‘All the World’s a Stage-Painting’: Scenery, Optics, and Greek Epistemology

M. F. Burnyeat

Anaxarchus of Abdera is not your most famous Greek philosopher, and I have no wish to argue that he deserves better. He was known as the Happiness Man (ho eudaimonikos), either from his equable temperament or, according to other sources, because he led something called the Happiness School (hoi eudaimonikoi),¹ and he features prominently in the biography of Alexander the Great, at whose kingly presence he aims a number of caustic remarks. But all that is forgiven, I hope, in return for the splendid saying which I have borrowed for my title ‘All the world’s a stage-painting’.

I have of course reconstructed the saying from the paraphrase in Text 1, which comes from a history of theories of knowledge compiled in the first century bc, probably by the sceptic philosopher Aenesidemus, and later used by Sextus Empiricus.

[T1] οὐκ ὀλίγοι δὲ ἦσαν, ὡς προεῖπον, οἱ καὶ τοὺς περὶ Μητρόδωρον καὶ Ἀνάξαρχον ἔτι δὲ Μόνιμον ϕήσαντες ἀνῃρηκέναι τὸ κριτήριον, ἀλλὰ Μητρόδωρον μὲν ὅτι εἶπεν “οὐδὲν ἴσμεν, οὐδ ᾿ αὐτὸ τοῦτο ἴσμεν ὅτι οὐδὲν ἴσμεν”, ᾿Ανάξαρχον δὲ καὶ Μόνιμον ὅτι σκηνογραϕίᾳ ἀπείκασαν τὰ ὄντα, τοῖς τε κατὰ ὕπνους ἢ μανίαν προσπίπτουσι ταῦτα ὡμοιῶσθαι ὑπέλαβον. (S.E. M. 7. 87–8)

As I said above, there have been not a few who have asserted that Metrodorus and Anaxarchus, and also Monimus, abolished the criterion—Metrodorus because he said ‘We know nothing, nor do we even know the very fact that we know nothing’; and Anaxarchus and Monimus because they likened existing things to a scene-painting and supposed them to resemble the impressions experienced in sleep or madness. (trans. Bury, emphasis added)

In the first sentence of the excerpt before you, Sextus is the ‘I’ of hōs proeipon, Aenesidemus the ouk oligoi from whom Sextus’ information comes. But if Sextus’ paraphrase lacks the Shakespearian splendour which I am imagining the original to have had, nonetheless his report gives us valuable information about the philosophical context in which the saying made its mark. For the passage mentions two other philosophers beside Anaxarchus, Metrodorus and Monimus. Let me introduce them, two minor characters in what will, I fear, be quite a large cast of players.

Metrodorus is the best known of several fourth-century followers of Democritus who pushed to extremes the sceptical tendencies in the epistemology of fifth-century Atomism; when Epicurus at the end of the fourth century set himself to give Atomism a more secure epistemological foundation, Metrodorus was a target of particular concern.² Monimus is a Cynic, a follower of Diogenes in his barrel, and he is one of a number of Cynics who seem to have used Democritean sayings more for their moral than for their epistemological value;³ Text 2 suggests that in his mouth ‘All the world’s a stage-painting’ would have meant ‘All is vanity’ in the spirit of Ecclesiastes, rather than ‘It’s all a big illusion’ in the spirit of Descartes’s First Meditation.

[T2] Ξενιάδης μὲν οὖν ὁ Κορίνθιος, ὡς ἀνώτερον ὑπεδείκνυμεν, μηθὲν εἶναί ϕησιν ἀληθές· τάχα δὲ καὶ Μόνιμος ὁ κύων, τῦϕον εἰπὼν τὰ πάντα, ὅπερ οἴησίς ἐστι τῶν οὐκ ὄντων ὡς ὄντων. (S. E. M. 8. 5)

Thus Xeniades the Corinthian, as we indicated above, declares that nothing is true; and so also, perhaps, Monimus the Cynic when he said that ‘All things are vanity’ (that is to say, a vain fancy that nonexistents are existent). (trans. Bury)

In between comes Anaxarchus, a vigorous moralizer too, but also an Atomist; this is presupposed by the anecdote in which he discourses to Alexander about the innumerable worlds postulated by Democritus’ cosmology and Alexander bursts into tears because he is not yet master of one world:

[T3] ᾿Αλέξανδρος ᾿Αναξάρχου περὶ κόσμων ἀπειρίας ἀκούων ἐδάκρυε, καὶ τῶν ϕίλων ἐρωτώντων ὅ τι πέπονθεν, “οὐκ ἄξιον” ἔϕη “δακρύειν, εἰ κόσμων ὄντων ἀπείρων ἑνὸς οὐδέπω κύριοι γεγόναμεν;” (Plut. Tr. an. 466 d = Anaxarch. 72 A 11 DK)

Alexander wept when he heard Anaxarchus discourse about an infinite number of worlds, and when his friends inquired what ailed him, ‘Is it not worthy of tears,’ he said, ‘that, when the number of worlds is infinite, we have not yet become lords of a single one?’ (trans. *Helmbold)

iam Alexandri pectus insatiabile laudis, qui Anaxarcho comiti suo ex auctoritate Democriti praeceptoris innumerabiles mundos esse referenti: ‘heu me’ inquit ‘me miserum quod ne uno quidem adhuc potitus sum’. (Val. Max. 8. 14, ext. 2 = Anaxarch. 72 A 11 DK)

Alexander’s appetite for fame was insatiable. He said to his companion Anaxarchus, who was retailing on the authority of his teacher Democritus the existence of innumerable worlds: ‘Alas for me, I have not yet made myself master of one!’ (trans. *Shackleton Bailey)

Having sketched a context for Anaxarchus’ saying ‘All the world’s a stage-painting’, I can now ask the question which this lecture will try to answer. Why, if you are a fourth-century Atomist looking for a memorable way to express, whether for epistemological or for moralistic purposes, the sceptical outlook that Metrodorus has made fashionable—why do you choose the image of scene-painting?

This is, after all, one of the earliest extant occurrences of the word skēnographia. The only other certainly fourth-century occurrence⁴ is that famously truncated sentence of Aristotle’s Poetics (Text 4) about Sophocles introducing the third actor and skēnographia:

[T4] καὶ τό τε τῶν ὑποκριτῶν πλῆθος ἐξ ἑνὸς εἰς δύο πρῶτος Αἰσχύλος ἤγαγε καὶ τὰ τοῦ χοροῦ ἠλάττωσε καὶ τὸν λόγον πρωταγωνιστὴν παρεσκεύασεν, τρεῖς δὲ καὶ σκηνογραϕίαν Σοϕοκλῆς. (Arist. Poet. 1449^a15–19)

Aeschylus first introduced a second actor; he diminished the importance of the Chorus, and assigned the leading part to the dialogue. Sophocles 〈raised the number of actors to〉 three, and 〈added〉 scenepainting. (trans. Butcher, brackets added)

So, innocent souls who believe Aristotle incapable of writing such a sentence—quite a number of scholars wish to regard it as an interpolation—must look to Anaxarchus as the first known user of the term.⁵ Thus scene-painting is not such a common topic of discussion that there is no need to ask why it should turn up on the lips of a fourth-century Abderite Atomist who trails around the world in the court of a conquering king.

However, these fourth-century Atomists are not, so far as one can tell, particularly original. Their material derives from Democritus, even if they sharpen the scepticism and put a harder, more Cynic edge on the morals. More likely than not, Anaxarchus is developing a thought that was already formulated by the fifth-century Atomist Democritus.

This brings me to a well-known passage in Vitruvius, which has received a great deal of discussion from historians of the tragic stage, and even more from historians of ancient painting, but very little discussion, I regret to say, from historians of ancient philosophy, even though (in the second paragraph) it mentions both Democritus and Anaxagoras—in that order, which is the reverse of the chronological order, for we have Democritus’ own word for it that he was younger than Anaxagoras by some forty years.⁶ At present the order of names will seem a rather small and insignificant detail, but then this passage of Vitruvius, vital as it is for the early history of the tragic stage, has never been approached from the direction I am coming from, with a view to answering a question about Anaxarchus.

[T5]

10 ego vero, Caesar, neque alienis indicibus mutatis interposito nomine meo id profero corpus neque ullius cogitata vituperans institui ex eo me adprobare, sed omnibus scriptoribus infinitas ago gratias, quod egregiis ingeniorum sollertiis ex aevo conlatis abundantes alius alio genere copias praeparaverunt, unde nos uti fontibus haurientes aquam et ad propria proposita traducentes facundiores et expeditiores habemus ad scribendum facultates talibusque confidentes auctoribus audemus institutiones novas comparare.

11 igitur tales ingressus eorum quia ad propositi mei rationes animadverti praeparatos, inde sumendo progredi coepi. namque primum Agatharchus Athenis Aeschylo docente tragoediam scaenam⁷ fecit et de ea commentarium reliquit. ex eo moniti Democritus et Anaxagoras de eadem re scripserunt,

quemadmodum oporteat, ad aciem oculorum radiorumque extentionem certo loco centro constituto, lineas⁸ ratione naturali respondere,

uti de incerta re certae⁹ imagines aedificiorum in scaenarum picturis redderent speciem et, quae in directis planisque frontibus sint figurata, alia abscedentia, alia prominentia esse videantur.

12 postea Silenus de symmetriis doricorum edidit volumen; de aede Iunionis, quae est Sami dorica, Theodorus;¹⁰ ionice Ephesi quae est Dianae, Chersiphron et Metagenes … (Vitruv. 7, praef. 10–12)

10 But for my part, Caesar, I am not bringing forward the present treatise after changing the titles of other men’s books and inserting my own name, nor has it been my plan to win approbation by finding fault with the ideas of another. On the contrary, I express unlimited thanks to all the authors that have in the past, by compiling from antiquity remarkable instances of the skill shown by genius, provided us with abundant materials of different kinds. Drawing from them as it were water from springs, and converting them to our own purposes, we find our powers of writing rendered more fluent and easy, and, relying upon such authorities, we venture to produce new systems of instruction.

11 Hence, as I saw that such beginnings on their part formed an introduction suited to the nature of my own purpose, I set out to draw from them, and to go somewhat further. In the first place Agatharchus, in Athens, when Aeschylus was bringing out a tragedy, painted a scene,¹¹ and left a commentary about it. This led Democritus and Anaxagoras to write on the same subject,

showing how, given a centre in a definite place, the lines should naturally correspond with due regard to the point of sight and the divergence of the visual rays,¹²

so that by this deception a faithful representation of the appearance of buildings might be given in painted scenery,¹³ and so that, though all is drawn on a vertical flat façade, some parts may seem to be withdrawing into the background, and others to be standing out in front.

12 Afterwards Silenus published a book on the proportions of Doric structures; Theodorus, on the Doric temple of Juno which is in Samos; Chersiphron and Metagenes, on the Ionic temple at Ephesus which is Diana’s … (trans. Morgan)

This paper is therefore going to be a plea for the unity of the Classics. We have become so divided into specialisms that we seldom share our knowledge or consider whether our specialized perspectives are complementary rather than conflicting. Perhaps a historian of philosophy can help with the problems that have exercised other branches of the Classics in connection with this vexed passage of Vitruvius. I am sure that the attempt should be made, even though it will mean that I must ask you to forgive my talking with a shameless abandon about lots of things whose larger frameworks I know nothing about. For I hope also to show that, in return, the history of ancient philosophy has much to gain from considering the history of the tragic stage and the development of painting.

Let us then begin the approach to Vitruvius. Anaxarchus, as I said, points us to Democritus, whose name precedes that of the older philosopher Anaxagoras. But it is immediately obvious that Democritus’ place in Vitruvius’ story can only be understood by reference to the previous sentence about Agatharchus, whom we know from other sources to have enjoyed a certain fame as a painter in fifth-century Athens. We must therefore start with the sentence about Agatharchus.

It is remarkable how often this sentence is reported, sometimes even translated, by reputable scholars as saying that Agatharchus invented perspective¹⁴ or that he was the first to do scene-painting.¹⁵ We shall be talking in a minute about the word primum. For the moment let us take scaenam fecit by itself and, being warned by the variant translations recorded in notes 11–13, let us not take it for granted that this means, as Morgan’s translation has it, that Agatharchus painted, or made, the scenery or the scenic background for a play of Aeschylus. It is not that the words could not mean this in Vitruvius’ Latin. They certainly could (cf. 7. 5. 5). But it will become increasingly likely as we proceed that in this context the words represent the Greek skēnēn epoiēse in a Greek source, and skēnēn epoiēse means that Agatharchus made the stage-building for a play of Aeschylus. Granger’s alternative text ad scaenam fecit¹⁶ means that Agatharchus busied himself with the stage-building. But we shall find reason later to think that Granger’s textual judgement is not to be trusted. And if we adopt as the safest, because minimal, translation, ‘he made the stage’, then two results follow, both of which will be confirmed shortly.

The first result is that it is wrong to think, as many scholars have done, that Vitruvius’ report is at odds with the statement in Aristotle’s Poetics (Text 4) that scene-painting began with Sophocles rather than with a play of Aeschylus.¹⁷ So far, given our translation, Vitruvius has said nothing about scene-painting. (We will see in a moment that even on the Morgan translation there is no inconsistency with Aristotle.) The second result, more germane to my question about Anaxarchus, is that whatever Agatharchus did to interest Democritus and Anaxagoras, it was not that he produced a striking tour de force of perspective for a tragic festival which they saw or heard about.

Vitruvius makes it perfectly clear what, according to him, interested the two philosophers. It was the written word, not the painted scenery. Ex eo moniti, he says, where eo is the commentarium, hupomnēma, or treatise which Agatharchus wrote de ea, where ea on the safe minimal translation of scaenam fecit is the stage-building he made or busied himself with. Agatharchus, according to this testimony, wrote something about making or organizing the stage-building, not a book specifically devoted to scene-painting, still less a treatise on perspective. No doubt he included some remarks about scene-painting (recently introduced by Sophocles, who first won the tragedy prize in 468). Agatharchus was after all a painter. Perhaps he even made some remarks about perspectival effects in painted scenery, for the remainder of our Vitruvius text leaves us in no doubt that it was something said about scene-painting that interested Democritus and Anaxagoras. But the point I wish to emphasize for the moment is that it is to something said, and said in writing, that we must look for the beginning of the story that leads to Anaxarchus saying ‘All the world’s a stage-painting’, not to something done with brush and paint.

This is confirmed by the fact that what Vitruvius wishes to tell us is that Agatharchus was the first to write on any architectural matter, the matter in his case being stage construction. The weight of primum does not fall on (ad) scaenam fecit but on commentarium reliquit. Primum contrasts with postea at the beginning of the next paragraph. Agatharchus wrote first, then came Silenus’ book on the proportions of Doric structures. Hence scaenam fecit describes what Agatharchus was the first to write about, not what he was the first to do. This confirms that there is no conflict between Vitruvius and the statement in the Poetics that Sophocles was the first to introduce scene-painting. There is no conflict even if scaenam fecit means that he made or painted the scenery, for Vitruvius is not saying that he was the first to do so. He is saying that he was the first to write about it or, if you like, the first to combine doing it and writing about it. The entire context in Vitruvius is concerned not with the history of architectural achievements, still less with the history of painting, but with the history of architectural writings.

At this point it is only prudent for me to remind myself that prose writings from the early fifth century do not come at two a drachma. The evidence for their existence is often suspect. What reason have I to think that Vitruvius is to be believed when he writes in the twenties or thirties of the first century bc about Agatharchus writing in the fifth century bc?¹⁸

Vitruvius certainly does not claim to have seen this work of Agatharchus’ himself. What he says, in the first paragraph, is that he is immensely grateful to earlier writers who collected from still earlier ages (ex aevo conlatis) the ‘stores’ (copias) from which he can draw off material for his own use. Like Sextus in the text we began from (Text 1), he is relying on an earlier compilation, and his value as a source is dependent on the value of the compilation he is using. We know quite a lot about the value of the compilation Sextus uses, for on a number of occasions he preserves a mention by Aenesidemus (assuming he was the author of the compilation) of the particular first-century bc authority he is drawing on for his account of a given philosopher. Vitruvius is not quite so helpful, but he tells us in Text 6, in connection with writers on symmetry, that he has had access to both Greek and Latin compilations.

[T6] quorum ex commentariis, quae utilia esse his rebus animadverti, [collecta in unum coegi corpus, et ideo maxime, quod animadverti] in ea re ab Graecis volumina plura edita, ab nostris oppido quam pauca. Fufidius enim mirum de his rebus primus instituit edere volumen, item Terentius Varro de novem disciplinis unum de architectura, P. Septimius duo. (Vitruv. 7, praef. 14)

As to the useful contributions to our subject which I found in their commentaries, many volumes have been published by the Greeks, exceedingly few by our own writers. For Fufidius curiously enough was the first to publish a volume on these topics. Further, Varro included one volume in his work On the Nine Disciplines; Publius Septimius wrote two volumes. (trans. Granger)

He has clearly read about Agatharchus’ writing,¹⁹ and it would be unfair to dismiss his testimony, as some scholars do, without investigating the possibility that his source had access through the Hellenistic libraries to information which is lost to us. Quite apart from the further writings on architectural themes which Vitruvius lists (last paragraph of the passage), we can point to early prose writings about other technical subjects—music, for example, on which the first written treatise, by Lasus of Hermione, quite certainly goes back to the sixth century bc.²⁰ The technical treatise on how to do something is one of the earliest prose genres, and some fifth-century examples of the genre—Polyclitus’ Canon, for instance, and Damon on music—won attention from intellectuals generally.

The following two paragraphs were deleted in the revised version—Ed.

⟦This is where a historian of philosophy can perhaps offer some help. For Democritus and Anaxagoras were not the only ancient philosophers to be interested in Agatharchus. The evidence of these other philosophers is seldom even mentioned, and it has certainly never been analysed, by historians of the tragic stage or of ancient painting. But it does, I shall argue, preserve a faint but valuable trace not only of the actual existence of Agatharchus’ treatise, but also of something he wrote. It will take quite a lengthy digression to analyse this evidence before I can return to Vitruvius, Democritus, and Anaxagoras. But I believe that the analysis is worth undertaking both for its intrinsic interest and because it illustrates another, equally important aspect of the unity of the Classics: unity through time. All too many scholarly footnotes use the phrase ‘a late source’ dismissively, as if to be a late source was some kind of sin, or a rather ugly disease. The truth, of course, as we all in our heart of hearts know perfectly well, is that late sources are a manifestation of the extraordinary continuity of interest and discussion, in every department of knowledge, which characterized the ancient world from the sixth century bc or before to the sixth century ad. As such, a late source is something to be understood, not something to be spurned, and understanding it involves trying to appreciate the way one period of antiquity regarded another. Let us therefore plunge into the sixth century ad.

In the sixth century ad the Neoplatonist philosopher Olympiodorus wrote a commentary on Plato’s Phaedo. Not all of what has come down to us in the manuscript under that title is by Olympiodorus, however. The part that will interest us is probably due to a somewhat earlier Neoplatonist, Damascius.²¹ In various sections of this Damascius portion are to be found what purport to be excerpts from Plutarch. In one of these sections, under the heading ‘Further proofs from Plutarch of Chaeronea’, we find Text 7, which mentions Agatharchus. And it so happens that the context in which Agatharchus is mentioned is one that we know quite a lot about, because it was set up by Strato of Lampsacus.⟧

Strato was a pupil of Theophrastus, then until c.285 bc tutor to the future king of Egypt Ptolemy Philadelphus, subsequently head of the Peripatetic school at Athens. He wrote among other things a rather cogently argued critique of Plato’s Phaedo. A number of his objections are preserved in the portion of the Olympiodorus commentary which is probably due to Damascius. Text 7 contains one of them, an objection to the Platonic theory of recollection. Text 7 also contains a reply to Strato’s objection, and it is the first part of the reply that brings in Agatharchus.

[T7] ὅτι Στράτων ἠπόρει, εἰ ἔστιν ἀνάμνησις, πῶς ἄνευ ἀποδείξεων οὐ γιγνόμεθα ἐπιστήμονες· πῶς δ ᾿ οὐδεὶς αὐλητὴς ἢ κιθαριστὴς γέγονεν ἄνευ μελέτης. ἢ μάλιστα μὲν γεγόνασί τινες αὐτοδίδακτοι· ῾ Ηράκλειτος, ὁ Αἰγύπτιος γεωργός, Φήμιος ὁ ῾ Ομήρου, ᾿Αγάθαρχος ὁ γραϕεύς. εἶτα καὶ αἱ ψυχαὶ πολλῷ τῷ κάρῳ κατεχόμεναι τῆς γενέσεως πολλῆς πρὸς ἀνάμνησιν δέονται τῆς μοχλείας· διὸ καὶ τῶν αἰσθητῶν χρῄζουσιν. ([Olymp.] In Phaed. 158. 6–12 Norvin = Plut. fr. 216g Sandbach = Strato fr. 14c Gottschalk; ὅτι … μελέτης = Strato fr. 126 Wehrli)

That Strato raised this difficulty: if ‘remembering’ is a fact, how is it that we do not become possessed of knowledge without demonstrative proof? And how is it that no-one has become a flute-player or a harp-player without practice? Or have there in fact been some self-taught men—Heraclitus, the Egyptian farmer, Homer’s Phemius, the painter Agatharchus? Then souls are overcome by much drowsiness at birth and need much therapeutic exercise if they are to recollect. And this is why they require sense-objects. (trans. Sandbach)

Strato’s objection is that it takes an awful lot of hard work to get knowledge: for theoretical knowledge you have to master proofs, for practical skills such as flute-playing you need to practise, and this (it is implied) is something that the theory of recollection cannot account for. The reply is that young souls are drowsy and need much therapeutic exercise: that is, the theory of recollection positively predicts the hard work that bothers Strato, so the objection fails. The other part of the reply (placed first) is a suggestion that there have in fact been some self-taught men, autodidaktoi, viz. Heraclitus, the Egyptian farmer, Homer’s Phemius, the painter Agatharchus. Who, we may ask, composed this list of four people who provide, as it were, empirical proof of the Platonic theory of recollection? When was the list composed? And what did Agatharchus do to merit inclusion in the gang of four?

We can, I think, exclude the possibility that the list derives directly from Damascius. This theory, which is put forward by Gottschalk in his book on Strato, seems to be ruled out by item (e) in Text 8.²²

[T8] ἐπιχειρημάτων διαϕόρων συναγωγὴ δεικνύντων ἀναμνήσεις εἶναι τὰς μαθήσεις ἐκ τῶν τοῦ Χαιρωνέως Πλουτάρχου·

(a) εἰ ἀϕ ᾿ ἑτέρου ἕτερον ἐννοοῦμεν. οὐκ ἂν εἰ μὴ προέγνωστο. τὸ ἐπιχείρημα Πλατωνικόν.

(b) εἰ προστίθεμεν τὸ ἐλλεῖπον τοῖς αἰσθητοῖς· καὶ αὐτὸ Πλατωνικόν.

(d) εἰ ἄλλοι πρὸς ἄλλο μάθημα ἐπιτηδειότεροι.

(e) εἰ πολλοὶ αὐτοδίδακτοι ὅλων τεχνῶν.([Olymp.], In Phaed. 212. 1–11 Norvin = Plut. fr. 217 Sandbach)

A collection of various arguments to show that acts of learning are acts of remembering, from Plutarch of Chaeronea:

(a) Whether we think of one thing from another. We should not unless it had been known previously. The argument is Platonic.

(b) Whether we mentally add to percepts that by which they are deficient. This too is Platonic.

(c) Whether children are quicker to learn, as being nearer to the previous existence, in which memory was retained. The argument is an obvious one.

(d) Whether men differ in their capacity for different kinds of learning.

(e) Whether many men have taught themselves complete skills.

(trans. Sandbach)

This is Plutarch fragment 217 (our previous excerpt was fragment 216), and fragment 217 is a separate set of Plutarch excerpts in the commentary; item (e) in fragment 217 indicates that the commentator found autodidaktoi cited already in the source he calls Plutarch. (This parallel is indeed what persuades Sandbach, as editor of the Plutarch fragments, to extend fragment 216 as far as (g) = our Text 7. Wyttenbach extends the fragment much further, for the commentary gives no indication of where the excerpts of fragment 216 come to a stop;²³ the asterisk of hesitation in Sandbach’s text signals that he was somewhat reluctant to go even as far as (g), but that he was impelled to do so—rightly, I am sure—by the parallel with 217(e).) So I shall take it that Damascius is not the original author of our list.

To make further progress than this elimination of just one suspect, we need to examine the list itself and try to see the relevance of these four names to the philosophical point in dispute between Strato and his Platonist opponents. Recall that the theory of recollection, as stated in the Phaedo, is that we have latent within ourselves knowledge and orthos logos:

[T9] ἑνὶ μὲν λόγῳ, ἔϕη ὁ Κέβης, καλλίστῳ, ὅτι ἐρωτώμενοι οἱ ἄνθρωποι, ἐάν τις καλῶς ἐρωτᾷ, αὐτοὶ λέγουσιν πάντα ᾗ ἔχει· καίτοι εἰ μὴ ἐτύγχανεν αὐτοῖς ἐπιστήμη ἐνοῦσα καὶ ὀρθὸς λόγος, οὐκ ἂν οἷοί τ ᾿ ἦσαν τοῦτο ποιῆσαι. (Plato, Phaedo 73 a 7–10)

‘I can give you one excellent reason,’ said Cebes. ‘When people are asked something, if the question is well put, they themselves explain everything—and yet if they hadn’t got knowledge and a right account of the matter stored away inside them, they couldn’t do that.’ (trans. Bluck, emphasis added)

Logos comes in here because, for Plato, to be able to give the correct account of something is the proof and the expression of knowledge, but we should bear in mind that the term logos can also cover speech and discourse of any kind. Such being Plato’s theory, how could our four names be thought to exemplify it?

It would be quite irrelevant to the philosophical point at issue to offer in support of Plato a list of inventors of the prōtos heuretēs type. Necessity can be the mother of invention, and was often said to be, meaning that men are driven to invent things by exposure to the uncomfortable world we live in. This counted as a point for empiricism, for the idea that knowledge comes from our interaction with the empirical world, whereas Plato held that we have knowledge already present within the soul, prior to and independently of our dealings with the empirical world. Besides, Phemius, the bard in Homer’s Odyssey, is not the inventor of song.

[T10] αὐτοδίδακτος δ᾿ εἰμί, θεὸς δέ μοι ἐν ϕρεσὶν οἴμας

παντοίας ἐνέϕυσεν· ἔοικα δέ τοι παραείδειν

ὥς τε θεῷ· τῶ μή με λιλαίεο δειροτομῆσαι. (Hom. Od. 22. 347–9)

Self-taught am I, and the god has planted in my heart

all manner of lays, and worthy am I to sing to thee

as to a god; wherefore be not eager to cut my throat. (trans. Murray)

Phemius explains himself, in Text 10, the sense in which he is autodidaktos: the paths of song are divinely implanted within him and derive from his own phrenes. Not a bad model for the Platonic theory of recollection if you allow poetic discourse to count as logos.

Similarly, Heraclitus would be horrified to be called the inventor of the logos which is the central theme of his philosophy. Heraclitus’ inclusion in the list is to be explained by Text 11, a Hellenistic interpretation of the well-known fragment (22 B 101 DK = Plut. Adv. Colot. 1118 c) in which he says edizēsamēn emeōuton, ‘I enquired of myself’:

[T11] ἤκουσέ τ᾿ οὐδενός, ἀλλ᾿ αὑτὸν ἔϕη διζήσασθαι καὶ μαθεῖν πάντα παρ᾿ ἑαυτοῦ. (D.L. 9. 5)

He was nobody’s pupil, but he declared that he ‘inquired of himself’, and learned everything from himself. (trans. Hicks)

When this is understood as meaning that he learnt the logos from himself, without becoming the pupil of any teacher, Heraclitus becomes a perfect example of someone who finds orthos logos within his own soul.

It should now be clear where this investigation is tending. With both Phemius and Heraclitus the list-maker has got his examples from written texts, and the basis for his interpretation is something said within the texts themselves. If the Egyptian farmer resembles Phemius and Heraclitus in the philosophically relevant respect, namely, he illustrates the logos within, and if the basis for including him on the list is something said in writing, there will be a strong case for viewing Agatharchus in the same light. So, who is the Egyptian farmer?

Norvin, the editor of Olympiodorus, followed by Sandbach in the Teubner Plutarch, cites a passage of Dionysius Periegetes (Text 12) according to which the Egyptians (plural) invented agriculture as well as astronomy.

[T12]	κεῖθεν δ᾿ ἐς βορέην τετανυσμένος ἄλλυδις ἄλλος,	inde ad boream prolapsus varius varias in partes
	ἑπτὰ διὰ στομάτων εἱλιγμένος εἰς ἅλα πίπτει,	septem per ostia distractus in mare devolvitur,
	ὕδασι πιαίνων λιπαρὸν πέδον Αἰγύπτοιο.	opimum Aegypti solum aquis pinguefaciens.
	οὐ γάρ τις ποταμῶν ἐναλίγκιος ἔπλετο Νείλῳ,	non enim fluviorum quisquam cum Nilo comparandus,
	οὔτ᾿ ἰλὺν βαλέειν, οὔτε χθονὸς ὄλβον ἀέξειν·	ad limum immittendum terraeque fertilitatem augendam,
	ὅς ῥά τε καὶ Λιβύην ἀποτέμνεται Ἀσίδος αἴης,	qui et Libyam dirimit a regione Asiatica,
	ἐς λίβα μὲν Λιβύην, ἐς δ᾿ αὐγὰς Ἀσίδα γαῖαν.	ita ut in libem sit Libya, in ortum vero Asia.
	τῷ πάρα ναιετάουσιν ἀριπρεπέων γένος ἀνδρῶν,	hunc insignium virorum gens accolit,
	οἳ πρῶτοι βιότοιο διεστήσαντο κελεύθους,	qui primi vitae distinxerunt ordines,
	πρῶτοι δ᾿ ἱμερόεντος ἐπειρήθησαν ἀρότρου,	primique jucundum experti sunt aratrum,
	καὶ σπόρον ἰθυτάτης ὑπὲρ αὔλακος ἁπλώσαντο,	ac sementem per directum sulcum disperserunt,
	πρῶτοι δὲ γραμμῇσι πόλον διεμετρήσαντο,	iidemque primi lineis coelum sunt emensi,
	θυμῷ ϕρασσάμενοι λοξὸν δρόμον ἠελίοιο.	quum mente obliquum solis cursum percepissent.

(Dionysius Periegetes, Orbis descriptio 225–37 Müller)
From then, stretched sprawlingly towards the north,
Through seven mouths flexed, it falls into the sea,
Making Egypt’s rich plain pregnant with its waters.
No river can be likened to the Nile
For laying mud, for bringing forth earth’s tilth.
It sunders Libya from the Asian land,
Libya westwards, Asia to the east.
A race of famous men live thereabouts,
Who first contrived the proper means to live,
Were first to make trial of the lovely plough,
And over the straight furrow broadcast seed,
And demarcated first the pole with lines,
Having discovered the sun’s slanted course.

(trans. *Lightfoot)

They suggest that the Egyptian farmer is supposed to be the inventor of agriculture.

This won’t do, and not just because the singular is essential to the philosophical argument. We have already seen that our list is not a list of inventors and would not be relevant to Plato if it was. The Egyptian farmer ought to be someone who produces some sort of logos from within himself.

Let me then tell you the Tale of the Eloquent Peasant.

There was once a man called Khunanup, who was a peasant²⁴ of the Wadi Natroun—i.e. the Valley of Salt, an oasis in the desert due south of Alexandria and north-west of Cairo. Now this peasant went down to Egypt, as the story puts it, with his donkeys laden with the goodly products of the Wadi Natroun, intending to trade his products for provisions. When he had reached the Nile and was journeying along it, he was seen by a man called Djehutinakhte, who thought it would be a good idea to grab the donkeys. Having conceived this dastardly plan, Djehutinakhte spread a cloth all the way across the narrow riverside path along which Khunanup was coming. On the one side, then, the river Nile, on the other, a field of barley belonging to Djehutinakhte. The peasant had no choice, if he was to avoid the cloth, but to take his donkeys up along the bank of the field. At once Djehutinakhte accused him of trampling his crops, an altercation broke out, and while the two men were arguing one of the donkeys ate a wisp of barley. ‘Aha’, said Djehutinakhte, ‘I shall confiscate your donkey for eating my barley’, and he did.

For the next ten days the peasant pleads with Djehutinakhte, but in vain. So he goes to the High Steward Rensi, to appeal to him. Now when Rensi hears the peasant’s appeal he goes to the King to tell him of the quite extraordinary eloquence this uneducated peasant has displayed. ‘Then keep him talking’, replies the King, ‘and have someone write it all down for me.’ Nine times the peasant comes before the High Steward and makes his appeal: nine different speeches, each of amazing Gorgianic grandiloquence. Nine times his words are written down for the King’s delight—and the peasant is sent away without an answer. Finally, when the King has a papyrus roll full of marvellous speeches, Rensi is permitted to give judgement: Djehutinakhte’s house and goods are confiscated and given to the peasant Khunanup. End of story.²⁵

The Tale of the Eloquent Peasant was popular in Egypt during the Middle Kingdom, which lasted from 2040 to the middle of the seventeenth century bc. The surviving manuscripts are XIIth—XIIIth Dynasty, i.e. from a period between about 1991 and 1633 bc. The only other Egyptian trace of the story is an inaccurate quotation on a Ramesside ostracon about five hundred years later,²⁶ which still leaves a gap of some eight hundred years until we reach Strato or twelve hundred years to Plutarch, who was born around ad 46. For an Egyptian story these Greek philosophers are a very late source indeed. But Egyptian stories were translated into Greek, and in one case, the ‘Legend of Tefnut’, we have the text in both languages. And I trust you will agree with me that Khunanup is a splendid example of a man whose rhetorical logoi come from within in a manner perfectly designed to illustrate the Platonic theory of recollection as that theory is understood in the debate between Strato and his opponent. In fact, the example is so splendid, the parallel with Phemius and Heraclitus is so close, that there can, I submit, be no doubt that we have now identified the person who is referred to as the Egyptian farmer in a collection of truncated excerpts from Plutarch which has come down to us in what purports to be a commentary on Plato’s Phaedo written by Olympiodorus in the sixth century ad, over a millennium and a half later than the last known trace of the Tale itself. Moreover, like Phemius and Heraclitus, the Egyptian farmer earns his place in the gang of four in virtue of a written text and what is said in that text. The only remaining question is: which Greek got to know, or to know about, that text, and when?

The obvious candidate is Plutarch himself. He spent time in Egypt,²⁷ his work on Isis and Osiris shows him to have been an Egyptologist of sorts, and he was, of course, a Platonist who is independently known to have worked seriously at defending the theory of recollection. But Strato too spent time in Alexandria, in his tutoring days, and we cannot exclude the possibility that Text 7 derives entirely from Strato, weighing up the arguments for the theory of recollection as well as the arguments against, this being the proper Peripatetic way to approach a philosophical problem. And there are any number of other possibilities. But even if we have to suspend judgement on the ultimate origin of our list, it is enough for my purposes that we can at least be confident that the first three names are, all of them, examples of the logos within. We can see, moreover, that each example illustrates a distinct genre of logos: Phemius represents poetry, Heraclitus philosophy, Khunanup rhetoric. It becomes overwhelmingly tempting to extrapolate to Agatharchus as follows:

(1) He represents technical logoi of the genre from which Vitruvius lists so many specimens.

(2) Agatharchus’ logos or hupomnēma on stage construction could be read, or more probably read about, in the libraries of the Hellenistic period.

(3) The list-maker’s evidence for considering Agatharchus to be self-taught in the peculiar sense signified by a Platonist’s use of the term autodidaktos was of the same type as his evidence in the other three cases, namely, something said in the text in question. So the next question is whether we can conjecture what that written something was that the list-maker took as evidence for Agatharchus being autodidaktos.

I suggest that it was the opening words of Agatharchus’ treatise. Hecataeus of Miletus began a very early genealogical treatise with the sentence ‘I write these things, as they seem to me to be true; for the logoi of the Greeks, in my view, are both many and absurd.’²⁸ Heraclitus a little later starts his logos by telling his readers that they haven’t a hope of understanding it. Several Hippocratic treatises open with a similarly aggressive bid for the reader’s attention. Suppose, then, that Agatharchus started by proclaiming his originality as a writer in some such terms as this: I am the first to write on these matters, I have had neither teacher nor predecessor in writing about the things I am going to tell you about. Hellenistic compilations love to quote the opening words of earlier books—the quotation from Metrodorus in Text 1 is a case in point, for we know from other citations that these were his opening words—and if Agatharchus did trumpet his originality as an author in the manner I have suggested, we have the perfect explanation for Vitruvius thinking he knows that Agatharchus was the first to write on an architectural theme. He knows this because the compiler he is following took Agatharchus’ own word for it.

The situation is actually a little more complicated than this, and complicated in a way that tends to confirm my hypothesis. Agatharchus could well have claimed, and been recorded by some compiler as claiming, to be the first to write about the stage-building. For if Agatharchus wrote within the lifetime of Aeschylus, the skēnē or stage-building had not long been in existence for anyone to write about. But if Agatharchus presented himself in the way Vitruvius presents him, as the first to write on any architectural matter at all, his claim is in conflict with the testimony of Vitruvius himself in the next paragraph of our passage. About Silenus we know nothing but what is recorded here, but the temples and authors next mentioned can be dated independently of Vitruvius to the sixth century bc. I very much doubt one can get around this conflict by giving primum … postea a non-temporal classificatory sense, for a few pages later, in Text 13, when the temple at Ephesus is mentioned again, the coupling of primum and postea is clearly temporal, as indeed it is taken to be by all translators of our main passage.

[T13] primumque aedes Ephesi Dianae ionico genere ab Chersiphrone Gnosio et filio eius Metagene est instituta, quam postea Demetrius, ipsius Dianae servos, et Paeonius Ephesius dicuntur perfecisse. (Vitruv. 7, praef. 16)

First of all the temple of Diana at Ephesus was planned in the Ionic style by Chersiphron of Cnossus and his son Metagenes; afterwards Demetrius, a temple warden of Diana, and Paeonius of Ephesus are said to have completed it. (trans. Granger)

Consequently, we must look for an explanation which accepts that Vitruvius has made a mistake. I propose that the most likely explanation is this: that Vitruvius or his compiler was hazy about the chronology of archaic Greek temples and overlooked the restricted scope of Agatharchus’ original claim when incorporating it into the wider context of a history of writings on architecture generally. Anyone who deals regularly with ancient doxography will know similar cases. Viewed in this light, Vitruvius’ inadvertent slip strongly suggests that Agatharchus was recorded as claiming originality as an author.

It is even possible that Agatharchus expressed his claim to originality with the word autodidaktos itself. The word is not common. Its only extant occurrence in early literature outside the Odyssey (Text 10) is in the Agamemnon of Aeschylus (Text 14).

[T14] τὸν δ᾿ ἄνευ λύρας ὅμως ὑμνῳδεῖ

θρῆνον ᾿ Ερινύος αὐτοδίδακτος ἔσωθεν

θυμός …

(Aesch. Ag. 990–2)

The thumos sings, without a lyre,

from within, self-taught,

the song of Erinys …

(trans. Burnyeat)

I hesitate to build one conjecture on another, but after all that Oliver Taplin²⁹ has done to make good Wilamowitz’s theory that the Oresteia of 458 bc is where Aeschylus first makes use of the recently introduced stage-building, and first exploits the dramatic resonance of the idea of the house which the skēnē represents, it would certainly seem to be a most appropriate occasion for the stage-designer to write a book about—a book telling of his own contribution to that memorable occasion and perhaps—I say again, perhaps—borrowing from his playwright the poetic word autodidaktos, with its associations of divine inspiration, in order to proclaim that technical literature too, not just poetry, can be divinely inspired.³⁰ For we can be sure that if all of that did happen, Agatharchus would have virtually written his own certificate of worthiness to be included, along with Phemius, Heraclitus, and the Egyptian farmer, in the list we have been examining. All it would take for him to be put on the list is that the Hellenistic compilers should preserve his opening words as they preserved the opening words of countless other authors.

I cannot, of course, prove that the Oresteia is the tragoedia of Aeschylus to which Vitruvius refers, as those scholars who trust Vitruvius have regularly assumed, but the main rival theory is Rumpf’s influential suggestion—adopted by Webster and Taplin among historians of the stage³¹ and by Robertson and Pollitt among historians of art³²—that the phrase Aeschylo docente does not mean what the words plainly say, but refers to a revival of an Aeschylean drama around 430, well after Aeschylus’ death in 456/5.³³ This late dating is motivated in part by the thought that 458 would be too early for Agatharchus to invent a sophisticated system of perspective. But we have already seen that it is not in fact as the inventor of perspectives but as a writer that Agatharchus comes into Vitruvius’ story. And it is important to realize that Rumpf’s only evidence for postponing Agatharchus’ activity until around 430 is Text 15,³⁴ a story originating with some fourth-century orators—who tell it all too plainly as a story—in which Alcibiades locks Agatharchus in his house to make sure he finishes the painting, or alternatively—and this is the version Rumpf relies on—he locks him up because he has found the painter making love to his mistress.

[T15] εἷρξεν Ἀγάθαρχον τὸν γραϕέα· καὶ γὰρ ταῦτα λέγουσιν. λαβών γέ τι πλημμελοῦνθ᾿, ὥς ϕασιν, ὅπερ οὐδ᾿ ὀνειδίζειν ἄξιον. (Dem. In Mid. 147)

Another story is that he imprisoned the painter Agatharchus. Yes, but he had caught him in an act of trespass, or so we are told; so that it is unfair to blame him for that. (trans. Vince, emphasis added)

εἷρξεν ᾿Αγάθαρχον] ζωγράϕος οὗτος ὢν ἐϕωράθη συνὼν τῇ παλλακίδι τοῦ Ἀλκιβιάδου, ὃν λαβὼν καθεῖρξε καὶ κατὰ τοῦτο ἥμαρτε μὴ παραδοὺς τοῖς νόμοις ἀπεσιώπησε δὲ τὸ ἁμάρτημα ὡς μικρόν, ἵνα μεῖζόν τι ὑποπτεύσωσιν οἱ δικάζοντες. τὸ ἀμϕισβητούμενον οὖν ὡς ὁμολογούμενον ἔθηκεν. ἔχει γὰρ αὑτῷ συναγωνιζομένην τὴν κατάγνωσιν τοῦ δήμου. τινὲς δὲ τὸ “εἷρξεν” οἷον “ἐξέωσε τοῦ θεάτρου ὡς ἀσελγῆ”. διὸ καὶ εἶπεν “ὅπερ οὐδὲ ὀνειδίζειν ἄξιον”. (Scholium to the above, 21. 506 Dilts)

‘He locked up Agatharchus.’ This painter was caught having sex with the concubine of Alcibiades, who seized and confined him, thereby doing wrong by not turning him over to the law. [Alcibiades] concealed how minor the wrong was, so that the jurors would suspect something greater. He thus stipulated something that was in dispute as though it were agreed upon, since he had on his side the censure of the people. Some take ‘locked up’ to mean something like ‘he threw him out of the theatre on the grounds that he was behaving outrageously’, for which reason [Demosthenes] adds ‘which is not even worth criticizing’. (trans. Ed.)

ὃς εἰς τοσοῦτον ἐλήλυθε τόλμης, ὥστε πείσας Ἀγάθαρχον τὸν γραϕέα συνεισελθεῖν [οἴκαδε] τὴν οἰκίαν ἐπηνάγκασε γράϕειν, δεομένου δὲ καὶ προϕάσεις ἀληθεῖς λέγοντος, ὡς οὐκ ἂν δύναιτο ταῦτα πράττειν ἤδη διὰ τὸ συγγραϕὰς ἔχειν παρ᾿ ἑτέρων, προεῖπεν αὐτῷ δήσειν, εἰ μὴ πάνυ ταχέως γράϕοι. ὅπερ ἐποίησε· καὶ οὐ πρότερον ἀπηλλάγη, πρὶν ἀποδρὰς ᾤχετο τετάρτῳ μηνί, τοὺς ϕύλακας λαθών, ὥσπερ παρὰ βασιλέως. ([Andoc.] In Alc. 17)³⁵

Why, there are no limits to his impudence. He persuaded Agatharchus, the artist, to accompany him home, and then forced him to paint; and when Agatharchus appealed to him, stating with perfect truth that he could not oblige him at the moment because he had other engagements, Alcibiades threatened him with imprisonment, unless he started painting straight away. And he carried out his threat. Agatharchus only made his escape three months later, by slipping past his guards and running away as he might have done from the king of Persia. (trans. Maidment)

If we take this anecdote seriously—and I suggest that we are in fact dealing merely with the echo of a comic plot—Agatharchus must be alive and well in the decade after 430, when Alcibiades emerges into the limelight. Fine, but this does not even begin to show that Agatharchus could not have worked for Aeschylus in 458—at the age of thirty, say—unless you believe that a man of, say, sixty plus would not be capable of the part assigned to him in the story. I trust that you will agree that Rumpf’s argument does not stand up.

Besides, Anaxagoras probably died in 428/7, far away in Lampsacus, after residing there long enough to inspire the Lampsacenes to pay him extensive honours at his death. Anaxagoras’ response to Agatharchus’ book, once we take that seriously, favours an earlier rather than a later date for Agatharchus’ writing.

Let us then turn to Anaxagoras. What could Agatharchus have said that might interest him? It was not the opening words of his book, Vitruvius implies (in Text 5), but something about which Democritus and Anaxagoras also wrote (de eadem re scripserunt). Vitruvius proceeds to identify the topic on which all three writers had something to say, in a long complex sentence which needs to be divided into two parts. The part which I have indented is the explanation of the phenomenon described from uti (‘so that’) onwards.

namque primum Agatharchus Athenis Aeschylo docente tragoediam scaenam fecit, et de ea commentarium reliquit. Ex eo moniti Democritus et Anaxagoras de eadem re scripserunt,

quemadmodum oporteat, ad aciem oculorum radiorumque extentionem certo loco centro constituto, lineas ratione naturali respondere,

uti … (Vitruv. 7, praef. 11)

In the first place Agatharchus, in Athens, when Aeschylus was bringing out a tragedy, painted a scene, and left a commentary about it. This led Democritus and Anaxagoras to write on the same subject,

showing how, given a centre in a definite place, the lines should naturally correspond with due regard to the point of sight and the divergence of the visual rays,

so that … (trans. Morgan)

The indented explanation is extraordinarily sophisticated. Thanks to John White’s careful analysis,³⁶ we are now able to regard it as one of just two ancient texts—the other being Text 16, again from Vitruvius—to formulate explicitly, even if somewhat clumsily, the idea of a centralized system of pictorial perspective complete with a vanishing point:

[T16] item scaenographia est frontis et laterum abscedentium adumbratio ad circinique centrum omnium linearum responsus. (Vitruv. 1. 2. 2)

In like manner, scenography is the sketching of the front and of the retreating sides and the correspondence of all the lines to the point of the compasses. (trans. White)

As such, it may have been known to Alberti and may have played its part in the Renaissance rediscovery of perspective. But by the same token no one doubts—no one sensible at least³⁷—that no such idea could have been formulated in a fifth-century treatise, by Agatharchus or by Anaxagoras. It is well in advance of Euclid’s Optics of around 300 bc. It must have been inserted into the doxographical tradition some time between 300 and the compilation Vitruvius is using, with the result that the fifth-century writers’ observations are served up, in a manner which will again be familiar to those who have worked regularly with doxography, in the illuminating light of the most modern explanation.

Accordingly, to discover what Anaxagoras or Democritus learnt from Agatharchus, we must turn to the second half of the sentence, the bit after the indentation (in Text 5), beginning with uti or ‘so that’.

uti de incerta re certae imagines aedificiorum in scaenarum picturis redderent speciem et, quae in directis planisque frontibus sint figurata, alia abscedentia, alia prominentia esse videantur. (Vitruv. 7, praef. 11)

so that by this deception a faithful representation of the appearance of buildings might be given in painted scenery, and so that, though all is drawn on a vertical flat façade, some parts may seem to be withdrawing into the background, and others to be standing out in front. (trans. Morgan)

At once we meet a problem of translation again, compounded by a problem about the text: should we read certae or incertae imagines? I propose that we tackle the translation problem first, for the moment just leaving the adjective (in)certae out of the reckoning.

That there is a problem of translation is evident from Morgan’s absurd rendering of the Latin de incerta re as ‘by this deception’.³⁸ Notice also the dots of despair in Erika Simon’s version quoted in my English apparatus (n. 13). But the rival translations in the apparatus have a dangling ‘from an uncertain object’ which is equally puzzling. For example, White translates ‘so that from an uncertain object certain images may render the appearance of buildings’ (see again n. 13). What uncertain object do the images come from? Surely they don’t come from anywhere; they are just there, painted on the wall of the stage-building.

The difficulty arises, I suggest, from the assumption that de means ‘from’. Given this assumption, de incerta re must depend upon imagines and then aedificiorum is left to depend on speciem. But if de could mean ‘concerning’ or ‘about’, as it did just a bit earlier in the phrase de eadem re, and as Erich Frank alone is prepared to take it (see n. 13), then aedificiorum could be detached from speciem and made to depend on imagines, which is after all the word it is adjacent to. We would then have this: ‘concerning an unclear object, the images of buildings in stage-paintings give a view’. De incerta re would then have an introductory function, and would be connected in sense, though not in grammar, with speciem instead of imagines.

Before asking you to accept this translation, I want to ask you to try putting it into Greek. Alberti remarked that Vitruvius is ‘Greek to the Romans and Roman to the Greeks’,³⁹ and we saw in Text 6 that Vitruvius is using Greek compilations as well as Latin ones; in any case the Latin doxographers will themselves have been borrowing from Greek predecessors. So what would our sentence have looked like in Greek?

Well, in Cicero’s philosophical writings he commonly uses incertus where we know his Greek source would have had adēlos, ‘nonevident’. There’s a good example in Text 17:

[T17] interdum enim cum adhibemus ad eos orationem eius modi, si ea quae disputentur vera sint, tum omnia fore incerta, respondent: ‘quid ergo istud ad nos? num nostra culpa est? naturam accusa, quae in profundo veritatem, ut ait Democritus, penitus abstruserit.’ alii autem elegantius, qui etiam queruntur quod eos insimulemus omnia incerta dicere, quantumque intersit inter incertum et id quod percipi non possit docere conantur eaque distinguere. cum his igitur agamus qui haec distinguunt, illos qui omnia sic incerta dicunt ut stellarum numerus par an impar sit quasi desperatos aliquos relinquamus. (Cic. Acad. 2. 32)

For sometimes when we address them in this sort of language, ‘If your contentions are true, then everything will be uncertain’, they reply, ‘Well, what has that to do with us? surely it is not our fault; blame nature for having hidden truth quite away, in an abyss, as Democritus says.’ But others make a more elaborate answer, and actually complain because we charge them with saying that everything is uncertain, and they try to explain the difference between what is uncertain and what cannot be grasped, and to distinguish between them. Let us therefore deal with those who make this distinction, and leave on one side as a hopeless sort of persons the others who say that all things are as uncertain as whether the number of the stars is odd or even. (trans Rackham, emphasis added)

—and species is the Latin for opsis. With two such clues it is hardly necessary to add Text 18 to show that imago can render phantasia, one of several Greek nouns connected with the verb phainesthai.

[T18] quare capiendae sunt illae, de quibus dixi, rerum imagines, quas vocari ϕαντασίας indicavimus, omniaque, de quibus dicturi erimus, personae, quaestiones, spes, metus habenda in oculis, in adfectus recipienda. (Quint. 10. 7. 15)

Consequently those vivid conceptions of which I spoke and which, as I remarked, are called ϕαντασίαι, together with everything that we intend to say, the persons and questions involved, and the hopes and fears to which they give rise, must be kept clearly before our eyes and admitted to our hearts. (trans. Butler, emphasis added)

quas ϕαντασίας Graeci vocant, nos sane visiones appellemus, per quas imagines rerum absentium ita repraesentantur animo, ut eas cernere oculis ac praesentes habere videamur. has quisquis bene conceperit, is erit in adfectibus potentissimus. hunc quidam dicunt εὐϕαντασίωτον, qui sibi res, voces, actus secundum verum optime finget; quod quidem nobis volentibus facile continget. (Quint. 6. 2. 29–30)

There are certain experiences which the Greeks call ϕαντασίαι, and the Romans visions, whereby things absent are presented to our imagination with such extreme vividness that they seem actually to be before our very eyes. It is the man who is really sensitive to such impressions who will have the greatest power over the emotions. Some writers describe the possessor of this power of vivid imagination, whereby things, words and actions are presented in the most realistic manner, by the Greek word εὐϕαντασίωτος; and it is a power which all may readily acquire if they will. (trans. Butler)

Hidden for all to see in Vitruvius’ Latin is a close relative of that famous dictum of Anaxagoras, opsis tōn adēlōn ta phainomena, ‘Appearances are a sight of things unseen’ (59 B 21a DK).

Now look at Text 19, paragraph 140, but for which we would never have known that Anaxagoras said that famous thing. (And by the way, the angled brackets in the Greek text are misleading: the words opsis tōn adēlōn ta phainomena are not an insertion conjectured by the editor, but an omission in manuscripts other than N.)

[T19] καὶ δὴ ἐν μὲν τούτοις πᾶσαν σχεδὸν κινεῖ κατάληψιν, εἰ καὶ μόνων ἐξαιρέτως

138 καθάπτεται τῶν αἰσθήσεων· ἐν δὲ τοῖς κανόσι δύο ϕησὶν εἶναι γνώσεις, τὴν μὲν διὰ τῶν αἰσθήσεων τὴν δὲ διὰ τῆς διανοίας, ὧν τὴν μὲν διὰ τῆς διανοίας γνησίην καλεῖ, προσμαρτυρῶν αὐτῇ τὸ πιστὸν εἰς ἀληθείας κρίσιν, τὴν δὲ διὰ τῶν αἰσθήσεων σκοτίην ὀνομάζει, ἀϕαιρούμενος αὐτῆς

139 τὸ πρὸς διάγνωσιν τοῦ ἀληθοῦς ἀπλανές. λέγει δὲ κατὰ λέξιν “γνώμης δὲ δύο εἰσὶν ἰδέαι, ἡ μὲν γνησίη ἡ δὲ σκοτίη· καὶ σκοτίης μὲν τάδε σύμπαντα, ὄψις ἀκοὴ ὀδμὴ γεῦσις ψαῦσις, ἡ δὲ γνησίη, ἀποκεκριμένη δὲ ταύτης.” εἶτα προκρίνων τῆς σκοτίης τὴν γνησίην ἐπιϕέρει λέγων “ὅταν ἡ σκοτίη μηκέτι δύναται μήτε ὁρῆν ἐπ᾿ ἔλαττον μήτε ἀκούειν μήτε ὀδμᾶσθαι μήτε γεύεσθαι μήτε ἐν τῇ ψαύσει αἰσθάνεσθαι, ἀλλ᾿ ἐπὶ λεπτότερον.” οὐκοῦν καὶ κατὰ τοῦτον ὁ λόγος ἐστὶ κριτήριον, ὃν γνησίην γνώμην καλεῖ.

140 Διότιμος δὲ τρία κατ᾿ αὐτὸν ἔλεγεν εἶναι κριτήρια, τῆς μὲν τῶν ἀδήλων καταλήψεως τὰ ϕαινόμενα, 〈ὄψις γὰρ τῶν ἀδήλων τὰ ϕαινόμενα,〉 ὡς ϕησὶν ᾿Αναξαγόρας, ὃν ἐπὶ τούτῳ Δημόκριτος ἐπαινεῖ, ζητήσεως δὲ τὴν ἔννοιαν (περὶ παντὸς γάρ, ὦ παῖ, μία ἀρχὴ τὸ εἰδέναι περὶ ὅτου ἔστιν ἡ ζήτησις), αἱρέσεως δὲ καὶ ϕυγῆς τὰ πάθη· … (S.E. M. 7. 137–40)

Now in these passages he almost rejects apprehension altogether, although

138 it is the senses only that he specially attacks. But in his ‘Canons’ he says that there are two kinds of knowledge, one by means of the senses, the other by means of the intelligence; and of these he calls that by means of the intelligence ‘genuine,’ ascribing to it trustworthiness in the judgement of truth, but that by means of the senses he terms ‘bastard,’ denying it inerrancy in the distinguishing

139 of what is true. He expressly declares—‘Of knowledge there are two forms, the genuine and the bastard; and to the bastard belong all these—sight, hearing, smell, taste, touch; but the other form is distinct from this and genuine.’ Then, while thus preferring the genuine to the bastard, he proceeds: ‘Whenever the bastard kind is unable any longer to see what has become too small, or to hear or smell or taste or perceive it by touch, 〈one must have recourse to〉 another and finer 〈instrument〉.’ Thus, according to this man also, reason is the

140 criterion, and he calls it ‘genuine knowledge’. But Diotimus used to say that according to Democritus there are three criteria—namely, the criterion of the apprehension of things non-evident, which is the things apparent; for, as Anaxagoras says (and Democritus commends him for it), the things apparent are the vision of the things non-evident; and the criterion of investigation, which is the conception—‘for in every case, my child, the one starting-point is to know what the subject of investigation is’ [cf. Plato, Phaedr. 237 b]; and the criterion of choice and aversion, which is the affections (trans. Bury, emphasis added)

The anti-chronological order of the names in Vitruvius, first Democritus and then Anaxagoras, becomes understandable when we learn that Democritus commended Anaxagoras for the famous dictum. We learn this on the word of one Diotimus, who is independently attested in Text 20 as an expositor of Democritus.

[T20] ἔτι πρὸς τούτοις [sc. Democritus, Hecataeus, Apollodorus, Nausiphanes] Διότιμος τὴν παντέλειαν τῶν ἀγαθῶν ἣν εὐεστὼ προσαγορεύεσθαι [68 B 4. 140], τέλος ἀπέϕηνεν. (Diotimus 76 A 2 DK = Clem. Strom. 2. 130. 6 Stählin)

In addition to them [sc. Democritus, Hecataeus, Apollodorus, Nausiphanes], Diotimus declared the goal to be a full complement of the things that are good, which he called well-being. (trans. *Ferguson, slightly modified)

The chances are that Diotimus got this information by studying the works of Democritus himself: hence the order of names. We do not know Diotimus’ date, but he uses the language of Hellenistic philosophy and quotes Plato’s Phaedrus. He writes about Democritus more as a historian than as a follower—he may in fact be identical with a second-century bc Stoic Diotimus—and he comes into Aenesidemus’ history as the author of an empiricist interpretation of Democritus which diverges from the more rationalistic interpretation of the preceding paragraphs 138–9, which give the view of Aenesidemus’ main authority for Democritus, who can, I believe, be shown to be Posidonius. But if the best we can do for Diotimus is to say that he is a Hellenistic scholar writing before Posidonius, that is enough to show that there is information in the Hellenistic libraries about Democritus responding to Anaxagoras on the issue that concerns us, information which is hard to explain except on the supposition that it comes ultimately from Democritus himself. The important thing is not to trace routes between Vitruvius and Democritus, but to satisfy ourselves that one or more routes are available. For I hope it is now looking very plausible indeed that what Agatharchus said to interest Anaxagoras and Democritus was closely connected with the philosophical dictum ‘Appearances are a sight of things unseen’.

The most economical hypothesis is simply this. Agatharchus’ treatise included, among all the other topics to do with stage construction, a section on scene-painting. (We may even be tempted to wonder if it wasn’t this section that told Aristotle that scene-painting had recently been introduced by Sophocles; at any rate the possibility cannot now be excluded.) In this section, be it long or short, one thing Agatharchus could have said was—well, more or less what we find in Vitruvius: concerning what is not evident to the audience, you can give them a view, i.e. let them as it were ‘see’ it, by painting images of buildings on the stage wall, and doing it in such a way that although all is drawn on a vertical flat façade, some parts seem to be receding into the background and others to be standing out in front. Painted scenery gives the audience a sight of the unseen.

That would be an exciting and relevant contribution, from the technician’s point of view, to all the play with the seen and the unseen that goes on in the Oresteia and subsequent tragedies. Yet it does not presuppose more than the simplest and most primitive indications of recession in the painting itself. Four columns under a roof, with the inner columns smaller than the outer ones, would be quite enough to give rise to the written boast which is what, as we have seen, inspired Anaxagoras. As for the actual scene-painting that Agatharchus is boasting about, we should not look ahead, as the indented explanation inserted by Vitruvius’ compiler tells us to do, but back—back to the lost mural paintings of the early fifth century. These can only be studied through descriptions in Pausanias and the distorting mirror of vase-paintings, but the evidence leads art historians such as John Barron to talk of ‘multi-level compositions’,⁴⁰ and Martin Robertson of the beginning of the idea of pictorial space, of the picture as ‘a window opening on a feigned world’.⁴¹ In so speaking they are referring to murals by Polygnotus and Mikon painted in the 470s—well before the Oresteia in 458. Agatharchus, then, was not the first Greek painter to start exploring the third dimension; but he could well, compatibly with the art-historical evidence, be the first to boast in writing about its application to the stage.

We can now solve the textual problem which we left aside earlier. Granger, as you see from the textual notes to Text 5 (n. 9), reads incertae on the grounds that it is required by Democritean doctrines. But even supposing he interprets Democritus’ views about images correctly, it is not Democritus’ view that matters here but Agatharchus’. Agatharchus would certainly want his painted columns to stand out bright and clear for the audience to see. That, then, is how he would describe them in the book. Alternatively, by Hellenistic times opsis tōn adēlōn ta phainomena has become a common tag which is quoted all over the place, and Hellenistic phainomena are invariably enargē or dēla, clear (see Text 21), never adēla.

[T21] ἔτι τῶν ὄντων τὰ μέν ἐστι πρόδηλα τὰ δὲ ἄδηλα, ὡς αὐτοί ϕασιν, καὶ σημαίνοντα μὲν τὰ ϕαινόμενα, σημαινόμενα δὲ ὑπὸ τῶν ϕαινομένων τὰ ἄδηλα· ὄψις γὰρ κατ ᾿ αὐτοὺς τῶν ἀδήλων τὰ ϕαινόμενα. (S.E. PH 1. 138)

Further, some existing things are ‘pre-evident’, as they say, others non-evident; and the apparent things are significant, but the non-evident signified by the apparent; for according to them ‘the things apparent are the vision of the non-evident’. (trans. Bury, emphasis added)

Thus whether the adjective stems from Agatharchus or the Hellenistic compiler, certae is the reading we want. Granger’s mistaken choice, which he actually uses as evidence for his reconstruction of the history of the Vitruvius manuscripts, is just one of the eccentricities which mar his widely used edition of an important author.

I have laboured long—I hope not too wearisomely—to establish, with varying degrees of probability, the existence, the date, the gist of the opening words, and the gist of one other fragment, of a written work on stage construction by Agatharchus. The historian can seldom supply proof beyond reasonable doubt. All I have aimed for is to win on the balance of a set of interconnected probabilities. If I have succeeded, the gain for the history of philosophy is considerable. We have a new and richer context for Anaxagoras’ famous dictum ‘Appearances are a sight of things unseen’. A context which involves reference to the powers of scene-painting.

I think we can see at once why Agatharchus’ remark about scene-painting would have caught the attention of Anaxagoras. The central postulate of Anaxagorean physics is that there is no break between the macroscopic world of ordinary observation and the microscopic world which explains it. Sensible qualities such as colour do not cease to exist below the threshold of perception: they continue all the way down in the sense that any portion of matter, however small, has colour, even if our eyes are too weak to see it:

[T22] ἔνθεν ὁ μὲν ϕυσικώτατος ᾿Αναξαγόρας ὡς ἀσθενεῖς διαβάλλων τὰς αἰσθήσεις “ὑπὸ ἀϕαυρότητος αὐτῶν” ϕησὶν “οὐ δυνατοί ἐσμεν κρίνειν τἀληθές.” τίθησί τε πίστιν αὐτῶν τῆς ἀπιστίας τὴν παρὰ μικρὸν τῶν χρωμάτων ἐξαλλαγήν· εἰ γὰρ δύο λάβοιμεν χρώματα, μέλαν καὶ λευκόν, εἶτα ἐκ θατέρου εἰς θάτερον κατὰ σταγόνα παρεκχέοιμεν, οὐ δυνήσεται ἡ ὄψις διακρίνειν τὰς παρὰ μικρὸν μεταβολάς, καίπερ πρὸς τὴν ϕύσιν ὑποκειμένας. (S.E. M. 7. 90)

Hence the greatest of the Physicists, Anaxagoras, in disparaging the senses on the ground of their weakness, says, ‘Owing to their infirmity we are unable to judge what is true.’ And as an assurance of their lack of sureness he alleges the gradual change in colours: for if we were to take two colours, black and white, and pour some of the one into the other drop by drop, our sense of sight will be unable to distinguish the gradual alterations although they subsist as actual facts. (trans. Bury)

Anaxagoras goes further still and maintains the puzzling thesis that any portion of matter, however small, contains every colour and every other quality as well. What we perceive, in those portions of matter which are large enough for us to perceive them, is the colours and qualities which predominate and which, because they predominate, are the qualities most evident to us:

[T23] ἕτερον δὲ οὐδέν ἐστιν ὅμοιον οὐδενί, ἀλλ᾿ ὅτων πλεῖστα ἔνι, ταῦτα ἐνδηλότατα ἓν ἕκαστόν ἐστι καὶ ἦν. (Anaxagoras 59 B 12 DK = Simpl. In Phys. 157. 3–4 Diels)

nothing else is like anything else, but each single body is and was most plainly those things of which it contains most. (trans. KRS 476)

To use a theatrical metaphor inspired by Lucretius’ account of Anaxagoras in Text 24 (the last line), the qualities we perceive in things are those that are ‘out front’.

[T24] linquitur hic quaedam latitandi copia tenvis,

id quod Anaxagoras sibi sumit, ut omnibus omnis

res putet inmixtas rebus latitare, sed illud

apparere unum cuius sint plurima mixta

et magis in promptu primaque in fronte locata. (Lucr. 1. 876–80)

Herein there is left a slight chance of hiding from justice,

which Anaxagoras grasps for himself, to hold that all things

are mingled, though in hiding, in all things, but that that one

thing comes out clear, whereof there are most parts mingled in,

stationed more ready to view and in the forefront.

(trans. Bailey, emphasis added)

Thus for Anaxagoras the world as we observe it is very like a scene-painting such as Agatharchus describes. We cannot actually see all the parts as they recede into the microscopic background, but a simple experiment with a black and a white paint pot (as in Text 22) will show that the recession we do see implies further recessions that we don’t see. Concerning what is not evident, the bits of paint you do see give you a view, as Agatharchus said, provided—and this is the point that will, I hope, convince the art historians—provided the perspectival effect alluded to is rather primitive. The marvellous perspectival achievements from Vitruvius’ own day in the House of Augustus (see Figure 1), or even in Boscoreale, would be much less suitable illustrations for the philosopher’s message. For the Roman paintings leave too little to the visual imagination. We see, in fact, and see with perfect clarity, all there is to be seen from where we stand in relation to the house represented in the painting. We are not challenged to complete the picture in our mind’s eye, for visually speaking it is already complete.

Fig. 1. House of Augustus, Room of the Perspectival Wall Paintings (Room 11)

Drawing by L. Pannuti, from Gianfilippo Carettoni, ‘La decorazione pittorica della Casa di Augusto sul Palatino’, Mitteilungen des Deutschen archäologischen Instituts, Römische Abteilung, 90/2 (1983), 373–419, fig. 4 (pp. 386–7).

I am suggesting, then, that the first philosopher to compare the observed world to a scene-painting was in fact Anaxagoras. And indeed what better model could he have found for his peculiar physics than the front of the skēnē or stage-building, that great centre of Athenian attention inside which the actors, unseen, changed their clothes and masks before coming out front to create a world in which who you are depends on which mask and costume is the predominating factor in your appearance? The front of the skēnē is just a wooden wall. But it is a wall rich in implications for a physicist to ponder.

Let us now imagine Democritus pondering this model, as he found it in Anaxagoras’ book. For if scene-painting was to Anaxagoras just a model for his physics, it becomes quite unnecessary to postulate, as some optimistic scholars have done, a treatise by Anaxagoras on perspective over and above the work On Nature which could be bought in the marketplace at Athens for a drachma.⁴² Not only is it unnecessary, but it is contrary to the explicit statement of D.L. 1. 16 numbering Anaxagoras among those philosophers who left just one book. More interesting is a tradition that Anaxagoras was the first to publish a book with diagrams.⁴³ But to return to Democritus. The great disagreement between Democritus and Anaxagoras concerns precisely the question whether the microscopic world is continuous with the macroscopic. Democritus says it is not. The atoms have no colour. As you cut smaller, nearly all the sensible qualities of ordinary observation disappear, until you are left just with atoms and void, whose properties must be determined by reason, not by extrapolating from information provided by the senses.

This rationalist, anti-empirical epistemology may surprise the modem reader, who is used to the idea that good scientific theories—and we all think of Atomism as one of the very best theories that antiquity produced—good scientific theories gain their support, even if indirectly, from the evidence of observation. But there is plenty of evidence that Democritean Atomism was based on a priori reasoning, not on observation, and the quotations in paragraphs 138–9 of Text 19 are clear evidence that his epistemology had a thoroughly rationalist character. This means that Diotimus’ empiricist interpretation of what Democritus meant by opsis tōn adēlōn ta phainomena must be wrong. (Incidentally, this is not the first time in the compilation Sextus is using that a rival interpretation is introduced—as we introduce the views of other scholars—precisely in order to show that it is wrong.⁴⁴) Contrary to Diotimus’ interpretation, Democritus does not think that appearances give us a sight or grasp (katalēpsis) of things unseen. So how could Democritus have commended Anaxagoras for his dictum?

Well, the word opsis has a passive sense as well as an active one. Suppose Democritus said, with a touch of irony, Yes, appearances are the opsis of things unseen—the phainomena are just the façade, the visible show put on by atoms and the void—using opsis in the passive and potentially pejorative sense it bears in Aristotle’s Poetics. His point would be that only reason can discover what happens behind the façade in the microscopic world. Given a scene-painting such as Agatharchus boasted about, you have to complete the picture with your mind, your reason; the senses don’t lead you on, there is no such thing as seeing the four columns as implying more of the same sort. Look at Democritus’ own words in paragraph 139 of Text 19:

γνώμης δὲ δύο εἰσὶν ἰδέαι, ἡ μὲν γνησίη ἡ δὲ σκοτίη· καὶ σκοτίης μὲν τάδε σύμπαντα, ὄψις ἀκοὴ ὀδμὴ γεῦσις ψαῦσις, ἡ δὲ γνησίη, ἀποκεκριμένη δὲ ταύτης.

There are two kinds of gnōmē (which means judgement or cognitive capacity rather than knowledge), the genuine kind and the bastard. And to the bastard belong all these: sight, hearing, smell, taste, touch; while the other kind is genuine and cut off from these.⁴⁵

He goes on:

ὅταν ἡ σκοτίη μηκέτι δύναται μήτε ὁρῆν ἐπ᾿ ἔλαττον μήτε ἀκούειν μήτε ὀδμᾶσθαι μήτε γεύεσθαι μήτε ἐν τῇ ψαύσει αἰσθάνεσθαι, ἀλλ᾿ ἐπὶ λεπτότερον …

When the bastard kind is unable any longer to see smaller or hear or smell or taste or perceive by touch, but for finer …

And there the Greek text stops, but as you see from the English translation, every reader’s mind easily completes the thought by postulating some unseen letters to complete and explain the whole of which the visible letters are just the bit that shows. (Please don’t misunderstand me. I am not seriously suggesting that the sentence is written to exemplify its own message, so that we need postulate no lacuna after all’ epi leptoteron. Any sober classical scholar will prefer to say there is an omission in the manuscripts just like the omission of opsis tōn adēlōn ta phainomena in all manuscripts but N lower down in Text 19.) The point I wish to emphasize is that Democritus’ view that reason is cut off from the senses is the epistemological correlate to his view that the macroscopic is discontinuous with the microscopic. Hence his saying, in Text 25, from the previous page of Sextus’ history,

[T25] ἐν δὲ τῷ περὶ ἰδεῶν “γιγνώσκειν τε χρή” ϕησὶν “ἄνθρωπον τῷδε τῷ κανόνι ὅτι ἐτεῆς ἀπήλλακται.” (S.E. M. 7. 137)

And in his book Concerning Forms he says, ‘Man must get knowledge by this rule, that he is cut off from truth/reality.’ (trans. Bury, with modifications)

What cuts man off from the truth is precisely the scene-painting, the façade between us and the real world of atoms and void.

One final step will take us back, at last, to Anaxarchus. If Democritus’ attitude to the skēnographia comparison was as I have reconstructed it, the difference between ‘All the world’s a stage-painting’ as approved by Democritus and the sceptical slant it acquires on the lips of Anaxarchus is the difference between an original thinker who believes that reason has the power to penetrate behind the scenes and a follower who has lost that faith, who, unlike Alexander, is Cynically certain that you can’t conquer the secrets of the world you live in. This difference between Democritus and Anaxarchus is easily explained. Between the two comes the generalized sceptical doubt of Metrodorus (Text 1), who introduced an Atomist cosmology with opening words that were carefully preserved by the Hellenistic compilers: ‘We know nothing, not even this very thing, that we know nothing.’

I have now answered the question I began with. Anaxarchus chose the model of scene-painting to express his scepticism because skēnographia was already the name and the image of the epistemological problem in the debate between Democritus and Anaxagoras. It remains only to explain why I have so far said virtually nothing about optics.

The reason is that I believe that, where Anaxagoras and Democritus are concerned, there is nothing to say. Many scholars have credited them with writing treatises on perspective. Eva Keuls, in her book Plato and Greek Painting, goes as far as to make Democritus’ theory of perspective a major influence on the Sicyonian school of painters.⁴⁶ In this lecture I have tried to show that a careful reading of Vitruvius removes the basis for thinking that either Democritus or Anaxagoras had a mathematician’s interest in skēnographia. For them, skēnographia was a topic in general epistemology. It is only later that skēnographia becomes a recognized branch of mathematical optics. And before that can happen, mathematical optics must itself come into existence.

It would take another paper, with more analysis of so-called ‘late sources’ and more balancing of probabilities, to show that mathematical optics does not develop in Athens or Abdera, but in South Italy—the area of the Würzburg vase reproduced in Figure 2 (mid-fourth-century bc) and other illustrations in Trendall and Webster which show, from vases, the advances made by the art of scene-painting in the Greek cities of fourth-century Italy.⁴⁷ It is also in fourth-century Italy, when the Pythagoreans begin at last to do some real mathematics under the leadership of Archytas, that mathematical optics is first attested, and from there it reaches the Academy at Athens. But while Aristotle refers frequently to mathematical optics, it seems to have had no effect on his theory of vision. It is only with the Hellenistic philosophers that optics begins to leave its trace on general epistemology. Text 26 shows Chrysippus incorporating the Euclidean concept of the visual cone into his account of sight.

Fig. 2. Reconstruction of the Würzburg skēnographia

From Apulian calyx krater fragments in several colours, c.350 bc, Martin von Wagner Museum der Universität Würzburg, inv. H 4696+H 4701. Drawing by Brinna Otto, Universität Innsbruck, from E. Simon and B. Otto, ‘Eine neue Rekonstruktion der Würzburger Skenographie’, Archäologischer Anzeiger, 88/1 (1973), 121–31, fig. 3 (p. 125). Reproduced by kind permisson of Professor Otto

[T26] ὁρᾶν δὲ τοῦ μεταξὺ τῆς ὁράσεως καὶ τοῦ ὑποκειμένου ϕωτὸς ἐντεινομένου κωνοειδῶς, καθά ϕησι Χρύσιππος ἐν δευτέρῳ τῶν Φυσικῶν καὶ Απολλόδωρος. γίνεσθαι μέντοι τὸ κωνοειδὲς τοῦ ἀέρος πρὸς τῇ ὄψει, τὴν δὲ βάσιν πρὸς τῷ ὁρωμένῳ· ὡς διὰ βακτηρίας οὖν τοῦ ταθέντος ἀέρος τὸ βλεπόμενον ἀναγγέλλεσθαι. (D.L. 7. 157)

They hold that we see when the light between the visual organ and the object stretches in the form of a cone: so Chrysippus in the second book of his Physics and Apollodorus. The apex of the cone in the air is at the eye, the base at the object seen. Thus the thing seen is reported to us by the medium of the air stretching out towards it, as if by a stick. (trans. Hicks, emphasis added)

But interestingly enough Epicurus is the one who makes most use of the newly developed mathematical understanding of visual perspective—just because, I would suggest, in the Atomist tradition skēnographia had been, as I said, the name and the image of the problem.

Epicurus’ task was to refute the scepticism of Metrodorus and establish Atomism on a secure epistemological foundation in the evidence of the senses. To do this he must explain standard examples of perceptual illusion, such as the square tower which appears round at a distance. It turns out that mathematical optics is the tool that enables him to succeed. All he needs to add is his famous theory of eidōla (simulacra in Lucretius) so as to provide a suitable physical embodiment for the diagrams with which the mathematician works out his theorems. To many people this will seem a strong and surprising claim. Lucretius’ account, in book 4 of De rerum natura, of the way we see things through the very thin simu-lacra or films that bodies are constantly giving off strikes most readers as one of the more curious and unpromising products of ancient speculation. But that reaction, I believe, is at least partly due to our neglect of a third dimension of the unity of the Classics, the unity of art and science in the ancient world. Obviously, I cannot expand on this theme now. But I can offer one little titbit to advertise its importance.

Anyone who comes to Lucretius book 4 after reading the surviving ancient literature on mathematical optics will find it obvious that in the last line of Text 27 in aedibus should be translated ‘in the temple’, not ‘in the house’.

[T27] fit quoque de speculo in speculum ut tradatur imago,

quinque etiam 〈aut〉 sex ut fieri simulacra süerint.

nam quaecumque retro parte interiore latebunt,

inde tamen, quamvis torte penitusque remota,

omnia per flexos aditus educta licebit

pluribus haec speculis videantur in aedibus esse. (Lucr. 4. 302–7)

It comes to pass too that the image is handed on from mirror

to mirror, so that even five or six idols are wont to be made.

For even when things are hidden far back in an inner part of

the room, yet, however far distant from the sight along a

twisting path, it may be that they will all be brought out thence

by winding passages, and, thanks to the several mirrors, be

seen to be in the house. (trans. Bailey, emphasis added)

Lucretius is not describing a phenomenon you might come across anywhere as you stroll through the streets of Rome. In good Epicurean style, he is debunking, by giving a rational explanation of, a piece of temple magic, the mathematics of which is beautifully set out in the Catoptrics of Heron of Alexandria (Catoptr. 18). One diagram shows what you see as you approach the temple, the other shows how the trick is effected (Figure 3). It is in a parallel way, I believe, that the general Epicurean theory of vision sets out, with the aid of Euclidean optics, to debunk, by giving a rational explanation of, the perceptual illusions which foment scepticism. Eventually, skēnographia ceases to be the name of the problem and becomes the name of the solution.

Fig. 3. Sketches illustrating Heron, Catoptrica, prop. 18

From Heronis Alexandrini opera quae supersunt omnia, ed. L. Nix and W. Schmidt, ii/1. Catoptrica (Leipzig, 1900), figs. 91a (p. 358) and 91b (p. 361)

But to show all that in enough detail to be convincing would, as I said, require another paper. For this occasion I shall be content if the philosophical perspectives I have offered on one passage of Vitruvius contribute to a better understanding and appreciation, in Classical circles generally, of a technical author who not only changed the face of Europe, but who also expressed his own vision of the unity of the Classics when he wrote that an architect needs to be philosopher as well, otherwise he won’t know how to get the air bubbles out of his pipework (1. 7).

All Souls College, Oxford

BIBLIOGRAPHY

Bailey, C. (trans.), Lucretius: On the Nature of Things (Oxford, 1921).

Baldry, H. C., The Greek Tragic Theatre (London, 1971).

Barron, J. P., ‘New Light on Old Walls: The Murals of the Theseion’, Journal of Hellenic Studies, 92 (1972), 20–45.

Beutler, R., ‘Olympiodorus’, in Paulys Real-Encyclopädie der classischen Altertumswissenschaft, xviii/1 (1959), 217 ff.

Bluck, R. S. (trans.), Plato: Phaedo. A Translation with Introduction, Notes and Appendices (London, 1955).

Brown, A. L., ‘Three and Scene-Painting Sophocles’, Proceedings of the Cambridge Philological Society, 30 (1984), 1–17.

Burkert, W., Lore and Science in Ancient Pythagoreanism, trans. E. L. Minar (Cambridge, Mass., 1972).

Burnet, J. (ed.), Plato: Euthyphro, Apology of Socrates, and Crito, with Notes (1924; repr. Oxford, 1977).

Burnyeat, M. F., ‘The Upside-Down Back-to-Front Sceptic of Lucretius IV 472’, Philologus, 122/1 (1978), 197–206.

Bury, R. G. (trans.), Sextus Empiricus: Against the Logicians, text with English translation (Cambridge, Mass., 1935).

Bury, R. G. (trans.), Sextus Empiricus: Outlines of Pyrrhonism, text with English translation (Cambridge, Mass., 1933).

Butcher, S. H., Aristotle’s Theory of Poetry and Fine Art, 4th edn. (London, 1920).

Butler, H. E. (trans.), Quintilian: Institutio oratoria, text with English translation, 4 vols. (Cambridge, Mass., 1920–2).

Clement, P. A. and H. B. Hoffleit (trans.), Plutarch: Moralia, vol. vii, text with English translation (Cambridge, Mass., 1969).

Dyer, L., ‘Vitruvius’ Account of the Greek Stage’, Journal of Hellenic Studies, 12 (1891), 356–65.

Else, G. F., ‘Aristotle and Satyr-Play: I’, Transactions and Publications of the American Philological Society, 70 (1939), 139–57.

Else, G. F., Aristotle’s Poetics: The Argument (Leiden, 1957).

Fensterbusch, C. (ed.), Vitruv: Zehn Bücher über Architektur/Vitruvii De architectura libri decem (Darmstadt, 1976).

Ferguson, J. (trans.), Clement of Alexandria: Stromateis Books 1–3 (Washington, 1991).

Frank, E., Platon und die sogenannten Pythagoreer: Ein Kapitel aus der Geschichte des griechischen Geistes [Platon] (Halle, 1923).

Gardiner, A. H., ‘The Eloquent Peasant’, Journal of Egyptian Archaeology, 9 (1923), 5–25.

Gardner, P., ‘The Scenery of the Greek Stage’, Journal of Hellenic Studies, 19 (1899), 252–64.

Gogos, S., ‘Bühnenarchitektur und antike Bühnenmalerei: Zwei Rekonstruktionsversuche nach griechischen Vasen’ [‘Bühnenarchitektur’], Jahreshefte des Österreichischen archäologischen Instituts in Wien, 54 (1983), 59–86.

Gottschalk, H. B. (ed.), Strato of Lampsacus: Some Texts Edited with a Commentary (Leeds, 1965).

Granger, F. (trans.), Vitruvius: On Architecture, Books 6–10, text with English translation, vol. ii (Cambridge, Mass., 1934).

Helmbold, W. C. (trans.), Plutarch: On Tranquility of Mind, text with an English translation (Cambridge, Mass., 1939).

Hicks, R. D. (trans.), Diogenes Laertius: Lives of Eminent Philosophers, text with English translation, 2 vols. (Cambridge, Mass., 1931).

Keuls, E., Plato and Greek Painting (Leiden, 1978).

Kirk, G. S, Raven, J. E., and Schofield, M., The Presocratic Philosophers: A Critical History with a Selection of Texts [KRS], 2nd edn. (Cambridge, 1983).

Lefebvre, G., ‘Le conte de l’Oasien (le Paysan)’ [‘Paysan’], in Romans et contes égyptiens de l’époque pharaonique: traduction avec introduction, notices et commentaire (Paris, 1949), 41–69.

Lefkowitz, M. R., ‘Aristophanes and Other Historians of the Fifth-Century Theater’ [‘Historians’], Hermes, 112 (1984), 143–53.

Lightfoot, J. L. (ed., trans., and comm.), Dionysius Periegetes: Description of the Known World (Oxford, 2014).

MacDowell, D. M. (trans. and comm.), Demosthenes: Against Meidias (Oration 21) (Oxford, 1990).

Maidment, K. J. (trans.), Minor Attic Orators: Antiphon and Andocides, text with English translation (Cambridge, Mass., 1941).

Morgan, M. H. (trans.), Vitruvius: The Ten Books on Architecture (1914; repr. New York, 1960).

Murray, A. T. (trans.), Homer: The Odyssey, 2 vols. (Cambridge, Mass., 1919).

Nix, L., and Schmidt, W. (eds.), Heron: Catoptrica, in Heronis Alexandrini opera quae supersunt omnia, vol. ii/1 (Leipzig, 1900).

Pollitt, J. J., The Ancient View of Greek Art [View] (New Haven, 1974).

Rackham, H. (trans.), Cicero: De natura deorum and Academica, text with English translation (Cambridge, Mass., 1951).

Robertson, M., A History of Greek Art, 2 vols. (London, 1975).

Robertson, M., Greek Painting (London, 1959).

Rumpf, A., ‘Classical and Post-Classical Greek Painting’ [‘Painting’], Journal of Hellenic Studies, 67 (1947), 10–21.

Sandbach, F. H. (trans.), Plutarch: Moralia, vol. xv, text with English translation (Cambridge, Mass., 1969).

Schneider, G. (ed. and comm.), Marci Vitruvii Pollionis De architectura, 3 vols. (Leipzig, 1808).

Sedley, D. N., ‘Epicurus’ Refutation of Determinisim’, in Συζήτησις: studi sull’epicureismo greco e latino offerto a Marcello Gigante, 2 vols. (Naples, 1983), i. 11–51.

Shackleton Bailey, D. R. (trans.), Valerius Maximus: Memorable Doings and Sayings, text with English translation (Cambridge, Mass., 2000).

Simon, E., The Ancient Theatre [Theatre], trans. C. E. Vafopoulou-Richardson (London, 1982).

Simon, E., and Otto, B., ‘Eine neue Rekonstruktion der Würzburger Skenographie’, Archäologischer Anzeiger, 88/1 (1973), 121–31.

Stewart, Z., ‘Democritus and the Cynics’, Harvard Studies in Classical Philology, 63 (1958), 179–91.

Taplin, O., The Stagecraft of Aeschylus [Stagecraft] (Oxford, 1989).

Thielscher, P., ‘Vitruvius’, in Paulys Real-Encyclopädie der classischen Altertumswissenschaft, ixa/1 (1961), 431–2.

Trendall, A. D., and Webster, T. B. L., Illustrations of Greek Drama (London, 1971).

Vince, J. H. (trans.), Demosthenes: Orations XXI–XXVI, text with English translation (Cambridge, Mass., 1935).

Webster, T. B. L., Greek Art and Literature 530–400 B.C. (Oxford, 1939).

Webster, T. B. L., Greek Theatre Production [Production] (Norwich, 1970).

White, J., Perspective in Ancient Drawing and Painting [Perspective] (London, 1956).

Wyttenbach, D. (ed.), Plutarchi Chaeronensis Moralia, id est, opera, exceptis Vitis, reliqua, vol. v/2 (Leipzig, 1832).

The following is the text of an unpublished lecture which Professor Burnyeat gave frequently in the late 1980s and in the 1990s. Like many of his unpublished papers, he requested that it not be circulated without his express permission, and so as a result relatively few people had access to it, sometimes through samizdat copies. But as it continues to be regarded highly and cited regularly, I felt it was valuable to make it available to a wider audience. I am grateful to Professor Burnyeat, who has kindly given his permission to publish it in this form, to Professors Sedley and Hobbs for their advice, and to my assistant David Morphew for his meticulous work in tracking down the references.

Professor Burnyeat has not revised the typescript further for publication here and has not been involved in its editing. The manuscript is very polished, however, and appears to have reached a stable form by the late 1990s: the only variations we found in two typescripts were that a small number of handwritten marginalia in one had been incorporated into the text of the other, along with the deletion of two paragraphs (which I have retained, but placed in double brackets). Only a few footnotes were left blank or with promissory notes. Otherwise I have kept the paper as it was, including many of the hallmarks of oral delivery, limiting myself to filling in the citations and references, sometimes with a more recent edition, all without comment. Any supplement editorially added has been placed in square brackets, ending with ‘—Ed.’

Had Professor Burnyeat revised the lecture for publication, he would almost certainly have provided translations of his own. I have instead provided the texts as they occur on his handout, which, in keeping with his usual practice, was mostly composed from xeroxes, cut and pasted from bilingual editions such as the Loeb Classical Library, with his own annotations or modifications (which we have included, using italic and underlining). Where he did not provide a translation, I have simply supplied the Loeb translation, when available, in keeping with his practice in the rest of the lecture, or another recent translation, with the translator’s name always preceded by an asterisk (*).

¹ The first according to D.L. 9. 60, but references to the second in LSJ s.v. ‘εὐδαιμονικός’ (D.L. 1. 17; Clearchus fr. 60 Wehrli = Athen. 12, 548 b). See also S.E. M. 7. 48.

² For citations and discussion see M. F. Burnyeat, ‘The Upside-Down Back-to-Front Sceptic of Lucretius IV 472’, Philologus, 122/1 (1978), 197–206 at 204 and n. 25; D. N. Sedley, ‘Epicurus’ Refutation of Determinisim’, in Συζήτησις: studi sull’epicureismo greco e latino offerto a Marcello Gigante, 2 vols. (Naples, 1983), i. 11–51 at 33.

³ See Z. Stewart, ‘Democritus and the Cynics’, Harvard Studies in Classical Philology, 63 (1958), 179–91 at 180–8; Demetrius ap. Seneca (‘truth in the depths’, Ben. 7. 1. 5). [Demetrius is the Cynic speaker in Seneca’s De beneficiis 7. 1. 5, who is presenting Democritean material. For discussion of this passage see Stewart, op. cit. 181.—Ed.]

⁴ Timaeus, FGrH 566F7 (= Polyb. 12. 28a. 1), is 350–250 bc.

⁵ A. L. Brown, ‘Three and Scene-Painting Sophocles’, Proceedings of the Cambridge Philological Society, 30 (1984), 1–17 at 1–8; O. Taplin, The Stagecraft of Aeschylus [Stagecraft] (Oxford, 1989), 457–8 n. 4. Answer A. L. Brown and G. F. Else (Brown, op. cit. 1–2 and 5–6; G. F. Else, Aristotle’s Poetics: The Argument (Leiden, 1957), 164–79; id., ‘Aristotle and Satyr-Play: I’, Transactions and Publications of the American Philological Society, 70 (1939), 139–57) on ὀψέ by reference to the life of tragedy and Plato, Laws on ὀψέ ποτε (Laws 7, 819 d 5–9), i.e. late in its own development, not lately relative to the time of Aristotle’s writing. Correct A. L. Brown (op. cit. 8 and n. 31) on whether in this instance Sextus Empiricus is a ‘late source’. [The last promissory note appears to have gone unfulfilled.—Ed.]

⁶ KRS 542 = D.L. 9. 34.

⁷ ad scaenam e₂, Granger [All references to Granger are to his Loeb edition of Vitruvius.—Ed.]

⁸ adlineas codd.: corr. Schneider: ad lineas Granger

⁹ incerta re~ certae H: dein certarem certae c: incerta re incertae h: incerta re certae alii: de certa re certae Krohn

Granger notes that the flourish attached to ‘re’ in H (the oldest and primary manuscript, from which he contends, against both previous and subsequent editors, that all others derive) is like the symbol for ‘m’ (cf. c’s ‘certarem’) and suggests that this ‘m’ is the scribe’s misreading of ‘in’. He prints ‘incertae’, basically because he believes it to be required by Democritean doctrine. The alternative is to suppose that c and h exhibit progressive corruption of the original ‘certae’. Professor Michael Reeve has looked at H for me and reports as follows: ‘The passage reads in all clarity (spacing excepted) uti de incerta re certe imagines. The squiggle or flourish after re is pure imagination: the centre of e protrudes with a modest flourish, but so it does in many other places on the same page.’

¹⁰ de aede ionica Iunonis quae est Sami Rhoecus et Theodorus conj. Granger, on grounds of truth.

¹¹ ‘eine perspektivische Bühnendekoration gemalt’— E. Frank, Platon und die sogenannten Pythagoreer: Ein Kapitel aus der Geschichte des griechischen Geistes [Platon] (Halle, 1923), 234; ‘paint[ed] a perspective stage set’—E. Simon (paraphrase), The Ancient Theatre [Theatre], trans. C. E. Vafopoulou-Richardson (London, 1982), 22; ‘was in control of the stage’—Granger; ‘set the stage’— J. White, Perspective in Ancient Drawing and Painting [Perspective] (London, 1956), 46.

¹² ‘how it is necessary that, a fixed centre being established, the lines correspond by natural law to the sight of the eyes and the extension of the rays’—White, loc. cit.

¹³ ‘so that from an uncertain object certain images may render the appearance of buildings in the paintings of the stages’—White, loc. cit.; ‘such that from an uncertain object, uncertain images may give the appearance of buildings in the scenery of the stage’—Granger (see n. 9 above); ‘so that … clear pictures can reproduce the appearance of buildings in stage-scenery’—Simon, loc. cit.; ‘damit von der undeutlichen Sache her deutliche Bilder den Anblick von Gebäuden bei den Bühnenmalereien wiedergeben’— C. Fensterbusch (ed.), Vitruv: Zehn Bücher über Architektur/ Vitruvii De architectura libri decem (Darmstadt, 1976); ‘so dass über eine ungenaue Sache genaue Abbilder die Gestalt der Gebäude auf der Bühnendekoration wiedergeben’—Frank, Platon, 235 [emphasis added by Professor Burnyeat.—Ed.]

¹⁴ J. J. Pollitt, Ancient View of Greek Art [View] (New Haven, 1974), 240–7.

¹⁵ [This note was left blank by the author, to be completed at a later date. The relevant claim can, however, be found in the following works: P. Gardner, ‘The Scenery of the Greek Stage’, Journal of Hellenic Studies, 19 (1899), 252–64 at 253; T. B. L. Webster, Greek Art and Literature 530–400 B.C. (Oxford, 1939), 74; M. Robertson, Greek Painting [Greek Painting] (London, 1959), 164; H. C. Baldry, The Greek Tragic Theatre (London, 1971), 46–7.—Ed.]

¹⁶ Petronius 62. 4, ‘homo meus coepit ad stelas facere’. Ox. Lat. Dict. says ‘perh(aps)’ under ‘facio’ 27b.

¹⁷ Inconsistent: S. Gogos, ‘Bühnenarchitektur und antike Bühnenmalerei: Zwei Rekonstruktionsversuche nach griechischen Vasen’ [‘Bühnenarchitektur’], Jahreshefte des Österreichischen archäologischen Instituts in Wien, 54 (1983), 59–86 at 71; M. R. Lefkowitz, ‘Aristophanes and Other Historians of the Fifth-Century Theater’ [‘Historians’], Hermes, 112 (1984), 143–53 at 152–3. Not inconsistent: T. B. L. Webster, Greek Theatre Production [Production] (Norwich, 1970), 13–14; Simon, Theatre, 22 and n. 84; Pollitt, View, 245.

¹⁸ P. Thielscher, ‘Vitruvius’, in Paulys Real-Encyclopädie der classischen Altertumswissenschaft, ixa/1 (1961), 431–2: wrote in the thirties, published in the twenties, on account of De architectura 1. 1. [The dedication to Augustus that opens De architectura 1. 1 makes reference to his victory at Actium (31 bc) and ascendancy to imperial rule, and to Vitruvius’ deferral of publication until later.—Ed.]

¹⁹ Contra Lefkowitz’s ‘wishful memory’ (‘Historians’, 153).

²⁰ Suda, s.v. ‘Λάσος’; W. Burkert, Lore and Science in Ancient Pythagoreanism, trans. E. L. Minar (Cambridge, Mass., 1972), 372 n. 12 and 378.

²¹ R. Beutler, ‘Olympiodorus’, in Paulys Real-Encyclopädie der classischen Altertumswissenschaft, xviii/1, (1959), 217 ff.

²² H. B. Gottschalk (ed.), Strato of Lampsacus: Some Texts Edited with a Commentary (Leeds, 1965), 164.

²³ D. Wyttenbach (ed.), Plutarchi Chaeronensis Moralia, id est, opera, exceptis Vitis, reliqua, vol. v/2 (Leipzig, 1832), 738–9.

²⁴ Doubted by G. Lefebvre, ‘Le conte de l’Oasien (le Paysan)’ [‘Paysan’], in Romans et contes égyptiens de l’époque pharaonique, traduction avec introduction, notices et commentaire (Paris, 1949), 41–69 at 41.

²⁵ Discussions in Lefebvre, ‘Paysan’.

²⁶ A. H. Gardiner, ‘The Eloquent Peasant’, Journal of Egyptian Archaeology, 9 (1923), 5–25 at 25.

²⁷ Plut. Mor. 678 c: on ‘my return from Alexandria’ (trans. Clement-Hoffleit).

²⁸ [FGrH 1F1a = Demetrius, De elocutione 12.—Ed.]

²⁹ Taplin, Stagecraft, Appendix C, ‘The Skene in Aeschylus’, 452–9.

³⁰ Taplin: Aeschylus’ use is itself a recall of Homer and his odd word. [We were unable to locate a reference for this claim. When consulted, Professor Taplin assured us that he has never made such a claim in print, but thought it possible that he might have discussed the issue with Professor Burnyeat on some occasion.—Ed.]

³¹ Webster, Production, 13–14; Taplin, Stagecraft, Appendix C, ‘The Skene in Aeschylus’, esp. 457–8 n. 4; see also ibid. 107 and 438 n. 2.

³² M. Robertson, A History of Greek Art, 2 vols. (London, 1975), 414–15; Pollitt, View, 244–5.

³³ A. Rumpf, ‘Classical and Post-Classical Greek Painting’ [‘Painting’], Journal of Hellenic Studies, 67 (1947), 10–21 at 13; Marm. Par. ep. 59 Jacoby; Σ ad Ar. Ach. 10 C Wilson.

³⁴ Rumpf, ‘Painting’, 13.

³⁵ Cf. also Plut. Alc. 4. [The precise interpretation of the events behind these reports is much disputed. For a brief survey of the issues see D. M. MacDowell (trans. and comm.), Demosthenes: Against Meidias (Oration 21) (Oxford, 1990), 362–3.—Ed.]

³⁶ White, Perspective, 43–69.

³⁷ Unsensible: Gogos, ‘Bühnenarchitektur’, 71–86, esp. 84–5.

³⁸ Cf. the elaborately odd interpretation in Pollitt, View, 241–2, close to Schneider ad loc. [Pollitt writes, ‘I take the meaning of the phrase to be that an assemblage of drawn lines, which, when viewed without regard to its illusionistic quality, has no meaning or significance in itself (incerta res), becomes, when understood illusionistically, a definite thing, certa res, for example, a building’ (View, 242).—Ed.]

³⁹ L. Dyer, ‘Vitruvius’ Account of the Greek Stage’, Journal of Hellenic Studies, 12 (1891), 356–65 at 364.

⁴⁰ J. P. Barron, ‘New Light on Old Walls: The Murals of the Theseion’, Journal of Hellenic Studies, 92 (1972), 20–45 at 25, 28, 31–44.

⁴¹ Robertson, Greek Painting, 122.

⁴² Burnet ad Plato, Ap. 26 d 10.

⁴³ D.L. 2. 11 (see Hicks’s note b ad loc.); Plut. Nic. 23; Clem. Strom. 1. 78. 4 Stählin.

⁴⁴ The paragraphs on Empedocles. [Professor Burnyeat presumably intended to fill this note out further, but in any case the passages he has in mind from Sextus seem to be M. 7. 115 and 122.—Ed.]

⁴⁵ [Burnyeat has slightly altered the translation printed earlier in Text 19: instead of Bury’s ‘distinct from’ for the word ἀποκεκριμένη emphasized there, he uses ‘cut off from’, a phrase that recurs below.—Ed.]

⁴⁶ E. Keuls, Plato and Greek Painting (Leiden, 1978), 140–50, esp. 140–1 and 147–50.

⁴⁷ A. D. Trendall and T. B. L. Webster, Illustrations of Greek Drama (London, 1971).