Dramatic principle of classification — was inherited by Thrasyllus from Aristophanes.

Authority of the Alexandrine library — editions of Plato published, with the Alexandrine critical marks.

That Thrasyllus followed Aristophanes in the principle of his classification, is manifest: that he adopted the dramatic ground and principle of classification (while amending its details), not because he was himself guided by it, but because he found it already in use and sanctioned by the high authority of the Alexandrines — is also manifest, because he himself constructed and tacked to it a better classification, founded upon principles new and incongruous with the dramatic. In all this we trace the established ascendancy of the Alexandrine library and its eminent literati. Of which ascendancy a farther illustration appears, when we read in Diogenes Laertius that editions of Plato were published, carrying along with the text the special marks of annotation applied by the Alexandrines to Homer and other poets: the obelus to indicate a spurious passage, the obelus with two dots to denote a passage which had been improperly declared spurious, the X to signify peculiar locutions, the double line or Diplê to mark important or characteristic opinions of Plato — and others in like manner. A special price was paid for manuscripts of Plato with these illustrative appendages:48 which must have been applied either by Alexandrines themselves, or by others trained in their school. When Thrasyllus set himself to edit and re-distribute the Platonic works, we may be sure that he must have consulted one or more public libraries, either at Alexandria, Athens, Rome, Tarsus, or elsewhere. Nowhere else could he find all the works together. Now the proceedings ascribed to him show that he attached himself to the Alexandrine library, and to the authority of its most eminent critics.

Thrasyllus followed the Alexandrine library and Aristophanes, as to genuine Platonic works.

Probably it was this same authority that Thrasyllus followed in determining which were the real works of Plato, and in setting aside pretended works. He accepted the collection of Platonic compositions sanctioned by Aristophanes and recognised as such in the Alexandrine library. As far as our positive knowledge goes, it fully bears out what is here stated: all the compositions recognised by Aristophanes (unfortunately Diogenes does not give a complete enumeration of those which he recognised) are to be found in the catalogue of Thrasyllus. And the evidentiary value of this fact is so much the greater, because the most questionable compositions (I mean, those which modern critics reject or even despise) are expressly included in the recognition of Aristophanes, and passed from him to Thrasyllus — Leges, Epinomis, Minos, Epistolæ, Sophistês, Politikus. Exactly on those points on which the authority of Thrasyllus requires to be fortified against modern objectors, it receives all the support which coincidence with Aristophanes can impart. When we know that Thrasyllus adhered to Aristophanes on so many disputable points of the catalogue, we may infer pretty certainly that he adhered to him in the remainder. In regard to the question, Which were Plato’s genuine works? it was perfectly natural that Thrasyllus should accept the recognition of the greatest library then existing: a library, the written records of which could be traced back to Demetrius Phalereus. He followed this external authority: he did not take each dialogue to pieces, to try whether it conformed to a certain internal standard — a “platonisches Gefühl” — of his own.

Ten spurious dialogues, rejected by all other critics as well as by Thrasyllus — evidence that these critics followed the common authority of the Alexandrine library.

That the question between genuine and spurious Platonic dialogues was tried in the days of Thrasyllus, by external authority and not by internal feeling — we may see farther by the way in which Diogenes Laertius speaks of the spurious dialogues. “The following dialogues (he says) are declared to be spurious by common consent: 1. Eryxias or Erasistratus. 2. Akephali or Sisyphus. 3. Demodokus. 4. Axiochus. 5. Halkyon. 6. Midon or Hippotrophus. 7. Phæakes. 8. Chelidon. 9. Hebdomê. 10. Epimenides.”49 There was, then, unanimity, so far as the knowledge of Diogenes Laertius reached, as to genuine and spurious. All the critics whom he valued, Thrasyllus among them, pronounced the above ten dialogues to be spurious: all of them agreed also in accepting the dialogues in the list of Thrasyllus as genuine.50 Of course the ten spurious dialogues must have been talked of by some persons, or must have got footing in some editions or libraries, as real works of Plato: otherwise there could have been no trial had or sentence passed upon them. But what Diogenes affirms is, that Thrasyllus and all the critics whose opinion he esteemed, concurred in rejecting them. We may surely presume that this unanimity among the critics, both as to all that they accepted and all that they rejected, arose from common acquiescence in the authority of the Alexandrine library.51 The ten rejected dialogues were not in the Alexandrine library — or at least not among the rolls therein recognised as Platonic.

Thrasyllus did not follow an internal sentiment of his own in rejecting dialogues as spurious.

If Thrasyllus and the others did not proceed upon this evidence in rejecting the ten dialogues, and did not find in them any marks of time such as to exclude the supposition of Platonic authorship — they decided upon what is called internal evidence: a critical sentiment, which satisfied them that these dialogues did not possess the Platonic character, style, manner, doctrines, merits, &c. Now I think it highly improbable that Thrasyllus could have proceeded upon any such sentiment. For when we survey the catalogue of works which he recognised as genuine, we see that it includes the widest diversity of style, manner, doctrine, purpose, and merits: that the disparate epithets, which he justly applies to discriminate the various dialogues, cannot be generalised so as to leave any intelligible “Platonic character” common to all. Now since Thrasyllus reckoned among the genuine works of Plato, compositions so unlike, and so unequal in merit, as the Republic, Protagoras, Gorgias, Lysis, Parmenidês, Symposion, Philêbus, Menexenus, Leges, Epinomis, Hipparchus, Minos, Theagês, Epistolæ, &c., not to mention a composition obviously unfinished, such as the Kritias — he could have little scruple in believing that Plato also composed the Eryxias, Sisyphus, Demodokus, and Halkyon. These last-mentioned dialogues still exist, and can be appreciated.52 Allowing, for the sake of argument, that we are entitled to assume our own sense of worth as a test of what is really Plato’s composition, it is impossible to deny, that if these dialogues are not worthy of the author of Republic and Protagoras, they are at least worthy of the author of the Leges, Epinomis, Hipparchus, Minos, &c. Accordingly, if the internal sentiment of Thrasyllus did not lead him to reject these last four, neither would it lead him to reject the Eryxias, Sisyphus, and Halkyon. I conclude therefore that if he, and all the other critics whom Diogenes esteemed, agreed in rejecting the ten dialogues as spurious — their verdict depended not upon any internal sentiment, but upon the authority of the Alexandrine library.53

Results as to the trustworthiness of the Thrasyllean Canon.

1. The Canon rests on the authority of the Alexandrine library and its erudite librarians;54 whose written records went back to the days of Ptolemy Soter, and Demetrius Phalereus, within a generation after the death of Plato.

2. The manuscripts of Plato at his death were preserved in the school which he founded; where they continued for more than thirty years under the care of Speusippus and Xenokrates, who possessed personal knowledge of all that Plato had really written. After Xenokrates, they came under the care of Polemon and the succeeding Scholarchs, from whom Demetrius Phalereus probably obtained permission to take copies of them for the nascent museum or library at Alexandria or through whom at least (if he purchased from booksellers) he could easily ascertain which were Plato’s works, and which, if any, were spurious.

3. They were received into that library without any known canonical order, prescribed system, or interdependence essential to their being properly understood. Kallimachus or Aristophanes devised an order of arrangement for themselves, such as they thought suitable.

CHAPTER VII.

The Canon of Thrasyllus continued to be generally acknowledged, by the Neo-Platonists, as well as by Ficinus and the succeeding critics after the revival of learning.

But the scholars of the sixteenth, seventeenth, and eighteenth centuries, speaking generally — though not neglecting these neo-platonic refinements, were disposed to seek out, wherever they could find it, a more literal interpretation of the Platonic text, correctly presented and improved. The next great edition of the works of Plato was published by Serranus and Stephens, in the latter portion of the sixteenth century.

Serranus — his six Syzygies — left the aggregate Canon unchanged, Tennemann — importance assigned to the Phædrus.

Serranus, while modifying the distribution of the Platonic works, left the entire Canon very much as he found it. So it remained throughout the seventeenth and eighteenth centuries: the scholars who devoted themselves to Plato were content with improvement of the text, philological illustration, and citations from the ancient commentators. But the powerful impulse, given by Kant to the speculative mind of Europe during the last quarter of the eighteenth century, materially affected the point of view from which Plato was regarded. Tennemann, both in his System of the Platonic Philosophy, and in dealing with Plato as a portion of his general history of philosophy, applied the doctrines of Kant largely and even excessively to the exposition of ancient doctrines. Much of his comment is instructive, greatly surpassing his predecessors. Without altering the Platonic Canon, he took a new view of the general purposes of Plato, and especially he brought forward the dialogue Phædrus into a prominence which had never before belonged to it, as an index or key-note (ἐνδόσιμον) to the whole Platonic series. Shortly after Tennemann, came Schleiermacher, who introduced a theory of his own, ingenious as well as original, which has given a new turn to all the subsequent Platonic criticism.

Schleiermacher — new theory about the purposes of Plato. One philosophical scheme, conceived by Plato from the beginning — essential order and interdependence of the dialogues, as contributing to the full execution of this scheme. Some dialogues not constituent items in the series, but lying alongside of it. Order of arrangement.

To those who understand Plato, the dialogues themselves reveal (so Schleiermacher affirms) their own essential order of sequence — their own mutual relations of antecedent and consequent. Each presupposes those which go before: each prepares for those which follow. Accordingly, Schleiermacher distributes the Platonic dialogues into three groups: the first, or elementary, beginning with Phædrus, followed by Lysis, Protagoras, Lachês, Charmidês, Euthyphron, Parmenidês: the second, or preparatory, comprising Gorgias, Theætêtus, Menon, Euthydêmus, Kratylus, Sophistês, Politikus, Symposion, Phædon, Philêbus: the third, or constructive, including Republic, Timæus, and Kritias. These groups or files are all supposed to be marshalled under Platonic authority: both the entire files as first, second, third and the dialogues composing each file, carrying their own place in the order, imprinted in visible characters. But to each file, there is attached what Schleiermacher terms an Appendix, containing one or more dialogues, each a composition by itself, and lying not in the series, but alongside of it (Nebenwerke). The Appendix to the first file includes Apologia, Kriton, Ion, Hippias II., Hipparchus, Minos, Alkibiadês II. The Appendix to the second file consists of Theagês, Erastæ, Alkibiadês I., Menexenus, Hippias I., Kleitophon. That of the third file consists of the Leges. The Appendix is not supposed to imply any common positive character in the dialogues which it includes, but simply the negative attribute of not belonging to the main philosophical column, besides a greater harmony with the file to which it is attached than with the other two files. Some dialogues assigned to the Appendixes are considered by Schleiermacher as spurious; some however he treats as compositions on special occasions, or adjuncts to the regular series. To this latter category belong the Apologia, Kriton, and Leges. Schleiermacher considers the Charmidês to have been composed during the time of the Anarchy, B.C. 404: the Phædrus (earliest of all), in Olymp. 93 (B.C. 406), two years before:1 the Lysis, Protagoras, and Lachês, to lie between them in respect of date.

Theory of Ast — he denies the reality of any preconceived scheme — considers the dialogues as distinct philosophical dramas.

Such is the general theory of Schleiermacher, which presents to us Plato in the character of a Demiurgus, contemplating from the first an Idea of philosophy, and constructing a series of dialogues (like a Kosmos of Schleiermacher), with the express purpose of giving embodiment to it as far as practicable. We next come to Ast, who denies this theory altogether. According to Ast, there never was any philosophical system, to the exposition and communication of which each successive dialogue was deliberately intended to contribute: there is no scientific or intentional connection between the dialogues, — no progressive arrangement of first and second, of foundation and superstructure: there is no other unity or connecting principle between them than that which they involve as all emanating from the same age, country, and author, and the same general view of the world (Welt-Ansicht) or critical estimate of man and nature.2 The dialogues are dramatic (Ast affirms), not merely in their external form, but in their internal character: each is in truth a philosophical drama.3 Their purpose is very diverse and many-sided: we mistake if we imagine the philosophical purpose to stand alone. If that were so (Ast argues), how can we explain the fact, that in most of the dialogues there is no philosophical result at all? Nothing but a discussion without definite end, which leaves every point unsettled.4 Plato is poet, artist, philosopher, blended in one. He does not profess to lay down positive opinions. Still less does he proclaim his own opinions as exclusive orthodoxy, to be poured ready-prepared into the minds of recipient pupils. He seeks to urge the pupils to think and investigate for themselves. He employs the form of dialogue, as indispensable to generate in their minds this impulse of active research, and to arm them with the power of pursuing it effectively.5 But each Platonic dialogue is a separate composition in itself, and each of the greater dialogues is a finished and symmetrical whole, like a living organism.6

His order of arrangement. He admits only fourteen dialogues as genuine, rejecting all the rest.

Though Ast differs thus pointedly from Schleiermacher in the enunciation of his general principle, yet he approximates to him more nearly when he comes to detail: for he recognises three classes of dialogues, succeeding each other in a chronological order verifiable (as he thinks) by the dialogues themselves. His first class (in which he declares the poetical and dramatic element to be predominant) consists of Protagoras, Phædrus, Gorgias, Phædon. His second class, distinguished by the dialectic element, includes Theætêtus, Sophistês, Politikus, Parmenidês, Kratylus. His third class, wherein the poetical and dialectic element are found both combined, embraces Philêbus, Symposion, Republic, Timæus, Kritias. These fourteen dialogues, in Ast’s view, constitute the whole of the genuine Platonic works. All the rest he pronounces to be spurious. He rejects Leges, Epinomis, Menon, Euthydêmus, Lachês, Charmidês, Lysis, Alkibiades I. and II., Hippias I. and II., Ion, Erastæ, Theages, Kleitophon, Apologia, Kriton, Minos, Epistolæ — together with all the other dialogues which were rejected in antiquity by Thrasyllus. Lastly, Ast considers the Protagoras to have been composed in 408 B.C., when Plato was not more than 21 years of age — the Phædrus in 407 B.C. — the Gorgias in 404 B.C.7

Socher agrees with Ast in denying preconceived scheme — his arrangement of the dialogues, differing from both Ast and Schleiermacher — he rejects as spurious Parmenidês, Sophistês, Politikus, Kritias, with many others.

Socher agrees with Ast in rejecting the fundamental hypothesis of Schleiermacher — that of a preconceived scheme systematically worked out by Plato. But on many points he differs from Ast no less than from Schleiermacher. He assigns the earliest Platonic composition (which he supposes to be Theagês), to a date preceding the battle of Arginusæ, in 406 B.C., when Plato was about 22-23 years of age.8 Assuming it is certain that Plato composed dialogues during the lifetime of Sokrates, he conceives that the earliest of them would naturally be the most purely Sokratic in respect of theme, as well as the least copious, comprehensive, and ideal, in manner of handling. During the six and a half years between the battle of Arginusæ and the death of Sokrates, Socher registers the following succession of Platonic compositions: Theagês, Lachês, Hippias II., Alkibiadês I., Dialogus de Virtute (usually printed with the spurious, but supposed by Socher to be a sort of preparatory sketch for the Menon), Menon, Kratylus, Euthyphron. These three last he supposes to precede very shortly the death of Sokrates. After that event, and very shortly after, were composed the Apologia, Kriton, and Phædon.

These eleven dialogues fill up what Socher regards as the first period of Plato’s life, ending when he was somewhat more than thirty years of age. The second period extends to the commencement of his teaching at the Academy, when about 41 or 42 years old (B.C. 386). In this second period were composed Ion, Euthydêmus, Hippias I, Protagoras, Theætêtus, Gorgias, Philêbus — in the order here set forth. During the third period of Plato’s life, continuing until he was 65 or more, he composed Phædrus, Menexenus, Symposion, Republic, Timæus. To the fourth and last period, that of extreme old age, belongs the composition of the Leges.9

Socher rejects as spurious Hipparchus, Minos, Kleitophon, Alkibiadês II., Erastæ, Epinomis, Epistolæ, Parmenidês, Sophistês, Politikus, Kritias: also Charmidês, and Lysis, these two last however not quite so decisively.

Schleiermacher and Ast both consider Phædrus and Protagoras as early compositions — Socher puts Protagoras into the second period, Phædrus into the third.

K. F. Hermann — Stallbaum — both of them consider the Phædrus as a late dialogue — both of them deny preconceived order and system — their arrangements of the dialogues — they admit new and varying philosophical points of view.

From Schleiermacher, Ast, and Socher, we pass to K. F. Hermann10 — and to Stallbaum, who has prefixed Prolegomena to his edition of each dialogue. Both these critics protest against Socher’s rejection of the four dialogues last indicated: but they agree with Socher and Ast in denying the reality of any preconceived system, present to Plato’s mind in his first dialogue, and advanced by regular steps throughout each of the succeeding dialogues. The polemical tone of K. F. Hermann against this theory, and against Schleiermacher, its author, is strenuous and even unwarrantably bitter.11 Especially the position laid down by Schleiermacher — that Phædrus is the earliest of Plato’s dialogues, written when he was 22 or 23 years of age, and that the general system presiding over all the future dialogues is indicated therein as even then present to his mind, afterwards to be worked out — is controverted by Hermann and Stallbaum no less than by Ast and Socher. All three concur in the tripartite distribution of the life of Plato. But Hermann thinks that Plato acquired gradually and successively, new points of view, with enlarged philosophical development: and that the dialogues as successively composed are expressions of these varying phases. Moreover, Hermann thinks that such variations in Plato’s philosophy may be accounted for by external circumstances. He reckons Plato’s first period as ending with the death of Sokrates, or rather at an epoch not long after the death of Sokrates: the second as ending with the commencement of Plato’s teaching at the Academy, after his return from Sicily — about 385 B.C.: the third, as extending from thence to his old age. To the first, or Sokratic stadium, Hermann assigns the smaller dialogues: the earliest of which he declares to be — Hippias II., Ion, Alkibiadês I., Lysis, Charmidês, Lachês: after which come Protagoras and Euthydêmus, wherein the batteries are opened against the Sophists, shortly before the death of Sokrates. Immediately after the last mentioned event, come a series of dialogues reflecting the strong and fresh impression left by it upon Plato’s mind — Apologia, Kriton, Gorgias, Euthyphron, Menon, Hippias I. — occupying a sort of transition stage between the first and the second period. We now enter upon the second or dialectic period; passed by Plato greatly at Megara, and influenced by the philosophical intercourse which he there enjoyed, and characterised by the composition of Theætêtus, Kratylus, Sophistês, Politikus, Parmenidês.12 To the third, or constructive period, greatly determined by the influence of the Pythagorean philosophy, belong Phædrus, Menexenus, Symposion, Phædon, Philêbus, Republic, Timæus, Kritias: a series composed during Plato’s teaching at the Academy, and commencing with Phædrus, which last Hermann considers to be a sort of (Antritts-Programme) inauguratory composition for the opening of his school of oral discourse or colloquy. Lastly, during the final years of the philosopher, after all the three periods, come the Leges or treatise de Legibus: placed by itself as the composition of his old age.