(Beiträge zur Altertumskunde 341) Gabriele Cornelli - Plato’s Styles and Characters Between Literature and Philosophy-Walter de Gruyter (2016)
particular cases are explored in the interest of general ideas. But Plato’s dia-
logues are not dramas engagés just as they are not plain philosophical treatises.
Although his dialogues deal with ideas directly, these are explored not ‘objective-
ly’, but from the points of view of the characters involved in the conversation at
each of its stages.
There were antecedents to Plato’s dialogues, by now well known. There was,
after the death of Socrates, a spate of logoi sokratikoi;⁴ Plato was apparently not
Galileo 1632; Berkeley 1713.
But Plato’s chronology is notoriously inconsistent. See below, p. 7.
With few exceptions, such as the boy in Meno, the Eleatic Stranger of the Sophist or the Athe-
nian of the Laws.
See Giannantoni 1983.
the first to write such logoi. But this is immaterial to our present interest. What-
ever his historical position in this respect, he certainly brought the genre to its
perfection – with an important difference to boot, which will occupy us as we
go along. The other logoi sokratikoi exhibited Socrates’ modus philosophandi
and, at most, conveyed a moral message detachable from the dialogical form,
or were of merely protreptic intent. As a matter of fact, if anything can be said
about the historical Socrates, his intent seems indeed to have been mostly pro-
However, for our purposes, we are interested not in the historical Socrates,
whoever he may have been, but in the Socrates of Plato’s dialogues. Yet, Socrates
in the Republic says things that in the Apology he would not say, or flatly contra-
dicts them: immortality of the soul (whatever that means), ideas, etc. This is not
a chronological distinction, not the venerable one corresponding to the ‘early’
and ‘later’ dialogues, respectively. It is rather the distinction between the aporet-
ic Socrates, such as that in Laches or Euthyphro, and the ‘positive’ Socrates, such
as the one in Phaedo or the Republic. Socrates of Plato’s dialogues is a complex
figure. Elsewhere, I have called these two strains of Socrates of the dialogues
‘Plato’s Socrates’ and ‘the platonic Socrates’: the agnostic, aporetic, and the py-
thagorizing, mythologizing, metaphysical Socrates. In the same dialogue both
may be there at different times. Euthydemus, Phaedrus, even Phaedo and the Re-
public have both. Traits of the first appear also in ‘later’ dialogues, such as The-
aetetus, and of the second also in ‘early’ dialogues, as Protagoras. Here, I am in-
terested mainly in the platonic Socrates, although I shall be referring
occasionally to Plato’s Socrates. The difference should be clear from the context.
Neither of these is necessarily Plato’s mouthpiece. Socrates often says different
things in different dialogues and even sometimes, if we take his words at face
value, contradicts himself from one dialogue to another and even in the same
One might say that, as in good drama, Plato is behind all of his characters
and, what amounts to the same, in none of them. Or perhaps he is in his char-
acters not only as their creator but also as voicing through them views that he
himself held or, at least, considered. Shakespeare did not put part of himself
into Iago, but Callicles may have a good claim to a part of Plato, or so the argu-
A playwright does not commit himself to consistency over different pieces. It
does happen sometimes, particularly when the author has a social or political
agenda, but this is certainly not the rule. The same has been sometimes said
of Plato: each of his dialogue is to be read on its own, without taking over
ideas from one to another. But Plato himself quotes from one dialogue in anoth-
er, even if sometimes he does it not quite accurately or straightforwardly mis-
Plato’s dialogue form is not just a beautiful outer shell but part and parcel of
the philosophical matter. As it has often been remarked but seldom paid atten-
tion to sufficiently, Plato’s philosophy is dialogical. In most cases, this has been
taken to mean that Plato uses dialogue as a pedagogical or philosophical meth-
od, the results of the enquiry being either merely protreptic or else independent
of the didactical process. Romantics favoured the search itself over its possible
results, while others, more dogmatically or analytically oriented, strove to sur-
face from the dialogues with explicit ‘doctrines’.⁵ To my mind, both these read-
ings of the dialogues fall short of giving a satisfactory explanation of the full sig-
nificance of this peculiarly platonic literary and philosophical form.
Plato’s dialogues do not seem, at first look, to be all merely protreptic and
exploratory, at least not such dialogues as the Symposium and the Republic –
and even these were not always above suspicion; even the ‘later’ dialogues
have sometimes been said to be merely tentative. Perhaps, then, there are, never-
theless, philosophical doctrines in Plato? Of course, in Phaedo, Symposium, Re-
public and other dialogues, there are the ideas, the immortality of the soul (what-
ever that means), the distinction between doxa and episteme. Are not these firm
doctrines to be learned from those dialogues? But, of these supposed doctrines,
the first two are presented as provisional hypotheses,⁶ and the third seems to be
gratuitously assumed and it is by no means evident; Protagoras, for once, negat-
ed it squarely. Parmenides and Euthydemus, to mention only two dialogues, put
these doctrines in serious doubt. And do, e. g., the Sophist and the Laws go back
on them? The chronology of Timaeus was hotly disputed because it explicitly
If we cannot make much headway with the content of the dialogues, let us
try another approach: let us investigate their form. On the literary level, Plato’s
dialogues may be classified as either diegetic, i.e. narrated, or mimetic, i.e. di-
rectly presented. Some, like Theaetetus or Parmenides, are mixed and some
are in the form of a narration framing a direct dialogue. In the Republic, Plato
prefers narration to mimesis.⁸ Narration interposes the narrator between the con-
tent narrated and his audience, thus allowing the author to distance himself
from the prima facie veracity of the content.
E.g., Crombie 1962. For a variety of approaches, see now Press 1993.
See Phaedo 107b.
Cf. Owen 1953 and Cherniss 1957.
Republic iii 394d ff.
Beyond Language and Literature
In both cases, Plato’s dialogues are incomplete. Aporetic dialogues are such
by their own nature. But, narrated or not, his dialogues contain purposeful in-
consistencies in chronology or other disturbing details precisely in order to
warn us against taking at face value all that is said. And one should never forget
that the narrator is himself a dramatis persona. As such, he has his own interests
and point of view and his report is never to be taken without more ado as nar-
rating things as they presumably were.
But, one may say, when Socrates is the narrator, he is surely to be trusted. –
Or is he? Take the case of Euthydemus. Twice in the course of the dialogue, Soc-
rates is on the verge of admitting that his story is not well told. At the third time,
Crito himself challenges Socrates’ narrative to the point that Clinias – Socrates’
young respondent, credited by him of quite unlikely answers – immediately
drops out of the dialogue altogether.⁹
Euthydemus teaches us something important about Socrates’ method. While
the two sophists, Euthydemus and Dionysodorus, put their questions from their
own point of view, without any preparation of their interlocutor,¹⁰ Socrates ar-
gues dialectically. In this dialogue as in all the other dialogues, he starts always
from the point where his interlocutor stands and aims at disproving him and
sometimes, if possible, at leading him gradually to change his views. In so argu-
ing, Socrates is necessarily being ironical. He takes on positions that are not his,
but his method brings him to develop them only for the sake of the argument.
One should, thus, be very careful when ascribing to Socrates views that he pres-
ents during the conversation.
In Meno 80, Plato introduces his hypothetical method, to be fully explained
in Phaedo 100 –101.¹¹ This method, inspired by the geometrical method of anal-
ysis, starts from an accepted conclusion and looks for the minor premise that
will support it. For Plato, philosophy is not primarily a deductive science.
Against the rules of good logical reasoning, accepted at least since Parmenides,
the conclusion is taken to be stronger than its premises. The hypothetical method
goes ‘the wrong way’: ¹² it goes from the conclusion to the premise, which is now
The source of that premise is irrelevant. It could be a few lines from a poem,
an accepted scientific theory, a myth (invented ad hoc or pre-existing), a purport-
ed dream. If that premise is not consistent with other opinions accepted by the
interlocutor, either it or the opinions that conflict with it are to be discarded as
Cf. Scolnicov 1975.
Cf. Cornford 1932.
false;¹³ if no such contradiction is found, that premise is considered provisional-
ly true until proved otherwise and is now itself in need of support ‘in the same
One consequence of the hypothetical method is directly relevant to our pres-
ent concerns: it can be successfully applied only in dialogues in which the inter-
locutor shares with Socrates the primary intuition expressed by the desired con-
clusion. Meno, as a matter of fact, is not convinced that learning is reminiscing
and asks to hear from Socrates what he has to say on the matter, rather than find
the truth for himself; the dialogue with him fails. Simmias and Cebes are well
predisposed towards philosophia and the dialogue can proceed to a positive out-
come, even if provisional and in need of further support.¹⁵
Plato’s fundamental intuition is the primacy of ethics, i.e. of philosophia as
he understood it, continuing Socrates’ rejection of the unexamined life. All his
effort is directed at supplying philosophia its foundation. Even in Parmenides,
the question of philosophia is there in a crucial passage.¹⁶
This means that the platonic Socrates is preaching to the converted or to
youngsters as yet unformed and uncrystallized. With others he is a dismal edu-
cational failure. But this is a necessary consequence of Plato’s method of hypoth-
esis. If the method takes its start from accepted conclusions, going ‘upwards’
from them to their premises, it is inevitable that for those who do not share,
not even for the sake of argument, those conclusions, the hypothetical move-
ment has no starting point.
On the other hand, socratic-platonic dialectic is essentially ironic, especially
in the elenctic dialogues. The dialectical procedure starts from where the inter-
locutor stands. Socrates accepts the interlocutor’s view pro tempore and only ar-
gumenti gratia. Otherwise, dialectic cannot proceed. But, then, this means that
Socrates’ initial positions cannot ever be trusted as his own.
How much of Plato’s irony is Socrates’? Again, I try strenuously not to say
anything about the ‘historical’ Socrates. By ‘Socrates’ I mean Socrates of Plato’s
dialogues. Yet, if Socrates (even the ‘historical’ one, for once) asked questions
and refused to answer – this is all the irony we need for a start. Xenophon’s Soc-
rates gives good advice and is very useful to have around. That Plato is ironical
Phaedo 92c, 100a, 101d.
Cf. Phaedo 107a.
Beyond Language and Literature
goes without saying. Is Socrates’ irony all of it Plato’s? Not impossible, but I per-
sonally find it hard to believe.¹⁷
Yet, socratic irony never discloses what Socrates means. It only makes ex-
ceedingly clear what he does not mean. In this, it is not like ordinary irony, or
antiphrasis, which says the opposite of what is meant.¹⁸ Nor is it like romantic
irony, or Vlastos’s complex irony, which leaves us undecided between ‘yes’
and ‘no’.¹⁹ It is open irony, in which the opposite pole is never given.²⁰
Socratic irony cannot tell us, not even indirectly, Socrates’ meaning. Socratic
irony involves a different understanding of the very words used. Using other
words would be of no avail. Language is essentially ambiguous, as Euthydemus
and Cratylus teach us. Time and again, Socrates’ argumentation brings us back to
what seems to be the accepted virtues.²¹ The words are the same, but, for Socra-
tes, they bear a different meaning, for which there are no other words than the
ones he and everybody else use.
Here an important distinction is in order: the distinction between utterance
and proposition. An utterance is a unit of speech, long or short, the actual token
of words emitted by the speaker at a given moment, essentially dependent on
him as the speaker. A proposition is the content expressed in the utterance, in-
dependent of him who produced it or of the language in which it was produced.
(For our purposes, the modern distinction between sentence and proposition is
irrelevant.) The dialogue format presents us with purported utterances, not with
propositions. All that is said is dependent of he who says it and in which context
and is inextricable from the sayer and from its context.
In Euthydemus, Plato makes a firm stand on the view that words have no
meaning in themselves. He is careful not to say that such is the meaning of
this word; instead he says that people use this word to convey such a meaning.
Words are tools²² used by the speaker to convey what he wants. Any word can
bear any meaning. I may, of course, use ‘dog’ to mean cat, if I disregard the
risk of being misunderstood. The meaning of the words is not in themselves
but in the soul of the speaker (and of the hearer).²³
Holger Thesleff, in private correspondence, pushed me towards taking a stand on this ques-
On Socrates’ open irony, see Scolnicov 2004.
Cf. Desjardins 1988.
Cratylus 388c ff.
Euthydemus 295b5, e8.
Thus, dialogue is absolutely necessary. Without it, we can never know
whether our words were properly understood.²⁴ And, a fortiori, the written
word is little less than useless – again, if the reader is not among the converted.
But, if he is not, how can he ever be won over? First, by trying to disabuse him of
the meaning he attaches to his words; then, by guiding him, very slowly, to grasp
the meaning Socrates attaches to his words, with no guarantee of success. After
much time spent together, a spark leaps from soul to soul, if it ever does, and
kindles a fire that will now nourish itself.²⁵
This is not a matter of unwritten doctrines²⁶ that could be put in words and
one must be careful not to disclose them to the uninitiated. The philosophical
core of Plato’s thought cannot be put in words, since words can never be trusted.
This is true not only of the written word. What cannot be written cannot be said
either. Words, written or spoken, can teach us only what we already know, or lit-
tle else.²⁷ A spark leaps, not through words.
One could say, then, as it has been said more than once, that all of Plato’s
dialogues are merely protreptic. Indeed, it is not the death of Socrates one
should mourn, but the death of the logos, if it comes to pass. This is Socrates’
last will and testament, in the all-important central excursus of Phaedo: ‘Keep
the logos alive!’²⁸ But there are in the dialogues no ‘gaps’ to be filled in oral dis-
cussion. This is not to say that no discussion or clarification followed their first
‘publication’, whatever that was.²⁹ His dialogues were not, most of them, written
to be read by all. But, in them, Plato said exactly what he wanted to say. The writ-
ten (or spoken) platonic dialogue goes as far as Plato believed language can go.
But the rest is not silence. The rest is a Gestalt switch. True, an inner circle of
initiated would be more likely to undergo that switch than the ordinary man-
in-the-street. And even that is doubtful, as the end of Euthydemus shows. But
if they could understand him, this was not because they had more information,
rather because they shared, at least partially, Socrates’/Plato’s view on nous and
This logos to be continued, in Plato’s interpretation, is not an open discus-
sion. It has a very definite direction. To keep the logos alive is to ‘save’ it, to
give it its metaphysical foundation, so that it can withstand the attacks of Demo-
Seventh Letter 341c.
As held, in modern times, among others by Krämer 1959 and Szlezák 1993.
Phaedo 89b. The place of the passage in the almost exact middle of the dialogue is a prime
example of ‘pedimental structure’, as it called by Thesleff 1966, in Thesleff 2009, 28 and passim.
Thesleff 2000, 241–550.
Beyond Language and Literature
critus’ mechanicalism and Protagoras’ relativistic ethics. If a real distinction be-
tween doxa and episteme cannot be supported, Protagoras won the day. Plato (as
well as his own Socrates of the ‘early’ dialogues) had a positive view about the
normativity of reason, but, due to the constraints of language, he could not have
possibly put it in words other than those in common use, with a different mean-
ing, understandable only to those who shared his views. No convincing the un-
convinced. No esotericism either. Just a profound mistrust of language.
The fulcrum of Plato’s thought is beyond language. Its meaning cannot con-
vincingly be put in words. Nor can it be proved, for a proof would have to be
based on some more primitive assumption, which would, again, be open to
the same difficulty. That archimedean point has to be directly intuited. But it
is not a matter of mysticism or extra-sensory perception. There is nothing myste-
rious about it. The arkhe anupothetos, or the idea of the Good, or the Beautiful of
the Symposium are not irrational; on the contrary, they are emphatically brought
up as the pinnacle of rationality. But the object of that primary intuition is not,
as it has sometimes been assumed, the unhypothetical beginning. If it were, phi-
losophy would be a deductive affair, as it was for Parmenides, with all its ensu-
ing aporiai. That arkhe comes, paradoxically, at the end of the dialectical process
and once achieved, all hypothetical steps leading to it are confirmed – but not
until then. Until the principle is achieved – if it ever is – all these steps only
have the epistemic status of opinions (doxai).
Therefore, if the distinction between doxa and episteme is to be kept, the
idea of the Good, even if it is the result of an hypothetical procedure, cannot
be a mere idea in the kantian sense. It cannot be only a postulate of reason,
or else we are thrown back onto a transcendental³⁰ variation of Protagoras’ ho-
momensura, only this time not relative to the individual man but to a transcen-
dental subject, as Kant has it – a price Plato could not be prepared to pay.
That mainspring of Plato’s philosophy cannot be dependent on language nor
can it be an intuited but unsupported premise of a series of deductions. The
starting point of the philosophical enterprise, faithful to Plato’s hypothetical
method, must be an intuition of the conclusion, not of the first principle. It
must be outside the dialogues, since dialogues are inevitably tied up with lan-
guage and can go no further than the utterances pronounced by the participants
in a particular occasion. That starting point, which demands the conviction of
the participants in the philosophical enterprise, cannot be summarized in a
proposition, discursively stated. That necessary basic conviction is elicited by
an event outside the dialogues.
In a kantian sense, as distinct from transcendent.
The formative event in Plato’s thought is Socrates’ death. Those present at
Socrates’ death understood what he meant by saying that ‘the unexamined
life is not worth living for man’. Socrates made these words meaningful on his
own person. For those who were not there, Plato (like other Socratic authors)
tried to convey the significance of the moment and the emotion attached to it,
as best he could, in the last scene of Phaedo.
All of Plato’s dialogues go back to that event. Socrates is present in all of
them, if not personally, at least by implication. Socrates’ death shows us what
meaning he gave to his words, as Plato saw it. A life in which reason is the serv-
ant of the passions or of mere empirical utility is not worth for man to live it. Not
the death of Socrates should be mourned, but the death of the logos, if he dies
and we cannot save it.
The unexamined life is not worth living for man. Was this just Socrates’ pri-
vate opinion? Those present felt its truth. But that opinion needs a metaphysical
basis that can vouch for it, that can support it and turn it into real episteme. Pla-
to’s metaphysics from Phaedo onwards is the salvation of the logos; it is Socrates
put on what was, for Plato, a secure basis.
For Socrates, as Plato understood him and portrayed him, reason was nor-
mative, in and of itself. The rational examination of one’s life needs no justifica-
tion. Reason is its own justification, and a human life without reason as the su-
preme criterion of goodness is not worth living. Not for Plato; not the Socrates of
Phaedo or the Symposium. But that is already matter for another paper.
Allen, R E (ed.) 1965, Studies in Plato’s Metaphysics, Routledege & Kegan Paul, London.
Berkeley, George 1713, Three Dialogues between Hylas and Philonous.
Cherniss, H F 1957, ‘The relation of the Timaeus to Plato’s later dialogues’, in Allen 1965.
Cornford, F M 1932, ‘Mathematics and dialectic in the Republic vi-vii’, in 1965.
Crombie, Ian M 1962, An Examination of Plato’s Doctrines, Routledge & K. Paul, London.
Desjardins, Rosemarie 1988, ‘Why dialogues?’, in Griswold Jr. 1988, 110–15.
Galileo Galilei 1632, Dialogo dei due massimi sistemi del mondo.
Giannantoni, Gabriele 1983, Socraticorum reliquiae, Bibliopolis, Roma.
Griswold Jr., C.L. (ed.) 1988, Platonic writings / Platonic readings Routledge, London.
Hoffmann, Ernst 1947, ‘Die literarische Voraussetzungen des Platonsverständnisses’,
Zeitschrift für philosophische Forschung 2, 465–480
Krämer, Martin 1959, Arete bei Plato und Aristoteles, Heidelberg.
Owen, G E L 1953, ‘The place of the Timaeus in Plato’s dialogues’, in Allen 1965.
Press, Gerald A 1993, Plato’s dialogues: New studies and interpretations, Rowman &
Littlefield, Lanham, Md.
Quintilianus, Marcus Fabius, Institutio oratoria.
Beyond Language and Literature
Scolnicov, Samuel 1975, ‘Hypothetical method and rationality in Plato’, Kant-Studien 66,
Scolnicov, Samuel 2004, ‘Plato’s ethics of irony’, in Maurizio Migliori (ed.), Plato ethicus
Academia Verlag, St. Augustin, 289–300.
Szlezák, Thomas A 1993, Platon lesen, frommann-holzboog, Stuttgart.
Thesleff, Holger 2000, ‘Plato and his public’ (2000), Studies in the styles of Plato, 241–250.
Thesleff, Holger 2009, Platonic patterns: A collection of studies, Parmenides Publishing,
Thesleff, Holger 1966, Studies in the styles of Plato (1966), now reprinted in Thesleff 2009,
28 and passim.
Vlastos, Gregory 1991, ‘Socratic irony’, in Socrates: Ironist and moral philosopher, Cambridge
University Press Cambridge, 21–44.
The Three Waves of Dialectic in the
The alleged lack of unity of the Republic has become increasingly questioned.
There are several strategies to show that this unity exists. One of the most com-
mon ones is to recognize in Republic I the anticipation of positions that will not
be developed thoroughly until further on.² Thus H.-J. Krämer (1959, p. 53ff.) has
recognized in Socrates’ three final arguments in Republic I – the arguments
about pleonexia (348b8–350c11); about cooperation (350c12–352d2); and
about the relationship between natural or artificial beings, their arete and func-
tion (352d2–354a7) – an anticipation of the fundamental ontological positions
developed in the following books. However, if we change perspective, one
might claim that those arguments are nothing more than a draft and, thus, an
image of different aspects of the notion of justice.³ Accordingly, the correspond-
ence between natural or artificial beings, their arete and their function, can be
understood as an image of the correspondence between a physis and an
ergon,⁴ since this correspondence is, in turn, explicitly qualified by Socrates
not only as an eidōlon of psychic justice (443c4–5), but also as an archē and
a typos of his project of forming citizens for the just city (443c1), which, as
such, leaves its stamp on the soul of the receiver, that is, on the soul of someone
I would like to thank Alexandra Alván for translating this paper and Peter Simpson for
carefully reading the translation. My aim here is to explain in more detail what I presented in
Gutiérrez, R. (2009). So I set to one side questions about the koinōnia of Ideas and the Par-
menides. I had already in the cited paper considered the three waves as hypotheses, each of
which relied on a superior hypothesis that explained it. Here I further develop the analysis of the
third wave following closely Benson’s (2008) analysis of the third wave. I was not aware of
Benson’s paper when I wrote my earlier paper.
Cf. Kahn, C. (1993); Wilson, J.R.S. (1995). Wilson presents Thrasymachus as an illustration of
the thymoeides. The starting point for his argument is R. 411e1, where, as at 439b4, he is said to
be ὥσπερ θηρίον, referring, according to Wilson, to the thymoeides (59). In the first passage,
however, the soul’s trichotomy has not yet been fully developed (the epithymētikon is not men-
tioned), and it is not until the 439b4 passage that this characterization is referred to the epithy-
mētikon, the determinant one. The reference to Thrasymachus’ “leonine” aspect and to his
search of εὐδoκιμεῖν and φιλονικεῖν (338a6–7), which are without doubt present in Thrasyma-
chus, favours Wilson’s interpretation, but it seems to me that both his calculating and spirited