SOURCE: "The Rationality of Political Speech: An Interpretion of Aristotle's Rhetoric," in Interpretation: A Journal of Political Philosophy, Vol. 9, Nos. 2 & 3, September, 1981, pp. 141-54.

[In the following essay, Arnhart maintains that Aristotle uses the concept of enthymeme (a logical argument, or syllogism, in which one of two conclusion-supporting premises is unexpressed) to defend the legitimacy of rhetorical discourse and to distinguish rhetoric from both science and sophistry.]

I

Is rhetoric some form of rational discourse about the intelligible reality of politics? Or is it merely a means for verbally manipulating men through fallacious arguments and appeals to irrational impulses? In short, can rhetoric be distinguished from sophistry?

One might say that the rhetorician—by his use of public speech to interpret, evaluate, and deliberate about political action—maintains somehow the rule of reason in political affairs. Does not rhetoric require political men to talk about and thereby to think about what they have done, are doing, or will do? Does not rhetoric thus elevate politics by bringing thought to bear upon action? "We weigh what we undertake and apprehend it perfectly in our minds," Pericles declared in his funeral oration, "not accounting words for a hindrance of action but that it is rather a hindrance to action to come to it without instruction of words before" (Thucydides, Peloponnesian War II 40).

But rhetoric also has a darker side. Does not the rhetorician sometimes employ emotional appeals and deceptive arguments to move his listeners to whatever position he wishes? Indeed, does not rhetoric consist of techniques that can be used as easily for the wrong as for the right side of any issue? In other words, there surely is some justification for the ancient criticism of rhetoric as permitting speakers to make the weaker argument appear to be the stronger. As Gorgias boasted, "Many are the men who shape a false argument and persuade and have persuaded many men about many things" (Helen 11).

So the problem is that, while rhetoric seems in some respects to be the means by which reason guides political action; it often seems to be an art of deception that hinders rational deliberation. Furthermore, to the extent that rhetoric is the primary mode of political reasoning, how one decides this question as to whether or not rhetoric is a genuine form of reasoning will determine the place of reason in political life.

The rationality of rhetoric becomes especially dubious if scientific demonstration is taken to be the sole model of valid reasoning. For it is obvious that rhetorical argument cannot attain the exactness and certainty that is possible in scientific inquiry. And therefore if only scientific demonstration is truly rational, rhetoric must be irrational. As a result, rhetoric becomes virtually indistinguishable from sophistry. For since there are no rational standards for political discourse, the power of rhetoric must depend upon manipulation through verbal deception and not upon any pervasive intelligibility of the speech itself. As a further consequence, the political itself becomes irrational. Since the ordinary discourse of citizens about political things has little to do with scientifically demonstrable knowledge, the political life of men must be understood to be guided by opinions with little foundation in reason.

But could one save the rationality of political speech—and of the political realm as a whole—by viewing rhetoric as occupying some middle ground between science and sophistry? This could be done if one could show that the realm of reason extends beyond the confines of scientific demonstration, and therefore that rhetorical argument can be in some sense truly rational even though it lacks the certainty and exactness of scientific knowledge. In this way one would restore the meaning of rhetoric as rational discourse.

And in fact this would seem to be the project that Aristotle sets for himself in the Rhetoric. For he criticizes the sophistical rhetoricians, whose common practice is to use purely emotional appeals to distract their listeners from the subject at hand, for failing to see that the true art of rhetoric is essentially a mode of reasoning, although without the rigor of apodictic proof. He explains rhetorical reasoning as reasoning through enthymemes, and it is in his conception of the enthymeme that his theory of rhetoric is most fully embodied. My claim, therefore, in this essay, is that Aristotle's rhetorical theory is an account of the rationality of political speech. To fully substantiate this interpretation of the Rhetoric would require a much more extensive commentary on the text than is possible here.¹ But I can at least state some of the major points.

How Aristotle uses his theory of the enthymeme to differentiate rhetoric from science on the one hand and from sophistry on the other, becomes clearer in the light of four tripartite distinctions. First, persuasion, which is the aim of the enthymeme, differs, both from instruction and from compulsion. Second, opinion, which provides premises for the enthymeme, does not conform to absolute truth, but neither is it absolute falsehood. Third, the probability characteristic of most enthymematic inferences falls somewhere between necessity and mere randomness or chance. Finally, the enthymeme itself differs from a strict demonstration but without being a sophistical fallacy. I shall comment briefly on each of these points.

II

That men are by nature both rational and political is manifest in the natural human capacity for speech. Men are naturally more political than gregarious animals, Aristotle says in the Politics (1253al5-18), because human community rests upon a union in discourse and thought. Other animals may signify to one another with their voices their sensations of pleasure and pain but men through rational speech … can share with one another their concepts of expediency, justice, and goodness. Human beings achieve a more intimate community among themselves than is possible for other creatures, because only human beings can found their association on mutual understanding through speech.

One might conclude from this that rhetoric—the artful practice of public speech—is the fundamental activity of politics, and that politics expresses the rational nature of men insofar as political activity is founded upon rhetoric. But does rhetoric encompass the whole of politics? Or is it perhaps important for only a limited realm of political life? That Aristotle does not simply identify politics with rhetoric is clear from his remarks at the end of the Nicomachean Ethics. Speeches or arguments …, he explains, are not sufficient to make men virtuous (1179b4-1180b28). At best they are effective with youths who because of some natural endowment or good moral training have a love of the noble. Most men, especially in their youth, live by passion and the pleasures of the body, and hence they can be controlled by force but not by arguments. For these people it is necessary that the laws coercively habituate them from their youth to do those virtuous things that they would never choose to do on their own. Thus the moral training of a community requires that the legislator apply legal compulsion where moral persuasion would be futile.

It is at this point that Aristotle criticizes the sophists for showing their ignorance of politics by making it the same as, or lower than, rhetoric (Nicomachean Ethics 1181al2-16).² This is often taken to indicate that Aristotle thought the sophistical view of politics to be too cynical, but from the context one might infer something quite different: the sophistical assumption that the art of persuasion can govern all political activity manifests a naive blindness to the true harshness of political life. Rhetorical reasoning displays the nobler side of politics, that area of political activity governed by persuasion through speeches. But most men respond not to persuasion but to force, and therefore the greater part of politics must be concerned with compelling men, and through repetition habituating them, to do without thought what they could never be persuaded to do. The success of rhetoric, Aristotle implies, presupposes the formation by the laws of an ethos in the community that makes people open to persuasion. The taming of the most irrational impulses demands force rather than argument; but once the lowest part of the soul has been subdued, the rhetorician can appeal to that part of the soul that can be persuaded by reason. Rhetoric is therefore subordinate to politics since the multitude of men would never be amenable to rhetorical reasoning unless they were first properly habituated by the laws.

Hence rhetoric introduces the rule of reason into human affairs since it moves men by persuasion rather than by force. And yet Aristotle makes it clear that rhetoric fails to attain the highest level of reasoning insofar as rhetorical persuasion falls short of scientific or philosophic instruction (R 1355a22-29). The exact knowledge and complex demonstrations necessary for scientific instruction are rarely effective in political speeches. To be persuasive, the rhetorician must draw the premises of his enthymemes not from the first principles of the particular sciences, but from the common opinions of his audience. And he must simplify and abbreviate his line of reasoning so that ordinary citizens can grasp it quickly and easily (R 1357a8-23, 1395b24-30, 1419al8-19). Thus the good rhetorician can persuade, but he cannot instruct.

Since the premises of the enthymeme are derived from common opinions, and since opinion surely differs from truth, it might seem that the enthymeme is a false form of reasoning, and therefore that all rhetoric is sophistical. But in fact Aristotle regards the common opinions that enter the enthymeme as being for the most part neither completely true nor completely false but at least partially true (R 1357b21-25, 1361a25-27; NE1098b26-30, 1145bl-7; EE 1216b28-35). Therefore, although this reliance on opinions does impose certain limits on enthymematic argumentation, this does not prevent the. enthymeme from being a valid form of reasoning. Although the "reputable opinions"… on any particular subject are usually confused and even apparently contradictory, Aristotle assumes that in most cases they manifest at least a partial grasp of the truth and therefore that any serious inquiry into moral or political subjects must start from them. So while Aristotle treats certain subjects differently in the Rhetoric than he does in the Politics or in his ethical treatises, since rhetoric involves opinions in their original state without the refinements of philosophic examination, his expositions in the Rhetoric still reflect in some fundamental manner those in his other works. For example, the account of "happiness" … in the Rhetoric clearly reflects, even if somewhat dimly, the philosophic understanding of "happiness" set forth in the Nicomachean Ethics (compare R 1360b15-19 with NE 1097b7-21, 1176b4-7; see also P 1323b21-1324a4, 1325bl4-31).

Furthermore, in its dependence on common opinions, rhetoric is distinguished both from science and from sophistry. Each science begins not with common opinions, but with the primary truths that are fundamental to the science (R 1358al7-27, PoA 71bl8-72a6, SR172al2-172b4). (But as I shall indicate later, even these scientific truths depend ultimately on some commonsense understanding of things.) And sophistry consists either of arguing from what appear to be common opinions but are not, or of making something appear to follow necessarily from common opinions when it does not(SR 165a37-165bl2, 176b29-177a8). Moreover, the fact that sophistical arguments cannot be truly derived from common opinions confirms the epistemological solidity of these opinions.

One of the limitations of common opinions, however, is that they usually hold for the most part but not in every case. Therefore, enthymemes have probable but not necessary validity, since the conclusions are true in most cases but not in all. Enthymemes, then, rarely achieve the necessity of scientific demonstrations. The one exception noted by Aristotle is the enthymeme founded on a "necessary sign" … (R 1357a24-34).

That enthymematic reasoning usually involves probability rather than necessity does not make the reasoning invalid. For, according to Aristotle, both the things that happen always or by necessity and those that happen as a rule or for the most part, can be objects of knowledge. Probability must be distinguished from chance, because unlike probable things those things that happen only rarely or by chance cannot be known (PoA 87b l9-28). That rhetoric should rest upon probabilities is consistent with the Aristotelian principle that one should demand only that degree of certitude that is appropriate to the subject matter. For like ethics and politics the subject of rhetoric is human action, and the regularities of human action can be known with probability but not with absolute certainty (R 1356al4-17, 24-33, 1402b21-37).

Since the enthymeme rests upon opinion rather than absolute truth, since its premises and conclusion are probable rather than necessary, and since its final aim is persuasion rather than instruction, enthymematic reasoning lacks the rigor of scientific demonstration. And yet rhetorical argument is still a valid form of reasoning, and therefore it provides an alternative to sophistry. Popular opinions manifest a commonsense grasp of reality that cannot be dismissed as simply false. Probabilities are fit objects of reason because they presuppose regularities in things, which are not random or by chance. And, finally, the persuasion for which the rhetorician strives requires an appeal to reason rather than force.

But to support the claim that Aristotle's Rhetoric is a theory of rhetoric as truly rational discourse, one must answer the serious objections that can be made to this interpretation. In particular, the following four points deserve attention. (1) It could be argued that enthymemes cannot be valid because Aristotle defines them as incomplete or otherwise defective syllogisms. (2) Furthermore, even if the enthymeme were a genuine syllogism, it could still be argued that Aristotle's discussions of persuasion through the character of the speaker and through the passions of the audience would show the reliance of rhetoric on irrational appeals. (3) Also, since Aristotle insists that rhetoric includes apparent as well as genuine "proofs," and since he describes it as a neutral instrument that may be used on either side of any issue, one might infer that he does not clearly distinguish rhetoric from sophistry. (4) Finally, Aristotle's remarks in Book Three of the Rhetoric on the style and arrangement of speeches seem to be further evidence that he does not view rhetoric as founded on rational argument. r shall reply to each of these objections.

III

Aristotle's enthymeme is a true syllogism; and therefore it is not, as has been commonly assumed, an incomplete syllogism. For if the enthymeme were an invalid or incomplete syllogism—to cite only one argument from the text—why would Aristotle distinguish between apparent and true enthymemes and declare that apparent enthymemes "are not enthymemes since they are not syllogisms" (R 1397a3)?

Aristotle refers to the enthymeme as "a sort of syllogism" … (R 1355a9-10), and some readers have taken this use of (tis] as implying that the enthymeme is not a true or complete syllogism. But the falsity of this interpretation is made evident by a passage in the Prior Analytics (24a10-16, 25b26-31). Here Aristotle explains that his theory of the syllogism in the Prior Analytics is more general than his theory of "demonstration" … in the Posterior Analytics: "for demonstration is a kind of syllogism …, but not every syllogism is a demonstration." Since there is no reason to believe that a "demonstration" is anything less than a true syllogism, it is clear that the phrase [sullogismos tis] is intended only to indicate that a "demonstration" is one kind of syllogism to be differentiated from other kinds (see also Poetics 1450al8). Likewise, the enthymeme can be one distinctive type of syllogism without being syllogistically defective, which is born out by Aristotle's repeated references to the syllogistic character of the enthymeme (see, for example, R 1362b29-30, 1394a9-11, 1400b25-33; PrA 68b8-14; PoA 71al-11). Since the premises and therefore the conclusion of the enthymeme are founded on common opinions and are probable but not absolutely certain, the enthymeme differs from the scientific syllogism; and since the enthymeme must be simple enough to be understood by the ordinary man, it differs from the dialectical syllogism. But neither of these points entails that the enthymeme be an invalid or incomplete syllogism.

Enthymematic reasoning is popular because by providing listeners with "quick learning," it satisfies their natural desire for learning (R 1400b25-33, 1410b6-35). For this reason, the enthymeme should be neither too superficial and obvious nor too long and complex. It should be simple enough to be quickly grasped, but at the same time it should give the listeners the pleasure of learning something new: it should be informative without being esoteric.

One of the ways to make the enthymeme an instrument of "quick learning" is to abbreviate it by leaving unstated whatever the listeners can be expected to add on their own (R 1356al9, 1357al7-23). But this practical rule is not part of the definition of the enthymeme; and furthermore, even when it is abbreviated, the enthymeme is a complete syllogism as stated in thought despite its incompleteness as stated verbally (PoA 76b23-28). Even in the most rigorously demonstrative reasoning, Aristotle suggests, premises that are clear or well known need not be explicitly stated (PoA 76b l-23). Moreover, the abbreviation of enthymemes is a tribute to the love of learning found in the audience. For when a speaker leaves unstated those steps in the reasoning that the listeners can easily supply themselves, he allows them to help construct the very arguments by which they are persuaded; and thus he gives them the satisfaction of thinking through the reasoning on their own.

IV

Aristotle begins the Rhetoric by condemning those sophistical rhetoricians who rely exclusively on exciting the passions of their listeners and thereby preventing them from making a rational judgment about the issues at hand. These speakers ignore the enthymeme, which is "the body of proof for rhetoric. But when Aristotle sets out the three "proofs" …³ of rhetoric, he includes appeals based on "character" … (that is, the "character" of the speaker) and "passion" … as supplementary to persuasion through the "speech" or "argument" itself…; and in Book Two he carefully delineates the passions with which the rhetorician must deal. Thus Aristotle seems to throw into doubt the rationality of rhetorical argument by introducing the same techniques for moving audiences through their passions that he initially condemns.

A closer examination, however, will show that Aristotle's emphasis on the enthymeme is consonant with his treatment of the passions, because the enthymeme combines reason and passion. Since it is "the body of proof," the enthymeme is the vehicle not just for one of the three "proofs" … but for all three … (R 1354al2-16, 1354b20-21, 1396b28-1397a6, 1403a34-1403bl). Enthymemes may be used not only to establish a conclusion as a probable truth, but also to alter the emotions of the listeners or to develop their confidence in the character of the speaker. Aristotle denounces the sophistical rhetoricians not because they appeal to the passions of the audience, but because they do this in a defective manner. Their solicitation of the passions would be acceptable if it were an integral part of an enthymematic argument pertinent to the subject under examination, but their exclusive reliance on the passions with no connection to any form of argument only distracts the listeners with things irrelevant to the matter at hand (R 1354a13-15, 1354bl8-22, 1356a9-19). The sophist excites the passions to divert his listeners from rational deliberation, but the Aristotelian speaker controls the passions of his listeners by reasoning with them.

Aristotle assumes that the passions are in some sense rational, and that a rhetorician can talk an audience into or out of a passion by convincing them that the passion is or is not a reasonable response to the circumstances at hand (R 1378a20-31, 1380b30-33, 1382al6-18, 1385a29-35, 1387bl8-21, 1403a34-1403bl). Since a passion is always about something, since it always refers to some object, it is reasonable if it represents its object correctly or unreasonable if it does not. Men's passions are not always reasonable, but they must always believe that they are: they have reasons for their passions although their reasons are not always good ones. The passions may often arise from false judgments about reality, but the mere fact that passions require judgments, whether true or false, suffices to show the rational character of the passions. And a rhetorician who understands this can learn to change the passions of his listeners by changing their minds.

It is the rationality of the passions that distinguishes them from purely bodily sensations and appetities. It would be ridiculous to judge an itch or a pang of hunger as true or false, reasonable or unreasonable; and it would be equally absurd to argue with a man who felt an itch or a sensation of hunger in order to convince him that his feelings were unjustified. But it is not ridiculous to judge a man's anger as reasonable or unreasonable or to try to argue with him when his anger is unjustified. A man's anger depends upon his belief that anger is a proper response to something that has occurred, but a man's sensations or physical appetites do not require that he believe this or that (NE 1149a25-1149b3).

The passions are rational in that they are founded on judgments of what the world is like, but they are less than perfectly reasonable to the extent that they are founded on shortsighted, partial, biased, or hastily formulated judgments. Yet the fact that the passions often depend on defective reasoning should not obscure the fact that they do require some sort of reasoning, and it is this element of reasoning that gives the rhetorician a lever for controlling the passions.

That enthymemes are often directed to the emotions of the listeners indicates again the difference between enthymematic and demonstrative reasoning. Emotions are irrelevant to scientific demonstration; but since enthymematic argumentation is a practical form of reasoning, its aim is to move men not just to think, but also to act; and argument cannot move men to action unless it somehow elicits the motivational power of emotion.

V

The interpretation of the enthymeme that I have advanced here suggests that Aristotle considered rhetorical argument to be governed by definite epistemological standards. But his treatment of enthymemes includes a study of "apparent enthymemes"—that is, fallacious arguments—and there are other examples of the care with which Aristotle instructs the rhetorician in the techniques of verbal deception. Indeed, the art of rhetoric is said to provide the power to be persuasive on the opposite sides of every question. So what is to prevent this art from being used to advance falsehood and injustice rather than their opposites? In other words, what is to keep the Rhetoric from being a handbook for sophists?

First of all, it may be answered that Aristotle recognizes that if the rhetorician is to be well armed, he must know all the tricks of sophistry so that he can properly defend himself. The Aristotelian rhetorician might even have to employ such tricks himself in those cases where otherwise bad means are justified by their advancement of good ends (R 1355a29-34, 1407a32-1407b7). Presumably, Aristotle would have the rhetorician follow the example of the dialectician: although he prefers to speak only with those who maintain discussion at a high level, the dialectician is able to defend himself in debates with unscrupulous opponents by using their own sophistical weapons against them, even to the point of showing himself more skillful with their weapons than they are themselves (T 108a33-37, 164b8-15; SR 175a32-175b3; compare R 1407a32-1407b7).⁴

In some cases, Aristotle does instruct the rhetorician in arguing opposite sides of an issue depending upon which side is most favorable to his position at the moment. But this is not a sophistical exercise, because in each case there is something valid to be said on both sides. In practical matters there is sometimes equally strong support for opposing arguments, and the prudent man must recognize this (see, for example, R 1375a25-1376b31).

It should also be said that, although the rhetorical art in itself is a morally and epistemologically neutral instrument, rhetoric tends to serve the true and the just. Even though there are no ends intrinsic to the art itself, ends are prescribed by the rhetorical situation—the speaker, the subject matter, and the audience. Since speakers who display good character are more persuasive, the noble rhetorician has an advantage over the sophist, who must attempt to hide his bad character (R 1356a6-13, 1378a6-19). Also, the sophistical speaker is restrained by the nature of the subject matter and by the opinions of the audience. With respect to the subject matter, it is generally the case that the true and the just are naturally more easily argued and more persuasive; and the opinions of the audience generally display this same tendency (R 1354a21-26, 1355al2-23, 36-38, 1371b5-11, 1373b3-13, 1396a4-1396bl9, 1409a35-1409bl2, 1410b9-35). Thus, in most cases, a speaker who has something to hide is more vulnerable than one who has not (R 1397b23-25, 1402a23-28, 1419al3-17). It is difficult to give a good speech for a bad cause (see Thucydides, Peloponnesian War III 36-48). This is not to deny, however, that sometimes the weaker argument can be made to appear the stronger. But—and this seems to be Aristotle's point—is it not usually easier to make the stronger argument appear to be the stronger, especially when it is skillfully presented?

VI

Matters of style and composition seem extraneous to the rational content of rhetoric since they seem unnecessary for the substantive argument of issues. Indeed, when Aristotle takes up these matters in Book Three, he begins by complaining that a concern with such things is only a concession to corrupt audiences (R 1403bl5-1404al2, 1415a5-1416a2).

But in his treatment of these elements of rhetoric, Aristotle stresses the extent to which they contribute to rational argument. For Aristotle good style is not merely ornamentation, since the goodness of style is determined by how well it satisfies the natural desire of listeners for learning through reasoning (R 1404b 1-13, 1408b22-29, 1409a23-1409bl2, 1410a18-22, 1412b21-32, 1414a21-28). Metaphor, for example, the most important instrument of rhetorical style, provides listeners, in a manner similar to the enthymeme, with "quick learning" ((R 1405a5-12, 33-37, 1410b6-35, 1412b9-12, 18-28). And Aristotle insists that the best arrangement for a speech is that which presents the substantive argument as clearly and directly as possible: a speaker should first state his case and then prove it (R 1414a30-1414bl8). The Aristotelian rhetorician strives for the same end in his style and composition as he does in his enthymematic reasoning—to be clear but not commonplace and informative but not recondite.

VII

I have argued that Aristotle views rhetoric as rational discourse, and that he wishes to show that rhetoric is a form of reasoning to be distinguished from sophistry, even though rhetorical reasoning is less exact and less certain than scientific demonstration. Measured by the standards of strict, demonstrative logic, the political argumentation of citizens does not usually qualify as genuine reasoning at all. But such argumentation can be seen to be quite rational if it is judged according to the logical criteria of rhetoric. Aristotle's theory of rhetoric rests on the assumption that one should evaluate political arguments according to their degrees of plausibility without demanding absolute certainty or exactitude. Thus Aristotle's theory conforms to the logical practice of citizens, who are able to judge the plausibility of arguments despite the fundamental uncertainty of all practical reasoning.

But what would the modern political scientist say about all this? He might protest that rhetoric is surely not a valid form of reasoning since it violates even the most elementary rules of scientific rationality.⁵ The fundamental problem, he might explain, is that the rhetorician's arguments can be only as reliable as the commonsense political opinions from which he draws his premises, but common opinions are at best uncertain and inexact reflections of political reality and at worst unexamined prejudices with no claim to truth. In contrast to the rhetorician's dependence on the vague and deceptive impressions of common sense, the contemporary political scientist might appeal to the epistemological criteria of a scientific methodology for precise standards of political knowledge. But does the modern scientific method provide a better starting point for political inquiry than does rhetoric?

This question was first clearly posed by Thomas Hobbes. For he rejected Aristotelian political science and applied the scientific method to political study, and in doing so he became the founder of modern political science. Now Hobbes did admire Aristotle's Rhetoric for its psychological insights; but he certainly denied Aristotle's claim, which is essential for his rhetorical theory, that common opinions can be the foundation of political reasoning. Classical political philosophers such as Aristotle could never lead us to genuine political knowledge, Hobbes argued, because "in their writings and discourse they take for principles those opinions which are vulgarly received, whether true or false; being for the most part false."⁶ Instead of starting with political opinions, Hobbes's political science would start with exact definitions and axioms; and from these principles one would deduce a theoretical framework that would provide the certainty and precision of geometry. Thus did Hobbes initiate the project to which many political scientists today have devoted themselves.⁷

Is there anything to be said in favor of Aristotle's reliance on common political opinions as the foundation of political knowledge? On the one hand, Aristotle's theory of rhetoric as a valid form of political reasoning depends on the assumption that common opinions reflect a rational grasp of political life. But, on the other hand, Aristotle presents those opinions as often offering a confused, crude, and distorted view of political reality, thus falling short of the rigor, refinement, and comprehensiveness necessary for political philosophy. It seems that political opinions are the starting point for the Aristotelian political scientist, but they are only the starting point. That is to say, the respect that he gives to those opinions does not require an uncritical acceptance. Since the political theorist seeks to move from opinion to knowledge, he will not completely accept the answers given in political speech. And yet even in his movement beyond the common political opinions, he will be guided by the questions to which those opinions point: he will try to give an adequate answer to the questions that political opinions answer only inadequately.⁸

But if Aristotle finds common opinions so defective that he has to transcend them, why does he not reject them from the start in order to reason from scientific principles in the manner advocated by Hobbes? Aristotle might answer with two types of arguments. First, the phenomena studied by the political scientist differ from those studied by the natural scientist in ways that justify a difference in method. Second, all reasoning, even that of the natural scientist, depends ultimately upon the truth of our commonsense understanding of things.

Because political phenomena are contingent rather than necessary, and because they are essentially cognitive rather than physical, the political scientist, Aristotle might argue, must rely on commonsense opinions in a manner that would be inappropriate for the natural scientist. Political reality is contingent because it depends upon human choices that change from time to time and from one situation to another. The nature of political life will vary, for example, depending upon the type of regime in existence: oligarchic politics differs from democratic politics. A regime is a product of certain choices as to the organization and the goals of political rule. To understand these choices, the political scientist must study them as they are manifested in common opinions. And it would be a mistake to try to examine these things as if they were as unchangeable as the Pythagorean theorem or the motion of the planets. Moreover, political things are not physical objects that can be studied through sense perception. A political scientist who restricted himself to sense data would never see anything political. For political phenomena come into view only when one pays attention to what people think about politics as indicated by what they say about it. Thus, again, an appeal to political opinions is unavoidable.

But in the most fundamental respect, all reasoning—not just that of political science—depends upon commonsense opinions. This is so because all reasoning rests upon presuppositions drawn from our commonsense awareness of things. The rules of logic govern the deduction of conclusions from premises, but these rules cannot determine the truth or falsity of the first premises. Reasoning is grounded upon fundamental assumptions that cannot be proven because they are the source of all proofs. A conclusion is demonstrated when it is shown to follow from certain premises. And the premises may themselves be shown to follow as conclusions from other premises. But eventually one must reach principles that are taken as true without proof, these being the starting points of reasoning. Indeed, are not the rules of logic themselves assumptions that cannot be proven logically?

Even the most rigorous empirical science cannot avoid reliance on unprovable assumptions. Scientific induction, for example, rests on the presupposition that one may generalize from particular cases, which depends in turn on the broader assumption that nature falls into recurrent patterns: one must assume that the universe is governed by laws, and that these laws do not change arbitrarily from one moment to another.⁹ Thus does scientific knowledge presuppose a prescientific knowledge of things. This is what Aristotle means when he says that to examine the first principles of any science, one must appeal to the "common opinions" … that are the source of the principles (T lOOal8-100b22, 101a37-101b4). Werner Heisenberg, the great twentieth-century physicist, seems to make the same point when he observes: "the concepts of natural language, vaguely defined as they are, seem to be more stable in the expansion of knowledge than the precise terms of scientific language, derived as an idealization from only limited groups of phenomena." This is the case because, on the one hand, "the concepts of natural language are formed by the immediate connection with reality"; but, on the other hand, scientific concepts require idealization and precise definition through which "the immediate connection with reality is lost." So Heisenberg concludes: "We know that any understanding must be based finally upon the natural language because it is only there that we can be certain to touch reality, and hence we must be skeptical about any skepticism with regard to this natural language and its essential concepts."¹⁰ A similar line of thought is found in the writings of Alfred North Whitehead. For although he helped to formulate modern mathematical logic, he insisted: "Logic, conceived as an adequate analysis of the advance of thought, is a fake. It is a superb instrument, but it requires a background of common sense."¹¹

Our commonsense awareness of reality is more reliable than any epistemological theory could ever be. In fact, the truth of any epistemological theory will depend upon how well it accounts for our reliance on common sense.¹² The Hobbesian political scientist may think he can acquire political knowledge through a formal method that is totally abstracted from commonsense experience. But in practice his choice of assumptions will always be guided, even if unintentionally, by his own natural grasp of political reality. How could he even begin looking for political phenomena if he did not already somehow know what politics was like? As with Lewis Carroll's Alice, he must learn that if he is completely lost, he will never find his way; for if he does not know where he wants to get to, it does not matter which way he goes. But the Hobbesian political scientist knows more than he will admit. For like any sensible human being, he begins with a natural awareness of political things that directs his scientific inquiry. He is not completely lost after all. He knows at the start, even if only vaguely, where he wants to go; so it is not surprising that he usually finds a way to get there.

To fully understand the fundamental importance of commonsense experience for political reasoning, one must see the limits of the Hobbesian method, and one must recover the Aristotelian tradition of political science. Aristotle's Rhetoric is an essential part of that tradition. More clearly than any other Aristotelian text, it brings into view the common political opinions of human beings as the primary ground of political knowledge. Although the political scientist must eventually go beyond those opinions through a process of philosophical refinement, he must always look to them for guidance. For only by continually turning his attention to the political questions found in ordinary political speech, can the student of politics understand political things as they are in themselves.

Notes

¹ Here I can only sketch the outline of an argument that I have developed in detail in Aristotle on Political Reasoning: A Commentary on the "Rhetoric" (DeKalb: Northern Illinois University Press, 1981). I have applied Aristotle's rhetorical theory to American rhetoric in an unpublished paper, "The Federalist as Aristotelian Rhetoric" (presented at the 1979 Annual Meeting of the Midwest Political Science Association, Chicago, April 18-21).

² Henceforth I shall abbreviate my references to Aristotle's works as follows: Eudemian Ethics (EE), Nicomachean Ethics (NE), Politics (P), Posterior Analytics (PoA), Prior Analytics (PrA), Rhetoric (R), Sophistical Refutations (SR), Topics (T).

³ Since [pistis] can be translated as "belief or "trust," Aristotle's use of this word in connection with rhetoric has been interpreted by some commentators as suggesting the weakness of rhetorical reasoning. But in fact Aristotle employs the term to refer to any belief that arises from a syllogism or from induction (see PrA 72a26-72b4; T 100b I8-22; SR 165b3; NE 1139b32-34, 1142al8-21; P 1323a34-1323b7).

⁴ Does Aristotle discuss the tricks of the base to instruct good men and thus to dispel the smug assumption of bad men that good men must be naive? See R 1355a29-34; compare R 1373a3 with Xenophon, Anabasis II.vi.24-26; see also P 1313a34-1315b39.

⁵ From the perspective of the contemporary political scientist, rhetoric may appear to be nothing more than the manipulation of irrational symbols that do not reflect empirical reality. See, for example, Murray Edelman, The Symbolic Uses of Politics (Urbana: University of Illinois Press, 1964), pp. 18-21, 29.35, 41-42, 96-98, 115-17, 121, 124-25, 161, 172-73, 179-81; Edelman, Politics as Symbolic Action (New York: Academic Press, 1971), pp. 1-2.

⁶Elements of Law, 1.13.3. See John W. Danford, Wittgenstein and Political Philosophy (Chicago: University of Chicago Press, 1978), pp. 16-42.

⁷ See,, for example, Harold Lasswell and Abraham Kaplan, Power and Society: A Framework for Political Inquiry (New Haven: Yale University Press, 1950).

⁸ Here and elsewhere in these concluding remarks I have drawn ideas from Leo Strauss, "An Epilogue," in Essays on the Scientic Study of Politics, edited by Herbert J. Storing (New York: Holt, Rinehart and Winston, 1962), pp. 307-27; Wilhelm Hennis, Politik undpraktische Philosophie (Berlin: Luchterhand, 1963), pp. 89-115; and Eugene F. Miller, "Primary Questions in Politics," The Review of Politics, 39 (July, 1977), 298-331.

⁹ On the assumptions necessary for modern science, see A. D'Abro, The Rise of the New Physics, 2 vols. (New York: Dover, 1951), I, 14-27. See also my article, "Language and Nature in Wittgenstein's Philosophical Investigations," Journal of Thought, 10 (July, 1975), 194-99.

¹⁰Physics and Philosophy (New York: Harper & Row, 1958), pp. 200-202.

¹¹ "Immortality," in The Philosophy of Alfred North Whitehead, edited by Paul Arthur Schilpp, 2nd ed. (La Salle, Ill.: Open Court, 1951), p. 700. The importance of "common sense" for mathematics is a theme of my unpublished paper, "Mathematics and the Problem of Intelligibility."

¹² One should keep in mind here the long rhetorical tradition of speculation about the nature of "common sense." See, for example, Thomas Reid, Essays on the Intellectual Powers of Man (Cambridge: M.I.T. Press, 1969), pp. 556-68; and Hans-Georg Gadamer, Truth and Method (New York: The Seabury Press, 1975), pp. 19-29.