8 Between Variety and Unity

How to Deal with Plato’s

Dialectic1
Walter Mesch

Plato’s dialectic has many faces. Most prominent, of course, is the general
distinction between different methods displayed in his dialogues: critical
enquiry or elenchus, hypothetical ascension and the twofold procedure
of division and collection. But these methods also show a considerable
variety depending on the topics discussed, argumentative aims, dialogical
settings and the ontological assumptions at play, suggesting an even more
diversified picture. In view of these findings, it seems far from obvious
that there is one single unified conception of dialectic in Plato’s dialogues.
Quite the contrary, we find ourselves confronted with the threat of incon-
sistencies, even with respect to fundamental issues. It is not even possible
to give an unambiguous and comprehensive answer to the important
question whether, for Plato, dialectic leads to knowledge of transcendent
ideas, because in some dialogues this appears to be the case, and in others
not, or at least not clearly.
Against this background, interpreters since the 19th century have
worked out developmental models, assuming an evolution of Plato’s
thought and ascribing different methods and conceptions to different sets
of dialogues written in successive periods of time. At first sight such an
assumption seems natural, since philosophers usually write most of their
texts one after the other and not all of them simultaneously. In addi-
tion, they often tend to change their minds over time by acquiring new
insights or doubts. And apparent inconsistencies can, at least sometimes,
be explained on the basis of such changes. When it comes to Plato, how-
ever, developmental models are not at all unproblematic. In particular,
it is far from clear whether the dialogues show a development or evo-
lution in his thought and in which sense this could be the case. Many
assumptions made in this regard have been severely criticised, especially
during the last decades, and in particular concerning extreme versions of
developmentalism that presuppose outright conceptual breaks between
different stages.2
I consider much of this criticism well-­founded and think this also holds
for the developmental account of Plato’s dialectic. So even if there is a

DOI: 10.4324/9781003111429-9
170 Walter Mesch
certain kind of development, we still have to ask how the variety and
unity of dialectical methods in his dialogues can be understood. In this
chapter, I explain why many aspects of developmental interpretations are
problematic and suggest a more plausible account of Platonic dialectic.
The first and introductory section considers methodological matters, with
the primary aim of delineating in which sense the dialogues may actually
be said to show a certain development and in which senses this has to
be denied. The second and main section supplements this methodologi-
cal introduction with considerations regarding the objects of dialectical
methods. In view of the special status of transcendent forms, the primary
question here is whether certain methods are exclusively concerned with
certain objects. The third section finally tries to give an answer to my titu-
lar question, how to deal with Plato’s dialectic. I suggest that both variety
and unity are essential features of the dialectical enquiries we find in his
dialogues. Plato works with basically unchanging methods, but employs
them in a highly context-­sensitive way according to an ambitious scheme
of philosophical investigation and education.

Dialectical Methods in Plato’s Dialogues and their Context

According to a widespread interpretation, not restricted to extreme ver-
sions of developmentalism, we have to distinguish between at least three
different dialectical methods and assign them to three different sets of
dialogues, sorted according to major steps in Plato’s career (cf. Robinson
19532, 70). In the early dialogues, we find a method of critical enquiry
relying on question, answer and refutation. Professing not to know any-
thing important himself, Socrates only examines the knowledge claims of
his interlocutors, mainly concerning virtue, and shows why they are not
sound. Though searching for knowledge, he always ends at an impasse.
In the middle dialogues, this critical and negative enquiry is replaced by
a more affirmative and positive method, unfolding definitional knowl-
edge of virtues and leading from hypothetical assumptions to premises
or principles. The explanation of the famous simile of the line in the
Republic articulates the main way this method can be understood. Plato’s
new constructive approach is dominant throughout the middle dialogues.
Socrates no longer professes not to know anything important, but intro-
duces a theory of transcendent ideas that serves as the basis for an articu-
lated psychology and detailed conceptions of ethical and political virtues.
In the late dialogues, we finally encounter a twofold procedure of divi-
sion and collection referring no longer to ideas, but to concepts. With the
exception of the Philebus, Socrates either is altogether absent, or steps
aside in favour of other discussion leaders, who develop new ontological,
cosmological, ethical and political theories of a less transcendent nature.
Now, I do not think that everything in this familiar picture is false.
There are, of course, considerable differences between different dialogues,
 Between Variety and Unity 171
which have to be recognised and assessed, and I consider it appropriate to
distinguish different sets of dialogues along these familiar lines. But in my
view, it is extremely important not to overestimate and misinterpret these
differences. I just mention a few important aspects, without explaining
them in any detail.

(1) There is no clear-­cut distinction between different dialectical methods

simply following each other in different sets of dialogues. What we
rather find is a highly complex variation of different ways to combine
these methods. Refutations are not limited to early dialogues, but can
also be found later and here still play an important role, e.g. in the
critique of the Protagorean relativism we find in the Theaetetus (see
Larsen 2019). The hypothetical method, delineated in the Phaedo
and the Republic,3 is at least prepared by the Meno (Stemmer 1992,
262–7). And, most importantly, the method of division and collection
is not peculiar to the late dialogues. On the contrary, it seems to be
employed in all dialogues, albeit in different ways. This point is of
great importance, and I will come back to it later.
(2) The difference between the Socrates of the early and of the later dia-
logues should not be assimilated to or identified with the difference
between the historical Socrates and Plato, in the way suggested by
many developmental interpretations. This has often been pointed
out (e.g. in Kahn 1981, 305–6 arguing against Guthrie and Irwin).
There are mainly two reasons. Firstly, we do not know enough about
the historical Socrates to evaluate the accuracy of Plato’s account.
Although it is very likely that there is some similarity, we simply can-
not ascertain it in a substantial way, not even by comparison with
other accounts.4 Secondly, it is extremely unlikely that historical
accuracy was Plato’s aim at all. He rather tried to show how philoso-
phy has to be carried out, using the Socratic way of investigating as
a model. This at least holds when “Socratic” very broadly refers to a
dialogical method of questioning and examining knowledge claims
concerning the virtues (Wieland 1982, 50–70).
(3) One should not overlook that Plato’s presentation tries to mediate
obvious differences between his Socrates figures. In the face of this, it
is rather unconvincing to claim regarding the content and method of
their philosophies “that they contrast as sharply with one another as
with any third philosophy you care to mention”, as Gregory Vlastos
puts it (1991, 46). Socrates in the middle dialogues not only finds the
knowledge of virtue he was looking for in the early ones, but his new
methodical ways are also affirmative counterparts to his old critical
instruments. It is, therefore, unnecessary to suppose he mysteriously
turns from a sceptic into a dogmatist. What Socrates in the middle
dialogues considers as justifiable is rather what, if possible, defies any
refutation. In the Phaedo such a logos is called dysexelenktotatos
172 Walter Mesch
(85d; cf. Stemmer 1992, 194). At least in this fundamental respect,
the Eleatic Visitor and the Athenian, who replace Socrates in some
late dialogues, are still committed to a Socratic view. They may be
more assertive, but they still carry out a critical investigation aim-
ing at reliable knowledge and in this regard “there is nothing in
the actual processes of reflection exhibited by these speakers that
marks them off from Socrates” (Rowe 2006, 167). This construc-
tive move is already prepared in the early dialogues, since refutations
here are often ordered in stages, at least indirectly conveying impor-
tant insights, and combined with gentler dialectical procedures like
maieutic and induction. A good example of this is the Meno, which
has three rather illuminating refutations of three failing, but gradu-
ally improving attempts to define virtue (71d–9e).
(4) This methodical link is confirmed by the critical attitude towards
the sophists, not merely prominent in the early dialogues, but rather
running through Plato’s entire work. An obvious connection holds
between the Protagoras and the Theaetetus. In the early Protagoras
its title character could be regarded as the paradigmatic sophist pro-
viding a political education through effective rhetoric. Protagoras
here, after all, claims he is the first to pursue the old sophistical art
openly, calling himself a sophist (316c–17e). Yet for a critique of his
relativism we have to consult its later treatment in the Theaetetus
(161b–86e). If one considers the dramatic dates of the two dialogues
it seems as if Plato deliberately wanted to depict Socrates as criti-
cally engaged with this paradigmatic sophist from the beginning to
the end of his career, because in the Protagoras Socrates is still very
young while in the Theaetetus he is fast approaching death.5 Other
later dialogues, such as the Republic, the Sophist or the Philebus,
also contribute substantially to Plato’s understanding of the soph-
ists. Whereas in the early dialogues the critique is primarily aimed
either at their deceptive rhetoric, as in the Gorgias, or at their eristic
dialectic, as in the Euthydemus, in the later dialogues the focus shifts
to the fundamental attitude towards virtue, being and truth presup-
posed by both main expressions of sophistry, i.e. deceptive rhetoric
as well as eristic dialectic. The Republic explains how a bad educa-
tion can spoil philosophical souls, turning promising youngsters into
sophists (489d–496e) and the Philebus supplements this by referring
to the confusion the one and the many in our logoi can cause if its
inherent indeterminacy is not appropriately shaped through dialecti-
cal method (15d–17a: cf. Schmidt-­Wiborg 2005, 159–73).
(5) The critical attitude towards the sophists is not only articulated
explicitly and directly in passages criticising them, but also implicitly
and indirectly by way of the dialectical practice displayed throughout
the dialogues. The methodical and critical links are thus combined
with and reinforced by an important dramatic link. For the early
 Between Variety and Unity 173
dialogues this should be quite obvious, since here the lively dramatic
presentation can hardly be overlooked. Just consider the very tell-
ing change of roles occurring in the Gorgias, when the rhetorician
Polus fails to question Socrates dialectically, whereas Socrates in
turn presents a rather long speech concluding his methodical critique
of sophistical rhetoric and indicating how a philosophical rhetoric
could look (462b–466a: cf. Mesch 2007). Later dialogues are, of
course, in general less lively and dramatic. But things have not com-
pletely changed. In the Philebus we meet a tired hedonist so unfit for
the strains of dialectical discussions that a more skilled colleague has
to take over to get the depicted part of the dialogue started (11a–c:
cf. Wieland 1982, 73). And in the Sophist, the interlocutors, looking
for the sophist, “find” the philosopher first (253c). Maybe this is the
reason why there is no dialogue called the Philosopher next to the
Sophist and the Statesman. But even if there is no Philosopher, possi-
bly because, given the limitations of philosophical writings, it would
be inappropriate to write about the highest principles (Szlezák 2011,
32), philosophical dialectic is, without any doubt, a central topic in
the Sophist. And this not only holds for the famous passage 253d–e
determining the tasks of a dialectician in a very abbreviated and con-
troversial way (cf. Hoffmann 1996, 100–110), but also for the larger
context, where the Visitor discusses the communion of kinds and
shows how to deal with them.
(6) The methodological link between the dialogues originates in the
double-­sided conviction that there is no good life without examina-
tion, since virtues are not helpful without knowledge, and no worth-
while examination without a relation to the good life, since in the
end what we really aspire to is eudaimonia. The first aspect is explic-
itly expressed in the Apology (38a), the second is implied in the con-
centration on virtues prominent throughout the early dialogues. In
view of the Republic both convictions probably have to be taken in a
less strict manner. After all, in the ideal state only philosophers live an
examined life strictly speaking, although guardians and citizens also
are guided by it (cf. Kraut 2006, 232), and dialectic is the only true
science, although mathematical sciences are its preparatory exercises
(cf. Mittelstrass 1997, 230–43). Nonetheless, the Republic famously
discusses the good as the highest “object” of learning (megiston math­
ema: 505a), indicating that there is still an essential unity of practical
and theoretical perspectives. Against the background of this thematic
link, it would come as a surprise if there was any purely theoretical
investigation in the late dialogues. And, indeed, nothing of this sort
can be found. In the case of the Timaeus and the Sophist this may
be less clear than in the case of the Philebus, the Statesman and the
Laws, which mainly discuss ethical or political issues. The Timaeus,
however, is closely linked with the political theory of the Republic
174 Walter Mesch
by its introduction and by its cosmology focusing on a world for
human beings including their good lives (Brisson 1996). And even the
Sophist at least points to ethical problems through the discussion of
the sophistical practice of deception leading to the problem of nega-
tion and thereby to the core of the ontological problem.
(7) As explained above, Socrates in the middle dialogues shifts his atten-
tion from the critical to the constructive potentials of refutations:
What defies every possible refutation can be regarded as knowl-
edge. This methodological link is combined with the conceptual
link between the definitional question and the theory of ideas or
forms. As most scholars argue, there are no transcendent forms in
the early dialogues, or at least no evident examples thereof (Dancy
2004, 4–19). But the theory of forms we find in the middle dialogues
articulates what we actually refer to when we ask a definitional ques-
tion. “What is justice?” more precisely means “What is the form of
justice?”, thereby making it clear that we are not referring to single
instances, but to justice as such. Closely related to this is the twofold
theory of anamnesis, without reference to transcendent forms (Men.
80d–6c), and with reference to such forms (Phd. 72e–8a). How we
are to understand these versions has been a matter of controversy in
various respects. There are, however, good reasons not to consider
them incompatible (Williams 2002). Moreover, one should bear in
mind that already the early dialogues introduce certain forms. The
most famous example can be found in the Meno, where Socrates
asks for one single and identical form (eidos) of virtue due to which
(di’ ho) the many and various virtues are virtues (72c). There are
also other well-­known examples, notably in the Laches and in the
Euthyphro (cf. Kahn 1981, 313). Normally, these forms in the early
dialogues are considered to be either universals in a purely logical
sense or else concepts. But this leaves open the question how they are
to be understood ontologically. Calling them immanent, just because
they are not clearly stated to be transcendent, should be avoided.
After all, there has to be a real difference between these forms and
their instances, one could argue, not only a logical one, if they really
are essences (Dixsaut 2001, 28–33). Seen from this angle, the forms
of the early and of the middle dialogues are more closely related than
developmental accounts normally suggest. How the universals or
concepts of the late dialogues fit in I discuss in the following section.

To sum up: Instead of breaks, we find a considerable number of links

mediating between the different methods, conceptions and styles of the
dialogues. Hence, there is no need to presuppose a development of Plato’s
thought in order to explain these differences, although the assumption of
a certain kind of evolution may appear quite natural or even inevitable.
What we rather have to do, is to interpret the dialogues themselves and
 Between Variety and Unity 175
the manifold relations holding between them. There are, of course, many
different ways of doing this. Just consider the groupings of dialogues we
already find in Antiquity, assembling Trilogies or Tetralogies according
to competing pedagogic schemes, or the dramatic sequence of some dia-
logues ordered according to the life of the Socrates figure. For Plato’s dia-
lectic, however, it seems of prime importance to understand the relation
between his different dialectical methods. And for this relation, it again
seems crucial to find out whether or not they refer to a set of fundamen-
tally different objects, the early elenchus to universal forms either imma-
nent or without articulated ontological basis, the method of hypothesis of
the middle dialogues to transcendent forms and the late collections and
divisions to universal forms once again immanent or at least not clearly
transcendent. If there really was such a double switch first to transcen-
dent forms and then away from them again, it would be hard to defend
a more unitarian interpretation of Plato’s dialectical methods. Therefore,
in the main section I want to concentrate not only on these methods, but
also on their objects in order to reinforce my critique of developmental-
ism and to work out an alternative way of dealing with Platonic dialectic.
Given the great number of relevant texts and their enormous complexity,
I have to restrict myself to some vital aspects.

Objects and Methods of Dialectic

As we have already seen, the link between the definitional question and
the theory of transcendent forms speaks strongly against a conceptual
break between the early and the middle dialogues. Even if there are no
clearly transcendent forms in the early dialogues, one should not assume
that they are clearly immanent. What we rather find is that the onto-
logical basis for the definitional question here has not yet been explicitly
articulated, although transcendent forms, at least in some passages, like
in the famous example from the Meno I mentioned above, do not seem
too far away. The late dialogues, however, are much more problematic in
this respect. Here we have to presuppose that Plato had already worked
out his theory of transcendent forms, but this theory no longer seems
to play a major role after having been confronted with famous objec-
tions, formulated in the Parmenides, and altered or alluded to at other
places, like in the Philebus. In view of these objections, many revisionist
interpreters have argued that Plato gave up his theory of transcendent
forms altogether (e.g. Ryle 1939) or at least transcendent forms as para-
digms (e.g. Owen 1953). Now, I do not think this can be true. But before
explaining why, the situation in the middle dialogues first has to be con-
sidered in a little more detail. In doing so, I want to show (1) that even
here transcendent forms are just the primary, not the only objects of dia-
lectic and (2) that in referring to transcendent forms and to other dialecti-
cal objects the method of hypothesis works together with elenchus and
176 Walter Mesch
division. This not only helps to bridge the gap between the middle and
the late dialogues, but also allows a better understanding of the relation
between the early and the middle ones. I begin with aspect (2).
There are two main passages explaining hypothetical ascension and
its reference to transcendent forms, one in the Republic and one in the
Phaedo. I mainly concentrate on the Republic, where the relevant con-
nections are worked out in more detail. Interpreting the simile of the
line’s famous ascension to the unhypothetical principle and subsequent
descent is very controversial, since the explanations for these procedures
given in the text are extremely sparse. As I understand it, there are good
reasons for thinking that the unhypothetical principle must be identified
with the idea of the good and that it is impossible to understand the good
here exclusively as a practical or epistemological principle. We also have
to understand the ontological relation holding between the forms and the
good. The most promising way of doing so seems not to consist in claim-
ing a formal causality of the good for ideal attributes of forms, but rather
in taking up the mathematical examples pointing to the central role of
unity as a basis for the constitution of numbers and, more generally, of
order (cf. Strobel 2004, 130–4). At any rate, there can be no reasonable
doubt that the method of hypothesis according to Socrates’ explanation
starts with forms taken as hypotheses, climbs up to higher ones and,
after having seized the unhypothetical principle, climbs down again to
forms without using anything perceptible, but only these transcendent
forms as such (all’ eidesin autois di’ autōn eis auta: 511c). Basically the
same picture emerges from the Phaedo, because in the relevant passage
the method of hypothesis also undoubtedly refers to transcendent forms
(99d–102a, esp. 100b).
Against this well-­known background, I want to point out first that
already in the Republic and the Phaedo, hypothesis is combined with
elenchus and division, at least if divisions are taken in a broader sense
not restricted to special features of the late dialogues. For the Republic
this becomes clear in the discussion of the dialectician following the three
similes. As Socrates asserts, we call “someone dialectical who grasps the
explanation (logos) for the essence (ousia) of each thing”. Immediately
afterwards, this is also said to apply to the good. “Unless someone is
able to separate the idea of the good from all other things (apo ton allōn
pantōn aphelōn) and to distinguish it in the explanation (dihorisasthai
tō[i] logō[i]), and, going through every [assault or] refutation (dia pantōn
elenkchōn), as it were in battle, eager to examine (elenkchein) things not
in accordance with opinion, but with being, to get his explanation intact
through all this, you will not claim that he knows the good itself or any
other good” (R. 534b–c, my translation). As Dorothea Frede argues, the
method of the dialectician in this passage is characterised by three fea-
tures: (1) Dialectic works with definitions, since giving the explanation
of the essence obviously means to define a thing. (2) This presupposes a
 Between Variety and Unity 177
distinction of the relevant thing from all other things, suggesting a method
of systematic division and collection as presented in Phaedrus 265c-­e,
even if it is not certain that such a method had already been worked out
systematically. (3) The results have to be confirmed through assault and
defence. Thus, refutations according to this explanation of dialectic still
constitute a part of its methods (Frede 2004, 149). To my mind, all of this
is convincing. For the idea of the good, feature (1) could appear problem-
atic, because it is not clear in which sense there really can be a definition
of the good, if definitions concern essences or being in the strict sense, i.e.
transcendent forms, and the good has to be located “beyond being” (R.
509b). But it should be clear that Socrates here considers it at least pos-
sible by separating the good from everything else.
In the Phaedo, we find essentially the same combination of methods
with reference to transcendent forms, although the relevant passage
mainly focuses on forms as formal causes for the attributes of things
participating in them. Socrates here obviously looks for the strongest
(errōmenestaton: 100a) or safest (asphalestaton: 100d) hypothesis that
allows him to get through a discussion without being refuted or falling
(pesein: 100d). Thus, the relevance of refutations appears obvious. It may
be doubted, however, whether definitions are really needed, since one
can know that everything becomes what it is through participating in its
form or proper being (idios ousia: 101c) without already knowing what
this form is. By all means Socrates could confidently claim that every
beautiful thing becomes beautiful through participating in beauty with-
out already having its definition. The same may still be true in climbing
up to higher hypotheses finally reaching something satisfactory (hikanon:
101e), at least if this is only something satisfactory for the actual interloc-
utors in a given context and not the same as the unhypothetical principle
of the Republic (Stemmer 1992, 266). If we interpret the passage this
way we should rather compare it with the Meno where the hypothesis
that virtue can be taught is related to the higher hypothesis that virtue is
knowledge and then again to the still higher, but not absolutely certain
hypothesis that virtue is something good (Stemmer 1992, 264). And for
such an ascent we absolutely do not need definitions, since the failure of
Meno’s definitional attempts was precisely the reason for falling back on
hypotheses in the first place. Nevertheless, a certain version of division
would still be needed. After all, it is hardly possible to decide whether or
not to accept that virtue is knowledge or something good without distin-
guishing knowledge from ignorance and good from bad. Thus, we can
conclude that in the Phaedo transcendent forms are objects of a dialecti-
cal investigation also working with all three main methods.
In what follows, we will see that such a combination of methods
likewise applies to other objects of dialectic. This can hardly come as
a surprise since we have already found important methodological simi-
larities between the Phaedo and the Meno, while only the Phaedo clearly
178 Walter Mesch
presupposes transcendent forms. But in order to spell this out in a more
systematic way I first have to say something about aspect (1) and explain
why in the middle dialogues transcendent forms are only the primary, but
not the only object of dialectic. At least the starting point for this expla-
nation is fairly obvious: In the middle dialogues only transcendent forms
are strictly, permanently and in every respect what they are, and thereby
qualify as true objects of reliable knowledge on the basis of an intellec-
tual activity, whereas things participating in them never strictly, perma-
nently and in every respect possess the attributes they acquire through
participation, and therefore do not allow more than unreliable opinions
on the basis of perceptions. Only perfect being (pantelōs on) is perfectly
knowable (pantelōs gnōston: R. 477a). Socrates points this out in a con-
siderable number of well-­known passages, not always mentioning all rel-
evant aspects but arguing very consistently, referring sometimes to the
difference between knowledge and opinion (R. 476c–480a), sometimes
to the cognitive presuppositions of acquiring knowledge (Phd. 73b–76d)
and sometimes to a single example of a form and things participating in
it, such as beauty (Smp. 209e–12c). And if dialectic is the tool for acquir-
ing this knowledge—as we have already seen in the dialectical passages
from the Republic and the Phaedo—, it evidently has to relate primar-
ily to transcendent forms and help us acquire knowledge of them. In
spite of all this, such forms cannot be the only objects of dialectic. On
the one hand, it appears to be impossible to attain knowledge of forms
without considering perceptible things participating in them. And, hence,
dialectic also has to relate to such things, although they cannot be known
in a strict sense. On the other hand, there seem to be clear instances
of dialectical procedures leading to a kind of definitional knowledge, in
which these procedures do not relate to forms. Crucial examples are the
Socratic determinations of the human soul we find in the Phaedo and the
Republic. I consider both points briefly in turn.
Defining a form like virtue in the Meno or justice in the Republic
requires reference to an intelligible unity that is not to be confused with
the perceptible plurality of different instances. Nevertheless, recognising
a certain similarity or identity between instances of the defined form can
be helpful for finding the definition. Maybe it is not entirely impossible
to find a definition without any explicit inductive component, since tran-
scendent forms are ontologically independent of participating things, and
anamnesis, at least in the Phaedo, has a strong a priori foundation in these
forms. Yet such a definitional procedure already requires access to the a
priori foundation of knowledge. And this is more than we can expect for
every definitional enquiry and every enquirer. In a situation in which we
still find it difficult to remember what we already know, it hardly appears
possible to do without a certain inductive component in the definitional
enquiry, because then we have to access the forms through the things
participating in them. It is true, opinions rely ultimately on perceptible
 Between Variety and Unity 179
things, in which, according to the Republic, being and non-­being are
mixed, whereas knowledge relies ultimately on intelligible forms or
unmixed being (476c–480a). But in order to attain knowledge, there has
to be a transition from perceptible things to intelligible forms, even if,
in this transition, the epistemological distinction between opinion and
knowledge necessarily appears to get somewhat blurred, raising doubts
about the underlying ontological distinction between perceptibles and
forms (cf. Horn 1997, 291–2). I think Plato could very well claim that,
considered appropriately, there are no opinions with respect to forms and
that what appears as such an opinion really is nothing but an unfinished
transition to the knowledge of forms. In any case, there has to be such a
transition, and since it is a central task of dialectic to achieve this, dia-
lectic has to refer both to things participating in forms and to forms that
things participate in. Nevertheless, for Plato there is no real knowledge of
perceptibles. Only forms can be objects of knowledge in the strict sense.
But if an investigation of perceptibles can contribute to the knowledge
of forms and if it is a central task of dialectic to manage this transition,
perceptibles must at least be regarded as objects of dialectic in a broader
or secondary sense. Although transcendent forms are its strict or primary
objects, there is no dialectic without some cognitive grasp of perceptibles.
Moreover, even if we had already attained full-­fledged access to transcen-
dent forms, including the capacity to understand how these participate
in other forms and to define them on the basis of this understanding, it
would be an odd thing to assume that such knowledge of forms was pos-
sible in the absence of any grasp of or acquaintance with participating
things. It may, then, not be necessary to rely explicitly on any induction
or to refer explicitly to participating things. But a dialectician should
still be able to take account of the participation, without confusing the
form and what participates in it (R. 476c7–476d5). And this is enough
for rejecting any strictly exclusive reference of dialectic to transcendent
forms. So, even in the middle dialogues, there is really no knowledge of
these forms without any grasp of or acquaintance with perceptible things
participating in them.
In addition, Socrates clearly develops different versions of a definitional
knowledge of the soul, without assuming that soul itself could be a form.
It may be true that he nowhere gives an obvious definition of soul in the
same way he gives a definition of justice. But he certainly formulates,
discusses and affirms its essential properties, which could become part of
a definition, like the tripartite structure in the Republic (R. 435a–441c)
or the absence of any corporeal composition and the essential liveliness
in the Phaedo (78b–80c, 105b–107a). Moreover, these properties of the
soul here are clearly not marginal cases that dialectic can also apply to,
but central topics of Socrates’ investigations. As is well-­known, in the
Republic it proves necessary to work out the tripartite structure of the
soul in order to show how human souls can be just. And in the Phaedo
180 Walter Mesch
the immortality of the soul even provides the leading question of the
whole dialogue. Nevertheless, the human soul is only a source of our
bodily life or our cognitive capacity, which, through understanding of
transcendent forms, can appropriately govern the human body. Thus, on
the one hand, it is impossible to deny that in these dialogues it is an
important task of dialectic to investigate and determine the nature of
souls.6 And on the other, soul itself clearly cannot be a transcendent form.
Soul is only a close relative of forms imitating some of their essential
properties. Just consider the third immortality proof of the Phaedo (78b–
80e). It is quite obvious that here again different dialectical procedures
are combined, because Socrates, in arguing for a fundamental similarity
or affinity of souls and forms, combines a presupposed division and a
hypothetical argument. So, this passage treats of the object-­aspect (1)
and the method-­aspect (2) at the same time. Souls are either composite
like bodies or incomposite like forms—this is the presupposed division
(again in the broad sense not limited to the stricter procedure of the late
dialogues)—, and they can only be mortal if they are composite since
only what is composite can perish or die by falling apart—this is the
hypothetical part of the argument. The justification of this hypothesis,
notoriously, is not really shown, but it is only made plausible by certain
indications, pointing towards an underlying similarity between souls and
forms (Frede 1999, 64–5). Whereas bodies are visible, souls are invis-
ible like forms. Bodies change permanently, but forms always remain the
same, and through understanding forms, souls adjust themselves to their
unchanging nature. Forms are of a divine nature governing bodies, and
in understanding forms, souls also imitate this divine governance. These
indications make it quite clear that the nature of the human soul, under-
stood appropriately, is meant to be ontologically dependent on the forms.
This ontological dependence is further worked out in the fourth and
concluding proof arguing that the soul is essentially alive (102a–7a)
which comes after the more detailed explanation of forms I already men-
tioned above. And this dependence also has an epistemological dimen-
sion, already present in the second proof referring to immortality as a
precondition for anamnesis (72e–7a). Hence, investigating and deter-
mining the nature of the soul dialectically is not inconsistent with forms
being primary objects of dialectic, since it is precisely the soul’s depen­
dence on the forms which makes this possible. The fact that souls are
also objects of dialectic cannot put into question the status of forms as
primary objects of dialectic, because on the one hand they themselves can
only be treated dialectically in the face of their similarity with forms, and
on the other they convey formal structures to bodies, thereby govern-
ing them. It is just through these formal structures and their similarity
with transcendent forms that souls can become important objects of dia-
lectic. In working out these structures Socrates not only uses the meth-
ods of hypothesis and division (in a broad sense), as pointed out for the
 Between Variety and Unity 181
third proof, but also aims at refutations. Here again in the Platonic text
both aspects, object and method, are important at the same time. This
allows me to conclude the consideration of soul as an object of dialectic
with a short digression concerning the relevant methods. From the first
proof arguing for a circular movement of life and death (70d–2e) until
the last one, Socrates always tries to refute objections made by Cebes or
Simmias. Even before the immortality question has been raised, Cebes’
initial doubts concerning the life and the death of a philosopher already
give the decisive motivation for the ensuing discussion (61d). Later he not
only puts forth widespread doubts about immortality (70a), but together
with Simmias also tries to refine them (77a–e). Against this background,
everything Socrates argues for in the whole dialogue can be seen as an
attempt at persuading them (anapeithein: 77e), implying a successive refu­
tation of their objections.
Let us now come briefly to the late dialogues. I must restrict myself
to some remarks (1) concerning the previous critique of forms in the
Parmenides, (2) the occurrence of forms in the late dialogues, mainly in
the Timaeus, (3) the relevance of cosmological innovations for dialectic,
(4) the status of ontological and dialectical innovations, mainly in the
Sophist, and (5) the connection of these with the middle dialogues.
There are no good reasons for assuming that Plato gave up his theory
of transcendent forms, since the critique formulated in the Parmenides is
less relevant and much weaker than often supposed. To begin with, the
old Eleatic here criticises a theory of transcendent forms introduced by a
young Socrates, who obviously still has to learn quite a bit,7 not the sub-
tler theory of the Republic or the Phaedo, and still less the supplemented
versions of it we find in the late dialogues (cf. Prior 1985, 51–84). As
Socrates claims in the Parmenides against the Eleatic position, perceptible
things can at one and the same time be similar and dissimilar by par-
ticipating in the different forms of similarity and dissimilarity. It would
only be surprising if the form of similarity was dissimilar and the form of
dissimilarity similar (129a–b). These are obviously abstract forms char-
acterising not only what is common to things belonging to a certain spe-
cies, but rather mutually exclusive universal determinations, like one and
many, being and not being, identity and difference. Nevertheless, Socrates
denies that forms “can be mixed among themselves and again be sepa-
rated” (129d). And if this is meant to exclude not only a corporeal mixing
of forms in the way this might wrongly be assumed to occur in perceptible
things, but also any formal mixing such as participation of forms in other
forms, then hypothetical ascension, division and collection and refuta-
tion referring to forms would be impossible. Is Socrates really unaware
of this? Hard to know. In any case, Socrates does not realise that many
aspects of Parmenides’ criticism concerning the participation of things
and the separation of forms are very problematic. Particularly striking is
the assumption that the form has to be in the thing in order to allow for
182 Walter Mesch
the thing to participate in the form, because it is obviously unnecessary
for and fatal to a theory of transcendent forms. In the Phaedo, Socrates
also speaks of presence and community referring to the participation of
things in transcendent forms, but avoids any commitment to a literal
understanding of immanence (110d). Later he even mentions a greatness
and smallness “in us” (en hēmin: 102d–e), thereby indicating that there
are immanent forms. But, of course, these immanent forms cannot be
simply identified with transcendent ones, as we shall see presently on the
basis of the Timaeus.
Transcendent forms are still important for the late dialogues. This is
particularly clear in the case of the Timaeus, because in its cosmology the
ontological distinction of “what is always the same, but never becoming,
and what is always becoming, but never truly being” plays a fundamen-
tal role, together with the epistemological distinction between thinking
and opinion (27d–28a). Already here the determinations of transcendent
forms from the middle dialogues are taken up almost verbatim.8 Yet
Timaeus later in his speech, while discussing the three genera—becoming
(gignomenon), form (eidos) and receptacle (dechomenon)—, even more
explicitly tries to show why there have to be forms in the sense of some-
thing strictly in itself (auto kath´ hauto: 51b), incapable of entering into
any perceptible thing (52a). They are necessary for the difference between
true opinion and knowledge and, hence, indispensable for thinking (51d).
Moreover, the paradigm of the visible cosmos, the so-­called perfect living
being (pantelōs zōon), is strictly transcendent. No part of it can enter into
the extended matter of the cosmos, but only its images, imitations, or rep-
resentations (eikones, mimēmata or aphomoiōmata: 50c, 51a). Thus, we
here clearly have to recognise different versions of forms, transcendent
ones on the level of the paradigm and immanent ones within the cosmos,
worked out on the basis of a demiurgical model. Quite close to this is the
technical model of the Philebus, distinguishing limit, unlimited, cause and
mixture (peras, apeiron, aitia and meikton) both in view of the cosmos
and of our own life, although it is again less clear whether limit refers to
transcendent or immanent forms. Regarding ethical purposes Socrates
probably does not consider it necessary to decide the matter. But earlier
in the text, there are also passages that unambiguously mention problems
of transcendent forms without dismissing them (15a–c). And close to the
end of the text Socrates explicitly claims that dialectic (hē tou dialeges­
thai dynamis: 57e) is the most exact of all sciences since it refers to being
in the strict sense (peri to on kai to ontōs kai to kata tauton aei pephykos:
58a), whereas all disciplines referring to becoming (gignomena) only lead
to opinion (doxa: 59a).9
The cosmology of the Timaeus aims at bridging the gap between tran-
scendent forms and perceptible things by explaining perceptible things
as parts of a visible cosmos which itself is explained as an image of tran-
scendent forms (29b). After all, the demiurge throughout aims at making
 Between Variety and Unity 183
the cosmos as similar to its perfect paradigm as possible (29e), equipping
it, for this purpose, with a perfect soul governing its body as perfectly as
possible (30b). And the culmination of similarity is reached in the produc-
tion of time as an arithmetically moving image of eternity resting in the
one (Mesch 20162, 167–86). This cosmological mediation of an ontologi-
cal gap on the one hand has important consequences for the ethics of a
mixed life in the Philebus, according to which the aim is to shape our life
in a way that makes it similar to the structures of a mixed cosmos. On
the other hand, it has important consequences for the science of dialectic,
since the different levels of forms distinguished in this cosmological con-
text allow us to explain why dialectical methods primarily working on
the basis of transcendent forms can reach further down into the sensible
realm and also refer to souls and to perceptible particulars. Dialectic can
achieve this because souls and perceptible particulars are constituted by
immanent forms shaping corporeal matter on the basis of transcendent
forms.10 The central mediating entity here is the world soul governing
the world body with reasonable and regular movements, being imitated
much less perfectly by the human soul governing its own body. And the
world soul can achieve this because it is constituted by a mixture of the
indivisible and the divisible, i.e. transcendent forms and corporeal matter,
which are again divided into being (on), the same (tauton), and the dif-
ferent (heteron). As has often been observed, this basic structure is very
much reminiscent of the greatest kinds of the Sophist (Brisson 1996, 233).
But when it comes to the Sophist itself, what about the ontology we
find there? Is it really a radically new one, dismissing transcendent forms
and replacing them by a logical analysis referring to the different senses
of being? I think both claims are highly unlikely. The first claim requires
that we interpret the critique of the friends of the forms, put forward by
the Eleatic Visitor in his famous discussion of the so-­called battle between
Giants and Gods or Gigantomachia, as the author’s self-­critique, refer-
ring to transcendent forms in the middle dialogues, and this makes no
sense in the dramatic context of the Sophist (Taylor 1961, 44–7). Instead,
one has to consider it as a reference to one of the Italian schools, i.e. to a
Pythagorean or Eleatic theory of forms (Ebert 1998, 86). This, of course,
does not preclude any similarity with Plato’s middle dialogues. But there
is no simple identity, as I will explain in the next paragraph. The second
claim is problematic since the text, even though it may be read as distin-
guishing between different senses of being, gives no explicit analysis of
them (Fronterotta 2011, 54). What we rather find is an explanation of
the highest or most common forms that we already presupposed when
we indicated Socrates’ deficits in the Parmenides and which are also men-
tioned in the Theaetetus (184–6). While Socrates in the Parmenides uncon-
vincingly denies that forms can be mixed with other forms or participate
in them, the famous theory of the greatest kinds (megista genē) in the
Sophist shows how forms can participate in one another, focusing mainly
184 Walter Mesch
on difference in order to explain sophistic deceptions. And the overall
significance of this seems to consist in the formal analysis of a structure
presupposed by all transcendent forms. If we understand the greatest
kinds along these lines they have to be transcendent forms themselves,
even though they cannot relate to one another as species and genera, like
ordinary forms (Frede 2004, 148). The forms occurring in the definitions
of the sophist or the statesman, it may be argued, belong to a lower level
since they are given according to a well-­known subordination of genera
and species.11 Do all the divisions we find in the late dialogues refer to
transcendent forms? Probably not, but most of them do. Referring only
to linguistic distinctions, in any case, will not be sufficient for the pur-
poses of dialectic—we have to agree not only about the name (onoma),
but also about the thing (pragma: Sph. 218c). This is also clearly ruled
out by the critique of such divisions in the Statesman (262a–3b). Dividing
humans into Greeks and barbarians must be considered inappropriate
because there is no form of barbarity. Quite to the contrary, barbarians
are only a part of human beings singled out by a name (Szlezák 2011,
28). Considered this way there is, as far as I can see, nothing in the divi-
sions of the late dialogues or in the theory of the greatest kinds speaking
against the assumption that transcendent forms here still are the primary
objects of dialectical methods. But what about the famous critique of the
friends of the forms we find in the Gigantomachia? Does it really rule
out that transcendent forms contribute anything to Plato’s ontology in
the late dialogues? I do not think this is the case and want to address this
widespread assumption in the last step of this section.
The Giants in the Sophist hold that there are only corporeal things
(246a–b), and the Gods, or friends of the forms, hold that intelligible
and incorporeal forms are true being (alethinē ousia), whereas corporeal
things possess “moving becoming” (genesis pheromenē: 246b–c). As the
Visitor points out against the Giants or corporealists, souls, which they
also regard as corporeal (246e), can only become just or wise through the
possession and presence (hexis kai parousia) of justice or wisdom, both
of which are incorporeal. And these incorporeal entities are as well, since
whatever has the power (dynamis) of acting or suffering must be (247e).
Speaking of “power” may sound anachronistic, but can be justified, since
“capacity”, “potentiality” or “possibility” clearly are too passive and neg-
ative (Gonzalez 2011, 67–8). The Visitor’s arguments against the Gods or
friends of the forms are more complicated and controversial because here,
in using the same criterion of power, he first argues that being insofar as
it is intelligible has to possess motion, given that knowing means act-
ing and being known suffering (248d–e), and then in a second step goes
even further by claiming that perfect being (pantelōs on) has to possess
life, soul and intelligence (248e). Trying to avoid absurd consequences
following from these two related assumptions, several commentators
have suggested that motion here should not be taken as real, but only as
 Between Variety and Unity 185

something accidental occurring in a thinking soul, and that pantelōs on

should not be understood as perfect, but only as comprehensive being,
including not only immoveable entities, but also things and souls together
with corporeal motion and cognitive activities. But this is unnecessary
and misleading (cf. already de Vogel 1953 and, more recently, Gonzalez
2011 or Mesch 2011). On the one hand, such a comprehensive under-
standing would not represent any progress with respect to the position
of the friends of the forms, who admit from the very beginning that there
are not only forms, but also participating things (246b–c) and knowing
souls (248a). Hence, pantelōs on has to be understood with reference
to transcendent forms as in the Republic and as a close parallel to the
pantelōs zōon of the Timaeus. On the other hand, motion can be under-
stood in an intelligible sense referring to dialectical procedures. And this
allows us to avoid the absurd consequence that something in becoming
known should change essential properties. Against the background of the
middle dialogues another important aspect should be taken into account.
The Phaedo “already shows the inadequacy of the position of the friends
of the forms”, since a sharp distinction between the changing corrupt-
ible and the unchanging incorruptible is not really adequate for the soul
(Gonzalez 2011, 78). Relying on such a distinction the third immortality
proof already implies a close vicinity of soul and transcendent forms.
And the fourth proof finally tries to answer the objections of Cebes and
Simmias in advancing the claim that the soul essentially lives. Here, it is
true, Socrates does not explicitly draw the consequence that transcendent
forms as such have to be alive to provide an adequate ontological basis
for the soul. But this consequence already seems to be implied. In any
case, recognising that already in the middle dialogues transcendent forms
are the primary, but not the only object of dialectic helps us understand
why there is no break between them and the late dialogues.

How to Deal with Plato’s Dialectic

Coming back to my leading question I want to draw some final conclu-
sions in the light of the previous sections.

(1) Although it is natural to assume that Plato, just as any other phi-
losopher, acquired insights over time and to a certain extent changed
already established positions, there are no methodological, ontologi-
cal or other relevant breaks in his dialogues suggesting that his concep-
tion of dialectic can only be interpreted consistently by presupposing
a development of his thought in a psychological or intellectual sense.
Therefore, one should avoid making strong developmentalist claims.
(2) The only philosophically meaningful development when it comes
to Plato’s dialogues is the ongoing philosophical articulation of the
186 Walter Mesch
single and unitary conception of dialectic they themselves present.
In this articulation, we find first an emphasis on refutation, shifting
then to hypothesis and finally to division and collection. But this
emphasis should not be confused with a successive relevance of only
one method at a given time, because in one way or the other they are
very often employed in combination.
(3) Insofar as there is no definition without division and collection, at
least in the standard case of definitions presupposing a scheme of
subordinated concepts or forms, this has to be considered the cen-
tral method throughout. But the other methods are also important.
Refutation is necessary for dealing with inappropriate positions and
hypothesis for the combination or justification of definitions, which
are either still being investigated or already known. This structure
is quite varied in the dialogues, but relevant throughout. In order to
adequately understand the methods being employed it is as necessary
to discover the reason for the variation in each case (which is mainly
found in the topic of the given dialogue and its interlocutors), as it is
to see the underlying unity.
(4) This methodological unity, together with the other links discussed
in the long main section two above, have an ontological basis in the
link between the definitional question and transcendent forms. This
link is still relevant in the late dialogues, although here the empha-
sis lies less on the transcendent status of these forms than on their
formal presuppositions and on the possibilities of their application.
Nowhere in the late dialogues are transcendent forms abandoned or
missing altogether, not even in the very concrete and practically ori-
ented discussions of the Laws. The members of the nocturnal council
must know not only that the good and the beautiful are manifold,
but also in which sense they come together in a unity (966a). And the
same is required in the case of other important questions including
the unity of all virtues. Hence, it is hard to deny that the members
of the nocturnal council have to master key competences of dialectic
concerning transcendent forms (Frede 2004, 154).
(5) The plan for this continuous articulation can be seen best in the simi-
les of the Republic, although even they present only a very abbrevi-
ated sketch. Thus, in the simile of the line, dialectic is only mentioned
as a dialectical capacity of the intellect concerning transcendent
forms whereas perceptible things and their images seem to lie beyond
its scope. But it is, of course, the dialectician Socrates who is present-
ing all sections of the line and the corresponding way of education in
the simile of the cave. This speaks in favour of a broader understand-
ing of dialectic not limited exclusively to the upper half of the line,
but also relating to the lower half. It is far from clear how we should
understand the way up to the good as the unhypothetical principle
and the way down to the sensible world. The cave, however, shows at
 Between Variety and Unity 187
least that education through dialectic leading first up to the good and
then down again does not start and end with transcendent forms, but
on the lowest levels with perceptible things and their images. Such a
broad understanding of dialectic fits better with the practice of the
middle dialogues, because it regards dialectic as primarily concerned
with transcendent forms without being limited to them.
(6) According to this scientific and educational scheme, the early dia-
logues start paving the way to knowledge through definitional ques-
tions and by refuting inadequate answers, thus pointing out the
deficiencies of opinions. But part of this task still has to be realised in
the middle and late dialogues. Another relevant part here is the hypo-
thetical ascension to higher stages already prepared in the Meno and
then suggested mainly by the Phaedo and the Republic. This ascent
is, however, not accomplished completely in the middle dialogues.
As we have seen, there are e.g. good reasons to interpret the greatest
kinds in the Sophist as the highest or most common forms reflecting
formal features essentially belonging to all transcendent forms. And
if this is correct, it seems natural to understand this formal reflec-
tion on transcendent forms as a late contribution to the ascent men-
tioned in the simile of the line. The double task of deducing from
higher principles and of descending back into the cave again is also
carried out in many different versions. Parallel to the simile of the
line, but not identical with it, deduction plays a special role in the
Phaedo, where Socrates suggests that one tests hypotheses not only
by ascending to higher ones but also by investigating whether their
logical consequences lead to contradictions (101d: cf. Stemmer 1992,
263). Moreover, for getting back into the cave, ethically or politi-
cally, the Republic and the Laws work out two different versions,
the first one more abstract, the second one more concrete. Thus, in
dealing with the leading scheme provided by the simile of the line, an
exclusivist allocation of certain dialogues to certain stages would be
inappropriate.
(7) The similes of the Republic and other key passages of Plato’s dia-
logues explicitly focusing on dialectic are of primary importance for
understanding his thought and have to be preferred to developmen-
tal or systematic presuppositions relying on external considerations
advanced by scholars. But even these passages should not be isolated
from the rest of the dialogues, since they only articulate general and
abstract schemes of dialectical methods. And such general articula-
tions are necessarily less rich or concrete than what the dialogues
show through the ongoing practice of these methods. How dialec-
tic is related not only to transcendent forms, but also to perceptible
things and souls only becomes clear when this dialectical practice is
taken into account. This is the reason why I did not limit myself to
well-­known key passages on dialectic but also tried to bring in and
188 Walter Mesch
explain illuminating examples we find elsewhere in the dialogues.
As we have seen from these examples there is no need to assume a
tension between Plato’s explicit articulations of dialectic and his dia-
lectical practice. It simply has to be acknowledged that the concrete
understanding of dialectic the dialogues show throughout in their
ongoing practice is richer than the theoretical articulations of dialec-
tic. This is particularly clear with regard to the context-­sensitive vari-
ations of the fundamental dialectical capacities. As I see it, however,
very much the same could be shown for the dialogical realisation
of dialectical methods not explicitly discussed in the key passages
on dialectic. I cannot enter into this issue here, but I think it is also
unnecessary to assume a conflict between dialectical methods and
their dialogical settings identified by some scholars (Roochnik 2003,
133–151). What we find is rather that the key passages on dialectic
concentrate on the theoretical foundations of dialectic and not on its
dialogical practice shown throughout Plato’s texts.
(8) The practice of dialectic can be considered important without relying
on problematic assumptions of an external systematic or develop-
mental origin. Before I finish, I would like to explain this in order to
point out where I mainly differ from Wolfgang Wieland and Charles
Kahn when it comes to the dialectical practice and its ongoing articu-
lation in Plato’s dialogues. Concerning systematic issues, it is neither
necessary to assume that the practical key competence of a dialec-
tician, being essentially non-­ propositional, cannot be articulated
theoretically in propositional statements, nor that dialogues, differ-
ing from treatises, are completely immune to the critique of writing
in the Phaedrus, as Wieland maintains (1982, 53 and 234–6). The
advantages of dialogues can be defended without such a preference
for practical competence alien to the Platonic unity of practice and
theory. Concerning developmental issues, it is not necessary to tie the
philosophical articulation of Plato’s dialectic to a temporal succession
in the way Kahn’s proleptic reading does. Kahn criticises historical
assumptions concerning Plato’s relation to Socrates and psychologi-
cal assumptions concerning Plato’s development, but still presupposes
a problem-­answer-­relation apparent to a so-­called “original reader”
(Griswold 1990, 248–50). This is problematic since we know neither
exactly when the dialogues were written and appeared nor what an
original reader could have grasped. In the book summarising his pre-
vious articles, Kahn explicitly grants the first point, but still wants to
suggest an ideal reading order for his proleptic or ingressive interpre-
tation of the early dialogues referring to their relative distance from
the Republic (Kahn 1996, 48). I am not sure whether there is anything
like this ideal reading order, but I also think that many problems of
the early dialogues are solved in the middle ones. And it undoubt-
edly is better to interpret Plato’s articulation or gradual disclosure of
 Between Variety and Unity 189
his dialectic without strong developmental or temporal assumptions
(Kahn 1996, 63–5). As I have tried to show, this also holds for the late
dialogues. What I want to recommend, in any case, is not to build a
hermeneutical or dramatical interpretation upon developmental pre-
suppositions concerning the writing and reading of Plato’s dialogues
but to combine it with systematic or argumentative accounts of the
represented methods and objects themselves.

Notes
1 I would like to thank Vivil Valvik Haraldsen, Kristian Larsen and Justin
Vlasits for their helpful criticisms and suggestions.
2 Classical examples of developmentalism are Stenzel 1931 and Robinson
19532. While Julius Stenzel primarily stresses the emergence of a new ver-
sion of division in the Sophist, hereby distinguishing ethical and theoreti-
cal dialogues (1–2, 45–71), Richard Robinson concentrates on the difference
between the elenchus of the early and the hypothesis of the middle dia-
logues and claims a “substantial alteration” for the meaning of “dialectic” in
Plato’s early, middle and late dialogues (70). Other influential developmental
interpretations are proposed by W. K. C. Guthrie (1975) and Terence Irwin
(1995). Gregory Vlastos (1991) presents a particularly extreme version to be
discussed later. Among works criticising developmentalism I found Wieland
1982, Kahn 1981, 1996, and Howland 1991 most helpful.
3 Even if a hypothesis sufficient in a given context (ti hikanon: Phd. 101e) may
not be quite the same as a completely unhypothetical principle (archē anhy­
pothetos: R. 511b).
4 This view is developed in detail in Klaus Döring’s thorough, but sobering
study, according to which the modern Socrates research has not yielded any
universally or widely accepted results (1998, 141). For the many changes in
the views of Socrates worked out and defended since the 18th century, com-
pare also Hayden Ausland (2006).
5 I owe this point to Kristian Larsen.
6 Against the background of the Timaeus it may seem more adequate to say
that souls belonging to a visible cosmos are proper objects of an eikos logos
not to be identified with any dialectical inquiry. For Plato’s cosmology, this
clearly is an important restriction to be taken seriously (cf. Mesch 20162,
147–67). But when it comes to Plato’s psychology I do not find the conse-
quences of this cosmological perspective completely decisive. To my mind, it
is hard to deny that in the Republic and the Phaedo Socrates not only inves-
tigates forms dialectically, but also souls, and that he does this on the basis
of an essential similarity between souls and forms. Is there an inconsistency
between this dialectical treatment of souls and the cosmological treatment
we find in the Timaeus? I do not think this is necessarily so, since there is a
cosmological contribution to the ontological problem of methexis relying on
a mediating function of the soul. I will briefly come back to this later.
7 For more on the significance of Socrates’ youth in the Parmenides, see the
contribution by Francisco Gonzalez to this anthology.
8 In the famous controversy between Harold Cherniss (1957) and G. E. L.
Owen (1953) this was the main, if not the only reason for Owen to reject
the standard view that the Timaeus has to be considered a late dialogue.
Since he presupposed that the critique of forms we find in the Parmenides
190 Walter Mesch
was criticising Plato’s own theory of the middle dialogues and fatal for it, he
thought the Timaeus had to be placed before the Parmenides.
9 I thank Justin Vlasits for pointing out the importance of this passage to me.
10 As I see it, this is the ultimate basis for Socrates’ dialectical investigations of
the nature of human souls we find in the Republic and the Phaedo. And it
is also the ultimate basis for the role perceptibles play in the inductive argu-
ments of the early dialogues.
11 For the methods of dialectic this clearly means that divisions here play the
central part. Nevertheless, the method of the Visitor, who is introduced as
a god of refutation (theos elenktikos: 216b), also has a strong refutational
dimension apparent mainly in the section on being (Brown 1998, 182). And,
I think, the parallel definitions of the sophist, leaving it open at first which
one is the most reliable, also show an important hypothetical dimension. On
the basis of a rather formal similarity with the angler, the Visitor first starts
with the assumption that the sophist has a certain art which in the follow-
ing six definitions is spelled out in different ways, explaining the sophist as a
sort of hunter of young men, an importer and exporter dealing with lessons
for the soul, a retailer of the same things, a salesman selling his own lessons,
an eristic specialising in verbal battles or an expert at purifying the soul of
false beliefs. The sixth definition obviously has a special status since it rather
refers to Socrates than to sophists. But the other five are also doubtful, not
only because they differ, but also since they all presuppose that the sophist is
an artist, which someone having read the early dialogues can hardly find con-
vincing (Cf. e.g. Grg. 462b–c where rhetoric clearly is no art, but only a sort of
experience, and the same has to hold for sophistry: 465c). All these definitions
have the general form “If the sophist is an artist/has an art …” and specify it
according to different possible versions of it. This is, however, a problematic
hypothesis. In the Sophist the underlying problems become explicit at the
latest after it has been clarified that sophists claim to know everything which
in fact is not possible at all (233a). And this leads to the seventh definition,
focusing on the production of deceptive images. At first, the Visitor still seems
to maintain that we here find a deceptive art (techne: 236c). But after the
long ontological discussion when he comes back to the sophist he stresses
once again and even more that sophists do not have knowledge (267b–268d).
Thus, in the end, I think, the hypothesis that the sophist is an artist producing
deceptive images has to be questioned. In fact, he does not really have an art
but only a certain experience, routine or dexterity in producing such images.
If this is convincing, it must be important for the outcome of the dialogue to
reflect on the hypothetical status of the whole argument. So, here once again
all main dialectical methods are combined. We find not only the dominating
divisions but also hypothetical and refutational aspects.

References
Ausland, Hayden W. 2006. “Socrates’ Definitional Inquiries and the History of
Philosophy.” In A Companion to Socrates, edited by Sara Ahbel-­Rappe and
Rachana Kamtekar, 493–510. Oxford: Blackwell.
Brisson, Luc. 1996. “Den Kosmos betrachten, um richtig zu leben: Timaios.”
In Platon. Seine Dialoge in der Sicht neuer Forschungen, edited by Burkhard
Mojsisch, 229–248. Darmstadt: Wissenschaftliche Buchgesellschaft.
Brown, Lesley. 1998. “Innovation and Continuity. The Battle of Gods and Giants,
Sophist 245–249.” In Method in Ancient Philosophy, edited by Jyl Gentzler,
181–208. Oxford: Clarendon Press.
 Between Variety and Unity 191
Cherniss, Harold F. 1957. “The Relation of the Timaeus to Plato´s Later
Dialogues.” American Journal of Philology 78: 225–66.
Dancy, Russell M. 2004. Plato’s Introduction of Forms. Cambridge: Cambridge
University Press.
De Vogel, Cornelia 1953. “Platon a-­t-­il ou n’a-­t-­il pas Introduit Le Mouvement
dans Son Monde Intelligible?” Actes du 13ième Congrès International de
Philosophie 12: 61–7.
Dixsaut, Monique. 2001. Métamorphoses de la Dialectique dans les Dialogues
de Platon. Paris: Vrin.
Döring, Klaus. 1998. “Sokrates, die Sokratiker und die von ihnen begründeten
Traditionen: § 13. Sokrates.” In Grundriss der Geschichte der Philosophie
(Ueberweg). Die Philosophie der Antike 2/1, edited by Hellmuth Flashar, 141–
78. Basel: Schwabe Verlag.
Ebert, Theodor. 1998. “Wer sind die Ideenfreunde in Platons Sophistes?“ In
Amicus Plato (Festschrift Wolfgang Wieland), edited by Rainer Enskat, 82–
100. Berlin: De Gruyter.
Frede, Dorothea. 1999. Platons “Phaidon”. Der Traum von der Unsterblichkeit
der Seele. Darmstadt: Wissenschaftliche Buchgesellschaft.
———. 2004. “Dialektik in Platons Spätdialogen.” In Platon verstehen. Themen
und Perspektiven, edited by Marcel van Ackeren, 147–67. Darmstadt:
Wissenschaftliche Buchgesellschaft.
Fronterotta, Francesco. 2011. “Some Remarks on the Senses of Being in
the Sophist.” In Plato’s Sophist, Proceedings of the Seventh Symposium
Platonicum Pragenese, edited by Aleš Havliček and Filip Karfik, 35–62. Prague:
OIKOYMENH Publishers.
Gonzalez, Francisco J. 2011. “Being as Power in Plato’s Sophist and Beyond.” In
Plato’s Sophist, Proceedings of the Seventh Symposium Platonicum Pragenese,
edited by Aleš Havliček and Filip Karfik, 63–95. Prague: OIKOYMENH
Publishers.
Griswold, Charles L. 1990. “Unifying Plato: Charles Kahn on Platonic Prolepsis.”
Ancient Philosophy 10: 243–62.
Guthrie, W. K. C. 1975. A History of Greek Philosophy, Volume IV: Plato, The
Man and His Dialogues. Cambridge: Cambridge University Press.
Hoffmann, Michael. 1996. Die Entstehung von Ordnung. Zur Bestimmung von
Sein, Erkennen und Handlung in der späteren Philosophie Platons. Berlin/
Boston: De Gruyter.
Horn, Christoph. 1997. “Platons epistēmē-­doxa-­Unterscheidung und die
Ideentheorie.” In Platon. Politeia, edited by Otfried Höffe, 291–312. Berlin:
Akademie Verlag.
Howland, Jacob. 1991. “Re-­Reading Plato: The Problem of Platonic Chronology.”
Phoenix 45 (3): 189–214.
Irwin, Terence. 1995. Plato’s Ethics. Oxford: Oxford University Press.
Kahn, Charles H. 1981. “Did Plato Write Socratic Dialogues?” The Classical
Quarterly 31 (2): 305–20.
———. 1996: Plato and the Socratic Dialogue. The Philosophical Use of a
Literary Form. Cambridge: Cambridge University Press.
Kraut, Richard. 2006. “The Examined Life.” In A Companion to Socrates, edited
by Sara Ahbel-­ Rappe and Rachana Kamtekar, 228–42. Malden: Blackwell
Publishing.
192 Walter Mesch
Larsen, Kristian. 2019. “Measuring Humans against Gods: On the Digression of
Plato’s Theaetetus.” Archiv für Geschichte der Philosophie 101 (1): 1–29.
Mesch, Walter. 2007. “Analogien und Antistrophen. Zur Bestimmung der
Rhetorik in Platons Gorgias.” In Gorgias—Menon. Proceedings of the VII
Symposium Platonicum. Selected Papers, edited by Michael Erler and Luc
Brisson, 149–157. Sankt Augustin: Academia Verlag.
———. 2011. “Die Bewegung des Seienden in Platons Sophistes.” In Plato’s
Sophist, Proceedings of the Seventh Symposium Platonicum Pragenese, edited
by Aleš Havliček and Filip Karfik, 96–119. Prague: OIKOYMENH Publishers.
———. 20162. Reflektierte Gegenwart. Eine Studie über Zeit und Ewigkeit bei
Platon, Aristoteles, Plotin und Augustinus. Frankfurt am Main: Klostermann.
Mittelstrass, Jürgen. 1997. “Die Dialektik und ihre wissenschaftlichen
Vorübungen.” In Platon. Politeia, edited by Otfried Höffe, 229–229. Berlin:
Akademie Verlag.
Owen, G. E. L. 1953. “The Place of the Timaeus in Plato’s Dialogues.” Classical
Quarterly NS 3: 79–95.
Prior, William James. 1985. Unity and Development in Plato’s Metaphysics.
London and Sydney: Croom Helm.
Robinson, Richard. 19532. Plato’s Earlier Dialectic. Oxford: Clarendon Press.
Roochnik, David. 2003. Beautiful City. The Dialectical Character of Plato’s
Republic. Ithaca and London: Cornell University Press.
Rowe, Christopher. 2006. “Socrates in Plato’s Dialogues.” In A Companion
to Socrates, edited by Sara Ahbel-­ Rappe and Rachana Kamtekar, 159–70.
Malden: Blackwell Publishing.
Ryle, Gilbert. 1939. “Plato’s Parmenides.” Mind 48: 129–51, 302–25.
Schmidt-­Wiborg, Petra. 2005. Dialektik in Platons Philebos. Tübingen: Mohr
Siebeck.
Stemmer, Peter. 1992. Platons Dialektik. Die frühen und mittleren Dialoge.
Berlin/New York: De Gruyter.
Stenzel, Julius. 1931. Studien zur Entwicklung der platonischen Dialektik von
Sokrates zu Aristoteles. Leipzig und Berlin: Teubner.
Strobel, Benedikt. 2004. “Attribute der Formen und die Form des Guten.
Gerasimos Santas über die metaphysische Theorie des Guten in Platons
Politeia.” In Platon verstehen. Themen und Perspektiven, edited by Marcel van
Ackeren, 111–34. Darmstadt: Wissenschaftliche Buchgesellschaft.
Szlezák, Thomas Alexander. 2011. “Zur Bedeutung der Eingangsszene des
Sophistes (216a–218a).” In Plato’s Sophist, Proceedings of the Seventh
Symposium Platonicum Pragenese, edited by Aleš Havliček and Filip Karfik,
11–34. Prague: OIKOYMENH Publishers.
Taylor, Alfred Edward. 1961. Plato: The Sophist and the Statesman, edited by
Raymond Klibansky and Elisabeth Anscombe. London: Thomas Nelson.
Vlastos, Gregory. 1991. Socrates. Ironist and Moral Philosopher. Cambridge:
Cambridge University Press.
Wieland, Wolfgang. 1982. Platon und die Formen des Wissens. Göttingen:
Vandenhoeck & Ruprecht.
Williams, Thomas. 2002. “Two Aspects of Platonic Recollection.” Apeiron 35
(2): 131–152.