by Scott McGill
Table 28.1 Consentius’s De barbarismis et metaplasmis.
uitium uirtus
barbarismus metaplasmus
gruit ~ ruit
tottum ~ totum
quandius ~ quandiu
onor ~ honor
uila ~ uilla
magi ~ magis
prelum ~ plerum
bobis ~ uobis per adiectionem– prothesis
– epenthesis
– paragoge
per detractionem– aphaeresis
– syncope
– apocope
per transmutationem
per immutationem
gnatoque ~ natoque (Virg. Aen. 3,12)
Mauortis ~ Martis (Virg. Aen. 8,630)
admittier ~ admitti (Virg. Aen. 9,231)
mitte ~ omitte (Ter. Ad. 185)
repostum ~ repositum (Virg. Aen. 1,26)
do ~ domum (Enn. ann. 576 V.2 = 587 S.)
Euandre ~ Euander (Virg. Aen. 11,55)
olli ~ illi
Isolated sections that are specifically devoted to rhetoric (de rhetorica) include an extract of Marius Victorinus’s rhetorica placed at the end of his Ars (RhLM 238,20–239,19 = Mariotti ed. 1967, 96), the Frg. Sangermanensis on the synecdoche and the hysteron proteron (GL 5,328,12–20), part of a Frg. Vaticanus on the chria (GL 6,273,7–25), and, most notably, an entire chapter on the figures of thought in Charisius’s fourth book (GL 1,283,15–287,16 = Barwick ed. 19642, 371,29‐375,9). In fact, not only did Charisius, like Donatus, explain the Gorgian figures of speech by the same Ennian quotations that already Rhetorica ad Herennium (4,18) used to this purpose, but he also preserved ancient materials for all the procedures of the ornatus. Parallels with the exegetical field prove that grammatical commentaries to literary authors influenced Charisius’s wide‐ranging choice, including, among others, dialogismos (cf. Donatus Ter. Eun. 46–48), paralipsis (cf. Donatus Ter. Hec. 420,1; 473; Phorm. 168,1; Servius ad Aen. 8, 483), ethologia – prosopopoeia (cf. Donatus Ter. Eun. 15; ad. 308; Hec. 148; Andr. 286,1), and epitrope (cf. Donatus Ter. Ad. 132; 134; 991). This also shows that in the “third part” of the late Artes grammaticae “a double frontier…broke up: the frontier between grammar as system of language and grammar as text explanation; the frontier between grammar and rhetoric” (Baratin and Desbordes 1986, p. 231/2007, p. 81, my translation).
28.5 Syntax
In antiquity, the difference between morphology and syntax was not as clear‐cut as it is today. Syntactic remarks could, therefore, find a place in the morpholexical sections of the grammars, like those on the septimus casus, on verbal constructions, on the use of some conjunctions with the accusative or the ablative, and so forth. Charisius’s De qualitatibus Latini sermonis et temporibus (GL 1,262,25–264,16 = Barwick ed. 19642, 347,1–349,15) and Diomedes’s De declinatione exercitationis chriarum and De coniunctione temporum (GL 1,310,1–29; 388,10–395,10) are particularly lengthy developments within de uerbo and de nomine chapters. An autonomous work, Arusianus Messius’s Exempla elocutionis (GL 7,449–514 = Di Stefano ed. 2011, end of the fourth century CE), classified different nominal and verbal constructions with the help of many examples taken from Terence, Cicero, Sallust, and Virgil (the so‐called quadriga Messii). In the last two books of his Ars, Priscian (end of the fifth–early sixth century CE; Baratin 1989, pp. 365–485) offers the most extensive treatment of syntactic, instead of rhetorical‐stylistic aspects of the oratio (see above, Vices and Virtues). He here explicitly followed Apollonius Dyscolus’s Syntax, but his work became more comprehensive (see below, Lexicon, Lexica) and offered a reinterpretation of the original, Stoic pattern in light of the contemporary, Neoplatonic debate on the Aristotelian categories (Luhtala 2005). This topic became widespread thanks to, among others, the commentaries on Minucianus and Hermogenes and to the work of people involved in framing both the grammatical and philosophical discussion, like Philoponus, a sixth‐century commentator of Aristotle and author of several grammatical treatises. Thus, for example, Priscian systematically used the words substantia and accidens, as only Boethius did, and rewrote his sources in order to show the reason why the noun precedes the verb in the list of the parts of speech (Ars 17, GL 3,116,25–27). In his eyes, action and passion (agere et pati, actio et passio) are the properties of the substance (substantia), whereas Apollonius (constr. 1,16 GG 2/2,18,5–9) took into consideration the dispositions of the body. The minimal utterance (GL 3,164,16–20) is also interpreted as the combination of a noun, that is, a substantia (≃ sôma, ousia, huparxis) and a verb, that is, an accidens (≃ episumbainon). This pattern is so extensively adopted that Priscian assigned accidental determinations not only to the substance but also to its dispositions (GL 3,123,13–124,3 ≃ Apollonius constr. 1,35 GG 2/2,32,9–33,8): So substantia ipsius actus realizes the shift of the verb from its status of accident to that of a substance, whose accidents are the adverbs (Ars grammatica group 2010, pp. 85–105 with notes).
28.6 Lexicon, Lexica
Different semantic problems, from appropriateness (de differentiis, de synonymis) to valence in a bilingual perspective (de idiomatibus), are treated in cumulative lists, contradicting the idea of grammar as structured knowledge (Baratin 1994, pp. 153–154). Only Macrobius wrote a comprehensive essay on verbal morphology from a systematic perspective: His De uerborum Graeci et Latini differentiis uel societatibus (Stoppie, Swiggers, and Wouters 2007) is partially preserved in the Excerpta Bobiensia of the MS Napoli BN lat. 2 (GL 631–633), but it is also the source of the extracts taken by Johannes Scotus Eriugena in order to learn the Greek inflection (GL 5,599–630) and of an anonymous Latin treatise De uerbo ad Seuerianum from Bobbio (De Paolis ed. 1990 combines the first two and puts the third one on the opposite page). If Diomedes treats idioms especially in his section de consensu uerborum cum casibus (GL 1,310,30–320,9; see Baratin 1989, pp. 117–201), the MS Napoli BN lat. IV.A.8 adds at the end of Charisius’s grammar many idiomata, synonyma, and differentiae. These sections, regarded as authentic by Barwick (1922, pp. 56–61) and accordingly edited by him (ed. 19642, 387‐480), reveal indeed some significant parallels with the corresponding passages in Dositheus (see below) and Ars Bobiensis (GL 1,552,23–554,33 = De Nonno ed. 1982, 32,1–35,33). While this is a clear sign of a Charisian origin, Charisius certainly did not conceive these passages in their actual form (De Nonno ed. 1982, 59 n. 7). As for Priscian, he concludes his Ars with a list of about 340 examples extracted from a monolingual syntactic lexicon of Attic authors (like that of anecd. Gr. 1,117–180 Bekker = Petrova ed. 2006) and enriched with Latin parallels (Martorelli 2014): for example, GL 3,281,7 = Rosellini ed. 2015, 10,3, ausculto tibi or te (with Terence Andr. 209 and 536) – ἀκρῶμαι σοῦ or σέ (with Aeschines 3,192). Besides the anonymous De dubiis nominibus (GL 5,571–594 = Glorie ed. 1968: 755–820; mid‐seventh century CE) and Pseudo‐Caper’s de uerbis dubiis (GL 7,107,4–112, a list of differentiae including late words like atramentarium and lignarium), independent works on this topic are a typical product of the late Latin and medieval culture (Brugnoli 1955; Magallón García 1996), which could stem from different grammatical sources and from commentaries to literary authors (Stok 2008).
The bilingual Ars grammatica written by Dositheus (fourth century CE) is followed by a long list of Latin verbs with their Greek parallels (GL 7,430,1–436,4 = Tolkiehn ed. 1913, 95,7–104,10, Bonnet ed. 2005 leaves out this section). After this work, the MS Sankt Gallen StiftsB 902 also gives a collection of bilingual texts, usually referred to as Hermeneumata Pseudodositheana and known in nine different versions: the eight that are edited in volume 3 of the Corpus glossariorum Latinorum and the ninth of the MS Wien ÖNB suppl. gr. 43, containing the transcription of a lost manuscript by Conrad Celtes (1495). Parts of these texts were conceived as a tool for learning both elementary Latin and elementary Greek; others were addressed to adults who needed to learn Latin as the lang
uage of the bureaucracy in the eastern part of the Empire or Greek in provinces like Egypt. Eventually, they were used in the West by a more erudite and select Latin‐speaking public, who were specifically interested in Greek. The Hermeneumata are formed by a multiplicity of elements: Alphabetical and thematic (capitula) bilingual glossaries share many resemblances with the Appendix Probi (see above, Sublexical Level); the colloquia provide the story of a pupil’s everyday life as a bilingual conversational guide (Dickey 2012–2014); and other bilingual texts were added for reading practice.
The diffusion and mutual relationships of monolingual lexica in late antiquity are an extremely complicated issue (see the useful overview of the frequently obsolete and still incomplete editions of the Greek tradition by Dickey 2007, pp. 87–103; Dionisotti 1996 gives an excellent survey of the Latin corpora edited by Löwe and Goetz 1888–1923 and by Lindsay 1926–1931). Among them, Nonius Marcellus’s De compendiosa doctrina clearly stands out as the most important Latin glossematical work of the fourth century (Barabino and Bertini 1967–1997; Bertini 2000–2005), with a vivid interest in grammatical problems, synonyms, and anomalies (books 1–12), as well as in lexical families (books 13–20). The sources are quoted following recurrent patterns, because Nonius takes extracts from 40 authors, which tend to appear in the same order (Lindsay 1901, with the noteworthy corrections by Cadoni 1987). This work was used by professional grammarians, as is shown by Priscian’s use of Nonius’s book 11 as a source for many of the doublets of more or less rare adverbial forms in book 15 of his Ars (GL 3,70,4–71,6 and 77,7–12). Nonius always provides the literary attestations (mostly among the ueteres), since they prove that these forms are acceptable, whereas Priscian is mainly interested in morphological aspects; thus he rejects memore (77,9–12), which is parallel to memoriter, even if it is used by Pomponius (CRF3 109), because it does not fit into the pattern of the neutral adverbial forms ending in ‐e (Ars grammatica group 2013, pp. 52–54).
28.7 Metrics
Some of the Artes incorporate a section on metrics (De arte metrica: De Nonno 1990b; useful tools: Luque Moreno et al. 1987–2007; Morelli 2006). Sometimes related subjects are also developed, as in Diomedes’s classification of the poetic literary forms (De poematibus: GL 1,482,13–492,14). Some commentaries also include this type of information, as shown by Evanthius’s De fabula and the Excerpta de comoedia which precede Donatus’s Terence. Traditionally, distinct works were written on metrics that pertained more to music (Cristante 1987, pp. 64–67; see e.g. Pseudo‐Censorinus’s Epitoma disciplinarum, GL 6,607–617 = Sallmann ed. 1983, 71,7–86,15) and rhetoric than to grammar. Within the grammatical field, these works were devoted more to criticism than to normative linguistic teaching (see above, A Textual Typology). Eventually, the inclusive character of the late Latin grammar inspired comprehensive, systematic works: Sacerdos ended up adding a third book, De metris, to his previous ones, whereas Diomedes planned from the very beginning a third section of his work on feet, poems, and meters, as did Charisius, although this part of his Ars is lost. Similar phenomena proliferate in the textual transmission: Aphthonius’s monumental and erudite four books, De metris omnibus, were added to Victorinus’s grammar without any indication, since they formed one single, unbroken text (Keil’s edition still follows this incorrect presentation). Some overlaps with the rhetorical sections of the Artes were possible: So the treatise on prosody De ultimis syllabis ad Caelestinum ends with the section de metaplasmis (GL 6,262,18–264,16) and, on the other side, Consentius’s De barbarismis et metaplasmis ends with the section de scandendis uersibus (GL5,398,16–404,8 = Niedermann ed. 1937, 22,19–32,20). Since reading Horace (the only lyric poet who was taught at school starting from the second century) required multiple metrical competences, the commentary on Odes and Epodes of the pseudo‐Acronian corpus in MS Paris BNF 7900A indicates the name of the verse as well as its interpretation and metrical scansion before the analysis of each poem. This information is also found in monographs ad hoc (De Nonno 1998). So, for example, Persius’s friend Caesius Bassus first wrote a De metris on this topic (GL 6,255–272) and spread in the Roman world the “Pergamian” metrical system that obtained all metrical forms from the hexameter and the iambic trimeter through the quadripertita ratio (see above, Vices and Virtues); his work is transmitted with other metrical texts from Bobbio, especially Fortunatianus’s De metris Horatianis (GL 6,278–304 = Morelli ed. 2011–2012, fourth century). Servius, who frequently quoted Horace in his commentary to Virgil and wrote two other essays on prosody (de finalibus) and metrics (centimeter: GL 4,456–467 = Elice ed. 2013), is also the author of a de metris Horatii (GL 4,468–472). Another example of this genre is an apocryphal De metris Horatianis (GL 6,174–184) that follows Aphthonius’s work in the MSS. Priscian, who paid attention to the relationship between prosody and morphology and put a metrical précis at the beginning of his Partitiones (GL 3,459,3–461,14 = Passalacqua ed. 1999, 45,4–48,10; see below, Interpretari), participated in the debate on the ancient theater. In his De metris fabularum Terentii et aliorum comicorum (GL 3,418–429 = Passalacqua ed. 1987, 19–32), written with two other heterogeneous opuscula, he strove to demonstrate that Terence wrote in verse, not in prose, even if the average speaker lost a precise feeling of the vocalic quantities. To the same end, Rufinus of Antioch wrote a Commentarium in metra Terentiana, whose excerpts are conserved with those from his Commentarium de compositione et de numeris oratorum (GL 6,554–578 = D’Alessandro ed. 2004). Latin monographs on metrics, for their part, reveal a “boxing‐up” structure similar to that of the Artes grammaticae: The versified treatise De littera, de syllaba, de metris by the former African poet Terentianus Maurus (GL 6,325–413 = Cignolo ed. 2002) explicitly follows this pattern.
28.8 Interpretari
Considered as one of the tools of the grammar (see above, A Textual Typology), the “historical” analysis of the content of a literary work gives all the relevant information on topics like mythology, law, and chronology. Specific monographs were written on this type of erudite research. Anyone who is interested in the scanty fragments of this production can find in the Grammaticae Romanae fragmenta (edited by Funaioli 1907 and Mazzarino 19552) many examples of works de dis or περὶ θεῶν, de iure ciuili, chronica, as well as uitae of literary authors and works de poetis. This was bookish learning, conceived in relation to literary texts and extracted from them. Eventually, in late antiquity, it reappears in an atomized form, within word‐to‐word commentaries, which became sorts of condensed encyclopedias. Besides anonymous explanations on the margin of the manuscripts like Virgil’s scholia Veronensia and Terence’s scholia Bembina, vast commentaries addressed to the public of the school were written by renowned grammarians, who followed the long‐standing and illustrious Greek tradition (see Dickey 2007, pp. 18–71). Donatus wrote a commentary on Terence, which is still preserved, although in an abridged form that lacks Heautontimoroumenos. His now lost commentary on Virgil became the basis of Servius’s own, now preserved in two redactions in which the second one (Seruius auctus or Seruius Danielis) is a late expansion with different, intertwined stages and levels. At the top of this cultural work one can find philosophical, Neoplatonic‐oriented commentaries, like those of Tiberius Claudius Donatus on Virgil and of Marius Victorinus on Cicero’s De inuentione. On the other hand, a basic type of commentary, limited to a form of parsing grammar reminiscent of the Homeric epimerismoi, is represented by Priscian’s partitiones (GL 3,459–515 = Passalacqua ed. 1999, 45–128), an analysis of the first 12 lines of Virgil’s Aeneid in the same catechetic form as Donatus’s Ars minor (Glück 1967). Interestingly enough, after a grammatical section containing inflectional and prosodic regulae, the Frg. Bobiense de nomine includes a glossographic section (GL 7,541,26–544,44 = Mariotti ed. 1984, § 14–158), with a mixture of glosses and scholia to Virgil, Horace, Lucan, Juvenal, and Sallust, the most read authors of late antiquity (generally not explicitly referred to in t
he text). In the case of Virgil, it is possible to demonstrate that Servius and Seruius Danielis are the main sources, of which the Frg. Bobiense sometimes transmits the original, fuller form that has been lost in the direct tradition of these commentaries (compare, e.g., Frg. Bob. de nom. GL 7,544,43–44 = Mariotti ed. 1984, § 158 Epicurei definiun < t > quidquid uisu aut tactu subiacet, hoc corpus esse, unde et umbras corpora esse dicunt with Servius Aen. 6,303 “corpora cumba”: omne quod potest uideri, corpus dicitur; see also Mariotti ed. 1984, pp. 44–45).
Thanks to these works, reading a classical author like Virgil allowed people to obtain a form of universal knowledge, which the commentators developed and nourished (Ziolkowski and Putnam 2008). So, according to Servius, Virgil was totus quidem…scientia plenus (ad Aen. 6, prooem.), because uariam scientiam suo inserit carmini (ad Aen. 2,148). Besides conveying some linguistic models (Uhl 1998), resuscitating ancient information on institutions and realia of the classical world had an ideological resonance in a context of mixed Christian and pagan culture. It revealed a desire to preserve a heritage that risked being lost (Casali and Stok 2008; Bouquet, Ménial, and Ramires 2011; Stok 2013; Garcea, Lhommé, and Vallat 2016). In this period of transition, grammar ended up being the ideal medium for this preservation, both for its formal aspects and for its content. Christians were at least as interested in commenting on written texts as the imperial‐era grammarians. Throughout late antiquity Christians were still trained within a classical school system in which they increasingly played important roles as professors of grammar and rhetoric, like Saint Augustine. In this way, an immense heritage, even if it resolved itself into fragments, could be the object of a “reconversion” that at the same time permitted its survival.
