A tale of two languages: Macrobius on Greek and Latin

1.Comparing Latin and Greek

Intimate knowledge of Greek language and philosophy had formed part of the cultural baggage of a Roman intellectual from the first century BCE onward.11.Generally on bilingualism involving Latin see Adams (2003)Adams, James N. 2003 Bilingualism and the Latin language. Cambridge: Cambridge Univesity Press. Adams, James N. 2003 Bilingualism and the Latin language. Cambridge: Cambridge Univesity Press. , in particular his remarks on elite bilingualism (2003Adams, James N. 2003 Bilingualism and the Latin language. Cambridge: Cambridge Univesity Press. Adams, James N. 2003 Bilingualism and the Latin language. Cambridge: Cambridge Univesity Press. : 9–14). On the place of Greek in education in Rome see Swain (2002Swain, Simon 2002 “Bilingualism in Cicero? The evidence of code-switching”. Bilingualism in Ancient Society: Language contact and the written text ed. by James N. Adams, Mark Janse & Simon Swain, 128–167. Oxford: Oxford University Press. Swain, Simon 2002 “Bilingualism in Cicero? The evidence of code-switching”. Bilingualism in Ancient Society: Language contact and the written text ed. by James N. Adams, Mark Janse & Simon Swain, 128–167. Oxford: Oxford University Press. : 130–138). On the varying degrees of Greek-Latin bilingualism see Biville (2002)Biville, Frédérique 2002 “The Graeco-Romans and Graeco-Latin: A terminological framework for cases of bilingualism”. Bilingualism in Ancient Society: Language contact and the written text ed. by James N. Adams, Mark Janse & Simon Swain, 77–102. Oxford: Oxford University Press. Biville, Frédérique 2002 “The Graeco-Romans and Graeco-Latin: A terminological framework for cases of bilingualism”. Bilingualism in Ancient Society: Language contact and the written text ed. by James N. Adams, Mark Janse & Simon Swain, 77–102. Oxford: Oxford University Press. , with many examples, ranging from Romans erroneously believing they could speak Greek to Romans exaggerating their love for all things Greek. See also Rochette (2010)Rochette, Bruno 2010 “Greek and Latin bilingualism”. A Companion to the Ancient Greek Language ed. by Egbert J. Bakker, 281–293. New York: Blackwell. Rochette, Bruno 2010 “Greek and Latin bilingualism”. A Companion to the Ancient Greek Language ed. by Egbert J. Bakker, 281–293. New York: Blackwell. . The Greek language with which these intellectuals were familiar was the literary language, rather than the colloquial language of the Greek-speaking world. Valette-Cagnac (2005Valette-Cagnac, Emmanuelle 2005 “ ‘’Plus attique que la langue des Athéniens’: Le grec imaginaire des Romains”. Façons de parler grec à Rome ed. by Florence Dupont & Emmanuelle Valette-Cagnac, 37–80. Paris: Belin.Valette-Cagnac, Emmanuelle 2005 “ ‘’Plus attique que la langue des Athéniens’: Le grec imaginaire des Romains”. Façons de parler grec à Rome ed. by Florence Dupont & Emmanuelle Valette-Cagnac, 37–80. Paris: Belin.: 43) aptly characterizes the linguistic situation as follows: “... dans la même société, et à la même époque, l’usage du grec fut tantôt perçu comme un signe de culture et de raffinement, tantôt au contraire comme un indice de basse extraction sociale.”

The literary Greek language was the only variety of that language in which the Latin grammarians were interested.22.‘Vulgar’ Greek expressions were sometimes employed in literary prose (Holford-Strevens 2003Holford-Strevens, Leofranc 2003 Aulus Gellius: An Antonine scholar and his achievement. Revised edition Oxford: Oxford University Press. Holford-Strevens, Leofranc 2003 Aulus Gellius: An Antonine scholar and his achievement. Revised edition Oxford: Oxford University Press. : 226), but colloquial Greek was not within the speech repertoire of Roman intellectuals (Souris & Nigdelis 2001Souris, George A. & Pantelis Nigdelis 2001 “The parallel use of Greek and Latin in the Greco-Roman world”. A History of Ancient Greek, from the Beginnings to late Antiquity ed. by Anastassios-Fivos Christidis, 897–902. Cambridge: Cambridge University Press.Souris, George A. & Pantelis Nigdelis 2001 “The parallel use of Greek and Latin in the Greco-Roman world”. A History of Ancient Greek, from the Beginnings to late Antiquity ed. by Anastassios-Fivos Christidis, 897–902. Cambridge: Cambridge University Press.). In one way or another all of them referred to Greek in their treatises on Latin grammar, usually for didactic purposes. A large part of their audience consisted of non-native speakers wishing to learn Latin, so that the main function of comparing Latin and Greek was to facilitate the students’ acquisition of Latin as a second language. Translation was not used in the curriculum as a tool to compare the two languages, and it did not serve as an aid to practice the grammatical rules, either, but solely as one of the ways to understand the source text.

Among the treatises using Greek materials Dositheus’ (fourth century CE) Ars grammatica ( GL GL = Grammatici latini Ed. by Heinrich Keil. 8 vols. Leipzig: B.G. Teubner 1864 (Repr., Hildesheim: G. Olms 2007.)GL = Grammatici latini Ed. by Heinrich Keil. 81 vols. Leipzig: B.G. Teubner 1864 (Repr., Hildesheim: G. Olms 2007.) IV, 376–436) stands out because it consisted of a Latin text accompanied by a Greek translation, or as Bonnet (2007)Bonnet, Guillaume 2007 “Glose ou traduction? La version grecque de l’ars grammatica de Dosithée”. Bilinguisme et terminologie grammaticale gréco-latine. (= Orbis/ Supplementa, 27) ed. by Louis Basset et al., 191–199. Leuven, Paris & Dudley, MA: Peeters.Bonnet, Guillaume 2007 “Glose ou traduction? La version grecque de l’ars grammatica de Dosithée”. Bilinguisme et terminologie grammaticale gréco-latine. (=Orbis/ Supplementa, 27) ed. by Louis Basset et al., 191–199. Leuven, Paris & Dudley, MA: Peeters. explains, a Greek gloss intended for Greek students who wished to understand the Latin text of the grammar.33.For a comparison between Dositheus and Macrobius see Scappaticcio (2015Scappaticcio, Maria Chiara 2015 Artes Grammaticae in frammenti: I testi grammaticali latini e bilingui greco-latini su papiro. Edizione commentata. Berlin and Boston: de Gruyter. Scappaticcio, Maria Chiara 2015 Artes Grammaticae in frammenti: I testi grammaticali latini e bilingui greco-latini su papiro. Edizione commentata. Berlin and Boston: de Gruyter. : 38–39). Dickey (2016a)Dickey, Eleanor 2016a “The authorship of the Greek version of Dositheus’ grammar and what it tells us about the grammar’s original use”. The Latin of the Grammarians: Reflections about language in the Roman world ed. by Rolando Ferri & Anna Zago, 205–235. Turnhout: Brepols.Dickey, Eleanor 2016a “The authorship of the Greek version of Dositheus’ grammar and what it tells us about the grammar’s original use”. The Latin of the Grammarians: Reflections about language in the Roman world ed. by Rolando Ferri & Anna Zago, 205–235. Turnhout: Brepols. assumes that Dositheus was probably a native speaker of Greek, who used the Greek translation of parts of his Latin grammar to help beginning students to understand the Latin text.44.Not always with great success. Students learned from Dositheus’ translation that in Latin the past tense has three subdivisions, praeteritum imperfectum, praeteritum perfectum, praeteritum plusquamperfectum (Ars 406.11), and they were told that in Greek this corresponds to παρῳχημένος παρατατικός, παρῳχημένος ἀόριστος, παρῳχημένος ὑπερσυντελικός. While it is correct to liken the Latin perfect to the Greek aorist (see below), students must have wondered what then the equivalent of the Greek perfect was. She observes that those parts of the text that had been fully translated represented the beginners’ materials in the curriculum, when students needed more support, while in the later curriculum translation was used less often.

Comparisons between Latin and Greek started from the basic assumption that the two languages were closely related or, as Macrobius ( De diff Macrobius, De diff. = Macrobii Theodosii, De verborum graeci et latini differentiis vel societatibus excerpta Ed. by Paolo De Paolis. Urbino: Quattro Venti 1990.Macrobius, De diff. = Macrobii Theodosii, De verborum graeci et latini differentiis vel societatibus excerpta Ed. by Paolo De Paolis. Urbino: Quattro Venti 1990.. 5.6) phrases it, they had a coniunctissima cognatio. Some scholars even maintained that Latin derived from Greek, specifically from the Aeolic dialect, which was believed to have been brought to Italy by a group of Arcadians led by Euander.55. Dionysius of HalicarnassusDionysius of Halicarnassus, Antiquitates = Dionysius of Halicarnassus, Antiquitatum Romanarum quae supersunt Ed. by Karl Jacoby. 4 vols. Leipzig: B.G. Teubner 1885.Dionysius of Halicarnassus, Antiquitates = Dionysius of Halicarnassus, Antiquitatum Romanarum quae supersunt Ed. by Karl Jacoby. 41 vols. Leipzig: B.G. Teubner 1885. was the most explicit representative of this theory (Antiquitates I, 90.1). About the ‘Aeolian thesis’, see Schöpsdau (1992Schöpsdau, Klaus 1992 “Vergleiche zwischen Lateinisch und Griechisch in der antiken Welt”. Zum Umgang mit fremden Sprachen in der griechisch-römischen Welt ed. by Carl Werner Müller, Carl Werner, Kurt Sier & Jürgen Werner, 115–136. Stuttgart: F. Steiner.Schöpsdau, Klaus 1992 “Vergleiche zwischen Lateinisch und Griechisch in der antiken Welt”. Zum Umgang mit fremden Sprachen in der griechisch-römischen Welt ed. by Carl Werner Müller, Carl Werner, Kurt Sier & Jürgen Werner, 115–136. Stuttgart: F. Steiner.: 117–119); Stevens (2006)Stevens, Benjamin 2006 “Aeolism: Latin as a dialect of Greek”. The Classical Journal 102:2. 115–144.Stevens, Benjamin 2006 “Aeolism: Latin as a dialect of Greek”. The Classical Journal 102:2. 115–144.. Gitner (2015)Gitner, Adam 2015 “Varro Aeolicus: Latin’s Affinity with Greek”. Varro Varius: The polymath of the Roman world ed. by David J. Butterfield, 33–50. Cambridge: The Cambridge Philological Society.Gitner, Adam 2015 “Varro Aeolicus: Latin’s Affinity with Greek”. Varro Varius: The polymath of the Roman world ed. by David J. Butterfield, 33–50. Cambridge: The Cambridge Philological Society. traces the development of the theory from Varro to Macrobius and Priscian. De Paolis (2015)De Paolis, Paolo 2015 “La parentela linguistica fra greco e latino nella tradizione grammaticale latina”. Latin linguistics in the early 21st century: Acts of the 16th International Colloquium on Latin Linguistics, Uppsala, June 6th–11th, 2011 ed. by Gerd V. M. Haverling, 610–624. Uppsala: Uppsala Universitet.De Paolis, Paolo 2015 “La parentela linguistica fra greco e latino nella tradizione grammaticale latina”. Latin linguistics in the early 21st century: Acts of the 16th International Colloquium on Latin Linguistics, Uppsala, June 6th–11th, 20111 ed. by Gerd V. M. Haverling, 610–624. Uppsala: Uppsala Universitet. points out that all proponents of this thesis were Greek grammarians active in Rome, while grammarians in the east all but ignore it. When confronted with differences between Latin and Greek, Latin grammarians had to find a way of coping with these. Some of the differences involved the lacking of a Greek category in Latin. One solution was developed in the framework of the Aeolian thesis. In its evolution Latin was thought to have followed Aeolic Greek, either by losing, just like Aeolic, some of the categories of other Greek dialects, or inversely, by retaining a feature like the Aeolic digamma that the other dialects had lost.66.On the Aeolic digamma see Schöpsdau (1992Schöpsdau, Klaus 1992 “Vergleiche zwischen Lateinisch und Griechisch in der antiken Welt”. Zum Umgang mit fremden Sprachen in der griechisch-römischen Welt ed. by Carl Werner Müller, Carl Werner, Kurt Sier & Jürgen Werner, 115–136. Stuttgart: F. Steiner.Schöpsdau, Klaus 1992 “Vergleiche zwischen Lateinisch und Griechisch in der antiken Welt”. Zum Umgang mit fremden Sprachen in der griechisch-römischen Welt ed. by Carl Werner Müller, Carl Werner, Kurt Sier & Jürgen Werner, 115–136. Stuttgart: F. Steiner.: 119); Stevens (2006Stevens, Benjamin 2006 “Aeolism: Latin as a dialect of Greek”. The Classical Journal 102:2. 115–144.Stevens, Benjamin 2006 “Aeolism: Latin as a dialect of Greek”. The Classical Journal 102:2. 115–144.: 127). Macrobius ( De diff Macrobius, De diff. = Macrobii Theodosii, De verborum graeci et latini differentiis vel societatibus excerpta Ed. by Paolo De Paolis. Urbino: Quattro Venti 1990.Macrobius, De diff. = Macrobii Theodosii, De verborum graeci et latini differentiis vel societatibus excerpta Ed. by Paolo De Paolis. Urbino: Quattro Venti 1990.. 51–53) mentions the digamma of the Aeolians as the source of Latin F to explain why f in fallo can be reduplicated as f in fefelli because it is not an aspirated consonant. An alternative line of reasoning was proposed by those grammarians who, against the majority, claimed that the missing category did exist in Latin after all. They did so, for instance, by claiming that the Latin perfect did have a dual form, legēre, contrasting with the plural legerunt (Denecker 2023Denecker, Tim 2023 “T(w)o be or not t(w)o be: The dualis numerus according to Latin grammarians up to the Early Middle Ages”. Roman perspectives on linguistic diversity: Guardians of a changing language ed. by Adam Gitner, 131–142. Oxford: Oxford University Press. Denecker, Tim 2023 “T(w)o be or not t(w)o be: The dualis numerus according to Latin grammarians up to the Early Middle Ages”. Roman perspectives on linguistic diversity: Guardians of a changing language ed. by Adam Gitner, 131–142. Oxford: Oxford University Press. ).77.In the same way the absence of a dual in Latin is explained by the fact that this category was an innovation in Greek belonging to a later period, which did not reach the Aeolic dialect and was therefore also lacking in Latin (Choeroboscus, Scholia Choeroboscus, Scholia I = Georgius Choeroboscus, Prolegomena et scholia in Theodosii Alexandrini Canones isagogicas de flexione nominum Ed. by Alfred Hilgard, GG IV/1, 103–417. (Repr., Hildesheim: G. Olms 2007.)Choeroboscus, Scholia I = Georgius Choeroboscus, Prolegomena et scholia in Theodosii Alexandrini Canones isagogicas de flexione nominum Ed. by Alfred Hilgard, GG IV/11, 103–417. (Repr., Hildesheim: G. Olms 2007.) II, 34.4–13), cf. De Paolis (2015De Paolis, Paolo 2015 “La parentela linguistica fra greco e latino nella tradizione grammaticale latina”. Latin linguistics in the early 21st century: Acts of the 16th International Colloquium on Latin Linguistics, Uppsala, June 6th–11th, 2011 ed. by Gerd V. M. Haverling, 610–624. Uppsala: Uppsala Universitet.De Paolis, Paolo 2015 “La parentela linguistica fra greco e latino nella tradizione grammaticale latina”. Latin linguistics in the early 21st century: Acts of the 16th International Colloquium on Latin Linguistics, Uppsala, June 6th–11th, 20111 ed. by Gerd V. M. Haverling, 610–624. Uppsala: Uppsala Universitet.: 612).

An example of the opposite case is that of the Latin ablative which is lacking in Greek, where the ablative corresponds to either the dative or the genitive. In addition, there is the somewhat mysterious casus septimus, equally lacking in Greek, which has the same form as the ablative, but is used for different functions (e.g. Diomedes, Ars Diomedes, Ars = Diomedes, Artis grammaticae libri III Ed. by Heinrich Keil, GL I, 297–529. (Repr., Hildesheim: G. Olms 2007.)Diomedes, Ars = Diomedes, Artis grammaticae libri III Ed. by Heinrich Keil, GL I1, 297–529. (Repr., Hildesheim: G. Olms 2007.) 317.34–318.22).88.See Schöpsdau (1992Schöpsdau, Klaus 1992 “Vergleiche zwischen Lateinisch und Griechisch in der antiken Welt”. Zum Umgang mit fremden Sprachen in der griechisch-römischen Welt ed. by Carl Werner Müller, Carl Werner, Kurt Sier & Jürgen Werner, 115–136. Stuttgart: F. Steiner.Schöpsdau, Klaus 1992 “Vergleiche zwischen Lateinisch und Griechisch in der antiken Welt”. Zum Umgang mit fremden Sprachen in der griechisch-römischen Welt ed. by Carl Werner Müller, Carl Werner, Kurt Sier & Jürgen Werner, 115–136. Stuttgart: F. Steiner.: 130–131); Benedetti (2018)Benedetti, Marina 2018 “Specchiarsi nell’altro: I grammatici, oltre l’ablativo”. L’altro nel mondo antico: Riflessioni linguistiche ed. by Marina Benedetti, Carla Bruno, Felicia Logozzo & Liana Tronci, 7–26. Pisa: Pacini.Benedetti, Marina 2018 “Specchiarsi nell’altro: I grammatici, oltre l’ablativo”. L’altro nel mondo antico: Riflessioni linguistiche ed. by Marina Benedetti, Carla Bruno, Felicia Logozzo & Liana Tronci, 7–26. Pisa: Pacini.. Macrobius just mentions both the ablative and the septimus casus without further explanation: “The ablative or what they call the ‘seventh case’ is something Greece has not received” (ablativum enim vel quem dicunt septimum casum Graecia non recipit; De diff. 11.2).99.Note the choice of the verb recipit in this citation, which appears to imply that there was a period before the formation of the present-day languages when some categories were still being distributed.

2.Macrobius’ treatise comparing Greek and Latin

Compared with grammarians who occasionally referred to Greek in their treatises on Latin grammar, the fifth-century author Macrobius stands out as an exception when it comes to the purpose of such a comparison.1010.On Macrobius see Cameron (1966)Cameron, Alan 1966 “The date and identity of Macrobius”. Journal of Roman Studies 56:1/2. 25–38. Cameron, Alan 1966 “The date and identity of Macrobius”. Journal of Roman Studies 56:1/2. 25–38. . His full name was Ambrosius Theodosius Macrobius, but he is usually referred to as Macrobius. Zetzel deals with the literature about Macrobius’ main work, the Saturnalia (2018: 263–264), his commentary on the Somnium Scipionis (2018: 261), and his treatise on the differences and commonalities between the Latin and Greek verb (2018: 300–301). Kaster (1988Kaster, Robert A. 1988 Guardians of language: The grammarians and society in Late Antiquity. Berkeley, Cal.: University of California Press. Kaster, Robert A. 1988 Guardians of language: The grammarians and society in Late Antiquity. Berkeley, Cal.: University of California Press. : 60–62, 172–174) frequently refers to Macrobius, but does not include him in his prosopographical analysis of Latin grammarians precisely because he was not a professional grammarian and did not want to be seen as one (1988: 196). He was not the only and not even the first scholar to juxtapose the structure of Greek and Latin, but his goal differed from that of the grammarians preceding him. Macrobius was not a professional grammarian, but an educated intellectual, a man of learning who belonged to the cultural elite without being a member of the aristocracy (Kaster 1980Kaster, Robert A. 1980 “Macrobius and Servius: Verecundia and the grammarian’s function”. Harvard Studies in Classical Philology 84. 219–262. Kaster, Robert A. 1980 “Macrobius and Servius: Verecundia and the grammarian’s function”. Harvard Studies in Classical Philology 841. 219–262. : 226). His fame as an author rests on two works, the Saturnalia, a fictitious conversation between luminaries from an earlier age, and a commentary on Cicero’sCicero, Ad Quintum fratrem = M. Tullius Cicero, Epistulae ad Quintum fratrem et M. Brutum Ed. by David Roy Shackleton Bailey. Cambridge: Cambridge University Press 1980.Cicero, Ad Quintum fratrem = M. Tullius Cicero, Epistulae ad Quintum fratrem et M. Brutum Ed. by David Roy Shackleton Bailey. Cambridge: Cambridge University Press 1980. Somnium Scipionis, in which he engaged in Neo-Platonic observations. But he also published a short treatise on the differences and commonalities of the Greek and Latin verbal systems. Macrobius did not have a didactic purpose in mind when he compared the two languages, but he studied them purely as a matter of interest, as a true comparative linguist. Macrobius was convinced that fundamentally the two languages had a similar structure, “almost to the point where learning the grammar of one, would acquaint you with the grammar of the other” (ut propemodum qui utramvis artem didicerit ambas noverit).1111.This quote is found in the prooemium of Johannes Scotus’ version of De differentiis ( GL GL = Grammatici latini Ed. by Heinrich Keil. 8 vols. Leipzig: B.G. Teubner 1864 (Repr., Hildesheim: G. Olms 2007.)GL = Grammatici latini Ed. by Heinrich Keil. 81 vols. Leipzig: B.G. Teubner 1864 (Repr., Hildesheim: G. Olms 2007.) V, 599.8–10), as well as in the anonymous treatise De verboAnonymus, De verbo = Anonymus, De verbo Ed. by Heinrich Keil, GL V, 634–654; ed. by Marina Passalacqua, Tre testi grammaticali bobbiesi ( GL V 555–566; 634–654; IV 207–217 Keil). Rome: Edizioni di Storia e Letteratura 1984.Anonymus, De verbo = Anonymus, De verbo Ed. by Heinrich Keil, GL V1, 634–654; ed. by Marina Passalacqua, Tre testi grammaticali bobbiesi (GL V1 555–566; 634–654; IV1 207–217 Keil). Rome: Edizioni di Storia e Letteratura 1984. ( GL GL = Grammatici latini Ed. by Heinrich Keil. 8 vols. Leipzig: B.G. Teubner 1864 (Repr., Hildesheim: G. Olms 2007.)GL = Grammatici latini Ed. by Heinrich Keil. 81 vols. Leipzig: B.G. Teubner 1864 (Repr., Hildesheim: G. Olms 2007.) V, 631.8–9). This did not mean that the two languages were exactly the same: after all, when things are similar, there must always be points of difference, otherwise they would be identical. Such differences are referred to by the term idiomata, while proprietates indicate the special features of either language (quasdam proprietates, De diff. 7). Stoppie et al. (2007Stoppie, Karen, Pierre Swiggers & Alfons Wouters 2007 “La terminologie grammaticale en contexte bilingue: Macrobe et l’analyse de la diathèse verbale”. Bilinguisme et terminologie grammaticale gréco-latine. (= Orbis/ Supplementa, 27) ed. by Louis Basset et al., 201–224. Leuven, Paris & Dudley, MA: Peeters.Stoppie, Karen, Pierre Swiggers & Alfons Wouters 2007 “La terminologie grammaticale en contexte bilingue: Macrobe et l’analyse de la diathèse verbale”. Bilinguisme et terminologie grammaticale gréco-latine. (=Orbis/ Supplementa, 27) ed. by Louis Basset et al., 201–224. Leuven, Paris & Dudley, MA: Peeters.: 204) interpret the latter term normatively when they speak about “des constructions reçues dans l’une des deux langues, mais irrecevables dans l’autre”. It is doubtful, however, that this normative interpretation corresponds to what the grammarians believed. The absence of a category in one of the two languages was not a matter of being unacceptable but just meant that, coincidentally, one of the two had received a certain category, while the other had not.

Macrobius’ treatise has a complicated manuscript history. The part of the treatise published by Heinrich Keil under the title De verborum graeci et latini differentiis et societatibus excerpta ( GL GL = Grammatici latini Ed. by Heinrich Keil. 8 vols. Leipzig: B.G. Teubner 1864 (Repr., Hildesheim: G. Olms 2007.)GL = Grammatici latini Ed. by Heinrich Keil. 81 vols. Leipzig: B.G. Teubner 1864 (Repr., Hildesheim: G. Olms 2007.) V, 599–629; re-edited by De Paolis in 1990De Paolis, Paolo 1990 “Per una nuova edizione critica del De differentiis di Macrobio”. Dicti studiosus: Scritti di filologia offerti a Scevola Mariotti dai suoi allievi, 259–293. Urbino: QuattroVenti.De Paolis, Paolo 1990 “Per una nuova edizione critica del De differentiis di Macrobio”. Dicti studiosus: Scritti di filologia offerti a Scevola Mariotti dai suoi allievi, 259–293. Urbino: QuattroVenti.) is actually an excerpt by the ninth-century scholar Johannes Scotus Eriugena, who originally came from Scotland, but migrated to France, where he taught Greek at the court of Charles the Bald in Laon. For the most part, Johannes Scotus took over from Macrobius’ treatise the sections dealing with Greek grammar: information about Greek was an eagerly sought after commodity at his time.1212.For the scholarly drive in the Carolingian period to find information about the Greek language see also Herren (2015)Herren, Michael 2015 “Pelasgian fountains: Learning Greek in the early Middle Ages”. Learning Latin and Greek from Antiquity to the Present ed. by Elizabeth P. Archibald, William Brockliss & Jonathan Gnoza, 65–82. Cambridge: Cambridge University Press. Herren, Michael 2015 “Pelasgian fountains: Learning Greek in the early Middle Ages”. Learning Latin and Greek from Antiquity to the Present ed. by Elizabeth P. Archibald, William Brockliss & Jonathan Gnoza, 65–82. Cambridge: Cambridge University Press. . If this view is correct, the anonymous treatise De verboAnonymus, De verbo = Anonymus, De verbo Ed. by Heinrich Keil, GL V, 634–654; ed. by Marina Passalacqua, Tre testi grammaticali bobbiesi ( GL V 555–566; 634–654; IV 207–217 Keil). Rome: Edizioni di Storia e Letteratura 1984.Anonymus, De verbo = Anonymus, De verbo Ed. by Heinrich Keil, GL V1, 634–654; ed. by Marina Passalacqua, Tre testi grammaticali bobbiesi (GL V1 555–566; 634–654; IV1 207–217 Keil). Rome: Edizioni di Storia e Letteratura 1984. published by Keil ( GL GL = Grammatici latini Ed. by Heinrich Keil. 8 vols. Leipzig: B.G. Teubner 1864 (Repr., Hildesheim: G. Olms 2007.)GL = Grammatici latini Ed. by Heinrich Keil. 81 vols. Leipzig: B.G. Teubner 1864 (Repr., Hildesheim: G. Olms 2007.) V, 634–654, re-edited by Passalacqua 1984: 21–60, and reproduced in parallel with the edition of De differentiis by De Paolis in 1990De Paolis, Paolo 1990 “Per una nuova edizione critica del De differentiis di Macrobio”. Dicti studiosus: Scritti di filologia offerti a Scevola Mariotti dai suoi allievi, 259–293. Urbino: QuattroVenti.De Paolis, Paolo 1990 “Per una nuova edizione critica del De differentiis di Macrobio”. Dicti studiosus: Scritti di filologia offerti a Scevola Mariotti dai suoi allievi, 259–293. Urbino: QuattroVenti.) could then represent the parts of the treatise that dealt with Latin grammar. Later grammarians used the two parts for different purposes (Dionisotti 1988Dionisotti, A. Carlota 1988 “Greek grammars and dictionaries in Carolingian Europe”. The Sacred Nectar of the Greeks: The study of Greek in the West in the Early Middle Ages ed. by Michael W. Herren & Shirley Ann Brown, 1–56. London: University of London King’s College.Dionisotti, A. Carlota 1988 “Greek grammars and dictionaries in Carolingian Europe”. The Sacred Nectar of the Greeks: The study of Greek in the West in the Early Middle Ages ed. by Michael W. Herren & Shirley Ann Brown, 1–56. London: University of London King’s College.: 19–21). The difference between the Latin and the Greek part is one of focus (Dionisotti 1984Dionisotti, A. Carlotta 1984 “Latin Grammar for Greeks and Goths”. Journal of Roman Studies 74. 202–208. Dionisotti, A. Carlotta 1984 “Latin Grammar for Greeks and Goths”. Journal of Roman Studies 741. 202–208. : 206): in the Latin part the emphasis is on forms that deviate from the general rules, accompanied by copious quotations from literary texts, whereas the Greek part concentrates on the rules, most of which are illustrated by made-up examples (see Silva 2018Silva, Hudson Carlos Alves da 2018 Uma proposta pedagógica da língua grega em um tratado gramatical em latim: De differentiis de Erígena. Ph.D. thesis, Universidade de Juiz de Fora.Silva, Hudson Carlos Alves da 2018 Uma proposta pedagógica da língua grega em um tratado gramatical em latim: De differentiis de Erígena. Ph.D. thesis, Universidade de Juiz de Fora.: 94–95). Put together and completed by some shorter excerpts and fragments, the ensemble gives a reasonably clear idea of the treatise as a whole and of Macrobius’ aim in writing it.1313.For the relationship between the different excerpters see De Paolis (1990)De Paolis, Paolo 1990 “Per una nuova edizione critica del De differentiis di Macrobio”. Dicti studiosus: Scritti di filologia offerti a Scevola Mariotti dai suoi allievi, 259–293. Urbino: QuattroVenti.De Paolis, Paolo 1990 “Per una nuova edizione critica del De differentiis di Macrobio”. Dicti studiosus: Scritti di filologia offerti a Scevola Mariotti dai suoi allievi, 259–293. Urbino: QuattroVenti..

Some believe that Johannes Scotus had a far greater impact on the text than just that of an excerpter. Silva (2018)Silva, Hudson Carlos Alves da 2018 Uma proposta pedagógica da língua grega em um tratado gramatical em latim: De differentiis de Erígena. Ph.D. thesis, Universidade de Juiz de Fora.Silva, Hudson Carlos Alves da 2018 Uma proposta pedagógica da língua grega em um tratado gramatical em latim: De differentiis de Erígena. Ph.D. thesis, Universidade de Juiz de Fora. regards Scotus’ version as an autonomous text with, at the most, some parallels with the text of De verboAnonymus, De verbo = Anonymus, De verbo Ed. by Heinrich Keil, GL V, 634–654; ed. by Marina Passalacqua, Tre testi grammaticali bobbiesi ( GL V 555–566; 634–654; IV 207–217 Keil). Rome: Edizioni di Storia e Letteratura 1984.Anonymus, De verbo = Anonymus, De verbo Ed. by Heinrich Keil, GL V1, 634–654; ed. by Marina Passalacqua, Tre testi grammaticali bobbiesi (GL V1 555–566; 634–654; IV1 207–217 Keil). Rome: Edizioni di Storia e Letteratura 1984. , which he believes is closer to Macrobius’ original text. Yet, Dionisotti is certainly right when she concludes (1984Dionisotti, A. Carlotta 1984 “Latin Grammar for Greeks and Goths”. Journal of Roman Studies 74. 202–208. Dionisotti, A. Carlotta 1984 “Latin Grammar for Greeks and Goths”. Journal of Roman Studies 741. 202–208. : 206) that Johannes Scotus reproduced Macrobius’ text fairly faithfully, as he himself declares in an introduction to his excerpt in the codex Parisinus (De diff. 174). He may be critical of the way Macrobius presented his data, but he describes his own role in excerpting (deflorare) as one of basically following the text as he found it, “changing a few things, but mostly using the same phrasing as he did” (aliquando permutatis, saepius eisdem quibus et ille utendo sermonibus). This is confirmed by the many verbatim correspondences between his text and the treatise De verboAnonymus, De verbo = Anonymus, De verbo Ed. by Heinrich Keil, GL V, 634–654; ed. by Marina Passalacqua, Tre testi grammaticali bobbiesi ( GL V 555–566; 634–654; IV 207–217 Keil). Rome: Edizioni di Storia e Letteratura 1984.Anonymus, De verbo = Anonymus, De verbo Ed. by Heinrich Keil, GL V1, 634–654; ed. by Marina Passalacqua, Tre testi grammaticali bobbiesi (GL V1 555–566; 634–654; IV1 207–217 Keil). Rome: Edizioni di Storia e Letteratura 1984. (a list in Silva 2018Silva, Hudson Carlos Alves da 2018 Uma proposta pedagógica da língua grega em um tratado gramatical em latim: De differentiis de Erígena. Ph.D. thesis, Universidade de Juiz de Fora.Silva, Hudson Carlos Alves da 2018 Uma proposta pedagógica da língua grega em um tratado gramatical em latim: De differentiis de Erígena. Ph.D. thesis, Universidade de Juiz de Fora.: 60–64). Dionisotti finds the tone and content of both texts to be that of a “learned enquiry”, and perfectly in line with Macrobius’ general style in his other works, for instance in the Saturnalia, where the protagonists are intellectuals who were certainly competent in both languages. In this respect, there is a fundamental difference between Macrobius’ treatise and the Artes written by Latin grammarians like Charisius and Diomedes, whose readers were not likely to have an intimate knowledge of Latin (Gatto 2021Gatto, Biagio 2021 Priscian: A syntactic interpretation of a Graeco-Roman World. Ph.D. thesis, University College London.Gatto, Biagio 2021 Priscian: A syntactic interpretation of a Graeco-Roman World. Ph.D. thesis, University College London.: 55).

We should keep in mind here that Johannes Scotus’s search for information about Greek grammar was not entirely compatible with the unsystematic set-up of the work.1414.For this negative judgment see De Paolis’ comment (1990: xvii): “che la struttura del De diff. è assolutamente asistematica e quindi del tutto inadatta a un giovane che comincia a studiare il greco.” And elsewhere (1990: xxviii) he says that “ciò ha fatto parlare di originalità macrobiana anche se essa va intesa più nel senso di originalità di intento che di risultati.” He complained about the jumbled structure of the text, characterizing Macrobius’ style as one “in which in a varying and confused way he mixed the rules of the verbs” (in quo varie confuseque verborum regulas miscuerit, De diff. 174.3; Dionisotti 1984Dionisotti, A. Carlotta 1984 “Latin Grammar for Greeks and Goths”. Journal of Roman Studies 74. 202–208. Dionisotti, A. Carlotta 1984 “Latin Grammar for Greeks and Goths”. Journal of Roman Studies 741. 202–208. : 206). De Paolis (1990De Paolis, Paolo 1990 “Per una nuova edizione critica del De differentiis di Macrobio”. Dicti studiosus: Scritti di filologia offerti a Scevola Mariotti dai suoi allievi, 259–293. Urbino: QuattroVenti.De Paolis, Paolo 1990 “Per una nuova edizione critica del De differentiis di Macrobio”. Dicti studiosus: Scritti di filologia offerti a Scevola Mariotti dai suoi allievi, 259–293. Urbino: QuattroVenti.: xxx) believes that the original purpose of the work was scholarly rather than didactic.

That Macrobius did not intend to write a learner’s grammar of Greek is obvious from the way he skips the treatment of a number of categories that Greek possesses, while Latin does not, for instance and most flagrantly, the category of the aorist, which is simply left out. Macrobius mentions the aorist three times (De diff. 79.15, 17; 83.2; 111–113), always concerning its use in Greek grammar, but he points out that there is no aorist in Latin (Latinitas nescit),1515.He does list aorist forms in some places in the treatise, no doubt because the Greek canons he consulted mentioned these forms, see for instance De diff. 79.15, 17; 83.2. and explains why he therefore does not consider it appropriate to discuss it: “After the pluperfect it would have been logical to treat the infinite tense, i.e., the aorist, but the reason we skip it is that Latin lacks it” (post plusquamperfectum consequens erat ut de infinito tempore, id est περὶ ἀορίστου tractaremus; sed ideo praetermittimus, quia eo Latinitas caret, De diff. 57.14–17). Likewise, he does not treat the second Greek future (μετ’ ὀλίγον μέλλων) πεποιήσομαι, “which signifies something that will happen soon, not after a long time” (quod rem significat non multo post sed mox futurum, De diff. 73), since it is a species propria Graecorum.1616.On this form in Greek see Wouters (1994)Wouters, Alfons 1994 “The ancient Greek (and Latin) grammarians and the μετ’ ὀλίγον μέλλων tense”. Florilegium historiographiae linguisticae: Études d’historiographie de la linguistique et de grammaire comparée à la mémoire de M. Leroy ed. by Jan De Clercq and Piet Desmet, 97–129. Leuven: Peeters.Wouters, Alfons 1994 “The ancient Greek (and Latin) grammarians and the μετ’ ὀλίγον μέλλων tense”. Florilegium historiographiae linguisticae: Études d’historiographie de la linguistique et de grammaire comparée à la mémoire de M. Leroy ed. by Jan De Clercq and Piet Desmet, 97–129. Leuven: Peeters..

3.Macrobius’ sources

About the sources for Macrobius’ treatise, it is difficult to state anything with certainty. De Paolis (De diff. p. xxvii) concludes that Macrobius must have consulted Apollonius Dyscolus’ (second century CE) Rhematikon, fragments of which are preserved in Choeroboscus’ (ninth century CE) Scholia. Actually, the parallel passages in Choeroboscus’ Scholia Choeroboscus, Scholia I = Georgius Choeroboscus, Prolegomena et scholia in Theodosii Alexandrini Canones isagogicas de flexione nominum Ed. by Alfred Hilgard, GG IV/1, 103–417. (Repr., Hildesheim: G. Olms 2007.)Choeroboscus, Scholia I = Georgius Choeroboscus, Prolegomena et scholia in Theodosii Alexandrini Canones isagogicas de flexione nominum Ed. by Alfred Hilgard, GG IV/11, 103–417. (Repr., Hildesheim: G. Olms 2007.) indicated by De Paolis in the footnotes to his edition are not really textual parallels, but passages dealing with more or less the same topics as Macrobius’ treatise.

Macrobius’ representation of the rules and his listing of Greek forms resembles that of the Greek canons, which, as Bacigalupo (2025Bacigalupo, Valeria 2025 “La flessione verbale nella tradizione greca: Un sistema imperfetto?”. Rhesis. International Journal of Linguistics, Philology, and Literature, special issue Dall’Anomia alla Norma, 150–163.Bacigalupo, Valeria 2025 “La flessione verbale nella tradizione greca: Un sistema imperfetto?”. Rhesis. International Journal of Linguistics, Philology, and Literature, special issue Dall’Anomia alla Norma, 150–163.: 156, 160) observes, present a set of rules generating all theoretically possible forms, rather than representing the language as it was actually used. Several times Macrobius refers to principles of inherent rationality in morphology, for instance rationis necessitas (De diff. 105.7) and multiplex ratio (De diff. 107.22). He is careful to point out that apparent exceptions to the rule (regula) do not invalidate it, as in the case of the ‘irregular’ perfect οἶδα (De diff. 51.5–6),1717.A term used in this context is imperium to refer to the scope of a rule (De diff. 99.19). Macrobius emphasizes that the one exception of οἶδα does not invalidate the rule (vitiis obsessus non nocebit). Here, vitia probably corresponds to the Greek term πάθη, changes affecting the form of the word (see Siebenborn 1976Siebenborn, Elmar 1976 Die Lehre von der Sprachrichtigkeit und ihren Kriterien: Studien zur antiken normativen Grammatik. Amsterdam: B. R. Grüner.Siebenborn, Elmar 1976 Die Lehre von der Sprachrichtigkeit und ihren Kriterien: Studien zur antiken normativen Grammatik. Amsterdam: B. R. Grüner.: 51f., 150f.). but he assures the reader that the generalis regulae necessitas makes it easier to understand the linguistic facts (De diff. 95.24).

In some instances Macrobius’ dependence on a Greek source is manifest, for instance when he discusses the morphology of the verb ῥαίνω ‘to sprinkle’ (De diff. 133.9), whose active perfect infinitive is ἐρραγκέναι while the passive infinitive is ἐρράνθαι. Macrobius remarks that this infinitive shows that underlyingly (δυνάμει) γ is ν. A similar remark about the same verb is made by Choeroboscus ( Scholia Choeroboscus, Scholia I = Georgius Choeroboscus, Prolegomena et scholia in Theodosii Alexandrini Canones isagogicas de flexione nominum Ed. by Alfred Hilgard, GG IV/1, 103–417. (Repr., Hildesheim: G. Olms 2007.)Choeroboscus, Scholia I = Georgius Choeroboscus, Prolegomena et scholia in Theodosii Alexandrini Canones isagogicas de flexione nominum Ed. by Alfred Hilgard, GG IV/11, 103–417. (Repr., Hildesheim: G. Olms 2007.) II, 186.8), including the term δυνάμει, which is frequently used by Choe­roboscus’ main source Apollonius Dyscolus to refer to both syntactic and morphological underlying forms.1818.About the use of this term to designate an underlying form, see my remark in Versteegh (2000Versteegh, Kees 2000 “Review of: Jean Lallot, Apollonius Dyscole, De la construction (syntaxe) ”. 2 vols. Paris: J. Vrin 1997 Word 51. 304–308.Versteegh, Kees 2000 “Review of: Jean Lallot, Apollonius Dyscole, De la construction (syntaxe)”. 21 vols. Paris: J. Vrin 1997 Word 511. 304–308.: 307). Likewise, the discussion about the nature of the infinitive, dealing with the question of whether or not it is a verbal form (De diff. 123–125), has a parallel in Choeroboscus ( Scholia Choeroboscus, Scholia I = Georgius Choeroboscus, Prolegomena et scholia in Theodosii Alexandrini Canones isagogicas de flexione nominum Ed. by Alfred Hilgard, GG IV/1, 103–417. (Repr., Hildesheim: G. Olms 2007.)Choeroboscus, Scholia I = Georgius Choeroboscus, Prolegomena et scholia in Theodosii Alexandrini Canones isagogicas de flexione nominum Ed. by Alfred Hilgard, GG IV/11, 103–417. (Repr., Hildesheim: G. Olms 2007.) II, 6.28–7.3), which presumably also goes back to Apollonius Dyscolus.

Sometimes, the discussion of a grammatical point is represented incompletely in Latin grammatical texts, but transmitted by Macrobius in full, for instance the debate about the reason why there is no first person imperative. A logical argument states that one cannot order oneself to do something, which is the argument taken over by the Latin grammarians (e.g., Diomedes, Ars Diomedes, Ars = Diomedes, Artis grammaticae libri III Ed. by Heinrich Keil, GL I, 297–529. (Repr., Hildesheim: G. Olms 2007.)Diomedes, Ars = Diomedes, Artis grammaticae libri III Ed. by Heinrich Keil, GL I1, 297–529. (Repr., Hildesheim: G. Olms 2007.) 338). But there is a morphological argument as well: Macrobius ( De diff Macrobius, De diff. = Macrobii Theodosii, De verborum graeci et latini differentiis vel societatibus excerpta Ed. by Paolo De Paolis. Urbino: Quattro Venti 1990.Macrobius, De diff. = Macrobii Theodosii, De verborum graeci et latini differentiis vel societatibus excerpta Ed. by Paolo De Paolis. Urbino: Quattro Venti 1990.. 105–107) explains that a form like λέγωμεν cannot be an imperative as evidenced by the second person plural of the imperative λέγετε ‘say [pl.]!’, which demonstrates that an imperative must have the same vowel as the indicative. Therefore, λέγωμεν must be an exhortative, ‘let us say’. This discussion is found in Choeroboscus ( Scholia Choeroboscus, Scholia I = Georgius Choeroboscus, Prolegomena et scholia in Theodosii Alexandrini Canones isagogicas de flexione nominum Ed. by Alfred Hilgard, GG IV/1, 103–417. (Repr., Hildesheim: G. Olms 2007.)Choeroboscus, Scholia I = Georgius Choeroboscus, Prolegomena et scholia in Theodosii Alexandrini Canones isagogicas de flexione nominum Ed. by Alfred Hilgard, GG IV/11, 103–417. (Repr., Hildesheim: G. Olms 2007.) II, 232) and has a clear parallel in Apollonius Dyscolus (Synt. II, 363–364).

It goes beyond the scope of the present article to compare all verbal paradigms used by Macrobius with those used by Choeroboscus and other canonical writers. Scappaticcio (2015Scappaticcio, Maria Chiara 2015 Artes Grammaticae in frammenti: I testi grammaticali latini e bilingui greco-latini su papiro. Edizione commentata. Berlin and Boston: de Gruyter. Scappaticcio, Maria Chiara 2015 Artes Grammaticae in frammenti: I testi grammaticali latini e bilingui greco-latini su papiro. Edizione commentata. Berlin and Boston: de Gruyter. : 364) mentions the frequent use of γράφω ‘to write’ and of ἕλκω ‘to draw’ in Greek grammatical papyri, both of them used by Macrobius in several examples. The second verb is also used by Choeroboscus and its popularity can probably be traced back to Apollonius’ Rhematikon.1919.Scappaticcio (2015Scappaticcio, Maria Chiara 2015 Artes Grammaticae in frammenti: I testi grammaticali latini e bilingui greco-latini su papiro. Edizione commentata. Berlin and Boston: de Gruyter. Scappaticcio, Maria Chiara 2015 Artes Grammaticae in frammenti: I testi grammaticali latini e bilingui greco-latini su papiro. Edizione commentata. Berlin and Boston: de Gruyter. : 276; 2016Scappaticcio, Maria Chiara 2016 “Tra Declinationes verborum ed Hermeneuta: Le flessioni verbali greco-latine su papiro”. The Latin of the Grammarians: Reflections about language in the Roman world ed. by Rolando Ferri & Anna Zago, 181–204. Turnhout: Brepols.Scappaticcio, Maria Chiara 2016 “Tra Declinationes verborum ed Hermeneuta: Le flessioni verbali greco-latine su papiro”. The Latin of the Grammarians: Reflections about language in the Roman world ed. by Rolando Ferri & Anna Zago, 181–204. Turnhout: Brepols.); Macrobius uses this verb De diff. 39.13–16, 63.1–3, 97.1–5, 109.22–111.3, 133.24–25.

In one case it is not clear where Macrobius found his examples. When he discusses the rules of the reduplication of the first syllable in the Greek perfect, in particular the replacement of the aspirated consonants (δασέα σύμφωνα) by the corresponding non-aspirated ones (ἀντίστοιχα ψιλά), he cites as examples θαρρῶ ‘to be of good courage’, χρίω ‘to anoint’, and φονεύω ‘to murder’ (De diff. 51). The scholia on Dionysius ThraxDionysius Thrax, Techne = Dionysius Thrax, Ars grammatica Ed. by Gustav Uhlig, GG I/1, 1–101. Leipzig: B.G. Teubner 1901 (Repr., Hildesheim: G. Olms 1965.)Dionysius Thrax, Techne = Dionysius Thrax, Ars grammatica Ed. by Gustav Uhlig, GG I/11, 1–101. Leipzig: B.G. Teubner 1901 (Repr., Hildesheim: G. Olms 1965.) (339.25–29) mention this rule and motivate it by saying that pronouncing two aspirated consonants would slow down speech (ἵνα μὴ δύο δασέων ἐκφωνουμένων τραχυτέρα γένηται ἡ φωνή), even giving an example of how this would sound, *φεφίληκα. As the main motive for this rule Choeroboscus ( Scholia Choeroboscus, Scholia I = Georgius Choeroboscus, Prolegomena et scholia in Theodosii Alexandrini Canones isagogicas de flexione nominum Ed. by Alfred Hilgard, GG IV/1, 103–417. (Repr., Hildesheim: G. Olms 2007.)Choeroboscus, Scholia I = Georgius Choeroboscus, Prolegomena et scholia in Theodosii Alexandrini Canones isagogicas de flexione nominum Ed. by Alfred Hilgard, GG IV/11, 103–417. (Repr., Hildesheim: G. Olms 2007.) II, 75.29–30) adduces the κακοφωνία that would result from pronouncing two aspirated consonants. But his examples, θύω ‘to sacrifice’, φιλῶ ‘to love’, are different from those of Macrobius.

Macrobius must have been familiar with a number of Latin grammatical texts. He probably used a grammatical treatise from the so-called Charisius group,2020.This is how Barwick (1922)Barwick, Karl 1922 Remmius Palaemon und die römische Ars grammatica. Wiesbaden: Dieterich’sche Verlagsbuchhandlung. (Repr., Hildesheim: G. Olms 1967).Barwick, Karl 1922 Remmius Palaemon und die römische Ars grammatica. Wiesbaden: Dieterich’sche Verlagsbuchhandlung. (Repr., Hildesheim: G. Olms 1967). calls a group of grammatical treatises that all belong to the type of Schulgrammatik as against the type of regulae treatise (Law 1986Law, Vivien 1986 “Late Latin grammars in the Early Middle Ages: A typological history”. Historiographia Linguistica 13. 365–380. Law, Vivien 1986 “Late Latin grammars in the Early Middle Ages: A typological history”. Historiographia Linguistica 131. 365–380. ). The texts in the Charisius group have in common that they use a considerable amount of Greek material, which is likely to have served as a tool in instructing Greek-speaking students learning Latin. For the use of Greek in Charisius see Stoppie (2005)Stoppie, Karen 2005 “The role of the Greek in Charisius’ Ars grammatica ”. Rocniki Humanistyczne 53:3. 123–139.Stoppie, Karen 2005 “The role of the Greek in Charisius’ Ars grammatica”. Rocniki Humanistyczne 53:3. 123–139.; for Greek in Diomedes’ Ars Diomedes, Ars = Diomedes, Artis grammaticae libri III Ed. by Heinrich Keil, GL I, 297–529. (Repr., Hildesheim: G. Olms 2007.)Diomedes, Ars = Diomedes, Artis grammaticae libri III Ed. by Heinrich Keil, GL I1, 297–529. (Repr., Hildesheim: G. Olms 2007.) see Rochette (2023)Rochette, Bruno 2023 “Use of Greek in Diomedes’ Ars grammatica ”. Roman Perspectives on Linguistic Diversity: Guardians of a changing language ed. by Adam Gitner, 73–98. Oxford: Oxford University Press. Rochette, Bruno 2023 “Use of Greek in Diomedes’ Ars grammatica”. Roman Perspectives on Linguistic Diversity: Guardians of a changing language ed. by Adam Gitner, 73–98. Oxford: Oxford University Press. . which includes the artes of Charisius (fourth century), Diomedes (second half fourth century; see Dammer 2001Dammer, Raphael 2001 Diomedes grammaticus. Trier: Wissenschaftlicher Verlag.Dammer, Raphael 2001 Diomedes grammaticus. Trier: Wissenschaftlicher Verlag.: 21), and Dositheus (fourth century). Unlike most Latin grammarians, Macrobius was interested not only in Latin morphology, but also in the way the rules work for Greek grammar, which he most likely gathered from reading Apollonius’ text. A terminological link between Macrobius and Diomedes appears in Macrobius’ treatment of one of the species verborum, the species usurpativa (De diff. 157–159). This term refers to various forms of the gerund and the supine, which Macrobius regards as a special ornament of the Latin language since it does not exist in Greek. All loci for this term mentioned by Schad (2007Schad, Samantha 2007 A Lexicon of Latin Grammatical Terminology. Rome & Pisa: Fabrizio Serra.Schad, Samantha 2007 A Lexicon of Latin Grammatical Terminology. Rome & Pisa: Fabrizio Serra.: 429) are from Diomedes (Ars Diomedes, Ars = Diomedes, Artis grammaticae libri III Ed. by Heinrich Keil, GL I, 297–529. (Repr., Hildesheim: G. Olms 2007.)Diomedes, Ars = Diomedes, Artis grammaticae libri III Ed. by Heinrich Keil, GL I1, 297–529. (Repr., Hildesheim: G. Olms 2007.) 396) and Macrobius.2121.On the passage about the species verborum in Diomedes see Baratin (1989Baratin, Marc 1989 La naissance de la syntaxe à Rome. Paris: Éditions de Minuit.Baratin, Marc 1989 La naissance de la syntaxe à Rome. Paris: Éditions de Minuit.: 144–146). He explains usurpativa as ‘catachrestic’, i.e., ‘inappropriate’, which is one of the meanings of usurpatio: forms like the gerund can only be called verbal forms catachrestically.

The influence of Aulus Gellius’ (d. after 180 CE) Noctes Atticae on Macrobius is commonly acknowledged, although Macrobius does not mention him by name (Holford-Strevens 2003Holford-Strevens, Leofranc 2003 Aulus Gellius: An Antonine scholar and his achievement. Revised edition Oxford: Oxford University Press. Holford-Strevens, Leofranc 2003 Aulus Gellius: An Antonine scholar and his achievement. Revised edition Oxford: Oxford University Press. : 170). Grammar occupied an important place in the Noctes Atticae with a focus on etymology and the explanation of deviant word forms. Macrobius’ Saturnalia Macrobius, Saturnalia = Macrobii Ambrosii Theodosii Saturnalia Ed. by Robert A. Kaster. Oxford: Oxford University Press 2011.Macrobius, Saturnalia = Macrobii Ambrosii Theodosii Saturnalia Ed. by Robert A. Kaster. Oxford: Oxford University Press 2011. illustrate Gellius’ influence through the many etymological digressions, a typical example being Servius’ learned and lengthy exposé about the different kinds of nuts, with a special focus on the origin of their names (Saturnalia 3,18.2–13).2222.On the use of etymology as a means to explore the hidden meanings of Virgil’s text in Macrobius’ Saturnalia Macrobius, Saturnalia = Macrobii Ambrosii Theodosii Saturnalia Ed. by Robert A. Kaster. Oxford: Oxford University Press 2011.Macrobius, Saturnalia = Macrobii Ambrosii Theodosii Saturnalia Ed. by Robert A. Kaster. Oxford: Oxford University Press 2011. see Amsler (1989Amsler, Mark 1989 Etymology and grammatical discourse in late Antiquity and the early Middle Ages. Amsterdam & Philadelphia: John Benjamins. Amsler, Mark 1989 Etymology and grammatical discourse in late Antiquity and the early Middle Ages. Amsterdam & Philadelphia: John Benjamins. : 123–128). Like Gellius, Macrobius does not hide his loathing of common grammarians and schoolmasters (Cavazza 1987Cavazza, Franco 1987 “Gellio grammatico e i suoi rapporti con l’Ars grammatica romana”. The History of Linguistics in the Classical period ed. by Daniel J. Taylor, 85–105. Amsterdam & Philadelphia: John Benjamins. Cavazza, Franco 1987 “Gellio grammatico e i suoi rapporti con l’Ars grammatica romana”. The History of Linguistics in the Classical period ed. by Daniel J. Taylor, 85–105. Amsterdam & Philadelphia: John Benjamins. ). But Gellius did not formulate morphological rules, in the way the author of the De differentiis did.

4.Comparing the Greek and the Latin verbal system

Just like grammarians in other grammatical traditions, Greek and Latin grammarians interpreted verbal tenses as chronological markers. While verbal morphology in the languages of the world rarely follows chronological parameters, there seems to be a natural tendency on the part of both native speakers and professional grammarians to define verbal tenses on a timeline (Versteegh 2013Versteegh, Kees 2013 “A Perfect Mess: The distinction of tense and aspect in grammatical traditions”. Comparing ancient grammars: The Greek, Syriac and Arabic traditions ed. by Sara Eco Conti & Margherita Farina, 27–64. Pisa: Edizioni della Scuola Normale Superiore.Versteegh, Kees 2013 “A Perfect Mess: The distinction of tense and aspect in grammatical traditions”. Comparing ancient grammars: The Greek, Syriac and Arabic traditions ed. by Sara Eco Conti & Margherita Farina, 27–64. Pisa: Edizioni della Scuola Normale Superiore.). At first, Latin grammarians defined the meaning, structure, and use of the Latin verb within the framework of Greek grammatical doctrine. The standard classification of the Greek verbal tenses, as it is found in the scholia on Dionysius Thrax’ Techne Dionysius Thrax, Techne = Dionysius Thrax, Ars grammatica Ed. by Gustav Uhlig, GG I/1, 1–101. Leipzig: B.G. Teubner 1901 (Repr., Hildesheim: G. Olms 1965.)Dionysius Thrax, Techne = Dionysius Thrax, Ars grammatica Ed. by Gustav Uhlig, GG I/11, 1–101. Leipzig: B.G. Teubner 1901 (Repr., Hildesheim: G. Olms 1965.) (249, 27–32), distinguishes three tenses, past, present and future. The present and the future are single tenses, while the past is subdivided into three ‘differences’ (διαφοραί): imperfect (παρῳχημένος), perfect (παρακείμενος) and pluperfect (ὑπερσυντελικός), to which the aorist is added as an infinite tense (see Mastriani 2022Mastriani, Maria Camilla 2022 Lo studio del verbo nella tradizione grammaticale latina: Per un corpus comparativo. Ph.D. diss., Università degli Studi di Napoli.Mastriani, Maria Camilla 2022 Lo studio del verbo nella tradizione grammaticale latina: Per un corpus comparativo. Ph.D. diss., Università degli Studi di Napoli.: 272). According to the scholiasts, it is appropriate for the past to be subdivided because, unlike the present and the future, the events of the past are known so that it is possible to differentiate between them, whereas the exact course of events in the present and the future is still unknown, so that these tenses cannot be subdivided.

An early classification of the verbal tenses in Latin grammatical literature is Varro’s distinction of perfecta and infecta ( De L.L Varro, De L.L. = Marcus Terentius Varro, De lingua latina Ed. and transl. by Roland G. Kent. 3rd ed. 2 vols. Cambridge, Mass.: Harvard University Press and London: W. Heinemann 1958.Varro, De L.L. = Marcus Terentius Varro, De lingua latina Ed. and transl. by Roland G. Kent. 3rd ed. 21 vols. Cambridge, Mass.: Harvard University Press and London: W. Heinemann 1958.. X, 47–48). Within the category of infecta Varro cites three verbal forms legebam, lego, legam as examples of praeteritum, praesens, futurum; in the category of perfecta the corresponding forms are legeram, legi, legero. Note that Varro’s classification is linked with the analogy/anomaly debate, specifically with the rules for setting up a valid analogy.2323.Varro’s notion of analogy is analyzed by Taylor (1996Taylor, Daniel J. 1996 Varro — De lingua latina X. Amsterdam & Philadelphia: John Benjamins. Taylor, Daniel J. 1996 Varro — De lingua latina X. Amsterdam & Philadelphia: John Benjamins. : 18–30). In order for an analogy to be valid, the forms compared must belong to the same category; thus, it is not permitted to compare the forms legebam and legi because they belong to different categories and cannot be cited as evidence of anomaly in speech. Varro is not making a statement about the present relevance of the perfectum: he is simply comparing the forms morphologically and there is no doubt that he saw the function of the Latin perfect as that of a past tense (Serbat 1976Serbat, Guy 1976 “Les temps du verbe en latin. III. Le parfait de l’indicatif actif”. Revue des Etudes Latines 54. 308–352.Serbat, Guy 1976 “Les temps du verbe en latin. III. Le parfait de l’indicatif actif”. Revue des Etudes Latines 541. 308–352.: 314–315).

The overwhelming majority of the Latin grammarians distinguished three verbal tenses in Latin, corresponding to what they saw as the ‘natural’ order of time: praesens (or instans), praeteritum, futurum.2424.A full discussion of the grammarians’ treatment of the verbal tenses may be found in Mastriani (2022Mastriani, Maria Camilla 2022 Lo studio del verbo nella tradizione grammaticale latina: Per un corpus comparativo. Ph.D. diss., Università degli Studi di Napoli.Mastriani, Maria Camilla 2022 Lo studio del verbo nella tradizione grammaticale latina: Per un corpus comparativo. Ph.D. diss., Università degli Studi di Napoli.: 259–289). The past tense (praeteritum) was subdivided into three tenses: imperfectum (or minusquamperfectum, see below), perfectum, plusquamperfectum.2525.Charisius (Ars 214.14) is the only grammarian to mention four differentiae of the past tense, the fourth being that of the recordativae, e.g. legerim, which the other grammarians regarded as a subjunctive. Macrobius explains the form legerim as perfect subjunctive, and legero as perfect future, while the rest of their conjugation is the same (nec alia discretio est inter perfectum ac futurum; De verboAnonymus, De verbo = Anonymus, De verbo Ed. by Heinrich Keil, GL V, 634–654; ed. by Marina Passalacqua, Tre testi grammaticali bobbiesi ( GL V 555–566; 634–654; IV 207–217 Keil). Rome: Edizioni di Storia e Letteratura 1984.Anonymus, De verbo = Anonymus, De verbo Ed. by Heinrich Keil, GL V1, 634–654; ed. by Marina Passalacqua, Tre testi grammaticali bobbiesi (GL V1 555–566; 634–654; IV1 207–217 Keil). Rome: Edizioni di Storia e Letteratura 1984. 36 Passalacqua). According to Carchereux-Marchand (2010Carchereux-Marchand, Diane 2010 “L’accident verbal du tempus chez les grammairiens latins: Définition et description”. Relations spatio-temporelles en latin (= Revue de linguistique latine du centre Alfred Ernout “De lingua latina”, 5) ed. by Claire Blanche-Benveniste & Michèle Fruyt, 1–13. Paris: Presses Universitaires de Paris-Sorbonne.Carchereux-Marchand, Diane 2010 “L’accident verbal du tempus chez les grammairiens latins: Définition et description”. Relations spatio-temporelles en latin (= Revue de linguistique latine du centre Alfred Ernout “De lingua latina”, 5) ed. by Claire Blanche-Benveniste & Michèle Fruyt, 1–13. Paris: Presses Universitaires de Paris-Sorbonne.: 7) Charisius’ introduction of a fourth subdivision may have been inspired by the position of the aorist in Greek grammar as the infinite tense of the past.

There was one category in the Latin verbal system for which there was no simple correspondence with a Greek verbal form, namely that of the perfect. In Greek, verbal forms like λέλυκα refer to an event that took place in the past, while having a morphologically present form. In Stoic terminology, the perfect is accordingly called ἐνεστὼς συντελικός, i.e., a form that refers to the past, but has present relevance; συντελικός is the opposite of παρατατικός, within a framework distinguishing between perfect and infect tenses.2626.For the most recent discussion of the Stoic classification of the verbal tenses see Manetti (2022)Manetti, Giovanni 2022 “The debate on the question of ‘tense’ and ‘aspect’ in the Stoics’ linguistic theory”. Conceptions of time in Greek and Roman Antiquity ed. by Richard Faure, Simon-Pierre Valli & Arnaud Zucker, 21–48. Berlin & Boston: De Gruyter. Manetti, Giovanni 2022 “The debate on the question of ‘tense’ and ‘aspect’ in the Stoics’ linguistic theory”. Conceptions of time in Greek and Roman Antiquity ed. by Richard Faure, Simon-Pierre Valli & Arnaud Zucker, 21–48. Berlin & Boston: De Gruyter. . Later Greek grammarians used παρακείμενος for the perfect tense because of its alleged proximity to the present tense.2727.One of the anonymous reviewers points out that in later Greek perfect and aorist have merged, with a reference to Horrocks (1997)Horrocks, Geoffrey 1997 Greek: A history of the language and its speakers. London & New York: Longman.Horrocks, Geoffrey 1997 Greek: A history of the language and its speakers. London & New York: Longman.. I do not believe that this change is relevant for the linguistic analysis by the grammarians: their corpus consisted of Classical Greek texts and they strove to preserve this system intact. The accents, for instance, had disappeared as markers of pitch in Late Greek, but the grammarians persisted in formulating the rules for their use (Horrocks 1997Horrocks, Geoffrey 1997 Greek: A history of the language and its speakers. London & New York: Longman.Horrocks, Geoffrey 1997 Greek: A history of the language and its speakers. London & New York: Longman.: 67f.). Most of the examples of the merger of perfect and aorist in Horrocks (e.g. 77f., 118f., 247f.) concern the development of the spoken language, sometimes reflected in non-literary texts, which were ignored in the grammatical treatises.

In Latin grammar perfectum translates Greek συντελικός, but this created considerable confusion because the Latin perfectum did not share the feature of present relevance of the Greek form. Apart from some exceptional forms like odi ‘I hate’, memini ‘I remember’, and novi ‘I know’ (cf. Serbat 1976Serbat, Guy 1976 “Les temps du verbe en latin. III. Le parfait de l’indicatif actif”. Revue des Etudes Latines 54. 308–352.Serbat, Guy 1976 “Les temps du verbe en latin. III. Le parfait de l’indicatif actif”. Revue des Etudes Latines 541. 308–352.: 348–349), the function of the Latin perfect resembles that of the Greek ἀόριστος. The term aoristus itself was seldom used by the Latin grammarians (Mellet 2000Mellet, Sylvie 2000 “Le parfait latin, un ‘praeteritum perfectum’”. Cahiers Chronos 6. 95–106. Mellet, Sylvie 2000 “Le parfait latin, un ‘praeteritum perfectum’”. Cahiers Chronos 61. 95–106. : 97), but the equivalence of the Greek ἀόριστος and the Latin perfect is remarked upon by Diomedes, who in addition claims a perfect meaning for the Latin form: “Our perfect corresponds to [their] ἀόριστος and παρακείμενος” (perfectum apud nos pro ἀορίστῳ καὶ παρακειμένῳ valet; Ars I, 336.10). Apart from this passage in Diomedes, Schad (2007Schad, Samantha 2007 A Lexicon of Latin Grammatical Terminology. Rome & Pisa: Fabrizio Serra.Schad, Samantha 2007 A Lexicon of Latin Grammatical Terminology. Rome & Pisa: Fabrizio Serra.: 35–36) mentions four occurrences of the term aoristus in Priscian (sixth century), who uses as the Latin equivalent infinitus (Institutiones II, 417.1, 445.19)2828. Infinitum tempus is already used by Macrobius ( De diff Macrobius, De diff. = Macrobii Theodosii, De verborum graeci et latini differentiis vel societatibus excerpta Ed. by Paolo De Paolis. Urbino: Quattro Venti 1990.Macrobius, De diff. = Macrobii Theodosii, De verborum graeci et latini differentiis vel societatibus excerpta Ed. by Paolo De Paolis. Urbino: Quattro Venti 1990.. 57.15), see below. and adds that the Latin praeteritum perfectum has the force (vis) of the Greek aorist (Institutiones II, 415.26, 416.15).

In connection with the interpretation of the perfect as being simultaneously past and present, De Martino (2000)De Martino, Marcello 2000 “On the semantic interpretation of passive imperative sentences: A case of ‘voluntary’ interference between Ancient Greek and Latin in Latin grammarians”. Syntaxis: An International Journal of Syntactic Research 3. 111–143.De Martino, Marcello 2000 “On the semantic interpretation of passive imperative sentences: A case of ‘voluntary’ interference between Ancient Greek and Latin in Latin grammarians”. Syntaxis: An International Journal of Syntactic Research 31. 111–143. treats the problematic meaning of an imperativus perfecti in Greek. It is phrased as a command to someone to perform an act, but according to Apollonius Dyscolus an expression like κεκλείσθω ἡ θύρα should be interpreted as a wish for the closing of the door to be completed (De Martino 2000De Martino, Marcello 2000 “On the semantic interpretation of passive imperative sentences: A case of ‘voluntary’ interference between Ancient Greek and Latin in Latin grammarians”. Syntaxis: An International Journal of Syntactic Research 3. 111–143.De Martino, Marcello 2000 “On the semantic interpretation of passive imperative sentences: A case of ‘voluntary’ interference between Ancient Greek and Latin in Latin grammarians”. Syntaxis: An International Journal of Syntactic Research 31. 111–143.: 129; cf. Conti 2009Conti, Sara Eco 2009 “Reflections on the verb in Apollonius Dyscolus”. Quaderni del Laboratorio di Linguistica 8. 1–22.Conti, Sara Eco 2009 “Reflections on the verb in Apollonius Dyscolus”. Quaderni del Laboratorio di Linguistica 81. 1–22.: 7–8). According to De Martino, Priscian misinterpreted Apollonius’ explanation, because he viewed the Latin periphrastic form doctus sit as doceatur within a framework relying on temporal rather than aspectual meaning. Macrobius interpreted the perfect imperative as meaning impero, ut claudendi officium iam sit peractum ‘I order the task of closing to have been performed already’ (De diff. 111.13), which in Latin can be expressed by ostium clausum sit (De Martino 2000De Martino, Marcello 2000 “On the semantic interpretation of passive imperative sentences: A case of ‘voluntary’ interference between Ancient Greek and Latin in Latin grammarians”. Syntaxis: An International Journal of Syntactic Research 3. 111–143.De Martino, Marcello 2000 “On the semantic interpretation of passive imperative sentences: A case of ‘voluntary’ interference between Ancient Greek and Latin in Latin grammarians”. Syntaxis: An International Journal of Syntactic Research 31. 111–143.: 130).2929.The same explanation, with different examples, in De verboAnonymus, De verbo = Anonymus, De verbo Ed. by Heinrich Keil, GL V, 634–654; ed. by Marina Passalacqua, Tre testi grammaticali bobbiesi ( GL V 555–566; 634–654; IV 207–217 Keil). Rome: Edizioni di Storia e Letteratura 1984.Anonymus, De verbo = Anonymus, De verbo Ed. by Heinrich Keil, GL V1, 634–654; ed. by Marina Passalacqua, Tre testi grammaticali bobbiesi (GL V1 555–566; 634–654; IV1 207–217 Keil). Rome: Edizioni di Storia e Letteratura 1984. (30.19–24 Passalacqua).

The perfect also plays a role in the discussion about the possibility of making a wish about the past. According to Macrobius ( De diff Macrobius, De diff. = Macrobii Theodosii, De verborum graeci et latini differentiis vel societatibus excerpta Ed. by Paolo De Paolis. Urbino: Quattro Venti 1990.Macrobius, De diff. = Macrobii Theodosii, De verborum graeci et latini differentiis vel societatibus excerpta Ed. by Paolo De Paolis. Urbino: Quattro Venti 1990.. 119–120) in a wish like εἴθε ὁ υἱός μου νενικήκοι ‘Would that my son has won!’ the Greek optative is paralleled in Latin by an expression like utinam vicerim, analyzed by most grammarians as a subjunctive (see above, n. 22). The discussion by Macrobius has a direct parallel in Apollonius Dyscolus (Synt. 355.5–356.5), who confirms the possibility of wishing something about the past (Schöpsdau 1992Schöpsdau, Klaus 1992 “Vergleiche zwischen Lateinisch und Griechisch in der antiken Welt”. Zum Umgang mit fremden Sprachen in der griechisch-römischen Welt ed. by Carl Werner Müller, Carl Werner, Kurt Sier & Jürgen Werner, 115–136. Stuttgart: F. Steiner.Schöpsdau, Klaus 1992 “Vergleiche zwischen Lateinisch und Griechisch in der antiken Welt”. Zum Umgang mit fremden Sprachen in der griechisch-römischen Welt ed. by Carl Werner Müller, Carl Werner, Kurt Sier & Jürgen Werner, 115–136. Stuttgart: F. Steiner.: 124; Conti 2009Conti, Sara Eco 2009 “Reflections on the verb in Apollonius Dyscolus”. Quaderni del Laboratorio di Linguistica 8. 1–22.Conti, Sara Eco 2009 “Reflections on the verb in Apollonius Dyscolus”. Quaderni del Laboratorio di Linguistica 81. 1–22.: 3–7).

In modern studies of the Latin verbal system the present relevance of the perfect remains a contested issue. Mellet (2000)Mellet, Sylvie 2000 “Le parfait latin, un ‘praeteritum perfectum’”. Cahiers Chronos 6. 95–106. Mellet, Sylvie 2000 “Le parfait latin, un ‘praeteritum perfectum’”. Cahiers Chronos 61. 95–106. forcefully denies any such relevance, other than what she calls a “déduction contextuelle”.3030.See Serbat (1976Serbat, Guy 1976 “Les temps du verbe en latin. III. Le parfait de l’indicatif actif”. Revue des Etudes Latines 54. 308–352.Serbat, Guy 1976 “Les temps du verbe en latin. III. Le parfait de l’indicatif actif”. Revue des Etudes Latines 541. 308–352.: 330–340) for extensive criticism of Antoine Meillet’s (1866–1936) theory about the alleged perfective aspect of the Latin perfect; see Mastriani (2022Mastriani, Maria Camilla 2022 Lo studio del verbo nella tradizione grammaticale latina: Per un corpus comparativo. Ph.D. diss., Università degli Studi di Napoli.Mastriani, Maria Camilla 2022 Lo studio del verbo nella tradizione grammaticale latina: Per un corpus comparativo. Ph.D. diss., Università degli Studi di Napoli.: 273, n. 88). In her view, the perfect always refers to events in the past, which only through the pragmatic context may acquire perfect aspect, especially in the passive perfect constructed with the auxiliary verb esse. In this respect, the Latin perfect differs from the Greek aorist, precisely because of its ability to denote a resultant state, which the Greek aorist cannot. Crellin (2020)Crellin, Robert 2020 Perfects in Indo-European Languages and Beyond ed. by Robert Crellin & Thomas Jügel, 549–590. Amsterdam & Philadelphia: John Benjamins. Crellin, Robert 2020 Perfects in Indo-European Languages and Beyond ed. by Robert Crellin & Thomas Jügel, 549–590. Amsterdam & Philadelphia: John Benjamins. reconstructs the Latin perfect as referring to both past tense and perfective aspect. But for him, too, any present relevance, for instance in the persistence of a state, is to be ascribed to the context: when I proclaim ‘I have built you a house’, the house is likely to be still there, but this does not mean that the perfect verb may be assumed to have present relevance.

The Latin grammarians were also faced with a problem when trying to determine the relationship between the perfect and the pluperfect, which they usually define in terms of a timeline. The Greek grammarians had done so with the help of temporal adverbs (ἄρτι, πάλαι). In the Latin grammatical tradition, too, temporal adverbs like nuper, olim, paulo ante, iamdudum were used to describe the difference between past tenses (Serbat 1976Serbat, Guy 1976 “Les temps du verbe en latin. III. Le parfait de l’indicatif actif”. Revue des Etudes Latines 54. 308–352.Serbat, Guy 1976 “Les temps du verbe en latin. III. Le parfait de l’indicatif actif”. Revue des Etudes Latines 541. 308–352.: 317–318). The chronological fixation is particularly clear when the distance in time of the pluperfect is measured from the present, not from the perfect. Referring to Priscian’s definition of the plusquamperfectum (Institutiones II, 406.5–6): “It presents the thing as having been performed earlier” (iam pridem res perfecta ostenditur), i.e., longer ago from the present, Mastriani (2022Mastriani, Maria Camilla 2022 Lo studio del verbo nella tradizione grammaticale latina: Per un corpus comparativo. Ph.D. diss., Università degli Studi di Napoli.Mastriani, Maria Camilla 2022 Lo studio del verbo nella tradizione grammaticale latina: Per un corpus comparativo. Ph.D. diss., Università degli Studi di Napoli.: 275) claims that the grammarians regarded the pluperfect as “ben più lontana”. In this respect, Diomedes’ (Ars Diomedes, Ars = Diomedes, Artis grammaticae libri III Ed. by Heinrich Keil, GL I, 297–529. (Repr., Hildesheim: G. Olms 2007.)Diomedes, Ars = Diomedes, Artis grammaticae libri III Ed. by Heinrich Keil, GL I1, 297–529. (Repr., Hildesheim: G. Olms 2007.) I, 336.5–6) definition was perhaps more insightful since he situates the pluperfect in time before the perfect: “a tense completed earlier ... than something we did” (tempus iam pridem exactum ... quo quid egimus).3131.See also the quotation from the mid-fifth century grammarian CledoniusCledonius, Ars = Cledonii Constantinopolitani grammatici Ars Ed. by Heinrich Keil, GL V, 10–79. (Repr., Hildesheim: G. Olms 2007.)Cledonius, Ars = Cledonii Constantinopolitani grammatici Ars Ed. by Heinrich Keil, GL V1, 10–79. (Repr., Hildesheim: G. Olms 2007.): “What is done now, is perfect; what was done before, is pluperfect” (quod est factum nunc, perfectum est; quod ante factum, plusquamperfectum; Ars 61.2–3); cf. Mastriani (2022Mastriani, Maria Camilla 2022 Lo studio del verbo nella tradizione grammaticale latina: Per un corpus comparativo. Ph.D. diss., Università degli Studi di Napoli.Mastriani, Maria Camilla 2022 Lo studio del verbo nella tradizione grammaticale latina: Per un corpus comparativo. Ph.D. diss., Università degli Studi di Napoli.: 277, n. 109). At any rate, the function of the Latin plusquamperfectum differs markedly from the Greek corresponding form since it does not indicate a state, but an action.3232.Cf. Serbat (1976Serbat, Guy 1976 “Les temps du verbe en latin. III. Le parfait de l’indicatif actif”. Revue des Etudes Latines 54. 308–352.Serbat, Guy 1976 “Les temps du verbe en latin. III. Le parfait de l’indicatif actif”. Revue des Etudes Latines 541. 308–352.: 318); Mastriani (2022Mastriani, Maria Camilla 2022 Lo studio del verbo nella tradizione grammaticale latina: Per un corpus comparativo. Ph.D. diss., Università degli Studi di Napoli.Mastriani, Maria Camilla 2022 Lo studio del verbo nella tradizione grammaticale latina: Per un corpus comparativo. Ph.D. diss., Università degli Studi di Napoli.: 275, n. 102) and Diomedes (Ars Diomedes, Ars = Diomedes, Artis grammaticae libri III Ed. by Heinrich Keil, GL I, 297–529. (Repr., Hildesheim: G. Olms 2007.)Diomedes, Ars = Diomedes, Artis grammaticae libri III Ed. by Heinrich Keil, GL I1, 297–529. (Repr., Hildesheim: G. Olms 2007.) I, 336, 8–10), cf. Mastriani (2022Mastriani, Maria Camilla 2022 Lo studio del verbo nella tradizione grammaticale latina: Per un corpus comparativo. Ph.D. diss., Università degli Studi di Napoli.Mastriani, Maria Camilla 2022 Lo studio del verbo nella tradizione grammaticale latina: Per un corpus comparativo. Ph.D. diss., Università degli Studi di Napoli.: 276, n. 106).

The text of Macrobius’ treatise is not very helpful in determining the status of the perfect because he is mainly interested in the morphological structure. When he explains that the form of the Greek perfect is derived from that of the future, he uses a logical argument: before things can be said to have happened they must first exist in the future “anything done at first was something that had to be done” (omne enim quod factum est prius faciendum fuit; De diff. 47.7). As we have seen above, he discusses the possibility of using an imperative or a wish in the perfect tense, but this, too, is a pragmatic/logical issue rather than a semantic analysis of the meaning of the perfect.

5.Rules and paradigms

Macrobius’ intellectual interest as opposed to any didactic preoccupations is evident in the way he formulates rules for the morphology of the Greek and Latin verb. An example of a rule that is elegantly formulated but can hardly have been helpful for a beginning student is the one about the number of syllables in Latin and Greek active and passive verbs (De diff. 29 + 31). In Latin, active verbs in the indicativus praesentis usually have one syllable less in the second person than in the first person (cf. doceo/doces), while in the passive the number of syllables is the same (cf. doceor/doceris). In Greek, it is the other way round, active verbs having the same number of syllables in the first and second person (cf. λούω/λούεις), but the second person passive having one syllable less than the first person (cf. γράφομαι/γράφῃ). It is hard to see how a rule like this could be of use to second-language learners. Macrobius presents a cross-linguistic rule like this as an interesting piece of information, rather than being of any practical use. Some of the rules mentioned by him and dutifully copied by Johannes Scotus may have served as a pons asinorum in education, for instance when Macrobius explains that from a third person of the indicativus praesentis in Greek one can make an infinitive by adding the letter ν, for instance, ποιεῖ --> ποιεῖν, and by repositioning the accent one can form the imperative, for instance ποιεῖ --> ποίει (at least for this type of verbs). Such rules make a connection between different morphological forms of the verbal system, but the didactic usefulness of a multitude of disparate rules to be memorized by the students is doubtful.

Some rules show the derivation of the personal endings of the verbs, as when Greek passives with a first person ending in -μην are said to change this into ο in the second person: ποιοίμην ποιοίο, while first persons ending in -ην change the ν in σ: δοίην δοίης. To obtain the third person ο is changed into το: ποιοίο ποιοῖτο, for the first class of verbs, and σ is deleted in the second class of verbs: δοίης δοίη (De diff. 123). It is difficult to see how such canon-like rules could have served as didactic tools for students of Greek, who after all would probably find it easier to learn a paradigm by heart than internalizing an endless series of rules, without any coherence or underlying rational motivation.3333.Admittedly, Macrobius himself refers to the memorization of rules: “When this rule has been memorized, we’ll add another one” (hac regula memoriae mandata alteram subiciemus, De diff. 105). In other words, Macrobius’ treatise could not have served as a primer to teach Greek to speakers of Latin, as Silva believes (2018Silva, Hudson Carlos Alves da 2018 Uma proposta pedagógica da língua grega em um tratado gramatical em latim: De differentiis de Erígena. Ph.D. thesis, Universidade de Juiz de Fora.Silva, Hudson Carlos Alves da 2018 Uma proposta pedagógica da língua grega em um tratado gramatical em latim: De differentiis de Erígena. Ph.D. thesis, Universidade de Juiz de Fora.: 71, 84). For this purpose, there were better materials available, even at Macrobius’ time (Racine 2015Racine, Félix 2015 “Servius’ Greek lessons”. Learning Latin and Greek from Antiquity to the Present ed. by Elizabeth P. Archibald, William Brockliss & Jonathan Gnoza, 52–64. Cambridge: Cambridge University Press. Racine, Félix 2015 “Servius’ Greek lessons”. Learning Latin and Greek from Antiquity to the Present ed. by Elizabeth P. Archibald, William Brockliss & Jonathan Gnoza, 52–64. Cambridge: Cambridge University Press. ).

Dickey (2016b)Dickey, Eleanor 2016b Learning Latin the Ancient Way: Latin textbooks from the Ancient world. Cambridge: Cambridge University Press.Dickey, Eleanor 2016b Learning Latin the Ancient Way: Latin textbooks from the Ancient world. Cambridge: Cambridge University Press. explains that some vocabulary lists were at first meant for one language only but were subsequently adapted to both languages.3434.See also Gatto (2021Gatto, Biagio 2021 Priscian: A syntactic interpretation of a Graeco-Roman World. Ph.D. thesis, University College London.Gatto, Biagio 2021 Priscian: A syntactic interpretation of a Graeco-Roman World. Ph.D. thesis, University College London.: 136) about the development of bilingual from monolingual wordlists. Such a development cannot be assumed for Macrobius’ treatise. His primary audience most likely consisted of native speakers of Latin, well-read intellectuals who were curious about the differences and commonalities between the two languages, rather than being in need of instruction in either language. They were certainly competent in Greek, without being native speakers. Macrobius’ use of expressions like illi or apud Graecos when introducing features of Greek grammar makes quite clear from which perspective he is comparing the two languages.

The discussion about the medial diathesis in Macrobius is more extensive than that in the corpus of the Latin grammarians.3535.See Benedetti (2015)Benedetti, Marina 2015 “ De verborum Graeci et Latini differentiis vel societatibus: La diatesi media e il punto di vista latino”. Studi e Saggi Linguistici 53:2. 273–288.Benedetti, Marina 2015 “De verborum Graeci et Latini differentiis vel societatibus: La diatesi media e il punto di vista latino”. Studi e Saggi Linguistici 53:2. 273–288. and Stoppie et al. (2015). In fact, Benedetti (2015Benedetti, Marina 2015 “ De verborum Graeci et Latini differentiis vel societatibus: La diatesi media e il punto di vista latino”. Studi e Saggi Linguistici 53:2. 273–288.Benedetti, Marina 2015 “De verborum Graeci et Latini differentiis vel societatibus: La diatesi media e il punto di vista latino”. Studi e Saggi Linguistici 53:2. 273–288.: 274) claims that Macrobius “fornisce l’unica testimonianza della ricezione del medio in ambito romano”. The problem — even with respect to the treatment by Macrobius — was that the identification of the Greek category of medial (μέσα) verbs with the Latin communia did not work: in the Greek grammatical tradition medial verbs could have both active and passive forms, while according to the Latin definition the communia are passive forms with alternatively passive and active meaning.

The category of the οὐδετέρα verbs mentioned by Macrobius seems to be an innovation with respect to the Greek sources. Apollonius Dyscolus does not use the term, but it is used by Choeroboscus and in the scholia on Dionysius Thrax’ Techne Dionysius Thrax, Techne = Dionysius Thrax, Ars grammatica Ed. by Gustav Uhlig, GG I/1, 1–101. Leipzig: B.G. Teubner 1901 (Repr., Hildesheim: G. Olms 1965.)Dionysius Thrax, Techne = Dionysius Thrax, Ars grammatica Ed. by Gustav Uhlig, GG I/11, 1–101. Leipzig: B.G. Teubner 1901 (Repr., Hildesheim: G. Olms 1965.) (Stoppie et al. 2007Stoppie, Karen, Pierre Swiggers & Alfons Wouters 2007 “La terminologie grammaticale en contexte bilingue: Macrobe et l’analyse de la diathèse verbale”. Bilinguisme et terminologie grammaticale gréco-latine. (= Orbis/ Supplementa, 27) ed. by Louis Basset et al., 201–224. Leuven, Paris & Dudley, MA: Peeters.Stoppie, Karen, Pierre Swiggers & Alfons Wouters 2007 “La terminologie grammaticale en contexte bilingue: Macrobe et l’analyse de la diathèse verbale”. Bilinguisme et terminologie grammaticale gréco-latine. (=Orbis/ Supplementa, 27) ed. by Louis Basset et al., 201–224. Leuven, Paris & Dudley, MA: Peeters.: 214). The term could derive from the Stoic tradition, which habitually divided things into positive, negative, and neither, but it is also possible that it originated in an attempt by Latin grammarians, for instance by DositheusDositheus, Ars = Dosithei Magistri Ars grammatica Ed. by Heinrich Keil, GL VII, 376–436. (Repr., Hildesheim: G. Olms 2007.)Dositheus, Ars = Dosithei Magistri Ars grammatica Ed. by Heinrich Keil, GL VII1, 376–436. (Repr., Hildesheim: G. Olms 2007.) (Ars Greek text 406.10), to create a Greek equivalent for the term neutrum (Stoppie et al. 2007Stoppie, Karen, Pierre Swiggers & Alfons Wouters 2007 “La terminologie grammaticale en contexte bilingue: Macrobe et l’analyse de la diathèse verbale”. Bilinguisme et terminologie grammaticale gréco-latine. (= Orbis/ Supplementa, 27) ed. by Louis Basset et al., 201–224. Leuven, Paris & Dudley, MA: Peeters.Stoppie, Karen, Pierre Swiggers & Alfons Wouters 2007 “La terminologie grammaticale en contexte bilingue: Macrobe et l’analyse de la diathèse verbale”. Bilinguisme et terminologie grammaticale gréco-latine. (=Orbis/ Supplementa, 27) ed. by Louis Basset et al., 201–224. Leuven, Paris & Dudley, MA: Peeters.: 215).3636.Macrobius is the only Latin author to use the term medium for Greek μέσον (De diff. 27.5–6) in the combinations medium perfectum and medium plusquamperfectum (Schöpsdau 1992Schöpsdau, Klaus 1992 “Vergleiche zwischen Lateinisch und Griechisch in der antiken Welt”. Zum Umgang mit fremden Sprachen in der griechisch-römischen Welt ed. by Carl Werner Müller, Carl Werner, Kurt Sier & Jürgen Werner, 115–136. Stuttgart: F. Steiner.Schöpsdau, Klaus 1992 “Vergleiche zwischen Lateinisch und Griechisch in der antiken Welt”. Zum Umgang mit fremden Sprachen in der griechisch-römischen Welt ed. by Carl Werner Müller, Carl Werner, Kurt Sier & Jürgen Werner, 115–136. Stuttgart: F. Steiner.: 124, n. 60); the term is not documented in this sense by Schad (2007Schad, Samantha 2007 A Lexicon of Latin Grammatical Terminology. Rome & Pisa: Fabrizio Serra.Schad, Samantha 2007 A Lexicon of Latin Grammatical Terminology. Rome & Pisa: Fabrizio Serra.).

6.Terminology

Generally speaking, Latin grammarians tended to translate Greek technical terms into Latin, which explains why so few Greek terms are used in the Latin grammatical tradition (Schironi 2007Schironi, Francesca 2007 “ ΑΝΑΛΟΓΙΑ, analogia, proportio, ratio: Loanwords, calques, and reinterpretations of a Greek technical word”. Bilinguisme et terminologie grammaticale gréco-latine. (= Orbis/ Supplementa, 27) ed. by Louis Basset et al., 321–338. Leuven, Paris & Dudley, MA: Peeters.Schironi, Francesca 2007 “ΑΝΑΛΟΓΙΑ, analogia, proportio, ratio: Loanwords, calques, and reinterpretations of a Greek technical word”. Bilinguisme et terminologie grammaticale gréco-latine. (=Orbis/ Supplementa, 27) ed. by Louis Basset et al., 321–338. Leuven, Paris & Dudley, MA: Peeters.: 320).3737.Schironi also explains why, exceptionally, in the case of analogia the Greek term prevailed, rather than the Latin terms proportio or ratio. Other examples of Greek terms that were not translated include grammatica and etymologia (Coleman 2001Coleman, Robert G. G. 2001 “Greek and Latin”. A History of Ancient Greek, from the Beginnings to late Antiquity ed. by Anastassios-Fivos Christidis, 792–804. Cambridge: Cambridge University Press.Coleman, Robert G. G. 2001 “Greek and Latin”. A History of Ancient Greek, from the Beginnings to late Antiquity ed. by Anastassios-Fivos Christidis, 792–804. Cambridge: Cambridge University Press.: 792). For the translation of Greek terms by Gellius see Holford-Strevens (2003Holford-Strevens, Leofranc 2003 Aulus Gellius: An Antonine scholar and his achievement. Revised edition Oxford: Oxford University Press. Holford-Strevens, Leofranc 2003 Aulus Gellius: An Antonine scholar and his achievement. Revised edition Oxford: Oxford University Press. : 229). The discussion about verbal diathesis mentioned above brings to the fore some particulars of Macrobius’ approach and terminology. In distinguishing between active and passive he focuses on the semantic roles of the nominal constituents “because in both active and passive there must always be two persons, one performing the action and the other undergoing it” (nam et in activo et in passivo debent omni modo duae, et administrantis et sustinentis subesse personae; De diff. 163.10–12).3838.Cf. the passage in De verboAnonymus, De verbo = Anonymus, De verbo Ed. by Heinrich Keil, GL V, 634–654; ed. by Marina Passalacqua, Tre testi grammaticali bobbiesi ( GL V 555–566; 634–654; IV 207–217 Keil). Rome: Edizioni di Storia e Letteratura 1984.Anonymus, De verbo = Anonymus, De verbo Ed. by Heinrich Keil, GL V1, 634–654; ed. by Marina Passalacqua, Tre testi grammaticali bobbiesi (GL V1 555–566; 634–654; IV1 207–217 Keil). Rome: Edizioni di Storia e Letteratura 1984. (52 Passalacqua): “Active is when someone undergoes someone else’s action, passive is when the action belongs to one person while someone else undergoes it” (activum est ubi altero agente alter patitur, passivumque est quando altero patiente penes alterum est actio), which also focuses on the participants but with different terms. The passage in Apollonius Dyscolus (Synt. 296.1–4) about verbal diathesis focuses on the actions/passions, rather than the participants.

In Latin grammar, Diomedes’ definition resembles the one given by Macrobius: “Passive is when the action is performed by one person, while the other undergoes it” (passiva est cum alio patiente penes alium sit administratio; Ars I, 336.32–33), but with administratio instead of action (see Mastriani 2022Mastriani, Maria Camilla 2022 Lo studio del verbo nella tradizione grammaticale latina: Per un corpus comparativo. Ph.D. diss., Università degli Studi di Napoli.Mastriani, Maria Camilla 2022 Lo studio del verbo nella tradizione grammaticale latina: Per un corpus comparativo. Ph.D. diss., Università degli Studi di Napoli.: 148–149). Administratio rei for ‘performing an action’ also occurs in Charisius (Ars 209.24), but he refers to the participants with different terms: “Where there is someone who acts, there needs to be someone who undergoes the action” (ubi est qui facit, necesse est ut sit qui patitur; Ars 210.18–19).

Among the technical terms used in De differentiis there is one term that Macrobius shares with only one Latin grammarian, the author of the Instituta artium, commonly believed to be Valerius Probus (fourth century), namely minusquamperfectum ‘less-than-perfect’ as a synonym for imperfectum.3939.The authorship of the Instituta artium ( GL GL = Grammatici latini Ed. by Heinrich Keil. 8 vols. Leipzig: B.G. Teubner 1864 (Repr., Hildesheim: G. Olms 2007.)GL = Grammatici latini Ed. by Heinrich Keil. 81 vols. Leipzig: B.G. Teubner 1864 (Repr., Hildesheim: G. Olms 2007.) IV, 47–192) is a contested issue (Kaster 1988Kaster, Robert A. 1988 Guardians of language: The grammarians and society in Late Antiquity. Berkeley, Cal.: University of California Press. Kaster, Robert A. 1988 Guardians of language: The grammarians and society in Late Antiquity. Berkeley, Cal.: University of California Press. : 348–350; Zetzel 2018Zetzel, James E. G. 2018 Critics, Compilers, and Commentators: An introduction to Roman philology, 200 BCE-88 CE. Oxford: Oxford University Press.Zetzel, James E. G. 2018 Critics, Compilers, and Commentators: An introduction to Roman philology, 200 BCE-88 CE. Oxford: Oxford University Press.: 312–316). Outside these two treatises the term does not occur anywhere. Macrobius most likely found this term somewhere in his sources, and since the term is found nowhere else,4040.Note that the Thesaurus linguae Latinae (VIII, 1045, 50–54) gives only the loci in Macrobius and Probus, which means that the term is really unique. his source must somehow be related to the Instituta artium.

An important term in Macrobius’ analysis of verbal forms is prima positio, for instance when he explains the Greek perfect λέλυκα, where the syllable -λυ- is referred to as “the second syllable corresponding to the origin [of the verb], since it is [the syllable] with which the first position of the verb [i.e. the first person of the present indicative] begins” (secunda originis, quia prima verbi positio ab hac incipit, De diff. 49). The term prima positio is widely used in Latin grammatical literature, while its Greek equivalent, πρώτη θέσις, was originally used in discussions about the origin of speech. In morphological literature, it marked the distinction between root words and derivatives (e.g. in Dionysius Thrax, Techne Dionysius Thrax, Techne = Dionysius Thrax, Ars grammatica Ed. by Gustav Uhlig, GG I/1, 1–101. Leipzig: B.G. Teubner 1901 (Repr., Hildesheim: G. Olms 1965.)Dionysius Thrax, Techne = Dionysius Thrax, Ars grammatica Ed. by Gustav Uhlig, GG I/11, 1–101. Leipzig: B.G. Teubner 1901 (Repr., Hildesheim: G. Olms 1965.) 25.2).4141.See Siebenborn (1976Siebenborn, Elmar 1976 Die Lehre von der Sprachrichtigkeit und ihren Kriterien: Studien zur antiken normativen Grammatik. Amsterdam: B. R. Grüner.Siebenborn, Elmar 1976 Die Lehre von der Sprachrichtigkeit und ihren Kriterien: Studien zur antiken normativen Grammatik. Amsterdam: B. R. Grüner.: 75, 152). In this case Macrobius may have leaned more towards its use in Latin grammatical terminology.4242.Elsewhere Macrobius uses another term, thema, with the meaning of ‘stem’, for instance when he explains the future tense διδάξω by stating that it is derived from a present tense διδάχω (διδάξω a themate non διδάσκω sed διδάχω, cuius indicium est διδαχή; De diff. 169.12). The term thema is also used in the treatise De verboAnonymus, De verbo = Anonymus, De verbo Ed. by Heinrich Keil, GL V, 634–654; ed. by Marina Passalacqua, Tre testi grammaticali bobbiesi ( GL V 555–566; 634–654; IV 207–217 Keil). Rome: Edizioni di Storia e Letteratura 1984.Anonymus, De verbo = Anonymus, De verbo Ed. by Heinrich Keil, GL V1, 634–654; ed. by Marina Passalacqua, Tre testi grammaticali bobbiesi (GL V1 555–566; 634–654; IV1 207–217 Keil). Rome: Edizioni di Storia e Letteratura 1984. (27.2 Passalacqua omne thema verbi Latini). Schad (2007Schad, Samantha 2007 A Lexicon of Latin Grammatical Terminology. Rome & Pisa: Fabrizio Serra.Schad, Samantha 2007 A Lexicon of Latin Grammatical Terminology. Rome & Pisa: Fabrizio Serra.: 401) cites seven occurrences in Latin, all from Macrobius. In Greek grammatical terminology the term is not uncommon; it is mainly used for the first person of the present indicative of the verb (Bécares Botas 1985Bécares Botas, Vicente 1985 Diccionario de la terminología gramatical griega. Salamanca: Ediciones Universidad de Salamanca.Bécares Botas, Vicente 1985 Diccionario de la terminología gramatical griega. Salamanca: Ediciones Universidad de Salamanca.: 205–206).

In general, Macrobius seems to have borrowed profusely from the Greek tradition, for instance in the case of the terminological pair foris and intus declinare, which does not seem to have a parallel in Latin grammatical literature. Macrobius uses these terms to explain the place of the augment of the imperfect tense in compound verbs, either ‘inside’ (intus), between the preposition and the verbal stem, as in κατέγραφον ‘I registered, wrote down’, or ‘outside’ (foris), before the entire compound verb, as in ἐκιθαρῷδουν ‘I sang to the cithara’ (De diff. 19). Schad (2007Schad, Samantha 2007 A Lexicon of Latin Grammatical Terminology. Rome & Pisa: Fabrizio Serra.Schad, Samantha 2007 A Lexicon of Latin Grammatical Terminology. Rome & Pisa: Fabrizio Serra.: 175) lists Macrobius as the only writer using these expressions. In the Greek sources they are represented by the pair ἔσωθεν and ἔξωθεν, commonly used by Apollonius Dyscolus, and after him by Choeroboscus (see Bécares Botas 1985Bécares Botas, Vicente 1985 Diccionario de la terminología gramatical griega. Salamanca: Ediciones Universidad de Salamanca.Bécares Botas, Vicente 1985 Diccionario de la terminología gramatical griega. Salamanca: Ediciones Universidad de Salamanca.: 170, 191).

The larger topic of the function of the preposition in verbal compounds, to which Macrobius refers, is dealt with by Choeroboscus ( Scholia Choeroboscus, Scholia I = Georgius Choeroboscus, Prolegomena et scholia in Theodosii Alexandrini Canones isagogicas de flexione nominum Ed. by Alfred Hilgard, GG IV/1, 103–417. (Repr., Hildesheim: G. Olms 2007.)Choeroboscus, Scholia I = Georgius Choeroboscus, Prolegomena et scholia in Theodosii Alexandrini Canones isagogicas de flexione nominum Ed. by Alfred Hilgard, GG IV/11, 103–417. (Repr., Hildesheim: G. Olms 2007.) II, 54.3–11). He distinguishes between prepositions retaining their own sense when combined with a verb and those that do not change the meaning of the compound verb from that of the simple verb (see Macrobius, De diff. 43). Macrobius adds that the two languages differ in this respect: while in Greek a change of preposition often does not change the meaning of the verb, in Latin a change of preposition almost always indicates a change in meaning.4343. De diff. 21.7–10: “You may remember that you’ll hardly ever find in Latin a preposition that doesn’t add something to the meaning, unlike in Greek where a preposition often does not produce any change in meaning” (memineris nullam fere invenis apud Latinos praepositionem quae nihil addat sensui, sicut apud Graecos saepe praepositio nullam sensui facit permutationem (cf. Dionysius Thrax, Techne Dionysius Thrax, Techne = Dionysius Thrax, Ars grammatica Ed. by Gustav Uhlig, GG I/1, 1–101. Leipzig: B.G. Teubner 1901 (Repr., Hildesheim: G. Olms 1965.)Dionysius Thrax, Techne = Dionysius Thrax, Ars grammatica Ed. by Gustav Uhlig, GG I/11, 1–101. Leipzig: B.G. Teubner 1901 (Repr., Hildesheim: G. Olms 1965.) 57.5–59.2). Note the style of directly addressing the reader in the second person by memineris, about which see Scappaticcio (2015Scappaticcio, Maria Chiara 2015 Artes Grammaticae in frammenti: I testi grammaticali latini e bilingui greco-latini su papiro. Edizione commentata. Berlin and Boston: de Gruyter. Scappaticcio, Maria Chiara 2015 Artes Grammaticae in frammenti: I testi grammaticali latini e bilingui greco-latini su papiro. Edizione commentata. Berlin and Boston: de Gruyter. : 502).

7.Conclusion: Macrobius and the two languages

Since the text of the treatise De differentiis has been preserved in a fragmented way, it is unknown whether at some stage Macrobius may have provided more insight in his motive to compare Latin and Greek. At any rate, his approach differed from that of common grammarians, whom he held in low esteem (Kaster 1988Kaster, Robert A. 1988 Guardians of language: The grammarians and society in Late Antiquity. Berkeley, Cal.: University of California Press. Kaster, Robert A. 1988 Guardians of language: The grammarians and society in Late Antiquity. Berkeley, Cal.: University of California Press. : 62, 196, 206), just like his predecessor and model Aulus Gellius did (Holford-Strevens 2003Holford-Strevens, Leofranc 2003 Aulus Gellius: An Antonine scholar and his achievement. Revised edition Oxford: Oxford University Press. Holford-Strevens, Leofranc 2003 Aulus Gellius: An Antonine scholar and his achievement. Revised edition Oxford: Oxford University Press. : 172f.; 2023Holford-Strevens, Leofranc 2023 “Grammar and grammarians: Linguistic and social change from Gellius to Macrobius”. Roman Perspectives on Linguistic Diversity: Guardians of a changing language ed. by Adam Gitner, 191–198. Oxford: Oxford University Press. Holford-Strevens, Leofranc 2023 “Grammar and grammarians: Linguistic and social change from Gellius to Macrobius”. Roman Perspectives on Linguistic Diversity: Guardians of a changing language ed. by Adam Gitner, 191–198. Oxford: Oxford University Press. ). In the Saturnalia, Macrobius presents the Virgil commentator and grammarian Servius as a paragon of erudition, but he subtly indicates that this young man of low birth should know his place (verecundia) among the noble participants in the discussion (Kaster 1980Kaster, Robert A. 1980 “Macrobius and Servius: Verecundia and the grammarian’s function”. Harvard Studies in Classical Philology 84. 219–262. Kaster, Robert A. 1980 “Macrobius and Servius: Verecundia and the grammarian’s function”. Harvard Studies in Classical Philology 841. 219–262. ).

Macrobius aims to present Latin and Greek on an equal footing: the two languages are very much alike, and if you know one, you know the other too. He believed that there was a universal structure underlying the syntax of the two languages, so that the Greek grammatical framework could serve equally well for a better understanding of Latin.4444.The idea of a universal structure for the two languages fits in with Macrobius’ Neo-Platonic views (Petrilli 2009Petrilli, Raffaella 2009 “Macrobius”. Lexicon grammaticorum ed. by Harro Stammerjohann, II, 950–951. 2nd edition. Tübingen: Niemeyer.Petrilli, Raffaella 2009 “Macrobius”. Lexicon grammaticorum ed. by Harro Stammerjohann, II, 950–951. 2nd edition. Tübingen: Niemeyer.), with their emphasis on universal concepts being represented by different ordinary languages (Helmig 2008Helmig, Christoph 2008 “Proclus and other Neoplatonists on universals and predication”. Documenti e Studi sulla Tradizione Filosofica Medievale 19. 31–52.Helmig, Christoph 2008 “Proclus and other Neoplatonists on universals and predication”. Documenti e Studi sulla Tradizione Filosofica Medievale 191. 31–52.). Macrobius regarded their relationship as a natural given, not shared by any other language pair. This does not entail that the traditional components of Latin and Greek “merged together in a new linguistic entity”, as Gatto (2021Gatto, Biagio 2021 Priscian: A syntactic interpretation of a Graeco-Roman World. Ph.D. thesis, University College London.Gatto, Biagio 2021 Priscian: A syntactic interpretation of a Graeco-Roman World. Ph.D. thesis, University College London.: 144) suggests, because Macrobius treats the two languages as separate, though closely related, systems.

The close relationship between Latin and Greek was taken for granted by all Latin grammarians. Some of them went even further. Priscian, for instance, declares his indebtedness to Apollonius Dyscolus in his discussion of the parts of speech (in ante expositis libris de partibus orationis in plerisque Apollonii auctoritatem secuti sumus) and he announces that he will follow his traces in the part about syntax as well (Institutiones 107.24–25). Apparently, he saw his task as one of transferring Greek linguistic knowledge to Latin in order to get a grip on the structure of Latin (Gatto 2021Gatto, Biagio 2021 Priscian: A syntactic interpretation of a Graeco-Roman World. Ph.D. thesis, University College London.Gatto, Biagio 2021 Priscian: A syntactic interpretation of a Graeco-Roman World. Ph.D. thesis, University College London.: 91, 100). This had been the grammarians’ practice from the start, when they set up the framework of Latin grammar on the basis of the translation of Greek terminology. Priscian, however, did not stop at believing that the two languages had a common structure: he also viewed Latin as heavily dependent on Greek and even went so far as to look for a Greek origin of Latin idioms (Schöpsdau 1992Schöpsdau, Klaus 1992 “Vergleiche zwischen Lateinisch und Griechisch in der antiken Welt”. Zum Umgang mit fremden Sprachen in der griechisch-römischen Welt ed. by Carl Werner Müller, Carl Werner, Kurt Sier & Jürgen Werner, 115–136. Stuttgart: F. Steiner.Schöpsdau, Klaus 1992 “Vergleiche zwischen Lateinisch und Griechisch in der antiken Welt”. Zum Umgang mit fremden Sprachen in der griechisch-römischen Welt ed. by Carl Werner Müller, Carl Werner, Kurt Sier & Jürgen Werner, 115–136. Stuttgart: F. Steiner.: 132–135).

The text of Priscian’s Institutiones grammaticae Priscian, Institutiones = Prisciani grammatici Caesariensis Institutiones grammaticarum libri XVIII Ed. by Martin Hertz, GL II–III. Leipzig: B.G. Teubner 1859 (Repr., Hildesheim: G. Olms 2007.)Priscian, Institutiones = Prisciani grammatici Caesariensis Institutiones grammaticarum libri XVIII Ed. by Martin Hertz, GL II–III1. Leipzig: B.G. Teubner 1859 (Repr., Hildesheim: G. Olms 2007.) presents several parallels with the text of De differentiis, and Priscian is likely to have used Macrobius’ text. His audience in Constantinople was Greek-speaking and, since his purpose in the Institutiones was to teach his students Latin, he may have found the Latin portions of De differentiis useful. Some of the parallels between Macrobius and Priscian may be explained by the fact that they both used Apollonius Dyscolus as their main source of information about Greek grammar (Fortes 2019Fortes, Fábio 2019 “Priscian and the Graeco-Roman bilingualism: An example of cross-linguistic analysis within Priscian’s De constrvctione (Inst. gramm. XVII)”. Revista de Estudios Clásicos 47. 25–35.Fortes, Fábio 2019 “Priscian and the Graeco-Roman bilingualism: An example of cross-linguistic analysis within Priscian’s De constrvctione (Inst. gramm. XVII)”. Revista de Estudios Clásicos 471. 25–35.).

The fundamental equality of the two languages did not mean that, deep down, speakers of Latin could not have their predilections. In a few passages Macrobius’ pride of Latin surfaces, for instance when he speaks about the Latin gerundium and supinum, forms unknown in Greek. About the gerund he says: “This form does not only provide Latin with an adornment, but also makes us appear to possess something that, understandably, the Greeks would like to have, too” (haec forma Latinitati non solum praestat, sed illud quoque, ut aliquid habere videamur quod Graeci iure desiderent, De diff. 159). One of the differences between Greek and Latin — and here the Roman feeling of superiority surfaces again — is that Latin is simpler than the needlessly complicated Greek (contenta simplicibus Latinitate compendiis, De diff. 27.10).4545.The conciseness of Latin compared to Greek verbosity is a topos, cf. Saturnalia 2.12.16: "Our language happens to be regarded as more concise and brief than that of the Greeks" (solet nostra quam Graecorum lingua brevior et angustior aestimari) This is also manifested in the Greeks’ predilection for double grammatical terms such as the past perfect or the near future, not to mention the (redundant) aorist ( De verboAnonymus, De verbo = Anonymus, De verbo Ed. by Heinrich Keil, GL V, 634–654; ed. by Marina Passalacqua, Tre testi grammaticali bobbiesi ( GL V 555–566; 634–654; IV 207–217 Keil). Rome: Edizioni di Storia e Letteratura 1984.Anonymus, De verbo = Anonymus, De verbo Ed. by Heinrich Keil, GL V1, 634–654; ed. by Marina Passalacqua, Tre testi grammaticali bobbiesi (GL V1 555–566; 634–654; IV1 207–217 Keil). Rome: Edizioni di Storia e Letteratura 1984. 26). Apparently, the Greeks’ boasting (iactantia) about the richness of their language compared to Latin was a sore point, since Phocas ( Ars Phocas, Ars = Phocas, Ars de nomine et verbo Ed. by Heinrich Keil, GL V, 410–439. (Repr., Hildesheim: G. Olms 2007.)Phocas, Ars = Phocas, Ars de nomine et verbo Ed. by Heinrich Keil, GL V1, 410–439. (Repr., Hildesheim: G. Olms 2007.) 438.3) remarks that the absence of impersonal passives in Greek is a good counterargument against this claim. CiceroCicero, De fin. = M. Tullius Cicero, De finibus bonorum et malorum Ed. and transl. by Harris Rackham. 2nd ed. London: Heinemann 1931.Cicero, De fin. = M. Tullius Cicero, De finibus bonorum et malorum Ed. and transl. by Harris Rackham. 2nd ed. London: Heinemann 1931. believes that the Greeks’ claim is spurious anyway and boldly states that “not only are we not defeated by the Greeks in lexical richness, but we are even superior to them in this” (nos non modo non vinci a Graecis verborum copia, sed esse in ea etiam superiores; De fin. 3, 5; Weis 1992Weis, Rudolf 1992 ““Zur Kenntnis des Griechischen im Rom der republikanischen Zeit””. Zum Umgang mit fremden Sprachen in der griechisch-römischen Welt ed. by Carl Werner Müller, Carl Werner, Kurt Sier & Jürgen Werner, 137–142. Stuttgart: F. Steiner.Weis, Rudolf 1992 ““Zur Kenntnis des Griechischen im Rom der republikanischen Zeit””. Zum Umgang mit fremden Sprachen in der griechisch-römischen Welt ed. by Carl Werner Müller, Carl Werner, Kurt Sier & Jürgen Werner, 137–142. Stuttgart: F. Steiner.: 142).

The difference between Macrobius and other grammarians using Greek material in their treatises, like Diomedes and Dositheus and, at a later age, Priscian is that they were native speakers of Greek, writing for a Greek-speaking audience about the Latin language, whereas Macrobius was a native speaker of Latin, whose knowledge of Greek had been acquired during his education. His perspective was that of a comparativist, rather than a teacher of either language.

In the end, what Macrobius wished to demonstrate was the similarities between the underlying morphological rules of the two languages, even when the outcome produced by these rules was different. At the same time, he clearly wished to exhibit the ease with which he himself handled both languages and to show off his knowledge of Greek grammatical literature, which no doubt was superior to that of common grammarians. All of this was part of his status as a man of learning with a vast erudition and knowledge about Latin and Greek language and literature. Not constrained by any pedagogical considerations, he presented his knowledge about Latin and Greek grammar in a manner fit for an intellectual audience.