Logic in Ancient Rhetoric

Christof Rapp, Ludwig-Maximilians-University

1. Introduction

Early Greek rhetoricians dealt with a wide range of persuasive techniques: emotional appeals,
stylistic ornamentation, slander and eristic tricks were all part of their repertoire.
Notwithstanding this variety, there was a basic understanding that any technique of persuasion
has to incorporate elements of genuine argumentation. This is why even early systems of
rhetoric came to acknowledge the relevance of notions like proof, sign, probability, contrariety,
etc. Still, Aristotle was the first to conceive rhetoric as an endeavour that essentially relies on
arguments and, thus, requires some expertise in logic. He strikingly requires the rhetorician,
who has to deal with rhetorical proofs or arguments, to be an expert in all sorts of sullogismos
or deductive argument. At the time Aristotle wrote the core of his Art of Rhetoric, he seems to
have taken for granted that it is the dialectician who is the expert on all sorts of sullogismoi.
The logic that he thus adopts for his account of rhetorical arguments seems to be the same logic
that underlies his dialectic, which is unfolded in his Topics and Sophistical Refutations. The
underlying logic of these works includes a clear understanding of deductive arguments, the
difference between deductive and inductive arguments, and the role of premises, conclusions
and refutations. Most notably, it introduces the so-called topoi, argumentative schemes that
enable the dialectician to construct premises from which he can derive the intended conclusion.
These are the most important tools that Aristotle uses for reinterpreting the terminology of the
traditional rhetorical manuals. For example, he takes over the traditional notion of an
enthymeme, which had previously been used for condensed and antithetical formulations (see
section 2 below), and redefines it as a sullogismos that is used for rhetorical purposes. He also
tries to adapt the sullogismos as it was defined and used in dialectic to the peculiar
circumstances of a public speech, taking into account, for example, that the mostly contingent

1
and variable subject matter of rhetorical arguments seldom allows for necessary proofs. This
insight brings the role of the sullogismos, in certain cases, close to the role of the likelihoods
and probability arguments, which were prominent in early rhetoric; but, unlike his predecessors,
Aristotle is able to clearly distinguish between the modal quality of a premise and the (logical)
necessity of a conclusion. Equipped with these logical distinctions he reinterprets the use of
likelihoods in traditional rhetoric as a modal modification of the premises of a rhetorical
sullogismos.
Rhetoricians after Aristotle seemed happy to include parts of Aristotle’s terminology
and some of his definitions. However, they did not share Aristotle’s rigour in turning the core
elements of rhetoric into logical notions, thus forcing as many rhetorical tools as possible into
the straitjacket of the sullogismos. Often, they fused Aristotelian notions with the uses of these
terms in pre-Aristotelian rhetoric. Insofar as the notion of sullogismos was maintained, it was
either interpreted (as usual among Aristotle’s immediate successors) as referring to Aristotle’s
syllogistic theory, not to his dialectic, or was even mingled with the Stoic notion of sullogismos.
In general, the post-Aristotelian rhetoricians tended to disentangle rhetorical proofs and
arguments from their nexus with a certain account of the sullogismos and ended up using the
above-mentioned notions of proof, sign, likelihood etc. in a way often reminiscent of the pre-
Aristotelian rhetoricians who were not committed to an underlying logical theory. 1

2. Pre-Aristotelian accounts of rhetorical arguments

Among the main sources for pre-Aristotelian rhetoric the works of Gorgias and Isocrates
deserve to be highlighted. In his Encomium of Helen Gorgias (c. 480 - c. 380 BCE) reflects on
the nature and impact of rhetoric in general. Isocrates (436 - 338 BCE) saw his own interest in
the teaching of rhetoric as part of a comprehensive pedagogical programme, which he actually
labelled philosophia. 2 He thus offered one of the main competitors to the educational
programme of Plato’s Academy. In his Phaedrus (266d-269d) Plato gives an ironical summary
of the state of the art of rhetoric, in the course of which he makes many allusions to the Gorgian
and the Isocratean schools. In the same dialogue, Plato notoriously pictures these rhetoricians
as replacing knowledge and truth with similarity and probability. Ancient anecdotes report that

1
For surveys on classical ancient rhetoric see e.g. Kennedy (1994) and Gunderson (2009).
2
On Gorgias and the sophists see e.g. De Romilly (1992), Woodruff (1999), Consigny (2001), for an English
translation see Dillon/Gergel (2003). On Isocrates see e.g. Eucken (1983) and Papillon (2007), for an English
translation see Mirhady/Too (2000) and Papillon (2004).
2
the rivalry between Isocrates and the Academy was the reason why young Aristotle during his
time in the Academy started lecturing on rhetoric in the first place.
The most important source of pre-Aristotelian rhetoric is a work that has come down to
us as a part of the Aristotelian oeuvre under the title Rhetoric to Alexander. 3 This title derives
from a letter of dedication, which is nowadays agreed to be spurious. Although it is now
generally thought that Anaximenes (380 - 320 BCE) might have been the author of this work,
we have no reliable information about when it was written; it may have been written during
Aristotle’s lifetime (384 - 322 BCE), so that it is pre-Aristotelian only in the sense that it shows
no awareness of Aristotle's most important logical innovations.
The place to look for logical aspects of ancient rhetorical theory is the treatment of
rhetorical arguments, proofs or demonstrations in the ancient manuals. In traditional handbooks
the proof or argumentation was considered a specific part of a speech. In giving
recommendations for this particular section of the speech, rhetorical authors started to
distinguish various probative moves an orator could make. Notwithstanding the differences
between the various authors and schools most accounts of the rhetorical proofs converge in
thinking that enthymemes (whatever exactly this term means) are an important part of this
rhetorical effort. Originally the word derives from enthumeisthai, which means something like
‘considering’, ‘deliberating’, ‘pondering’. In its earlier uses an 'enthymeme' seems to be a
stylistic device, in particular a well-put thought, a trenchant formulation or a brief saying (see
e.g. Isocrates, Or. IX, 10, XII, 16 XV, 47). In the Rhetoric to Alexander the enthymeme is
treated as one of seven technical means of persuasion (pisteis). 4 It is said to derive from
contraries or contradictions in speech, in actions, and in anything else as well (10, 1430a23-4).
These seem to include (i.) contradictions in what someone says, (ii.) inconsistencies in one’s
behaviour, (iii.) contradictions between what someone says and what the same person does. The
generalising formula “and [contraries] in anything else as well” is illustrated by the remark that
one must check whether either the speech or the action is contrary “to the just things, the law,
the useful, the noble or to the possible, the easy, the probable or to the character of the speaker
or the nature of the circumstances” (1430a26-30). In addition, the author emphasises that in the
enthymeme these things must be expressed “as briefly as possible” and “in the fewest possible
words” (1430a38-9). It seems safe to say, then, that at this stage, the enthymeme is essentially
characterised by its antithetical structure and by its conciseness. This is attested by a remark in
Aristotle’s Rhetoric where, in the discussion of the fallacy from diction, he points out that the

3
On this work see e.g. Chiron (2007), for an English translation see Mirhady (2011).
4
On the division of pisteis see Kraus (2011).
3
hearers might be misled into thinking that a certain remark is an enthymeme if it is just
formulated in a concise or antithetical way (Rhet. II.24, 1401a4-6).
The Rhetoric to Alexander is also significant because it includes the most advanced
classification of rhetorical argumentation before Aristotle. Apart from the enthymeme it
discusses probability or plausibility (eikos), example (paradeigma), indication of inconsistency
(tekmērion), maxim (gnōmē), sign (sēmeion) and irrefutable proof (elenchos). All of these terms
are taken up by Aristotle’s Rhetoric – though in different senses. For the author of the Rhetoric
to Alexander something that is said is plausible (eikos) when the audience has examples of it in
their thoughts (1428a23-4). In a similar vein, Aristotle will recommend basing rhetorical
arguments on what is already accepted (endoxon) by the audience. The example (paradeigma),
by contrast, is in Anaximenes’ account evidence (e.g. a past deed or event) that must be
mentioned by the rhetorician if the audience is not yet convinced. The tekmērion is close to the
enthymeme in that it is meant to diagnose contradictions, however it does so only in a more
restricted field, i.e. with respect to words and actions. A maxim is defined as the enunciation of
a conviction concerning general matters (1430b1-2). It might be plausible or generally accepted
(endoxon) or counterintuitive (paradoxon); in the latter case it must be backed by a reason, in
the former case not. A sign (sēmeion) is, according to Anaximenes, “what usually happens
before another event, at the same time or after it” (1430b31-2); the best sign brings about
knowledge, the second best the “most persuasive opinion (doxan pithanōtatēn)”. An irrefutable
proof (elenchos) is always drawn from necessities or impossibilities. A natural necessity is for
instance that living beings need food; a necessity “as we claim it” is for instance that those who
are tortured consent to what the torturer wants them to say (1431a7-13).

3. The Aristotelian enthymeme as a sullogismos

Aristotle’s approach to rhetoric is developed in his Rhetoric. The first two books of this work
are dedicated to the task of exploring the three means of rhetorical persuasion (pisteis), the first
of which comes about through the character or, more precisely, the trustworthiness of the
speaker (ēthos), the second through the hearer or, more precisely, through the emotional
condition of the hearer (pathos), and the third through arguments, proofs or seeming proofs
(logoi). Obviously, the logical aspect of rhetoric is mostly connected with the third means of
persuasion, logos. The underlying theory of argumentative persuasion is presented in a fairly
condensed way in Rhetoric I.1-2, while the core of Book I (I.4-14) is dedicated to catalogues

4
of topoi that are said to be specific to the three different genres of public speech (i.e. judicial or
forensic, deliberative and epideictic). After the treatment of ēthos and pathos the second half of
Rhetoric II returns to the discussion of aspects of logos, including topoi for the
possible/impossible, past and future events (II.19), examples (II.20), maxims (II. 21),
enthymemes (II.22), topoi for the formulation of enthymemes in all genres of speech (II.23),
topoi for fallacious enthymemes (II.24), 5 and refutation (II.25).
In several formulations, Aristotle highlights the kinship between rhetoric and dialectic;
most notably, he says that rhetoric is a counterpart to dialectic (I.1, 1354a1), that rhetoric is like
an offshoot of dialectic (I.2, 1356a25-6), that it is a part of it (I.2, 1356a30-1) and that it is
similar to it (I.2, 1356a31). Since the notion of sullogismos is at the centre of Aristotle’s
dialectical method, it does not come as a surprise that he is keen to relate the central probative
means of persuasion within his dialectically conceived art of rhetoric to this notion of the
sullogismos. Indeed, he says in the Rhetoric that the enthymeme is a rhetorical proof and hence
a sullogismos, i.e. a deductive argument. 6 Drawing on the distinction between deductive and
inductive arguments from the Topics, Aristotle also distinguishes between deductive and
inductive versions of rhetorical arguments, pointing out that in rhetoric the enthymeme is a
deductive argument, while the example is an inductive form of argument (I.2, 1356a35-b4).
One can persuade either by giving proofs or by giving what only seems to be a proof (I.2,
1356a36); this distinction appears to correspond to the differences between genuine and merely
apparent sullogismoi and, in rhetoric, to the difference between genuine and merely apparent
enthymemes. The latter, though persuasive under certain circumstances, rely on fallacious or
deceptive patterns of argument (which Aristotle discusses in Rhet. II.24). According to
Aristotle’s presupposed moral psychology of persuasion, proofs or enthymemes are essential
for persuasion, since people are mostly or most easily persuaded if they take something to have
been proved (I.1, 1355a5-6). This is also why, in one passage (I.1, 1354a15), the enthymeme is
said to constitute the body of persuasion.
Trying to specify the precise requirements for the rhetorical use of deductive arguments,
Aristotle points out that the enthymeme should not mention premises that are self-evident and
that the enthymeme should be brief or compressed. This has been taken to imply that the
enthymeme is by definition shorter than an ordinary sullogismos and that this is the differentia

5
On Aristotle’s rhetorical topoi see Rubinelli (2003), Rambourg (2014), Havrda (2019).
6
I will stick to the Greek word sullogismos if the reference is to deductive premise-conclusion arguments as
defined e.g. in Topics I.1, 100a25-7, and Prior Analytics I.1, 24b18-20, in general, whereas I will speak of
“syllogism” only if the reference is to the categorical syllogism as defined by the syllogistic theory; see below,
section 3.2)
5
specifica of the rhetorical sullogismos. Since the notion of sullogismos was identified with the
syllogism as specified in the syllogistic theory of Aristotle's Prior Analytics, and since the
categorical syllogism in the syllogistic theory is an argument consisting of exactly two premises
(introducing exactly three general terms, one of which is the middle term that occurs in both
premises) and one conclusion, it has been inferred that the Aristotelian enthymeme is meant to
have fewer than two premises, i.e. only one premise. 7 This again was the origin for the long-
standing tradition according to which Aristotelian enthymemes are defined as shortened,
abbreviated, incomplete, imperfect or truncated syllogisms or as syllogisms in which one of the
logically required premises is suppressed (syllogismus truncatus or syllogismus imperfectus). 8
More recently (i) it has been doubted by an increasing number of authors that Aristotle ever
wished to define the enthymeme as an abbreviated or truncated syllogism; 9 alternatively, it has
been suggested that this account derives from the post-Aristotelian tradition. 10 (ii.) Among the
scholars who were willing to revive the developmental approach to Aristotle’s logic, some
argued that one of the presuppositions of the syllogismus truncatus-account, namely the
assumption that enthymemes are sullogismoi in the sense of the syllogistic theory, is not
warranted. 11 As alternatives to the traditional assumption (a) that the entire Rhetoric must be
read against the background of the syllogistic theory, 12 scholars have come to consider (b)
versions of the older thesis of Friedrich Solmsen, namely that the Rhetoric includes both pre-
and post-syllogistic layers of logic, i.e. layers that either predate or postdate the syllogistic
theory as presented in the Prior Analytics, 13 and even the view (c) that the core of Aristotle’s
account of the enthymeme includes no reference to syllogistic theory. 14 Another recent debate
concerns the question (iii.) whether the enthymeme is actually meant to be a genuine
sullogismos or whether it is only a sullogismos of a kind, i.e. a sullogismos in an attenuated
sense: Aristotelian enthymemes would include ‘relaxed’ inferences, 15 i.e. inferences that are
not logically valid. This view sometimes seems to be inspired by a tendency in contemporary
argumentation theory to warn against overrating the impact of deductively valid arguments
within the common practice of argumentation and persuasion.

7
See e.g. Cope (1867), 157-58.
8
For historical interpretations of the enthymeme see the excellent survey by Kraus (1994).
9
See e.g. De Quincey (1893), 145-46.
10
For the crucial steps see Burnyeat (1994), 39-50.
11
See Rapp (2002).
12
For a consistent modern defence of this traditional view see Sprute (1982).
13
See Solmsen (1927).
14
See Rapp (2002).
15
For this tendency see Burnyeat (1982), (1994), (1996).
6
 3.1 Is the enthymeme only a sullogismos of a kind?

The passage that first says that the enthymeme is a kind of sullogismos reads as follows:

Persuasion is a kind of proof (hê de pistis apodeixis tis) – for we are most strongly
persuaded, whenever we assume something to have been proven – the rhetorical proof
again·is the enthymeme (esti d’ apodeixis rhêtorikê enthumêma), and this is, to say it in
a word, the most authoritative among the means of persuasion; the enthymeme again is a
kind of sullogismos (to d’ enthymêma sullogismos tis), but it is the task of dialectic to deal
equally with each sort of sullogismos … (Rhetoric I.1, 1355a4-9.)

Usually the formulation “to d’ enthymêma sullogismos tis” is taken to mean that the enthymeme
is a kind of the sullogismos. The idea that there are several kinds of sullogismos in accordance
with several uses is familiar from the subdivision of sullogismoi into dialectical sullogismoi and
scientific demonstrations – depending on the epistemic status of the premises (see Topics I.1).
A corresponding reading that implies several kinds or species of sullogismos suggests itself for
the current passage, especially since the idea of different kinds of sullogismos is present in the
final remark that the dialectician is responsible peri de sullogismou homoiôs hapantos), i.e. for
each kind of sullogismos alike. Since Aristotle mentions only a few lines later (1355a13) that
there are certain (not yet specified) differences between the enthymeme and other sullogismoi,
it seems justified to read the passage as saying that the enthymeme is the kind of sullogismos
that is used in rhetoric and that there are specific differences in the rhetorical use of the
sullogismos that are to be defined in what follows. Nevertheless, it has been argued that the
formulation sullogismos tis does not necessarily indicate a proper species of a genus, for tis can
be taken as an “alienans qualification”, as Myles Burnyeat explains:

Or it might be an alienans qualification, meaning that an enthymeme is only a sort of

apodeixis, only a sort of sullogismos, not as it were your full–blooded specimen, not
something from which you can expect everything that you would normally expect from
an apodeixis or a sullogismos. 16

This suggestion has been widely accepted, presumably because it helps to explain the alleged
conundrum that, although Aristotle defines the enthymeme as a sullogismos, the logical form
of the enthymemes that are actually given as examples in the Rhetoric does not seem to conform
to that of the categorical syllogisms that we know from his Prior Analytics (see below, section
3.2). The attraction of Burnyeat’s suggestion notwithstanding, there are strong reasons for

16
Burnyeat (1994), 13.
7
abiding by the traditional reading of this passage. To begin with, the quoted passage aims at
illustrating that rhetoric properly understood must rely on dialectical competence; this point is
made by showing that the enthymeme, which is central to rhetoric, is nothing but a sullogismos
and that the consideration of all kinds of sullogismoi belongs the dialectician’s expertise. This
entire argument would be spoiled if Aristotle in the same breath were to say that, strictly
speaking, the enthymeme is no real sullogismos, but only something akin to it (which, in turn,
would imply that rhetoric does not require genuine dialectical competence, but only dialectical
competence of a sort). Moreover, there are only a few passages in the Rhetoric that use the
allegedly more cautious formulation sullogismos tis (e.g. 1400b37), whereas many other
passages straightforwardly render the enthymeme as a sullogismos (e.g. 1357a16, 1359a9-10,
1394a26-7, 1395b22) or as the rhetorical sullogismos (1356b4, 1357a23, 1358a11). Last but
not least, Aristotle assumes that there is an important difference between genuine and only
apparent enthymemes which, again, goes back to the difference between genuine and only
apparent sullogismoi (e.g. II.24, 1400b34-7); this difference would be obscured if even genuine
enthymemes failed to be genuine sullogismoi.

3.2 Sullogismos in which sense?

If the enthymeme is agreed to be a genuine kind of sullogismos, then the question arises in
which sense the enthymeme is a sullogismos. In Topics I.1, Aristotle defines the sullogismos
as:

… an argument (logos) in which, certain things being supposed (tethentôn tinôn)

something different from the suppositions (heteron ti tôn keimenôn) results of necessity
(ex anagkês sumbainei) through them (dia tôn keimenôn). (Topics I.1, 100a25-7)

According to this definition the sullogismos is (i.) an argument or, to be sure, an ordered
sequence of words (λόγος), one part of which (ii.) things posited, i.e. in a premise or in premises,
another part of which is (iii.) different from the things posited and (iv.) follows (v.) of necessity
and (vi.) through them, i.e. through the things posited. This definition clearly refers to a
premise-part and a conclusion-part of the argument; it requires that the conclusion follows by
necessity and through the premises. This is compatible with what we would call a deductive
premise-conclusion argument. However, the definition does not specify the format or the
number of the premises involved, nor does it regulate the format of the conclusion or the number
and kind of terms that are used in the argument. On the basis of this definition, then, it is

8
reasonable to expect that any sullogismos in the Aristotelian sense should satisfy the criteria of
a deductive premise-conclusion argument.
It is generally assumed that the most prominent achievement of the Prior Analytics,
namely the invention of the syllogistic theory, is (still) absent from the Topics and the
Sophistical Refutations. It is part of this theory to prove that all valid arguments can be reduced
to arguments of a certain logical forms, i.e. to the categorical syllogisms, each of which consists
in exactly two premises and one conclusion, which together involve three general terms, one of
which, the so-called “middle term”, occurs in both premises, but not in the conclusion. In the
course of this syllogistic theory Aristotle is the first to introduce letters as variables for the three
terms that are involved in a categorical syllogism, thus making plain that the validity of a valid
syllogism is independent of the content of the terms. The (presumably pre-syllogistic) Topics
together with the Sophistical Refutations show no awareness of the three syllogistic figures or
the middle term, do not use letters and do not try to regulate the number and form of the
premises of an argument. In this sense, it seems fair to say that the Topics and the Prior
Analytics represent two different approaches to the sullogismos or, if one prefers to put this in
developmental terms, that they represent different stages in the development of Aristotle’s
logic. 17
When Aristotle requires the enthymeme to be a kind of sullogismos (and not only a
sullogismos of a kind), it makes all the difference whether we think of the broad concept of a
deductive premise-conclusion argument or whether we expect the enthymeme to comply with
the logical form of a categorical syllogism in the sense of the syllogistic theory of the Analytics,
i.e. an argument that corresponds to one of the three syllogistic figures:

A belongs to B B belongs to A A belongs to B

B belongs to C B belongs to C C belongs to B
A belongs to C A belongs to C A belongs to C

In other terms, the question is whether Aristotle’s Rhetoric is more closely linked with the
Topics or with the Prior Analytics or, slightly more specifically, whether the theory of the
Rhetoric relies on the accounts given in the Topics or on those given in the Prior Analytics.

17
It is significant, however, that the definition of the sullogismos that the Prior Analytics gives (An. Pr. I.1,
24b18-20) does not significantly differ from the one given in the Topics.
9
 3.3 Aristotle’s Rhetoric between the Topics and the Prior Analytics

There is overwhelming evidence for the intimate connection between the Rhetoric and Topics
and between the projects that are carried out in these works. The connection is manifest, e.g.,
in the numerous references in the Rhetoric to the Topics (1355a28, 1356b12, 1358a29, 1396b4,
1398a28, 1399a6, 1402a35, 1403a32, 1419a24), or to “the works on dialectic” (1356b36,
1396b26, 1401a2, 1402a5). In addition, there is a considerable overlap between passages in
Rhetoric I.7 and passages in Topics III.2-3. The Topics announces that the dialectical method
is useful for “the encounter with the many”, i.e. for a philosopher’s encounter with the
democratic crowd of people (I.2, 101a30-3), and the Rhetoric responds with the remark that
arguments should be formed on the basis of common beliefs, “as we said in the Topics about
the encounter with the many” (I.1, 1355a27-9). The author of the Rhetoric is aware of the fact
that the project carried out in the Rhetoric is a dialectical one (I.4 1359b2-16), thus confirming
the kinship with the dialectical method from the Topics. In addition, many chapters of the
Rhetoric consist of catalogues of topoi and thus presuppose the method of topoi, which, again,
is introduced in the Topics.
By contrast, the Rhetoric does not mention any of the ingredients of the syllogistic
theory as presented in the Prior Analytics. For this reason, the attempt to reconstruct the
underlying logic of Aristotle’s Rhetoric as a whole on the basis of the Prior Analytics alone18
is not particularly promising. However, even if one accepts that the bulk of the Rhetoric is
guided by the underlying logical theory of the Topics and not by the syllogistic of the Prior
Analytics one has to face the fact that the Rhetoric nevertheless includes five references to “the
Analytics”. What should we make of that? One of these references (I.2, 1356b9) only concerns
the dichotomy of inductive and deductive arguments – an idea that is certainly not peculiar to
the Analytics. All the other cross-references are included in two passages I.2, 1357a22-1358a2
and II.25, 1402b13-1403a16. This unusual distribution is one reason for the conjecture that
these two passages might be later insertions, and were probably meant to revise the otherwise
pre-syllogistic treatise Rhetoric in the light of the Prior Analytics. 19 This thesis also helps to
make sense of the fact that three of these four cross-references are made in the relatively narrow
context of the discussion of sign-enthymemes, because the only discussion of enthymemes in
the Prior Analytics (in II.27, see below, section 4.3) is almost exclusively dedicated to sign-
enthymemes. It has been observed, though, that in spite of the cross-references, the treatment

18
See e.g. Sprute (1982).
19
This thesis was first formulated by Friedrich Solmsen (1929), 13–27 and 27–31, and was more recently
adopted by Myles Burnyeat (1994), 31-8 and James Allen (2001), 21.
10
of sign-enthymemes in the Rhetoric does not make actual use of the syllogistic tools that are
provided in Prior Analytics II.27, but reaches similar results without reference to syllogistic
figures. 20

4. Aristotle’s definition of the enthymeme

There is a passage in the Rhetoric that has received a great deal of attention because it defines
the enthymeme in terms that are similar to the definition of the sullogismos as we know it from
Topics I.1 and Prior Analytics I.1.

… and when, certain things being the case (to de tinôn ontôn), something different besides
them results (heteron ti [dia tauta] sumbainein para tauta) 21 by virtue of their being the
case (tôi tauta einai) either generally or for the most part (ê katholou ê hôs epi to polu),
this is called a sullogismos there (i.e. in dialectic) and an enthymeme here (i.e. in rhetoric).
(Rhet. I.2, 1356b15-7)

The formulation in this passage is a bit tricky, since Aristotle wishes to point out that the
enthymeme has the same – deductive – structure as the sullogismos in the realm of dialectic.
For this purpose, he clearly alludes to the definition of the sullogismos from Topics I.1, 100a25-
7 (or from An. Pr. I.1, 24b18-20). Just as in these passages, Aristotle refers here to premises (to
de tinôn ontôn), to a conclusion that must be different from the premises (heteron ti … para
tauta ἕτερόν τι), and to the logical consequence and emphasises that the conclusion follows
because of the premises.

4.1 Relaxed inferences?

The formulation “either generally or for the most part (ê katholou ê hôs epi to polu)” seems, at
first glance, to take the place of “ex anagkês”, i.e. “by necessity”. In the definition of the
sullogismos “by necessity’ qualified sumbainei ι, i.e. the consequence, which is usually taken
as an indication of logical necessity. If, in a similar vein, “either generally or for the most part
(ê katholou ê hôs epi to polu)” were taken as a qualification of the way in which the conclusion
follows from the premises, this would yield a surprising result. While one can plausibly read

20
See Rapp (2002), II 202-7.
21
The words dia tauta, i.e. because of them, are bracketed by Kassel (1976) as a later insertion, as he takes them
to reduplicate the point included in tôi tauta einai. Also, Kassel puts a comma after tôi tauta einai, while I,
following a suggestion made in oral conversation by Paolo Crivelli, prefer to put the comma after para tauta.
11
“results generally” as a mere variation of “results necessarily”, the adverbial qualification
“results for the most part”, would imply that the conclusion follows only for the most part and,
thus, would not follow by logical necessity at all. A number of scholars took that to mean that
in this definition Aristotle wishes to allow for the possibility that at least some enthymemes are
no sullogismoi at all, provided that “for the most part” qualifies the consequence as such.22
Aristotle’s point would then, as Allen puts it, 23 be “that an argument can fail to be a syllogism
but still qualify as an enthymeme, though its conclusion would still somehow have to follow
for the most part.”
This is one of the main pieces of evidence for the suggestion that the Aristotelian
enthymeme is not meant to be a genuine sullogismos (see above, section 3.1), at least not in
general, but rather a “relaxed argument”, i.e. an argument that, though failing to match the
requirements of a deductive argument, somehow qualifies as a reasonable argumentative move.
Others tried to avoid this consequence, e.g. by pointing out that the formulation “either
generally or for the most part (ê katholou ê hôs epi to polu)” is need not qualify the verb which
indicates the consequence (sumbainein, i.e. to result, to follow), but could equally be referred
to the immediately preceding words “by virtue of their being the case (tôi tauta einai)”. 24 This
would amount to saying that the things that are posited in the premise (or premises) are the case
either generally or (only) for the most part and that the conclusion comes about because of their
being either generally or for the most part the case. And this, again, would yield a perfectly
good sense, because a few lines later (1357a22-32) Aristotle points out that few of the things
that are the subject matter of public speeches are necessarily the case, but are the case (only)
for the most part and that, while necessary conclusions are derived from necessary premises,
conclusions about what happens (only) for the most part are derived from premises about what
happens (only) for the most part. It seems, then, that with this peculiarity of the premises of
rhetorical arguments already in mind (which, again, derives from the peculiar subject matter of
public speeches and, hence, of rhetoric), Aristotle wants to use the definition of the enthymeme
to make room for the special character of premises in rhetorical argumentation. Taken as a
formulation concerning the character of the premises, however, the definition of the enthymeme
does not include a recognition of “relaxed inferences”.

22
Grimaldi (1980), 50, Wörner (1990), 352-53, Burnyeat (1994), 19-20, Allen (2001), 32.
23
Allen (2001), 32.
24
See Rapp (2011), 404-8.
12
 4.2 The brevity of enthymemes

Aristotle acknowledges that enthymemes are traditionally expected to be brief or concise. In

his discussion of stylistic features in the third book of the Rhetoric, Aristotle is prepared to treat
conciseness as a quality that may make learning easier and more pleasant (Rhet. III.9, 1410a19-
21, III.10, 1410b17-21), which thus offers a certain retrospective justification for the traditional
conciseness requirement. With respect to the enthymeme, Aristotle complies with this
traditional expectation by saying that in the course of a rhetorical argument one should not
mention what is already well-known by the hearers (I.2, 1357a17-8) and that lengthiness makes
things unclear (II.22, 1395b25-6) and difficult to follow (I.2, 1357a11). However, there is no
passage that would present brevity or even incompleteness as a defining feature of enthymemes.
It seems instead that the enthymeme should be as brief as possible in order to best fulfil
its function. In De Motu Animalium (701a25) Aristotle associates the omission of an obvious
premise with dialectal reasoning in general and does not mention enthymemes at all. Hence it
is clear that neither brevity nor incompleteness was thought to be the differentia specifica of
enthymemes. In the Rhetoric the demand to formulate rhetorical arguments more concisely
than, say, dialectical ones is always justified by the reference to the limited intellectual
capacities of the typical audience of a public speech (Rhet. I.2, 1357a1-22, II.22, 1395b22-
1396a2). Their lack of experience in following arguments is manifest in their inability to follow
long trains of thought and to “see together” (sunoran) remote things, e.g. remote premises and
conclusions. The recommended brevity of rhetorical arguments is not a matter of omitting or
supressing logically required premises, but can be accomplished by the selection of suitable
premises, namely premises that do not need to be introduced by another deductive argument
(indeed, the key paragraph in Rhet. I.2 contrasts the enthymeme with a relatively long chain of
proof) and that are pertinent (and not too remote or too general) for the envisaged conclusion.
It is thus recommended to use premises already agreed upon or accepted (endoxa), which aligns
the enthymeme with the genus of dialectical argumentation – although the enthymeme is mostly
about things that can be otherwise and are not necessarily the case, but only for the most part,
and although it should avoid the length typical of dialectical chains of arguments.

13
 4.3 The enthymeme as an argument from signs or probabilities

In Rhet. I.2, 1357a32-3, Aristotle says that enthymemes are constructed from probable premises
and from signs. Is this meant to imply that all enthymemes are so constructed? If so, what about
deductive arguments in rhetoric that deploy neither probabilities nor signs? Indeed, the brief
discussion of enthymemes in Prior Analytics II.27 says that the enthymeme is a sullogismos
taken either from probabilities or from signs (70a10), which seems to come close to giving a
definition of the enthymeme. The fact that this conforms to the formulation in the Prior
Analytics, i.e. with a text that was probably written after the core of the Rhetoric (see above,
section, 3.3), has given rise to the suspicion that the formula that connects the enthymeme either
to signs or to probabilities in the Rhetoric is part of a later revision that tended to polish and
simplify the original and more complex account of the enthymeme. However that may be, it is
obvious that for Aristotle arguments taken from signs and probabilities make up a significant
part of enthymematic argumentation.
In An. Pr. II.27, 70a2-5, Aristotle points out that whatever people know to happen or
not to happen in a certain way for the most part (epi to polu) is probable (eikos); and this, again,
is why he calls the probable an accepted premise. Since there is a fallacy that consists in the
failure to specify the respect in which something is said to be probable (II.24, 1402a12-29),
Aristotle adds in the Rhetoric that the probable is what happens for the most part, but not in an
unqualified sense; rather it is related to that in respect to which it is probable as a universal is
related to a corresponding particular (I.2, 1357a34-b1). The adherents of “relaxed inferences”
(see above, section 4.1) have argued that enthymemes with probable premises should be
construed as inferences with a merely probable consequence (probabilitas consequentiae): if it
is likely that P, then it is likely to follow/then it will follow for the most part that Q. 25 However,
neither in the Posterior Analytics (87b23-5) nor in the Rhetoric (1370a3-4, 1371b28-30) is
Aristotle willing to construct probability arguments along those lines, but rather sticks to the
deductive model: If it is likely that P, then it follows by necessity that it is likely that Q.
As for sign-enthymemes, Aristotle distinguishes in Rhet. I.2, 1357b1-21 between
necessary and non-necessary signs. The former always yield a sullogismos and are irrefutable.
Adapting a term from traditional rhetoric, Aristotle labels the necessary type of sign tekmērion,
while he says that there is no distinguishing name for the non-necessary type. On closer
examination, it turns out that the non-necessary signs are used in two different forms, so that
we get three types of sign, from which three types of sign-enthymeme can be derived.

25
See e.g. Burnyeat (1994), 24-30.
14
(i.) She has given birth, since she has milk. (tekmērion, not refutable)

(ii.) This man has fever, since he is breathing rapidly. (refutable)

(iii.) Wise men are just, since Socrates was wise and just. (refutable)

The subdivision between two types of non-necessary signs seems to be triggered by the thought
that the non-necessary sign of type (ii.), just like the necessary one (i.) instantiates an inference
from the universal to the particular, while type (iii) of the non-necessary sign-argument
instantiates the inference from the particular to the universal. 26

In An. Pr. II.27, 70a3-38, Aristotle gives a discussion of signs in which he assigns the
different types of signs or, strictly speaking, the resulting sign-arguments, to the three syllogistic
figures, with the result that, again, there is one type of sign that yields a valid syllogism and
two types of signs that do not.

P1 All those with milk are pregnant. First Figure Example

P2 This woman has milk.
C This woman is pregnant.

P1 All those who are pregnant are pale. Second Figure Example
P2 This woman is pale.
C This woman is pregnant.

P1 Pittacus is wise. Third Figure Example

P2 Pittacus is virtuous.
C All wise men/people are virtuous.

Obviously, there is a neat correspondence between the first, second and third figure examples
from An. Pr. II.27 and the type (i.), type (ii.) and type (iii.) examples from Rhet. I.2. The second
and third figure examples in the An. Pr. are said to be refutable because they do not correspond
to any valid syllogistic mode within these two schemes (An. Pr. 70a35-6). The account of the
Rhetoric reaches more or less the same result without explicitly referring to syllogistic figures.
However, it is remarkable that the discussion of the sign-enthymeme in the Prior Analytics,
perhaps the only discussion of the enthymeme in the Aristotelian oeuvre that explicitly refers
to the syllogistic figures, mentions the phenomenon of implicit and explicit premises (80a19-

26
See Weidemann (1989).
15
20, 24-8); we get a sign, Aristotle points out, if we mention only one premise, but a syllogism
if we also make use of the other premise. If we apply this to the formulation of the necessary
sign-enthymeme in the Rhetoric (which, like most enthymemes in the Rhetoric, clearly includes
only one premise), one could wonder whether it qualifies as a sullogismos only in the sense that
it could be turned into one by articulating the implicit premise (She has given birth, since she
has milk and since whoever has milk has given birth).

What worries commentators most about Aristotle’s analysis of sign-enthymemes is that

he first introduces signs as sources for enthymemes (both in the Rhetoric and in the Prior
Analytics) just in order to acknowledge a few lines later that out of three types of signs only one
yields a sullogismos. How is this compatible with the claim that the enthymeme is a
sullogismos? What about enthymemes that are taken from refutable signs? With a view to these
latter types of sign-enthymemes, then, it seems only consistent that in a later part of the
Rhetoric, chapter II.24, Aristotle mentions signs among the topoi for fallacious arguments
(1401b9-14), i.e. for arguments that are merely apparent sullogismoi. However, in Rhet. I.2 (in
a passage that is often thought to derive from a later insertion; see above, section 3.3), he fails
to say in so many words that signs of the latter two types yield only seeming enthymemes. 27

5. Post-Aristotelian enthymemes

After Aristotle, the Peripatetic philosophers Theophrastus (c. 370 - c. 285 BCE) and Demetrius
of Phaleron (c. 350 - 280 BCE) worked on rhetoric, but we know little about their views on the
logical aspects of rhetoric. It seems that they placed special emphasis on diction or style
(although the important work On Style was falsely attributed to Demetrius). 28 Stoic
philosophers treated rhetoric, just like dialectic, as a sub-branch of logic, assuming that the main
difference between dialectic and rhetoric was a different style of presentation. 29 Unfortunately,
we know little about how the Stoics’ ground-breaking views on logic and on the nature of the
sullogismos affected their treatment of the central rhetorical tools such as the enthymeme.

27
And this absence of such a remark has led to the suggestion that at least in this, presumably late, passage
Aristotle is more appreciative of non-necessary sign-arguments: see Burnyeat (1994), 35-9, Allen (2001), 29-38.
28
On Peripateic Rhetoric in general see Fortenbaugh/Mirhady (1994) and Porter (1997), the Greek text plus
English translation of On Style is available in Innes (1995).
29
See e.g. Sextus Empiricus, Adversus Mathematicos 2.7, who refers in this passage to Zeno of Citium. For a
collection of testimonies concerning the Stoic views on dialectic and rhetoric see Long/Sedley (1987), section
31. For succinct discussion of Stoic rhetoric see Atherton (1988).
16
However, a certain impact of Stoic logic on the definition of the rhetorical sullogismos can
occasionally be observed in Cicero’s and Quintilian’s accounts of rhetorical arguments.
Occasional references to logical concepts can also be found in the fragments ascribed to
Hermagoras of Temnos (2nd century BCE) and in the early Roman work Rhetoric for Herennius
(early 1st century BCE), which was falsely attributed to Cicero. 30 Cicero (106 - 43 BCE) himself
composed several works on rhetoric or rhetoric-related topics – including De inventione (84 -
82 BCE), De oratore (55 BCE), Brutus (46 BCE), Orator (46 BCE), and Topica (44 BCE) –
which include reports on previous accounts of rhetorical terms and Cicero’s own definitions
deriving from these accounts. Marcus Fabius Quintilianus (35-100 CE) is the eminent figure of
rhetorical theory in the Roman imperial period. The fifth book of his Institutio Oratoria is
dedicated to argumentatio and considers several types of rhetorical arguments. 31

5.1 Different meanings of enthymeme

The occurrence of the enthymeme in the extant post-Aristotelian works is often connected with
an uncertainty about the exact meaning of the term ‘enthymeme’. Thus some works offer
several customary definitions. Pseudo-Demetrius in his work On Style, for example, mentions
that the enthymeme is “a kind of rhetorical sullogismos” (32, 1), but also defines it as “a kind
of thought derived either from incompatibility or from consequence (διάνοιά τις ἤτοι ἐκ μάχης
λεγομένη <ἢ> ἐν ἀκολουθίας σχήματι)” (30, 4-5). While the former definition seems to be in
line with the Aristotelian tradition, the second definition clearly follows the tradition of the
pseudo-Aristotelian Rhetoric to Alexander and includes Stoic vocabulary. Quintilian gives the
following list of different meanings of ‘enthymeme’ (V. 10. 1-3):

(i.) A thought, i.e. anything conceived in the mind.

(ii.) A maxim backed by a reason (sententia cum ratione).
(iii.) A certain inference (argumenti conclusio) derived either (iii.a) from consequents
(ex consequentibus) or (iii.b) from incompatibles (ex repugnantibus).
(iv.) A rhetorical syllogism (syllogismus rhetoricus or oratorius)
(v.) An incomplete (imperfectus) syllogism, “because it is concluded from parts that
are not distinct or not of the same number [as in other uses of the syllogismus]”.

30
For editions and French translations of these two works see Achard (1989) and Woerther (2012), for an
English translation of the latter works see Caplan (1954).
31
Cicero’s rhetorical treatises are available in various editions; for De Oratore May/Wisse (2001) is useful, for
Topica Reinhardt (2000) is exceptional. For Quintilian’s work see the Latin/English edition Russell (2001).
17
Meaning (i.) is common to many rhetorical writers; the enthymeme is an idea, something we
have in mind (sometimes as opposed to the linguistic expression, see Alcidamas, frg. 1, 117-
20), or, at least, derives from a bright idea. Meaning (ii.) might go back to Aristotle’s association
of maxims with enthymemes; he even defines the maxim as the part of an enthymeme (Rhet.
II.20, 1393a24-5), i.e. either a premise or a conclusion, if only the inferential connection
between them is omitted. Maxims backed by a reason are also treated in the Rhetoric to
Herennium (IV, 17, 24-5), but not under the heading of “enthymeme”. The bifurcation in (iii.)
has precursors in Pseudo-Demetrius and Cicero (Topica 53,1-2; see below, section 5.2). The
formulation argumenti conclusio that Quintilian uses is troublesome, as it might seem to refer
to the conclusion, as opposed to the entire argument. However, the same wording is used by
Cicero where it is either equated with the Greek ἀπόδειξις, proof (Luc. 26) or clearly refers to
the entire argument or inferential scheme (Topica 53). Quintilian reports with regard to (iii.)
that there is little consensus among the experts and that some regard (iii.a) as an epicheirema
(see below, section 5.3), while most would only accept option (iii.b), the antithetical argument,
as the proper meaning of “enthymeme”. This is clearly due to a revival of the pre-Aristotelian
tradition of the enthymeme as an antithetical figure of speech. This focus on antithetical
formulations again derives from the general interest of early post-Aristotelian rhetoricians in
diction and style (elocutio), so that even the enthymeme becomes part of the stylistic discussion.
This tendency is clearly manifest in pseudo-Demetrius and in the Rhetoric to Herennium, whose
author deals at length with reasoning from contraries (VI, 18, 25-6). As examples he mentions
rhetorical questions that include a somehow antithetical formulation, e.g. “How should you
expect one who has always been hostile to his own interests to be friendly to another’s?” This
tendency to treat enthymemes as rhetorical figures under the heading of elocutio is opposed to
Aristotle’s approach, which treats the enthymeme not as a stylistic move, but as part of what
was later labelled inventio, i.e. the discovery of the content of a speech. Accordingly, the interest
of many post-Aristotelian rhetoricians in genuinely logical questions fades away.
Meanings (iv.) and (v.), by contrast, are apparently inspired by the Aristotelian tradition,
although Quintilian is neither clear nor explicit about the meaning of syllogismus he
presupposes. It seems, however, that for him incompleteness is the defining feature of the
enthymeme. He reports, e.g., that (in accordance with meaning (v.)) the enthymeme is taken to
be a part of the syllogismus and comments on this by saying that, while the syllogismus fulfils
its purpose through all of its parts, an enthymeme is content simply to be understood (V, 14,
24). Take the following argument as a full-fledged syllogismus: “Virtue is the only good,

18
because that alone is good which no one can put to bad use; no one can put virtue to bad use,
therefore virtue is a good” (V, 14, 25). As the corresponding enthymeme from the
“consequents” (ex consequentibus) he offers: “Virtue, which no one can put to bad use, is a
good”, where the implicit reason is given in the relative clause. Another syllogismus runs as
follows: “Money is not a good; for what can be put to bad use is not a good; money can be put
to bad use; therefore money is not a good.” Again this corresponds to an enthymeme, this time
an enthymeme from the incompatibles (ex pugnantibus): “Is money, which can be put to bad
use, a good?” In addition to the premises and the conclusion, Quintilian’s examples for a full-
fledged syllogismus also include the statement of the thesis. The former enthymeme includes
no such statement and skips one premise, while the latter enthymeme, which is put as a
rhetorical question, does not mention the conclusion either. This illustrates the general point
that, although the incompleteness requirement is inspired by a certain interpretation of what
Aristotle says about the choice and articulation of premises, the Roman rhetoricians seemed
happy to extend this idea to the omission of the conclusion (which has no precedent in
Aristotle).

5.2 The enthymeme as a Stoic sullogismos

Quintilian’s examples of enthymemes mentioned in the previous section are clear examples of
incomplete or extremely condensed arguments. It is less clear, however, which account of
syllogismus is presupposed, when he characterises the enthymeme as incomplete syllogismus
or as part of the syllogismus. It is tempting to think, though, that he fuses the syllogismus
imperfectus-tradition, which ultimately derives from a certain reading of Aristotle, with a Stoic
approach to propositional logic. In its classical version, as formulated by Chrysippus, Stoic
logic acknowledges five so-called “indemonstrables”:

1. If P, then Q; but P; therefore Q.

2. If P, then Q, but not Q, therefore not P.
3. Not both P and Q; but P; therefore not Q.
4. P or Q; but P, therefore not Q.
5. P or Q; but not Q; therefore P.

A valid syllogism in the Stoic account either is, or can be reduced to, one of these
indemonstrables. In principle, Quintilian’s examples (see above, section 5.2) could be

19
reconstructed in Aristotelian term logic or in Stoic propositional logic, but the formulation he
prefers clearly hints at Stoic logic; in particular, the enthymeme ex pugnantibus seems to use
the scheme of the third indemonstrable: “Not both (something being good & something being
put to bad uses); but money can be put to bad uses; therefore not (money is good)”. An explicit
connection between this kind of enthymeme and Stoic logic is established in a passage of
Cicero’s Topica (53-55). Cicero introduces a place (locus) that, he says, is peculiar to the
dialecticians; it includes arguments from the consequents, from the antecedents and from the
incompatibles (ex repugnantibus). The consequents and antecedents are necessarily connected
with an item, while something incompatible can never be connected with an item. He then gives
a paraphrase of first three Stoic indemonstrables and assigns this distinction to the same
“dialecticians”. He does not further comment on the first two modes (leaving it to the reader to
conclude that they are suggested for the handling of consequents and antecedents), but applies
the third to the arguments from incompatibles, which, he says, are called “enthymemes” by the
rhetoricians. Cicero hastens to admit that there is a wider use of “enthymeme”, but that the term
has come to be properly applied to the arguments from the incompatibles, because they are
“most pointed (acutissima)”. In other words, Cicero takes up the old tradition of enthymemes
as antithetical formulations (as exhibited, e.g. by the author of the Rhetoric to Alexander, the
author of the Rhetoric to Herennius, by the author of On Style and as confirmed by Quintilian),
reinterprets contrariety as incompatibility, combines it with the view that the enthymeme is a
sort of sullogismos and specifies the kind of sullogismos that most deserves to be called an
“enthymeme”, namely the one that corresponds to the third Stoic indemonstrable. Quintilian’s
enthymeme from incompatibles discussed above in section 5.1 seems to be meant as an instance
of exactly this kind of enthymeme.

5.3 The enthymeme and the epicheirema

In his early work, On Invention, Cicero takes up a different tradition that associates the
enthymeme with a type of deductive argument (ratiocinatio) that consists of five steps.
According to Cicero the five steps are called propositio, approbatio propositionis, assumptio,
assumptionis approbatio, complexio (I, 37, 67). Clearly, the final part amounts to the conclusion
or summary of the argument. It is also clear that the second and fourth steps are both thought
to offer support for the first and third steps respectively. It is controversial, however, whether
the first and third steps are thought to be the two premises, i.e. the major and the minor premises,
of an Aristotelian syllogism – in this case the five-step scheme would amount to a mere

20
amplification of a syllogism – or whether the first step, the propositio, is thought to be the
announcement of a general thesis or principle. The author of the Rhetoric to Herennius, by
comparison, mentions a five-step argument – allegedly the most complete and perfect among
the deductive arguments – in which the first step, the propositio, is followed by a reason (ratio),
which is followed by a support for the reason (rationis confirmatio), which again is followed
by embellishment (exornatio) and a résumé (complexio); here (II, 18, 28) it seems clear that the
first step introduces the thesis (intension) which is backed up by the two following steps. Cicero,
at any rate, refers his five-step scheme back (falsely) to the syllogism used by Aristotle and
Theophrastus and tells us that it was taken up by the most accomplished teachers of rhetoric
(presumably Hermagoras?). What both Cicero and the author of the Rhetoric to Herennius
describe is elsewhere known as epicheirema, and the latter, again, is thought to be a rhetorical
form of argument and is regularly associated, sometimes even identified with the enthymeme.
Aristotle used the word epicheirema for the attack in a dialectical dispute, but the origin of the
use of this word for complex rhetorical arguments is entirely obscure. Quintilian defends against
Cicero a three-step version of the epicheirema and distinguishes it from the enthymeme (V,
14); its main difference from the syllogismus lies in the fact that the latter derives truth from
truth, whereas the former is used for propositions that are only probable (V, 14, 14). In the work
On Invention that is ascribed to Hermogenes, the enthymeme is recommended as additional
support for a previously formulated epicheirema (III, 8, 151-2). 32

32
For sections 5.1-5.3 I am indebted to the well-informed surveys in Kraus (1994) and Klein (1994).
21
 Bibliography

Achard, G. (1989). Rhétorique À Herennius. Paris: Les Belles Lettres.

Allen J. (2001). Inferences From Signs. Ancient Debates about the Nature of Evidence. Oxford:
Clarendon Press.
Allen J. (2007). Rhetoric and Logic. In: I. Worthington (ed.), A Companion to Greek Rhetoric.
Malden/Oxford/Victoria: Blackwell Publishing, pp. 350-64.
Allen, J. (2014). Aristotle on the Value of “Probalility,” Persuasivenes, and Verisimilitude in
Rhetorical Argument. In: V. Wohl (ed.), Probabilities, Hypotheticals, and Counterfactuals in
Ancient Greek Thought. Cambridge: Cambridge University Press, pp. 47-64.
Atherton, C. (1988). Hand over Fist: The Failure of Stoic Rhetoric. The Classical Quarterly
38/2, pp. 392-427.
Burnyeat, M.F. (1982). The Origins of Non-deductive Inference. In: J. Barnes, J. Brunschwig,
M.F. Burnyeat, M. Schofield (eds.), Science and Speculation: Studies in Hellenistic Theory and
Practice. Cambridge/Paris: Cambridge University Press and Éditions de la Maison des Sciences
De l’Homme, pp. 193-238.
Burnyeat, M.F. (1994). Enthymeme: Aristotle on the Logic of Persuasion. In: D.J. Furley and
A. Nehamas (eds.), Aristotle’s Rhetoric. Philosophical Essays, Princeton: Princeton University
Press, pp. 3-55.
Burnyeat, M.F. (1996). Enthymeme: Aristotle on the Rationality of Rhetoric. In: A.O. Rorty
(ed.), Essays on Aristotle’s Rhetoric. Berkeley, Los Angeles, London: University of California
Press, pp. 88-115.
Caplan, H. (1954). [Cicero] Rhetorica Ad Herennium. Cambridge, Mass. and London: Harvard
University Press.
Cope, E.M. (1867). An Introduction to Aristotle's Rhetoric. London and Cambridge:
Macmillan. Repr. Hildesheim: Olms 1970.
Chiron, P. (2007). The Rhetoric to Alexander. In: I. Worthington (ed.), A Companion to Greek
Rhetoric. Malden/Oxford/Victoria: Blackwell Publishing, pp. 90-106.
Consigny, S. P. (2001). Gorgias. Sophist and Artist. Columbia SC: University of South
De Quincey, Thomas (1893). Essays on Style, Rhetoric and Language, ed. by F.N. Scott,
Boston: Allyn & Bacon, pp. 145-46.
De Romilly, J. (1992). The Great Sophists in Periclean Athens, Oxford: Clarendon Press.
Dillon J. and Gergel T. (2003). The Greek Sophists. London: Penguin Classics.
Eucken, C. (1983). Isokrates: Seine Positionen in der Auseinandersetzung mit den
Zeitgenössischen Philosophen, Berlin/Boston: De Gruyter.
Fortenbaugh, W.W. and Mirhady, D.C. (eds.) (1994). Peripatetic Rhetoric After Aristotle. New
Brunswick and London: Transaction Publishers.

22
Grimaldi, W.M.A. (1980). Aristotle, Rhetoric I. A Commentary, New York: Fordham
University Press.
Grimaldi, W.M.A. (1988). Aristotle, Rhetoric II. A Commentary, New York: Fordham
University Press.
Gunderson, E. (2009). The Cambridge Companion to Ancient Rhetoric. Cambridge: Cambridge
University Press.
Havrda, M. (2019). Does Aristotle Distinguish Between Common and Specific Topoi in the
Rhetoric? Eirene 55, pp. 179-97.
Innes, D.C. (1995). Demetrius, On Style (= Loeb Classical Library, Aristotle XXIII, pp. 307-
530). Cambridge, Mass. and London: Harvard University Press.
Kassel, R. (1976). Aristotelis Ars Rhetorica. Berlin: De Gruyter.
Kennedy, G.A. (1994). A New History of Classical Rhetoric. Princeton: Princeton University
Press.
Kennedy, G.A. (2007). Aristotle, On Rhetoric. A Theory of Civic Discourse. 2nd edition. New
York and Berlin: Oxford University Press.
Klein, J. (1994). Epicheirem. In: G. Ueding and W. Barner (eds.), Historisches Wörterbuch der
Rhetorik, Vol. 2, Tübingen: Niemeyer, pp. 1251-1258.
Kraus, M. (1994). Enthymem. In: G. Ueding and W. Barner (eds.), Historisches Wörterbuch
der Rhetorik, Vol. 2, Tübingen: Niemeyer, pp. 1197-1222.
Kraus, M (2011). How to Classify Means of Persuasion: The Rhetoric to Alexander and
Aristotle on pisteis. Rhetorica 29/3, pp. 263-79.
Long, A.A. and Sedley, D.N. (eds.) (1987). The Hellenistic Philosophers. 2 vols. Cambridge:
Cambridge University Press.
May, J. and Wisse, J. (2001). Cicero: On the Ideal Orator. New York: Oxford University Press.
Mirhady, D.C. and Too, Y.L. (2000). Isocrates I. Austin: University of Texas Press.
Mirhady, D.C. (2011). Rhetoric to Alexander (= Loeb Classical Library, Aristotle XVI, pp. 449-
656). Cambridge, Mass. and London: Harvard University Press.
Papillon T.L. (2004). Isocrates II. Austin: University of Texas Press.
Papillon T.L. (2007). Isocrates. In: I. Worthington (ed.), A Companion to Greek Rhetoric.
Malden/Oxford/Victoria: Blackwell Publishing, pp. 58-74.
Porter, S.E. (1997). Handbook of Classical Rhetoric in the Hellenistic Period, 330BC-AD400.
Leiden: Brill.
Rambourg, C. (2014). Topos. Les Premières Méthodes D'Argumentation Dans La Rhètorique
Grecque des Ve–IVe Siècles. Paris: Vrin.
Rapp, C. (2002). Aristoteles: Rhetorik. Übersetzung, Einleitung, Erläuterungen. 2 vols. Berlin:
Akademie Verlag.

23
Rapp, C. (2011). Aristotelische Grundbegriffe in der Theorie der juridischen Argumentation.
Rechtstheorie 42, pp. 383-415.
Reinhardt, T. (2000). Marcus Tullius Cicero: Topica. Oxford: Oxford University Press.
Rubinelli, S.R. (2003). Topoi e idia nella Retorica di Aristotele. Phronesis 48, pp. 238-47.
Russell, D.A. (2001). Quintilian, The Orator’s Education, 5 vols. Cambridge, Mass. and
London: Harvard University Press.
Solmsen, F. (1929). Die Entwicklung der Aristotelischen Logik und Rhetorik. Berlin:
Weidmannsche Buchhandlung.
Sprute, J. (1982). Die Enthymemtheorie der aristotelischen Rhetorik, Göttingen: Vandenhoeck
& Ruprecht.
Weidemann, H. (1989). Aristotle on Inferences from Signs (Rhetoric I 2, 1357b1-25).
Phronesis 34, pp. 343-351.
Woodruff, P. (1999). Rhetoric and Relativism: Protagoras and Gorgias. In: A.A. Long (ed.),
The Cambridge Companion to Early Greek Philosophy. Cambridge: Cambridge University
Press, pp. 290-310.
Wörner, M. (1990). Das Ethische in der Rhetorik des Aristoteles, Freiburg/Munich: Alber.
Woerther, F. (2012). Hermagoras. Paris: Les Belles Lettres.

24