The discussions of Platonic style, most notably in the Demosthenes and the Letter to Pompeius Geminus, by the Atticist critic Dionysius of Halicarnassus, writing in Rome in the time of Augustus, demand attention on several fronts. They offer the fullest discussion of the subject in a classical author,¹ are a central witness to a debate which we know to have stretched from shortly after Plato's death (if not indeed before) to the neo-Platonism of late antiquity, one which puts on display several very typical features of the ancient discussion of written style, and are crucial texts in the history of Greek Atticism; moreover, they also provide one of the few classical cases where both an original text and the author's response to criticism of it survive.² In this chapter I will use Dionysius’ treatment of the Phaedrus in his essay on Demosthenes as a way into a world of ancient critical debate which both has its own interest and also manifests much of what seems to have been typical of ancient stylistic discussion.

The Phaedrus was, of course, of the greatest importance to the ancient rhetorical tradition in general, but Dionysius, whose principal interest was the (at least theoretical) training of young men to become practising orators, may well have felt a particular involvement, and need to engage, with a work which openly discussed rhetorical technique and style, which seemed to demand in the rhetorician areas of knowledge, such as knowledge of the soul, with which Dionysius had nothing to do and appeared irrelevant to his concerns, which appeared to conclude that oratory, including the writing of πολιτικοὶ λόγοι, which were very close to Dionysius’ heart, and the study of oratory were little more than ‘amusement’ (παιδιά), and in which Socrates openly wonders whether he and Phaedrus have not ‘rather too ignorantly heaped abuse on the technê of logoi’ (260d3–4). Socrates brings the Phaedrus to its conclusion by observing that he and Phaedrus have now had an appropriate amount of amusement on the subject of logoi (278b7), but logoi were no laughing matter for Dionysius. In the Demosthenes Dionysius indeed demonstrates Plato's incompetence in the field of πολιτικοὶ λόγοι (chaps. 23–32 on the Menexenus) – Plato should have stuck to Socratic dialogues.

Most significantly of all perhaps, Lysias, for Dionysius the embodiment of very much that was great in the classical, particularly the Atticist, ideal,³ is a central absent presence in the Phaedrus, and this was always going to require a response; Socrates’ prayer to Eros that the god should turn Lysias away from rhetoric, as he practised it, to philosophy, as (so we are to understand) Plato practised it (257b), must have seemed a terrible misjudgement. To make matters worse, Plato concludes the dialogue with a gesture towards a synkrisis between Lysias and Isocrates (278e–9b), another crucial figure for Dionysius in both his rhetorical and cultural works, and not only was the synkrisis of orators Dionysius’ business, but Plato's apparent view of Isocrates’ general superiority to Lysias was perhaps a provocation too far; with hindsight one could see that, not only was the answer to Socrates’ opening ‘Where are you coming from?’ question to Phaedrus ‘from Lysias’,⁴ but one answer to the ‘Where are you going?’ question was ‘to Isocrates’. Although Plato is not solely responsible for the fact that much of Dionysius’ treatise on Isocrates is structured as a synkrisis between Isocrates and Lysias, he bears a fair share of responsibility; this closing passage of the Phaedrus is specifically alluded to, and without hostility, at Isocrates 12.2. Dionysius’ stress in the Isocrates on the superiority of Lysias’ ‘natural’ style over the artificial beauties of Isocrates (3.4, 12.3–4) may be seen as a response, at the level of style, to Plato's claim (Phaedrus 279a2–3) of the superiority of Isocrates’ φύσις, just as Plato's reference to Isocrates’ ἦθος γεννικώτερον (279a4) may be an ancestor of Dionysius’ acknowledgement of Isocrates’ superiority over Lysias in dealing with grand subjects (Isocrates 3.5–7, 11.4). Where, for Dionysius, Isocrates really triumphs over Plato is in the subjects he tackles, subjects of the highest public importance – his were real πολιτικοὶ λόγοι (Isocrates 4–10, 12). Plato, ironically or otherwise, had observed that Isocrates might be destined for higher things (μείζω) because ‘there was innately some philosophic instinct in the man's manner of thought’ (279a8–10).⁵ Dionysius could not agree more; for him, Isocrates is required study for anyone with any concern for ‘true philosophy’, for anyone who enjoys ‘not just speculation (τὸ θεωρητικόν), but also thought which can lead to action (τὸ πρακτικόν), who chooses a course which will benefit many people, not one which will produce a carefree life for himself’ (Isocrates 4.4); one does not need the wisdom of Socrates to see the target of these barbs.

The beginning of the Demosthenes is lost,⁶ but when our text opens Dionysius is discussing Gorgias as an example, and perhaps the originator, of a grand, ‘poetic’ style, which is ‘beyond the expected and strange and elaborated (ἐξηλλαγμένη καὶ περιττὴ καὶ ἐγκατάσκευος) and full of every kind of accessory adornment’ (1.3);⁷ the ‘standard and paradigm’ of this style is Thucydides. Opposed to this style is a ‘straightforward, simple’ style, which draws its power from its resemblance to ordinary speech (2.1); this was the style of, inter alios, ‘the philosophers of nature and the creators of ethical dialogues (οἱ τὰ φυσικὰ φιλοσοφήσαντες καὶ οἱ τῶν ἠθικῶν διαλόγων ποιηταί), such as the whole Socratic school except Plato’, and its most perfect embodiment is the style of Lysias. ‘Ethical dialogues’, such as – we must suppose – Xenophon's Memorabilia, were supposed to, and largely succeeded in, capturing the flavour of real speech, just as (according to Dionysius and other Atticists, such as Caecilius of Calacte), Lysias’ ‘purity’ was one created from ‘ordinary’ words. The contrast between the two styles is, with some difference of nuance, essentially the same structure which we find in the opening chapters of Dionysius’ treatise on Lysias, where Dionysius makes the point that those who preceded Lysias, such as, above all, Gorgias and Thucydides in the speeches, wrote in an inflated, poetic style; when in that treatise, to anticipate later discussion, Dionysius claims that Gorgias often spoke ‘in an almost dithyrambic manner’ (οὐ πόρρω διθυράμβων, Lysias 3.4), quoting Phaedrus 238d2–3, he virtually equates Gorgias and the ‘enthused’ Socrates of the Phaedrus in a manner which was typical of much of the critical tradition.

In the Demosthenes there is a third style which is ‘mixed and put together from the other two’ (3.1), which Dionysius traces back to Thrasymachusand of which the finest examples are Isocrates and Plato. In both of these latter cases, however, the style as a whole is not a true mixture, but rather the writers exemplify both the plain and the grand, but in different passages and with different levels of success. Whereas Isocrates uses the two styles in accordance with his purpose, the plain, Lysianic style ‘for instructing’ (διδάξαι) the audience’ and the grand style for ‘amazing’ it (καταπλήξασθαι) (Dem. 4.3), Dionysius offers no reason, beyond authorial choice, for Plato's stylistic variation: ‘when, on the one hand, he uses the plain, simple, unartificial style…when, on the other, he launches himself without restraint, as he often does, into elaborated and decorated language’ (Dem. 5.2, 4). If Isocrates comes out the better from this comparison, so too Dionysius’ criticisms of his Gorgianising (Dem. 4.3–4) are as nothing compared with what he will throw at Plato. It is perhaps worth noting, in passing, that some two centuries later Hermogenes was to use Socrates’ famous comparison in the Phaedrus of a logos to a living creature ‘with its own body’ and with all its parts ‘appropriate to each other and to the whole’ (Phaedrus 264c2–5) as a model for how, not just different themes and subjects, but also the different styles used in a logos should be blended together into a harmonious whole (296.15–7.21 Rabe).⁸ Thus, from the perspective of a Dionysius, Plato did not practise what he preached. Later in antiquity, at least, the ‘mixture’ of styles in Plato was theorised and taught. In the extant Prolegomena to Plato from (probably) the sixth century ad,⁹ it is noted (17.2–18) that one can have a mixture of the two opposed styles either ‘by a blending’ (κατὰ κρᾶσιν) or ‘by juxtaposition’ (κατὰ παράθεσιν). The author notes that Plato uses the two styles to suit his subject: the grand for ‘theological dialogues’,¹⁰ a ‘blended’ style, which is a moderated version of the grand or the simple, for ‘the ethical dialogues about virtue’,¹¹ whereas a ‘mixture by juxtaposition’ is what the author finds in the Gorgias, where different parts of a work belong to different styles. Dionysius’ account of the Phaedrus would, as we shall see, fall readily into the category of ‘mixture by juxtaposition’, though such scholastic categorisation would not, in his view, excuse the excesses of the Phaedrus, grand style or not.¹²

For the ‘plain, simple’ style in Plato, Dionysius has nothing but praise:

As has been frequently observed,¹⁴ this metaphorical and, from a modern perspective, impressionistic description clearly evokes the locus amoenus of the Phaedrus, and Dionysius subsequently (7.1–2) explicitly adduces the opening section of that work, up to the point where Socrates invokes the Muses before he responds to the Lysianic speech (237a), as a model of ‘grace and charm’.¹⁵ The ‘impressionistic’ character of this description is in fact typical, not just for Dionysius, but for much ancient criticism in general, particularly where praise is being distributed.¹⁶ However traditional it might have been, Dionysius’ choice of passage here draws, as was normal in antiquity, a very close link between subject (a lovely piece of nature) and style (a ‘lovely’, ‘natural’ one), and not only shows how the Platonic text might be read as self-referential on a stylistic as well as a structural level, but also seems to foreshadow some of the programmatic imagery of Hellenistic and Roman poetry, together with modern ‘poetological’ readings of that poetry – descriptions of ‘small’ things invoke the thin, slender style of modern poetry etc.; the more we know about Hellenistic criticism, the more complex seem the relations between what was produced and how it was described. It is thus not just Phaedrus who is discovered ‘coming from Lysias’ at the opening of the work (227a1–2), but also the Phaedrus which begins in the plain, simple style of which Lysias was the perfect paradigm. For an Atticist such as Dionysius, Plato could not have done better than quite literally to put a speech of Lysias into his own text.

The echoes of the locus amoenus of the Phaedrus seem, perhaps appropriately, deceptively obvious. If Plato's simple style is ἐκτόπως ἡδεῖα (5.2), the fresh air of the grove in the Phaedrus is σφόδρα ἡδύ (230c2).¹⁷ Dionysius’ phrase varies the Platonic one, presumably under the influence of ἀτοπώτατος which follows almost immediately in Plato (230c6); Dionysius uses the adverb ἐκτόπως nowhere else, and the variation calls attention to the model, even though Dionysius has not yet specifically directed our attention to the Phaedrus.¹⁸ So too, as if to confirm our suspicions, καθαρὰ…καὶ διαυγής, ὥσπερ τὰ διαφανέστατα τῶν ναμάτων (5.2) picks up Socrates’ description of the stream at Phaedrus 229b7–8, χαρίεντα γοῦν καὶ καθαρὰ καὶ διαφανῆ τὰ ὑδάτια κτλ., and by the use of a comparison signals the borrowing and its purpose.¹⁹ ‘Purity and clarity’ are, for Dionysius, the hallmarks of this ‘Lysianic’ style; for Quintilian, Lysias’ style was ‘subtle and elegant…with nothing vacuous or artificial, though closer to a pure spring than to a mighty river’ (10.1.78), and this is exactly how Dionysius views the Lysianic Plato.²⁰ The description of style in terms of rushing or quietly flowing water is of course ubiquitous, both in critical writing and in poetry itself, but here it seems hard not to think of the ‘pure and unsullied’ water of the end of Callimachus’ Hymn to Apollo (vv. 111–12), just as the description of Plato's style as ἀκριβής τε καὶ λεπτή (5.2) may also call the Hellenistic poet to mind;²¹ behind those verses of the Hymn to Apollo will lie, inter alia, stylistic discussion of the kind to which Dionysius is also heir, and it is probably fair to use the similarity to push back in time the origins of Dionysius’ critical language and framework.

In very different mood, however, Dionysius later in the Demosthenes uses similar imagery to mock the description of the life of early man in Plato's Menexenus; here the thrust of the criticism is that some of the style is too plain, too unelaborated, so that it becomes ‘mean’ (ταπεινόν, εὐτελές) and unworthy of the subject which is being described, whereas elsewhere mean subjects are described with inappropriate elevation. The charge of ‘belittling’ a grand subject is a familiar critical move – we see it, for example, in ‘Longinus’’ treatment of Aratus’ Phainomena (On the sublime 10.6)²² – but Dionysius also launches his sarcasm at Plato's ‘noble language’ (γενναία φράσις):

The image of the stream picks up, not merely the familiar ‘water imagery’ of stylistic discussion, but probably something specifically Platonic. Critics used the notion of the ‘stream of language’ to describe both the flow of words within a particular text and the flow of language and ideas from one author to another; Plato himself uses the latter idea in the Phaedrus where Socrates, with some fairly patent irony, tells Phaedrus before his speech in rivalry with Lysias that he must have been filled with knowledge on these erotic subjects ‘like a jar’ filled from ‘other people's streams’ (235d1).²³ ‘Longinus’ famously describes Plato as ‘channelling off for himself countless tributaries from that famous Homeric stream (ἀπὸ τοῦ Ὁμηρικοῦ κείνου νάματος) (On the sublime 13.3);²⁴ the similarity to Dionysius’ expression may be chance, but it is as likely that common ‘sources’ stand behind both passages. To mock the drying up of Plato's fluency, Dionysius here echoes a famous self-description of the grand comic poet Cratinus from perhaps his most famous play, the Pytine:²⁵

No doubt Dionysius thought that putting Plato on a par with, on one hand, a drunk and comically grandiose comic poet and, on the other, a ‘super-Callimachus’ was a very sharp stiletto indeed.²⁶ Moreover, ὦ θεοὶ καὶ δαίμονες, which may be seen to vary Cratinus’ ἄναξ Ἄπολλον,²⁷ probably taunts Plato through an evocation of Diotima's distinction between gods and daimones at Symposium 202e–3a, a passage the style of which Dionysius would very likely not have admired.²⁸ Here too, however, the Phaedrus may have been in Dionysius’ mind. If any ancient writer could be described as having ‘the Ilissos in his throat’, it is the Plato of the Phaedrus.

The cicadas of the Phaedrus, with their λιγυρόν song (230c2, cf. further below), became not merely an emblem for Plato's style in this famous passage, but also stand behind the λιγὺς ἦχος of the cicada with which Callimachus associates the style of his Aitia (fr. 1.29–30), there too, as in Dionysius, contrasted with the ὄγκος of a louder, grander, less ‘clear’ style (cf. ὀγκήσαιτο, v. 31). For Dionysius, the paradigm of such translucent clarity is of course Lysias, and this is a clarity of both style and subject; when Dionysius goes on to describe the approved Platonic ‘simple’ style as ‘seeking after common words and striving for clarity (saphêneia)’ (5.3), he may as well have been describing Lysias (cf., e.g., Lysias 3–4), and there could be no clearer illustration of how the presence of Lysias in the Platonic text has influenced critical judgements of its style. The contrast which lurks behind such judgements is, of course, with the grand style; much, for example, in Thucydides and Demosthenes, for all their power, remains δυσείκαστα…καὶ ἀσαφῆ καὶ δεόμενα ἐξηγητῶν, ‘hard to understand…and unclear and requiring interpreters’ (Lysias 4.2). As often, Dionysius there rewrites a famous snippet of classical poetry to make his point: he chooses an allegorical passage about obscure language from the grandest of poets to point the contrast between such grandeur and the limpid clarity of Lysias:²⁹

In the Demosthenes too, as we shall see, Dionysius uses the grandeur of Pindar as a touchstone against which to highlight the stylistic virtues and vices of the authors he treats.

The patina of ‘archaicness’ which runs ‘gently (ἠρέμα) and imper-ceptibly’³⁰ over the simple Platonic style itself recalls the ‘gently (ἠρέμα)’ sloping grass of the Phaedrus (230c3), and Dionysius uses it to introduce the rewriting of Phaedrus 230b–c as a description of this simple style as a locus amoenus. The gestures which this passage makes to the diction of the Phaedrus (ἠρέμα, ἄνθος varying Plato's rare (and archaic?) ἄνθη, εὐωδεστάτων) are themselves relatively unobtrusive, or at least Dionysius has rearranged the echoes so that there is no sense – as indeed in a writer such as Dionysius would not have been expected – of a simple ‘taking over’ of the Phaedrus passage. The two qualities which mark this style of Plato in particular are τὸ λιγυρόν and τὸ κομψόν (5.4);³¹ the former picks up the θερινόν τε καὶ λιγυρόν sound of the cicadas in the Phaedrus (230c2),³² and the latter, a term which Dionysius very probably used of Lysias himself, recalls the κομψότατον of the sloping grass (230c2).³³ λιγυρός is found in a number of rhetorical and stylistic contexts which emphasise musicality and attractiveness.³⁴ The Aristotelian treatise De audibilibus makes cicadas, grasshoppers and nightingales the models for this quality, which is associated with sounds and voices which are λεπταί and with qualities such as ὀξύτης and ἀκρίβεια, and which explicitly has nothing to do with ὄγκος or βάρος (804a22–32). That passage makes very clear that there was a shared vocabulary for the analysis of sound and style; the alignment and linkages between terms may shift between different contexts, but a broadly consistent set of dichotomies emerges, and it is a pattern which we find not just in Hellenistic criticism, but also – as we have seen – in Hellenistic poetry. In the current instance, for example, we may think that τὸ λιγυρόν is to be opposed to the παχύτης of Plato's unsuccessfully grand style (5.4),³⁵ an opposition which may (again) recall the Callimachean Reply to the Telchines. A very interesting parallel to this instance occurs in the second part of the treatise on Demosthenes, in which Dionysius seems to have moved from an analysis of ‘styles’ to that of the three ἁρμονίαι familiar also from his De compositione verborum. Here the two ends of the spectrum are represented by, on one side, the ‘austere and archaising harmonia which aims at dignity (τὸ σεμνόν) rather than elegant wit (τὸ κομψόν)’ (38.1) and, on the other, by the ‘polished and theatrical (γλαφυρὰ καὶ θεατρική) harmonia which chooses elegance over dignity’ and which privileges ‘euphony and musicality and the pleasure which arises from them’ (40.1). The latter harmonia, which is ‘polished and λιγυρά and theatrical and which gives much emphasis to elegant wit (τὸ κομψόν) and softness (μαλακόν)’ is the harmonia by which ‘festival crowds and large gatherings of the uneducated (ὁ συμφορητὸς ὄχλος) are bewitched (κηλοῦνται)’ (36.5). Two points are worthy of note in the present context. Whereas in the earlier part of the treatise, in the discussion of Plato's ‘simple’ style, τὸ θεατρικόν seems to be a fault into which the pursuit of elegant wit (τὸ κομψόν) can lead (5.4),³⁶ thus ‘showiness’ perhaps, here the two terms work together (and with τὸ λιγυρόν) and are applied to the same harmonia; there is, however, a world of difference between ‘large festival crowds’ and Plato's audience, and the avoidance of τὸ θεατρικόν in the latter case is readily understandable. Secondly, this passage may perhaps evoke Socrates’ description of ‘the many’ who doze in the heat, lulled (κηλουμένους) by the sound of the cicadas ‘because of their sluggishness of mind’ (Phaedrus 259a1–4).

How early we may trace such stylistic analysis of Plato's genus dicendi amoenum is uncertain, but one further fragment of the early reception of Plato may be relevant and may help to confirm the implications of passages such as the end of Callimachus’ Hymn to Apollo (above). The third-century poet and philosopher Timon also connected Plato with cicadas:

The cicadas are taken from the description of the Trojan elders at Iliad 3.150–2, although, as Di Marco notes in his commentary on Timon, it is hard to believe that there is not also a reference to the Phaedrus. Be that as it may, for ἀγορηταί | ἐσθλοί of that Homeric passage Timon substitutes ἀγορητὴς | ἡδυεπής from the famous Homeric introduction of someone even older than the Trojan elders, Nestor:

The ‘pleasurable words’ which Timon ascribes to Plato may not necessarily refer to the ‘beautiful’ passages of the Phaedrus, but the evocation of the Homeric Nestor is of some interest in this context. Later rhetorical theory at least, which regularly sought the origin for all forms of discourse in Homer, saw in this passage of Iliad 1, together with Antenor's account of the Greek embassy to Troy at Iliad 3.203–24, the origins of stylistic difference within oratory, with Nestor, Menelaos and Odysseus each embodying a different style;³⁹ Socrates’ jesting reference in the Phaedrus itself to rhetorical treatises by Nestor, Odysseus and Palamedes (261b6–8) perhaps pushes these ideas back into Plato's own time. However that may be, the ‘pleasantness’ and ‘honey-sweetness’ of the Homeric Nestor's style share obvious common ground with Dionysius’ description of Plato's simple and pure style, and Quintilian associates that honeyed sweetness with his ‘middle style’, of which Nestor is the Homeric paradigm (12.10.64); this style is ‘like a gentle and clear river, shaded by green banks on both sides’ (12.10.60), a description which once again brings the Phaedrus irresistibly to mind.⁴⁰ Other aspects of Quintilian's description of this middle style map less well on to Dionysius’ account of Plato, but we should not expect, either in the nature of things or in the state of our evidence, a very neat fit between any two stylistic systems. What is important in the present context are the suggestive links between Timon's satirical presentation of Plato, a presentation which seems specifically to evoke the Phaedrus, and a stylistic discourse, three centuries before Dionysius, which also used Plato as one of its prime witnesses.⁴¹

In 5.4–6 Dionysius proceeds to describe Plato's far less successful ‘grand style’:

Here the verbal evocation of the Phaedrus apparently fades, though it will be from Socrates’ speeches in that dialogue that this disapproved style will shortly be illustrated (7.3). The principal charge here, apart from that of excessive verbiage, is – to oversimplify – of writing like a poet, and that was indeed the central stylistic charge brought against Plato throughout antiquity,⁴² and one which Plato himself clearly encourages explicitly at various points through the Phaedrus; at 238d2, for example, Socrates describes part of his first speech as ‘no longer far from dithyrambic’ and at 257a5 he calls the diction of the palinode ‘poetic’, as indeed its model was the ‘palinode’ of Stesichorus.⁴³ That parts of the Phaedrus are in fact replete with allusions to and language redolent of archaic poetry is well known.⁴⁴ Cassius Longinus later described Plato's grandeur (ὄγκος) as ‘too poetic for prose’ (fr. 49.108–9 Patillon–Brisson), a judgement which is very close to how Dionysius expresses the matter in the Letter to Pompeius, ‘he brought the grandeur of poetic style into philosophical discourse’ (2.2). Plato is here the very opposite of Dionysius’ Lysias, who used ‘figured speech’ (τροπικὴ λέξις) ‘very little indeed’ (Lysias 3.1) and was (paradoxically) the true ‘poet of discourse’ (ποιητὴς λόγων) in ordinary words without metre, but with a real harmonia (‘arrangement’) which bestowed brilliant ornament to his language (κοσμεῖ τε καὶ ἡδύνει, words regularly used of poets) and contained nothing tumid or vulgar (Lysias 3.8). This tradition about Plato's unfortunate poeticism may be seen, from another perspective, as one (among several) negative spins put upon the association between Homer and Plato, to which ancient criticism returned time and again. It was, as we have seen, an association which Plato himself encouraged.⁴⁵

In his discussion of Plato's ‘bad’ style, Dionysius has not really left the Phaedrus behind, even momentarily, for it was the ‘dithyrambic’ parts of that dialogue which were always the paradigm case for such artificial and over-elaborated writing. The tradition goes back at least to Aristotle with his firmly theorised insistence on the stylistic differences between poetry and prose. In the discussion of style in the third book of the Rhetoric, Aristotle notes that the key to all rhetorical usages is ‘appropriateness’, τὸ εὐκαίρως ἢ μὴ εὐκαίρως χρῆσθαι (1408a36); for Dionysius the grandiose Plato is ἄκαιρος in his use of metonymy and his ‘allegories’ preserve neither measure nor kairos (Dem. 5.5–6). For Aristotle, one instance of the importance of appropriateness relates to ‘compound-words and accumulation of epithets and strange words (ξένα)’ (Rhet. 1408b10–11, cf. Dion. Hal. Dem. 5.5 above); these are particularly appropriate to emotional speech, when orators speak ἐνθουσιάζοντες (1408b17), and such a style is therefore also appropriate to poetry, ‘for poetry is ἔνθεον’. One appropriate context for such a style is ‘when speaking ironically, as Gorgias used to do or like the passages in the Phaedrus’ (1408b19–20). Aristotle here echoes the enth-ousiasmos which Socrates claims to feel in the Phaedrus (241e5, 263d2), as well as, with perhaps a touch of ‘speaking ironically’ himself, a particularly Platonic view of poetry (cf. esp. Ion 533e2–5a1).⁴⁶ Aristotle's pupil, Dicaearchus, is reported as reproving the manner of the Phaedrus as ‘vulgar’ (φορτικόν, Diog. Laert. 3.38 = Dicaearchus fr. 42 W = 48 Fortenbaugh–Schütrumpf),⁴⁷ the same word which, along with many others, Dionysius, who follows Aristotle in linking Gorgias and the Phaedrus, applies not just to the ‘dithyrambic and vulgar parts’ of Gorgias’ speeches (Lysias 3.4) and of Plato's Menexenus (Dem. 29.5), but also (unsurprisingly) to the hated ‘Asianic’ rhetoric described in the proem to On the ancient orators (1.4, 2.4);⁴⁸ that Dionysius saw similarities between Plato's grand style and this degenerate ‘modern’ rhetoric will have been not the least reason for his anxieties.

From a modern perspective, perhaps one of the most striking aspects of Dionysius’ discussion, particularly after Aristotle's recognition of Platonic ‘irony’, is the apparent absence of any consideration at all of purpose, tone or irony: everything, the good and the bad, in Plato is an example of ‘Platonic style’, and open to praise or censure on that basis.⁴⁹ Just as Dionysius proceeds to use Socrates’ references to his own dithyrambic style in the Phaedrus (238d3, 241e2) as an acknowledgement by Plato of ‘his own lack of taste (ἀπειροκαλία)’ in writing those passages (6.4, cf. Letter to Pompeius 2.2), so Dionysius would presumably have seen Socrates’, to us very obviously ‘ironic’, likening of Agathon's style in the Symposium to that of Gorgias (198c), with particular respect precisely to τὸ καλλιεπεῖν, as another acknowledgement by Plato of his own stylistic failings. It is, however, important that Dionysius’ critical blindness (if that is what it is) was by no means universal. Thus, at the end of antiquity, the neo-Platonist Hermeias, who wrote a kind of commentary on the Phaedrus which seems to have drawn to some degree at least on Hermeias’ teacher Syrianus,⁵⁰ listed three charges (ἐγκλήματα) which had been brought against the work:⁵¹

Two of Hermeias’ responses to these charges are worth noting in the present context. To the second charge he notes that Socrates was trying to save Phaedrus from the noxious (and immoral) attentions of Lysias, by showing him the dangers of Lysias’ verbal style and thoughts, and was thus drawing the young man up ‘from the superficial and apparent beauty of Lysias’ empty (ψιλοῖς) and godless words towards what is beautiful in the soul and intellect’ (10.2–13 Couvreur). ψιλός here probably means ‘empty, without real substance’,⁵² but we can perhaps also sense a Platonist argument in which Lysias’ highly praised verbal and stylistic ‘simplicity’ and ‘artlessness’ are turned against the orator: Lysias may be simple and unadorned, but there is no philosophical substance there either. This would lead well into Hermeias’ defence of the Phaedrus against the third charge:

Hermeias’ neo-Platonism here determines his attitude to Plato's ‘higher’ language for describing truths only to be grasped with the mind, but he presumably also reflects a long rhetorical, as well as philosophical, tradition in searching for a contextual reason for Plato's use of a particular style,⁵⁴ and it may be possible to trace this (again) back to the Phaedrus itself.

In his prayer to Eros at the conclusion of his second speech, Socrates says that the diction of the speech was of necessity rather poetic ‘because of Phaedrus’ (257a5). Various explanations for this wording have been proposed,⁵⁵ but it would be easy enough to understand ‘for the sake of Phaedrus’ and this could then well lead to the kind of explanation which Hermeias (above) proposes. This would indeed be strengthened by the fact that, in joining in Socrates’ prayer, Phaedrus immediately expresses his wonder that Socrates’ second speech was so much ‘more beautiful’ (καλλίων) than the first, to such an extent that he might even suspect that Lysias’ speech was ‘banal’, ταπεινός (257c2–3). ‘Banality’, ταπεινότης, of both thought and diction was precisely the danger to which the plain style was exposed,⁵⁶ and thus here too the sense that Socrates’ style was imposed by the necessity of competing with Lysias is suggested by the Phaedrus itself. As for the grand style, according to Dionysius, a Thucydides or an Isocrates used this style ‘to amaze’ (εἰς τὸ καταπλήξασθαι, Dem. 2.5, 4.3, cf. Isocrates 4.3), just as Lysias’ predecessors, like Gorgias, used their exotic style ‘to amaze the ordinary man’ (καταπληττόμενοι τὸν ἰδιώτην, Lysias 3.3). So too, Hermogenes, for whom Plato represents the peak of achievement in πανηγυρικὸς λόγος, notes that there is a special point in the way in which Socrates draws attention to the style in which he is going to speak, so as to create a contrast with the speech of Lysias (387.17–20 Rabe). When, by contrast, at 234d1 Socrates tells Phaedrus that he thought Lysias’ speech ‘marvellous…so that I was knocked out’ (ἐκπλαγῆναι), it is possible that the word is already chosen to draw attention to Lysias’ plain style; being ‘knocked out’ is precisely the wrong reaction to such a style, but Phaedrus may not have picked up this amusing nuance.⁵⁷

In what was in part an explanation for its stylistic ‘excesses’, the Phaedrus was held by some critics to have been Plato's first work, a μειρακιῶδες iuuenile in fact.⁵⁸ Again, the late antique Prolegomena to Plato sum up a very rich critical tradition:

Here the idea of a μειρακιῶδες style combines with the familiar notion that Plato wrote poetry before he took to philosophy. The closest Dionysius comes, at least in the Demosthenes,⁶⁰ to an explanation for the two Platonic styles is also essentially biographical, but is of a quite different – indeed almost the very opposite – kind. Here too the Phaedrus may be involved:

The plain simplicity of the ‘Socratic dialogues’ of men such as Aeschines of Sphettos and Antisthenes was a familiar tenet of ancient stylistic criticism,⁶¹ which, for what it is worth, is not contradicted by the occasionally substantial fragments of such writers which have been preserved. Behind Dionysius’ claim here will also lie the (almost certainly correct) notion, which surfaces in various guises in the later tradition, that Plato was not the first to write ‘Socratic dialogues’; he did indeed change, stylistically and in every other way, a literary practice which had at least taken recognisable shape before he began to write.⁶² Elsewhere, in discussing criteria for deciding whether a work ascribed to a classical writer is genuine or not, Dionysius notes that ‘those who claim to imitate Plato and who cannot catch his archaic and sublime flavour, his charm and beauty, introduce dithyrambic and vulgar (φορτικά) words and in this way are easily exposed’ (Dinarchus 8.1). Such ‘dithyrambic’ language was thus thought to be a hallmark of ‘the Platonic’; Dionysius’ point here may be that, in the genuine Plato, this style is limited to particular passages, whereas unskilled imitators throw in such words indiscriminately, regardless of context, because this is ‘Platonic’. We have already noted that Dionysius counts Plato as the only exception to the ‘plain style’ of ‘the whole Socratic school’ (Dem. 2.2); Plato in fact absorbed this style from earlier Socratic writers, and preserved it in some dialogues, such as the Philebus (Dem. 23.4), but allowed himself also to be attracted to the dangerous charms of the elaborate, poetical style.

When Dionysius comes to illustrate this style in the Phaedrus, he first turns his attention to Socrates’ prayer before his first speech, as the place where the rot sets in:

An invocation of the Muses and a snatch of poetic diction (ξύμ μοι λάβεσθε) was more than enough to raise Dionysius’ hackles, even without the fact that Plato is now setting himself to compete with Lysias.⁶³Dionysius’ imagery here is, however, of considerable interest. The stylistic switch comes, according to Dionysius, ‘like a strong wind bursting from a clear and still sky’ (7.3). The image is pointed. The Lysianic style of the opening part of the work is indeed, for Dionysius, ‘clear and calm’ (εὔδιος καὶ σταθερός), and καθαρός may be used of a cloudless sky as well as of purity of language;⁶⁴ ταράττει suggests the disorder of things caused by strong wind and hence the lack of properly ordered clarity, characteristic, in Dionysius’ view, of Plato's grand style (cf. 5.4–5).⁶⁵ The familiar association of storms with grandiose poetry such as epic resonates here. Dionysius’ image is intended to sound poetic – it enacts what it is talking about – just as his earlier claim that ‘[Plato] becomes particularly storm-tossed (χειμάζεται) in the matter of figurative speech’ (5.5) reinforces through metaphor a point about the use of tropes such as metaphor.⁶⁶ The technique is not unusual in ancient criticism. We may be reminded of the general points made (?) slightly later by ‘Longinus’ about the origins of puerility, τὸ μειρακιῶδες (cf. Dion. Hal. Dem. 5.6), which arises when writers ‘aim at the unusual and the artificial and above all at what brings pleasure, and thus run aground (ἐξοκέλλοντες)⁶⁷ on vulgarity and affectation’ (On the sublime 3.4). More particularly, at 32.7 ‘Longinus’ notes that tropes such as metaphors always carry the risk of excess (τὸ ἄμετρον, cf. Dion. Hal. Dem. 5.4–5)⁶⁸ and, having just heaped great praise upon Plato's famous image of the body as a state in the Timaeus, he then uses a notorious wine-mixing image from the Laws to illustrate a less flattering feature of the critical tradition about Plato ‘the poet’:

Some of the language and ideas are familiar from Dionysius – with ἐκφερόμενον cf. Dem. 7.3 (above) and with the charge of ‘allegory’ cf. Dem. 5.6, 7.7 – but here too ‘Longinus’’ critical images clearly evoke Plato's own ‘wine’ image; as well as ἀκράτους and βακχείας, it is tempting to think that διασύρουσι evokes a Dionysiac sparagmos.⁶⁹ It is often thought that (again) ‘Longinus’ has particularly in mind Dionysius’ colleague and fellow-Atticist Caecilius of Calacte, whom he proceeds to cite in 32.8.⁷⁰

The passage from Socrates’ first speech which Dionysius proceeds to single out as particularly dithyrambically overblown and signifying nothing is, unsurprisingly, 238c, on the origins of erôs in rhômê (‘strength’), which occurs immediately before Socrates’ first recognition of the heightened quality of his speech:

This passage perhaps also suggests that the choice of ἐρᾶν to describe what happened to Plato when he ‘fell in love with’ the grandiose style of Gorgias and Thucydides (Dem. 6.4, above) was not a random one; the verb certainly suggests the exaggerated emotional effects at which the elaborated, grand style aims (effects of ψυχαγωγία and κατάπληξις in fact), but it also evokes the ἐρωτικοὶ λόγοι (7.1) of the Symposium and particularly of the Phaedrus, which is, for Dionysius, the principal witness to this unhappy stylistic development. ‘Irrational desire’, which overcomes τὸ ὀρθόν, and an unreasonable pursuit of the pleasures of beauty, κάλλος, which has both a physical and a stylistic sense (cf. Plato's immoderate ὁρμή towards τὸ καλλιεπεῖν, Dem. 5.4 cited above), lead not just the lover astray, but also may be seen to have wrought havoc with Plato's stylistic judgement. We see again how Plato's own words are given stylistically programmatic force within the critical tradition.

Dionysius’ next selection (Dem. 7.6) of poetic material from the Phaedrus is the celestial parade of Socrates’ great speech:

In antiquity this was one of the most frequently cited extracts of the Phaedrus; Hermogenes refers to it three times as a model of grandeur and σεμνότης,⁷¹ and more than one imperial writer exploits its allegorical potential.⁷² The cosmic subject and the imaginative vision, as well as the fact that what is being described is high above us, would in fact make such a passage an excellent candidate for the Longinian sublime, but for Dionysius, whose model of stylistic tact is Lysias, this is simply too close to high poetry and therefore a serious stylistic misjudgement; the cosmic visions of the Phaedrus are of course explicitly an alternative (in every way, including style) to the ‘Lysianic’ opening section. Zeus’ chariot and the chariot of the soul, which both use and seek to obscure Homeric poetry, are iconic images, as indeed Plato presumably intended them to be, for what ‘philosophical poetic prose’ might look like.⁷³ Dionysius will have been as aware as Plato of the very close association of chariots and verse, and a common later way of figuring the historical change from poetry to prose was of language ‘stepping down from the chariot’.⁷⁴ So too, Lucian reverses Dionysius’ history of the dialogue form, or at least takes it a stage further, when his personified Dialogue, citing Zeus’ chariot from the Phaedrus, complains that he used to be grandiosely philosophical and poetical, but ‘Lycinus’ has brought him down to earth (Bis accusatus 33–4).⁷⁵ For Dionysius, however, such a passage as the celestial parade of the Phaedrus has overstepped the mark; if it was metrical, it would be poetry (Dem. 7.7). Such a ‘high flown’ passage threatens the very essence of the πεζὸς λόγος; Plato's extended comparison of the soul to a chariot is one of those ‘lengthy allegories which lack due measure and proportion’ (Dem. 5.6).

Dionysius proves the ‘poetic’ nature of this passage by juxtaposing it to (what we call) the opening of the ninth Paean of Pindar, one of the grandest and most sublime of poets:

ἀκτὶς ἀελίου, τί πολύσκοπε μήσεαι,

ὦ μᾶτερ ὀμμάτων, ἄστρον ὑπέρτατον

ἐν ἁμέραι κλεπτόμενον; <τί δ’> ἔθηκας ἀμάχανον

ἰσχύν <τ’> ἀνδράσι καὶ σοφίας ὁδόν,

ἐπίσκοτον ἀτραπὸν ἐσσυμένα; 5

ἐλαύνεις τι νεώτερον ἢ πάρος;

ἀλλά σε πρὸς Διός, ἱπποσόα θοάς,

ἱκετεύω, ἀπήμονα

εἰς ὄλβον τινὰ τράποιο Θήβαις,

ὦ πότνια, πάγκοινον τέρας 10

]ρα[–––––]

[– – – – – – –]

]ῶνος [– –], πολέμοιο δὲ σᾶμα φέρεις τινός,

ἢ καρποῦ φθίσιν, ἢ νιφετοῦ σθένος

ὑπέρφατον, ἢ στάσιν οὐλομέναν 15

ἢ πόντου κενεώσιας ἂμ πέδον,

ἢ παγετὸν χθονός, ἢ νότιον θέρος

ὕδατι ζακότωι ῥέον,

ἢ γαῖαν κατακλύσαισα θήσεις

ἀνδρῶν νέον ἐξ ἀρχᾶς γένος; 20

ὀλοφύ<ρομαι οὐ>δέν, ὅ τι πάντων μέτα πείσομαι…

Pindar, Paean 9. 1–21⁷⁶

If Plato's Socrates speaks of Zeus driving (ἐλαύνων) a ‘winged chariot’, Pindar addresses the sun, the ‘highest star’, who is a ‘swift driver of horses’ (v. 7) but who is ‘setting in motion’ (ἐλαύνεις) something new. Moreover, the Pindaric subject matter, an eclipse which betokens cosmic upheaval, literalises one of the claims which Dionysius has made about the Platonic ‘elaborated’ style, namely that it ‘darkens what is clear and makes it like gloomy night’ (5.4);⁷⁷ the eclipse is a sign which is anything but ‘clear’ – it ‘requires interpreters’ (cf. Lysias 4.2 above) – and there is here no saphêneia of any kind. For Dionysius, Plato's high-flown style was indeed inimical to ‘the path of sophia’ (v. 4). For the neo-Platonist Hermeias, of course, the very opposite was true, and it is instructive to set Dionysius’ analysis alongside Hermeias’ contrast between the Lysianic speech and Socrates’ first speech in the Phaedrus:

Not only does an eclipse destroy saphêneia, but the very number of possible interpretations of it offered in the Pindaric passage, as by implication the very length of Socrates’ speech in the Phaedrus, seem also to pick up Dionysius’ complaints about how Plato's elaborated style is marked by verbosity (5.4) and by ‘a flooding out (ἐκχεῖται) into tasteless periphrases and a meaningless wealth of words’ (5.5).⁷⁸ The Platonic flood is picked up by the cosmic floods perhaps portended by Pindar's eclipse (vv. 16, 19), but there may also be a particular point to Dionysius’ choice of words here. ‘Longinus’ too notes that Plato's periphrases are not always successful:

This passage shares with Dionysius the charge of Platonic παχύτης, but if the standard example of unfortunate periphrasis in Plato was indeed a phrase featuring the word ‘wealth’, then we would have here another example of a Platonic text being taken almost as stylistically self-reflexive; we may recall Dionysius’ mockery of the drying up of the ‘rich (πλούσιον) Platonic stream’ (Demosthenes 28.7, above pp. 157–8). The Pindaric style – the ‘mighty strength of snow’ in place of ‘snow’ – is thus made to reflect (adversely) on Plato;⁷⁹ we may be reminded of this in the later discussion of the Menexenus, when Dionysius mocks Plato for writing πηγὰς τροφῆς (‘streams of nourishment’) instead of γάλα (‘milk’), to describe what a mother offers her new born child (28.7, Menexenus 237e).

Behind, and indeed in front of, Dionysius here lies a critical language of some interest. When, for example, after a lengthy lacuna, the text of On the sublime 12 resumes its discussion of auxêsis, ‘Longinus’ is describing Plato:

The liquid image of style found here is, of course, very common – ‘Longinus’ proceeds to contrast the χύσις of Cicero with the intensity of Demosthenes⁸⁰ – but we have already noticed that ‘wealth’ might have a particular resonance in discussions of Platonic style. If Dionysius associates the Platonic flood with Pindar's cosmic upheaval, it may again be Plato himself who shaped the critical discussion of Platonic style, as we witness it in ‘Longinus’. Just as the Phaedrus lent itself to later critics for self-referential readings, so too would the Symposium, in which Diotima/Socrates’ great speech about ‘giving birth to logoi’ could very easily have been taken to refer to Plato's own writing; if anyone ‘overflows (εὐπορεῖ) with words about virtue’ (Symp. 209b8) it is certainly Plato himself. In the presence of the beautiful, that which is pregnant ‘floods (διαχεῖται) in its joy and gives birth and generates’ (206d2–3), and when finally the lover is ‘turned towards and observes the great sea of the beautiful, he gives birth to many beautiful and grand (μεγαλοπρεπεῖς) logoi and thoughts’ (210d4–5).⁸¹ It is not difficult to guess what ‘spin’ a Dionysius would have given to the ‘beautiful grandeur’ of Diotima's own speech; a stylistic reading is all but invited by Plato himself, and it is perhaps not too rash to speculate that the invitation was indeed taken up. Dionysius reports Demetrius of Phalerum as having quipped of Plato's ‘Gorgianic’ passages that ‘there is a lot of the mystery-priest (τελετής) in such passages of his’ (Demosthenes 5.6 = Demetrius fr. 170 W = 133 Fortenbaugh–Schütrumpf), and perhaps the two best candidates for the object of such a jest⁸² are Diotima's ‘initiation’ of Socrates and Socrates’ description of the philosopher in the ‘palinode’ of the Phaedrus:

Here there is indeed a ‘lot of the tel-’ in Plato. Finally, it is likely that the ‘flooding sea’ of Platonic style has something to do with one of the later rival explanations for Plato's change of name from Aristocles, namely that Πλάτων is a nickname derived ‘from the breadth (πλατύτης) of his style’.⁸³

Dionysius’ comparison of Plato to Pindar may have been one stone in a mosaic of critical claim and counter-claim, and we may be able to trace a response to it, or at least to something very like it. After concluding his comparisons of Plato to Lysias and Demosthenes to Hypereides, ‘Longinus’ notes that nature draws man towards the grand and spectacular – this is what we were born for: a pure, little stream will never seize our wonder as do the Nile or the Danube. He then moves to a description of Mt Etna:

As is well recognised, ‘Longinus’ here echoes the famous and much-imitated description of Mt Etna in Pindar's First Pythian (1.21–7). Is ‘Longinus’ then answering Atticist criticism of allegedly overblown writing by matching one Pindaric τέρας with another (Pyth. 1.26, Paean 9.10)? When ‘Longinus’ notes that, just as the grand writers sometimes fall short, so the fires of heaven are ‘often darkened’ (πολλάκις ἐπισκοτουμένων), are we to think of eclipses, Pindar's ἐπίσκοτος ἀτραπός (Paean 9.5), as much as of the fact that the stars are often obscured. Both the eclipse and the volcano turn day into night, making ‘Longinus’’ riposte, if this speculation is correct, a very pointed one. It is in fact tempting to conclude from this that, in his comparison of Plato's ‘high style’ to Pindar, Dionysius may be echoing something in his colleague Caecilius, the direct target of On the sublime; that of course is neither strictly necessary nor provable, and yet it is very clear that much mutual polemic lies under the surface of the texts which happen to have survived.

Virgil was indebted to Pindar's First Pythian for a very grand and sublime description of Mt Etna at Aeneid 3.570–87,⁸⁴ and Favorinus’ unfavourable comparison of Virgil's version to that of Pindar, which is reported for us by Aulus Gellius (NA 17.10.8–19), has noteworthy points of contact with Dionysius’ criticism of Plato's grand style. For Favorinus, Virgil has so ‘piled up expressions and words that, at least in this passage, he is more abnormal (insolentior) and grandiose (tumidior) even than Pindar himself, who has been thought to have a style which is too rich and luxuriant (nimis opima pinguique…facundia)’. Virgil, in his quest for grandiose sound effects, has ‘confused’ the clarity of Pindaric expression; he has piled up (congessit…accumulauit)⁸⁵ words ‘clumsily and immoderately’ (crasse et inmodice) and ‘vainly and to no purpose’ (uacanter…inaniter, cf. Plato's ‘empty wealth of words’, Dem. 5.5), so that, like Dionysius’ Plato, all is merely ‘sound and nonsense’ (cf. Dem. 7.4, 7.6). When Favorinus concludes that Virgil's description is omnium, quae monstra dicuntur, monstruosissimum, he appears to pick up the language of τέρας in the Pindaric passage, and make Virgil's passage stylistically self-referential, as Dionysius had done to Plato. This is by no means an isolated example where language used in Greek criticism to contrast two authors or styles is taken over by Latin critics to describe differences between a Latin text and its Greek model,⁸⁶ but it is singularly instructive for the tradition in which Dionysius is placed and against which ‘Longinus’ reacts. In Favorinus’ criticism, Pindar becomes a model of clarity (luculente 17.10.13) and realism (ueritati magis obsecutus id dixit quod res erat), almost – paradoxically enough – a Lysias.

After breaking off (8.1) the discussion of Plato, so as not to commit the (Platonic) faults of περιττολογία and ἀκαιρία, Dionysius returns to the philosopher in Chapter 23. The bulk of the following chapters is taken up with an analysis of the Menexenus and the contrast between it and Demo-sthenic oratory; here Dionysius is particularly concerned with ‘political’ speeches, which should be the work of genuine πολιτικοί and ῥήτορες, such as Demosthenes, whereas Plato should have stuck to ‘Socratic dialogues’ (23.5). In summing up the contrast, Dionysius returns to the Phaedrus for a further impressionistic description of Platonic style, although this time it is not the Phaedrus about which he is explicitly writing:

Dionysius here contrasts the now familiar image of Plato's style as a locus amoenus to the really valuable ‘agricultural land’ of Demosthenes,⁸⁷ which brings both profit and delight, but he also adds a further contrast which again seems drawn from the Phaedrus. Demosthenes is to Plato as ‘bodies used to sunlight and hard work’ are to ‘those that know only shade and luxurious ease’; here Dionysius takes over, with close verbal adaptation, the distinction in Socrates’ first speech in the Phaedrus between what each of the two kinds of lover seeks in an erômenos:

Plato's words are not merely (once again) used as weapons against himself, but it is he who becomes the stylist of εἴδωλα rather than of ‘true visions’, a contrast which clearly evokes and pokes fun at Platonic metaphysics, and it is he who is concerned with surface beauty, not with ‘the useful and the real’; the Platonic locus amoenus offers only passing pleasures where one may briefly rest,⁸⁸ whereas Demosthenes offers nourishment for life.⁸⁹ Moreover, the use which Dionysius makes of this passage of the Phaedrus well illustrates how the physical language of the body and its adornment was taken over into the language of stylistic criticism: μαλθακόν, στερεόν, ἁπαλόν, χρῶμα, κόσμος all have stylistic as well as ‘corporeal’ senses, as of course does ἀνθηρόν, which for Dionysius describes Platonic style both literally and metaphorically (32.2). The imagistic language is of a very familiar kind, but a similar evocation of the Phaedrus may have contributed to a rather different use of these ideas in Cicero's description of the oratory of Demetrius of Phalerum:

Plato too avoided public life almost entirely and, according to Dionysius, the style of the Menexenus shows that all too clearly;⁹⁰ as the language of Phaedrus 239c–d is made by Dionysius to reveal, Plato's style is – not to put too fine a word on it – feminine and unmanly. Plato, the sheltered philosopher pursuing a life of theory, is – in this construction – no more at home in ‘the real world’ of πόνοι than Aphrodite, the most feminine of all goddesses, is on the field of war (23.5).⁹¹

Where the Platonic language of Dionysius’ contrast might lead is well shown in some satirical chapters of Lucian's Teacher of Rhetoric. Lucian there blends this stylistic contrast between types of oratory with Hesiod's famous ‘hard path of Virtue’ and another reworking of the Prodican ‘Choice of Heracles’ to outline the two different approaches possible to the mastery of Rhetoric.⁹² On the one hand, there is the steep and narrow path of πόνοι and sweat, a path followed by Demosthenes and Plato, here joined – rather than distinguished – as two of the paradigms of classical prose;⁹³ the teacher who offers you that road will be ‘wiry, manly of gait, and with a deep suntan’. The other path is shaded and level and passes through flowery meadows, a path of ease (Rhet. Praecep. 14, 24, 26) which takes very little time to complete, and the guide along this road is correspondingly effeminate (and sexually effeminised), given to excessive concern with his appearance and the appearance, rather than the substance, of what he has to say. The polemical language of ancient stylistics thus remained stable over centuries; the targets, however, were rearranged to suit changing circumstances and needs.

In the contrast between Plato and Demosthenes in Chapter 32, Dionysius echoes inter alia, as we have seen, a contrast in the Phaedrus, but behind the intellectual structure of the contrast seem to lie further Platonic texts which give bite to Dionysius’ criticism. For the Plato of Republic 10 all poets are ‘imitators of phantoms (εἴδωλα) of virtue…and have no part of truth’ (600e), which brings us quite close to Dionysius’ description here of Plato's own style, but it is rather two passages from the Gorgias and (again) the Phaedrus which demand attention in the present context. In the Gorgias Socrates claims that rhêtorikê is, like cookery, sophistikê and ‘beautification’ (κομμωτική), a part of flattery (κολακεία), specifically in fact ‘a phantom (εἴδωλον) of a part of politikê’ (463b–d). As is very clear from, for example, Aelius Aristides’ lengthy reply to the Gorgias in ‘On Rhetoric’ (Oration 2 Behr), this description of rhetoric was, perhaps unsurprisingly, particularly provocative to later theoreticians and practisers of oratory (cf. Ael. Arist. 2.148, 172, 234, 4.46), and we should perhaps sense its influence in Demosthenes 32 also.⁹⁴ The skills which belong to ‘flattery’ have, like Plato's style, only τὸ ἡδύ as their method and goal (Gorgias 464d2, 465a2). In particular, Plato's description of how ‘beautification’ deceives through ‘forms (σχήματα) and colours (χρώματα) and making smooth and putting on clothes’ (465b4) very clearly lends itself to an application to style,⁹⁵ and we may in fact trace its descendants in such things as Ovid's frequent play with the various senses of cultus; Plato's style, which, according to Dionysius, is concerned only with a εὐμορφία based on insubstantial appearances, is itself little better than the product of κομμωτική, whereas it is the orator, in this case Demosthenes, who offers what is useful, true and fruitful. In his reply to the Gorgias, Aristides notes that it would make more sense to replace ‘cookery’ in Plato's exposition by ‘philosophy’ (2.25, 29 Behr); in a sense, this is just what Dionysius has done: Plato's barbs against rhetoric have been turned against the philosopher's own style, and here too Plato himself has shown how best to do it. The defence of oratory and sophistic mounted by Callicles in the Gorgias makes use of the contrast from Euripides’ Antiope of the brothers Zethos and Amphion, the former representing the life of seriousness and toil, the latter the pursuit of mousikê and pleasure;⁹⁶ it may in fact be that this contrast also resonates behind that of Phaedrus 239c–d. Dionysius at any rate follows Callicles in casting the philosopher as the effete aesthete in pursuit of pleasure (cf. Euripides, Antiope frr. 187.6, 196, 199.1 Kannicht), with the public man, in this case Demosthenes, as the Zethos, whose praise of toil and agricultural labour (fr. 188 Kannicht) is perhaps recalled in the final image of Demosthenes 32.2 (Demosthenes as the productive field providing life's necessities). If this reuse of the Gorgias remains implicit in Dionysius, Aristides explicitly takes Amphion's side (2.394 Behr) in telling a version of Protagoras’ history of mankind from Plato's Protagoras, but one in which it is rhêtorikê, freely given to mankind by Zeus as a result of Prometheus’ just pleading, which is responsible for human progress. Moreover, at Phaedrus 276b written discourse is said to be the εἴδωλον of the living discourse of him who knows, and this is then followed by the famous contrast between the παιδιά of the short-lived pleasures of the ‘Gardens of Adonis’ (i.e. written logoi) and the slower but lasting benefit of properly farmed and productive fields (i.e. true dialectic). Dionysius has turned the Platonic distinction around so that Plato now occupies the place of παιδιά, whereas it is Demosthenes who represents the σπουδή which for Plato was that of true dialectic.⁹⁷

Just as Aelius Aristides was later to feel compelled (or bound by literary tradition?) to defend in Oration 4 Behr the attack of ‘On Rhetoric’ upon Plato,⁹⁸ so in the Letter to Pompeius Dionysius felt compelled to defend the attack upon Platonic style of the opening chapters of the Demosthenes, and he does so by protesting that this was in fact the proper ‘critical’ attitude to one of the great figures of the past and must not be understood as implying any disrespect or lack of acknowledgement of Plato's greatness (Plato is indeed ‘near to the nature of divinity’, 2.2).⁹⁹ He reveals his respect for the philosopher (and the depth of his knowledge of his works) by littering the opening chapters with echoes of Platonic texts and using Plato (again) as the chief witness against Plato.¹⁰⁰ Thus his technique of contrasting Plato (to his detriment) with Demosthenes was in fact ‘authorised’ by Plato in the Phaedrus, in the contrast which the philosopher drew between Socrates’ own speeches and that of Lysias and in the criticism of Lysias’ speech in which Socrates indulges (1.9–11). Moreover, Plato was fond of making fun (κωμωιδεῖν) of his own predecessors such as Parmenides, Protagoras, Gorgias, Thrasymachus and so forth;¹⁰¹ in this the spirit which moved him was one of jealous rivalry (φιλοτιμία, ζηλοτυπία, 1.13–14). Dionysius’ allegedly dispassionate criticism of Plato here crosses the path of the very rich ancient anti-Platonic traditions, which are perhaps best on display in the collection of material at Athenaeus 5.215c–21a and 11. 504e–9e,¹⁰² and to which Dionysius refers (with winning disingenuousness) at 1.15–17. Plato's innate malice and jealousy, particularly towards other followers of Socrates (most notably Xenophon), and his mockery (κωμωιδεῖν, διαβάλλειν, κακολογεῖν etc.) of those he portrays or mentions in his dialogues are standard themes of these traditions.

Dionysius illustrates the spirit of rivalry (τὸ φιλότιμον) which characterised Plato from his obvious jealousy (ζηλοτυπία) against Homer, whom the philosopher had notoriously excluded from his ideal state (1.13).¹⁰³Dionysius’ critical practice is of some interest here. What he appears to have done is to have taken a traditional critical motif, namely Plato's ζῆλος (‘creative imitation’) of Homer, a commonplace of ancient criticism,¹⁰⁴ and blended it with a motif from the anti-Platonic tradition, namely Plato's ζηλοτυπία. ‘Longinus’ was to make Plato a paradigm case for the idea that μίμησίς τε καὶ ζήλωσις of the great writers of the past is one way to reach the sublime heights (De subl. 13.2, above p. 43), and when he accuses Caecilius of ‘hatred’ for Plato and a spirit of φιλονεικία (De subl. 32.8), he may again be turning the Atticists’ criticisms back upon themselves. Be that as it may, Dionysius closes his self-defence by insisting that Plato's mistakes (τὰ διαμαρτανόμενα) were only a tiny fraction of his successes (τὰ κατορθούμενα); here Dionysius is in fact close to ‘Longinus’, and we may perhaps sense an awkward suspicion on Dionysius’ part that his criticisms of ‘the divine Plato’ may have gone too far. Plato, like Homer, may sometimes nod, but too harsh a spotlight on those moments of alleged slumber might all too easily be taken for signs of the same moral failings in the critic as had just been imputed to Plato himself.