Bryn Mawr Classical Review 2003.03.24
Michela Rosellini, Ps. Remmii Palaemonis Regulae. Introduzione, testo e commento. Collectanea Grammatica Latina, 1. Hildesheim: OLMS-Weidmann, 2001. Pp. lvii, 165. ISBN 3-487-11307-4. EUR 29.80 (pb).
Reviewed by Leofranc Holford-Strevens, Oxford University Press (email@example.com)
Word count: 1489 words
The eminent grammarian, sexual nonconformist, and agricultural entrepreneur Q. Remmius Palaemon of Vicetia, who taught at Rome in the first century AD, enjoyed a posthumous reputation sufficient for his name to be set over various works with which he had nothing to do, including a grammatical treatise complied in late antiquity, rediscovered by Pontanus, and first printed in 1503. Hitherto scholars have relied on the edition by Keil,1 who held the work in low esteem as likely to be Pontanus' own concoction. Michela Rosellini, already known for her edition of Julius Valerius,2 has recovered the Regulae from two medieval manuscripts in Angers and Oxford, which show not only that Keil's suspicion was unfounded but that the recension found in the seven late-Quattrocento manuscripts and the editio princeps had undergone serious interpolation. In addition to presenting the first edition of the authentic text, Rosellini (hereafter R.) offers a full introduction on the history and nature of the work and a detailed commentary. Readers with an interest in ancient grammar will find this edition of the highest value.
The work comprises six sections: 'De nomine', 'De pronomine', 'De verbo', 'De participio', 'De adverbiis', and 'De praepositionibus'. In accordance with its title, it is more concerned with rules of thumb than with the systematic exposition appropriate to an ars; indeed the final section, having set out the basic distinction between motion and position, closes 'Abunde dictum sit de praepositionibus. Iam cetera in artibus habes plenissime dicta'. The longest section is that on the noun,3 the rules on which are designed to help the reader who knows one case to find the others; for instance, if the ablative singular ends in a or o the genitive plural will end in rum, the dative and ablative plural in is.4 Although the verb declinare and the nomen actionis declinatio are used, there is no trace in the original recension of the familiar five declensions; two references are inserted in the interpolated version, but without changing the general emphasis. This is slightly less marked in the treatment of the verb, inasmuch as the genera verbi are listed and three (not four) conjugations recognized (as in many of the ancient grammarians); on the other hand, only present-stems forms are considered, and not even all these, the most notable absentees being the imperfect tense and the subjunctive mood. The style is plain, with less than a handful of poetic examples and no sign of Greek beyond a few technical terms; however, the author allows himself one flight of fancy (the student who pays heed to the verbal conjugations 'quasi ex vado vitiorum emergens citius ad portum Latinitatis accedet'), unparalleled in the ancients but gladly taken up by the Irish Anonymus ad Cuimnanum, and one original metaphor, used twice: those who had adjusted the language for euphony's sake had taken care 'musicare Latinitatem'.
In her introduction, R. begins by discussing the textual transmission: having described the individual manuscripts (pp. xx-xxiv), she offers a stemma together with arguments showing that all extant manuscripts are derived from a single archetype and that the late manuscripts are derived from an interpolated lost intermediary μ, itself sharing a hyparchetype α with the Angers manuscript (A) but not with the Oxford manuscript (O). It follows that, whereas agreement between O and either the later manuscripts against A or A against the later manuscripts establishes the paradosis, when the later manuscripts side with A against O, or when in the absence of the defective A the later manuscripts differ from O, the two variants are equipollent and must be judged on their merits case by case.5 These are clearly the principles on which R. operates; in fact, however, agreement between O and μ against A is very rare, though when all three differ, μ may have a text closer to O than to A.
Having discussed the transmission, R. summarizes the work's 'struttura e caratteristiche'; most notable amongst the latter is the author's penchant, shared to a far lesser extent by 'Augustine', for inventing incorrect forms to be forbidden by his rules. R. then turns to the place of the Regulae in Roman grammatical tradition, reiterating that it has nothing to do with Q. Remmius and pointing out close connections with the so-called Regulae Augustini; she places 'Palaemon' no earlier than 'Augustine' and 'Probus'. This section is brief, since she had given a fuller exposition elsewhere.6
R.'s editorial judgement is generally sound, for instance when she considers that the demonstrative hic haec hoc is more likely to have been added in α and again in μ than suppressed in O. The text of the interpolations is also edited in accordance with her stemma, though I cannot understand why, having offered cogent arguments in her introduction (pp. XXXIX-XLI) for deriving the manuscripts R and X respectively from B and W, she does not eliminate these codices descripti but cites their readings in her apparatus. In her opinion the interpolations in μ were made either between the lines or in the margins in such a way that subsequent copyists could interpret them in different ways; thus within the expansion of 1. 61 one phrase is wrongly transferred in one manuscript to the next section, where it is completely out of place. When, however, the manuscripts are divided between futurum tempus and promissiuum (3. 25), alternative terms in this treatise for the same thing, one wonders whether the cause is rather scribal whim -- not that this affects R.'s stemma.
R. normalizes spelling and imposes the vowel-consonant distinction between u and v, even in reporting MS readings, so that at p. 39, l. 7 O's datuos for datiuos becomes 'datvos'. This obscures the nature of the error, the loss of a minim. That apart, the transcription of the readings, to judge by spot checks against the Oxford manuscript, shows a high degree of accuracy; use of microfilm is suggested by her note at p. 25, l. 2 'ab eo quod sunt . . . fere evan. in O' (they are not at all faded, merely traversed by a fold in the parchment) and no doubt explains the lapse at p. 31, l. 10 'dat(ivu)m p(lura)li' for 'dat(ivus) in p(lura)li'.7
The original recension is printed on recto pages; where the alterations in the interpolated version cannot conveniently be accommodated in the apparatus criticus, the text is set out on the facing verso. Presentation is attractive, but on p. 41 the marginal line-number 5 is placed against line 3, and 10 against line 7; however, the line-references in the apparatus pertain to the correct, not the printed, lineation.
The remainder of the book, nearly a hundred pages, is taken up with an extensive commentary; notes on the interpolations, locating their sources and attempting to explain their purpose, are included as separate paragraphs in square brackets. Although the author's own Latinity is not neglected, R.'s chief concern is with parallels to his doctrines in other Latin grammatical writings. Several notes bear witness to the close relation between this text and the Regulae Augustini; many of the interpolations are taken from Priscian. Such attention is not paid to written or spoken usage in the language the grammarians purported to describe: when at pp. 108-10 R. discusses the repudiation of hoc ossum ('De nomine' 52-4), Cn. Gellius fr. 26 Peter = 27 Chassignet appears by virtue of being cited in Charisius, but there is no reference to Augustine's use of this form in clarifying the ambiguous os meum of Ps. 138: 15,8 nor indeed to the genitive ossi at Apul. Apol. 49. 49 or the vulgar attestations of ossus (e.g. Mulomedicina Chironis 691). Despite the comment on euphonus ('De pronomine' 11) at p. 131, which cites a letter by Marcus Caesar to Fronto, there is very little on non-grammatical parallels to grammatical diction: pronomina minus quam finita do not evoke leges minus quam perfectae (see Tit. Ulp 2). Nor does R. notice the Greek grammarians, so that at pp. 122-3 the question whether ego may have a vocative ('De pronomine' 4) is considered without reference to their firm denial of that case to the first and third persons.10 Indeed, R. gazes far enough ahead to cite Julian of Toledo, but there is no trace of the immortal debate on the subject between Galbungus et Terrentius in the slightly earlier Virgilius Maro (Epist. 2. 4).
Such complaints, however, seem churlish when so much has been provided. The user of this commentary will find a wealth of parallel passages that, without attempting to cite every derivative statement traceable through the indexes of GL, enable every paragraph of the Regulae Palaemonis to be put in context and will be a convenient point of reference for any scholar concerned with grammatical doctrine on particular questions even if he or she is not concerned with Ps.-Palaemon as such. It is fully worthy of this most valuable edition, which may be confidently recommended to all students of the Roman grammarians.
1. Text in Grammatici Latini, v. 533-47; in his preliminary discussion he asserted (p. 528) that 'Palaemon' offered nothing not found in other grammarians. While originality is not to be expected in such a work, that goes too far.
2. Iulius Valerius: Res gestae Alexandri Macedonis. Adhibitis schedis Roberti Calderan edidit Michaela Rosellini (Stuttgart and Leipzig: B. G. Teubner, 1993).
3. Meaning of course both substantive and adjective, not confined as in the sloppy modern English use to the former.
4. The reader is assumed to know enough of the language not to need telling that the former is added to the final vowel, the latter substituted for it. The exceptional forms in abus, however, are mentioned: in the original recension deabus and filiabus 'discernendi sexus gratia', in the interpolated version this has been replaced by the assertion that such ablatives 'contra artem suscepta sunt', and the list is expanded by the addition of five other substantives together with the numerals duabus and ambabus.
5. It also follows that, when A and the later manuscripts agree, their reading should bear the siglum for their common hyparchetype, namely α; but in fact it is regularly marked with Aμ, as if the two stood on the same footing as O. Readers may be confused at p. 41, l. 12 by the double reference to O, but it is quite correct, for in that manuscript the Regulae Augustini follow directly after Ps.-Palaemon; on fol. 78r it reads significati us with a wide space between the two, on fol. 85r significatiuus.
6. 'Le Regulae Palaemonis e le Regulae Augustini: ipotesi su una relazione ambigua', RFIC 126 (1998), 414-45. R. had also written on the textual transmission: 'Sul testo delle Regulae Palaemonis (GL V 533-547)', in Mario De Nonno, Paolo de Paolis, and Louis Holtz (eds.), Manuscripts and Tradition of Grammatical Texts from Antiquity to the Renaissance: Proceedings of a Conference held at Erice, 16-23 October 1997, 2 vols. (Cassino: Università degli Studi di Cassino, 2000), i. 223-41.
7. At p. 41, l. 17 'quo O ut vid.', O does indeed read quo, but with an abbreviation-mark like an acute accent as if for quoniam; at p. 51, l. 21, text in 'bo', O's reading does not look to me like in bum but iubo with the same symbol, but both the latter and the o are in erasure, with an abbreviation like that for bus, resembling a semicolon, tight against the b. However, in both cases O must be wrong anyway. At p. 55, l. 23 she is right to discern faciat; the ia is written over something else, perhaps &.
8. De doctrina Christiana 3. 3. 7. 14, 4. 10. 23. 65, Enarr. in Ps. 138, section 20. To be fair, these passages will be found in the paragraph of TLL that R. cites.
9. Also in the MSS of Celsus, De medicina 8. 10E, but emended away to match the consistent os (for which Par. lat. 7028 sometimes substitutes ossum.
10. e.g. 'Dion. Thrax' 67. 4 Uhlig with Schol. 81. 3-5, 87. 33-88. 2, 264. 32-265. 12, 384. 8* ad fin. Hilgard; Apoll. Dysc. Synt. 3. 105, Pron. pp. 6. 9, 21. 4-5 Schneider.