Perelman's Interpretation of Reverse Probability Arguments

as a Dialectical Mise en Abyme

Manfred Kraus

Philosophy & Rhetoric, Volume 43, Number 4, 2010, pp. 362-382 (Article)

Published by Penn State University Press

DOI: https://doi.org/10.1353/par.2010.0005

For additional information about this article

https://muse.jhu.edu/article/403026

Access provided by University of Toledo (29 Jun 2018 23:29 GMT)

 Perelman’s Interpretation of Reverse
Probability Arguments as a Dialectical
Mise en Abyme

Manfred Kraus

introduction
Imagine the following situation: an act of violent assault has been committed.
And there are only two possible suspects, of which one is a small and weak
man and the other a big and strong man. The weak man will plead that he
is not strong enough and therefore not likely to have committed the crime,
which seems reasonable straight away. But there will also be a loophole
for the strong man, as Aristotle tells us, who reports exactly that story in
book 2, chapter 24, of his Rhetoric. And he also has a name to assign to the
inventor of that kind of argument (1402a17–20):

The Art of Corax is composed of this topic. For if a man is not

likely to be guilty of what he is accused of, for instance if, being
weak, he is accused of assault and battery, his defence will be that
the crime is not probable [eikós]; but if he is likely to be guilty, for
instance, if he is strong, it may be argued again that the crime is
not probable [eikós], for the very reason that it was bound to appear
probable [eikós]. (Aristotle 1926, 335, translation modified).

Philosophy and Rhetoric, Vol. 43, No. 4, 2010

Copyright © 2010 The Pennsylvania State University, University Park, PA

PR 43.4_04.indd 362 10/28/10 10:16:16 AM

 perelman’s interpretation of reverse probability arguments

Plato refers to the same example, though with slight differences, in the
Phaedrus (273b–c), where he ascribes it to Tisias. In fact, the invention of
such arguments by probability (eikós) has always been associated with the
legendary founders of rhetoric, the Sicilians Corax and Tisias (see Hinks
1940, 63–66; Kuebler 1944, 15; Kennedy 1963, 26–51; Goebel 1989, 41–42;
Gagarin 2002, 29), and their unscrupulous exploitation with sophists such
as Protagoras and Gorgias (Aristotle, Rhetoric 1402a25; Plato, Phaedrus
267a). For our present purpose, we may confine our attention to Aristotle’s
version, since only this version contains the additional twist of the strong
man’s argument, and we may be confident—with Michael Gagarin—that
“Aristotle’s version is closer to Tisias’ original version than Plato’s” (Gagarin
2007, 33; see also 1994, 51). It matters little whether we ascribe it to Corax
or Tisias. Tisias is regarded as Corax’s student and may have recorded his
master’s teachings. But it has also been suggested that they may have been
one and the same person, Corax (“Crow”) being a nickname for Tisias
(Cole 1991).
The argument of the strong man, which wittily turns the tables on the
simple argument from probability as used by the weak man, has been aptly
dubbed the “reverse eikós-argument” or “reverse-probability argument” by
Michael Gagarin (1990, 30; 1994, 51; 1997, 14; 2002, 112–14; 2007, 25). Yet, in
no more than two pages in §96 of their New Rhetoric—as only very few
scholars have ever noticed—Chaïm Perelman and Lucie Olbrechts-Tyteca
offer an intriguing expanded account of this “reverse probability argument”
(1969, 458–59), which is the subject of the following analysis.
I first give a brief account of the background and operating mode
of ancient probability arguments and of the particular variant of reverse
probability that the strong man’s argument represents and then analyze
Perelman and Olbrechts-Tyteca’s formulation, which I try to interpret as
an instance of a dialectical application of the technique of mise en abyme
and as a quasi-logical counterpart to logical paradoxes such as the Cretan
liar and Russell’s paradox.

probability and reverse probability

Simple Probability Arguments
In the fifth and fourth centuries BCE in particular, arguments from
probability played a considerable role in Greek oratorical practice. Not
only sophists such as Gorgias (Kuebler 1944, 26–36; Anastassiou 1981;

363

PR 43.4_04.indd 363 10/28/10 10:16:16 AM

 manfred kraus

Spatharas 2001, 394–98) but also orators such as Antiphon, Lysias, Isocrates,
and others made ample use of them (Kuebler 1944, 36–61; Synodinou 1981,
118–28; Gagarin 1989, 47–56; 1990, 29–31; 2002, 112–18, 153–54). In view of
their importance in early rhetoric and oratory, the concept of eikós has
been subjected to closer scrutiny in a significant number of recent studies
(Kuebler 1944; Turrini 1977, 1979; Synodinou 1981; Goebel 1989; Gagarin
1994; 1997, 13–15; 2002, 112–18; Poulakos 1995, 179–81; Jacob 1996; Carawan
1998, 184–92; Schmitz 2000; Kraus 2006, 2007; Hoffman 2008).
The common translation of eikós as “probability” should not, how-
ever, make us think of anything like the modern notion of statistical or
frequency-based probability, which only emerged in the seventeenth century
in conjunction with inductive logic (see Hacking 1975; Hoffman 2008, 4–6).
Not even Aristotle’s standard definition of eikós as that which holds “for the
most part” (hôs epì to polú) with respect to “things that may as well be other
than they are” (Rhetoric 1357a34–b1; 1402b14–16; Prior Analytics 70a3–6)
refers to probability in a “statistical” or frequency-based sense, but it has
been shown to refer to what will happen naturally unless impeded by some
internal or external interference (see Winter 1997).
Furthermore, a lot would seem to depend on the linguistic interpreta-
tion of the Greek term eikós itself. Recent studies elicit two basic mean-
ings: a “comparative” one (Synodinou 1981, e.g., 11, 34; Turrini 1977, 544–50),
according to which the Greek verb éoika, of which eikós is the participle,
would mean “to be similar,” “to resemble,” “to look like,” and a “normative”
one (Synodinou 1981, 2–27; Turrini 1977, 550–57) that conveys the sense of
“to be appropriate, fitting, suitable, proper.” But which of these meanings is
the primary one? In a couple of recent papers (2006, 2007), I have tried to
make a case in favor of the “appropriateness” meaning, against strong resis-
tance from many authors from Plato till David Hoffman, who all advocate
the “likeness” meaning (2008; see also Frisk 1960–70, 1:530; Turrini 1977,
545, 557; and—with false etymological derivation—Dzialo 1998, 225–26). An
ultimate decision may prove unnecessary after all, as both meanings can
be further reduced to a common core such as “closeness,” or “matching-
ness.” Indo-European linguists, however, insist that the meaning of the root
*ueik- is opaque and cannot be unraveled (Frisk 1960–70, 1: 530; Chantraine
1968–80, 2:355; Pokorny 1989, 1129); unfortunately, any link to the root in
the Greek word oikos, “house,” also present in oikeios, “of the same house
or family, kin, related” but also “fitting, suitable,” appears to be precluded
by the velar nature of the guttural plosive in eikós as opposed to its palatal

364

PR 43.4_04.indd 364 10/28/10 10:16:16 AM

 perelman’s interpretation of reverse probability arguments

character in oikos (Pokorny 1989, 1129, 1131; for other possible roots of *ueik-,
see Pokorny 1989, 1128–31).
Of even greater importance is the fact that the effectiveness of
probability arguments in a rhetorical context is strongly audience-
dependent. In these arguments, audience adherence is secured by the
coherence of the speaker’s arguments with what an audience can reasonably
be expected to assume as natural, with their own everyday experience,
their own preconceived opinions about nature and human life, their own
emotional predispositions and behavioral habits (Kraus 2006, 141). Such
an audience-dependent view is perfectly reflected in the fourth-century
BCE Rhetoric to Alexander, in which probability is defined as “a statement
supported by examples present in the minds of the audience” (1428a25–26;
see Kraus 2006, 145–46). Similar views have recently also been acknowledged
by contemporary argumentation scholars. As Douglas Walton puts it,
“plausible or so-called eikotic arguments are based on a person’s subjective
understanding of how something can normally be expected to go in a
familiar situation” (2001, 104). They are based on something we would
nowadays call “empathy,” on stereotypical action frames, for which Roger C.
Schank and Robert P. Abelson have introduced the term “scripts” (1977; see
Walton 2001, 108–10), on what Christopher Tindale describes as a sharing
of “cognitive environments” between speaker and audience that ultimately
leads to the “acceptability” of arguments (1999, 101–15), on what Michael
Billig formulates as the principle of common sense based on “communal
links, foremost among which are shared values or beliefs” (1996, 226), or on
what Walton himself calls “common knowledge” (2001, 108–9). Hoffman,
in the same sense, prefers to speak of “social expectations” or “profiles”
(2008, 15–22). Probability arguments thus clearly draw on common ground
shared by speaker and audience (Kraus 2007). If this is correct, then what is
reflected in rhetorical eikós arguments is plausibility rather than probability
(Rescher 1976; Vega Renon 1998; Walton 2000, 729; 2001, 98; see also Walton
1992, 1999). In practice, in accordance with Aristotle’s remarks in Rhetoric
1:10–12, the eikós arguments found in ancient Greek judicial speeches turn
out to basically focus on three main issues: a) motive, b) means or occasion,
and c) the prospect of concealment (Kraus 2006, 139–40), all of which
draw on what the audience in question would accept as a reasonable and
hence plausible account. If, however, such an argument were to be made
plausible to any audience instead of some particular audience, something
like Perelman’s concept of a universal audience would indeed be called for
(see Crosswhite 1989).

365

PR 43.4_04.indd 365 10/28/10 10:16:16 AM

 manfred kraus

Reverse Probability
Probability arguments clearly were a central feature in practical oratory in
the fifth and fourth centuries BCE. But how about reverse probability, that
is, the strong man’s argument? The persuasive force of such a reverse eikós
argument is evidently much weaker than that of a straightforward proba-
bility argument. Consequently, although Walton—as well as Perelman and
Olbrechts-Tyteca (1969, 459), and to a certain extent also Hoffman (2003,
506)—assumes that this was a “kind of argumentation typically used in
legal reasoning” (2000, 730), in reality the reverse eikós appears not to have
been widely used in practical oratory. As Gagarin and others have pointed
out, apart from the Corax story there is only one further attested example,
in Antiphon’s First Tetralogy (2.3, 2.6), a piece that is, however, a model exer-
cise in eikós argumentation. Yet “we never find it in an actual court speech”
(Gagarin 2001, 284; see also 1997, 14; 2007, 33). Hence—despite Hoffman’s
defense of its limited “psychological” validity in terms of the reflectivity and
self-referentiality of human thought (2003, 504–6)—the reverse eikós may
after all have been rather a highlight in demonstration and instruction or
an example of witty sophistic “antilogies” (Gagarin 2001, 282–85) or “intel-
lectually interesting arguments” (Gagarin 2007, 33) rather than a technique
of oratorical argumentation that was used.
In Antiphon’s First Tetralogy, too, a man accused of murder defends
himself by arguing that “it was still more natural for me to foresee before
committing the crime that suspicion would settle upon me as it has done”
(2.2.3). Gagarin notes that in this case “the reverse eikos-argument . . . would
ultimately lead by a reductio ad absurdum to the conclusion that the least
likely suspect was most likely to be the murderer (because he would never
be suspected), even though he would have no motive at all” (1994, 138).
At first sight, this does indeed appear paradoxical. But the motif is per-
fectly well known from detective stories, in books as well as on TV. In such
murder mysteries the most likely suspect is virtually never guilty, and the
villain instead will be always the innocent-looking guy (or lady). In fact, the
reverse eikós is a highly common device in fiction, and nobody would ever
take offense at it there. Aristotle, too, was apparently aware of this fact. In
the Rhetoric (1402a7–14) he says:

In Rhetoric, an apparent enthymeme may arise from that which

is not absolutely [haplôs] probable but only in particular cases [ti].
But this is not to be understood absolutely [katholou], as Agathon

366

PR 43.4_04.indd 366 10/28/10 10:16:16 AM

 perelman’s interpretation of reverse probability arguments

says: “One might perhaps say that this very thing is probable, that
many things happen to men that are not probable;” for that which
is contrary to probability nevertheless does happen, so that that
which is contrary to probability is probable. If this is so, that which
is improbable will be probable. (Aristotle 1926, 335)

This is then illustrated by the story of the weak and strong man argument.
Thus for Aristotle for an argument to hold “for the most part” requires
some absolute preponderance of one side over the other, and for this reason
in his view the reverse eikós argument is not acceptable but a clear case of
“making the weaker argument the stronger” (see Schiappa 2003, 103–16), as
he explicitly states only a few lines later (1402a22–24). It should be noted,
however, that for illustration of the opposite position, Aristotle quotes a
tragedian. He thus appears to be willing to grant greater license in the
use of eikós to poets, who write fictional narratives, than to orators, who
deal with real life, as has been aptly pointed out by Jane Sutton (1991). In
fact, Aristotle alludes twice to the Agathon quote in the Poetics (1456a23–25,
1461b15) in his explanation of the role of the unexpected and the inconsis-
tent in fiction.
How then does a reverse eikós argument work? In the Corax example,
the weak man’s argument draws on the lack of means (cf. Aristotle, Rhetoric
1372a22). The strong man’s reverse argument, on the other hand, clearly
draws on the lack of prospect of getting away unsuspected, and it does so
by anticipating audience expectation, and then frustrating that expectation
with a countermove. This is indeed the typical move a reverse probability
argument makes: anticipation and frustration of the audience’s expectations,
or, as Hoffman would put it, contradiction of a “stereotype” or “prejudice”
(2003, 506).

multiple reverse probability: the corax

Perelman and Olbrechts-Tyteca’s Interpretation of the Corax
What works once may also work twice, and so another countermove may
be possible. The opposite party might say: “Well, you might have foreseen
that you would not be suspected owing to the fact that you would certainly
have foreseen you would be suspected; so you may very well have com-
mitted the crime.” To this, however, the defendant might in turn retort:
“But I would undoubtedly have foreseen that I would after all be suspected

367

PR 43.4_04.indd 367 10/28/10 10:16:16 AM

 manfred kraus

owing to the fact that I would certainly have foreseen that I would not
be suspected owing to the fact that I would certainly have foreseen that
I would be suspected; therefore I would never have committed the crime,”
and so on ad infinitum (see Hoffman 2003, 505). Arguments would start
dialectically swinging back and forth endlessly and spinning around like a
spinning wheel, and we would end up in what Plato would call an íliggos, a
vertigo or paralyzing dizziness (see Protagoras 339e2; Gorgias 486b1, 527a2;
Republic 407c2; Laws 892e7; Seventh Letter 325e3).
This is exactly the description given of this type of argument by Perel-
man and Olbrechts-Tyteca in §96 of the New Rhetoric. They treat it in their
chapter on arguments from dissociation of ideas (pt. 3, chap. 4, §§89–96),
which both authors thought was one of the New Rhetoric’s most original
contributions to the theory of argumentation (see, e.g., Olbrechts-Tyteca
1963; Perelman 1982, 126–37), and they analyze it in terms of their broader
concept of the dissociative “philosophical pair” of “appearance” and “reality”
(1969, 415–16), in which a devalued term 1 (“appearance”) is both dissociated
from and related to a valued term 2 (“reality”) so as to avoid or terminate
incompatibilities that may appear between different aspects of term 1.
In §96, under the heading of “Rhetoric as a Process” (1969, 450), they
maintain that this pair in its particular manifestation as “device” and “reality”
is basic to rhetorical argumentation (the corresponding terms in the original
French are “procédé” and “réalité” [1958, 607]; the French term “procédé”—
first mentioned as a pejorative term 1 in §92 [1958, 577]—is rendered vari-
ably as “process” or “device” in the English version). In their interpretation,
rhetorical means of persuasion are purposefully employed “devices.” Yet
this is a disqualifying term (designating term 1 of a philosophical pair),
and synonymous with fallacious appearance (1969, 450). “When treated as
oratorical or rhetorical devices, the means of persuasion are pronounced to
be artificial, formal, and verbal—terms 1 characteristic of the pairs artificial/
natural, form/substance, verbal/real” (1969, 451). Hence these devices will
only be efficient as long as they are not exposed as such, but accepted by the
audience as terms 2, that is, as natural, substantial, and real.
This evokes the similar pair of dissimulation and sincerity. In the inter-
est of efficiency, the artificiality of rhetorical devices must be concealed by
the appearance of sincerity. Since “everything that furnishes an argument
against the thesis being defended by the speaker, including objections to
his own hypotheses, becomes an indication of sincerity and straightfor-
wardness and increases the hearers’ confidence” (1969, 457), the multiple
reverse probability argument, by explicitly building on a counterargument,

368

PR 43.4_04.indd 368 10/28/10 10:16:16 AM

 perelman’s interpretation of reverse probability arguments

becomes one convenient means of creating this appearance. Obviously, the

technique consists in turning the counterargument into a term 1 (“appar-
ent probability”) and relating it to a new term 2 (“real probability”). This
creates a potentially infinite series of pendulum swings. “In the absence of
a criterion [for sincerity] that is beyond argument the dissociation device/
reality can operate indefinitely and contradictorily. The use of this opposi-
tion seems to be a feature of that extremely ancient technē attributed to
Corax” (1969, 458). Perelman and Olbrechts-Tyteca even have an apt name
for this pattern of argument: they dub it the “corax” (1969, 459).
Perelman and Olbrechts-Tyteca’s overall assessment that “the corax is
simply an application of the dissociation device/reality to the field of con-
jecture” (1969, 459) is consistent with the fact that the primary scope of
probability arguments is indeed what in rhetoric is called the conjectural
status, that is, the issue of whether or not a certain act was in fact commit-
ted by a certain person. The argument is also clearly dialectical, because it
consists in a repeated “batting back and forth” (1969, 459) and “incites one to
perform an act precisely because it is improbable and, for converse reasons,
diminishes the chances that probable acts will be committed” (1969, 459).
But, most importantly of all, Perelman and Olbrechts-Tyteca underscore
its rhetorical character: “The corax is a typically rhetorical device [“procédé”
(1958, 609)] in argument because it is based on the possibility of a wide vari-
ety of interpretations. It is characteristic of nonformal discourse and can be
imagined only in an ambiguous situation” (1969, 459).
Before they turn to quoting all the classical passages that are typically
quoted in accounts of eikós, including Aristotle, Plato, and Antiphon,
Perelman and Olbrechts-Tyteca invoke a modern instance in the form of
letters to the editor published in the New York Herald Tribune between
24 April and 4 May 1948. One of the readers who commented on an earlier
letter, which exhibited a profascist bent and insulted the United States, saw
in this letter a subtle form of communist propaganda (the profascist stance
being used as a dissimulative “device”), whereupon other readers in turn
wondered if it was not perhaps written by a fascist who would have antici-
pated that the letter would be taken for communist propaganda, and so
forth. “This game of interpretation, alternately attributing pro-communist
and pro-fascist views to the writer of the letter could go on indefinitely”
(1969, 458).
Although Hoffman (2003, 505) acknowledges this potential infinity of
dialectical pendulum swings of reciprocal arguments, he does not mention
Perelman’s earlier analysis and does not seem to be aware of it. Bernard Jacob,

369

PR 43.4_04.indd 369 10/28/10 10:16:16 AM

 manfred kraus

on the other hand, who appears to be the only one who does refer to The
New Rhetoric in this context, merely thinks that “Perelman and Olbrechts-
Tyteca see the comic possibilities of this way of proceeding” and asserts that
“the speaker who identifies someone as using the corax is raising questions
about the user’s sincerity because the user of the corax has essentially begun
an infinite, but idle batting back and forth of alternative readings” (1996,
241). “Thus Aristotle’s version is comic, as Perelman and Olbrechts-Tyteca
suggest” (1996, 241). I don’t think this is what they suggest. On the contrary,
they think it is in principle a serious and common argument, “often used
in legal proceedings” (Perelman and Olbrechts-Tyteca 1969, 459). They do
indeed speak about the indefinite “batting back and forth in the corax” (1969,
459), but in their eyes this is anything but “idle,” which is Jacob’s addition.
Moreover, their primary concern is not with the refutability of the corax
but with its utility in creating the appearance of sincerity. What they do say,
however, is that at a certain point in this infinite loop the argument “will
lose almost all its persuasive force and become merely comical, because it
implies an excessive capacity for foresight” (1969, 459). They even pinpoint
the reason why it must stop at a certain point, namely the incapacity of
human speakers and audiences for infinite foresight.

Comic Effects
I’ve already mentioned that reverse probability is less unusual in fiction. One
classic instance of an argument of the corax type can be found in the 1973
fantasy novel and corresponding 1987 movie The Princess Bride by William
Goldman. In a famous scene from this novel that takes place on the Cliffs
of Insanity, the Man in Black (the disguised hero) challenges his antagonist,
the Sicilian criminal genius Vizzini, to a “battle of wits” (Goldman 2007,
175–79; 1987, 43–48). The Man in Black dares Vizzini to choose from two
goblets filled with wine, one of which he alleges is poisoned. Vizzini, proud
of his sharp intellect (“you are no match for my brains,” 2007, 174; 1987, 43),
enthusiastically accepts. He announces that he will not guess but think,
ponder and deduce (2007, 176).
His reasoning starts out as follows: “Now a great fool . . . would place
the poison in his own goblet, because he would know that only another
great fool would reach first for what he was given. I am clearly not a great
fool, so I will clearly not reach for your wine. . . . [Y]ou knew I was not a
great fool, so you would know that I would never fall for such a trick. You
would count on it. So I will clearly not reach for mine either . . . We have

370

PR 43.4_04.indd 370 10/28/10 10:16:16 AM

 perelman’s interpretation of reverse probability arguments

now decided the poisoned cup is most likely in front of you” (2007, 177; see
also the slight modifications in 1987, 45). The line of reasoning breaks off at
this point but starts afresh in a similar way from a different assumption,
that the Man in Black must be a criminal because the poison is made from
a powder that he could only have gotten in Australia (2007, 177; cf. 1987, 45):
“Australia, as everyone knows, is peopled with criminals and criminals are
used to having people not trust them, as I don’t trust you, which means I can
clearly not choose the wine in front of you. . . . But, again, you must have
suspected I knew the origins of iocane, so you would have known I knew
about the criminals and criminal behavior, and therefore I can clearly not
choose the wine in front of me.” He adopts this line of reasoning yet a
third time from a still different perspective, the Man in Black’s exceptional
bodily strength and intellectual power (2007, 178; 1987, 46). This form of
argument clearly recalls the kind Perelman and Olbrechts-Tyteca describe
(particularly by virtue of the progressive anticipation of the opposite party’s
expectations). But the fact that the turning of tables is each time quickly
broken off is telling.
The passage is full of (intentional or unintentional?) allusions to the
ancient background of the corax. Remember that Vizzini is a Sicilian (“Never
go in against a Sicilian when death is on the line” [2007, 179; 1987, 48]). In
the movie version he evokes Plato, Aristotle, and Socrates (1987, 43). And
the Man in Black’s tongue-in-cheek reaction after the first two runs of the
argument (“Truly you have a dizzying intellect” [2007, 178; 1987, 45]) adverts
to the vertiginous effect of such reasoning. Yet the ironic and even comic
undertone of the narrative is unmistakable, particularly so since Vizzini’s
reasoning proves futile after all: both goblets are poisoned, and he dies.
As in Vizzini’s case, such a dialectical argument can be and is often
conducted in a soliloquious and self-reflective manner (see also Hoffman
2003, 505). The pattern will then be something like: “I should like to act in
a certain way, but I can’t, since I know that everybody will expect me to
act in exactly this way; yet, on second thought, they will know that I know
that they will expect this; hence they will expect me not to act in this way;
yet I know that they know that I know, etc.; and they again will know that
I know that they know that I know, etc.” This produces just the kind of
neurotic argument that would be typical of a Woody Allen character. And
it would ultimately lead to perpetual indecision by infinite pondering, as in
the case of Buridan’s ass.
One cannot help perceiving the comic effect of such reasoning when it
reaches epic proportions, and Perelman and Olbrechts-Tyteca, given their

371

PR 43.4_04.indd 371 10/28/10 10:16:16 AM

 manfred kraus

awareness of the role of the ridiculous and of irony in rebutting arguments

(1969, 205–10), were not exceptions. Yet they also emphasize the courage
required to defy such ridicule and to actively attack accepted rules, using
the technique of dissociation (1969, 209–10). But, in fact, the entire pat-
tern itself turns out to be a standard pattern of a particular type of Jewish
joke (William Goldman, it is worth noting, comes from a Jewish family)
that was already described by Sigmund Freud. As a psychoanalyst, Freud
was greatly interested in the psychological background of jokes. In 1905, he
published a volume titled Jokes and Their Relation to the Unconscious. One
of the jokes, which Freud describes as of the “rarest” category and rates as
particularly “excellent,” runs as follows: “Two Jews met in a railway car-
riage at a station in Galicia. ‘Where are you going?’ asked one. ‘To Cracow,’
was the answer. ‘What a liar you are!’ broke out the other. ‘If you say you’re
going to Cracow, you want me to believe you’re going to Lemberg. But I
know that in fact you’re going to Cracow. So why are you lying to me?’”
(Freud 1960, 115). The familiar pattern is instantly recognizable. It is a clear
case of a corax, since it is based on anticipation and deliberate frustration
of an interlocutor’s expectations, and the tables could also easily be turned
time and again to trigger the spinning wheel. Yet the example is instructive,
since it highlights the problem of truth and lie involved in such arguments,
which is why Freud chooses to call this category “‘sceptical’ jokes” (1960,
115), and that in turn clearly touches on Perelman and Olbrechts-Tyteca’s
point about sincerity in rhetorical arguments. When Freud asks, “Does not
genuine truth consist in taking the hearer into account and giving him a
faithful picture of our own knowledge?” (1960, 115), he almost anticipates
a Perelmanian dissociative pair (“genuine” vs. “apparent” truth) and the
essential appeal to the audience.

explanatory approaches
The Corax as Mise en Abyme
For a deeper analysis of this type of argument, a concept that originally
derives from literary criticism proves helpful. This is the concept of mise en
abyme, or “placing into the abyss,” which describes a technique in which a
pictorial or narrative artifact embeds within itself as in a frame a smaller
reflection of itself, the sequence potentially appearing to recur infinitely as
in parallel mirrors.

372

PR 43.4_04.indd 372 10/28/10 10:16:16 AM

 perelman’s interpretation of reverse probability arguments

After André Gide had first described the phenomenon using the
expression “mettre . . . en abyme” in his 1893 Journal (1948, 41), the term “mise
en abyme” was coined and successfully established as a technical term—
against rival expressions, such as “composition” or “construction en abyme”
advocated by Pierre Lafille (1954, 25, 206)—by Claude-Edmonde Magny
(1950, 269–78). The metaphor was clearly suggested to Gide by his passion
for heraldry. In heraldry, “abyme” is the technical term for the center or heart
of a shield (escutcheon), in which a smaller shield (inescutcheon), which
may or may not reproduce the design of the larger shield, may be placed.
The Swiss critic Lucien Dällenbach, in his magisterial book The Mirror
in the Text (Le récit spéculaire), criticizes the ambiguities and inconsistencies
inherent in Gide’s description and examples and in Magny’s and Lafille’s
analyses, as well as the confusion of the mirror and shield imagery (1989,
7–26). He points out that the term is in fact being used equivocally in a
number of different senses. He proposes a threefold classification of dis-
tinct types of mise en abyme (1989, 24, 35–38, 107–13, esp. 110). The first type
is simple duplication, in which the reflection reflects a similar object, and
the whole is reproduced only once in the embedded part (such as in the play
within the play in Hamlet or in paintings by Hans Memling and others that
show a mirror that reflects the entire scene). The second is infinite duplica-
tion, in which the reflection mimetically reproduces the same object, which
inevitably creates an infinite regress in that the embedded reproduction will
itself necessarily embed its own reproduction and so forth, such as in the
identical miniature reproduction of a heraldic shield in its own abyme (it
is evident that the reproduction can only be exact if the miniature shield
will again bear its own reproduction in still smaller size in its own center
and so forth ad infinitum) (1989, 111) or on the famous boxes of Dutch
Droste cocoa powder, which display a nurse carrying a serving tray with a
box of the same brand, decorated with a smaller replica of the same picture.
This picture again necessarily embeds a still smaller reproduction and so on
(1989, 195–96n6; see Magny 1950, 271–72, 276, for a discussion of the Quaker
Oats box, an exact counterpart), which is why this effect is also known as
the Droste effect. It can easily be created by parallel mirrors reflecting each
other (see Magny 1950, 270; Dällenbach, 1989, 22) or by filming a televi-
sion screen while displaying the output on the same screen; it creates the
visual effect of a descent into a fathomless abyss, which the etymology of
abyme would in fact also connote (Magny 1950, 277). The third kind of mise
en abyme is paradoxical or aporetic duplication, which applies when the

373

PR 43.4_04.indd 373 10/28/10 10:16:16 AM

 manfred kraus

reflection reflects the object itself so that the embedded part simultaneously
embeds the whole. This generates a closed self-referential loop similar to a
Möbius strip and results in a logical paradox.
The relevance of this concept of mise en abyme to the pattern of the
corax as described in The New Rhetoric is evident. First and foremost, it
offers an appropriate theoretical model for the potential infinity of reit-
erations, a feature that most clearly distinguishes the Perelmanian corax
from its ancient predecessors. Moreover, the technique of mise en abyme
can also be used as a deliberate “device.” Magny, in her analysis, repeatedly
refers to it as a “procédé” (1950, 269–70, 273–74), and she also observes that
its immanent infinite reiterations will create in the observer something like
a vertigo (vertige) (1950, 270, 273, 277; cf. Dällenbach, 1989, 22: “dizzying”),
which, as has been demonstrated, is also the case in any excessive employ-
ment of the corax.
The corax is clearly not the first kind of mise en abyme described by
Dällenbach, since the reproduction in the argument is neither necessarily
single nor only similar. But the other two categories most certainly could
both be relevant. The question that remains to be answered is: what does
a corax ultimately produce, an infinite series of progressive repetitions or
a circular self-referential loop? Hoffman strongly emphasizes the element
of self-referentiality as basic to this argument pattern (2003, 504). But its
most distinctive feature clearly is the potentially infinite accretion of the
argument, which would seem to make a case for type 2 (infinite regress)
rather than type 3 (closed loop). Either way, we still need another model to
explain the corax, since the mise en abyme can account for the infinity of
the repetitive process but cannot account for the dialecticity of the recipro-
cal pendulum swings.

The Corax as Quasi-Logical Argument

Since Perelman and Olbrechts-Tyteca consider arguments from probability
to be “quasi-logical arguments” (1969, 255–60, esp. 257 and 260)—though,
it must be admitted, they rely on a fairly quantitative and statistical inter-
pretation of probability—it would seem only natural to try to interpret the
corax, too, in this way. Quasi-logical arguments, as they explain, “lay claim
to a certain power of conviction, in the degree that they claim to be similar
to the formal reasoning of logic or mathematics” (1969, 193); “since there
are formal proofs of recognized validity, quasi-logical arguments derive
their persuasive strength from their similarity with these well-established

374

PR 43.4_04.indd 374 10/28/10 10:16:16 AM

 perelman’s interpretation of reverse probability arguments

modes of reasoning” (1969, 193; Dearin 1982, 79–81). Such logical relations
include, for example, contradiction, identity, and transitivity (Perelman and
Olbrechts-Tyteca 1969, 194; Dearin 1982, 81–83). Quasi-logical arguments—
in Perelman and Olbrechts-Tyteca’s view—include, among others, not
only arguments from probability but also from ridiculousness (§49), from
reciprocity (§53), and from division (§56), which all seem to address essen-
tial components of the corax. While these techniques appear to share the
qualities of formal demonstrations, Perelman notes in an essay also titled
“The New Rhetoric” that for all quasi-logical approaches, “complementary,
nonformal hypotheses are necessary to render the argument compelling”
(Perelman 2001, 1398).
What, then, might be the logical or mathematical structures on which
the corax could be modeled as a quasi-logical argument? In order to
defend its validity on psychological grounds, Hoffman maintains that self-
referential logic is typical of the human mind even if it is circular (2003, 505)
and in support of this idea invokes two classical self-referential paradoxes
that in his view “one must mention” (2003, 504): Russell’s paradox (2003,
504–5) and the Cretan liar (2003, 506). Both are in fact related to each
other.
The Cretan liar paradox, which found its way even into Douglas R.
Hofstadter’s discussion of self-reference in his famous book Gödel, Escher,
Bach (1979, 15–17), runs as follows: Epimenides the Cretan once stated: “All
Cretans are liars.” Is this statement true or false? Originally, Epimenides
may have meant “Cretans other than myself,” so that there was no self-
reference and thus no logical problem. It can be pointed out, however, that
even if he literally meant “all Cretans,” the statement may still be consis-
tently false, if there existed at least one Cretan other than Epimenides who
at least once spoke the truth. Epimenides’ statement would then simply be
false without implying its own truth. It cannot, of course, be consistently
true. The paradox only arises if the statement is taken to be strictly self-
referential, if it is understood as in fact asserting its own falsehood. In this
case, Epimenides’ statement, if deemed true, would imply its own falsehood,
yet its falsehood would inversely imply its own truth and so forth ad infini-
tum, which creates a paradox. Other variants of the same paradox include
statements such as “This statement is false” or “This statement is not true”
(known as the strengthened liar and attributed to Eubulides of Miletus)
or more elaborate, related formulations such as what is known as Quine’s
paradox (Quine 1976). Hofstadter’s conclusion is that “the Epimenides
paradox is a one-step Strange Loop, like Escher’s Print Gallery” (1979, 17)

375

PR 43.4_04.indd 375 10/28/10 10:16:16 AM

 manfred kraus

or, for that matter, Gödel’s incompleteness theorems. Such a strange loop
indeed generates an infinite dialectical batting back and forth of two recip-
rocal arguments just as in the corax, which is also essentially related to the
problem of truth and lie.
Various solutions have been proposed for the Epimenides paradox.
A number of philosophers have tried to solve the problem either by inter-
preting the statement as internally self-contradictory and thus a priori
false (Prior 1958) or by suspending the applicability of a dual system of
truth values (Priest 1979, 1984; Barwise and Etchemendy 1987; Mills 1998).
Alfred Tarski, on the other hand, proposes that self-referentiality in the
discussion of truth values can be avoided by postulating levels of languages,
each of which can predicate truth (or falsity) only of languages at a lower
level (1956).
The Epimenides paradox is cited explicitly by Bertrand Russell in
“Mathematical Logic as Based on the Theory of Types” (1908, 222). Russell
uses the paradox as a point of departure for the discussion of mathematical
problems, including the paradox known as Russell’s paradox. This paradox,
discovered by Russell in 1901, is the most famous of the set-theoretical par-
adoxes. The paradox arises within naive set theory (such as Gottlob Frege’s)
if one considers the set of all sets that are not members of themselves. For
such a set will be a member of itself if and only if it is not a member of itself,
from whence the paradox arises. One of the popularized variants of this
paradox is the barber’s paradox. The barber shaves all and only those men
who do not shave themselves. Will he or will he not shave himself?
Russell himself responded to the paradox by developing his theory of
types. Recognizing that self-reference was at the bottom of the paradox,
Russell argued that self-referentiality can be avoided by arranging all prop-
ositions into a hierarchy. The lowest level of this hierarchy would consist
of sentences about individuals, the next lowest of sentences about sets of
individuals, the next lowest of sentences about sets of sets of individuals,
and so forth. It would then be possible to universally refer to all objects for
which a given predicate holds only if they are all at the same level or of the
same “type” (Russell 1903, 1908).
There are many more cognate logical paradoxes that arise from the
problem of self-reference. What they all have in common is the triggering
of an infinite dialectical swinging back and forth between two reciprocal
statements or their truth values. All of them in some way concern the
problem of truth and falsity. This would apparently make them the perfect
logical models onto which to project the corax as a quasi-logical argument.

376

PR 43.4_04.indd 376 10/28/10 10:16:16 AM

 perelman’s interpretation of reverse probability arguments

But they inevitably lead to an impasse, since this perpetual reciprocating

process cannot be stopped at any point. They can thus be described as
instances of a mise en abyme of the third category.

A Procedural Solution
As we have seen, Russell’s and Tarski’s solutions to logical paradoxes of
this type amount to stipulating different hierarchic levels of statements.
How would this apply to the classical corax as a rhetorical argument? For
although the paradoxes have an undeniable explanatory force when it comes
to the basic structure of the argument, the corax pattern is different from all
these logical or mathematical paradoxes in one crucial point. Whereas in
these paradoxes, as the tables are turned back and forth, the argument basi-
cally always remains the same (step 3—and any odd-numbered step—will
be completely identical with step 1, and step 4—and any even-numbered
step—will be the same as step 2), the situation is much different in the
case of the corax. Here the argument piles one step on the other and grows
longer with each single step; step 3 is by no means the same as step 1, nor
is step 4 the same as step 2. Instead of merely swinging back and forth
between two stationary terminals as in the logical paradoxes, the argument
as it were spirals outward from its initial center of gravity (the initial simple
probability argument), gets progressively longer and more and more cum-
bersome, and thus gradually ceases to be persuasive and becomes ridiculous.
In this respect, the corax is again closer to a mise en abyme of Dällenbach’s
second category. While in the logical paradoxes there is simply an inescap-
able impasse, here there is a procedural progress. As a corollary, the hierar-
chy that has to be introduced into the logical and mathematical paradoxes
as a means of breaking the impasse is in the case of the corax already part of
the pattern. Thus, to save the corax as an expedient argument, a procedural
solution is called for.
The difference is also obvious from an ancient example of a paradox
ostensibly cognate to the corax, which is also invoked by Hoffman (2003,
504; see Cole 1991, 65–66; Schiappa 2003, 227), namely the story that Tisias
refused to pay the tuition fee to his teacher Corax before he had won his
first case in court; he then failed to go on to practice judicial speech and so
was consequently taken to court by Corax. Tisias argued: “If I win, I need
not pay, since I have won the case; and if I lose, I ought not pay, because I
have not been well taught.” On which Corax countered: “On the contrary,
if you lose, you have to pay, since you lost the case; but if you win, you ought

377

PR 43.4_04.indd 377 10/28/10 10:16:16 AM

 manfred kraus

to pay, because you have been taught so well.” This example, however, is
only apparently similar to the corax, since there is one important differ-
ence: there is no element of anticipation of audience expectation involved,
which is the essentially rhetorical ingredient in the corax argument. It is
simply a logical paradox that leads to an impasse. Any further reply will
only repeat what has been said already. Significantly, in the story the judges
are reported to have proved unable to decide, their solution being to throw
both Corax and Tisias out of court with an outcry.
In the corax as a rhetorical argument, however, the process of pendulum
swings clearly has to stop at some point. For how many moves and counter-
moves can one foresee and recalculate before one goes mad? As Perelman
and Olbrechts-Tyteca observe, in principle such a descent into the abyss
can go on forever (“Can the batting back and forth in the corax be con-
tinued indefinitely? It can” [1969, 459]), as in fact it does in purely logical
paradoxes. But just as the process of repetition in the visual Droste effect
can in theory go on forever but practically can go on only as long as the
resolution of the picture allows (which is not long at all, since each iteration
exponentially reduces the picture’s size), so too in a corax argument at some
point the real world will inevitably intrude; it can therefore be cut short at
any time.
The crucial factor in any rhetorical argument is audience. As early as
Aristotle we find the warning that an ordinary audience will not be capable
of following an all-too-lengthy argument (Rhetoric 1357a3–4, 10–12, 15–16,
1395b25–26) and the observation that for this reason rhetorical arguments
have to be kept brief. Hoffman, too, indicates that there is a “practical limit
of reversal (any further reversal would strain both the understanding and
credibility)” (2003, 506). Therefore, an argument of the kind described
by Perelman and Olbrechts-Tyteca will certainly qualify as a rhetorical
argument, since it will be based on the audience’s reasonable expectations,
but it will cease to be a good rhetorical argument as soon as it overstrains
the average audience’s memory and mental capacities. So it must needs be
broken off at a timely stage.
But how, in practice, to break the chain and prevent a further rever-
sal? Perelman and Olbrechts-Tyteca have a tentative solution that brings
sincerity back into the picture. They suggest that since the corax is based
on assumptions about the knowledge the arguer might have of what the
opposite party will expect, “a possible countermove is to allege ignorance
of the criteria serving as a basis for this determination” (1969, 459), that is,
to disavow one’s capacity for far-reaching foresight and thus to rule out the

378

PR 43.4_04.indd 378 10/28/10 10:16:17 AM

 perelman’s interpretation of reverse probability arguments

argument’s interpretation as an artificially calculated “device.” Such may be

the “complementary, nonformal hypotheses” that Perelman in “The New
Rhetoric” declares necessary to render a quasi-logical argument compel-
ling (2001, 1398). One might also be tempted to think of the role played
particularly in legal reasoning and decision making by what Aristotle calls
“atechnic” proof (Rhetoric 1375a22–25) or what the Rhetoric to Alexander
(1428a23–24) calls “supplementary” proof (i.e., witnesses, documents, etc.),
but apparently Perelman and Olbrechts-Tyteca have in mind a number of
more “technically” rhetorical parameters that would mark “the difference
between quasi-logical argument and formal proof ” such as “the context, the
evaluation of the situation, [and] the determination of the end” (1969, 241).

conclusion
Perelman and Olbrechts-Tyteca’s interpretation of the multiple reverse
probability argument yields a highly complex and intricate phenomenon.
As a rhetorical “device,” it belongs to the category of arguments based on
the dissociation of appearance and reality, of sincerity and dissimulation.
In this particular case in The New Rhetoric, consideration of audience
response and formal analysis of argument techniques are for once happily
interlaced.
Perelman and Olbrechts-Tyteca somewhat overrate the importance
of this type of argument in legal disputes, but they are clearly aware of
the pragmatic constraints that the limited capacity of human speakers and
audiences for foresight imposes on its use and of the hazard of ridiculous-
ness. But possibly what ultimately intrigued them about it may after all
have been the pattern of a particularly witty Jewish joke about truth and lie
that would seem to lurk behind it.

Department of Classics
University of Tübingen

works cited
Anastassiou, Jannis. 1981. “Die Wahrscheinlichkeitsargumentation als Beweismethode bei
Gorgias.” Deukalion 36:355–71.
Aristotle. 1926. Rhetoric. Trans. John Henry Freese. Cambridge, MA: Harvard University
Press.
Barwise, Jon, and John Etchemendy. 1987. The Liar: An Essay on Truth and Circularity.
Oxford: Oxford University Press.

379

PR 43.4_04.indd 379 10/28/10 10:16:17 AM

 manfred kraus

Billig, Michael. 1996. Arguing and Thinking: A Rhetorical Approach to Social Psychology.
New ed. Cambridge: Cambridge University Press.
Carawan, Edwin. 1998. Rhetoric and the Law of Draco. Oxford, UK: Clarendon Press.
Chantraine, Pierre. 1968–80. Dictionnaire étymologique de la langue grecque: Histoire des mots.
4 vols. Paris: Klincksieck.
Cole, Thomas. 1991. “Who was Corax?” Illinois Classical Studies 16 (1–2): 65–84.
Crosswhite, James. 1989. “Universality in Rhetoric: Perelman’s Universal Audience.”
Philosophy and Rhetoric 22 (3): 157–72.
Dällenbach, Lucien. 1977. Le récit spéculaire: Essai sur la mise en abyme. Paris: Seuil.
———. 1989. The Mirror in the Text. Trans. Jeremy Whiteley and Emma Hughes. Chicago:
University of Chicago Press.
Dearin, Ray D. 1982. “Perelman’s Concept of ‘Quasi-Logical’ Argument: A Critical
Elaboration.” In Advances in Argumentation Theory and Research, ed. J. Robert
Cox and Charles Arthur Willard, 78–94. Carbondale: Southern Illinois
University Press.
Dzialo, Michael G. 1998. “Legal and Philosophical Fictions: At the Line Where the Two
Become One.” Argumentation 12 (2): 217–32.
Freud, Sigmund. 1960. Jokes and Their Relation to the Unconscious. Trans. and ed. James
Strachey. London: Routledge and Kegan Paul.
Frisk, Hjalmar. 1960–70. Griechisches etymologisches Wörterbuch. 2 vols. Heidelberg: Winter.
Gagarin, Michael. 1989. The Murder of Herodes: A Study of Antiphon 5. Frankfurt am Main:
Peter Lang.
———. 1990. “The Nature of Proofs in Antiphon.” Classical Philology 85 (1): 22–32.
———. 1994. “Probability and Persuasion: Plato and Early Greek Rhetoric.” In Persuasion:
Greek Rhetoric in Action, ed. Ian Worthington, 46–68. London: Routledge.
———. 1997. Antiphon: The Speeches. Cambridge: Cambridge University Press.
———. 2001. “Did the Sophists Aim to Persuade?” Rhetorica 19 (3): 275–91.
———. 2002. Antiphon the Athenian: Oratory, Law, and Justice in the Age of the Sophists.
Austin: University of Texas Press.
———. 2007. “Background and Origins: Oratory and Rhetoric Before the Sophists.”
In A Companion to Greek Rhetoric, ed. Ian Worthington, 27–36. Oxford, UK:
Blackwell.
Gide, André. 1948. Journal, 1889–1939. Paris: Seuil.
Goebel, George H. 1989. “Probability in the Earliest Rhetorical Theory.” Mnemosyne
42 (1–2): 41–53.
Goldman, William. 1987. The Princess Bride. Screenplay. www.dailyscript.com/scripts/
princess_bride.html.
———. 2007 [1973]. The Princess Bride: S. Morgenstern’s Classic Tale of True Love and High
Adventure. New York: Harcourt Brace Jovanovich.
Hacking, Ian. 1975. The Emergence of Probability: A Philosophical Study of Early Ideas
About Probability, Induction, and Statistical Inference. Cambridge: Cambridge
University Press.
Hinks, D. A. G. 1940. “Tisias and Corax and the Invention of Rhetoric.” Classical Quarterly
34 (1–2): 61–69.
Hoffman, David C. 2003. “Reversing Perceptions of Probability Through Self-Referential
Argument: Interpretation and Analysis of Protagoras’ Stronger/Weaker

380

PR 43.4_04.indd 380 10/28/10 10:16:17 AM

 perelman’s interpretation of reverse probability arguments

Fragment.” In Proceedings of the Fifth Conference of the International Society for

the Study of Argumentation, ed. Frans H. van Eemeren, J. Anthony Blair, Charles
A. Willard, and A. Francisca Snoeck Henkemans, 503–7. Amsterdam: Sic Sat.
———. 2008. “Concerning Eikos: Social Expectation and Verisimilitude in Early Attic
Rhetoric.” Rhetorica 26 (1): 1–29.
Hofstadter, Douglas R. 1979. Gödel, Escher, Bach: An Eternal Golden Braid. New York: Basic
Books.
Jacob, Bernard E. 1996. “What If Aristotle Took Sophists Seriously? New Readings in
Aristotle’s Rhetoric.” Rhetoric Review 14 (2): 237–52.
Kennedy, George A. 1963. The Art of Persuasion in Greece. Princeton, NJ: Princeton
University Press.
Kraus, Manfred. 2006. “Nothing to Do with Truth? Eikós in Early Greek Rhetoric and
Philosophy.” Papers on Rhetoric 7:129–50.
———. 2007. “Early Greek Probability Arguments and Common Ground in Dissensus.”
In Dissensus and the Search for Common Ground, ed. Hans V. Hansen,
Christopher W. Tindale, J. Anthony Blair, Ralph H. Johnson, and David M.
Godden. CD-ROM. Windsor, Ontario: Ontario Society for the Study of
Argumentation.
Kuebler, Clark George. 1944. The Argument from Probability in Early Attic Oratory.
PhD diss., University of Chicago.
Lafille, Pierre. 1954. André Gide, romancier. Paris: Hachette.
Magny, Claude-Edmonde. 1950. Histoire du roman français depuis 1918. Paris: Seuil.
Mills, Eugene. 1998. “A Simple Solution to the Liar.” Philosophical Studies 89 (2–3): 197–212.
Olbrechts-Tyteca, Lucie. 1963. “Rencontre avec la rhétorique.” Logique et analyse 3:1–15.
Perelman, Chaïm. 2001 [1970]. “The New Rhetoric: A Theory of Practical Reasoning.”
Trans. Evelyne Griffin-Collart and Otto Bird. In The Rhetorical Tradition, ed.
Patricia Bizzell and Bruce Herzberg, 1384–409. 2nd ed. New York: Bedford
Books.
———. 1982. The Realm of Rhetoric. Trans William Kluback. Notre Dame, IN: University
of Notre Dame Press.
Perelman, Chaïm, and Lucie Olbrechts-Tyteca. 1958. Traité de l’argumentation: La nouvelle
rhétorique. Paris: Presses Universitaires de France.
———. 1969. The New Rhetoric: A Treatise in Argumentation. Trans. John Wilkinson and
Purcell Weaver. Notre Dame, IN: University of Notre Dame Press.
Pokorny, Julius. 1989. Indogermanisches etymologisches Wörterbuch. Vol. 1. 2nd ed. Bern:
Francke.
Poulakos, John. 1995. Sophistical Rhetoric in Classical Greece. Columbia: University of South
Carolina Press.
Priest, Graham. 1979. “The Logic of Paradox.” Journal of Philosophical Logic 8 (1): 219–41.
———. 1984. “Logic of Paradox Revisited.” Journal of Philosophical Logic 13 (2): 153–79.
Prior, Arthur N. 1958. “Epimenides the Cretan.” Journal of Symbolic Logic 23 (3): 261–66.
Quine, Willard Van Orman. 1976. “The Ways of Paradox.” The Ways of Paradox and Other
Essays, 1–21. Rev. ed. Cambridge, MA: Harvard University Press.
Rescher, Nicholas. 1976. Plausible Reasoning. Assen: Van Gorcum.
Russell, Bertrand. 1903. “Appendix B: The Doctrine of Types.” Principles of Mathematics,
523–28. Cambridge: Cambridge University Press.

381

PR 43.4_04.indd 381 10/28/10 10:16:17 AM

 manfred kraus

———. 1908. “Mathematical Logic as Based on the Theory of Types.” American Journal
of Mathematics 30:222–62.
Schank, Roger C., and Robert P. Abelson. 1977. Scripts, Plans, Goals, and Understanding.
Hillsdale, NJ: Lawrence Erlbaum.
Schiappa, Edward. 2003. Protagoras and Logos: A Study in Greek Philosophy and Rhetoric.
2nd ed. Columbia: University of South Carolina Press.
Schmitz, Thomas A. 2000. “Plausibility in the Greek Orators.” American Journal of Philology
121 (1): 47–77.
Spatharas, Dimos G. 2001. “Patterns of Argumentation in Gorgias.” Mnemosyne 54 (4):
393–408.
Sutton, Jane. 1991. “Rereading Sophistical Arguments: A Political Intervention.” Argumen-
tation 5 (2): 141–57.
Synodinou, Katerina. 1981. Eoika–eikós kai syggenikà apò ton Hómēro hōs ton Aristophánē:
Sēmasiologikè melétē. Ioannina: Ioseph Ph. Goura.
Tarski, Alfred. 1956. “The Concept of Truth in Formalized Languages.” Logic, Semantics,
Metamathematics, 152–278. Oxford, UK: Clarendon Press.
Tindale, Christopher W. 1999. Acts of Arguing: A Rhetorical Model of Argument. Albany:
State University of New York Press.
Turrini, Guido. 1977. “Contributo all’analisi del termine eikos: I. L’età arcaica.” Acme
30:541–58.
———. 1979. “Contributo all’analisi del termine eikos: II. Linguaggio, verosimiglianza e
immagine in Platone.” Acme 32:299–323.
Vega Renon, Luis. 1998. “Aristotle’s Endoxa and Plausible Argumentation.” Argumentation
12 (1): 95–113.
Walton, Douglas. 1992. Plausible Argument in Everyday Conversation. Albany: State
University of New York Press.
———. 1999. Appeal to Popular Opinion. University Park: Pennsylvania State University
Press.
———. 2000. “Argumentation and Theory of Evidence.” In New Trends in Criminal
Investigation and Evidence, ed. Caroline M. Breur et al., 2:711–32. Antwerp:
Intersentia.
———. 2001. “Enthymemes, Common Knowledge, and Plausible Inference.” Philosophy
and Rhetoric 34 (2): 93–112.
Winter, Michael. 1997. “Aristotle, hōs epi to polu Relations, and a Demonstrative Science of
Ethics. Phronesis 42 (2): 163–89.

382

PR 43.4_04.indd 382 10/28/10 10:16:17 AM