SOURCE: Schmitz, Thomas A. “Plausibility in the Greek Orators.” American Journal of Philology 121, no. 1 (2000): 47-77.

[In the following essay, Schmitz analyzes the rhetorical strategies Antiphon and other orators used to convince judges of the accuracy of their arguments in representing reality.]

When Tzvetan Todorov edited a special issue of the journal Communications on vraisemblance (verisimilitude) in 1968, he described the origin of the concept as follows (I paraphrase):

One day during the fifth century b. c., there was a trial in some Sicilian city. Neither the plaintiff nor the defendant could produce witnesses or any other form of evidence to corroborate their version of the events, so they had to convince the judges by the sheer power of their arguments. This was the day the principle of vraisemblance was discovered.

Todorov's tongue-in- explanation refers to a concept which in Greek was called εἰaός. The word combines two aspects that modern logic differentiates: the “objective” meaning of probability (what is probable by the laws of nature) and the “subjective” meaning of plausibility (what the addressee of a message considers plausible). Aristotle's use of the word shows that the Greek concept does not draw a sharp distinction between these two elements.¹ Their intricate connection becomes evident in, e.g., his Rhetoric (2.25 1402b14-16) where he defines enthymemes. They are drawn from probability, he explains, when they are based upon things which occur regularly or are perceived to occur regularly (ἔsτι δὲ τὰ μὲν ἐa τῶν ὡς ἐπὶ τὸ πολὺ e ὄντων e δοaούντων sυνηγμὲνα ἐνθυμήματα ἐa τῶν εἰaότων). The same definition combines objective, purely statistical probability (τὰ ὡς ἐπὶ τὸ πολὺ ὄντα) and subjective plausibility (τὰ ὡς ἐπὶ τὸ πολὺ δοaοῦντα).

Ancient rhetorical theory in general highlighted the subjective aspect, as is evident in the definition of εἰaός that Anaximenes (?) gives in his Rhetorica ad Alexandrum (7.3.4 1428a27-35):

Εἰaὸς μὲν ονν ἐsτιν, ον λεγομννου παϱαδείγματα ἐν ταῖς διανοίαις ἔχουsιν οἱ ἀaούοντες. λέγω δ' οεον εἴ τις φαίη τὴν πατϱίδα βούλεsθαι μεγάλην εὲναι aαὶ τοὺς οἰaείους εν πϱάττειν aαὶ τοὺς ἐχθϱοὺς ἀτυχεῖν aαὶ τὰ τούτοις ὅμοια, sυλλήβδην εἰaότα δόξει εiναι. εaαsτος γὰϱ τῶν ἀaουόντων sύνοιδεν αὐτὸς αὑτὲ πεϱὶ τούτων aαὶ τῶν τούτοις ὁμοιοτϱόπων ἔχοντι τοιαύτας ἐπιθυμίας. ὥsτε τοῦτο δεῖ παϱατηϱεῖν ἡμaς ἐν τοῖς λόγοις ἀεί, εἰ τοὺς ἀaούοντας sυνειδότας ληψόμεθα πεϱὶ τοῦ πϱάγματος ον λεγομεν· τούτοις γὰϱ αὐτοὺς εἰaός ἐsτι μάλιsτα πιsτεύειν.

Plausibility is achieved if the listeners have examples of what is said in mind. To wit, if somebody were to say that he wants his fatherland to be great and his friends to be lucky and his enemies to be unlucky and similar things, this, in short, will seem plausible. For every listener is aware that he himself has similar desires about these and like matters. Hence we must always take care in our speeches to secure our listeners' concurrence in the matters about which we are speaking. For they are likely to trust these statements most.

Anaximenes thus defines εἰaός as an attentiveness to the public's assumptions rather than an objective quality inherent in certain facts. His aim is to train future orators. Whereas a philosopher such as Aristotle wants to explore the very foundations of rhetorical persuasion and thus is interested in all aspects of εἰaός, orators will be concerned with persuading their audiences instead of speculating about the laws of nature. Yet it could be argued that the modern distinction is largely illusory. Even the most “objective” probability must appear probable to someone; if there are no observers bringing in their own experiences and presuppositions, there is no probability.

From its earliest period, Greek oratory made extensive use of this concept. It is impossible to pinpoint its exact origin²—it probably was “discovered” several times. Anastassiou's suggestion (1981, 358-60) that εἰaός was first used as a means of reaching a decision in cases of insufficient evidence and was not employed by orators as a strategy of persuasion until later³ strains the imagination. As soon as there was a case of insufficient evidence, both parties were sure to use εἰaός to further their point of view, trying to persuade the judges of the plausibility of their version. From early times, εἰaός certainly was no less a means of discourse than a tool of knowledge about the world.

One of our earliest examples for this rhetorical use comes from the Homeric Hymn to Hermes. When Apollo accuses Hermes of having stolen his cattle, Hermes answers that he, a mere baby, is not likely to be a cattle thief (οὐδὲ ἔοιaα, 265). So far, discussions of this passage⁴ have seldom mentioned one important feature: Hermes is lying. Against all appearances and probabilities, the mere baby did indeed steal Apollo's cattle. Thus not only is the author of the hymn familiar with this particular form of argument, he also knows that it can be employed deceptively. His use of the device appears to be a playful parody of arguments that he considered typical of court speeches. This seems to presuppose that at the time of the hymn's composition (which most critics date to the beginning of the fifth century),⁵ the argument from εἰaός was already well understood and had originated at least one or two generations earlier. In the speeches of the oldest Attic orator, Antiphon, we find the concept fully developed.⁶

One example will suffice to demonstrate how ancient orators utilized this form of argument. In a speech that Antiphon wrote sometime between 422 and 413 b. c.,⁷ the speaker (a man from Lesbian Mytilene whose name has not been transmitted) defends himself against allegations of having killed an Athenian named Herodes. His accusers assert that after the crime he disposed of the corpse with the help of a slave. Among the many arguments that he adduces to prove his defendant's innocence Antiphon makes him deliver this reflection (Or. 5.43):

aαίτοι τὸ εἰaός sύμμαχόν μοί ἐsτιν. οὐ γὰϱ δήπου οὕτω aαaοδαίμων ἐγώ, ὥsτε τὸ μὲν ἀποaτεῖναι τὸν ἄνδϱα πϱοὐνοηsάμην μόνος, ἵνα μοι μηδεὶς sυνειδείη, ἐν o μοι ὁ πaς aίνδυνος ἦν, ἤδη δὲ πεπϱαγμένου μοι τοῦ ἔϱγου μάϱτυϱας aαὶ sυμβούλους ἐποιούμην.

And yet, probability (εἰaός) is on my side. For I am certainly not so wretched as to plan the actual killing of the man alone in order to have no accessories while I take the whole risk, but create witnesses and advisers after I had already perpetrated the deed.

This passage is typical of εἰaός considerations in a number of ways. Its logical force rests on an enthymeme: “People are likely to seek help not after, but in (i.e., during) a dangerous situation. My opponents' version states that I did exactly the contrary of this general rule. Therefore, it is improbable: probability is on my side and against them.” Assumptions about human behavior in general are tacitly understood in the reasoning. The combination of particles οὐ γὰϱ δήπου tries to drive home the purported logical weight of the argument by “supporting a positive statement by an appeal to the impossibility of its opposite” (Denniston 1954, 267). This also shows that most forms of εἰaός considerations appeal to basic abilities of storytelling. The judges⁸ are asked to imagine a situation (the indicative verbs πϱοὐνοηsάμην and ἐποιούμην designate the unreality of this line of thought) whose implausibility is then used to convince them of the truth of the speaker's version.⁹

Εἰaός was not only mastered practically, its theory had been included in rhetorical handbooks at a very early stage. Passages in Plato and Aristotle relate that the rhetorical treatises of Corax and Tisias included discussions of εἰaός. These detailed accounts display a level of abstraction which demonstrates that the concept itself must have been well understood at the time of their writing.¹⁰ We have already seen that later rhetorical treatises such as Anaximenes' Rhetorica ad Alexandrum or Aristotle's Rhetoric give detailed accounts of εἰaός.

Modern scholarship has examined these ancient theories and rhetorical practice on a number of occasions.¹¹ The focus of these studies, however, has hitherto been exclusively on explicit εἰaός considerations such as the passage in Antiphon quoted above: the speaker himself points out that the argument he is making depends on considerations about the plausibility of events and stories. But this use is just the tip of the iceberg; εἰaός covers a wider ground than these argumentations. When we realize that εἰaός means “probable” as well as “plausible,” we must recognize that it is part of a larger network of devices that try to render speeches persuasive and acceptable.

My object here is to advance our understanding of the device by exploring its broader context. I propose that the concept of εἰaός needs to be examined in conjunction with theories of plausibility that have been developed for other literary and nonliterary types of discourse, especially fictional narrative. As was evident in the example quoted above, most εἰaός considerations explicitly or implicitly appeal to the audience's ability to interpret, or analogize from, storytelling. The connection between strategies of verisimilitude in fiction and in oratory was indeed perceived in antiquity. As Eden has shown in her study of Aristotelian theories (1989, 18-20; see also 69-70), Aristotle regarded probability in rhetoric and in tragic poetry as related. And Pratt's recent (1993) book analyzes the concept of verisimilitude, of “plausible lying,” in the Homeric epics.¹² Modern literary theory has also made important contributions to the field of study of verisimilitude and plausibility.¹³ In applying such ideas, I aim at a more thorough understanding of εἰaός than has hitherto been achieved.

The first section of my discussion regards the logical status of court speeches as compared to fiction. These preliminary observations show that the strategies of verisimilitude utilized in these genres are indeed comparable. In the second section I indicate how these new approaches may be productive for the assessment of εἰaός in the Greek orators, and distinguish various strategies of achieving plausibility. Finally, I suggest that the word εἰaός was particularly apt for use in rhetorical argumentation precisely because of its wide range of meaning.

ATTIC COURT SPEECHES AND FICTION

In the most systematic comparison of “factual” and fictional narrative that has been undertaken so far, Genette (1993, 54-84) comes to the conclusion that on the level of discourse there are surprisingly few factors that allow us to distinguish one from the other. However, the pragmatic differences between these two types are obviously paramount. Although Gorgias and the anonymous author of the διssοὶ λόγοι both claim that tragedy tries to deceive (ἀπατaν) its public,¹⁴ we can be certain that no fictional text is deceptive in the same manner as a nonfictional text can be. While the latter may imply an absolute claim to its own trustworthiness, fiction does not purport to be true in the same way.¹⁵ Even if we accept Samuel Taylor Coleridge's definition of poetic faith, his “willing suspension of disbelief for the moment” does not imply that for the time of our reading or listening, we believe in fiction the same way we believe in statements of fact. The social setting of the actual performance marks a crucial difference, as even those would concede who accept the (in my view, considerably problematic) theory that archaic Greek poetry was not fictional in our modern sense but rather a form of truth statement.¹⁶ A defendant before a law court had to assert the truth of what he was saying with a degree of intensity that was unknown to an epic bard or rhapsode. In many cases, he would literally speak for his life, and his chances rose or fell depending on the plausibility of his narration.

This important difference between court speeches and fictional texts can be described as one of logical status. Whereas fictional texts produce their objects instead of referring to existing things—in Genette's terminology (1993, 25), they are characterized by “pseudoreference, or [by] denotation without denotata”—the speeches are, at least ideally, referential, that is, they presuppose an extratextual reality to which they refer. However, this difference between referential and fictional texts is less clear-cut than it may appear at first sight. There exist a considerable number of borderline cases that deconstruct this nice dichotomy. Take the example of mythological characters. On the one hand, the Greeks believed that these had somehow existed and were historical persons. On the other hand, the fact that storytellers felt free to alter the traditional versions and that their audiences accepted these modifications shows that no Athenian would have viewed these myths to be as “real” (i.e., referential) as communications about everyday life. Comedy is another borderline case which poses difficult logical problems. Side by side on the same stage, the Athenian audience saw clearly fictitious characters such as Trygaeus (who flies into Olympus on the back of a dung beetle) and representations of real persons such as the poet Euripides or the general Lamachus. From a strictly logical point of view these cases present troublesome problems, because degrees of reality do not exist: something is, or is not, real. However, in our everyday lives, we master such borderline cases with ease,¹⁷ and obviously so did ancient Athenians.

Even the court speeches themselves are not free from this kind of fluctuation. At an early stage of Greek oratory, Gorgias used the same methods of argumentation that are normally employed by court speeches in his fictitious pleas for mythological characters such as Palamedes or Helen.¹⁸ Antiphon's three Tetralogies picture wholly imaginary lawsuits. Incidentally, this genre of fictional court speech, later called μελέτη, was to be immensely successful during the Hellenistic and imperial eras.¹⁹ The authenticity of the Tetralogies has been questioned many times.²⁰ As yet, no compelling reasons for regarding them as spurious have been adduced, and I suspect that many critics just feel uneasy about the strange status of these texts. Fictitious court speeches are secondary in their logical and chronological condition; they presuppose the existence of real law courts and the procedures of pleading customary in them. Hence critics instinctively regard these fictitious pleas as late phenomena (like the Hellenistic and imperial μελέται) and find it hard to believe that the oldest preserved documents of Attic oratory should be fictitious speeches. And yet, despite misgivings, this seems to be the case.

These borderline instances demonstrate that the distinction between fictional and referential texts may be problematical,²¹ but there are other factors too that bring court speeches close to fictional literature. The most important of these can be described in terms of the situation in which the speeches were delivered and received. When Athenian citizens of the fifth or fourth century convened as jurors for a trial, some of them probably knew the accused and/or the plaintiff personally (all the more so in trials which featured well-known politicians). However, only very few of them will have had in advance any peculiar knowledge of the facts about which the trial was to proceed (even though a challenge for bias does not seem to have existed in the Athenian law system).²² This meant that the reality of the facts and circumstances was accessible to the overwhelming majority of the jury only via the text of the pleadings that both parties held.²³ The only elements of the speeches that the jurors could immediately perceive as referential were mentions of people and things present in the law court, most typically in the form of οντος or οντοι, “my opponent(s) whom you see over there.” All other parts of the text merely claimed to be referential, a claim that could, due to the nature of Athenian lawsuits, not be checked against external evidence: all trials were “she says, he says” cases, as we would call it today.

It is this quality of court speeches that allows us to compare them to fictional texts and to draw on definitions of plausibility and verisimilitude that have been discussed for fiction. Many positions in modern literary theory are marked by a deep-rooted mistrust concerning the relation between texts and extratextual reality. De Man, for instance, always insisted on the rhetorical, opaque character of language, which frustrates all attempts to reach the real world via linguistically encoded information.²⁴ According to proponents of such theories, some texts appear to be more “realistic” or convey the impression of being more lifelike than other texts, not because they are somehow closer to reality itself but because they make use of certain textual strategies. These procedures, if employed successfully, create what Barthes has labeled the “referential illusion.”²⁵ They make the audience believe that the words of the text are a direct representation of the world as in itself it really is—or, to be more precise, a representation of the world such as the readers or listeners of the text believe it to be. Brinker has rightly insisted that seemingly verisimilitudinous representations always imply and presuppose the beliefs of the public.²⁶

If we take this definition seriously, verisimilitude should be described as a relation not between texts and reality, but between different sorts of discourse, as Kristeva has argued.²⁷ “True” discourse is a representation of reality, verisimilitudinous discourse a representation of this true discourse. Obviously this relation presupposes one important notion: we have to accept that there is such a thing as true discourse. If an audience takes the extreme skeptical stance that reality is totally inaccessible to human thought and speech, the notion of “truth” makes no sense anymore, and with it goes the concept of verisimilitude.

How do these reflections apply to the Attic court speeches? The ancient jurors certainly believed in the existence of truth—otherwise, the whole Athenian system of jurisdiction would have been absurd.²⁸ Yet this optimistic belief that the truth exists and can be reached via thorough examination of the existing evidence was offset by a profound skepticism toward the litigants' speeches. Both parties in a trial would attempt to depict their own discourse as honest and trustworthy while denouncing their opponents' as simply trying to simulate this truthfulness. If their own position was corroborated by εἰaός considerations, speakers would emphasize that there was no other hope of attaining the truth.²⁹ The plaintiff in Antiphon's First Tetralogy, for example, stresses that crimes are difficult to prove if their planning was careful. Hence, knowing this difficulty, the judges ought to trust every bit of plausibility of which they can catch a glimpse (Tetr. 1 α 2, γιγνώsaοντας ονν ὑμaς χϱή ταῦτα, aἄν ὁτιοῦν εἰaὸς παϱαλάβητε, sφόδϱα πιsτεύειν αὐτω). A similar point is further elaborated in Demosthenes' speech Against Androtion (22.22):

ἔsτι τοίνυν ἀνάγaη τοὺς ἐλέγχοντας e τεaμήϱια δειaνύναι δι' ων ἐμφανιοῦsι τὸ πιsτὸν ὑμῖν, e τὰ εἰaόα φϱάzειν, e μάϱτυϱας παϱρχεsθαι· οὐ γὰϱ οiόν τ' ἐνίων αὐτόπτας ὑμaς ἐsτι aαταsτῆsαι, ἀλλ' ἐὰν ἐπιδειaνύῃ τίς τι τούτων, ἱaανὸν νομίzετ' ἔλεγχον ἔχειν ὑμεῖς εἰaότως τῆς ἀληθείας ἑaάsτοτε.

By necessity, those who want to prove something must either provide evidence by which they can show you what is reliable, or tell plausible facts or provide witnesses. For in some cases, it is not possible to make you see the evidence yourselves. Accordingly, if someone manages to show any of these, you rightly consider this a sufficient evidence of the truth.

On the other hand, speakers tended to disparage εἰaός considerations which their opponents had used and point out that there was an immense difference between reality and probability. Thus the defendant in Antiphon's First Tetralogy is eager to oppose the real (ὄντως) and the probable (εἰaότως) perpetrator (as at 1 β 10, “It may be probable that I killed this man, but in reality, I did not do it,” εἰ aαὶ εἰaότως μέν, ὄντως δὲ μὴ ἀπέaτεινα τὸν ἄνδϱα),³⁰ whereas Isaeus in his speech For the Heirs of Nicostratus (4.18) distinguishes between the precise facts (τὸ ἀaϱιβές) and the plausibilities (εἰaός).

This strategy is explicitly mentioned in an interesting passage in Aristotle's Rhetoric. In the section on artless proofs he outlines a strategy of how to strengthen one's position against opponents whether they have witnesses or not.³¹ Either case requires a special kind of argument (1.15 1376a17-23):

πιsτώματα δὲ πεϱὶ μαϱτυϱιῶν μάϱτυϱας μὲν μὴ ἔχοντι, ὅτι ἐa τῶν εἰaότων δεῖ aϱίνειν aαὶ τοῦτ' ἐsτὶ τὸ “γνώμη τη ἀϱίsτη,” aαὶ ὅτι οὐa ἔsτιν ἐξαπατῆsαι τὰ εἰaότα ἐπὶ ἀϱγυϱίῳ, aαὶ ὅτὶ οὐχ ἁλίsaεται τὰ εἰaότα ψευδομαϱτυϱιῶν· ἔχοντι δὲ πϱὸς μὴ ἔχοντα, ὅτι οὐχ ὑπόδιaα τὰ εἰaότα, aαὶ ὅτι οὐδὲν ἂν ἔδει μαϱτυϱιῶν, εἰ ἐa τῶν λόγων ἱaανὸν ων θεωϱῆsαι.

For those who do not have witnesses, the strategy is to say that one has to judge according to probabilities and that this is the meaning of the phrase “according to my best judgment,”³² and that probabilities cannot be bribed or convicted of bearing false witness. For those who have witnesses and are pitted against those who do not, the strategy is to say that probabilities incur no responsibility and that there would be no need for witnesses if one could investigate the truth by looking at the arguments alone.

We can be certain that the Athenian judges would have heard both of these arguments over and over again, and that in most cases these appeals to different standards of credibility would cancel each other out. Unfortunately, today we rarely have the arguments of both sides available, but instead have to reconstruct the other side's case by using the transmitted speech.³³ Athenian judges, however, confronted with two diverging accounts and assessments of the circumstances, knew precisely that at least one of the parties (and frequently both, we may assume)³⁴ was perfectly apt and inclined to distort the facts in order to win the case. Accordingly, jurors certainly were experienced and shrewd listeners and evaluators of court speeches and hardly let themselves be deceived by versimilitudinous narratives, no matter how plausible and convincing they might appear.³⁵ It is therefore hardly an exaggeration to say that Athenian judges were no less skeptical in their evaluation of the relation between text and reality than postmodern literary critics. (Even today, it is frequently said that because of their profession, jurists inevitably turn cynical.)

Professional speechwriters knew this, of course, and therefore tried to anticipate this skepticism in their orations. Lysias' third speech contains an incredible tale of strife about a young male prostitute, his abduction, and the ensuing hubbub that ended in a free-for-all. At one point, the speaker self-consciously comments about his own narrative (3.37): “It is not the case that these words are plausible, but the events were completely different, no …,” οὐ τοίνυν ταῦτα εἰaότα < μέν>, ἄλλως δὲ πεϱὶ αὐτῶν πέπϱαaται, ἀλλὰ … Only after this interruption does he resume his tale. The speaker (and hence, Lysias) is thus aware that jurors might resent a narrative that was too perfect and smooth because they were conscious of how much tampering with the bare facts such an account might require.

In general, however, litigants would aim at maximum plausibility in their speeches. Therefore, they exerted great effort that their discourse would appear a “truthful” one. Our next step is to investigate how this was achieved. Precisely which devices made a narrative appear to be similar to a true discourse? In this case, too, concepts developed for the analysis of fictional texts can help us understand how Athenian speechwriters worked toward giving this desirable impression.

STRATEGIES OF VERISIMILITUDE

First of all, it is obvious that every narrative aiming at verisimilitude has to fulfill certain basic expectations about the behavior of human beings and about the material world in general. In an illuminating discussion of the subject Culler tried to establish two sets of such rules, differentiating between “natural” and “cultural” vraisemblance. He himself acknowledged that the latter is “in some cases difficult to distinguish from the first.”³⁶ Rather, we should consider these categories as a continuum, any partition of which is bound to be arbitrary. At one extreme are notions that really seem so universal as to transcend any given culture: “We speak of people as having minds and bodies, as thinking, imagining, remembering, feeling pain, loving and hating, etc., and do not have to justify such discourse by adducing philosophical arguments” (Culler 1975, 140). At the other extreme are convictions that are valid for an identifiable group of people only. For those who believe in the existence of UFOs, flying saucers comprise as much a part of reality as trees and stones. Most cases, however, are somewhere in between these extremes, as a brief look at the Attic orators will show.

In one of his speeches (23.15) Lysias makes the argument that no citizen of Plataea would ever choose to live in Thebes because of the deep-seated hostility between these two cities. This clearly is a topical argument, valid only in certain historical, social, and geographical circumstances, and we may presume that the speaker and the audience were aware of this limitation. On the other hand, when Attic orators adduced family relationships as motivations for certain behavior,³⁷ they likely assumed that they were referring to eternal laws of human nature. But the very fact that the speechwriters mention these plausibilities quite so often and quite so emphatically³⁸ should make us wary. In this respect, fourth-century Athens was probably not very different from our own culture. Family values are easy to utilize in manipulating the public's emotions; their real importance in everyday life, however, is not always easy to assess. Culler's distinction between “natural” and “cultural” verisimilitude, then, is not always easy to grasp; but what counted in Athens was the acceptance that both forms of plausibility found with the judges. When they were listening to the speeches in court, they had no time for lengthy philosophical speculation about whether a narrative's verisimilitude was based on natural or cultural factors. If they could be brought to accept, as being reasonable, a description of laws governing human actions, the speaker had succeeded in his task.

Most of the time these general laws are tacitly implied in the narratives, yet we also find them explicitly quoted in the course of the argument. A similar phenomenon in modern fiction was analyzed by Genette in a paper first published in 1968: “The verisimilitudinous narrative is a narrative whose actions correspond, like applications or examples, to a body of maxims that the public accepts as true. Yet most often, these maxims, precisely because they are accepted, are merely implied. The relation between the verisimilitudinous narrative and the system of verisimilitude to which it refers is hence essentially mute.”³⁹ Let us consider two examples that show such accepted maxims at work in order to produce verisimilitude. Lysias' twentieth speech concludes with an emotional appeal to the judges (36): “We would indeed suffer horrible things if we were saved by our enemies who were likely (εἰaός) to preclude our rescue, but will not even obtain salvation from you” (δεινὰ δ' ἂν πάθοιμεν, εἰ ὑπὸ τῶν πολεμίων μὲν ἐsώθημεν, οuς εἰaὸς εν διαaωλύειν μὴ sozεsθαι, παϱ' ὑμῶν δὲ μηδὲ εὑϱηsόμεθα τὸ sωθῆναι). In his speech Against Euthynus Isocrates uses an intricate εἰaός consideration to demonstrate that Nicias, on whose behalf he is speaking, did indeed deposit a large sum of money with Euthynus (21.6):

'Αλλὰ μὴν aαὶ ἐξ αὐτοῦ ἄν τις τοῦ πϱάγματος γνοίη ὅτι πολὺ μaλλον εἰaὸς ηέν Εὐθύνουν λαβόντα ἐξαϱνεῖsθαι e Nιaίαν μὴ δόντα αἰτιasθαι. δῆλον γὰϱ ὅτι πάντες aεϱδους ενεa' ἀδιaοῦsιν. οἱ μὲν οεν ἀποsτεϱοῦντες ενπεϱ ὲνεa' ἀδιaοῦsιν ἔχουsιν, οἱ δ' ἐγaαλοῦντες οὐδ' εἰ λήψεsθαι μελλουsιν ἴsαsιν.

The matter itself shows that Euthynus is much more likely to have received the money and denied it than Nicias [is] to accuse him although he did not deposit it. For it is obvious that everybody commits crimes because of the profit. Well, those who steal something hold the object of their deeds while those who complain about it do not even know if they are going to recover it.

These passages exemplify the reference to rules which both speaker and audience unquestioningly accept as normative and compulsory, as being completely obvious. (In fact, “obvious” would be a good translation for εἰaός in these sentences.) To use the formulation of Anaximenes' quoted earlier: Every listener immediately “had examples of these general maxims in mind.” “Helping friends and harming enemies” is such a rule, as is the assumption that people will commit a crime in order to reap profit from it.

However, there is a major difference between these cases and Genette's analysis. Whereas fictional narratives generally prefer merely to imply these rules, the Attic orators frequently quote the maxims that underlie their reflections. One may wonder why this is the case. Developing an idea of Barthes's, Culler has shown (1975, 195-96) that among other things, these references to implicit maxims can be interpreted as a way “of evoking and solidifying the contract with the reader, insisting that the narrator is only a more knowledgeable version of the reader and that they share the same world to which the language of the novel refers.” This “voice of reading” effectively merges the reader's and the narrator's points of view, thus achieving verisimilitude. This mechanism could not have been employed in the court speeches, where the “narrator” is not a textual function but was an actual person that narratees (the audience, judge, jury) could see and hear. Instead of creating the illusion that the text spoke with the voice of reading, Attic speechwriters were at pains to convince their audiences that the speaker was one of their own kind. Hence explicit reference to these rules served a double purpose. First, there was the logical import. By showing that their cause was supported by such (literally commonsense) arguments, speakers demonstrated that they were obviously right. But I would argue that the psychological impact is at least as important: the citation of these unwritten rules emphasized that both the speaker and the jurors shared common ground, that their whole outlook on the world and on human behavior was similar. Emphasizing this shared ideology will not only have rendered the narratives and arguments plausible, it will also have created a feeling of togetherness and sympathy.

Of course, such references to commonly accepted moral rules are particularly suitable for manipulating the listeners. An especially conspicuous case is the allusion to the “laws of our fathers” or “the customs of our ancestors.” Appeals to traditional values seem to have impressed Athenian audiences to a great extent, leaving no room for disagreement or doubt.⁴⁰ Thus orators were apt to use these references precisely at those points where their argumentation was quite weak. At one point in his speech On the False Embassy Aeschines quotes a detail of procedure which “our fathers” (οἱ πατρϱες ἡμῶν, 2.87) had instituted for lawsuits at the Palladion. This has nothing whatsoever to do with the case at hand, but the speaker probably hoped that the unassailability of the “fathers” would delude the jurors, thus obscuring his faulty logic and preparing the way for the ensuing abuse against his opponent (Demosthenes). Similar strategies are sometimes applied by modern politicians.

Our last strategy to consider for producing verisimilitude is of a somewhat different nature. It is particularly conspicuous in nineteenth-century realist fiction. The narrative dwells at length on small details that seem completely superfluous for the plot—descriptions of clothing or furniture are an example of this. Such elements pretend to elude the clear functionality that readers expect of narrative, thereby proclaiming that they represent not artificially planned narrative, but reality itself. Barthes, who described this “reality effect” in a famous article published in 1968, was of course not the first to see this method of achieving realism, but he analyzed for the first time the underlying reasoning: seemingly nonfunctional descriptions and characterizations appeal to the audience's knowledge that in reality, most things are accidental.⁴¹ By a curious yet explainable reversal of strict logic, this knowledge in turn leads to the conviction that accidental details are a sign of reality.

Of course, the knowledge that vivid details endow a tale with verisimilitude has existed ever since the beginning of narrative. Homer's Odysseus understood this principle thoroughly when he embellished his deceptive tales with all kinds of fictitious details. Appearently, it was to this quality of the Homeric epics that Aristotle referred in his Poetics when he wrote (24 1460a18-27) that Homer “has taught the other poets how to lie fittingly.” This, at least, seems to be the most plausible explanation of a passage which is rather difficult to understand because of Aristotle's brevity of expression.⁴²

Aristotle's contemporary Aeschines gives further proof that the function of graphic, unexpected details in lending verisimilitude to fictional narratives was well understood in that time. Two passages in his speeches against Demosthenes accuse his opponent of counterfeiting “truthful” discourse by imitating its procedures (2.153):

ἡγεῖται δ' ὅταν τι ψεύδηται τῶν λόγων ὅϱaος aατὰ τῶν ἀναιsχύντων ὀφθαλμῶν, aαὶ τὐ <μὴ> γεγενημένα οὐ μόνον ὡς ἔsτι λέγει, ἀλλὰ aαὶ τὴν ἡμεϱαν, ἐν εφηsι γενέsθαι· aαὶ πϱοsτίθηsίν τινος ὄνομα πλαsάμενος, ὡς ἔτυχε παϱών, μιμούμενος τοὺς τἀληθῆ λέγοντας.

Whenever he tells a lie, his speech begins with an oath on his shameless eyes, and not only does he assert the existence of what never happened, but he also tells the day on which it supposedly happened. And he adds somebody's name whom he claims was present, making it up, and so imitates those who tell the truth.

The second passage is similar (indeed, it repeats some sentences of this first). Here, Aeschines contrasts Demosthenes' behavior with that of other liars (3.99):

οἱ μὲν γὰϱ ἄλλοι ἀλαzόνες, ὅταν τι ψεύδωνται, ἀόϱιsτα aαὶ ἀsαφῆ πειϱῶνται λέγειν, φοβούμενοι τὸν ἔλεγχον· Δημοsθένης δ' ὅταν ἀλαzονεύηται πϱῶτον μὲν μεθ' ὅϱaου ψεύδεται, ἐξώλειαν ἐπαϱώμενος ἑαυτω, δεύτεϱον δε, ἅ εν οηδεν οὐδiποτε ἐsόμενα, τολμέ λέγειν ἀϱιθμῶν εἰς ὁπότ' ἔsται, aαὶ αν τὰ sώματα οὐχ ἑώϱαaε, τούτων τὰ ὀνόματα λέγει, aλέπτων τὴν ἀaϱόαsιν aαὶ μιμούμενος τοὺς τἀληθῆ λεγοντας.

When the other boasters tell lies, they try to make their speech vague and imprecise because they are afraid of being disproved, but whenever Demosthenes boasts, first he tells his lies under oath, conjuring destruction upon himself; second, he dares to tell what he knows will never happen and actually calculates the time when it will happen, and he tells the names of people whose bodies he has not seen, deceiving his audience and imitating those who tell the truth.

Aeschines' charges show that the strategy of achieving verisimilitude by adding small, significant details was not only mastered in practice but also understood in theory. He accuses Demosthenes of making nonreferential language look like truthful discourse, or, in the words of Homer and Hesiod, of inventing “lies that are similar to the truth.” The strategy of mentioning specific dates and places and giving specific names is one of the most persistent ways of achieving verisimilitude, employed in countless fictional narratives from antiquity to the present day.⁴³

Later (post-Aristotelian) rhetorical and poetical theory identified this device as one of the main virtues of the “simple style” (λόγος ἀφελής) and termed it ἐνάϱγεια.⁴⁴ This quality is usually defined as “a power of conveying the things about which one speaks to the senses of the audience” (Dion. Hal. Lys. 7, δύναμίς τις ὑπὸ τὰς αἰsθήsεις ἄγουsα τὰ λεγόμενα),⁴⁵ and the theorists frequently affirm that this power is achieved by giving important details of an event. Dionysius explains that ἐνάϱγεια “is achieved by a grasp of circumstantial detail” (γίγνεται δ' ἐa τῆς τῶν παϱαaολουθούντων λήψεως). “Demetrius” makes the similar statement that “ἐνάϱγεια is produced by precision and neither omitting nor taking away anything” (On Style 209, γίνεται δ' ἡ ἐνάϱγεια πϱῶτα μὲν ἐξ ἀaϱιβολογίας aαὶ τοῦ παϱαλείπειν μηδὲν μήδ' ἐaEτὲμνειν).⁴⁶

Barthes's definition of the “reality effect,” however, goes beyond graphic and surprising details. He has rightly insisted on the seeming lack of narrative function: “the ‘real’ is supposed to be self-sufficient, … it is strong enough to belie any notion of ‘function’” (1986, 147). This particular aspect, however, seems to be missing from Attic oratory. Despite Aeschines' assertions that orators may counterfeit such details just for the sake of verisimilitude, a closer look at elements that seem to belong in this category shows that this is not quite the case. Details always seem to have more functions than merely achieving the reality effect.

Let us consider an example from Lysias' speech Against Agoratus. When the speaker mentions a trial under the Thirty, he gives a vivid description of the scene. The Thirty were sitting as judges on an elevated platform; before them, two tables had been set.⁴⁷ At first this graphical description might be understood as being nonfunctional in the narrative, thus serving as an instance of reality effect. But further on, it turns out that the setting served to prevent a secret ballot. Every judge had to cast his vote openly on the tables, so that the Thirty could observe every detail from their raised platform. The description, then, is functional in the accuser's plea; it serves to demonstrate that the Thirty (and nobody but the Thirty) were fully responsible for every sentence that was passed by this jury. Thus what at first sight seems to be a superfluous detail really is functional in the argumentation.

Other elements may indeed be nonfunctional in regard to the narrative and the logical structure but still serve well-defined purposes. Lysias' speech Against Eratosthenes (the only one he made in his own name) depicts how the Thirty robbed his family of all their possessions. One of them, Melobius, did not even refrain from ripping a pair of golden earrings from the ears of the wife of Polemarchus, Lysias' brother. This detail is clearly intended to create an emotional climax: the earrings are certainly not necessary in the narrative, and they do not prove any of Lysias' charges. They simply serve to highlight the Thirty's cruelty and brutality, which did not even stop at robbing a woman. This emotional function is further strengthened by the emphasis which is put on the history of those earrings: Polemarchus' wife had been wearing them when she entered his house for the first time—that is, on the day of their wedding (Lys. 12.19, τῆς γὰϱ Πολεμάϱχου γυναιaὸς χϱυsοῦς ἑλιaτῆϱας οuς ἔχουsα ἐτύγχανεν ὅτε τὸ πϱῶτον ἦλθεν εἰς τὴν οἰaίαν, Μηλόβιος ἐa τῶν Ὤτων ἐξείλετο).⁴⁸ Here again, this detail serves no purpose in the logical progress of the narration, but it is functional as an appeal to the jurors' pity and indignation.

We must assume that emotional passages such as this one were highly efficacious. Athenian jurors were not professionally trained jurists, so they may have been especially susceptible to emotional appeals.⁴⁹ This also explains the very large amount of personal abuse that was common in the court speeches.⁵⁰ Not only did it create an atmosphere that was favorable to the speaker, it also suggested why such an evil creature as the opponent was apt to commit crimes or to accuse maliciously. This could again be described as an enthymeme or internal εἰaός consideration. But it is obvious that, at the same time, the tear-jerking story of the earrings is meant to enhance the verisimilitude of the tale: it is one of those small details that, as Diderot said, are the hallmark of truth.⁵¹ What we find here, then, is not an instance of reality effect; it is rather an example of an element that functions on more than one level.

In general, it seems more typical of the Attic court speeches to display these multifunctional elements, rather than aiming at pure reality effect. One last example, from Demosthenes, will demonstrate this point. In his speech Against Conon the speaker, Ariston, accuses Conon and his son Ctesias of beating him and taking away his cloak. He pictures the scene in a vivid narrative. He was going for a walk in the agora with a friend when they encountered Ctesias. On their way back, they were attacked by several men who had been lying in wait for them. When they had beaten up Ariston, Conon and his son reviled him as he was lying in the mud. Then the narrative emphasizes one particularly graphic detail (Dem. 54.9):

aαὶ τὰ μὲν ἄλλα aαὶ βλαsφημίαν ἔχει τινὰ aαὶ λέγειν ὀaνήsαιμ' ἂν ἐν ὑμῖν ἔνια, ὃ δὲ τῆς ὕβϱεώς ἐsτι τῆς τούτου sημεῖον aαὶ τεaμήϱιον τοῦ πaν τὸ πϱaγμ' ὑπὸ τούτου γεγενῆsθαι, τοῦθ' ὑμῖν ἐϱῶ· eδε γὰϱ τοὺς ἀλεaτϱυόνας μιμούμενος τοὺς νενιaηaότας, οἱ δὲ aϱοτεῖν τοῖς ἀγaῶsιν αὐτὸν ἠξίουν ἀντὶ πτεϱύγων τὰς πλευϱάς.

The rest was so obscene that I do not want to repeat some of it before you, but I will say one thing which is a demonstration of his insolence and a proof that the whole affair has been done by him. He crowed, imitating fighting cocks that have won a battle, and his friends made him beat his elbows against his sides like wings.

The speaker is eager to point out that this especially vivid description is not there for its own sake, but actually serves a purpose, or rather, several purposes. On the one hand, Conon's derisive gesture is a token of his overall character, his violence (ὕβϱις). This function of this detail can be attributed to the emotive level. It will predispose the jurors to condemn such a man, and simultaneously it will make more plausible the assumption that he is indeed capable of performing the deeds with which he is being charged (thus serving as an implicit εἰaός consideration). Furthermore, this telling detail shows that Conon was indeed responsible for this misdeed, that he had orchestrated the whole affair. This would make it functional on the strictly logical and juridical level. Ariston of course does not mention that the vividness of this picture is also meant to enhance the verisimilitude of his tale, but it seems obvious that this was achieved, too. The detail of Conon crowing like a cock and flapping his arms rings utterly true: it appears just too freakish to be invented. In view of the theoretical awareness of similar procedures displayed in the passages from Aeschines quoted above, it appears improbable that this was just a coincidental by-product of a logically necessary detail. Instead, we must assume that Demosthenes went out of his way to emphasize the other (seemingly more acceptable) functions of this element in order to disguise the effect he ultimately intended.

Several reasons could be adduced to explain why the Attic speechwriters used these multifunctional elements rather than instances of straight reality effect. First of all, the limited time that was assigned to the speakers in a lawsuit made it impossible to develop narratives at a leisurely pace. They had to proceed briskly; every element had to be functional.⁵² Digressions from the main points of the case (ἔξω τοῦ πϱάγματος λέγειν) were formally forbidden before the Areopagus, and apparently even in private cases the parties had to swear an oath to keep to the point.⁵³ It is unclear whether (and how) this rule could be enforced, but it may have contributed to the avoidance of nonfunctional elements. Finally, the importance of emotionally affective pleas may explain why speechwriters would include vivid details: they could simultaneously achieve other emotive aims with these elements.

THE MEANINGS OF EIKOE

Thus far I have presented εἰaός in a broader context by showing that the concept is one element in a network of strategies that lend plausibility to court speeches. In illustrating one last aspect of εἰaός, I now concentrate on the Greek term itself rather than on the abstract concept. The words ἔοιaε and εἰaός ἐsτι have a wide range of meanings.⁵⁴ This quality helps explain why εἰaός considerations indeed played such an important role in the Attic court speeches.

Synodinou (1981) would distinguish two major meanings of εἰaός: the phenomenological and the deontological. The first of these would simply describe outward appearances, the second that which is morally right. But a reviewer (Dalfen 1985) rightly pointed out that these distinctions are alien to the Greek concept. Depending on the context, εἰaός ἐsτι can mean anything from “it appears” to “it is obvious” or even “it is proper,” “it is legitimate.” The reason for this breadth of meaning seems to be that things which occur very often are also likely to occur, and their occurrence comes to be regarded as justified. Apparently the psychological connection between the word εἰaός and the idea of justice was so close that it induced the speechwriters to combine the terms δίaαιον and εἰaός quite often (see, e.g., Ant. 5.49; Isoc. 18.37; Dem. 18.294, 20.36, 40.5).⁵⁵

Although these combinations might have arisen from subconscious associations, the Attic orators were also capable of taking full, conscious advantage of this connection. They liked to make stealthy transitions from one of these meanings to the other. Two examples will illustrate this usage. In his eighth speech Isaeus tries to support his clients' claim to an inheritance; this involves (as usually in his speeches) a complicated question of kinship. Here is how the speaker attempts to prove that he is far more credible than his opponents (8.14):

Τίνας εἰaὸς εἰδέναι τὰ παλαιά; δῆλον ὅτι τοὺς χϱωμένους τω πάππῳ. μεμαϱτυϱήaαsι τοίνυν ἀaοὴν οέτοι. τίνας <δ'> εἰδεναι τὰ πεϱὶ τὴν ἔaδοsιν τῆς μητϱὸς ἀνάγaη; τοὺς ἐγγυηsαμένους aαὶ τοὺς ἐaείνοις παϱόντας ὅτε ἠγγυῶντο. μεμαϱτυϱήaαsι τοίνυν οἵ τε Nαυsιμενους πϱοsήaοντες aαὶ οἱ τοῦ ἐμοῦ πατϱός. τίνες δὲ οἱ τϱεφομενην ἔνδον aαὶ θυγατεϱα οεsαν εἰδότες γνηsίαν Κίϱωνος; οἱ νῦν ἀμφιsβητοῦντες ἔϱγῳ φανεϱῶς μαϱτυϱοῦsιν ὅτι ταῦτ' ἐsτὶν ἀληθῆ, φεύγοντες τὴν βάsανον. ὥsτε οὐ δήπου τοῖς ἡμετέϱοις ἂν ἀπιsτήsαιτε εἰaότως, ἀλλὰ πολὺ μaλλον τοῖς τούτων μάϱτυsιν.

Who is likely (εἰaός) to know about these events of the past? Obviously those who were intimate with my grandfather. Well, they have given evidence of what they have heard. Who is likely to know about my mother's giving in marriage? Obviously those who betrothed her and those who were present at the betrothal. Well, the relatives of Nausimenes and my father's relatives have submitted testimony. Who is likely to know that she was reared in the house and that she is a legitimate daughter of Kiron? Those who try to deny this fact now are indeed giving evidence of its truth because they shy away from having their slaves questioned. Therefore, you definitely ought to (εἰaότως) disbelieve not our witnesses, but rather theirs.

The opening sentence of this section carries the keyword εἰaός, which suggests to the listeners that a consideration of probability will follow. At first, this assumption seems to be justified. The two εἰaός considerations that ensue are rather commonplace: eyewitnesses are most likely to have knowledge of what happened when they were present. The third element is a bit more complicated but still remains within the limits of probability: if someone denies a fact but is not willing to have his slaves questioned under torture (βάsανυς),⁵⁶ although they certainly know about this particular fact, he is likely to be concealing something and to be lying. These three sentences all use εἰaός in its phenomenological sense. The concluding occurrence, however, is different. The speaker still tries to make it look as though he was continuing his εἰaός arguments. He begins the sentence with ὥsτε, thus insinuating that it represents a logical conclusion of the foregoing arguments, and he uses the particle δήπου, which we encounter very often in εἰaός considerations.⁵⁷ The following εἰaότως, however, has shifted its meaning. It no longer expresses what is likely, but what ought to be done; the potential ἂν ἀπιsτήsαιτε is really a hidden request to the judges.⁵⁸ The speaker thus intimates that the foregoing reflections on εἰaός make it inevitable for the judges to accept his version of the case and to give judgment in his favor.

We may observe a similar strategy in the speech which Andocides pronounced in order to be allowed to return from exile after the oligarchic coup of 410 had failed. He had to convince the assembly of the newly restored democracy that he was not an enemy of the regime. He could not claim to have deserved well of democracy himself, but he was able to refer to one of his forebears who had played an active role in fighting against the “tyrants” (2.26):⁵⁹

ὁ τοῦ ἐμοῦ πατϱὸς πϱόπαππος Lεωγόϱας sταsιάsας πϱὸς τοὺς τυϱάννους ὑπὲϱ τοῦ δήμου, ἐξὸν αὐτω διαλλαχθέντι τῆς ἔχθϱας aαὶ γενομηνῳ aηδεsτη ἄϱξαι μετ' ἐaείνων τῶν ἀνδϱῶν τῆς πόλεως, εἱλετο μaλλον ἐaπεsεῖν μετὰ τοῦ δήμου aαὶ φεύγων aαaοπαθεῖν μaλλον e πϱοδότης αὐτοῦ aαταsτῆναι. ὥsτ' ἔμοιγε aαὶ δaὰ τὰ τῶν πϱογόνων ἔϱγα εἰaότως ὑπάϱχει δημοτιaω εεναι, εἴπεϱ τι ἀλλὰ νῦν γε φϱονῶν τυγχόνω. εν aαὶ ενεaα εἰaὸς ὑμaς, ἐὰν χϱηsτὸς oν ἀνὴϱ εἰς ὑμaς φαίνωμαι, πϱοθυμότεϱόν μου ἀποδὲχεsθαι τὰ πϱαττόμενα.

Leogoras, my father's great-grandfather, led a democratic revolt against the tyrants. And even though he could have chosen to come to terms with them, to become their in-law and to rule the city together with these men, he preferred to go into exile together with the democrats and to suffer the hardships of banishment rather than become a traitor to his citizens. Accordingly, on account of my ancestors' deeds, it is probable (εἰaότως) that I, too, am of democratic convictions, if I have finally regained my senses. Therefore, it is fitting (εἰaός), if it is obvious that my attitude toward you has been excellent, that you should be more inclined to approve my policy.

As was the case in the previous example, the last sentence of this passage begins with an εἰaός consideration, the first word (ὥsτε) again implying that this reflection is merely a conclusion of the foregoing narrative. The democratic convictions of Leogoras, his forefather, make it probable that Andocides himself should side with the democrats. By the introductory ων aαὶ ενεaα, the following sentence is again marked as giving the logical result of what preceded. But again, the following εἐaός merely feigns continuation of the same train of thought: its actual meaning here is “it is legitimate,” “it is fitting.”

In both cases (to which other occurrences could be added),⁶⁰ the speakers use this device to manipulate their listeners. The repetition of the word εἐaός and the emphasis on logical markers such as ὥsτε or εν aαὶ ωνεaα mask the fact that there is only the slightest logical connection between these steps. We can safely assume that this rhetorical device, which was made possible by the wide range of meanings the concept εἰaός conveyed, was rather successful. Judges would not realize that they were deceived by a specious reasoning. If a speaker could prove in this way that he was supported by εἰaός, that εἰaός was on his side, as Antiphon expressed it (τὸ εἰaὸς sύμμαχόν μοί ἐsτιν, 5, 43), this meant that he was supported by probability as well as by justice. Of course, this alone would never be decisive for the outcome of a trial, but it helps explain why the Attic orators used arguments from probability so often and why Greek rhetoric paid so much attention to them.

CONCLUSION

The concept of εἰaός was utilized by the Attic speechwriters in various ways to obtain various effects. Court speeches can indeed be compared to fictional texts because their performance renders dubious the very referentiality on which their persuasion depends. Convincing Athenian judges that the pleadings they were listening to represented reality certainly was no easy task. The orators employed textual strategies to achieve this. Three such strategies have been analyzed here: (1) invoking “natural” and “cultural” expectations about reality and human behavior, (2) the quotation of accepted maxims which lend authority to the speaker and his claims, and (3) introducing into the presentation significant details that work on several levels of the discourse to enhance its credibility and emotional impact. Finally, we have seen how speechwriters exploited the broad range of meanings of the term εἰaός to manipulate their audiences' response.⁶¹

These pages are only a small contribution to an immensely wide field. Recently, Attic oratory has risen from a long phase of neglect, and work on these texts with modern methodologies has only just started. I suggest that one line for future research could indeed be the question how the speechwriters convinced their audiences by creating plausible narratives. Given the importance of εἰaός in the Greek judicial system, this would teach us not only about oratory itself but also about Greek culture at large and the way it perceived its reality.

Notes

1. Pace Eden, who claims (1989, 115-19) that Aristotle's use of εἰaός was almost exclusively based on objective probability. Against, see Warnick 1989; Halliwell 1986; O'Sullivan 1995. Cf. also Aristotle's definition of εἰaός in Analytica Priora 2.27 70a4-6.

   I should note, here at the outset, that for my texts I have drawn on standard editions from a number of publishers. The translations are all my own.

2. See, e.g., Johnstone 1996, 12; or López Eire (1994, 48-49), who tries to demonstrate that the concept goes back to Parmenides and the Eleatic school of thought. Our evidence is just too slight to warrant such conclusions.

3. A similar argument has been advanced by Cole (1991a, 96-97).

4. See, e.g., Kennedy 1963, 40-41.

5. It is difficult to give a precise date for the Hymn to Hermes; see Görgemanns 1976, 115-17. Szepes 1980 fails to convince me.

6. Solmsen (1931) was thus wrong to argue that these speeches indicate a transitional period from an earlier stage of oratory, which relied exclusively on nonartistic proofs such as oaths or witness accounts, to the later system, which used arguments from probability. Cf. Cole 1991a, 96-97. Against Solmsen's thesis see also Due 1980, 11-14; Goebel 1983, 49-55; Gagarin 1990.

7. For the date see Schindel 1979; on the speech and the underlying case see further Erbse 1977; Heitsch 1984, 33-89; Manuwald 1995; Carawan 1998, 316-21.

8. I am aware of the difficulties in translating the Greek διaαsτής: both “judge” and “juror” are misleading, because they suggest a similarity to modern institutions (see Todd 1993, 69, 82-83). However, the inelegance of retaining the untranslated Greek term seemed to outweigh this imprecision; hence I ask my readers to keep the problem in mind whenever they find the words “judges” and “jurors.”

9. Οὐ γὰϱ δήπου is regularly used in this sense. Denniston (1954, 267) notes that in Plato it occurs most often with a following γε; in the orators, however, it seems to be more frequent without the γε. For examples with γε see, e.g., Dem. 20.35, [49.52], 53.1 (a passage very similar to the one discussed above), 57.48; Lys. 4.15. For examples without γε Dem. 6.36, 18.13, 19.204, 21.121, 23.86, 27.45, 37.18, 38.8, 39.6 (again, a very similar use), 57.25, [58.56]; Lys. 10.8, 12.27, 13.18, 14.36, 21.18, 24.13, 31.28; Isoc. 16.44; Is. 1.46, 6.46. In all these occurrences it is clear that the listeners are invited to imagine an alternative story line.

10. The passages are Phaedr. 272d-273c and Rhet. 2.24 1402a17-28. Plato attributes these ideas to Tisias; Aristotle, to Corax. Much ink has been spilled on the question whether what we have here are references to two different passages or whether one of our authorities blundered; see Goebel 1983, 109-35; Goebel 1989, 45-53; Gagarin 1994, 50-57; López Eire 1994, 37-40. An ingenious solution to the mystery has been offered by Cole (1991b), who suggests that Tisias and Corax really were the same person; see also Kennedy 1994, 34. I cannot enter here into the argument whether Tisias and Corax wrote rhetorical handbooks proper or merely published examples of devices and model speeches.

11. Esp. Fairchild 1979; Anastassiou 1981; Goebel 1983; Goebel 1989; Warnick 1989; Gagarin 1990, 20-30; Gagarin 1994; López Eire 1994, 47-52; O'Sullivan 1995; Grimaldi 1996, 38; Gondos 1996.

12. See Pratt 1993, 91, on Od. 19.203, where Odysseus' tales are described as “lies which are similar to the truth” (ψεύδεα ἐτύμοιsιν ὁμοῖα): “this is an accurate description of the verisimilitude that lends them [lies] plausibility.” See also Pratt 106-13 on Hes. Theog. 22-35; also Bowie 1993.

13. For general discussions see, e.g., the special issue of Communications 11 (1968) mentioned above (opening paragraph), which includes papers by T. Todorov, G. Genette, C. Metz, G. Genot, J. Kristeva, R. Barthes, V. Morin, J. Gritti, O. Burgelin, and M.-C. Boons; see also the articles by T. Todorov, I. Watt, L. Bersani, R. Barthes, M. Riffaterre, and P. Hamon collected in Littérature et réalité (1982); also Culler 1975, 131-60; Riffaterre 1990.

14. Gorgias, fr. B23 VS, ὅ τε ἀπατήsας διaαιότεϱος τοῦ μὴ ἀπατήsαντος aαὶ ὁ ἀπατηθεὶς sοφώτεϱος τοῦ μὴ ἀπατηθεντος, “He who deceives is more righteous than he who does not deceive, and he who is deceived is wiser than he who is not deceived”; διssοὶ λόγοι, 3.10, ἐν γὰϱ τϱαγῳδοποιίᾳ aαὶ zωγϱαφίᾳ ὅsτις πλεῖsτα ἐξαπατε ὅμοια τοῖς ἀληθινοῖς ποιρων, οατος ἄϱιsτος, “Whoever deceives most in the art of tragedy and of painting, producing works that are similar to reality, is the best.” On these passages see Finkelberg 1998, 12, 177-81.

15. A vast amount of work has been published on the question in what respect the term “truth” can be applied to fictional works of art. Here I must refer to only a few items which I consider particularly enlightening: Ingarden 1973, 300-304; Searle 1979, 58-75; Walton 1990, esp. 70-105; Riffaterre 1990; Genette 1993. The discussion in one of the most recent contributions, Lamarque and Olsen 1994, did not strike me as very helpful.

16. See esp. Detienne 1996; Rösler 1980. Against this view see the compelling arguments in Bowie 1993; Finkelberg 1998; Pratt 1993, esp. 113: “We tend to behave in speaking of the invention of fictionality as though the history of narrative proceeds from fundamentalism to an ability to appreciate invented narrative, but in fact literal-mindedness may be the deadly product of our own culture.” See also Walton 1990, 95-98, an excellent discussion of the truth value in myths from a philosophical perspective.

17. For a highly entertaining overview see Watzlawick 1976.

18. On the Encomium of Helen see now Schiappa 1996; on Gorgias' skeptical view of the relation between language and reality see Jarratt 1991, 53-57.

19. On μελέται see Kennedy 1974; Russell 1983; Swain 1996; Schmitz 1997.

20. The most recent contributions are (against authenticity) Sealey 1984; Carawan 1993; Carawan 1998, 171-215; (for authenticity) Goebel 1983, 15-16; Goebel 1989, 47 n. 16; Eucken 1996; Gagarin 1997, 8-9.

21. More examples of such borderline cases are analyzed in Degering 1983; Currie 1985, 388-90; Cohn 1989; Walton 1990, 89-95. See also Genette 1993, 79-84.

22. Although the jurors swore an oath to the effect that they would judge “according to their justest judgment, without regard to favor or enmity” (Dem. 57.63, τὸ ψηφιεῖsθαι γνώμῃ τω διaαιοτάτῃ aαὶ οὔτε χάϱιτος ενεa' οὔτ' ἔχθϱας; see Harrison 1998, 48), they could not be held responsible for their decisions.

23. Additional testimonies were themselves carefully drafted and thus apt to distort reality; see Carey 1994, 97-106.

24. De Man 1986, 11: “Literature is fiction not because it somehow refuses to acknowledge ‘reality,’ but because it is not a priori certain that language functions according to principles which are those, or which are like those, of the phenomenal world. It is therefore not a priori certain that literature is a reliable source of information about anything but its own language.”

25. Barthes 1986, 148. Riffaterre (1978) speaks of “referential fallacy.”

26. Brinker 1983. He defines this quality as follows: “Verisimilitude may be described as the factor by which a representation is recognized as a possible model of a given (nonartistic) reality when the representation in question seems to us to be modeled upon this very reality” (261). A similar argument had been made by Chatman (1981, 28-29); see also Jakobson 1987, 23 (in the article “On Realism in Art,” first published in 1921); Orr 1981, 204-5; Barthes 1986, 147. See also the definition of εἰaός in Anaximenes' treatise, quoted above.

27. Kristeva 1969, 150; cf. Culler 1975, 138-40.

28. Indeed, every judicial system depends on the assumption that objective truth exists and can be reached (at least approximately) by human understanding—as has been powerfully argued by Farber and Sherry (1997).

29. Cf. Gondos 1996, 86-87.

30. See further 1 δ 10, “They pretend to prove my guilt by showing the probabilities and assert that I am the real, not the likely murderer of this man,” ἐa δὲ τῶν εἰaότων πϱοsποιούμενοί με ἐλέγχειν, οὐa εἰaότως ἀλλ' ὄντως φονέα μέ φαsι τοῦ ἀνδϱὸς εiναι, referring to 1 γ 8, φάsaων δὲ οὐ τοὺς εἰaότως, ἀλλ' ὄντως ἀποaτείναντας φονέας εέναι. On the use of εἰaός in the Tetralogies see Carawan 1998, 184-92.

31. This passage should not be read as a judgment endorsed by Aristotle, but rather as a description of strategies used by speakers in the courts, possibly derived from older handbooks of rhetoric; cf. Radermacher 1951, 213-14. On the basis of this passage Warnick makes the claim that “in the Greek courts, probabilities were often given more credence than physical evidence and the testimony of witnesses. Physical evidence could be corrupted or prearranged, and witnesses could be bribed, but probabilities were viewed as constant and universally acceptable” (1989, 308). However, this is neglecting the context: Aristotle is describing the tactics of lawyers, not a general attitude toward types of evidence. Cf. also the debate about the role of torture in the Athenian law courts cited below, note 56.

32. This expression was part of the jurors' oath; see above, note 22.

33. For the few cases of “matching pairs” see Todd 1993, 37 with n. 13.

34. See Todd 1990, 172: “an orator will lie his head off at the slightest opportunity.” And Todd 1993, 38: “It is often possible to show … that a speaker is lying, but it is dangerous to infer from this that his opponent is telling the truth; even when you have a case that is both legally strong and likely to win the favour of the jury, it may be worth ‘improving’ it.”

35. Aristophanes' Philocleon is rather cynical about the credibility of speakers in the law courts: see Wasps 560-75.

36. Culler 1975, 139-45 (quotation, 140).

37. For instance, Andocides claims that it is natural to love one's father (1.50, τὸν πατέϱα, ὃν εἰaός ἐsτί sε μάλιsτα φιλεῖν). Isaeus says it is εἰaός that relatives perform religious rituals together (8.15, οiα γὰϱ εἰaὸς παίδων ὄντων ἐξ ἑαυτοῦ θυγατϱός, οὐδεπώποτε θυsίαν ἄνευ ἡμῶν οὐδεμίαν ἐποίηsεν).

38. See further Ant. 1.2, 5.74; Dem. 44.32; Is. 2.18, 4.1, 6.41.

39. Genette [1968] 1969, 76: “Le récit vraisemblable est … un récit dont les actions répondent, comme autant d'applications ou de cas particuliers, à un corps de maximes reçues comme vraies par le public auquel il s'adresse; mais ces maximes, du fait même qu'elles sont admises, restent le plus souvent implicites. Le rapport entre le récit vraisemblable et le système de vraisemblance auquel il s'astreint est donc essentiellement muet.” See also Riffaterre 1990, 8-10.

40. On the orators' attempts to manipulate the jurors by referring to “the laws” see Hillgruber 1988, 104-20.

41. Barthes [1968] 1986, 146: “The pure and simple ‘representation’ of the ‘real,’ the naked relation of ‘what is’ (or has been) thus appears as a resistance to meaning; this resistance confirms the great mythic opposition of the true-to-life (the lifelike) and the intelligible …, as if, by some statutory exclusion, what is alive cannot signify—and vice versa.” See also Striedter 1989, 38.

42. On this passage see Lucas 1968, 229: “What the deception of the audience by παϱαλογιsμός comes down to is that, the more realistic details are invented in an account of an improbable event, the more probable it is made to seem.” See also Else 1957, 625-26; Puelma 1995, 140 with n. 64.

43. For an interesting study cf. Rommel 1995.

44. On ἐνάϱγεια see Zanker 1981.

45. Similar definitions can be found in Herm. Progymn. 10 p. 23.9-11 Rabe; Theon Progymn. 11 p. 119.27-29 Spengel; or Anon. Seguerianus 96; cf. Kroll 1950, 1111.

46. See also ps.-Aristid. Ars 2.99 p. 112 Schmid.

47. Lys. 13.37, οἱ μὲν γάϱ τϱιάaοντα ἐaάθηντο ἐπὶ τῶν βάθϱων, ον νῦν οi πϱυτάνεις aαθέzονται· δύο δὲ τϱάπεzαι ἐν τω πϱόsθεν τῶν τϱιάaοντα ἐaείsθην.

48. On the earrings and their history see Wooten 1988; Borthwick 1990; Bons 1993.

49. On the differences between adversarial and inquisitorial systems of justice see Harrison (1998, 163-64) and Todd (1993, 67-68). Todd rightly describes “Athenian justice as the adversary system taken to its extremes.” For someone brought up in a strongly inquisitorial system (I am German), the Athenian use of emotion is very surprising; but I assume that even Americans (whose system is strongly adversarial and allows similar strategies to some extent) would find most Athenian instances unacceptable.

50. On abuse in Attic oratory see Voegelin 1943; Carey 1989, 162; Harding 1994, 212-18.

51. Diderot developed his thoughts on “small common circumstances” producing verisimilitude in chapter 10 of his De la poésie dramatique ([1758] 1980, 356-61).

52. Cf. the complaint about scarcity of time in, e.g., ps.-Dem. 43.8-9.

53. See Harrison 1998, 163; MacDowell 1978, 249.

54. Good observations on this can be found in Westlake 1958.

55. See also Lys. 9.19, aατὰ τοὺς νόμους aαὶ aατὰ τὸ εἰaός; 13.1, εἰaότως … διaαίως, 25.6. At Ant. Tetr. 1 β 3 the manuscripts are divided between διaαίως (N) and δ' εἰaότως (A).

56. On βάsανος see now Mirhady 1996 and the convincing reply to his hypothesis in Thür 1996.

57. To give but a few examples see Is. 9.21; Dem. 21.28, 21.120, 24.198, 36.25; Andocides 2.19, 5.38, 5.43; Lys. 12.27, 13.18; Ant. 1.11. Cf. above, note. 9.

58. See Goodwin 1890, 79 § 237; Kühner and Gerth 1898-1904, I 233-34.

59. On the question of which political fights this passage alludes to see Albini 1961, 26.

60. See, e.g., Dem. 34.40; Hyperid. 5.26; Lys. 12.27-28; Ant. Tetr. 1 δ 5.

61. Earlier versions of this essay were presented at the University of North Carolina at Chapel Hill and at New York University. Useful comments were made by the audience in both instances, as well as by the anonymous referee for this Journal. Bill Race read an earlier draft and made many helpful suggestions. I am indebted to all these colleagues and friends.

Bibliography

Albini, Umberto, ed. 1961. Andocide: De Reditu. Introduzione e commento. Florence: Le Monnier.

Anastassiou, Jannis. 1981. “Die Wahrscheinlichkeitsargumentation als Beweis-methode bei Gorgias.” Deucalion 36:355-71.

Barthes, Roland. 1968. Writing Degree Zero. Translated by Annette Lavers and Colin Smith. London: Hill & Wang. (First published in French, Paris: Seuil, 1953.)

———. 1986. The Rustle of Language. Translated by Richard Howard. New York: Hill & Wang. (First published in French, Paris: Seuil, 1984.)

Bons, J. A. E. 1993. “Lysias 12, 19: The Earrings Again.” Hermes 121:365-67.

Borthwick, E. K. 1990. “Two Emotional Climaxes in Lysias' Against Eratosthenes.” CW [Classical World] 84:44-46.

Bowie, Ewen L. 1993. “Lies, Fiction and Slander in Early Greek Poetry.” In Lies and Fiction in the Ancient World, edited by Christopher Gill and T. P. Wiseman, 1-37. Exeter: University of Exeter Press.

Brinker, Menachem. 1983. “Verisimilitude, Conventions, and Beliefs.” New Literary History 14:253-67.

Carawan, Edwin. 1993. “The Tetralogies and Athenian Homicide Trials.” AJP [American Journal of Philology] 114:235-70.

———. 1998. Rhetoric and the Law of Draco. Oxford: Oxford University Press.

Carey, Christopher, ed. 1989. Lysias: Selected Speeches. Cambridge: Cambridge University Press.

———. 1994. “‘Artless’ Proofs in Aristotle and the Orators.” BICS [Bulletin of the Institute of Classical Studies] 39:95-106.

Chatman, Seymour. 1981. “How Do We Establish New Codes of Verisimilitude?” In The Sign in Music and Literature, edited by Wendy Steiner, 26-38. Austin: University of Texas Press.

Cohn, Dorrit. 1989. “Fictional versus Historical Lives: Borderlines and Borderline Cases.” Journal of Narrative Technique 19:3-24.

Cole, Thomas. 1991a. The Origins of Rhetoric in Ancient Greece. Baltimore: The Johns Hopkins University Press.

———. 1991b. “Who Was Corax?” ICS 16:65-84.

Culler, Jonathan. 1975. Structuralist Poetics: Structuralism, Linguistics, and the Study of Literature. Ithaca: Cornell University Press.

Currie, Gregory. 1985. “What is Fiction?” Journal of Aesthetics and Art Criticism 43:385-92.

Dalfen, Joachim. 1985. Review of Synodinou 1981. Gnomon 57:1-4.

Degering, Klaus. 1983. “‘Fiction as Fact’: Eine Analyse des Pulitzer Fraud (1980-1981).” Poetica 15:151-78.

De Man, Paul. 1986. The Resistance to Theory. Theory and History of Literature, 33. Minneapolis: University of Minnesota Press.

Denniston, J. D. 1954. The Greek Particles. 2d ed. Oxford: Oxford University Press.

Detienne, Marcel. 1996. The Masters of Truth in Ancient Greece. Translated by Janet Lloyd. Cambridge: MIT Press. (First published in French, Paris: Maspero, 1967.)

Diderot, Denis. [1758] 1980. De la poésie dramatique. In Oeuvres complètes. Vol. X, Le drame bourgeois. Edited by Jacques Chouillet and Anne-Marie Chouillet. Paris: Hermann.

Due, Bodil. 1980. Antiphon: A Study in Argumentation. Opuscula Graecolatina, 17. Copenhagen: Museum Tusculanum.

Eden, Kathy. 1989. Poetic and Legal Fiction in the Aristotelian Tradition. Princeton: Princeton University Press.

Else, Gerald F. 1957. Aristotle's Poetics: The Argument. Cambridge: Harvard University Press.

Erbse, Hartmut. 1977. “Antiphons Rede (Or. 5) über die Ermordung des Herodes.” RhM 120:209-27.

Eucken, Christoph. 1996. “Das Tötungsgesetz des Antiphon und der Sinn seiner Tetralogien.” MH 53:73-82.

Fairchild, William D. 1979. “The Argument from Probability in Lysias.” Classical Bulletin 55:49-54.

Farber, Daniel A., and Suzanna Sherry. 1997. Beyond All Reason: The Radical Assault on Truth in American Law. New York: Oxford University Press.

Finkelberg, Margalit. 1998. The Birth of Literary Fiction in Ancient Greece. Oxford: Clarendon Press.

Gagarin, Michael. 1990. “The Nature of Proofs in Antiphon.” CJ 85:22-32.

———. 1994. “Probability and Persuasion: Plato and Early Greek Rhetoric.” In Persuasion: Greek Rhetoric in Action, edited by lan Worthington, 46-68. London: Routledge.

———, ed. 1997. Antiphon: The Speeches. Cambridge: Cambridge University Press.

Genette, Gérard. 1969. Figures II. Paris: Seuil.

———. 1993. Fiction and Diction. Translated by Catherine Porter. Ithaca: Cornell University Press. (First published in French, Paris: Seuil, 1991.)

Goebel, George Harry. 1983. Early Greek Rhetorical Theory and Practice: Proof and Arrangement in the Speeches of Antiphon and Euripides. Diss. University of Wisconsin, Madison.

———. 1989. “Probability in the Earliest Rhetorical Theory.” Mnemosyne 42: 41-53.

Gondos, Elizabeth Ann. 1996. Auf dem Weg zur rhetorischen Theorie: Rhetorische Reflexionen im ausgehenden fünften Jahrhundert v. Chr. Tübingen: Niemeyer.

Goodwin, William Watson. 1890. Syntax of the Moods and Tenses of the Greek Verb. 2d ed. Boston: Ginn.

Görgemanns, Herwig. 1976. “Rhetorik und Poetik im homerischen Hermeshymnos.” In Studien zum antiken Epos, edited by Herwig Görgemanns and Ernst A. Schmidt, 113-28. Beiträge zur Klassischen Philologie, 72. Meisenheim am Glan: Hain.

Grimaldi, William M. A. 1996. “How Do We Get from Corax-Tisias to Plato-Aristotle in Greek Rhetorical Theory?” In Theory, Text, Context: Issues in Greek Rhetoric and Oratory, edited by Christopher Lyle Johnstone, 19-43. New York: State University of New York Press.

Halliwell, Stephen. 1986. Aristotle's Poetics. London: Duckworth.

Harding, Phillip. 1994. “Comedy and Rhetoric.” In Persuasion: Greek Rhetoric in Action, edited by Ian Worthington, 196-221. London: Routledge.

Harrison, A. R. W. 1998. The Law of Athens. 2d ed. Vol. II, Procedure. London: Duckworth.

Heitsch, Ernst. 1984. Antiphon aus Rhamnus. AAWM 1984.3. Mainz: Steiner.

Hillgruber, Michael, ed. 1988. Die zehnte Rede des Lysias. Einleitung, Text und Kommentar mit einem Anhang über die Gesetzesinterpretation bei den attischen Rednern. Untersuchungen zur Antiken Literatur und Geschichte, 29. Berlin: de Gruyter.

Ingarden, Roman. 1973. The Literary Work of Art: An Investigation on the Borderlines of Ontology, Logic, and Theory of Literature. Translated by George G. Grabowitz. Evanston: Northwestern University Press. (First published in German, Tübingen: Niemeyer, 3d ed., 1965.)

Jakobson, Roman. 1987. Language in Literature. Cambridge: Belknap Press.

Jarratt, Susan C. 1991. Rereading the Sophists: Classical Rhetoric Refigured. Carbondale: Southern Illinois University Press.

Johnstone, Christopher Lyle. 1996. “Introduction: The Origins of the Rhetorical in Archaic Greece.” In Theory, Text, Context: Issues in Greek Rhetoric and Oratory, edited by Christopher Lyle Johnstone, 1-18. New York: State University of New York Press.

Kennedy, George Alexander. 1963. The Art of Persuasion in Greece. Princeton: Princeton University Press.

———. 1974. “The Sophists as Declaimers.” In Approaches to the Second Sophistic: Papers Presented at the 105th Annual Meeting of the American Philological Association, edited by Glen W. Bowersock, 17-22. University Park, Pa.: American Philological Association.

———. 1994. A New History of Classical Rhetoric. Princeton: Princeton University Press.

Kristeva, Julia. 1969. Σημειωτιaὴ: Recherches pour une sémanalyse. Paris: Seuil.

Kroll, Wilhelm. 1950. “Rhetorik.” RE Suppl. 7, 1039-1138.

Kühner, Raphael, and Bernhard Gerth. 1898-1904. Ausführliche Grammatik der griechischen Sprache: Satzlehre. 3d ed. 2 vols. Hannover: Hahn. (Reprinted 1992.)

Lamarque, Peter, and Stein Haugom Olsen. 1994. Truth, Fiction, and Literature: A Philosophical Perspective. Oxford: Oxford University Press.

Littérature et réalité = Collection Points 142. 1982. Paris: Seuil.

López Eire, Antonio. 1994. “Literatura y vida publica: Orígenes de la oratoria.” In Los Orígenes de la oratoria y la historiografía en la grecia clásica, edited by Antonio López Eire and Carlos Schrader. Monografías de Filología Griega, 5. Zaragoza: Departamento de Ciencias de la Antigüedad, Universidad de Zaragoza.

Lucas, D. W., ed. 1968. Aristotle: Poetics. With an introduction, commentary, and appendices. Oxford: Oxford University Press.

MacDowell, Douglas M. 1978. The Law in Classical Athens. Ithaca: Cornell University Press.

Manuwald, Bernd. 1995. “Zur rechtlichen Problematik von Antiphon, Or. 5.” RhM 138:41-59.

Mirhady, David C. 1996. “Torture and Rhetoric.” JHS 116:119-31.

Orr, Leonard. 1981. “Vraisemblance and Alienation Techniques: The Basis for Reflexivity in Fiction.” Journal of Narrative Technique 11:199-215.

O'Sullivan, Neill. 1995. “Aristotle on Dramatic Probability.” CJ 91:47-63.

Pratt, Louise H. 1993. Lying and Poetry from Homer to Pindar: Falsehood and Deception in Archaic Greek Poetics. Ann Arbor: University of Michigan Press.

Puelma, Mario. 1995. “Der Dichter und die Wahrheit in der griechischen Poetik von Homer bis Aristoteles.” In his Labor et Lima: Kleine Schriften und Nachträge, 111-51. Basel: Schwabe. (Reprinted with additions from MH 46 [1989] 65-100.)

Radermacher, Ludwig, ed. 1951. Artium Scriptores (Reste der voraristotelischen Rhetorik). SAWW 227.3. Vienna: Rohrer.

Riffaterre, Michael. 1978. “The Referential Fallacy.” Columbia Review 57, no. 2: 21-35.

———. 1990. Fictional Truth. Baltimore: The Johns Hopkins University Press.

Rommel, Thomas. 1995. “Aspects of Verisimilitude: Temporal and Topographical References in Robinson Crusoe.” Literary and Linguistic Computing 10:279-85.

Rösler, Wolfgang. 1980. “Die Entdeckung der Fiktionalität in der Antike.” Poetica 12:283-319.

Russell, Donald Andrew. 1983. Greek Declamation. Cambridge: Cambridge University Press.

Schiappa, Edward. 1996. “Toward a Predisciplinary Analysis of Gorgias' Helen.” In Theory, Text, Context: Issues in Greek Rhetoric and Oratory, edited by Christopher Lyle Johnstone, 65-86. New York: State University of New York Press.

Schindel, Ulrich. 1979. Der Mordfall Herodes: Zur 5. Rede Antiphons. NAWG 1979.9. Göttingen: Vandenhoeck & Ruprecht.

Schmitz, Thomas. 1997. Bildung und Macht: Zur sozialen und politischen Funktion der zweiten Sophistik in der griechischen Welt der Kaiserzeit. Zetemata 97. Munich: Beck.

Sealey, R. 1984. “The Tetralogies Ascribed to Antiphon.” TAPA 114:71-85.

Searle, John R. 1979. Expression and Meaning: Studies in the Theory of Speech Acts. Cambridge: Cambridge University Press.

Solmsen, Friedrich. 1931. Antiphonstudien: Untersuchungen zur Entstehung der attischen Gerichtsrede. Neue Philologische Untersuchungen, 8. Berlin: Weidmann.

Striedter, Jurij. 1989. Literary Structure, Evolution, and Value: Russian Formalism and Czech Structuralism Reconsidered. Harvard Studies in Comparative Literature, 38. Cambridge: Harvard University Press.

Swain, Simon. 1996. Hellenism and Empire: Language, Classicism, and Power in the Greek World A.D. 50-250. Oxford: Oxford University Press.

Synodinou, Katerina. 1981. '′Εοιaα-εἰaὸς aαὶ sυγγενιaὰ ἀπὸ τὸν ‘′Ομηϱο ὡς τὸν 'Αϱιsτοφάνη. Σημαsιολιγιaὴ μελέτη. Ioannina: Panepistemio Ioanninon.

Szepes, Erika. 1980. “Humour of the Homeric Hermes Hymn.” Homonoia 2: 5-56.

Thür, Gerhard. 1996. Reply to Mirhady 1996. JHS 116:132-34.

Todd, Stephen C. 1990. “The Use and Abuse of the Attic Orators.” G&R 37: 159-78.

———. 1993. The Shape of Athenian Law. Oxford: Oxford University Press.

Todorov, Tzvetan. 1968. “Introduction.” Communications 11:1-4.

Voegelin, W. 1943. Die Diabole bei Lysias. Basel: Schwabe.

Walton, Kendall L. 1990. Mimesis as Make-Believe: On the Foundations of the Representational Arts. Cambridge: Harvard University Press.

Warnick, Barbara. 1989. “Judgment, Probability, and Aristotle's Rhetoric.” Quarterly Journal of Speech 75:299-311.

Warning, Rainer. 1969. “Fiktion und Wirklichkeit in Sternes Tristram Shandy und Diderots Jacques le fataliste.” In Nachahmung und Illusion: Kolloquium Giessen Juni 1963. Vorlagen und Verhandlungen, edited by Hans Robert Jauss. 2d ed. Poetik und Hermeneutik, 1. Munich: Fink.

Watzlawick, Paul. 1976. How Real Is Real? Confusion, Disinformation, Communication. New York: Random House.

Westlake, H. D. 1958. ‘“Ως εἰaός in Thucydides.” Hermes 86:447-52.

Wooten, Cecil W. 1988. “The Earrings of Polemarchus' Wife (Lysias 12.19).” CW 82:29-31.

Zanker, Graham. 1981. “Enargeia in the Ancient Criticism of Poetry.” RhM 124: 297-311.