Reviewed by James Zetzel, Columbia University (firstname.lastname@example.org)
Even among classicists, Pomponius Porphyrio is not a household name. The real Porphyrio was a scholar who worked in the early third century. The one useful chronological indication (in Charisius) shows that he made use of the dictionary of Festus and was in turn used by Julius Romanus; hence probably he was active around 200 CE. What survives under his name is a commentary on Horace which contains the usual amalgam of useful knowledge, idiocy, and misinformation and also shows an interesting taste (appropriate to his date) for archaic literature, displaying (or having taken from someone else) a broad and fairly precise knowledge of Plautus and Lucilius. His comments on rhetoric and even on grammar are sometimes useful. But 'Porphyrio' is not exactly by the real Pomponius Porphyrio: it is generally agreed, with some variations, that the extant text stems from the fifth century, not the third, and that it has been to some (not altogether clear) degree altered both by addition and by simplification, as is usual in the transmission of commentaries. That does not diminish the usefulness of what we have, but it does raise questions about the patterns of scholarship and its transmission in late antiquity.
The commentaries on Horace that survive from late antiquity are a mess, and one that could still use sorting out. Traditionally, two related 'authors' have been recognized, Porphyrio and Pseudo-Acro (Acro was a very good scholar, earlier than Porphyrio, but what goes under his name has not for a long time been thought to have anything to do with him): Porphyrio is relatively terse, and was pillaged by Pseudo-Acro, who is often wordy and even less to the point than most ancient commentaries. But about 40 years ago it was argued by G. Noske that Pseudo-Acro is not one pseudepigraphic work, but at least two: the commentary on the Odes is not part of the same work as the commentary on the hexameter poetry (it is more complicated than that, but let that pass).1 In retrospect, that discovery is hardly surprising: the commentary on the Odes is transmitted in a different manuscript (Parisinus Latinus 7900A, one of the most interesting manuscripts of Latin poetic commentary) written in a different country (Italy, probably Milan) from the many busy northern European manuscripts of scholia and commentary on the rest of Horace. One might also mention that there are lesser sets of scholia too in the margins of many manuscripts, as well as the mysterious 'commentator Cruquianus,' a set of notes printed in a sixteenth century edition of Horace based on a lost manuscript, but containing unique and unquestionably ancient material. Faced with this mess, one heaves a deep sigh, picks up an axe, and heads into the tangled thicket.
As far as Porphyrio is concerned, the first (and until recently pretty much the only) serious attempt to distinguish the early layer from later accretions was Paul Wessner's 1893 dissertation.2 Wessner established some useful criteria for recognizing interpolations; his arguments are summarized in the first chapter of Antonina Kalinina's monograph. Wessner was by far the most intelligent student of Latin scholia in the past century and more; but, as his edition of the scholia vetera on Juvenal shows, he believed strongly in carving an 'original' commentary out of the mass of late antique and Carolingian accretions. In the case of Juvenal, it works fairly well because of the existence of trustworthy guides to the ancient commentary in the Bobbio fragment and the scholia of the Pithoeanus. For Porphyrio, the fates have been less kind: there are a few Carolingian manuscripts that need to be balanced against a humanistic tradition. Nothing provides an easy way to pry the extant commentary apart.
Prying apart the text of Porphyrio, however, is just what K. wants, in a relatively restricted area, to do. She examines just one aspect of the language of the text, technical rhetorical terminology. In particular, she looks at the uses of the words allegoria, translatio, metaphora, metonymia, and figura together with other words derived from them (allegorice, figurate and the like); she examines their meanings, uses, and contexts; and she tries to argue from inconsistencies among those uses that certain passages should be seen as interpolations. By removing such inconsistencies (by emendation or deletion), she wants to get a little closer to the ur-Porphyrio.
K.'s analysis is careful and painstaking, but it is not altogether successful, for several reasons. In the first place, simple numbers: there are only 16 uses of translatio or translatiue in the commentary, nine of metonymia or metonymicos; and while there are rather more instances of allegoria and figura, there are hardly enough to be statistically significant. One simply cannot say that one usage is right, the other wrong, when they can be counted on the figures of two hands. More important is her assumption (with which Wessner would have agreed) that a given commentator will use one and only one form of words to describe the same stylistic or linguistic phenomenon. The rigidity and circularity of such an assumption is obvious: only if one assumes that each person uses only one form of words can one use that to emend the text to prove that each person uses only one form of words. Some people are probably like that, but it is not a particularly realistic assumption, and in fact, the very flexible transmission of commentaries between the fourth and the ninth centuries suggests both that multiple sources fed into single texts and that single works were used by multiple followers. No extant text of ancient commentary is a monolithic whole, and it is not likely that we can ever reconstruct or emend our way to a single ur-text. It just doesn't work that way.
Beyond the basic methodological problem, however, is the simple fact that K. is working with a very small body of text (and with very few particular lexical items) and has relatively little to correct. When she does have enough to work with, moreover, the results are sometimes very odd, as with the largest group of passages she discusses as a set, those in which the two phrases figurata elocutio and figura elocutionis appear. The first of these appears in five passages (and one closely related), the second in three (and one closely related). K.'s decision is that in all instances figurata elocutio should be emended, and that figura elocutionis is the 'correct' expression for Porphyrio. There is an argument to be made for K.'s view: figura elocutionis is certainly the less common expression in commentaries, and as a rhetorical term it fits what we can observe about Porphyrio's interests. The difficulty is that the emendations seem at best unnecessary, at worst quite wrong. Thus, at Carm. 2.9.17-18, she emends 'Graeca elocutione figuratum est' to 'Graeca figura'. The latter phrase is certainly common, but there are parallels for the former: Servius (Comm. in Don. 4.411.24K) uses 'Graeca elocutio' and (on Aen. 12.25) 'elocutio figurata de Graeco'. At Carm. 1.6.3-4, 'quam rem cumque ferox . . . miles te duce gesserit' not only is the emendation unnecessary, but K. misunderstands the text. The note (unemended) reads: 'Figurata elocutio: nam cum dixisset "scriberis", intulit "Quam rem miles gesserit te duce." Simplex autem haec elocutio erat: scribentur res, quascumque miles gesserit te duce.' K. (following Diederich) thinks that the figure is the tmesis of quam...cumque, but the opposition between 'figurata' and 'simplex' shows that the figure is the switch from second person to third. And in emending 'figurata elocutio' to 'figura elocutionis', K diminishes (as she recognizes) the balance in the note between 'figurata' and 'simplex'.
To say that usage of technical terms should never be regularized would be excessively conservative, but to say that they should always be regularized fails to recognize the amorphousness of the tradition. It is possible that figura elocutionis represents an early layer of the tradition (perhaps Porphyrio himself) while figurata elocutio is a later one; but it is also possible either that the two phrases represent different sources amalgamated at a later stage of the tradition or even that they are used by the same person in slightly different contexts. What K. does not seem to realize is that the two in fact do come from very different contexts: figurata elocutio is a frequent phrase in scholiastic language, and in that context elocutio merely means 'choice of expression'--other similar uses are absoluta elocutio, honesta elocutio, or antiqua elocutio, and even Latinior elocutio, all found in Servius. Figura elocutionis, however, appears nowhere in Servius, and except in Porphyrio only (as far as I can see from consulting PHI and TLL) in rhetorical texts where it means, as one expects, 'figure of speech' as opposed to 'figure of thought.' When Porphyrio uses it, it should probably be understood as meaning 'this passage displays a (rhetorical) figure of speech,' while figurata elocutio means 'this word or construction is used in an interesting or irregular way.'
In this respect, K. is not only too dogmatic about the meaning of technical terms, she is also curiously uninterested in parallels from other scholiastic texts that might help locate the various layers of Porphyrio in the tradition of late antique scholarship. In many ways, this lack of interest is surprising, as she is the only scholar I know to mention a curious feature that occurs only once in Porphyrio but frequently (as K. notes) in Donatus' commentary on Terence, the numbering of rhetorical figures, as in apodosis prima and anacolouthon tertium. This is, as K. notes, an otherwise unknown (and as far as I know, completely unstudied) system of classifying rhetorical figures; it very much deserves closer study. But if Porphyrio shares a rare feature with Donatus and shares some of the technical language of stylistic analysis with Donatus' pupil Servius, then it would be worth exploring further whether the extant (fifth-century) Porphyrio is indebted to Servius' methods as is, for instance, Pseudo-Asconius on Cicero, or if the original (third-century) Porphyrio was someone to whom Servius and Donatus were themselves indebted.
If K.'s attempt to use a limited range of technical terms to emend and to analyze the text of Porphyrio is not very successful, her examination of the distribution of these terms over the text of Porphyrio leads to one very interesting and significant discovery, namely that the terms she studies are found throughout the commentary, except in the commentaries on the Carmen Saeculare and Ars Poetica, where they never appear. The absence from the commentary on the Carmen may be simple coincidence--it is a short text--but it is clearly not that for the Ars, and K. reveals a completely different set of Greek critical terms found in the commentary on the Ars and nowhere else in Porphyrio. While it has long been known that various ancient scholars commented on one or another separate work of Horace, this is the first evidence I know that such commentaries have left traces of their individuality in one of the extant texts. K. studies the commentary on the Ars carefully in terms of both language and structure, and that is in itself a worthwhile contribution.
In sum, K.'s major argument is careful and clearly set out, but is much too dogmatic and ultimately does not lead very far. The discovery about the commentary on the Ars, however, is of genuine importance. If K. had been willing to follow her own original observations and given up on the textual archaeology of Porphyrio, she could have written a much more satisfying monograph.
1. G. Noske, Quaestiones Pseudoacroneae (Munich, 1969).
2. P. Wessner, Quaestiones Porphyrioneae in Commentationes philologae Ienenses 5 (1894) 153-96.