In Rome, probably in the late third century CE, a teacher probably named Marius Plotius Sacerdos wrote three books of artes grammaticae. The first book contained the standard elements of the ars grammatica (basic concepts such as “the letter” or “the syllable”; extended discussion of the eight parts of speech; and a section on virtutes et vitia orationis), while the second contained catholica (morphology of the noun and verb, plus a short and remarkable section on clausulae, contrasting Ciceronian usage with that of the author’s own time) and the third contained a discussion of metrics. What survives of the first two books (now seriously incomplete) is found in a fifth-century manuscript discovered at Bobbio in 1493 and now in Naples (Naples, Lat. 2); the third book is preserved separately, in three manuscripts of the ninth and tenth centuries.
Sacerdos is not often read and not often edited: the first two books, included in Andrea Bramanti’s edition, have been printed only twice before, the last time by Heinrich Keil in Grammatici Latini 6, published in 1874. Even among grammatical texts, Sacerdos’ work is dry: curious details are generally lacking, and the vast majority of his quotations come from school authors. Two elements of Sacerdos’ work, however, deserve attention. In the first place, his is the earliest surviving Latin grammar, at least half a century older than the better-known Donatus or Charisius. In the second place, his work has the peculiar distinction of having been plagiarized at a very early stage, and by a text that also still survives and in the same manuscript: the work De catholicis ascribed to “Probus” and included in Bramanti’s edition is almost identical, where both works survive, to Book 2 of Sacerdos. It thus provides important (if ambiguous) evidence about the growth and development of grammatical writing in Rome.
Even though Bramanti does not include Book 3 of Sacerdos (as it is on a different subject and transmitted separately), his edition is massive: nearly 300 pages of introduction (of which nearly 100 are bibliography) followed by 240 pages of text (with Sacerdos Book 2 and Probus De catholicis on facing pages) and, in the second volume, nearly 700 pages of commentary. His text does not differ greatly from Keil’s (other than including a previously lost section of Book 2 identified in 1983 by Mario de Nonno) but it is set out much more readably, while the introduction and especially the commentary to Book 1 represent a very significant contribution. The commentary is generally exegetical rather than textual; Bramanti’s aim is to elucidate Sacerdos’ place in the grammatical tradition, and he sets out the evidence very clearly and copiously, giving his readers all they need to make their own assessment.
Bramanti’s work is, in effect, a commentary on the whole tradition of school grammar. For much of Book 1 (the basic ars grammatica, unfortunately missing a great deal at the beginning), Bramanti introduces his commentary on each section with a list of loci similes in other grammars; for some of the more vexed issues he provides a synopsis of all the major approaches found in the tradition. Thus, on the septimus casus (the name for varying sets of uses of the ablative, itself sometimes called sextus casus), Bramanti identifies and gives the evidence for four different approaches found in the grammarians. So too with conjunctions: in order to elucidate Sacerdos’ views, he lists the classifications of conjunctions in all the other major (and some quite minor) grammatical texts. The same is true of the whole series of problems of classification and nomenclature for solecism and barbarism and for the section usually given the title de ceteris vitiis (sc. orationis) which follows, in Sacerdos entitled De metaplasmis vel figuris: Bramanti anatomizes Sacerdos’ huge list of rhetorical figures which includes both vitia and virtutes without clear distinction. Bramanti is careful and (so far as I can judge) thorough in providing information both about the ancient analyses of grammar and about modern interpretations of them, making good use of the extensive recent scholarship about ancient scholarship; one can easily use his book as a starting point for research on a great many aspects of the Roman grammatical tradition.
In his introduction (CCVI-CCXIV), Bramanti gives a brief explanation of Sacerdos’ work: that it is one of the earliest attempts (from our point of view, the earliest) to move grammar as an intellectual endeavor away from scholars and antiquarians and into the schools, where something much simpler than what we find in Charisius or Diomedes was needed. Sacerdos, even more than most of the grammarians, emphasizes morphology over semantics; he eschews recondite quotations and rabbit holes in favor of giving his pupils no more than they can absorb; and he tries to include everything needed, forms as well as theory, in a relatively short space. You can read Sacerdos much more quickly than the eclectic and rambling Charisius, and you can read him with much less theoretical savvy than the relentlessly abstract Donatus: closer to Wheelock than to Kühner-Stegmann. Modern scholars are therefore wrong, Bramanti argues, to treat Sacerdos as an incompetent version of either Charisius or Donatus: he is doing something else, and he is performing a real service (to his students, if not to us).
Bramanti’s interpretation has much to commend it. He rightly rejects a teleological account of the Roman grammarians which makes Sacerdos an imperfect stage on the path to grammatical excellence, just as he rightly rejects (as does much recent scholarship) the overly rigid Quellenforschung which led Karl Barwick to trace everything interesting in Roman grammar back to Remmius Palaemon and to treat everything diverging from his reconstructed Palaemon as of lesser worth or intelligence. But in his desire to create a coherent and purposeful Sacerdos, Bramanti tries to explain away or dismiss the real problems of this text as by-products of its faulty transmission rather than authorial imperfection. And while that works for some things, it leaves a lot to be desired.
Even in Book 1, which exists only in one form, there are real problems. Why does Sacerdos have a separate section on the septimus casus that comes well after he has finished talking about the noun? Why is a small set of rhetorical figures discussed before solecism and barbarism (and why does solecism come first?) and then repeated in the longer set of schemata that concludes the book? Bramanti suggests that whoever put together the extant text had a damaged copy and assembled what he had in what was not the original order. Damage there certainly was, but the role of the compiler that Bramanti constructs is implausible, and even if it is right, it has further implications.
Put very simply, is “Sacerdos” Sacerdos? We know from comparing the two versions of Book 2 (Sacerdos and “Probus”) that both of them contain some real nonsense; each one needs to be corrected from the other, and in some cases there are clearly two legitimate versions, each of which has divagated in some respects from the posited original of Sacerdos. And even without two versions, we can see that Book 1 is not what it presumably once was: if this is what it looked like when the real Sacerdos finished writing it, then he deserves the scorn he receives from modern scholars. The grammar of Sacerdos had a real, if not necessarily strong, influence on the later grammatical tradition to judge both from citations and the use of his name in grammatical examples; but, as with other famous grammatical names (e.g. Scaurus and Probus), the texts attributed to them remained less stable than the names themselves. Bramanti, particularly in editing Book 2, hesitates between hoping that one version is original and accepting that neither one is; between aiming at consistency between Sacerdos and “Probus” by emendation and maintaining divergences among the two versions, even if they are trivial or wrong. That is always a problem in editing texts, like grammars, that are simultaneously tralatitious and variable: every manuscript—particularly a manuscript as early as the fifth century—has a claim to be considered a work in its own right, no matter how odd or erroneous some of it may seem; but at the same time, no matter how early the manuscript is, Naples, Lat. 2 still presents an adaptation of an earlier (and perhaps more coherent?) work which may or may not have been by the real “Sacerdos.”
I respect Bramanti for his willingness to face the problems a text like this poses, but even if there is no perfectly satisfactory solution, an editor needs to be clear and consistent in picking one method of presenting the text. The opening of the discussion of the verb in Book 2 (200-204 in this edition; 6.483-84 and 4.33 in Keil) is a mess in both versions; but Bramanti fluctuates between cross-emendation and leaving the two alone. Keil frequently emended for consistency and intelligibility, perhaps too much; but his text of this section makes sense (I think) while parts of both of Bramanti’s versions seem incoherent to me—as unfortunately does the apparatus, which is clogged with irrelevant detail. His commentary is very helpful; his text much less so.
And that brings me to a final, and unfortunate aspect of Bramanti’s work: while he is a good and careful exegete, he is not a careful editor. His apparatus is unnecessarily verbose: he gives not only credit for correct emendations and readings to earlier editors, but points out every minor misreading too, and while this makes his apparatus look exhaustive, in fact there are unexplained differences from Keil (e.g. 33.15-17=6.444.5-7K, where the word order is significantly different). He fails to regularize the text (e.g. on 234-35, where in quotations of Cicero Verres’ praenomen is abbreviated as G. three times and as C. once, and clausulae are printed with quantities marked on one page, but not on the other). What is worse is that the facing texts of Sacerdos Book 2 and de catholicis are frequently not aligned with one another, thus considerably diminishing the value of a synoptic text. And one final comment is unfortunately necessary: aside from typographical errors (which are not excessively numerous and not, so far as I can see, in the text itself), there are numerous false references to his own text: as a sample, I checked the dozen places Bramanti lists on p. CCXXXV in which he has preferred the readings of the Carolingian manuscript Paris, Lat. 7520 (which contains only a portion of the text) to those of Naples, Lat. 2; not a single one of the references is correct. I checked a few others at random; I found only one that is correct. Even the name he gives the grammarian on the title page is incorrect: the manuscripts give as his name M. Claudius Sacerdos or Marius Plotius Sacerdos, but what we get here instead is an unattested blend, M. Plotius Sacerdos.
Bramanti has good and valuable things to say about the place of Sacerdos in the grammatical tradition. Drawing on and complementing the valuable interpretations of ancient scholarship that have been developed (largely in France and Italy) in the last few decades, he rightly emphasizes the variety and creativity of the Latin grammatical tradition. His presentation of the evidence for the range of available approaches to numerous grammatical problems, both in antiquity and in modern scholarship, is clear and extremely useful; many sections of his commentary can serve as good starting-points for someone trying to find a way into this complicated tradition. Exegesis rather than editing, however, is Bramanti’s strength; more attention to presentation and proofreading would have made his work far more valuable.
 On the vexed count of the schemata, in addition to the commentary, see the discussion at CCXXVI-CCXXX.