SOURCE: "Callimachus and the Generation of his Pupils," in History of Classical Scholarship: From the Beginning to the End of the Hellenistic Age, Oxford at the Clarendon Press, 1968, pp. 123-51.

[One of the most-cited Callimachus scholars, Pfeiffer presents an in-depth study of ancient Greek scholarship in his History of Classical Scholarship: From the Beginning to the End of the Hellenistic Age. The chapter excerpted below looks at Callimachus as a primary contributor to the scholarship, especially in his role as cataloguer for the Alexandrian royal library. Pfeiffer also offers a detailed view of the Pinakes, or bibliographies, Callimachus prepared for the library, and considers their impact on Callimachus's poetry.]

There was no distinguished textual critic in the generation after Zenodotus; only Aristophanes of Byzantium at the end of the third century was his equal if not his superior in this field. The outstanding representatives of scholarship between Zenodotus and Aristophanes were two men from Cyrene, Callimachus and Eratosthenes.

After Alexander's death Ptolemy I ruled over the old Dorian colony of Cyrene as the western part of his Egyptian kingdom (perhaps 322 B.C.); then his stepson Magas was given a kind of independent regency (about 300 B.C.?), and there was a time of considerable trouble between Egypt and Cyrene in the seventies. But at length the only daughter of Magas and Apame, Berenice, was betrothed to the son of Ptolemy II, and on their marriage and accession in 247/6 B.C. Cyrene was finally united with Egypt. Although we cannot fix a precise date for the arrival of the two Cyreneans in Alexandria, there is no doubt that it was after the lonians had started the 'new movement'. For literary men were attracted, not all at once—but in the course of several generations—by the splendour of the new capital and the patronage of its kings. Callimachus' Encomion on Sosibius (fr. 384) may have been one of his earliest elegiac poems, written under Ptolemy I in Alexandria; the only well-attested facts are that he celebrated the marriage of Ptolemy II to his sister Arsinoe (between 278 and 273, perhaps 276/5 B.C.) by an epic, and the apotheosis of the queen (shortly after July 270 B.C.) by a lyric poem. This was apparently in the prime of his life; towards its end he composed the Lock of Berenice (246/5 B.C.) in honour of the Cyrenean princess recently married to Ptolemy III. It was this king who sent for the other native of Cyrene, Eratosthenes, called a 'pupil' of Callimachus, to be librarian and probably tutor to his son. Both the Cyreneans, very different from each other in age and spirit, seem to have been peculiar favourites of the young royal pair.

There is a complete unity of the creative poet and the reflective scholar in Callimachus. We found this combination first in Philitas. Between him and Callimachus, however, Zenodotus had made a contribution of a new kind to scholarship, and institutions for its promotion had been founded by the kings and especially favoured by a king who was the pupil of Philitas and Zenodotus; so the younger generation started from a better position and was enabled to reach a higher degree of that unity than the older one. There is every reason to believe that Callimachus began to write poetry in his early years in Cyrene. We read on Cyrenean coins of the end of the fourth and the beginning of the third century the same names of members of a noble family as in one of his epigrams in which he mourned their misfortunes. He was apparently still in his mother country when, as he tells us himself, he first put a writing tablet on his knees, and the Lycian Apollo addressed him as 'poet' and 'dear friend' and advised him on the art of poetry. A few lines later he implies that he is one of those 'on whom the Muses have not looked askance in their childhood'. In the poem to his greatest poetical achievement, the four books of the Aitia, he pictures himself transferred in dream from 'Libya' to Mount Helicon 'when his beard was just sprouting'; and 'Libya'—supposing that the anonymous epigram quotes him exactly—can mean Cyrene more easily than Alexandria. When and why he left Cyrene for Alexandria we do not know; we are only told that he started modestly as a schoolmaster in a suburb of the Egyptian capital called Eleusis. This may have been under Ptolemy l; since in the seventies, during the reign of Ptolemy 11 and his sister Arsinoe, Callimachus already moved in the court circle, celebrating royalty in the two poems we have mentioned, and he was probably still a 'young man' of the court when he was given a responsible commission in the royal library. This swift career seems to have been due entirely to the extraordinary gifts of a masterful personality.

Callimachus' poems, in spite of their novelty, were informed by an exact and wide knowledge of the earlier poetry from which he drew his models. Practising his craft and reflecting on it went together. This reflection quite naturally extended to the literature of the past, to all the various forms of metre and language, and to the recondite sources of its subject-matter. Only the most passionate study could result in exquisite poetical workmanship, and only boundless curiosity could open the untrodden ways (fr. 1. 28) to new fields of learning. Ironically the poet hints at the danger of 'much knowledge' … in certain cases; on the other hand, the mere pleasure of listening and learning is to him the least perishable of pleasures in human life.

Two points should be kept in mind. If his verse very often sounds like charming word-play, the poet is never tired of reminding us that everything he is going to tell is true because it is well attested …; the Muses, who once taught Hesiod and now answer Callimachus' questions, always utter the truth. In another case he refers to a local writer by name (fr. 75. 54) as his reliable source. In speaking of 'recondite sources', 'reliable source', we apply this word, which originally means the fountain of a stream or a river, figuratively to literature. In the beautiful finale of Callimachus' hymn to Apollo (hy. 11 108-12) the god contrasts the filthy water of a great river with the clear droplets the bees carry to Demeter from the pure and undefiled fountain-head. In these metaphorical lines spoken by Apollo the poet condemns the lengthy traditional poem with its conventional formulae, but praises brevity and novelty in verse. This meaning is quite obvious. But there seems to be implied another piece of advice, hardly recognized by modern interpreters of the hymn: poets should draw from the original pure source, not from its polluted derivatives. Callimachus was, as far as I can see, the first to use this image in a literary sense. This demand of the scholar poet applies equally to poetry and to scholarship. It became a favourite image in the age of humanism and a fundamental concept of scholarship in the modern world.

If we consider Callimachus' general attitude, occasionally revealed in some lines of his poems, the remarkable feat of scholarship that he achieved in the library is perhaps not quite incomprehensible. His task was to find a system for arranging the texts of all the writers collected for the first time in the royal library (or libraries). When we glanced at the prehistory and early history of script and book in Greece, we observed the oriental background and commented cautiously on the relations between the orient and Greece. Now in Alexandria a Greek library was founded on a grand scale; and this reminds us of the enormous Babylonian and Assyrian libraries of old. It is natural to inquire whether there may have been direct influence, since the door of the east had been opened by Alexander much wider than before, and recent research has at least put this question more urgently; but the answer so far is not very definite. The layout of the papyrus-rolls in the Alexandrian library seems to have resembled that of the clay tablets in the oriental libraries in one or perhaps two significant points. The title of a work was regularly placed at the end of the roll and of the tablet (in contrast for instance to the practice in the Egyptian papyri), and in 'catalogues' not only this title, but also the 'incipit' was cited. On tablets and rolls the number of lines was occasionally counted, and these 'stichometrical' figures were put at the end and sometimes as running figures in the margins; they appear again in library-catalogues. The earliest example of title and number of lines placed at the end of a roll turned up in a recent publication of Menander's Sicyonius; the date of the papyrus seems to be the last third of the third century B.C., very near to Callimachus' lifetime. Even a personal remark of the scribe in verse is added, and these notes altogether may be properly called a 'colophon'. There is very scanty evidence for libraries in the Ionic and Attic periods; but the same technical devices as in the east, or similar ones, may have been used in Greek private houses or in philosophical schools.

Whatever may have been achieved before the third century B.C., Callimachus had no real model for his immense undertaking. Though his task was probably not so much to create as to develop an appropriate method, he did it so successfully that his 'lists', called IIίνακες, were generally acknowledged as a model for the future. Besides the Pinakes, he assembled a variety of learned material helpful for the understanding of the ancient texts and invaluable for the writing of poetry in the new style; in these books he resumed the labours of the younger Sophists and the Peripatos with a new purpose.

For the IIίνακες Tzetzes is again our authority; after giving the number of books in the two libraries he goes on to say: ων τον̀ς πίναλας ν̀́στερον Kαλλίμαχoς απεγράψατο. This sentence is slightly enlarged in another later version: ώς (?) ó Kαλλίμαχoς νεείσκoς ὼ́ν τη̂ς αψ́λη̂ς ψ́στέρως μετὰ τὴν άνóπθωσιν τψ̀ς πίνακας αν́τω̂ν άπεγράψατo; then a reference to Eratosthenes follows and finally the remark: άλλά τά Kαλλιμάχoν καὶ τον̂ 'Eρατοσθέψοψς μετά βραχν́ν τινα χρóνον έγένετo τη̂ς σμναγωγη̂ς τω̂ν βίβλων, ώς ὲ́ϰην, καὶ διορθώσενς κὰ́ν έπ αν́τον̂ τον̂ IIτολεμαίον τον̂ φιλαδέλϰον. Obviously it is the sequence of events that is stressed in both versions of the Prolegomena: ν̀́στερον—ν́στέρως μετὰ τ.ά.—μετὰ βρξν́ν τινα ξρóνον. Therefore the change of ν̀στέρως to ιστορει̂ ò́ς, proposed by Dziatzko and accepted by most modern editors, is not justified. This conjecture would enormously enhance the authority of Tzetzes' report, as it makes Callimachus himself the ultimate source of at least a part of the Prolegomena. The unfortunate Italian humanist had no scruples about producing the following 'translation' on the margin of his Plautine codex: 'sicuti refers Callimacus aulicus regius bibliothecarius qui etiam singulis voluminibus titulos inscripsit.' Hinc illae lacrimae. Here we have Callimachus not only quoted as a literary authority, but also elevated to the official rank of court-librarian; there is no evidence that he held this position except this slip of the 'translator', and there is not even room for him in the well-known series of librarians.

Tzetzes apparently had in mind a sort of catalogue of books extant in the library. Hesychius-Suidas' biographical article Kαλλίμαχoς, once probably the introduction to an edition of Callimachus' collected poems (of which therefore very few titles were mentioned in the biography) points to a comprehensive 'bibliography': … 'Tables of all those who were eminent in any kind of literature and of their writings in 120 books'. The previous generation had done some quite respectable scholarly work in the library at least on the foremost poets, without waiting for catalogues and bibliographies, and this may have been very helpful now in the completion of the Pinakes. It is Suidas' description—as we should expect from his much better sources—not Tzetzes' that is correct; this is confirmed by the fragments still preserved. The distinction between a mere library catalogue and a critical inventory of Greek literature is sometimes obscured in modern literature on Callimachus' great work; it was certainly based on his knowledge of the books available in the library, but he also had regard to works only mentioned in earlier literature and to questions of authenticity.

The entire body of Greek literature, the πα̂σα παιδεειία, was divided into several classes: only three are attested by verbal quotations: ρητορικά (fr. 430-2, cf. 443-8), νóμι (fr. 433), παντοδαπὰ σνγγράμμαατα (fr. 434/5). From references to epic (fr. 452/3), lyric (fr. 441, 450), tragic (fr. 449?, 451), comic poets (fr. 439/40), to philosophers (fr. 438?, 442), historians (fr. 437), and medical writers (fr. 429?) registered in the Pinakes we may conclude that seven further classes existed; there were probably many more and a number of subdivisions. It is now fairly certain that the individual authors of every class were arranged in alphabetical order; each name was accompanied by a few biographical details, and later writers were sometimes disappointed by what they considered deficiencies (fr. 447). Less conscientious, even sensational, the vast biographical work of Hermippus of Smyrna, who is called 'peripateticus' as well as KαλλιμPχειoς, be regarded as a more popular supplement to the esoteric Pinakes. But we may doubt if his master Callimachus liked it; he had confined himself to the reliable evidence for the lives and works of literary men. The list of their writings which followed the biography cannot always have been arranged in the same way, but the alphabetical system seems to have prevailed. The little we know of some minor epics and all we know of the dramatic poems leads to this assumption, if indeed the order in the lists of later antiquity is derived from the Pinakes. The best example is the famous κατάλογoς τω̂ν Aισχν́λον δραμάτων which was once obviously an appendix to the life of the poet and still presents to us the titles of seventy-three plays, tragedies, and satyr-plays, in strictly alphabetical order. For Euripides there were only fragments of two inscriptional catalogues, until recently published papyri brought very welcome new evidence for titles of his plays arranged in order of the initial letter. In the most important of these papyri, which gives summaries of the plots, the title is followed by the formula οὗ (ἧς, ὧν) ὰρχή and the citation of the first line. This 'incipit' had been introduced by Callimachus in his Pinakes, for instance: έπικòν δὲ τò ποίημα, οὗ ή ὰρχή, followed by the opening verse of the poem (fr. 436). A mere title might have been ambiguous, particularly in the case of prose writings; the 'incipit' made the identification easier. A list similar to that of Aeschylus' plays is preserved in two manuscripts of Aristophanes, where brief details of his life are followed by an alphabetical catalogue of his comedies. Menander may have found a place in the Pinakes, like Alexis (fr. 439) and Diphilus (fr. 440); for the beginning of an alphabetical list of his plays (titles only) is preserved in a papyrus.

Cataloguing the lyric poetry … must have presented thorny problems. Callimachus divided the great triadic poems (which we usually call 'choral', although we often cannot tell whether they were actually sung by a choir) into special groups.… Simonides' songs of victory, for instance, were called έπίνικοι and subdivided according to the type of contest (foot-race, pentathlon, etc.); for we know that Callimachus (fr. 441) had described a part of the Epinicia as … 'for runners'. Pindar's Epinicia must also have been divided into several groups, but in a different way, according to the place of the contests (Olympia, Nemea, etc.); otherwise it would not have been possible to say that Callimachus (fr. 450) placed the second Pythian ode, as it was later named although it actually celebrated a local victory of Hieron, among the Nemean odes. Finally Bacchylides' dithyrambs seem to have been separated from his paeans; for Callimachus was blamed for having entered among the paeans a poem which Aristarchus declared to be a dithyramb and entitled Cassandra. We know these classifications from references to the editions of lyric texts, begun by Aristophanes of Byzantium, and from later grammatical sources; but it is easily forgotten that some fundamental terms and formulae were, if not coined, at least first attested in the Callimachean Pinakes. Although we can recognize certain groups of choral songs, we still cannot guess how the individual poems were arranged. No doubt they were registered somehow, as the references to Pindar's second Pythian ode and Bacchylides' Cassandra demonstrate. Because the dithyrambs and probably the νóμοι had titles like the plays, they could easily have been listed in alphabetical order. But what about all the others, especially the monostrophic poems of Sappho, Alcaeus, Anacreon? They had no titles; and therefore the only way to register them, it seems, was according to the 'incipit', a method still applied in modern indexes of lyric poems of an author or of an anthology.

Callimachus lavished his efforts also upon the classification of the prose writers; the different classes, as far as we can make out their names, have been noted above. In principle the arrangement followed the same lines as in the poetical section; but the difficulties were greater than in the Pinakes of the poets, as the case of Prodicus shows; Callimachus listed him with the orators (quite correctly, I should say), but others objected that he belonged to the philosophers (fr. 431). The names of writers in every class were given in the usual alphabetical order (fr. 435). The works of each author may have been subdivided into several groups, such as public and private speeches; subdivision was unavoidable in the case of πoλνείδεια (fr. 449; 429?). Individual speeches that had titles, for instance IIερί 'Aλοννήσον (fr. 443), IIερί τω̂ν σνμμοριω̂ν (fr. 432) of Demosthenes, or IIερί φερενίκον (fr.448) of Lysias, could be listed alphabetically, though the 'incipit' was usually added (fr. 443, 444). But in cases where there was no title, or where the authorship of speeches (fr. 444-7) or whole books (fr. 437) was a matter of dispute, we have no clue to the arrangement. It seems to be over-optimistic to see in the famous complete list of Theophrastus' writings (Diog. L. v 42-50) a sort of enlarged copy of Callimachus' Pinakes; the very complicated tradition does not recommend this simple solution. Neither can we trace the list of Aristotle's writings (Diog. L. v 22-27) back to Callimachus as the ultimate source. As regards the philosophers in the Pinakes our knowledge is deplorably poor. Our information is much more precise on the 'Miscellanea' (… fr. 434-5); under this heading a number of writings were registered that did not fit into the main categories of literature. For instance, Athenaeus in his Deipnosophists, being particularly concerned with books about 'dinners', has preserved an excerpt with the name of the author of one (a wellknown Athenian parasite), the title, the opening words and the number of lines (fr. 434), and another excerpt containing an alphabetical list of four writers on pastry cooking (fr. 435). The intention was clearly to omit nothing from this inventory of the πα̂σα παιδεία, not even books on cookery.

Besides the general Pinakes two special ones are known that differ totally from the main work in being one chronological, one linguistic. Both titles are extant only in Suidas' article. The first is … 'Table and register of the dramatic poets in chronological order and from the beginning'. This Pinax must have been based on Aristotle's διδασκαλίατ taken from the documents in the archon's archives. Alexander Aetolus and Lycophron had busied themselves with the tragic and comic texts in the Alexandrian library in the early third century; in its second half Eratosthenes and Aristophanes devoted major works to the Attic drama. Between them Callimachus compiled his record, the great scale of which we can still guess from fragments of three inscriptions found in Rome, where they had probably occupied a wall in a great library. Körte's suggestion that the inscriptions are a more or less exact apographon of the Callimachean Pinax has been universally accepted. The parts preserved enumerate the Dionysiac and Lenaian victories of comic poets from 440 to 352 B.C.; but if the title given by Suidas is correct the Pinax extended back to the άρχή, that is, to the introduction of comedy into each of the two festivals, the City Dionysia in 486 B.C. and the Lenaia in 442 B.C. The second special Pinax was apparently a list of glosses, and it is not surprising to find Callimachus following Philitas and Zenodotus as glossographer; what surprises is the wording in Suidas: IIίναξ τω̂ν Δημοκράτονς γλωσσω̂ν ααὶ σuνταγμάτων (?). Whatever is meant by σuντάγματα (probably 'writings'), its connexion with γλω̂σσαι is strange, as 'a list of writings' should belong to the great general Pinakes. It is, of course, easy to change the proper name to Δνμοκρίτον. Democritus was a bold innovator in the language of philosophy, but it can hardly be said that his own language is distinguished by obsolete words. We must also remember that he wrote something himself on Homer's language and his glosses, although only the title remains, as in the case of Callimachus' Democritean Pinax.

We have taken pains to call attention to many dry and sometimes baffling titles. Inconspicuous as the individual headings may look, the impression of the whole is overwhelming. To amass hundreds of thousands of rolls in the library would have been of little use without a sensible classification that enabled the prospective reader to find the books he needed. For the first time in history the Pinakes of Callimachus made the greatest treasures of literature accessible by dividing poetry and prose books into appropriate classes and by listing the authors in alphabetical order. Only the most passionate desire to save the complete literary heritage of the past from oblivion and to make it a permanent and fruitful possession for all ages could have provided strength and patience for this immense effort. Querulous critics of the scholar poets, Philitas, Callimachus, and their followers in ancient and modern times, may carp at the excessive learning of their poetry and at the amateurish deficiencies of their scholarship. But they should not undervalue the fervent devotion to learning that sprang from the enthusiasm of a great poet.

No doubt the 120 books of the Pinakes gave plenty of scope for additions and corrections; even our short quotations have revealed this again and again. Aristophanes of Byzantium published a whole book IIρòς τον̀ς Kαλλιμάχoν πίνααας. IIρóς is ambiguous and often means 'against' in titles, but there is not the slightest reason to assume that Aristophanes ever wrote 'Against Callimachus' Pinakes'; his book was meant to be a supplement, which certainly was very welcome about fifty years afterwards, and he made use of Callimachus' chronological tables of the Attic dramatists for the summaries of plays in his editions. This was the immediate effect; but everyone who needed biographical material, who undertook editions of texts, who wrote on any literary subject had to consult the great work; it has never been superseded by a better one. The anonymous IIίνακες of the rival library in Pergamum, very rarely quoted, once for a comic poet and twice for orators, did not compare in importance with the Alexandrian Pinakes of Callimachus upon which they were probably modelled.

A number of titles, some of them found only in Suidas' article, and some short quotations give an idea of the variety of learned books published under Callimachus' name; in preparing them he may have been assisted by friends and pupils. A throng of students was drawn to Alexandria by the new longing for unlimited knowledge and the fact that incomparably richer material was now offered there than ever before in Athens or elsewhere. The Sophists had had epideictic-oratorical aims in their treatment of literary, especially poetical, subjects, and the great Attic philosophers and their schools had had their philosophical purposes. Now for the first time we find wide literary knowledge being acquired for the sake of the literary tradition itself, that is, for the works to be written in the present age and for the preservation and understanding of the works written in past ages. This is the new separate discipline of scholarship.

The books of Callimachus the scholar … are often regarded as mechanical compilations of antiquities. As a matter of fact they are not restricted at all to antiquarian matter; we can apply our old scheme to them, though perhaps in a different sequence, briefly reviewing his books on antiquities, on language, and on literary criticism, and finally considering how far he may be regarded as an interpreter of earlier Greek poetry.

The Nóμτμα βαρβαρτκά were an antiquarian collection of 'Non-Greek Customs', possibly supplementing Aristotle's book with the same title. A general book IIερὶ άγώνων probably belongs to the same group, since some of the Sophists and Aristotle and his school frequently compiled material 'On games'. The forty-four excerpts in Antigonus of Carystus, Hist. mirab. 129-73, from Callimachus' IIαράδoξα show him as a writer on marvels; his keen curiosity for 'Incredibilities' led him to make this Collection of marvels in all the earth according to localities from historical, geographical, and antiquarian sources. There is no earlier example of paradoxagraphy as a distinct literary genre. Like Philitas and Zenodotus he was not scientifically minded, as this work reveals better than any other; there is no recognizable intercourse between science and scholarship in Alexandria before Eratosthenes.

From a book entitled … Local nomenclature, special names for fishes in different cities (Chalcedon, Thurii, Athens) are quoted; as there was a chapter on fishes, the arrangement of the whole must have been by subjects. Though unproven, it is not impossible that the titles IIερὶ άνέμωά (fr. 404), IIερὶ óρνέων (fr. 414-28), Mηνω̂ν πρσσηγορίαι κατὰ ὲ́θνoς καὶ πóλεις (p. 339, Local Month-names) are only the sub-titles of other chapters in the same comprehensive Onomastikon. This vocabulary was certainly not arranged in alphabetical order like Zenodotus' Glossai. The relation of names to things was a philosophical problem, discussed at length in Plato's Cratylus and also by Aristotle. But Callimachus listed and disposed all the names he could find for the purely literary reasons which we have just stated; it was the first vocabulary of its kind, as far as we know, and was eagerly used by Aristophanes of Byzantium and later generations. It can hardly be decided whether works entitled Kτίσεις νήσων καὶ μετονομαίαι (p. 339) and IIερὶ τω̂ν έν τη̂ oικονμένη πoταμω̂ν (fr. 457-9) belong to the books on antiquities or to the books on language; 'changes of name' rather point to the second group. There remain a few headings and fragments for which we are completely at a loss to find a place, or even to understand the titles. But the important fact is that we are able to find traces of nearly all the learned collections of Callimachus in his poems: fair sounding names of rivers and islands, of winds and nymphs and birds were picked out of them to embellish the verses, and a number of fine local stories was found in them and saved from oblivion.

One book has been left out of this cursory survey, Callimachus' Against Praxiphanes … (fr. 460); we mentioned it earlier, when we were pointing out non-Aristotelian features in the whole new movement in Alexandria. The only fragment quoted from this book is clear evidence of literary criticism in so far as it asserts the high poetical qualities of the work of his contemporary Aratus.… As we know of no other similar book by Callimachus, the polemics against the Peripatetic Praxiphanes may have included both his judgement on Plato's incompetence as a literary critic, (the more so as Praxiphanes' IIερὶ πoιητω̂ν was a dialogue between Plato and Isocrates), and also his famous maxim: τò μέγα βιβλίον ἶσον τω̂ μεγάλω κακω̂. Whatever βιβηίον here means, μέγα κακóν, a 'great evil', is a sort of old formula (O 134, ι 423), and μέγας with reference to literature is always vituperative; we may compare the filthy μέγας ρóoς in contrast to the pure óλίγη λιβάς (hy. II 108), or the μεγάλη γυνή of a poem contrasted with subtle small scale ones, κατὰ λεπτóν (fr. 1. 12). As in the case of Aratus the statement in the prose book has its exact parallel in an epigram, so also there are obvious parallels in the poems to the two other passages tentatively ascribed to the same prose writing. Plato was deemed an incompetent critic …; the reason was that he appreciated the poetry of Antimachus, whose Lyde Callimachus condemned in an epigram (fr. 398) as 'a fat and not lucid book'. The general disapproval of the μέγα βιβλίον uttered in the prose maxim is a common topic in Callimachus' poems and is the particular theme of his introductory elegy to the Aitia against his adversaries, whom he calls 'Telchines'.

A list of these adversaries compiled by a learned scholiast includes the name of Praxiphanes the Mytilenean; this is invaluable evidence for the opposition between the poet and a leading Peripatetic and shows that the ambiguous IIρóς in the title means 'against Praxiphanes'. There is no tradition that Praxiphanes had personally attacked Callimachus in his writings. The learned collections and also the Pinakes may give the impression of being rather Aristotelian in subject-matter, despite their new purpose; but in literary criticism Aristotle's theory and Callimachus' views are plainly incompatible. As the one relevant prose book is almost lost, we have to rely mainly on the poems. Again and again, charmingly as well as firmly, he put forward his clear and consistent opinions. He is never pedantic, but rather humorous and ironical or even of a lively aggressive spirit. Aristotle, we remember, in the severest of styles demanded organic unity of every artistic work: ὲ́ν, ò́λον, τέλoς, μέγεθoς were the decisive terms. All parts must have a definite relation to the whole work, which itself is distinguished by completeness and magnitude. The Iliad and Odyssey, but not the other epics, are living organisms of this kind; they and the masterpieces of Attic tragedy alone fulfil these requirements. If it were possible for any further poetical works to be produced at all, they must somehow conform to this standard prescribed by Aristotle. Now Callimachus regarded Homer with the same devotion and affection as Aristotle had done, in contrast to everything 'cyclic' (Ep. 28), which lacked organic unity, but abounded in traditional formulae. For that very reason he esteemed Homer inimitable, even unapproachable. It would be a vain ambition to vie with him and the other great poets of the past; if poetry lived on, it was bound to follow principles quite different from those inferred by Aristotle from the ancient poems.… For years poetical criticism had been in the hands of Sophists and theorizing philosophers; the time had come for a return to its originators, the practicing poets.

The new poetical school of Callimachus and his followers was ostentatiously anti-Aristotelian. Rejecting unity, completeness, and magnitude, it consciously aimed at a discontinuous form (fr. 1.3 …) in a more or less loose series of pieces of a few lines (fr. 1.9 …). The proper quality of a poem was to be … 'subtle'. It has been rightly noticed that this key term and a few other ones had already occurred in Aristophanes' comedies, especially in the critical passages of the Frogs: τέχνη / [κρίνετε]… τὴν σοϰίην (Call. fr. 1. 17 f.) is almost a verbal quotation. But the truth of this observation was overlaid by two hypotheses: namely that Aristophanes borrowed his phrases from a Sophistic source, probably Gorgias, and that Callimachus used a rhetorical source on the genera dicendi. No proof of these hypotheses has yet been produced; they remain a strange but typical example of the modern quest for hidden sources. The natural assumption is that the Hellenistic poets derived their critical terminology directly from the poets of the fifth century, whom they knew so well. Substantial parts of Callimachus' Iambi are indebted to Attic comedy; there is no need to invent intermediate handbooks. The meaning of the word λεπτóς underwent a characteristic change; while it was once used disapprovingly of over-refinement of spirit or diction, for instance, that of Euripides in contrast to the vigour that Aeschylus achieved through the magnitude … of his words, the Alexandrians, Callimachus, Hedylus, Leonidas, employed it as a term of the highest praise to describe the style they were eager to achieve in their poems. We find another significant epithet in the Praxiphanes pamphlet, where Aratus was praised as a poet of the highest rank: πολνμαθής. 'Much learning' was in archaic times a reproach against those who had no true wisdom; but this word also came to have the opposite connotation in the Hellenistic age; unlimited knowledge of subject-matter and language was now deemed an indispensable requisite for the new poetry called σοϰίη (Call. fr. 1. 18).

Looking back on Callimachus' own πoλνμαθίη amassed in his prose works, we may ask whether they can be assigned to a particular epoch of his life. When the epilogue of the Aitia came to light, the first editor saw in the concluding line Callimachus' 'formal farewell to poetry' and a declaration 'that he will now devote himself to prose'; indeed his appointment at the Alexandrian library was regarded as the point in his career at which he turned from poetry to prose. But αν́τὰρ ὲγὼ Moνσέων πεζòν [ὲ́]πετμι νομóν indicates the Musa pedestris of the Iambi which followed the Aitia in the final edition arranged by the poet himself; the pentameter gives no answer to this or any other question of chronology.

When we divided Callimachus' prose works into three groups, on antiquities, on language, and on literary criticism, we asked whether there was not a fourth one on interpretation. As far as we know, he never edited a text or wrote a commentary; the few fragments of his 'γπoμνήματα seem to indicate a collection of mythological, linguistic, and geographical material. But in many passages of his poems he discloses his acquaintance with the Iliad and Odyssey and occasionally allows us to guess not only what text he chose but also how he understood its meaning. In this sense only he may with reserve be called an 'interpreter' of Homer.

First of all we should like to know how far Callimachus used Zenodotus' new critical edition of Homer and how far he relied on pre-critical texts, … such as Timon recommended to Aratus. Several Callimachean readings of the Homeric text seem to agree with those known to us only as Zenodotean. The beautiful Naxian girl, Cydippe, took part in 'the dance of sleeping Ariede, 'Aριήδης / ὲς ξ]oρòν εν́ρον́σης Callimachus tells us (fr. 67. 13); in the famous Homeric passage to which he alludes, ξoρóν … οἶον … Δαίδαλoς ὴ́σκησεν… 'Aριάδνη (E 592), only Zenodotus read 'Aριήδη. This certainly is a most remarkable coincidence; but as Zenodotus constituted his text on earlier manuscripts that he found reliable, the same sources may have been accessible to Callimachus. Although it is possible or even very probable that he followed Zenodotus, the coincidence in this and about ten similar cases is not conclusive proof. At least one example proves that Callimachus also consulted other texts older than the Zenodotean edition: only the 'city-editions' had the unique variant reading νήσων ὲ́πι θηλντεράων (φ 454 and X 45), from which he transferred the epithet to another noun, θηλν́τατον πεδίον (fr. 548), 'the most fertile plain'. By connecting θηλψ́τατον with πεδίον Callimachus gave his 'interpretation' of the Homeric phrase: it does not mean 'island where females reign', like Lemnos and Imbros, but 'island that is … with good soil, fertile.' It is possible that he consulted the elementary explanatory notes that must have accompanied the Homeric text for a long time and finally became a substantial part of our so-called D-Scholia, in which they were mixed up with more learned grammatical comments. When he took over τοι̂Oς from H 231 in the sense of άγαθóς (fr. 627), his interpretation possibly was in accordance with Aeschylus, certainly with the glossographers; when he called a messenger άπον́ατoς (fr. 315), he was induced by whatever source not to read άπ' oν̀́ατoς in E 272, but a compound meaning 'bringing the tidings'. These may be rather odd examples of Homeric epithets; but there are more common epic words too that have been puzzling in all ages. We can still distinguish how Callimachus understood some of these adjectives …, nouns …, or verbs …, or controversial etymologies of proper names.…

We started from the fact that the creative epic poets were their own interpreters and that the rhapsodes continued the self-interpretation of the poets. The Sophists can be regarded as the heirs of the rhapsodes in so far as they tried to explain poetry for their new purpose, and the great Attic philosophers and their schools completed this development. Now once again poets became active in this field; there were no commentaries produced in the first generations of the Hellenistic age, but these poets were the immediate forerunners of the writers of continuous interpretations.… It is from Callimachus and his pupils that a line runs to the true έρμηνεία τω̂ ποιητω̂ν by the Alexandrian γραμματικοί of the following generations. In contrast to them, Crates and his pupils in Pergamum renewed in a way the ancient allegorical method and forced their own philosophical, particularly Stoic, views upon the Homeric and other poems. But the not infrequent quotations from Callimachean verses in our Scholia to Homer show how helpful they were for the other, the scholarly way of interpreting old epic poetry.