BMCR 2003.11.33

Cratete di Mallo: I frammenti. Edizione, introduzione e note

, , I frammenti. Pleiadi: studi sulla letteratura antica ; 2. La Spezia: Agorà, 2001. xciv, 357 pages ; 24 cm.. ISBN 888721834X EUR 30.00..

Crates of Mallos is surely one of the most important and least understood grammatical scholars from antiquity. A contemporary and rival of Aristarchus, the Alexandrian grammarian and celebrated editor of Homer, in later opinion he was remembered as Aristarchus’ equal in the study of Homer. Crates’ base of operations was Pergamum, where he possibly headed the library and (as Erika Simon once speculated) may have contributed to the design of the frieze of the magnificent Altar of Zeus at Pergamum, an idea that has been rejected by subsequent scholars and that depends entirely on Crates’ uncertain credentials as a Stoic.1 Less speculatively, we can say that Crates left his mark on the ancient tradition of Homeric interpretation, which remembered him above all with the nickname Homerikos, not least in honor of his trademark readings of Homer: Crates produced some of the most bizarre allegorical interpretations of the poet known to antiquity. But he wasn’t content to restrict himself to the Poet alone, and in this he showed himself a representative literary scholar of his age. His varied output attests an interest in the major literary genres (epic, lyric poetry, tragedy, and — perhaps, though this is doubted — comedy), but even that wasn’t enough for the busy-minded Crates, who prided himself on having attained to the level of a super-scholar. Philologos he found too constricting a title, so he dubbed himself a kritikos, something evidently higher and nobler, and possibly more philosophical and more original too. Whence his strong interests in grammar and linguistics, where he appears to have played a role in the shadowy analogy/anomaly debates. And whence the theory of poetic judgment and value he seems to have developed (more on this possible connection below).

In 168/7 BCE, while on an embassy to Rome, Crates was obliged to stay on (having broken his leg in a sewer). The result was fateful for the history of Roman letters. Suetonius testifies to his local impact, but one can only assume that what Crates accomplished in person he would have achieved through his writings even had he never come to Rome, given the way Italians began absorbing Greek influences in letters with an ever-increasing avidity in the centuries following and given the obvious place and importance of Crates in that heritage.

To judge from the scatter of ancient testimonia we have, some of which are glowing and others disparaging, Crates cut a controversial figure in antiquity. Gradually, over the centuries and then millennia, he was consigned to the dustbin of minor curiosities. His memory, or at least his name, vanished in all but the most technical ancient literature in antiquity, and by the time modernity discovered him again it was almost too late to rescue him from oblivion. F. A. Wolf regrets that we know as much about Crates’ “wisdom” as we do, but Crates’ stock began to rise as the systematizing urge in scholarship set out to hurl light on even the darkest corners of antiquity. The first stab at a collection of his remains was made in 1836 by C. F. Wegener, which was followed in 1860 by K. Wachsmuth, and then, in two specialized collections on the Homerica, by H. Helck in 1905 and 1914. The first monumental collection was undertaken, however, only in 1936 by H. J. Mette ( Sphairopoiia) and again in 1952 ( Parateresis).2 But useful as these last two volumes are, they are less collections than discursive studies of the fragments, compromised at times by tendentious readings. Nor are these two prior collections complete, and until now anyone working on Crates had to supplement Mette with Wachsmuth and Helck. Thus it happened that in 1968 Rudolph Pfeiffer could still lament the lack of “a complete collection” of the fragments of Crates.3 Now, finally, with Maria Broggiato’s splendid and handsome edition, Pfeiffer’s prayers can be deemed mostly answered. It is unlikely that her collection will be superseded in our lifetime.

Broggiato’s (B.’s) edition has every advantage over its predecessors. It is more complete than all the others, not least thanks to the added material from the latest editions and readings of the Herculaneum papyri (some of it based on her own autopsy). What is more, B.’s edition is distinguished by the eminently thoughtful, judicious, and all in all justifiably conservative manner in which it weighs all the available information we have on Crates and all the scholarship on him up to the last minute. The book contains a strong 69-page introduction that places Crates within the history of scholarship, ancient and modern; a full bibliography of works cited; some 125 pages of testimonia and fragments, replete with comparanda and critical apparatus (but no translations), organized by authors and genres, then by categories (poetics, grammar, the Attic dialect), and finally “Varia” (but no special grouping of the allegorical fragments; see below); some 150 more pages of detailed commentary; and indices. Each of these components is valuable, and equally so. B. offers no new grand theories about how all the various facets of Crates’ thought can be shown to cohere in their larger lineaments or in their troubling details, but she by no means shrinks from labeling the conclusions of her predecessors probable or improbable as she sees fit. And because B. plainly sees her own role to be that of an expositor but not a final arbiter, her edition is an ideal place to go for getting one’s own bearings as one ponders where next to go with one’s readings. For all this care and circumspection of judgment and for her infinite amounts of good-will towards earlier scholarship, B. is to be praised. These features make her edition of Crates the very best to date and guarantees it a long shelf-life to come.

That said, it isn’t at all clear that Pfeiffer’s prayers have been answered, in part because, one supposes, what he was really wishing for was not a more complete edition of the available fragments but simply for more of Crates to become available. Collections of fragments grow in three ways: new fragments are found, having been overlooked; or they are discovered for the first time; or they are retroactively found, or if you like invented. In the last case, one adds to a corpus by rewriting the texts we have, whether through emendation or by conjecture or both, inserting into a text a name or a clue to the fragment’s identity that wasn’t there before.4

In the case of Crates, Eureka -moments have occurred in each of these ways, but less has changed in the state of our evidence since the nineteenth century than one might have hoped. No secret hoards of Cratetean wisdom have yet come to light, with one major possible exception, the fragments from Philodemus mentioned earlier, which haven’t so much come to light as they have been increasingly illuminated by new scholarship. But these latter are slightly troublesome, for although they increase the fragment count by four since Mette and add several pages of columnar text to B.’s edition, the fact remains (although it is not one Philodemeans would readily breathe out loud, myself included) that, with one exception, the name “Crates” never appears unadorned by an editor’s conjectural brackets and/or (often even then) surrounded by dotted letters expressing uncertainty as to the reading of the original ink. And so B. rightly asterisks these accretions, as she does other either uncertain or doubtful attestations or readings of Crates’ name elsewhere in her edition, as for example in the by no means atypical instance of her new F *17: Κράτη]ς. (The ambiguity of the asterisk, familiar from Jacoby, is unfortunate. It might have been eliminated with the use of two distinct sigla, one for doubtful ascriptions, the other for paleographical uncertainty.)

Apart from the scraps from Herculaneum, only two of which Mette cites (and no earlier editor), two further papyrus fragments have come to light since Mette, one edited by Lobel in 1972 and another by Haslam in 1986. And so when all is said and done, of the 143 numbered fragments in B.’s collection, twelve appear in no prior collection. Of these nine are asterisked. And of this subset, two result from supplements (one made in 1899 and rejected by Helck, but included by B. for completeness’ sake, the other made in 1962); two are from Tzetzes, who ascribes a work on comedy to Crates (dubiously, in B.’s opinion, though Wachsmuth, had he seen them, would doubtless have considered them genuine attestations, as he does with two further fragments from Tzetzes asterisked by B.); one is from a scholium to Oppianus of Anazarbus, another doubtful attribution (here B. agrees with Wachsmuth, but includes the fragment among her dubia nonetheless); and four are from the Herculaneum papyri, the value of which needs to be discussed apart. That leaves three new more or less certain additions, two of which were discovered among the Oxyrhynchus papyri in the past thirty years and which no prior editor could possibly have known. In the case of the third novum, B. here catches out her predecessors, reproducing a scholium to the Iliad mentioned in no earlier edition and securely, in her mind, attributable to Crates. So far the fragments. As for the testimonia, these have grown by a few slivers, without however adding any new knowledge to our picture of Crates (several of these are redundant and prompt little or no commentary from B.). Sizing up the situation in purely quantitative terms, one has to conclude that the gains have been meager indeed.

The case of the Herculaneum papyri is worth dwelling on a little longer, for it is here that what is most novel and exciting about our current picture of Crates has spectacularly shown itself. It’s long been known, ever since the Herculaneum papyri were published in the nineteenth century, that Crates’ peculiar brand of physical allegory, his “theory of the sphere” or sphairikos logos, was familiar to Philodemus. Philodemus actually supplies an important notice that firmly ties Crates into the tradition of physical allegoresis, a separate line of allegorical interpretation that contrasts with the moral apologetics that run from Theagenes of Rhegium to ps.-Heraclitus’ Homeric Questions and Longinus. Philodemus calls such interpreters “palpably mad,” albeit in a seeming contrast with Crates. Philodemus seems to have in his sights here the likes of Metrodorus of Lampsacus, a fifth-century allegorist about whom only a few tantalizing details are known. Metrodorus seems to have compressed the whole plot of the Iliad into a single cosmic image of planets circling around a fixed point, with heroes assigned cosmic roles (thus, the sun chases the moon the way Achilles flies on the heels of Hector, etc.), and (if we accept the identification based on Philodemus’ text) with gods allegorized as bodily organs. Crates, less madly, merely ascribed to Homer knowledge of the spherical nature of the cosmos and the earth, a thesis he seems to have expounded in relentless lemmatic detail along the philological model of the Alexandrian scholars but to their discomfort, to judge from other critical notices in the Homeric scholia and elsewhere.

Everywhere else but Philodemus, that is. All that Philodemus says is that Crates practiced allegory, as did others, only his was called sphairopoiia (F 99, grouped together with the poetic fragments, but — symptomatically of this edition — not with those on Homer). We don’t have much of a context for this information, and there’s not even enough of a context to say whether Philodemus had more to divulge about Crates or exactly why he brought up the point where he did. And that is one of the problems with the fragments from Philodemus in B.’s edition, and in any edition that includes these fragments. There is no obvious way to connect up this one scrap of evidence with the others in the collection, although good guesses are possible.5 Nowhere else in the surviving Herculaneum papyri does Crates’ name come up in connection with his allegories of Homer, which at least have the virtue of enjoying a confirmation outside of Philodemus (Crates’ sphairopoiia is richly attested elsewhere). In a word, there is a gaping disconnect in our testimony about Crates, and that disconnect happens precisely in the place that most promises to bring the various facets of Crates’ personality into focus for us. In no other ancient writing do they so much as appear together. Alas, Philodemus, or what we have of him, leaves us pretty much in the lurch.

Scholars typically play down this disparity, but it is graver than they make out. The disconnection happens on two levels: textually and interpretatively. B. rightly notes, for instance, that the fragment on allegory just mentioned stands out for a reason: “Questo frammento è importante perché è l’unico tra quelli citati nella Poetica di Filodemo in cui il nome di Cratete è leggibile con certezza.” The fragment accordingly bears no asterisk. The others from the same source, Philodemus’ treatise On Poems, do, with one exception, possibly mistakenly. Here are the remaining mentions: (i) [! ! ! ! ! ! ! Κρά]τητος τε [! ! ! ! ! ! !]; (ii) ὑπὸ μὲ[ν Κρά | τη]τος;6 [Kαὶ Κράτ]ητος;7 (iii) ὁ Κράτης λέγει (rightly asterisked by B. because she suspects the reading to be contaminated by an underlying layer of papyrus showing through in the surface layer);8 and finally, the one unasterisked fragment (which includes six pages of surrounding material): (iv) τὰ π?[Aρὰ] τῶι Κράτη?T?[I | ! ! ! ! !]. How certain is the last reading? Less than may warrant the absence of an asterisk, I would suggest, if we follow the rule that seems to govern the other cases. The supplement is from Jensen (1923), who printed Κράτ[η | τι] but read κρατη? | τ?[!]. But Hausrath supplemented this fragment of a name differently: Κρατ[ίπ | πωι]. And although my own recent reading of the traces (summer 2000) might be forced into conformity with Κράτη? | τ?ι?, or less probably with Κράτη?τ?[ι], this is wishful (the traces are imperfect and very faint: to be safe I might print Κρατ! | ! !) and admittedly based on the same assumption that all editors including B. have learned to accept from Jensen, namely that Crates of Mallos is the subject of these columns, in the concluding portion of Book 5 (the final book of On Poems). Obviously, the decision to ascribe the columns to Crates depends on taking the other attributions into account; but those are in turn dependent upon this seemingly good-enough reading of the papyrus. Circular reasoning is inevitable. In the absence of the remaining supplements, how likely would this one supplement be? More pointedly, in the absence of the one secure attestation of Crates’ name, in the allegorical fragment (first published in 1890 by Gomperz), would anyone have thought to supply his name in the remaining passages?9

The question is all the more urgent given the nature of the Philodemean evidence. For the boon of these scraps is also their bane: they yield information about Crates of Mallos that is corroborated by no other ancient source. What we learn from Philodemus is that Crates held a theory of poetic excellence based on the sound of poems and thus that he was an exponent of poetic euphonism. Qualities of sound here are assessed independently of, and (at the theory’s most reductive expression) at the complete expense of, meaning and form. The value of poetry for a euphonist critic lies not in what poetry means but in the way it sounds. And Crates seems to have assembled the evidence of predecessors in the euphonist tradition of criticism from the end of the classical era (starting with Megaclides of Athens, in addition to several otherwise unknown entities presumably of later date), and then distinguished his view from theirs in some fashion. Philodemus seems to have gone to Crates to gain access to these bizarre critics, where he found their views conveniently grouped together and, again presumably, organized after a fashion (at least chronologically, but possibly also according to other principles: for example, the euphonist theory seems to be increasingly radicalized, viz., reductionist, over time, and tracking this may have been one of Crates’ interests).10 On this basis, Philodemus refuted them all.

The importance of this information is beyond question. Thanks to the Herculaneum papyri, all existing histories of ancient literary criticism and aesthetic theory need to be revised, at a minimum so as to include a new chapter in their historical overview of the field, though probably this won’t suffice by itself. Euphonism lives on in critics of the Roman era (Dionysius of Halicarnassus and ps.-Longinus both know this approach), and it is also a staple of much mainstream criticism after Aristotle, whenever critics turns their attention to attributes of sound. The role of euphonism in the debates over linguistic Hellenicity, or the purity of the Greek language ( hellenismos), and later, by analogy, of Latinity, is likewise implicated and would need to be brought into sharper focus. It wasn’t only the later lexicographers of the imperial era, such as Phrynichus, who were contending over the correct sounds of classical Greek (most often, Atticism, a topic to which Crates probably devoted a treatise; if so, it was one of the first of its kind [F 106-121* B.], but in the wake of possibly isolated treatments by earlier scholars: see B., xliii-xlv for excellent arguments about the early appearance of Atticism as a debating item among the Alexandrians). A text from Philodemus attributed by Richard Janko to one of Crates’ predecessors and presumably quoted by Crates (and so not found in B.’s edition) makes this connection explicit:

“So too, then, in the case of people speaking Greek the sound produces what is particular [ to idion ] with regard to our pleasure — wouldn’t it be dreadful for [the sound to be deprived [of this particularity of pleasure] because of speaking Greek [ dia ton hellênismon ]! —, but is perhaps distracted by some other (factors),” viz., by, or towards, the sense of the words, as opposed to their phonic qualities (Phld. On Poems 1, col. 100 Janko; cf. ibid., col. 94).

And once this connection to the perception of the Greek language is made, the formerly bizarre and arbitrary exponents of poetic sound suddenly become culturally intelligible, even necessary. Euphonism is a way of defending the intuition that Greek, correctly spoken, is aesthetically pleasing: Greek sounds good for the same reasons that it sounds right. So widespread a phenomenon won’t have issued out of the void: it must have existed in the earliest traditions of criticism, and in fact there is evidence to suggest it did.11 In a word, it would be surprising if a culture as steeped in orality as Greece failed to produce a strain of criticism that could explain its own cultural obsessions. The euphonists seem to fill this gap admirably. Indeed, if they hadn’t existed, somebody would have had to invent them.12

All this is well and good, but what does it have to do with Crates of Mallos? Here is where the disconnect mentioned earlier comes into view again. Not only is there a problem of attestation in Philodemus, as we saw, but nowhere outside of the Herculaneum papyri do we find even the slightest hint that Crates was interested in euphony as a criterion of poetic excellence. Nor does Philodemus himself ever link the “insane” Crates of On Poems 2, the Homerist and allegorist (see above), to the author of the euphonist theory elsewhere in the same treatise who has been identified with Crates. Worse still, in the one general notice we have on Crates’ programmatic self-description (from Sextus’ Against the Grammarians), Crates elevates the critical science to an airy height that leaves such mundane concerns as breathing, accents, and syllabic quantity — the bread and butter of euphonic criticism — trailing well behind: those are areas that the “Critic” ( kritikos) may share with the (lower-case) grammarian ( grammatikos), viz., the run-of-the-mill (Alexandrian) scholar, but they are his minor capacities (if they are his at all), while he devotes himself to a higher, more universal calling (whatever that is). If Crates really thought so little of prosodic analysis, why would he have devoted a treatise to the virtues of euphony? And how would he have connected euphony to the higher science of criticism?13

So now instead of one problem we have two (or three). The connection between Crates’ textual criticism, his allegorism, and his euphonism (attested nowhere else in antiquity) makes for a hard sell, not least because these activities resolve the question of poetic value along two or three seemingly incompatible axes. It won’t do, without further argument, to say that euphonism was part of Crates’ general critical arsenal, to be drawn upon at will (see B., p. xxxiii), because at first blush euphonism has no room for meaning as a positive criterion of poetic value, whereas textual and allegorical criticism do. (See F 78 B., where, as B. points out [p. xxxiii], Crates’ decision to athetize the proem to Hesiod’s Theogony was based on considerations of content — and, one should add, on content alone, without consideration of how the verses sounded. The examples could be multiplied.) Getting from textual criticism to allegorical reading is not a hard step to make. Getting from either of these places to euphonism, or vice versa, is. It’s not just that the move is difficult to make. The problem is that the different activities are strikingly at odds. In Philodemus, Crates claims (F 101, col. xxix) that “the judgment ( krisis) of excellent poems resides in the letter/sounds ( stoicheia).” Krisis advertises what is distinctive about Crates’ self-appointed calling as a kritikos. But the kritikos, Sextus tells us, transcends the world of small empirical things. The problem has been intensified by attempts since Jensen to brand Crates a Stoic philosopher (a tendency B. shares). But Crates nowhere describes himself as a philosophos, let alone as a Stoic: his proud label was that of a kritikos, and so far as we know it was firmly planted in the realm of literary criticism.14 Thus, merely subordinating criticism to philosophy is not the solution. Criticism, on the contrary, has to be conformable with whatever subordinates it, and ideally it ought to lead to it too.

I believe there is a way to connect these two theories (euphony and allegoresis), and to make room for the third (conventional textual criticism) besides.15 That is, I believe it is possible to save the identity of Crates in Philodemus’ remains. But I would be more comfortable about making the connection and saving his identity if I could find just one aspect of it ready-made in the ancient testimonies. On the other hand, there are some hints in the material on euphony that point to a good fit with Crates’ known method of allegoresis, not the least of which is his sly equivocation that the object of the critic is “to judge [poems] not without the meanings, but it is not to judge the meanings” either (F 101, col. xxviiii). Allegorical reading does just this: its object is what poems mean in some sense, though not what they mean in their usual sense, and so it is a reading that equivocally involves what poems “mean.” Or, as Philodemus puts it in the sphairopoiia fragment (F 99 B.), “though [allegorical interpreters] often agree that the words qua signifiers ( ta sêmainonta) [involve] ‘wrath,’ they want the poet to be representing other things” beyond the surface meaning of the poem. Assuming, that is, that allegoresis constitutes a kind of judgment (yet another issue), and assuming that it is a judgment that can somehow be arrived at through the empirical experience of how a poem sounds (a further requirement of F 101, col. xxviiii). And though it may be a small consolation, there is in fact no way to infer from the testimony of Sextus, or from the fragments on Atticism, or from many other grammatical fragments, that Crates of Mallos was interested in allegoresis either. This may just be an accident of the way the ancient traditions remembered him. Possibly Crates was not explicit about how the various facets of his critical activity dove-tailed. Perhaps they didn’t fit into one comprehensive theory, and he was either overambitious in his claims or else he never meant to claim that they do fit together. But to allow this last possibility would be an act of final desperation. One wants to see the various parts fit together in a comprehensive whole, and ideally one would want to see how each of the parts is motivated by the others and by the whole.16

The complications that arise out of treating Philodemus as a source brings to the fore one last point that affects the nature of any edition of fragments, and that touches B.’s edition as well. This has to do with the question of what in any given instance counts as a “fragment.” A good part of the “fragments” we have of Crates of Mallos are not fragments in any conventional sense, if by that we mean a piece of a larger, now lost text that contains the ipsissima verba of its author. On the contrary, the fragments are frequently of testimonial value (as is often the case with philosophical fragments generally). Philodemus presents a special case of this larger problem. For the writings of Philodemus are like so many filters superimposed upon his sources, most often in infinitives of indirect discourse, and often again in hypotheticals (“If N claims that-X, then he is wrong”). In some cases (exemplarily, Crates’), we find polemical refutation buried within polemical refutation: Philodemus refutes Crates and his adversaries, while reporting Crates’ disagreements with his predecessors. Views and opinions can be resuscitated from these distortive layers of polemics, but rarely with complete confidence. And if we follow Richard Janko’s persuasive reconstruction of the first two books of On Poems, there is much more to report about Crates than B. includes in her edition, if only by deductive inferences (a problem B. gets around by restricting herself mainly, but not exclusively, to those passages that come closest to attesting Crates’ name). Thus, any edition of the fragments of Crates is bound to be either too thin or too full. Here, with the survivals from Herculaneum, we have a true embarrassment of riches indeed.

Collections of fragments may grow thanks to new discoveries, but that is not their only value, or the only way in which they grow in usefulness. Sometimes they grow, paradoxically, by subtraction, for instance when choices made by earlier editors are rejected by later editors. B.’s edition does this, for instance, in rejecting the attribution by Mette of a work titled On Athenian Sacrifices that was ascribed by Preller and later by Jacoby to the lexicographer Crates of Athens (see p. xliii), and more dramatically in eliminating the extra material from Sextus, Varro, and other sources that pads out Mette’s arguments but blurs the lines between fragments, contexts, and parallels. The history of these waxings and wanings by itself is an interesting window onto the changing fashions and self-conceptions of philology. But other factors count too. Clarity of design and presentation, judiciousness, user-friendliness, and typographical improvements all contribute to the usability of editions. And on this score, B.’s edition marks a huge advance over its predecessors: it is an elegant production in every respect. Full translations of the Greek texts, many of which come from obscure and often difficult sources, would have guaranteed this book an even wider readership. And some kind of grouping of fragments having to do with sphairopoiia, even if only by way of an index specially designed for this purpose, would have allowed B. to displace Mette definitively as a collection of fragments.17 Nonetheless, anyone interested in finding an orientation in the fields touched on by Crates, and not only in learning about Crates, will absolutely want and have to use this work, and they will benefit greatly from it besides. B.’s edition will be an indispensable tool, as scholars continue to navigate the tricky fields of literary criticism, grammar, and the ancient practices of reading.18


1. E. Simon, Pergamon und Hesiod (Mainz am Rhein, 1975) and A. F. Stewart, “Narration and Allusion in the Hellenistic Baroque,” in Narrative and Event in Ancient Art, ed. P. J. Holliday (Cambridge, 1993), 130-74, neither of which is mentioned in the book under review.

2. See also the supplementary article by Hans Joachim Mette, “Krates von Pergamon 1953-1983,” Lustrum 26 (1984), not listed in Broggiato’s bibliography.

3. R. Pfeiffer, History of Classical Scholarship from the Beginnings to the End of the Hellenistic Age (Oxford, 1968), 239 n. 7.

4. For excellent discussion of the thrills and perils of fragment-hunting, see the essays in G. W. Most, ed., Collecting Fragments. Fragmente Sammeln (Göttingen 1997).

5. The surrounding columns are in tatters, but the context has to do with the question how we can derive pleasure and entertainment ( psuchagôgia) from verses when their meanings are obscure. Allegorical meaning is a special case of this kind of obscurity, but not obviously so, because it is not obvious why the surface meaning of Homer should be considered obscure. See further at n. 10 below.

6. But as Richard Janko points out to me it is indeed hard to come up with another name that would fit the available space.

7. The omicron should appear sub-dotted with an asterisk, to indicate that it has been supplied by an editor and that it contradicts the text or its witnesses, but this is technically impossible for the present review.

8. On the uncertainties of the readings of this entire text and its surroundings, see Philodemus, On Poems, Book One, ed. R. Janko (Oxford, 2000) 165-66 n. 6.

9. Dirk Obbink (personal communication) makes the convincing point that if the Crates in the allegorical fragment weren’t identical to the Crates mentioned elsewhere in the same work, Philodemus would have needed to distinguish them (which he fails to do).

10. See Janko, ed. (2000). This last speculation about the increasing reductionism of the euphonist tradition is my own; see J. I. Porter, ” Hoi Kritikoi : A Reassessment,” in Greek Literary Theory after Aristotle: A Collection of Papers in Honour of D.M. Schenkeveld, ed. J. G. J. Abbenes, S. R. Slings, and I. Sluiter (Amsterdam 1995), 83-109; 106-107.

11. Janko, ed. (2000) 136-37, 173-85; J. I. Porter, “Aristotle and the Origins of Euphony,” in Mneme e Charis. Studi in Memoria di Marcello Gigante, vol. 1, ed. S. Cerasuolo (Naples 2004).

12. For this argument, see J. I. Porter, “Feeling Classical: Classicism and Literary Criticism” (forthcoming).

13. The notice from Sextus is vague, and it may be misleading as well. I suspect it sheds light not on Crates’ activities, which in many respects were perfectly ordinary, but on his self-conception as a scholar and on his wish to position himself in a crowded field of competitors: grasping for a difference, he created a flashy label for himself and his school and thus laid claim, backed by etymology, to the domain of krisis poiêmatôn, viz., literary discernment, judgment, and criticism. See further D. Blank, ed., Sextus Empiricus, Against the Grammarians (Adversus Mathematicos I) (Oxford, 1998), 140. Blank (“The Organization of Grammar in Ancient Greece,” in History of the Language Sciences, ed. S. Auroux [Berlin and New York 2000] 400-417; here, 405) believes that Crates’ definition won’t allow the lower grammarian to accede to higher criticism and then assumes that this latter consists in “a judgment based on [literature’s] sound.”

14. His alleged “Stoicism” appears only in the Byzantine Suda, where the label is a desperate attempt to rationalize Crates’ bizarre literary critical approach. See J. I. Porter, “Hermeneutic Lines and Circles: Aristarchus and Crates on Homeric Exegesis,” in Homer’s Ancient Readers: The Hermeneutics of Greek Epic’s Earliest Exegetes, ed. R. Lamberton and J. J. Keaney (Princeton 1992), 67-114, esp. 85-87); cf. B, p. xvii (“Vita”). B. follows the view that the logika theôrêmata that Crates sees operating in poems reflect Stoic truths. But they need do this only if we assume Crates’ Stoicism. The mention, without closer definition, of logikê epistêmê in Sext. M 1.79 has led commentators to the same kind of assumption. But Sext. M 1.248, on Crates and his school, defines the logikon meros of kritikê as “preoccupied with language ( lexis) and grammatical figures ( tropoi),” while the remaining two parts have to do with differences in dialect and style (the tribikon meros) and with literary material (the historikon meros). I see no reason to resort to Stoic logic in order to fill in the meaning of logikê epistêmê in any of Sextus’ reports on Crates or his school.

15. See ” Hoi Kritikoi” (n. 9 above), 94-99 and 105, where I drew on some columns from Book 1 that I took (following an earlier stage of Janko’s research) as direct evidence of Crates’ views, but which now, in the light of Janko’s edition, I would want to claim reflect the kinds of arguments Crates must have made once he took them over from a certain Pausimachus of Miletus and refined them for his own ends, as he evidently did in other cases too.

16. A tacit assumption of interpreters (including myself) has been that Crates’ allegories of Homer culminated his general science of language, but there is no evidence for this. All we can really say is that Crates excelled at interpreting Homer in addition to making extravagant claims about his brand of philology.

17. Mette remains useful simply for gathering together so much relevant ancient testimony on sphairopoiia, only a fraction of which can be directly attributed to Crates but much of which is relevant to his theory or its afterlife in antiquity. The need for a separate category is all the more pressing if, as it may be speculated (although B. does not do so here), Philodemus’ tag in F 99 B. ( ta peri tês sphairopoiias) is a reference to an actual title by Crates ( Peri tês sphairopoiias, or On the Spherical Construction [ of the World ]), parallel to a second possible new title, ta peri tôn stoicheiôn in F 101 col. xxix B. (On the Letters). If this is right, Peri ought to be capitalized in both cases. Philodemus refers to Apollodorus’ On the Gods in the same way in On Piety : ta Peri theôn. See Philodemus, On Piety, Part 1, ed. D. Obbink (Oxford, 1996), 17 n. 2. (Thanks to Dirk Obbink for this parallel.) Indeed, this curious survival of the two titles — or are they sections of a single work? Philodemus’ apparent failure to refer to the allegorical theory except once in passing tells, I think, against this second possibility — would encapsulate the dilemma of the Philodemean information on Crates all too agonizingly: how did Crates get from the line to the circle, from euphony to allegory, in his readings of Homer?

18. My thanks to David Blank, Richard Janko, and Dirk Obbink for their comments, and their forbearance.