^{page 15 note 1} 514a 1, 515a 5.

^{page 15 note 2} See, e.g., 492e. This part of Book VI. is the most valuable commentary on the cave, although the application of the Cave to the line has obscured it. As in the first paper, Sun, Cave, and Line are used for the similes, and sun, line, cave for the objects themselves.

^{page 16 note 1} See especially 492e and 493a.c.

^{page 16 note 2} ‘A shadow-pantomime produced by throwing shadows of miniature figures on a wall or screen’ (O.E.D.). There is, so far as I know, no earlier description of the shadow-play, as distinguished from the puppet-play, in Europe.

^{page 16 note 3} 518d, oùκ òρθς δ τετραμμένῳ οὺδ βλποντι οῖ ἕδει: 519b 5.

^{page 16 note 4} ‘The gradual ascent,’ Shorey, loc. cit., p. 238, Adam on 532b.

^{page 16 note 5} Persae, 688.

^{page 16 note 6} 515d.

^{page 16 note 7} 516a, 517a, and see below.

^{page 17 note 1} 517d, π θεων θεωριν π τᜀ νθρώπεια. Cf. 518a, κ πωτς ες σκóτος.

^{page 17 note 2} The classification in the Sophist is only superficially like this one, because it does no more than distinguish between what man makes and what God makes. But this is a piece of symbolism, and must be interpreted in accordance with the requirements of the symbolism. In the Sophist fire and sun and shadows are all alike made by God; but this can have no bearing at all on a simile which turns upon the distinction between fire and sun, between one set of shadows and the other.

^{page 17 note 3} In the Epinomis (990d) a distinction is drawn between the art of land-surveying and pure geometry, similar to that which is made in the propaedeutic between the disciplines that lead to Being and the arts that are merely useful: δ δ θαμα οὐκ νθρώπινον λλ γεγονòς θεῖον, κ.τ.λ. For the utilitarian arts as serving the νθρώπινος βος in the narrow sense see Xenophon, Mem. IV. 7, 2. See also 500c, θεῳ δ κα κασμῳ ὅ γε θ;ιλσος μιλν κσμις τε κα θεῖος εἰς τ δυνατν νθρώπῳ γγνεται. See 500e, θεῳ παραδεγματι; 518e (an important passage), 589d.

^{page 17 note 4} E.g. Soph. 219a.

^{pate 18 note 1} 492b.c, 493d; cf. 516c, τιμα δ κα ἔπαινοι.

^{page 18 note 2} 491d.

^{page 18 note 3} 500b.c, 492c. Cf. 517c, τ τν ν θ ρ ώ π ω ν πρττειν.

^{page 18 note 4} 490e, 491c.

^{page 18 note 5} Βοι … 518a 7, κ θανεροτρου βου ἥκουσα (Cf. b 2); 520e 4, εἰ μν βον ξευρσεις μενω το ρχεῖν; 521b, Ἔχεις οὖν βον ἄλλον τινα π ο λ τ ι κ ν ρ χ ν καταπρονοντα ἢ τ ò ν τ ς λ η θ ι ν ς π ι λ ο σ ο π α ς; b 9, τ ι μ ς ἅλλας κα β ο ν μενω το π ο λ ι τ ι κ ο .

^{page 18 note 6} Τ ι μ α … 516c 8, 519d 6, 521b 9, 540d 5. The philosopher ‘knows other honours and a better “life,” but men like Kallikles believe that there is but one life’: ζηλν οὐκ λγχοντας ἄνδρας τ μικρ τατα, λλ' οἷς ἔστιν κα β ο ς κα δ ξ α [scil., τιμ] κα ἄλλα πολλ γαθ (Gorgias, 486c]. πιλοτιμα is also πιλονικα Rep. 516e 9, 517d 8, 521a. See especially 520c: ὡς νν αἱ πολλα [πλεις] ὑπò σκιαμαχοντων τε πρòς λλλους κα στασιαζντων περ το ρχεῖν οἰκονται, ὡς μεγλου τινς γ α θ ο ντος.

^{page 18 note 7} See Aristotle, Fr. 58 (Rose), for an application of the figure of θεωρα.

^{page 18 note 8} See 491–3, especially 493b: σοφαν τε καλσειεν κα ὡς τ χ ν η ν συστησμεος, κ.τ.λ., and 493d, comparing 516c. τς κεῖ σοφας.

^{page 18 note 9} 516d; Odyssey, XI. 489.

^{page 18 note 10} 516e, 'κεῖν τε δοξζειν κα κενως ζν … ζν κενως (ζν is the verb of βος). This phrase too, be it noted, is the equivalent of τιμωμνους τε κα νδυναστεοντας.

^{page 19 note 1} I wish to suggest a not unlikely meaning for another allusion to Hades. Those who are rescued from the cave are compared to men raised from Hades to the gods (521c). The names of some who did so ascend are collected by Adam; but, so far as Plato had any definite figure in mind, is not Pollux (or rather Castor) appropriate for the very reason that leads Adam to exclude him? Castor's life above was intermittent—‘si fratrem Pollux alterna morte redemit’—and he owed it to the self-sacrifice of Pollux. Each took his turn below. But the prisoners in their Hades can only be rescued by men who sacrifice the divine life for the time, and they must themselves take their turn in the cave. Cf. 520c, καταβατον οὖν ν μρει, 520d, ν μρει, 540b, ὅταν δ τò μρος ἥκῃ. The plot involves a κατβασις and νβασις by turns.

^{page 19 note 2} Olympiodorus, , In Phaeàonem, 101, 11 (Norwin)Google Scholar ; βαθμο τς ληθεας.

^{page 19 note 3} Pepys's Diary, December 21, 1661.

^{page 20 note 1} Works and Days, 287. For traces in our simile see 532e: ὥσπερ δο νπαυλα … κα τλος τς πορεας; 515e, δι τραχεας τς ναβσεως (cf. Hesiod, 291); 517b 4, 5; 521c; 531c; 533b 3, c 7. See also Part I., p. 146, n. 2.

For the applicability of the metaphor of θεατα to politicians see Cleon's speech in Thucydides, III. 38.

^{page 20 note 2} For θαυματοποια as a metaphor for some trifling pursuit, see Isocrates' attack on the cosmologists in Antidosis, 269, The galanty-show is a natural image for the vanity of φιλοτιμα. See The Dynasts, II. v., viii. (on Napoleon's marriage): ‘All day have they been waiting for their galanty-show.’

^{page 20 note 3} See Stocks, , Mind, 04, 1915Google Scholar .

^{page 21 note 1} 519c: μτερον δ ἔργον … τν οἰκιστν.

^{page 21 note 2} 515a 5.

^{page 21 note 3} Cf. 495d: ὑπ δ τν τ ε χ ν ν τε κα δημιουργιν σπερ τ σώματα λελώβηνται, οὕτω κα τς ψυχς συγκεκλασμνοι κ.τ.λ.; 535d 9: πρòς λ θειαν … νπηρον ψνχν. Plato suggests that the τχναι of the cave produce βναυσοι (495e, 522b). See the list of φθορα τς ψυχς in p. 491 of Book VI. The best illustration of the figure of the bonds is Theaetetus, 173a-b. For an γών that failed compare Plutarch's story about the calling in of Plato for the younger Dionysius: διαλελωβημνον παιδευσᾳ κα συντετριμμνον τò ἦθος (Dion. c. X.).

^{page 21 note 4} The terrestrial cavern of Empedokles and the myth of the Phaedo have given plausibility to the identification of this Hades with the earth and its inhabitants. If any particular place suggested the cave to Plato, it would seem to be the cave of Vari on Hymettus, which corresponds in all essential points to the description in the text. See Wright, , Harvard Studies in Class. Philology, XVIIGoogle Scholar .

^{page 21 note 5} It is important to observe from 515a-c that the shadow-play is the prisoners' whole world. Four questions are asked to emphasize this confinement in one plane, so to speak. (1) What do they see of themselves and of each other? The shadows. (2) And of the puppets? The shadows. Following Proclus, some modern writers have supposed that there are two kinds of shadows, those of the puppets and those of the prisoners. But c2 shows that this is not so: indeed it is manifest that the intrusion of large immobile shadows upon the moving show would spoil the illusion—an accident that we have all seen at a lantern entertainment. I take it that they are seated well below the line of the firelight (cf. 514b 2 and 4, and see Mr. Wright's plan of the cave of Vari). The point surely is that (1) the prisoners see even of themselves only what is presented to them by the showmen, and (2) that they can't tell the source of their illusion because it is behind them.

With (3) we come to a crux. The reading of ADM is: εἰ οὖν διαλγεσθαι οἶο τ' εἶεν πρòς λλλους, οὐ τατα (ταὐτ AFM) γῇ ἂν τ παρντα αὐτοὺς νομζειν νομζειν ἅπερ ρῷεν; F omits νομζειν; Iamblichus omits νομζειν and reads ντα for παρντα; Burnet's text reads ντα and omits νομζειν. I venture to state a case for the text as I have written it above. The prisoners are ‘in blinkers’; they only see what is before them and they do not see each other. Well, if they talk, to whom do they think they are talking? To the shadows, for it is the show that they imagine to be themselves and others. The four questions seem to be about particulars, and lead up to the general conclusion in 515c. Translate ‘If then they were able to talk to each other, do you not think that they would consider they were addressing those objects before them, the objects they saw?’ This reverses an interpretation of MrBury's, R. G. (C.R., 1903, p. 296)Google Scholar .He considers that the shadows seem to address the prisoners. But this view depends on the supposition that the two previous questions deal with two kinds of shadows, and that the second pair of questions by reason of symmetry parallels the first pair. But since Plato writes ‘if they could talk’ the first question is to whom do they think they are talking, not to whom are they listening. Doubtless the second belief is implied, though not explicitly stated by Plato. (4) If there is an echo, will the prisoners not think that the bearers' voices also come from the shadows?

There are thus two parallel groups of questions. The first and the third suggest that all the prisoners see and hear of themselves comes from the screen in front of them. The second and fourth show that the mechanism of the illusion is unknown to them. In short: what they might know of their own plane and of the showmen's plane is referred to the shadows of the puppets, Such is the conclusion of 515c: Παντπασι δ, ἦν δ' γώ, οἱ τοιοτοι οὐκ ἂν ἂλλο τι ν ο μ ζ ο ι ε ν τ λ η θ ς ἢ τς τν σκευαστν σκις.

Note the force of the interlaced construction: τατα … τ παρντα … ἅπερ ρῷεν. The received παριντα spoils this. I doubt whether Proclus' τν ρχν ντα νομζουσιν (In Rem Publicam, I. 293, 20) gives any clue to the reading. It seems better to take this phrase as his interpretation of the summing-up in c1–2 just quoted.

^{page 22 note 1} 533b 2, Cf. 532C 4. The Cave should be read in the light of the distinction that Plato is careful to draw at each stage between the arts as they serve politics and utilitarian ends and the arts that draw to Being.

^{page 22 note 2} 516c-d, 519a. Nettleship (on 516d) plausibly compares πομαντευομνῳ τò μλλον ἥξειν to εἰκασα. But while εἰκασα means inference from actual evidence to a stable original, the prisoners exercise mere political ‘divination’ about the future from their flickering shadows, the originals of which are unknown to them. This mantic art is pure riddle-guessing, not a grade of perception. See 493b: συνουσᾳ τε κα χρνου τριβῇ. For the sense of ‘groping’ in πομαντεεσθαι see 505e.

^{page 22 note 3} See also 519a-b, 521a.

^{page 23 note 1} The confusion starts from φσει in 515c: σκπει δ, ἦν δ' γώ, ατν λσιν τε κα ἴασιν τν τε δεσμν κα τς φροσνης, οἶα τις ἂν εἴη, εἰ φσει τοιδε συμβανοι ατος. It owe the true rendering to Professor Burnet. It is: εἰ φσει τοιᾱδε [scil., λσις τε κα ἴασις] συμβανοι ατοῖς (φσει τοιᾱδε = τοιδε τν φσιν). For the healing see Politicus, 296b.

^{page 23 note 2} The whole tone of the dialogue is decisive against the attempt to make the sophists into ‘purgers’ of the soul. They are like true educators as wolves are like dogs (Soph. 231a)— mimics and jugglers. Cf. Rep. 496a and 494e, also 493a and c.

^{page 23 note 3} Rep. 492a; Polit. 291a-b, 303b-c.

^{page 23 note 4} See 517d-e and 582 b-c. Compare Mr. Conrad's Arrow of Gold: ‘I was as stranger as the most hopeless castaway, stumbling in the dark upon a hut of natives and finding them in the grip of some situation appertaining to the mentalities, prejudices, and problems of an undiscovered country—of a country of which he had not even had one single clear glimpse before.’

^{page 24 note 1} See Part I., p. 145. In 520c Plato means that the rescuer will be able to relate the puppet to its shadow and to that which it counterfeits; the prisoner can do neither.

^{page 24 note 2} The phrase is taken from its context, which describes three main stages—the rescue, the propaedeutic, and the dialectic (532b). The first stage is the ‘loosening from the bonds and the turning from the shadows to the puppets and the light and ascent from the cave to the sun.’ A careful reading shows that the purpose is to change the light. Plato marks the break after the first stage by the words κα κεῖ (in the sun), and by placing the third stage (the originals) before the second (their shadows) in order to emphasize as strongly as possible the break with the cave and the distinction between the two kinds of shadows. The mark of the first stage is the use of medicinal force throughout. It is the force of a physician to a diseased patient. Then in the sunlight the natural and unconstrained study of truth can begin as the youth is able (536e). For the meaning of the resistance to the turning see VI. 494e, 492e.

^{page 24 note 3} They are truer, because they are what they are, without distortion, not what the shadows make them seem to be.

^{page 24 note 4} 234d. A passage from the lost Aristotelian Protrepticus (see Bywater, , J. Phil. II. 55Google Scholar ) in Iamblichus' tract of that name might almost be a reminiscence of our passage: Γνοη δ' ἄν τις τò αὐτò κα πò τοτων, εἰ θεωρσειεν ὑπ' αὐγς τòν νθρώπειον βον. εὐρσει γρ τ δοκοντα εἶναι μεγλ τοῖς νθρώποις πντα ντα σκιαγραφαν (c. VIII.). Cf. c. X.: α ὐ τ ν γρ στι θ ε α τ ς, λλ' οὐ μ ι μ η μ τ ω ν, and c. XV., end.

^{page 25 note 1} Soph. 235a; cf. Polit. 291c.

^{page 25 note 2} I may add that it is surely impossible for any part of the cave to represent the first education. Anything more unlike a region where the guardians have from childhood breathed the ‘air that carries health from happy lands’ cannot be imagined; nor is the γών consistent with this view. The problem here is purely to find an intellectual means of rescue, and the first question in the γών is, ‘What is it?’

May not πρòς τò φς, which is the keynote of the allegory, always mean ‘towards the sunlight,’ even in 515c? I take it that the prisoners sit in a sudden dip at the bottom of the cave (see Wright's plan of the cave of Vari), but that on standing up they might be able to see the wide mouth (514a) and the daylight. However that may be, πρòς τò φς in 515c and πρòς αὐτò τò φς in e 1 must mean the same light, as the prisoner has not moved in the meantime: αὐτ is used, not in contrast to another light, but to the puppets. The phrase may suggest initiation, as in Clouds, 632; but it must be initiation into sunlight.

^{page 27 note 1} Ἄλλ δ ἂν μóνον δηλοῖ πρòς τν ἕξιν σαφηνεᾳ λγει ν ψυχῇ (λγειν, FM; λγεις, A²). Adam, who expels the clause from the text, considers ἕξιν to be a trace of Stoic influence in an interpolation. But the word is simply a reminiscence of 511d 4, which introduces the four παθματα there. Ἕξις is of course δινοια, as in 511d. Nor can one reject a mutilated text on grounds of style. There is, I suggest, no reference to the Platonic doctrine of thought as the conversation of the soul—that is irrelevant—nor does Plato mean that we should be content if the word expresses our meaning clearly. The test of σαφνεια is applied, as in the Line, to the ἕξις of δινοια. The question is, as in the Line, how clear is the ἕξις in comparison with πιστμη. Then the proportion that is to determine this is at once begun. Does not π ρ ò ς τν ἕξιν, when combined with σαφηνεᾳ, suggest that the other term in a comparison has dropped out?

^{page 27 note 2} See Poetics 1457b 16, quoted above, Part I., p. 149. There is, I think, no reason why Plato should repeat the proportion in terms of the objects. All is said when the states are given, and it would only be multiplying λóγοι to repeat the same proportions over again. But it may be remarked that if the difference between the objects of δινοια and πιστμη is not of kind, but of limitation, then a proportion of objects would involve explanations which would render brevity difficult (see Glaucon's difficulty in 511c 4 sqq.).

^{page 28 note 1} 533b, 534c. It is the same insistence on the limitations of the mathematical disciplines that gives its point to the pun in 534d—Glaucon would not allow his spiritual children, λóγους ντας σπερ γραμμς, to have control of the greatest issues as rulers in the city. There is a topical allusion to Theaetetus' doctrine of irrationals (compare the stress laid on stereometry earlier). See Miss Sachs' dissertation, De Theaeteto Atheniensi Mathematico. As Theaetetus died only in 469, the reference may well be to a discovery just made.

^{page 28 note 2} Olymp. IX. Cf. 534c: κα ὥσπερ ν μχῃ δ ι π ν τ ω ν λ γ χ ω ν δ ι ε ξ ω ν … ν πσι τοὐτοις π τ τ ι τ ῷ λ γ ῳ δ ι α π ο ρ ε η τ α ι, οὔτε αὐτò τò γαθν θσεις εἰδναι, κ.τ.λ.