464c). After all, rhetoric does whatever it can to win its point, because it has itself no moral position to take (cf., e.g., Grg. 460e; and witness Gorgias' vacillations on whether or not rhetoric is just). So rhetors are nothing but Sophists, who care only for the correctness of words and never get to grips with "things" (ta pragmata, Euthd. 278b). And if the practice of rhetoric is flattery, an unsystematic knack, then a fortiori there can be no theory of rhetoric, no ars rhetorica. In that case, those who claim to teach rhetoric in fact do no such thing: the sophists are charlatans.

Plato's first objection to rhetoric, then, is an epistemological one. Philosophy aspires to absolute and objective standards (e.g., at Rep. 51lb); and once those standards are achieved, they provide a systematic science. But neither the practice of rhetoric nor its theory can lay claim to objective knowledge: and for that reason neither can lay claim to the title of a techne.

His second objection is an ethical one. Gorgias and his colleagues are interested in winning their cases in court, in persuading the assembly, in impressing their audience. Their activities aim at power and success. Socrates, on the other hand, denies that this is power at all; and he cares only for the inquiry into truth, which he conducts by a meticulous examination of his interlocutors one by one. Now the contrast between Socrates and the Sophists is not just a matter of good intentions; rather, Plato argues that the methods of rhetoric are fundamentally misguided, lacking as they do the proper ethical focus.

We may see this in two connected ways: i) in the argument of the Gorgias about power; and ii) in the Socratic method itself.

1. The art of rhetoric, the rhetors say, gives you power—over juries, over friends and enemies, over cities (Grg. 466b ff.). This kind of power is bound to be a good thing, because it lets you have whatever you want and do as you will (Grg. 466CI). Socrates disagrees: this is not power at all. For the tyrant may do what he thinks is best; but he does not do what he wants, because what he wants is what really turns out to be the best. So unless he knows what he is doing (Grg. 466dlO), he will end up getting only what he thinks is best—and then he will be wrong, disappointed, and not powerful at all.⁴

   Power, on Socrates' account, is about knowledge and getting things right. He contrasts two different levels of belief—"shallow" belief (what we think we want; what we are persuaded to want, what the orators convince us of) and "deep" belief (the deep structure of our beliefs and desires⁵). These two levels of belief (Socrates suggests) are usually at odds with each other; but only deep belief corresponds to what is true and objectively valuable (what we really want). Getting at our deep beliefs, therefore, is in our interests; disguising them with shallow ones is disastrous. So philosophy turns out to be a good thing, and rhetoric a bad one. If philosophy is power over ourselves, then rhetoric gives others power over us, without allowing them power over themselves either (because they don't know what they are doing, are not experts in anything, as Gorgias admitted). Power over ourselves is something we get from our own intellectual activity, and not something we can achieve either by manipulating others, or by being manipulated by them, or by manipulating ourselves. The only thing that matters, then, is the active philosophical life,⁶ the true life of the soul.

2. The methods of rhetoric compare unfavorably with the Socratic method of elenchus.⁷ A collection of premises is made (from various sources) which includes some proposition under scrutiny; an argument is mounted to investigate the coherence of the premise set; and in the end the premise set is shown to be inconsistent. The method is quite general, not (or not legitimately) to derive the contradictory of the hypothesis directly but rather to evaluate the coherence of the entire set of propositions. So the logical focus of the elenchus is consistency, which characterizes a collection of beliefs.

An elenchus also demands that the interlocutor "say what he believes."⁸ Thus, for example, Socrates encourages Charmides to discover hypotheses about self-control' from introspection—because Charmides is "obviously" self-controlled (Chrm. 159a). Or Callicles needs to reflect on what he really believes, in order to engage in argument with Socrates (Grg. 482). The investigation into what he really believes, Socrates suggests, will itself produce inconsistency with what he appears to believe—and hence he will arrive at impasse, aporia. Now aporia is not a thesis in an argument, it is a state of mind.⁹ Unlike ignorance, aporia is highly self-conscious—it has all kinds of trappings of embarrassment and shamefacedness, but it is not a condition that its patient can ignore. Socrates' position on this is famous, featured as it is in his self-defense—we are better off, he claims, being aware that we are ignorant, than wrongly thinking we know (Ap. 21b ff.). Someone whose premises have been shown inconsistent clearly does not know what to think; and in this respect he is aware of his own ignorance—so he is better off than he was at the beginning of the argument. And this positive result of the elenchus would be impossible if the interlocutor were insincere—because it would not be his views that ended up in difficulties.

What is the connection between the consistency of the belief-set and the sincerity with which those beliefs are held? In the elenchus, inconsistency is a bad fit of beliefs held by a particular person; and a person can only reveal what his or her beliefs are by meeting the sincerity condition. But then this same person will feel the discomfort of inconsistency; and the need for consistency then turns out to be a condition of the self, a disposition of the soul.

Hence recurs a theme of the Platonic Socrates' attack on the Sophists and their denial that consistency matters. "Who will you become if you are taught by Sophists?" Socrates asks (Prt. 31lb-c). "I need to know myself," he insists; "I need to make sure I do not contradict myself."¹⁰ Why does it matter whether I contradict myself or not? First of all, Socrates suggests that I care about consistency because finding myself inconsistent affects my personal integrity (compare Grg. 482a ff., Euthd. 283c ff., 287b ff.). This is why aporia is so unsettling—because I seem then to be in conflict with myself—and why I aspire always to eliminate inconsistency from my belief set. But this aspiration is not, on Plato's view, merely an attitude to the logical relations of my beliefs. On the contrary, the drive toward inconsistency is fundamentally ethical. For without psychological integrity, not only can I have no coherent life-plan, but even the "I" who is central to my having a life-plan at all is impaired.¹¹ Indeed, questions of power and success are trivial in comparison to this central ethical matter, of the harmony of my soul. So (Plato develops this point from Socrates, Rep. 436b ff.) to be psychologically harmonious is just what it is to be happy; and for this the consistency of my beliefs is a necessary condition. Indeed, if the only fully consistent set of beliefs is the set of true beliefs—and if those beliefs are only explicitly held when reason rules in the soul—then consistency is not only a necessary condition for a proper, ethical life; it is also a sufficient condition for a happy one.¹²

This radical acount of the relation between the logical matter of consistency and the ethical matter of happiness is, I suggest, fundamental to Plato's rejection of the practice of rhetoric (and thus of its art). For rhetorical persuasion can only mask the proper relations between my beliefs and interfere with the consistency of my soul. By contrast the elenchus (as would any other legitimate philosophical method) uncovers those relations and allows me to aspire toward a unified consciousness. And that—on Plato's thoroughgoing intellectualist account—-is all that matters.

All this would be true for any rhetoric that is not based on knowledge. On the other hand, we might imagine a "real" rhetoric based on knowledge (Phdr. 270ff.). This, both in theory and in practice, would be a genuine science, free of the shenanigans of the Sophists. And in such a case rhetoric would turn out to be philosophy. So the "real" rhetor would be, we may suppose, a philosopher-king; his task would be to understand the soul of the persons to whom he speaks (Phdr. 270e) and to tailor his speeches to each soul "explaining what each soul is like and showing by what speeches and for what reason one is necessarily persuaded, the other not" (Phdr. 271b). Has Plato changed his mind about rhetoric? I think not. The Phaedrus shows what kind of epistemic claims the Sophist would have to make in order to defend himself against the criticisms launched in the Gorgias. If, Socrates here suggests, the rhetor really knew what he was doing, then he would really ("necessarily") persuade his audience. But such knowledge, Socrates insists, must be concerned first and foremost with the soul; it is in this deep understanding of psychology that the ordinary orator totally fails.¹³

This account of what ideal rhetoric might be points to Plato's third, political objection to the rhetors. Philosophers care for their souls; philosopher-kings care for the souls of their subjects. This means that the proper relations between the philosopher-king (the "real" rhetor) and the citizens in the state he rules should be uncontaminated by persuasion, unmuddied by the confusions of rhetoric. Instead they should be, as it were, soul to soul. Just as the elenchus demands of me that I be honest about my own beliefs, so "real" rhetoric demands that I be honest with everyone else. Contrariwise, if I fail to face the truth I am not merely mistaken; I am actively a corrupting influence on the state in which I live.

And for that reason rhetoric turns out to be morally reprobate and philosophy to be good for us, because philosophy cares for the soul while rhetoric doesn't mind what you (or anyone else) really believe, just so long as you are persuaded. Socrates worries most of all about deep psychological health. But the Sophists fail to concern themselves with sincerity and consistency because they are just interested in what goes on at the surface—they are just interested in winning. That superficiality, that misdirection is why rhetoric and sophistry are fatally misguided—on, to recapitulate, three grounds: epistemological (rhetoric is based on no objective truth), ethical (rhetoric is self-deceiving and thus bad for its practitioners), and political (rhetoric deceives others and damages the interests of those it addresses). Anyone who would rescue rhetoric needs to answer all three charges.

The Rhetors' Challenge

Plato did not, however, have the last word. He claimed that the professional rhetors had nothing to teach, because rhetoric is a non-subject. The professionals, on the other hand, both teachers of rhetoric and Sophists, claimed that, so far from philosophy excluding rhetoric, rhetoric encompasses philosophy. Philosophy is no special discipline; instead, philosophy as rhetoric takes over the world.

Alcidamas (e.g., On Sophists 3¹⁴) attacks people who write their speeches. Instead we should speak extempore, because that is not only a sign of the highest intellect, but useful, too. For speaking off the cuff, we can seize the rhetorical "moment" (kairos) and defend ourselves against any point that happens to come up. The extempore speech is alive (On Sophists 27) and has a mind of its own; a written speech, which lacks that spontaneity, is not real at all.¹⁵

Isocrates offers a deeper, but connected, defense of rhetoric. Rhetoric is not a fixed set of rules (Against the Sophists 12) but an ability. It is the power¹⁶ to innovate, to be flexible in the face of the immediate circumstance, to seize the rhetorical moment. This is philosophy. Philosophy trains souls (Antidosis 181ff.), in a way complementary (antistrophos) to the skills of the gymnastic trainer. For after all, logos ("speech; argument") rules (Antid. 257)—it is the power (dunamis) by which we can secure the greatest goods. How? By conveying the truth as we can understand it (Antid. 271). What does that mean? Isocrates suggests that human truth is to be found in nomos (Antid. 82ff., 254ff.) not in nature. For that reason, the art of speaking which presents and discovers what is true by nomos is bound to be of fundamental importance and power.

The Sophists were held in contempt by Plato and Isocrates equally.¹⁷ Sophists care only for winning—they will use any kind of argument, any sleight of tongue to achieve that end.¹⁸ In the extreme cases they claim that whatever the opponent says they can tie him in knots; they can even argue that whatever they say must be true, whatever anyone else says cannot contradict them (Plato, Euthd. 283ff.). Thus they could be associated with a radical theory of truth: everything is true and contradiction is impossible (Plato, Tht. 151ff.); with an extreme logic: consistency does not matter (Plato, Euthd. 285); and with a moral nihilism that suggested that people are not to blame for anything they do (Gorgias, Helen).

Sophists like this are radical relativists. What of Isocrates? His job description commits him, it might appear, to the primacy of language: speech shapes our understanding of what there is.¹⁹ Perhaps this comes from cultural relativism—everything we think or do is mediated by ideology.²⁰ Or perhaps Isocrates offers a theory of truth: what is true is exactly what is interpreted by us. Or perhaps Alcidamas and Isocrates were talking about art: what we invent or create matters at least as much as what happens to us; art and creativity alone allow us to escape the determined world of causes and effects. Any of those claims might appeal to someone practicing forensic or deliberative oratory, or to an epideictic rhetor. In court, for example, maybe the facts of the matter are determined (rather than discovered) by the judge; up to that point, there are only points of view, represented by the opposing parties.²¹ That need not be out-and-out subjectivism; nor need it underpin a sophistic approach to forensic and political debate ("I'll do anything so long as I win my case"), which supposes that the end (winning) justifies any rhetorical means. On the contrary, the moral relativist like Isocrates may indeed be of unimpeachable character, only not convinced of the Platonic view that there are real, natural values unmediated by human thought. In that case, his claim to a skill or a science cannot be based on his access to objective truths; but then the sheer universality of rhetoric for someone like Isocrates might justify his claim that there are rules and principles to underpin his art of rhetoric. It is not, after all, true that a scientist cannot also be a relativist.

This view of rhetoric (which for brevity I shall call "Isocratean") might well be correct. First, maybe ideology does determine everything we say and do. Second, maybe the imagination is autonomous and thus an important manifestation of what it is to be human. Third (and perhaps consequently) value and virtue may indeed depend on the humans who employ them. Isocratean rhetoric, then, might turn out to be humanism and a thoroughly good thing; it may thus not be vulnerable to either the ethical or the political objections launched by Plato.

The Sophist and the Isocratean rhetor have one fundamental tenet in common: rhetoric is all-pervasive, nothing is free from rhetorical shape, nothing can be expressed without language, so language comes first. Rhetoric is communication—a relation between the rhetor and his audience. But in that case, the audience is an essential part of its definition. While Plato pays attention to truth and argument in souls, Isocrates thinks about the social context—the actual persons who give the speeches and the persons who hear them. Rhetoric is constituted by the interpersonal context in which it appears. And thus for Isocrates the Platonic distinction betwen real truth and the lying manipulation of the rhetors is misplaced—for there is only the truth that comes mediated by language. This may, indeed, be demonstrated pragmatically—by showing how there are always opposed arguments on any subject, by showing how people do disagree, with no serious hope of a decisive resolution. It is rhetoric itself that shows that no truth is unvarnished.

Isocrates, therefore, makes two points against Plato. Where Plato insists—almost solipsistically—on the integrity of persons, Isocrates insists on the social context of rhetoric, on the fact that rhetoric is communication between persons. So his first point, we might say, is sociological, a counter to Plato's ethical and political objections. His second point is about the nature of language and it counters Plato's epistemological objection that rhetorical practice is unsystematic. Where Plato insists that rhetoric is limited to the blandishments of dishonest speakers, Isocrates argues instead that rhetoric is all-pervasive: there is no speech (no meaningful speech) that is not communication. Rhetoric, that is, is just language: and it is therefore both systematic and a suitable object of a technē

The Strategy of Aristotle's Rhetoric

These opposed views on the status of rhetoric suggest that we either go for all of it, or not at all. Given that rhetoric is a phenomenon of the marketplace, the debate about whether it can be systematically studied and whether it can be justified is clearly important. The contrast between the Platonic view and the Isocratean one is extreme, as I have characterized them, but no less compelling for that. Such, moreover, is clearly Aristotle's view, because he uses this debate as the opening gambit of the Rhetoric; and, as I shall argue, he hangs the theory of the work as a whole on the resolution of these opposed arguments.

It might be objected, of course, that the accounts I have given of Plato and Isocrates are mere thumbnail sketches, hardly doing justice to the real views of the men themselves. If Aristotle attacks this Plato and this Isocrates, he is tilting at historical windmills and his whole enterprise is thereby rendered ridiculous. But to that objection there are two retorts:

First, as a matter of evidence, Aristotle himself sets up the dispute between Plato and the rhetorical tradition to which Isocrates was heir. For the opening of the Rhetoric invokes directly the imagery and the arguments of both these parties to the rhetorical debate, as I shall argue. So in what follows I shall treat this Plato and this Isocrates as theoretical constructs, part of the apparatus of Aristotle's developing argument.

Second, as a matter of method, think of Aristotle's standard procedure, dominated by two maxims: "save the phenomena" and "start with the endoxa."²² Now people do rhetoric all the time—in the street and in the schoolroom, in the agora and in court; here are the phenomena. But then the question arises how we are to explain the phenomena. When it comes to explanation, we should look first to the endoxa, the opinions of the many and the wise—especially so that we may find ourselves a good puzzle, one where the opposed opinions are clearly set out but still unresolved (Metaph. 3.1, 995a28ff.). On rhetoric, then, Aristotle should perhaps turn first to Plato and Isocrates, his concern not so much historical fidelity (he is no doxographer) but the setting out of a good puzzle—and between the positions of Plato and Isocrates, as I have presented them, there is just that sort of puzzle.

But if Aristotle starts with the appearances, he cannot simply reject the appearance that there are both rhetoric and philosophy, each separate from the other but both practiced (whether by the many or by the few) in a systematic way. Instead, he must negotiate the middle ground between his wise predecessors to arrive at a solution; the solution must fully explain why Plato and Isocrates disagreed, and how the appearances can be saved after all. The resolution is not just a matter of picking one side or the other; instead, progress consists in the discovery of a new approach or new understanding—not just adjudicating between the old. The test of Aristotle's success in this balancing act is whether we can see how the new explanation does indeed supplant the old, and resolve the opposed arguments on the way.

Aristotle first invokes his predecessors by echoing their language. He starts his account of rhetoric with a challenge, "Rhetoric is antistrophos to dialectic." This gauntlet is not merely thrown at the feet of Plato,²³ who wants to say that rhetoric is the counterpart of cookery, and so not a proper technē at all (Grg. 465d7-el). Isocrates too had claimed that rhetoric (philosophy) has a counterpart, the proper training of the body (Antid. 181-82). Isocrates and Plato both use analogy (rhetoric is antistrophos to something else) to show either that rhetoric is nothing at all (as is cookery) or that it is in fact proper philosophy (Oust as gymnastic is proper physical training). Aristotle, however, tries to turn the tables on both men, by offering not an analogy between rhetoric and some unrelated skill but instead a formal comparison between the argumentative techniques of rhetoric and those of dialectic. So there are both philosophy and rhetoric, similar and yet formally distinct. Aristotle must establish that distinction against both Plato and Isocrates.

"Enthymeme is the body of conviction" (Rhet. 1.1.3, 1354al5). This is Aristotle's claim for the argumentative core of rhetoric, and it exploits both Platonic imagery and a complex analogy offered by Alcidamas. For Plato, the comparison between rhetoric in its contemporary form and true philosophy is the contrast between the flattering clothes and cosmetics that disguise the reality, and the real beauty of the naked truth (Grg. 464c ff.). But when rhetoric is transmuted into philosophy (Phdr. 276a ff.), it has a body that can act on its own and defend its own ideas; it comes alive.²⁴ For Alcidamas, proper rhetoric—the immediate activity of improvisation—is alive, it is a real creature with a mind of its own; but the written word is just a fake, an image of reality. For Aristotle, enthymematic argument is the body of conviction, real or apparent, able to come to the defense of the truth (Rhet. 1.1.12, 1355a39). We may think it wears the clothes of politics (Rhet. 1.2.7, 1356a27); but true rhetoric is the body within (compare Rhet. 3.14.8, 1415b8-9), a part and the image of dialectic (Rhet. 1.2.7, 1356a31).

Aristotle's deployment of his precedessors' imagery reflects the play of his argument, where he resolves this dispute between the philosophers and the rhetors; and he appropriates their language for his own account of rhetoric. The argument of the opening chapters has the same character. The opening chapters of the Rhetoric may seem to vacillate between different points of view,²⁵ but once Aristotle's use of the endoxic method is remembered and once his deployment of rhetorical devices to invoke his opposed arguments is noticed, the apparent vacillation begins to look like a coherent argument. Consider the argument of the opening chapters in stages.

1. Is rhetoric a technē? Plato says it is not; Aristotle says it is, but a general one, about common principles. The truth of this is obvious from the fact that people employ rhetoric both automatically and as a result of deliberate practice. So there must be a technical procedure.²⁶ And yet current writers on rhetoric (Isocrates and co.?) miss the technical center of rhetoric; they write, instead, about the "clothes" of rhetoric. They fail to see that the enthymeme is the "body of conviction"; instead they occupy themselves with peripheral matters, such as how to affect the jurors. This must be wrong; well-run states forbid speaking "outside the matter." Oily if rhetoric is centered on argument is it morally correct; to try to affect the emotions of the judge is like trying to make a straight ruler crooked.²⁷ So the disputants have no other task than to prove the case (to pragma)—did it happen or did it not? The decision whether it was done well or badly, justly or otherwise is a matter that should be left to the judge (Rhet. 1.1. 1-6, 1354al-31).

   This opening salvo opposes the "body" of rhetoric to its clothes, rhetorical argument (the enthymeme) to improper attempts to sway the emotions of the jurors. So against Plato Aristotle first claims that rhetoric has a technical core, namely its arguments; against Isocrates that some rhetoric is illicit (whatever is outside the subject).

2. Aristotle goes on to consider the role of the judge (or the assembly). His opening gambit might have suggested that there is no need for a judge or an assembly (because the orator should properly discuss only the facts of the matter). But even the best-constituted states need judges—after all, the laws are general, and the job of the jury is to judge the particular. So the jury is needed to decide the facts of the matter in the particular case, but nothing further. And for this reason the rhetors' discussion of emotion is outside the subject, for this is about affecting the juror in a particular way; the rhetors should have discussed the technical convictions (entechnai pisteis) that make someone enthymematic (Rhet. 1.1.7-9, 1354a31-b22).

   Here Areistotle repudiates both the actual activities of contemporary rhetors and Isocrates' theoretical claim that rhetoric is all-pervasive and thus morally neutral—on the contrary, rhetors should not manipulate their audience. Like Plato, Aristotle wants to prescribe "true" rhetoric—as the non-manipulative presentation of a case to a judging audience. Unlike Plato, Aristotle allows the activities of ordinary judges (not ideal or Socratic ones) to be important and legitimate.

3. He returns to the attack on the technical writers. They concentrate on forensic oratory, because it provides a better opportunity for manipulating the audience, with the result that the jurors "surrender a decision," they do not judge. This is why some states prohibit speaking "outside the subject." Why are juries easily manipulated? Assemblies are deliberating about their own interests while juries are judging what does not concern them. The very detachment of juries from the case—Aristotle suggests—makes them easily swayed when the orator does introduce their interests (and when by doing so he embarks on what is outside the case; Rhet. 1.1.10, 1354b22-55a3)
4. Rhetoric is about conviction (pistis); and conviction is demonstration of a kind—for we are convinced most of all when we accept that the point has been demonstrated.²⁸ This is achieved by the enthymeme, which is the rhetorical equivalent of a dialectical argument and is the most authoritative of the convictions. So just as there are rules for dialectical arguments, there are rules for enthymematic argument; therefore there is,. a technique of rhetoric—

   namely the capacity to see "the truth and what is like it" (that is, for rhetoric no less than for dialectic, experts are able to see when the arguments have followed the rules).²⁹ And after all, we are naturally inclined toward the truth—so the same person can make a stab at³⁰ the truth and at the endoxa.³¹ (Rhet. 1.1. 11, 1355a3-18).

   Here Aristotle claims that rhetoric has at its center a series of argumentative (logical) rules; so, he concludes, there can be experts in rhetoric (which Plato denies), provided they concentrate on what rhetoric is really about—argument (which Isocrates and co. ignore). At this stage, he has turned the other way again, to repudiate the Platonic view that there is no subject matter for rhetoric; the theory of the enthymeme is designed to respond to exactly that difficulty.

5. Aristotle returns to the problem of the technical writers. Clearly they do write about what is outside the subject (that follows from the centrality of the enthymeme). But rhetoric is useful just because we are naturally predisposed in favor of the truth. Rhetoric allows the truth clear expression, without which it might shamefully be defeated. It substitutes for scientific demonstration in cases where the people in the audience are so appalling³² that no manner of scientific knowledge could convince them; and its ability to argue on both sides does not commit us to acting inconsistently but enables us to understand and resist the view of the opposition (Rhet. 1.1.12, 1355al9-33).

   Now rhetoric is persuasion, justified by the natural moral priority of the truth. Here, then, Aristotle seems to prefer Isocratean rhetoric to the austerity of Plato. In order to persuade, rhetors need to know how to argue on both sides of the case. Plato's objection was that thus the rhetor could persuade his audience of anything, whatever the truth of the matter. But Aristotle returns, to rebut this, to his comparison with dialectic. Rhetoric and dialectic are alone in being able to argue on both sides of a case. But this does not imply that each side is indifferent to the truth; rather, what is naturally true and what is naturally better are more easily argued and more convincing. So rhetoric allows us to defend ourselves verbally as well as physically. That it may be used for immoral ends is true; but it is a common feature of many things, that used properly they are of great benefit, while used improperly they do great harm (Rhet. 1.1.12-13, 1355a33-b7).

6. This is the first move in Aristotle's case against the view that rhetoric is manipulative and morally unsound. On the contrary, he argues, rhetoric is the means to protect the natural rightness of truth; the fact that it may be exploited for bad ends is no reason to throw out the whole of the persuasive art. So, Aristotle concludes, it is clear that rhetoric is a general technē like dialectic. What is more, it is valuable, not trivial nor manipulative, because it is there to discover the available possibilities in each case. So rhetoric is interested in what is actually plausible and what appears so,³³ just as dialectic is concerned with syllogisms and apparent syllogisms. Rhetoric differs from sophistic in its intention, not its capacity. (Rhet. 1.1.14, 1355b8-21).
7. In the second chapter Aristotle embarks on a more detailed analysis of the "convictions" and offers his tripartite analysis of the subject matter of rhetoric—the arguments, the speaker, and the audience. He opens with a new definition: "Rhetoric is the power of considering the possible plausibility in each case." For rhetoric differs from the other arts and skills, each of which both teach and persuade about their particular subject matter. Rhetoric, on the contrary, has as its subject matter what is plausible about any topic whatever—and thus it spans the subject matter of all the particular arts (Rhet. 1.2.1, 1355b25-34).
8. Aristotle returns to the notion of "technical convictions," which are to be contrasted with the convictions that are ready to hand (witnesses and tortures etc.). The technical convictions are the ones we supply or discover for ourselves, and they are three: the character of the speaker; the disposition of the audience; and the demonstration, or apparent demonstration, of the case. Of these, the conviction through the character of the speaker is the most authoritative, in those cases where the character of the speaker comes across through the speech itself—for in cases like that our moral inclination to believe the speaker speeds up the process of persuasion. The conviction that comes from the audience occurs when the audience is affected by something in the speech (after all, our reactions are different when we hate or love, are happy or sad); and this is the preoccupation of the technical writers. Convictions produced by the arguments are evidenced when we show a truth or an apparent truth from what is plausible in each case (Rhet. 1.2.2-6, 1355b25-56a20).
9. If rhetoric has indeed this tripartite structure, then it is the task of the rhetor to consider arguments and character and affections, so that rhetoric is an offshoot of dialectic and politics. It puts on the clothing of politics—but it is in fact a part and an analogue of dialectic, not having any particular subject matter other than argument itself (Rhet. 1.2.7, 1356a20-33).

Aristotle returns here to the language of Plato to reinforce the position of Rhetoric 1.1, occupying the middle ground between Plato and Isocrates. The universal nature of rhetoric ensures that the subject matter of the art of rhetoric is argument, not politics—it has a political appearance because these arguments are presented by speakers, to an audience. Rhetorical arguments are contextualized in the political arena. Yet Aristotle continues to suggest that argument, not politics, is the distinguishing mark of rhetoric, and to press the similarity between rhetoric and dialectic as he draws a parallel between the enthymeme and the syllogism, the example and induction: these alone are the means to produce conviction through argument. These arguments are plausible (Rhet. 1.2.8-10, 1356a34-b27).

Plausible, however, means plausible to someone whether directly or mediated by some argument. But if rhetoric is to be a technē, then like dialectic it must concern itself with the general, not the particular—with what people believe, not what appeals to Socrates or to Callias. Yet rhetoric and dialectic differ in their subject matter—dialectic is about what needs explaining, rhetoric about what needs deciding. So rhetoric is about what could be otherwise, what we can affect—and it takes place especially before audiences that cannot cope with long chains of reasoning (Rhet. 1.2.11-12, 1356b28-57a7).

The course of Aristotle's argument in these two chapters is hardly straightforward (so that it is tempting to blame it all on a muddle in the text). In the first chapter Aristotle works by contrasting opposed views of the definientia of rhetoric. There is (my section 1) a difference between rhetorical arguments (the most authoritative persuasions) and the manipulation of the emotions of the audience (by the devices of the technical writers—prologue etc.). Then (2-3) there is a contrast between the proper reaction of the people in the audience (a proper judgment) and their "surrendering" to the orator; this corresponds to cases where their interests have been left untouched, and those where they have been illicitly involved. And finally (4-6) there is the opposition of the moral and the immoral use of rhetorical technique by the speaker, where what is moral sides with the truth, what is immoral with the reverse. All of these contrasts work against the background of a play-off between Plato and Isocrates, as Aristotle gradually refines his position away from either extreme. He argues, that is, that the art of rhetoric is not to be collapsed into dialectic nor allowed to take over all speech. Instead he grants the practice of rhetoric a quite precise description (the enthymeme is its body) and a specific task (to defend the truth against fabricators). In that case, the practice of rhetoric can be formalized—there can be an art of rhetoric (this meets Plato's epistemological objection). At the same time, Aristotle fends off the Isocratean view (that rhetoric is all-pervasive) by showing that the manipulative part of rhetoric (what is off the point) is not a proper part of it; nonetheless, it does have a persuasive aspect (rather than a purely argumentative one) because the circumstances prohibit pure argument in many cases; and our duty to defend the truth positively urges us to use rhetorical techniques in its defense (this meets Plato's ethical objection).

In the second chapter (my 7-9), Aristotle seems to change direction. He continues to press the idea that rhetoric is the counterpart of dialectic because it is about the common features of speech. But his emphasis has altered. While the first chapter talked about argument, the second focuses instead on the business of persuasion (and it is here, no doubt, that the lines of fracture between two versions of the Rhetoric may appear). Indeed, the text as we have it seems to contain a direct inconsistency. In Rhetoric 1.1 the "most authoritative of the convictions" was the enthymeme; in 1.2 it is the character of the speaker; and the program of 1.2 is picked up in the rest of the work, which discusses not only the formality of the enthymeme, but also (with just as heavy emphasis) the character of the speaker and the emotions of the audience (not to mention the analysis of rhetorical or literary style offered in the final book). We may be forced to conclude that the interplay betwen Plato and Isocrates was a feature of the first chapter alone—while the rest of the work reflects a more workmanlike treatment of the rhetorical practices of Aristotle's time—Aristotle's own ars rhetorica purified of Platonic influence.

That, perhaps, is a counsel of despair; particularly because from a Platonic perspective Aristotle's own theory is vulnerable to criticism. Plato's objection to rhetoric was that since, both in theory and in practice, it fails the test for a technē it also fails the ethical (political) test for the proper care for the soul (of its practitioner or its audience or its pupil). Aristotle's first defense against Plato is to reinstate the technical status of rhetoric by giving it a subject matter (rhetorical arguments, especially enthymemes). But he elaborates this position by suggesting, first, that rhetoric also has an ethical function (the defense of the truth) and second that it has (what I have called) political status as well, because it is concerned both with the character of the speaker and the emotions of the audience.³⁴ How far, Plato might ask, are these different aspects of rhetoric unified into a single technē? He might object, first, that while his arguments in favor of philosophy provide a necessary connection between knowledge and happiness (or virtue), Aristotle has merely taken for granted the connection between enthymematic arguments and the natural value of the truth. Plato might complain, second, that unless there is some rigorous account to be given of the relation between rhetorical arguments and the virtue of the speaker then rhetoric is still open to abuse and vulnerable to the charge that if the orators are merely interested in winning the argument, the art of rhetoric just cares for showing them how to do it. Isocrates may argue the opposite case. If rhetoric as it appears in the market place is a relation between persons, and Aristotle insists that it is a system of formalized argument, has he saved the phenomena at all?

We may put this point a different way. As a result of the opposed arguments of Plato and Isocrates, Aristotle has isolated two aspects of the art of rhetoric: the argumentative one, in which it resembles dialectic, and the "persuasive" or sociological one, in which it resembles the activities of the technical writers. How far are these aspects of Aristotelian rhetoric properly connected? On the one hand, Aristotle may (disappointingly) simply have declared that these are two aspects of the same thing (rather than two quite separate and different notions of rhetoric); he saves the phenomena, then, by fiat and not by argument—and is himself the victim of the opposed arguments. On the other hand, he may be arguing that rhetoric is an organic unity, which displays these two aspects, and that both Plato and Isocrates were wrong, one-sided in their views. Persuasion and argument, in this view, are naturally one thing. But then he badly needs to explain what that unity would be. Unless Aristotle can show that there is some essential connection between rhetorical arguments, the character of the speaker, and the emotions of the audience, then his theory of rhetoric may just be a gluing together of his predecessors' views (just as the text itself may be badly joined) without any proper theoretical unity of its own. And this, of course, is exactly the challenge that Aristotle has set himself, by his strategy of presenting the opposed arguments of the wise.

In Rhet. 1.1 Aristotle moved from the point about rhetoric being a technē, through the charge that it is manipulative, to conclude that the method of enthymeme is thoroughly respectable. At each stage he resisted both the Platonic and the Isocratean account, because he insisted on a clear differentiation between rhetoric and dialectic; and at each stage he suggested that the proper ("inside the matter") methods employed by the orator are technical and susceptible to analysis. His distinction between "inside" and "outside" the matter is clearly crucial to his position; this allows him to reject the view that all communication is legitimate rhetoric, while continuing to assert that there is a technical core. The problem arises when we try to determine how much more than argument proper lies "inside" the subject of rhetoric. After all, the remainder of the Rhetoric deals not only with rhetorical argument forms, but also with the character of the speaker and the affection of the audience—if Aristotle's own work exemplifies his own definition of rhetoric, we should expect the definition to contain all three aspects. So from the outset, he needs to show that the context of rhetorical argument is as much one of its definientia as the arguments themselves. And he must explain then how rhetoric turns out to be a unified whole, rather than an accidental association of argument and context, and consequently how his own theory of rhetoric is not merely an uncomfortable amalgam of Plato with Isocrates, but a coherent unity.

He might begin with the old Platonic metaphor. Suppose our rhetor is standing up in a (well-constituted) court and hoping to persuade the jury of the innocence of his client. Over and over again Aristotle points out that the enthymeme alone is not enough to do that; and often he concedes that the enthymeme is indeed unsuitable for such a task. What then determines what is legitimate in such cases? The answer to that still must be "the enthymeme itself because it is the body of conviction. Suppose that the rhetor's argument can be understood in terms of an enthymeme (or a paradeigma)—even if it is not fully expressed that way—then what he says may be counted legitimate. If, on the other hand, there is no central argument holding his speech together, then what he says should be ruled out of court. If there is an argument beneath what he says, than anything that represents that argument to the audience is licit; whatever is not organically related to that argument is to be discounted as sheer manipulation. The metaphor of a body and its clothes is fundamental—enthymeme is the body of conviction; the reasonable use of ēthos and pathos will be genuine natural parts of the rhetorical episode, cohering by virtue of their relation to its body. Conversely, manipulative appeals to the emotions will be separable from the argumentative body, the clothes it can take off and do without. So what is licit and what is not are determined simply by whether they are related to, naturally connected with, or relevant to the enthymeme that lies at the heart of the episode (even if, as may happen in cases of particularly inadequate audiences, the enthymeme is not formally expressed). The enthymeme, then, is what holds the whole thing together, the formal (formalizable) argument at the heart of the speech. The context (the dispositions or whatever of the persons involved) is whatever is necessary to and inseparable from the central argument—all the rest is superfluous cosmetics and should be illegal.

Can this metaphor be supported by argument? I shall suggest that it can, for Aristotle's account of the enthymeme is such as to locate it firmly in the context, first of value (the interests of the various parties) and then in the twin issues of character and emotion; this takes up the discussion of the first two books of the Rhetoric.

Rhetorical Arguments

Aristotle's first defense of rhetoric against Plato is that there are rhetorical arguments that can be formalized; this then legitimates both the practice and the art of rhetoric. Rhetoric's arguments are enthymemes.

Now Aristotle's solution to the Platonic problem must meet two conditions: First, he must show that rhetoric's arguments are indeed specific to rhetoric (and not dialectic in disguise) but still general (so that rhetoric does not collapse into, say, physics). Second, he must demonstrate that their specific features are an intrinsic feature of the arguments themselves (an argumentative or logical feature), rather than part of the way they are presented. After all, if the enthymeme is just a dialectical argument presented in a persuasive way, there would after all be no special subject matter for rhetoric, only a special attitude to the subject matter of dialectic. So Aristotle needs an account of the enthymeme that is logical or epistemological, not psychological or rhetorical.

Accordingly, Aristotle contrasts two modes of argument—the arguments with which dialectic deals, and enthymemes. The former are about what is generally true; the latter are about pisteis (e.g., Rhet. 1.1, 1354al3,15, b21; 1355a3,4,5,7,28; 1.2, 1356al, b6); about what is pithanon (e.g., 1.1, 1355a38; 1.2, 1355bll,15, 26, 32; 1356a20); what is endoxon (e.g., 1.1, 1355al7; 1.2, 1356b33); what is endechomenon (e.g., 1.2, 1355bl3, 26; 1357al4, 24); what is eikos (e.g., 1.2, 1357a34). Several of these expressions are common to earlier discussions of rhetoric (e.g., pithanon, Grg. 456c, Phdr. 269c; eikos, Phdr. 267a ff.); but in those contexts they emphasize persuasion—as indeed do the translations we are inclined to use—"conviction," "plausible," "generally believed," "likely" (endechomenon is an important exception, "contingent"). How can persuasion be a condition for formal argument in the Aristotelian mode?

Consider the phenomena once again. Rhetoric happens in two special places, the court and the assembly. There what is at stake is not the eternal verities (as is true of scientific knowledge) but either individual matters of fact ("did he do it?") or general matters of practical policy ("should we exact a poll tax?"). Aristotle recognizes both these special features of rhetoric (cf., e.g., the problems of corruption in a jury trial, Rhet. 1.1.10, 1354b23ff.).

First, he argues that the reason for having judges and juries is that particular matters of fact cannot be decided by the legislator, who deals only in generalities. In the ideal situation, therefore, the judge is concerned with deciding (as little as possible, but) whether it happened or not, is or is not the case, will or will not be a consequence (Rhet. 1.1.7, 1354a31-bl6). Now these decisions are not taken at random, but with thought; but because they involve particular premises ("this is the man whose fingerprints are on the blunt instrument"), they are not susceptible to demonstration in the full sense of the word—they are not amenable to scientific understanding (whatever the state of the fingerprinting art).

Second, enthymemes are not about what is necessary, but about what is "for the most part," or what could be otherwise (e.g., Rhet. 1.2.14, 1357a22ff.). Why? Aristotle makes the same point about the subject matter of ethics.³⁵ If we are going to decide what to do, there is no point in deciding to do something that cannot be affected by us (Rhet. 1.2.12, 1357a4). Practical reasoning is not about the shape of the earth, but about how to get to the marketplace and why. Political reasoning is about practical reasoning, generally understood—so it is about what we can effect as a matter of general policy (poll tax for everyone, free school milk, regular symposia).³⁶ The conclusions of such reasoning will be general, but not necessary (some people may refuse to pay their poll tax, the last child in the queue may not get any milk, the symposia may be ruthlessly terminated). That they are "capable of being otherwise" does not make them irrational, but it does make them improper subjects for scientific reasoning (the rhetor's subject matter of rhetoric differs from the physicist's, insofar as physics is about the general rules that may have exceptions; rhetoric is especially about the things that could be otherwise—cf. Rhet. 1.4.1-3, 1359a35ff.).

If rhetoric is what happens in the lawcourts and in the assembly,³⁷ then rhetoric always involves either what is particular or what is general but could be otherwise (compare the discussion of equity at Rhet. 1.13.13, 1374a29ff.). How far are either of those two matters amenable to formal argument?

Particular matters of fact, we might easily concede, are objective, they happened, and the propositions describing them are objectively and straightforwardly true. They can be assimilated to argument in one of two ways—either by existential quantification, or by treatment as examples of a general law—and thus they can figure either as propositions in a deductive argument or as steps in an induction. But in neither case do they provide a very sound basis for any general claim (Aristotle dismisses their contribution to argument at Rhet. 1.2.18, 1357bl3); they cannot found a necessary proposition³⁸—at worst a contingent one (the existential quantification), at best a "for the most part" (a general law derived from induction). Individually, they are not enough for either induction or deduction proper; instead, they partake of the rhetorical forms enthymeme and example (Rhet. 1.2.10, 1356b21ff.)

General issues of policy may not fit easily into argumentative form for a quite different reason—here arguments contain not only matters of fact but matters of opinion and evaluations of their use or general desirability. That, at any rate, may be our view of value judgments—and it may also have been Isocrates'. But Aristotle's view, it seems, was rather different. His analysis of value (the good) in the Nicomachean Ethics³⁹ starts from a metaphysical point about essence and function, and a linguistic point about the performance of the word good (it is syncategorematic). Value, in Aristotle's view, is an expression of the essence or form of something; it is defined in terms of good functioning and not determined by the eye of the beholder or the view of the valuer.⁴⁰ In that case, it is perfectly possible to have general, objective value judgments operating as the premises in an argument (compare the discussion of natural justice, Rhet. 1.13.2-3, 1373b4ff.).

Yet, because value reflects what we pursue and avoid, such judgments must represent things we can choose to pursue or avoid; so even if they are general, they cannot be necessary, for otherwise there would be no point in deliberating about them at all (but that is what rhetoric must do, Rhet. 1.2.12, 1357al). Moreover, value is a matter of dispute (as Plato had observed, Euthyphro 4eff. and as Aristotle agrees—compare the decision procedures offered in Rhetoric 1.7). Hence deliberative rhetoric, because it is about value, may well be disputatious (Rhet. 1.3.5-6, 1358b20ff.); this may be the source of disagreement and opposed arguments (cf. Rhet. 1.6.17, 1362b29ff.). That value is a matter of dispute may lead us to suppose that it is in some sense relative; but this does not make value merely subjective. Consider two sorts of evaluative relativity: i) Helen may be beautiful relative to Quasimodo, ugly in comparison to Aphrodite; nonetheless, Helen's beauty may still be an intrinsic, natural feature of Helen, not something bestowed upon her by the onlooker. Relativity of value does not imply subjectivity. ii) Seawater may be good for fishes and poisonous for humans. This does not imply that its goodness is in the eye of the fish and so subjective, but rather that its goodness is relative to creatures with gills: value may be objective and still relative.

Can these two different sorts of proposition—the statement of particular fact and the value judgment—be incorporated into rhetorical logic? Aristotle seems to think so (Rhet. 1.2.14-15, 1357a30-b3 and a more systematic view at 1.9.40, 1368a25ff.; cf. 2.25.8-10, 1402b13ff.)

The particular statement is a "sign," a sēmeion; the general (but not necessary) proposition is a "likelihood," an eikos. Both may figure as the premises of enthymemes (Rhet. 2.25.8, 1402b13ff.); and that seems to be the distinctive feature of enthymematic argument (Rhet. 1.3.7, 1359a6-10). Some of these arguments could, perhaps, be formalized and set out in the same way as is the syllogism (cf. Rhet. 1.2.13, 1357al5).⁴¹ However, the modality of the propositions in rhetoric must be different from the modality of demonstration. For the propositions in demonstration are necessary, while rhetoric needs a modality of possibility or contingency (Rhet. 1.3.8, 1359allff.), to endechomenon⁴² (Rhet. 1.12.2, 1372a9ff., where this figures in the calculations of the criminal; 2.18.3, 1391b27ff.).

Now contingency and possibility are defined not in subjective terms (as what is plausible to someone or other) but in objective terms (what can happen but may not; what does follow but may not, 1.2.18, 1357bl9; 1.4.1, 1359a31ff. "Did he do it?" "Is this law worth the paper it's written on?" "Can I get away with it?" "probably"; "probably not.").⁴³ In that case, the logic of rhetoric is about propositions that are likely (eikos, Rhet. 1.2.15, 1357a34)—that is, likely to be true objectively.

Has Aristotle shown that the art of rhetoric is a techne in its own right, not to be collapsed into any other, whether general or specific? Rhetoric concerns ethics, politics, and jurisprudence; but the art of rhetoric is none of them, because it is about argument, not directly about value or justice or expediency. In that way it is general, like dialectic (and like nothing else). But dialectic is about things that need explaining, rhetoric is about things that need deciding.⁴⁴ So its generality is determined by its modality—what is contingent, possible, or even probable.⁴⁵ Rhetoric's treatment of contingency, however, differs from the way in which physics, for example, deals with the exceptions to the rules (what happens "for the most part").⁴⁶ In physics the exceptions are incorporated into the rules; they are explained by virtue of what happens generally, regularly, as a matter of natural course. In ethics, however, and so in rhetoric, "for the most part" points to the way in which the world can be affected by our action (cf, e.g., Rhet. 1.4.3, 1359a35ff.; 2.21.2, 1394a25ff.). In ethics possibility is about what could be otherwise; in physics things could be otherwise, but they occur according to natural regularities that are not otherwise. So while dialectic may deal with propositions that are universal or accidental, rhetoric deals with what is contingent per se.⁴⁷

"Rhetoric is antistrophos to dialectic"—Aristotle's programmatic remark should not be taken lightly. On the contrary, his opposition to the Platonic view of rhetoric (antistrophos to cookery) reflects his view of the real world. Rhetoric happens, so it cannot be denied; and it does have an argumentative center, governed by the modality of contingency. If any sense can be made of the modality of contingency or of Aristotle's version of it, then Aristotle can argue that the practice of rhetoric can be systematized into a technē; and he has his first point against Plato. What is more, his account of natural value allows him to suppose, against Isocrates, that the evaluative fabric of the world exists independently of us, prior to speech and understanding. In that case there is a difference between the objective core of rhetoric and communication in general; and Aristotle may deny that rhetoric is all-pervasive. On both grounds he is entitled to claim that there is after all an art of rhetoric.

Souls and Enthymemes

So Aristotle may resist an lsocratean account of rhetoric as (any) communication by insisting on an austere description of the logic of the enthymeme. He may also resist Plato's denial that the practice of rhetoric is systematic. But in the process he may lay himself open to Plato's ethical complaints—in a rather unexpected way.

For Plato, I suggested, arguments take place in souls (logic and ethics are thus inextricably linked). Aristotle, by contrast, has a thoroughly austere view of argument (Top. 1. 1, 100a25ff). Arguments are collections of propositions,⁴⁸ where the premises are causally related to the conclusion. Different sorts of premise generate different sorts of conclusion—the premises may be quantified differently, or modally qualified. But all arguments are significantly similar in one important respect—that the premises generate the conclusion.⁴⁹ Why?

Take the primary case—of scientific understanding. Here the premises generate the conclusion because they explain why the conclusion is true. What is more, in grasping the argument we understand why the conclusion is so. The explanation, as Aristotle puts it, makes the understanding (A Po. 1.2, 71b20-25). To cite causes is to explain, to get an explanation is to understand.⁵⁰ But causal relations are real relations in the world; so understanding comes from real states of affairs in the world (especially from essential features of the world). If p explains q, the reason is that the state of affairs that p describes is a "cause" (in the special Aristotelian sense, a "because") of q. Our understanding of that is a direct consequence of the argument that represents the facts of the matter—so the cognitive state (understanding) follows from the real state of affairs.⁵¹ Let us say that arguments are transitive—the facts are transmitted through the arguments to the mind that receives them.⁵²

On this view of the metaphysics of understanding, the mind's role is largely a passive one. Aristotle's account of the logical relations between propositions confirms this impression. His syllogistic system is based on the insight that validity can be determined by inspection in the paradigm cases.⁵³ An apodeictic syllogism in Barbara is obviously valid, because of the structure of the syllogism and the operators on the propositions it contains. But if validity is (sometimes) obvious, then validity is one of the phenomena that Aristotle is committed to save. It is an actual, objective relation between propositions—not something that we impose on argument or make up as we go along. Validity is a matter of logical fact, not a hypothesis of theory—and certainly not a piece of ideology.

Aristotle makes mileage from that. For if validity (a relation between propositions of particular types) is obvious, then formalization of arguments is possible; our systematic understanding may then (if argument is transitive) follow. But to understand arguments is then a very different thing from inventing them. Arguments just are valid; we see them as valid because they are so, not the other way about; and so we also see logical system because it is actually there—system is not imposed by the minds of those who understand it. Once again, then, and for logical reasons, arguments cause understanding; and the understander plays a relatively passive role. The same psychological account may be true for the enthymeme. Thus we are convinced when we take something to have been demonstrated (Rhet. 1.1. 1355a5); so we are primarily convinced by argument (not by the character of the speaker or by our own emotions). Conviction follows from demonstration⁵⁴—rhetorical arguments are transitive."⁵⁵

In that case, conviction is something that happens to us in the face of a rhetorical argument—we directly receive the argument itself (we perceive it, cf. Rhet. 2.23. 30, 1400b30ff.) Being convinced, on that view, is not like being persuaded ("Will I?" "Won't I?"); for being convinced does not reflect some decision procedure on our part. Nor is being convinced like feeling an emotion—for it is not a feeling at all but would better be understood as a purely cognitive mental event. Hence "conviction is a demonstration" (Rhet. 1.1.11, 1355a5), produced by the two sorts of rhetorical argument, paradeigma and enthymeme (1.2.8, 1356b6). So what happens—if rhetorical arguments are indeed transitive in this way—is that an argument about possibility can then be described as plausible to its audience. Thus rhetoric deals with what is possible and what is plausible (to pithanon). Plausibility turns up both as a general possibility (1.2.1, 1355b26) and as a particular issue (1.2.6, 1356a20); but in either case, what is genuinely plausible (cf. 3.1.3, 1403b19) should be distinguished from what merely appears so (1.1.14, 1355bl5).⁵⁶ So the job of rhetoric is to "see the existing plausibilities in each case" (1.1.14, 1355blO)—just as in ordinary logic validity is a formal feature of the argument itself.⁵⁷

But still what is plausible is plausible to someone (Rhet. 1.2.11, 1356b28). If we understand Aristotle's account of argument in the austere way I have described (arguments are transitive), then both the speaker of a rhetorical argument and its audience will be insignificant in the definition of what rhetoric is. The delivery of rhetorical arguments will be a series of bare events that happen to both speaker and hearer—and there will be nothing left to the rhetor's insistence that rhetoric is about communication between persons. There will be no agents in rhetoric, and no acts—merely events and things.⁵⁸ Plausibility will collapse into objective possibility, and persuasion will simply be the result of a rhetorical argument's happening to its audience.

If that were Aristotle's position, it would be an uncomfortable one. First of all, it would be hard to understand how, or why, arguments from contingency would be constructed at all. Unlike pure syllogisms, they cannot be purely formal, because they are embedded in the possibilities that interest us. So while a dialectician may care about formality for argument's sake, our interest in contingency comes from the hope, or the fear, that things may indeed turn out that way. Enthymemes, that is, cannot be formalised away from their context in the easy way that syllogisms can. To give a Platonic objection: Aristotle has, if he offers a theory of bare argument to explain rhetoric, ignored the ethical dimension of Plato's charge against the rhetors and the sophists.

Second, if the audience of an enthymeme, by hearing it, simply suffers it then Aristotle's account of the art of rhetoric will be vulnerable to Plato's charge that rhetoric exercises force majeure over its audience. And yet this conclusion is clearly one that Aristotle wants to avoid—consider, for example, the way he contrasts the passivity of a manipulated jury (they "surrender a verdict") with the activity of proper judgment.

Third, I have suggested that Aristotle treats plausibility as equivalent to possibility. But that—given his insistence that what is plausible is plausible to someone—surely will not do. Possibility may be construed (and formalized) in terms of actual and possible events in the world. Plausibility, however, is about how events or things or persons or arguments appear to us; plausibility must have an audience. Bare enthymemes (again on the transitive view of argument) have no audience—only a recipient, and any recipient would be as good as another. If rhetoric is genuinely about what is plausible, there must be some real role for the audience to play. Isocrates was not so far wrong.

The opening of the Rhetoric makes some play with the judicial proprieties. The technical writers go wrong because they allow what is beside the point, while what is needed in court is a proper judgment, not a surrendered verdict. It is the notion of "judgment" that bears the weight of Aristotle's defense against Plato's ethical objection; and it allows him to distinguish, against Isocrates, between cases where the people in the audience are properly involved in rhetorical argument and cases where they are manipulated.

Consider three examples of "judgment" from outside the Rhetoric.

1. The defense of the law of noncontradiction (Metaphysics 4.3-8). One sophistic strategy to demonstrate that truth is radically relative is the use of opposed arguments: whatever we say, we end up contradicting ourselves. For an extreme relativist of this sort, the "opposed arguments" manoeuvre shows not only that truth is radically relative, but also that consistency does not matter and that contradictions may both be true. This person might be Protagoras—and it could also be an extreme Isocratean, who goes for cultural relativism. Against Protagoras Aristotle insists that the physical world out there "comes first"; that it is stable, real and accessible to us.⁵⁹ But the dispute is not (as Aristotle himself realizes) easily resolved—because neither party concedes the basis of the other's argument. Aristotle's strategy is to demonstrate his position (that the law of noncontradiction holds) by refutation of his opponent (let him only "signify something" and he is done for). But his own position remains vulnerable to a relativist complaint: even if the real world out there is ever-present and stable, how is it ever-present to us? How is it that we have reliable access to what is real or true and are not at the mercy of relativism anyway, whatever the facts of the matter may be? Here Aristotle appeals to a different notion—the "proper judge," the need to find a correct judge for any question (Metaph. 4.6, 101la5-6). If consistency matters, and the law of noncontradiction is true, the reason is that there are facts of the matter (things); another reason is that the determinate nature of the world is reflected in what we say and think. There must be some way of resolving what we say and think in favor of the facts of the matter; and that resolution is a matter of judgment. Now this account repudiates hopeless relativism (and its cousin subjectivism) by continuing to insist on the primacy of objective truth; but it reassures us also that there is some active part to be played by the judge.
2. Dialectical progress. At Metaphysics 3.1, 995b3 we resolve opposed arguments by judging between the opponents. The imagery of the passage emphasizes first of all the effect the opposed arguments have on the judge. He starts out tied and trapped by the opposed arguments; only familiarity and practice allow him to untie the knot and escape; to begin with, the arguments are in control. Once again, arguments are thought of as obvious—so the dialectician deals with his perplexity by seeing all the difficulties; only then is the terminus of his search obvious to him. Similar metaphors of perception turn up throughout the Rhetoric to describe the activities of the judge (e.g., 1.1.7, 1354b10-11) who is, of course, also a spectator (1.3.1, 1358bl-2; 2.1.1, 1377b20ff.; 2.18.1, 1391bIO). And yet the release of the aporia in the Metaphysics describes a process of gaining control over opposed arguments—the judge (not the arguments) are winning by the time the knot is untied. Once again, the notion of judgment shows how the audience is crucial to the argument, and an active participant in the process.
3. Meno's paradox. Compare the way Aristotle handles some old epistemological chestnuts in the last chapter of the Posterior Analytics (2.19, 99b17ff.). How does the mind ever move from knowing nothing (no universal) to knowing something? Does it have ideas all along? Or does it acquire them somehow—if so, how? Plato said we remember from before—Aristotle says we perceive now. The universal is established in the mind when sense-impressions are retained in the memory and reinforced by repetition. So the mind is affected by the world, and universals are thereby acquired. Is the mind then totally passive? Not quite, Aristotle suggests, for it can judge. Judgment is the capacity of the mind to think universals (to identify one and discriminate it from another). Once again, the judge is in control.

Aristotle dovetails his metaphor of judgment elsewhere with the theory of the Rhetoric. Judgment is something that is done by the receptive mind; it is something that people in an audience can produce when the argument is properly presented to them, and they are dispassionately disposed. In the ideal case, therefore, the formal arguments of rhetoric must be contextualized by being assessed by the audience. Now this does not mean that any audience will do; on the contrary, Aristotle distinguishes between cases when there has been proper judgment and those where the audience has been improperly manipulated; and thus he resists the temptations of a universal Isocratean rhetoric. We might understand this in judicial terms—Aristotle wants to limit as far as possible the discretion of a judge or jury. In the ideal case, they should be concerned solely with what did or did not happen, is or is not true, will or will not occur. Judges should decide particular areas of factual doubt, then, but nothing else. So to judge the truth is to grasp or to have access to the facts. Judging is to be rid of extraneous influence; otherwise it is not judgment at all—but just handing over a verdict (Rhet. 1.1.10, 1354b34-55al). And in the latter case, the judge has been unwarrantably affected by the art of persuasion.

How far does this respond to the ethical objection Plato launched against rhetoric (as opposed, say, to some objection taken from the philosophy of mind)? Recall that opposed arguments may seem to be in equipoise—but in fact the balance is tipped in favor of natural truth (Rhet. 1.1.12, 1355a37). So practice in dealing with opposed arguments is a way of coming to see clearly what is true and what is false (Top. 1.2, 101a35). And to defend the truth, of course, is something to which we are naturally fitted (Rhet. 1.1.11-12, 1355a3-33)—and thus something that it is appropriate for us to do.

Aristotle introduces his long discussion of ēthos and pathos thus: "Since the art of rhetoric is for the sake of judgment—for people judge deliberative matters and judicial proceedings are a judgment—we should not only look to the argument, how it can best be demonstrative and trustworthy, but also that the speaker appears in a particular way, and the judge is affected in a particular way" (Rhet. 2.1.2, 1377b20-24). Here he shifts from the formal view of rhetorical arguments to pay attention to the act of judgment, and thus to the disposition of the audience and the character of the speaker. Judgment is now not only about argument, because it also concerns the appearances.⁶⁰ That is, it concerns how things appear to us—what is plausible is plausible to us. Even if rhetorical arguments can be formalized, they are arguments spoken by someone to someone else; they are contextualized. Always? Yes—the formalization of dialectical arguments can be context-free, but the discussion of what we can change, what we can avoid or pursue, what actually happened on a particular dark night always carries a reference to persons and context; the discussion of _thos and pathos brings that personal feature to the fore.

Arguments, if purely formal, would be transitive and their audiences passively affected by them. But judgments involve the active participation of the hearer—and hence the importance of the metaphor of judgment in other metaphysical or epistemological contexts. The proper exercise of rhetoric is constrained by the proper mental disposition of its audience. So the definition of rhetoric is not context-free: instead it demands the proper cognitive disposition of its audience. And that disposition is naturally inclined toward the truth. So proper judgment and ethical propriety go hand in hand.

Arguments in Context

The Rhetoric contains long discussions of the social context in which rhetorical arguments may appear. Thus Book 2 discusses at length the way in which the character of the speaker can influence the audience, or how the audience's emotions can be exploited for persuasive ends. And Book 3 shifts to a handbook of rhetorical technique: how to adopt a weighty style, how to use metaphor, how to avoid (or exploit) prejudice. With reference to the work as a whole, then, it is hard to argue that Aristotle completely replaces the phenomena of rhetoric with his formal analysis of rhetorical argument. Instead, he integrates the enthymeme into its context without relying merely on the metaphor "the enthymeme is the body of conviction." In justifying the claim that rhetorical arguments are essentially contextualized, he meets Plato's political objection against rhetorical practice (that it is only superficially interpersonal, never "soul-to-soul"); at the same time he makes some concessions to Isocrates' sociological account without conceding that rhetoric is all-pervasive.

Aristotle argues in three stages: first, he elaborates on the evaluative features of the enthymeme; second, he explains how these features are inextricably bound up with the disposition of the audience; and third, he shows that because emotions are attitudes toward persons, then the character of the speaker is an essential feature of the rhetorical occasion. The development of the claim that rhetorical arguments are contextualized by their very nature takes up the first two books of the work. But it depends on the initial characterization of the enthymeme—so that the Rhetoric turns out to be well-formed after all.

The art of rhetoric concerns itself with what is possible just because the practice of rhetoric is about actions and decisions; other sorts of possibilities hold no interest for us. In court, for example, we may wonder whether he did bash his wife with a blunt instrument; in the agora we may debate the question whether we should pull down all the theatres to make way for schools of philosophy. In either case we are talking about actions to be pursued or avoided—in either case we are interested in value. So the possibilities discussed in enthymemes are not dull (could this stone crack before it is worn away?) but interesting ones (could we have caviar and champagne for supper?). And if these possibilities are interesting, they are interesting to someone.

Aristotle outlines his theory of rhetoric in Rhetoric 1. 1-3 and then turns to the detailed discussion of deliberative rhetoric (which is, after all, the central feature of rhetoric in classical Athens). He starts with a complex discussion of "the good and bad things" with which the deliberative rhetor is concerned—these are limited to things over which we may have some control (Rhet. 1.4.1, 1359a30ff.). They are specified—they are "those things which by nature can be referred to us and which have the beginning of their generation in us" (Rhet. 1.4.3, 1359a38-39; compare a similar point in EN 3.3.7, 1112a30). Value, that is, is interesting because it is interesting to us. Hence in the case of practical reasoning I deliberate with myself about what is up to me; and I act when what is good or apparently good for me seems to be possible. The case of public deliberation is analogous; this is a matter of what is possible and what is interesting to the persons concerned. So the entire discussion of value in Rhetoric 1.4-9, including the discussion of virtue and vice (Rhet. 1.9.1-34), is predicated on the view that these values are relative to persons and that the rhetor aims to discuss what is valuable to the particular persons in front of him: "Speaking generally, there is a target both for individuals and for men acting together; this is what they aim at when they decide what to choose or avoid; and this, briefly, is eudaimonia and its parts. So for the sake of an example let us grasp what eudaimonia is, and what are its parts composed of; for all exhortation and dissuasion are about this and the things which conduce to it and their opposites" (Rhet. 1.5.2, 1360b4ff.).

But then the discussion of value and the discussion of what is possible are inextricably linked (Rhet. 1.4. -4, 1359a30ff.). Enthymemes, therefore, are not bare arguments at all; instead they have an ethical component all along, because they concern value, and value is relative to persons (to reemphasize the point, this does not imply that they are subjective; Aristotle does not fall into the sophistic trap here).

Where people have an interest, their interest is affected by their emotions. For example, "people choose to do the things just mentioned, and they choose to do bad things to their enemies and good things to their friends where they can" (Rhet. 1.6.26, 1363a20-21). Emotions are first of all dispositions toward our interests. Anger, for example, may be caused by an insult. If someone says to me, "You fool!" my pride, my amour propre or my timē (if I am a Greek) is affected and my interests have been damaged. Likewise anger may prompt me to action: "And some pleasure goes along with all anger, namely the anticipation of revenge; for it is pleasant to think that one may reach what one aims at, while no-one aims at what appears impossible, and the angry man aims at what is possible to him" (Rhet. 2.2.2, 1378blff.). So if the enthymeme is necessarily about what interests me (and is possible), it will be about what affects me; in that case it will only be effective if I am properly disposed—and that is a matter of the proper arrangement of my emotions. So not all emotion is provoked by force majeure—not all cases of emotion are cases where we are "swayed by our emotions." For example, we may be moved to pity by some apparently deserving person—and then change our minds when it is explained to us that he is in fact a thoroughly bad lot and worth no pity at all (Rhet. 2.9.16, 1387b16ff.).⁶¹ Second, and most important, emotions do not preclude active judgment; Aristotle denies the audience of the rhetor must be mere patients of his rhetoric, even if their emotions are involved. "The emotions are the things through which people change their judgments, and which are accompanied by pleasure and pain" (Rhet. 2.1.8, 1378a19ff.)—this does not mean that any decision that is made in a state of emotion is a faulty decision. But the function of the people in the audience, as he insists from the begining of the work, is to judge, to come to a decision—in this manner, Aristotle argues, they are actively involved and (to meet the Platonic objection) suitable moral agents.

Emotions are determined by three factors: the disposition, for example, of being angry; the occasions for anger; and the persons at whom the anger is directed (Rhet. 2.1.8, 1378a19ff.). Throughout Rhetoric 2.1-11 Aristotle makes it clear that emotion is an interpersonal affair—the emotions we feel, even in a legitimate way, to the proper degree and without undue pressure from a clever speaker, by their very nature involve us with others—we feel pity toward a victim, anger toward a criminal, fear toward an enemy, goodwill toward a friend. But then if the emotions are what relate us to others, not only are the emotions central to establishing the context of a rhetorical occasion; they are also fundamental to creating what I have called the political element in rhetoric—the element that shows that we are genuinely involved here with the others who share our community with us. So the discussion of the emotions is the complement of the discussion of the virtues of character, just because emotions are felt toward others as we perceive them; so how they represent themselves to us provokes our emotions, either properly or otherwise.

Plato, it will be recalled, claimed that only "real" rhetoric was clearly interested in the souls of the audience; for only rhetoric based on the high standards of objective knowledge he demanded could satisfactorily interact with the souls of others—only "real" rhetoric met the ideal demands of the political situation. But Aristotle has established against Plato that there is some system to the everyday practice of rhetoric (and thus some possibility of an art of rhetoric); and he has shown (I have argued) that the audiences of, say, the deliberative rhetor are not mere pawns in his nasty sophistic game but active moral agents to be taken with due seriousness. This aspect of the audience has its counterpart in the discussions of the character of the speaker himself. For the discussion of the rhetor's character (initiated in the discussion of value and virtue in Rhetoric 1.9) is developed in Aristotle's account of how the people in the audience believe the rhetor to be well-disposed.⁶²

Since rhetoric is about judgment … we should look not only to the argument and how that may be demonstrative and convincing, but also to how the speaker himself may be presented as of a particular character and how the judge may be disposed. For it makes a great difference when it comes to conviction, and particularly in deliberative rhetoric … that the speaker should seem to be of a particlar character and that the audience should suppose him to be disposed in a certain way towards them. (Rhet. 2.1.3, 1377b20-27)

This condition on rhetoric is elaborated in the detailed account of the virtues of character (1.9; 2.12-17). For first of all the apparent character of the rhetor is persuasive—if we are confronted with someone who appears to be of goodwill, we are likely to trust this person's recommendations as being genuinely in our interests. Correspondingly, we should expect the true rhetor actually to be a man of goodwill, genuinely offering advice that will benefit his audience. On the whole, the characters are described as they would appear to an audience; but Aristotle's point is not so much a contrast between appearance and-reality (that is what Plato latches on to). Instead Aristotle wants to insist on the intrinsic relations between the values that rhetorical argument concerns itself with, the emotions of the audience, and the character of the speaker (cf. Rhet. 1.9.1, 1366a23ff.). And his account of this last feature suggests that virtue and character are to be understood not absolutely but relatively to others; after all, it is by virtue of character "that we can make both ourselves and others trustworthy in respect of virtue" (Rhet. 1.9.1, 1366a27-28). The character of the speaker is therefore the third feature of the context in which rhetorical arguments appear; and it is a feature determined by the other two.

Rhetorical arguments, therefore, necessarily appear in a social context; and Plato's political objection to rhetoric, Aristotle can argue, fails. The rhetor has a system of arguments to practice, and these arguments are about possibility and value. If they are about value, they are relative to the beneficiaries of the value; if they are relative to the beneficiaries, then they also reflect the benefactors. All rhetoric, then, is contextual. But is, then, all contextualized speech rhetoric? Has Aristotle defeated Plato only to fall prey to Isocrates' all-pervasive rhetoric? Once again, he resists the relativist view. For, as the program of the opening of the Rhetoric makes clear, true rhetoric, the rhetoric that is "inside the matter," is focused on the enthymeme; arguments that lack that cohesive feature are not arguments at all but mere attempts at manipulation. But arguments focused on the enthymeme are not therefore devoid of ethical or political content. On the contrary, both the systematic nature and the subject matter of the enthymeme ensure first that these arguments are naturally best suited to presenting the truth, and second that the character of the speaker and the emotional disposition of the audience matter. In this way, Plato's objections to rhetoric are finally silenced, and Aristotle resists the Isocratean takeover.

So both Plato (who denied that Isocratean rhetoric was anything at all) and Isocrates (who said that rhetoric was all-pervasive) are wrong; or (to put the matter in an Aristotelian manner) they are in some ways right and in some ways wrong. Now one view might be that this compromise is effected by ambiguity. We might generalize and say that rhetoric is about "what is likely." Aristotle has shown that rhetorical arguments are about contingency, about what is in fact likely; so they concern to eikos in the sense of "what happens for the most part" (Rhet. 1.2.15, 1357a34). He has also argued that rhetorical devices are about persuasion, about what seems likely to the audience; so they concern to eikos in the sense of "what seems plausible" (whatever the facts of the matter may be) (e.g., Rhet. 1.15.17, 1376a17ff.). Rhetorical arguments aim at ta endoxa, the opinions of the many and the wise which are inclined toward natural truth (Rhet. 1.1.11, 1355a14ff.). Rhetorical devices exploit what people commonly believe—ta endoxa (e.g., Rhet. 2.25.2-3, 1402a34ff.)—or even what they find paradoxical (Rhet. 2.23.22, 1400a5ff.). We are convinced by rhetorical arguments because they are (in a loose sense) demonstrative (Rhet. 1.1.11, 1355a5ff.). We are convinced by rhetorical devices because they exploit ethos and pathos (Rhet. 1.2.4, 1356a4ff.). So "what is likely" may represent "what is possible" or "what is plausible to someone"—but the two senses are not necessarily connected. Is this ambiguity between possibility and plausibility hopeless?

Surely not. As so often elsewhere, Aristotle explains the wide scope of a word (or a collection of ideas) in terms of a central core of meaning. At the center of rhetoric is the formal account of argument understood without a context.⁶³ But this formal rhetoric both establishes and delimits its own context, in such a way that Aristotle can include the ordinary phenomena of rhetoric in his account of rhetorical arguments. Licit rhetoric is held together by the way in which the context is always focused on the argument,⁶⁴ so that the art of rhetoric is, after all, really about argument.⁶⁵ And the truth that lies at the center of the arguments has itself a natural priority, which explains the natural coherence of the body of rhetoric and its parts. Argument is the body of rhetoric because it is about things, it is objective (cf. Metaph. 4.4, 1006bl 1). This is where rhetoric is parallel to dialectic, the enthymeme to the syllogism. So opposed arguments can be resolved by objective logical standards—they are neither indeterminate (no answer) nor subjective (answered by the whim or the preferences of the speaker). But the audience of formal rhetorical argument—in parallel with Plato's emphasis on the psychological dimension of argument-is actively engaged in judgment. What is more, rhetorical arguments are contextualized: rhetorical arguments are essentially communicated. So in the ordinary business of forensic or deliberative rhetoric, people make decisions; they are influenced by the character of the speaker and swayed by their own emotions. To understand this aspect of rhetoric, we must understand ēthos and pathos as well as pure logos. Together they form a natural whole that allows for a genuine technē of rhetoric.

To reach this conclusion, Aristotle has rejected both the Platonic view of rhetoric (objectivity or bust) and the Isocratean one (rhetoric is all-pervasive). Instead, he argues, we need both a formal account of argument (so that rhetoric may be a techne) and a contextual one (to save the phenomena), arranged and unified by the central account of rhetorical argument. So the encounter in the first chapters between Plato and the rhetoricians is a model for the balancing act of the whole work. Aristotle then deploys the method of endoxa—treating the opposed arguments of the wise—to extricate us from the puzzle and provide us with a new account of rhetoric. Rhetoric is a natural whole, whose body is the arguments modally suited to the arena of rhetoric—arguments about what is possible. It is those arguments that organize the art of rhetoric: here Aristotle deploys imagery from nature suitable to his afternoon lectures. In the morning, perhaps, he would have told us about focal meaning.

Notes

¹ By F. Solmsen, Die Entwicklung der Aristotelischen Logik und Rhetorik (Neue Philologische Untersuchungen), Berlin: Wiedmann, 1929. ch. 3; then by J. Barnes, "Proof and the Syllogism" in Berti, E., ed., Aristotle on Science: the Posterior Analytics (8th Symposium Aristotelicum), Padova: Antenore, 1981, pp. 51ff.

² As I now think: in my original paper to the Symposium I gave a harsher verdict on Aristotle. My thanks to the other members of the symposium for their comments on my first version of this paper—and in particular to Aryeh Kosman, Geoffrey Lloyd, Michael Frede, John Cooper, Mario Mignucci, and David Charles; my thanks also to the editors and to an anonymous commentator, who sent me back to the drawing board.

³ … Grg. 464c ff., picked up at Rhet. 1.2.7, 1356a27.

⁴ Well, he might, anyway—there is a question here about whether Plato is suggesting that the tyrant will not invariably get what he does want; or whether, because he is ignorant, he will always get what he does not want.

⁵ Compare his point on what people really believe about justice at 474b.

⁶ Plato takes the notion of the active philosophical life seriously. Philosophy is what you do, not what you suffer; but we are helpless in the face of rhetoric's force majeure. In the Theaetetus Socrates insists that his method of philosophy brings the truth out of the souls of his interlocutors, rather than imposing it on them from without (Tht. 150b ff.); and that is why he denies the charge of teaching the young (anything at all—as at Ap. 24c ff.). Philosophy is done by minds (whether they be conceived as wax tablets or aviaries or scribes or painters), not done to them; and it is for that reason that the method of dialectic is understood as "soul-leading". … Phdr. 261a, cf. Rep. 523a ff.). It is souls who think and understand; so argument happens in souls (Sph. 263e), judgments are made by souls (Tht. 185d ff.). This factor in Plato's account of the philosophical life is significant when we come to consider Aristotle's account of judgment.

⁷ See G. Vlastos, "The Socratic Elenchus." Oxford Studies in Ancient Philosophy I (1983), 27-58.

⁸ Vlastos p. 35.

⁹ Compare, e.g., the discussion of [alogia] in the Meno (80a ff.), where the watershed of the experiment with the slave-boy is the moment at which he realizes that he does not know the answer to Socrates' question. Cf. Mackenzie, "The Virtues of Socratic Ignorance," Classical Quarterly 1988, pp. 331ff.

¹⁰ Cf., e.g., Socrates' dispute with the "other" Socrates, Hp.Ma. 293b, 304c; or his investigation of "know thyself at Chrm. 166d.

¹¹ I discuss Plato's account of the unity of persons in detail in Plato's Individuals (Princeton, forthcoming) ch. 9. It is an oddity of Plato's account of personal integrity (which contains both metaphysical and ethical components) that he can think of personal unity as something to which we aspire—being "one" is honorific.

¹² Cf Mackenzie, Plato on Punishment (Berkeley: 1981), ch. 10.

¹³ Some suppose (e.g., G. R. F. Ferrari, Listening to the Cicadas, A Study of Plato's Phaedrus, Cambridge: Cambridge University Press, 1987; C. Rowe, "The Argument and Structure of Plato's Phaedrus," Proceedings of the Cambridge Philological Society 1986) that the Phaedrus represents a modification of the Gorgias account of rhetoric—Plato has changed his mind. Consequently for my present purposes, there is no single "Platonic" view on rhetoric for Aristotle to consider. Against this complaint I offer three considerations: i) The strategy of Plato in the Phaedrus seems to be to appropriate rhetoric's terms of art ("rhetoric," "persuasion") for the legitimate activities of philosophy or dialectic—he is not making concessions to ordinary rhetoric; ii) even when he does this, he subverts his own conclusions (turns everything upside down, 272b) to show that nothing of what he says may be taken at face value, cf. Mackenzie, "Paradox in Plato's Phaedrus," Proceedings of the Cambridge Philological Society 1982, 64-76; and iii) as far as Aristotle's treatment of Plato is concerned, the view represented in the Gorgias is the Plato he takes on—there is no commitment in the method of [endoxa] to reflect with fidelity any changes of mind by the wise.

¹⁴ Text in L. Radermacher, Artium Magistri et Scriptores—Reste der voraristotelischen Rhetorik. Wien: Osterreichische Akademie der Wissenschaften, 1951, p. 135ff.

¹⁵ There is of course an echo—-or a twist—of this point in the Phaedrus (264c, 275d ff.). Not only Plato but Aristotle too picks up Alcidamas' imagery, as I shall suggest.

¹⁶ Cf.… Antid. passim, e.g., 271- and compare the discussion of rhetoric as power in Plato's Gorgias.

¹⁷ Compare Grg. e.g., 459b ff., or Sph. 234a ff. with Isoc. Soph. 1ff.; Antid. 148.

¹⁸ Compare here Isocrates' complaint … Soph. 7.

¹⁹ Grimaldi appears—implausibly—to attribute this thesis to Aristotle, to explain the purpose of the Rhetoric, in Studies in the philosophy of Aristotle's Rhetoric (Hermes Einzelschriften 25). Wiesbaden: Franz Steiner Verlag, 1972, pp. 5ff.

²⁰ By "ideology" here I mean the collection of inherited assumptions and norms that constitute the society and that, as such, remained unquestioned by it. For a discussion of the influence of ideology on Aristotle (a major issue when it comes to considering his endoxic method), see M. Schofield, "Ideology in Aristotle's Politics," in Patzig, G., ed., Aristotle's Politics (11th Symposium Aristotelicum), Berlin, 1990.

²¹ This, you might say, is a "mouthpiece" theory of legal counsel, where the advocate acts, not as judge of his client's case but as his client's mouthpiece. That view is essential to a thorough answer to the question, "Can you defend anyone you know to be guilty?" Guilt is a matter for the judge and is a fact only after the judgment takes place.… Here the connection between epistemological relativism and its ethical counterpart is very clear.

²² Cf., of course, G. E. L. Owen, "Tithenai ta phainomena," in his Logic, Science and Dialectic, London: Duckworth, 1986, pp. 239-51; and then, e.g., J. Barnes, "Aristotle and the Methods of Ethics," Revue Internationale de Philosophie, 133-34 (1980), 490-511.; M. C. Nussbaum, "Saving Aristotle's Appearances," in Schofield, M., and Nussbaum, M. C., eds., Language and Logos, Cambridge: Cambridge University Press, 1982; but now a much more complex account of Aristotle's approach to the business of philosophy, T. Irwin, Aristotle's First Principles, Oxford: Clarendon Press, 1989, pp.30ff.

²³ Plato is not mentioned by name, but the allusions are obvious; compare here W. Rhys Roberts, "References to Plato in Aristotle's Rhetoric." Classical Quarterly 18 (1924), 344-45.

²⁴ Compare Rhet. 1.1.12, 1355a38 with Plato's complaint about the helplessness of the written word, Phdr. 276.

²⁵ Cf. W. W. Fortenbaugh, "Aristotle's Platonic Attitude toward Delivery," Philosophy and Rhetoric 19 (1986), 242-54.

²⁶ Is the technical procedure one for doing rhetoric or for studying it (cf. Brunschwig, this volume)? At times Aristotle concentrates on the teachers of rhetoric and at times on the practicing rhetor. Fair enough—Plato's attack is both on the teachers of rhetoric, for failing to teach anything at all, and on its practitioners, for their lack of system and their moral decrepitude. I have suggested, however, that for Plato the two points are connected; Aristotle's defense of rhetoric correspondingly copes with both.

²⁷ Compare Aristotle's remarks, EN 5.10, 1137b29ff. on the use of a flexible rule in ethics (and building). Is there a gruesome joke lurking here? Aristotle talks about making a ruler crooked, … cf., e.g., Ar. Clouds 620, Antiphon 5.32.

²⁸ Cf. Burnyeat, this volume, on this passage.

²⁹ This comment, Rhet. 1.1.11, 1355al4-15, seems to make the point about what sort of skill an enthumematic person would have; it is comparable, I take it, to the skilled logician's being able to see validity. On this more below.

³⁰ … Cf.—Grg. 465a2; for Plato this word is pejorative-but Aristotle's usage seems to point rather to the contingent nature of the subject matter of rhetoric; this need not make rhetoric the disaster Plato takes it to be. Cf. here Cooper, this volume, and compare Rhet. 1.5.1, 1360b5 with EN 6.1, 1138b22ff.

³¹ I shall return to the sense of [endoxa] below.

³² We should be a little wary here—does Aristotle mean "stupid" or "reprobate"? If the issue is whether they can understand argument, the emphasis ought to be on the former, and on their character only insofar as they fail to perform proper human functions (i.e., to be sensible, not stupid). Compare here Top. 1.2, 101a30; and Rhet. 3.14.8, 1415b5-6, where Aristotle's complaint is that we have to deal with a lousy auditor (someone who is bad at hearing)—not a lousy person (someone who is bad at being a moral agent). Compare his own strictures against Plato, EN 1.6, 1096a19ff.

³³ The contrast between the plausible and the apparently plausible seems, by the analogy with the syllogism, to be a logical contrast between valid and invalid arguments, not a psychological one between what actually persuades and what merely appears to do so (indeed, it is hard to see what the latter would mean).

³⁴ Note, however, the reservation, "Rhetoric wears the clothes of politics." But "politics" may be said in many ways: as I have construed Plato's objection, rhetoric should be "self-to-self' and interpersonal, so political in that sense. This sort of "politics" is exactly what Aristotle goes on to elaborate in his account of what I describe as the "context" of rhetorical argument.

³⁵ Cf. EN 1.3, 1094bllff; 3.3, 1112a18ff.; 6.1, 1139a12ff.

³⁶ See Aristotle's recommendation that the deliberative orator should know about economics, military strategy, foreign policy, etc., Rhet 1.4.7, 1359b19ff.

³⁷ Epideictic rhetoric seems not to be of prime interest to Aristotle—see, e.g., Rhet. 1.1.5, 1354a22ff. Perhaps this is true because [epideiseis] often take the form of either forensic or deliberative speeches—compare, e.g., Gorgias' Helen, Isocrates' Antidosis, and Rhet. 1.9.35, 1367b37. Or perhaps Aristotle's lack of interest in epideictic may be a part of his emphasis on what is real as opposed to what is invented, and thus part of his stance against Isocratean rhetoric. In what follows my main focus of attention will be deliberative rhetoric—in keeping with the main target of Plato's criticisms.

³⁸ At least only insofar as past propositions are necessary, cf., e.g., Rhet. 1.3. 8, 1359a11ff.; 3.17.5, 1418a5; they do not have the necessity of general statements of essence.

³⁹ Aristotle argues against the Platonic view that value is uniform, but goes along along with Plato in the assumption that value is natural or objective; cf. EN 1.6-7.

⁴⁰ E.g., EN 1.7,1097a32, what is good is sought in itself Aristotle's naturalism about the good does not, of course, preclude his taking the view that goods are relative to persons or circumstances; hence the discussion of the mean, EN 2.6, 1106a26ff. Compare here also the suggestion that the judge should recognize … the justice or otherwise of a case, Rhet. 1.1.6, 1354a29ff. The metaphors of light and darkness at 1.1.7, 13 54b 10-11 imply that, though personal interest may affect a judge's decision, both interest and truth are nonetheless real features (of the world), not inventions or subjective views of the judge himself.

⁴¹ While I hope I shall have shown that the project of the Rhetoric is coherent, this allows for the possibility that the syllogistic framework is a later incorporation into the original series of afternoon lectures.

⁴² Cf above, n. 30. Plato rejected the stochastic approach, Grg. 463a, 465a. [To endekhomenon] inevitably is said in many ways, A Pr. 1.3, 25a37ff.; and compare the problems of the conversion of contingent propositions at Int. 12 and 13, 21a34ff. Cf. here W. and M. Kneale, The Development of Logic, Oxford: Clarendon Press, 1962, pp. 83ff.

⁴³ Cf., e.g., Int. 12, 21b IO; and compare the discussion of possibility and actuality in Metaph. 9, esp. the comparison with natural actuality, e.g., 9.8, 1049b8. Possibility is treated as a rhetorical, [topos] at Rhet. 2.19.1, 1392a8ff.

⁴⁴ There will still be [topoi] common to rhetoric and dialectic; cf. Rhet. 1.2.21, 1358alOff.; 1.3.8-9, 1359a16ff., put into practice at 1.7.1, 1363b5ff.; 2.18.2, 139lb22ff.

⁴⁵ Cf.… Rhet. 2.19.1, 1392a8ff.

⁴⁶ Notice the reference to A Pr. 1.8, 29b29ff.; 1.13, 32a15ff. at Rhet. 1.2.14, 1357a29.

⁴⁷ Look here at Rhet. 1.4.3, 1359a35ff. or at the discussion of chance at 1.10.12, 1369a32ff; or the contrast between habit, nature, and accident at 1.11.3, 1370a6.

⁴⁸ The notion of "syllogism" that appears in this passage in the Topics need only mean "argument," even of so rigorous a type that the conclusion follows "from necessity"; we need not be dealing with the canonical three-proposition syllogism of the Prior Analytics. Cf.… Rhet. 1.2.7, 1356a22 to describe enthymematic arguing. Cf. Burnyeat, this volume.

⁴⁹ Cf. here, e.g., A Pr. 1.1, 24b18ff.

⁵⁰ E.g., Ph. 2.3, 194b17ff.

⁵¹ Cf. here M. F. Burnyeat, "Aristotle on Understanding Knowledge," in Berti, E., ed., Aristotle on Science: The Posterior Analytics (8th Symposium Aristotelicum). Padova: Antenore, 1981, 97-139.

⁵² Compare the way the mind receives sense-data and form, De An. 3.4, 429a l Off.

⁵³ Look, e.g., at A Pr. 1.6, 28a36; 28b30 etc. etc. and cf. J. Lear, Aristotle, the Desire to Understand, Cambridge: Cambridge University Press, 1989, 21 1ff.

⁵⁴ Of a sort, cf. Burnyeat, this volume.

⁵⁵ Cf. here the idiom of sight to describe how the judge sees, or fails to see, the truth, e.g. at Rhet. 1.1.7, 1354blO-11; 1.1.11, 1355alO, 15, etc.

⁵⁶ Thus enthymemes are presented as real arguments by contrast with apparent enthymemes, which are fallacious (and thus deceitful), Rhet. 2.24.1, 1400b34-Olal.

⁵⁷ Some enthymematic kinds of argument (maxims) offer obvious explanations, Rhet. 2.21.6, 1394b21; cf. 2.23.30, l400b30ff.

⁵⁸ In what follows I shall not be suggesting that Aristotle' anticipated Austin or Searle; but see J. L. Austin, How to Do Things with Words. Oxford: Oxford University Press, 1976, e.g., p. 52 on the context of an utterance.

⁵⁹ Inevitably this argument is hugely controversial; see most recently Irwin pp. 179ff.

⁶⁰ Notice the emphatic use of [phainesthai] from Rhet. 2.1.3, 1377b26 onward.

⁶¹ This will generate a cognitivist account of emotion. Compare M. C. Nussbaum, The Fragility of Goodness Cambridge: Cambridge University Press, 1986 with Lear, "Katharsis," Phronesis 1988, 297-326.

⁶² See Cooper, this volume.

⁶³ "It is first by nature," Rhet. 3.1.3, 1403bl8-19.

⁶⁴ Compare Metaph. 4.2, 1003a33; 7.4, 1030al7, where the metaphysics of this strategy is announced. Cf. Owen, "Logic and Metaphysics in Some Earlier Works of Aristotle," in Logic, Science and Dialectic, pp. 180-99. Of course Aristotle could have deployed the strategy of focal meaning without having enunciated it (that is, the Rhetoric need not postdate the late books of the Metaphysics); and he could have used it in his afternoon lectures without subjecting his audience to the rigors of metaphysical argument.

⁶⁵ Argument is the most authoritative of the convictions, … Rhet. 1.1.11, 1355a7; cf. here, e.g., Metaph. 5.4, 1015al4; 8.6, 1045a36; De An. 2.1, 412b9; EN 6.13, 1144b3ff…

Bibliography

Austin, J. L. How to Do Things with Words. Oxford: Oxford University Press, 1976.

Barnes, J. "Proof and the Syllogism." In Berti, 17-59.

——. "Aristotle and the Methods of Ethics." Revue Internationale de Philosophie 133-34 (1980), 490-511.

Berti, E. Aristotle on Science: The Posterior Analytics (8th Symposium Aristotelicum). Padova: Antenore, 1981.

Bumyeat, M. F. "Aristotle on Understanding Knowledge." In Berti, 97-139.

Ferrari, G. R. F. Listening to the Cicadas, A Study of Plato's Phaedrus. Cambridge: Cambridge University Press, 1987.

Fortenbaugh, W. W. "Aristotle's Platonic Attitude toward Delivery." Philosophy and Rhetoric 19 (1986), 242-54.

Grimaldi, W. M. A., Studies in the Philosophy of Aristotle's Rhetoric (Hermes Einzelschriften 25). Wiesbaden: Franz Steiner Verlag, 1972.

Irwin, T. Aristotle's First Principles. Oxford: Clarendon Press, 1989.

Kneale, W., and M. Kneale. The Development of Logic. Oxford: Clarendon Press, 1962.

Lear, J. "Katharsis." Phronesis 1988, 297-326.

——. Aristotle, the Desire to Understand. Cambridge: Cambridge University Press, 1989.

Mackenzie, M. M. (McCabe). Plato on Punishment. Berkeley: University of California Press, 1981.

—"Paradox in Plato's Phaedrus." Proceedings of the Cambridge Philological Society 1982, 64-76.

—"The Virtues of Socratic Ignorance." Classical Quarterly 1988, 331-50.

McCabe, M. M. (Mackenzie). Plato's Individuals. Princeton: Princeton University Press (forthcoming).

Nussbaum, M. C. "Saving Aristotle's Appearances." In Schofield, M., and Nussbaum, M. C., eds., Language and Logos. Cambridge: Cambridge University Press, 1982.

—The Fragility of Goodness. Cambridge: Cambridge University Press, 1986.

Owen, G. E. L. "Tithenai ta phainomena." In his Logic, Science and Dialectic. London: Duckworth, 1986, pp. 239-51.

—"Logic and Metaphysics in Some Earlier Works of Aristotle." In Logic, Science and Dialectic, pp. 180-99.

Radermacher, L. Artium Magistri et Scriptores—Reste der voraristotelischen Rhetorik. Wien: Osterreichische Akademie der Wissenschaften, 1951.

Rowe, C. "The Argument and Structure of Plato's Phaedrus" Proceedings of the Cambridge Philological Society, 1986, pp. 106-25.

Rhys Roberts, W. "References to Plato in Aristotle's Rhetoric." Classical Quarterly 18 (1924), 344-45.

Schofield, M. "Ideology in Aristotle's Politics." In Patzig, G., ed., Aristotle's Politics (1lth Symposium Aristotelicum). Berlin: de Gruyter, 1990

Solmsen, F. Die Entwicklung der Aristotelischen Logik und Rhetorik (Neue Philologische Untersuchungen). Berlin: Wiedmann, 1929.

Vlastos, G. "The Socratic Elenchus." Oxford Studies in Ancient Philosophy 1 (1983), 27-58.