lii'' iti'lii: ■liii'iiiiili! 1 I t ifli* ! A^ PARTES OKTrnONIS AS DISCUSSED BY VIRGILIUS MARO GRAMMATICUS WITH SOME OBSERVATIONS UPON HIS INFLECTION AND SYNTAX BY CLEMENT ORESTES MEREDITH i: :,■ iil A DISSERTATION SUBMITTED TO THE BOARD OF UNIVERSITY STUDIES OF THE JOHNS HOPKINS UNIVERSITY IN CONFORMITY WITH THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY BALTIMORE 1912 THE PARTES ORATIONIS AS DISCUSSED BY VIRGILIUS MARO GRAMMATICUS WITH SOME OBSERVATIONS UPON HIS INFLECTION AND SYNTAX BY CLEMENT ORESTES MEREDITH A DISSERTATION SUBMITTED TO THE BOARD OF UNIVERSITY STUDIES OF THE JOHNS HOPKINS UNIVERSITY IN CONFORMITY WITH THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY BALTIMORE 1912 » • , : i CONTENTS Bibliography 5 Introduction 9 The Parts of Speech 15 Nomen 17 Qualitas 17 Comparatio 23 Genus 24 Numerus 26 Figura 27 Casus 29 Declinatio 30 Pronomen 34 Declinatio 37 Verbum 42 Qualitas (modi) • 45 Qualitas (formae) 52 Coniugatio 55 Numerus 60 Figura 61 Persona 61 Significatio, or Genus 62 Tempora 64 Participium 66 Adverbium 69 Praeposito 70 Coniunctio 70 Interiectio 71 Conclusion 73 Vita 75 NOTE: Page 30, heading Declaratio should be Dechnatio. 34662J BIBLIOGRAPHY Babbler, J. J., Beitrage zu einer Geschichte der lateinischen. Gram- matik im Mittelalter (Halle a. S., 1885). Bonnet, M., Le Latin de Gregoire de Tours (Paris, 1890). BUECHELER, F., Grundriss der lateinischen Declination (Bonn, 1879). COLLIGNON, A., Note sur une grammaire latine manuscrite du viiie sie- cle appartenant a la Bibliotheque de Nancy, contenant des fragments inedits de Virgilius Maro, Revue de Phi- lologie, Vol. VII. p. 13-22. DlEZ, F., Romanische Grammatik (Bonn, 1882). Draeger, a., Historische Syntax der lateinischen Sprache (Leipzig. 1878). Ernault, E., De Virgilio Marone Grammatico Tolosano (Paris, 1886). Geyer, p., Beitrage zur Kenntnis des gallischen Lateins, Archiv II. 25 f. Grammtici Graeci: Dionysii Thracis ars grammatica, ed. G. Uhlig (Leipzig, 1883) ; ApoUonii Dyscoli quae supersunt, ed. R. Schneider (Leipzig, 1910). Grammatici Latini, ex recensions Henrici Keilii (Lipsiae, 1857-1880. Huemer, J., Die Epitomae des Grarmiatikers Virgilius Maro nach dem Fragmentum Vindobonense 19556 (Wien, 1882). Jeep, L., Zur Geschichte der Lehre von den Redetheilen bei den latei- nischen Grammatikern (Leipzig, 1893). Job, L., De grammaticis vocabulis apud Latinos (Paris, 1893). Landgraf, G., Historische Grammatik de, lateinischen Sprache. Drit- ter Band, erstes Heft. (Leipzig, 1903). Manitius, M., Geschichte der lateinischen Literatur des Mittelalters. ErsterTeil (Munchen, 1911). Meyer-Lubke W., Grammatik der romanischen Sprachen (Leipzig- 1890-1894. Nettleship, H., The Study of Latin Grammar among the Romans in the First Century A. D. Journal of Philology, Vol. xv (1886), reprinted in Lectures and Essays, second series Oxford, 1895). Neue-Wagener, Lateinische Formenlehre (Berlin, 1902.) Rogers, M., L'enseignement des lettres classiques d'Ausonea Alcuin Paris, 1905. Chap. Ill, sec. 3, pp. 110-126, is devoted to 'Le grammarien Virgile'. ScHMALZ J. H., Lateinische Grammatik: Syntax und Stilistik [Pvliin- chen, lylOJ. Stangl, Th., Die grammatischen Schriften des GalliersVirgilius Maro auf Grund einer erstmaligen Vergleichung der Hand- schriften von Paris und Neapel (Miinchen, 1891). Strong, H. A., On the Grammarian, Virgilius Maro. Classical Review, XXV [1911] 70-71. This is merely a summary of an article by H. Zimmer. Thurot, C., Notices et extraits de divers manuscrits latins pour ser- vir a I'histoire des doctrines grammaticales au moyen age (Paris, 1868). Zimmer, H., Der Gasconger Virgilius Maro Grammaticus in Irland (Sitzungsberichte derKoniglich Preussischen Akademie der Wissenschaften, Jahrg. 1910, 1031-1098). Editions used: Huemer, J., Virgilii Maronis Grammatici Opera (Lip- siae, 1886). Mai, Aug., Classici Auctores, V, 1-149. THE PARTES ORATIONIS AS DISCUSSED BY VIRGILIUS MARO GRAMMATICUS WITH SOME OBSERVATIONS UPON HIS INFLECTION AND SYNTAX INTRODUCTION Very little is known of the late Latin grammarian who calls himself Virgilius Maro Grammaticus. Even his name was assumed, and his date can only be conjectured. Almost all that can be said of him with certainty is that he be- longed to Southwestern Gaul, and that he was a Christian. The name Virgilius Maro was bestowed upon him by his teacher 'Aeneas': (92,18) unde Aeneas cum me vidisset ingeniosum hoc me vocabulo iussit nominari dicens: hie filius mens Maro vocabitur, quia in eo antiqui Maronis spiritus redivivit. In this same passage (92, 21) he speaks of his grandfather 'Martulis': hie in arte grammatica di- ligentissimus. And at 28, 1 he mentions an uncle named 'Samminius'. He is commonly assigned to the seventh century, though various dates have been suggested for him, ranging all the way from the fifth century to the ninth. Angelo Mai (Class. Auct. V, ix) identified him with the Virgilius Maro who is mentioned by Ennodius, Carm. ii, 118, In tantum prisci defluxit fama Maronis, Ut te Virgilium saecula nostra darent, and ii, 120, Captivo stultus congaudet stemmate vatis. Non est Virgilius, dicitur iste tamen. [10] This, however, has met with very little favor from other scholars. F. Osann (Beitrage, i-ii, p. 126 f.) assigned him to the ninth century— on the inconclusive reasoning that his Latin is not very good, and that his contemporaries were fond of assuming the names of the great writers of antiquity. 1 L. Traube (Hermes, xxiv, 647) set a definite limit by finding in one of the letters of Aldhelmus a quo- tation which Virgilius had already used. That is, Aldhelmus (p. 95 Giles) quotes a line from one of the grammarian's authorities, Glengus: Si veroquippiam,inscitiasuppeditante, garrula frontose convincitur pagina prompsisse: ut versi- dicus ait: Digna fiat fante glingio guro fugax fambulo. And this is to be connected with Virgilius, 121, 9: verum- tamen ne in illud Glengi incedam, quod cuidam conflictum fugienti dicere fidenter ausus est gurgo inquit fugax fam- bulo dignus est, etc. M. Manitius has proposed a limit in the other direction, on the ground that many of Virgilius' etymologies are derived from Isidorus.^ His 'floruit' would thus fall later than Isidorus (d. 636) and earlier than Ald- helmus (d. 709), and he should probably be assigned to the middle of the seventh century. H. Zimmer, by an elaborate argument, has placed him, at the latest, at the end of the fifth century or about the beginning of the sixth, and set his 'floruit' about two generations later than Ausonius, one generation later than Sulpicius Severus.^ 1 So the group of scholars who were gathered at the Court of Charle- magne used to call each other by the names of great ancient poets. Alcuin was Flaccus, one of his disciples was Naso, Angilbert was Ho- merus. See Ad. Ebert, AUgemeine Geschichte der Literatur des Mittelalters im Abendlande, ii 64. - Geschichte der lateinischen Literatur des Mittelalters, Miinchen, 1911, p. 120. 3 Sitzungsberichte der Koniglich Preussischen Akademieder Wissen- schaften, Berlin, 1910, p. 1067. [11] Virgilius himself tells us indirectly that he was a Gaul: (137, 30) in quibusdam Gallorum nostrorum scriptis; and his expression 'bigerro sermone clefabo' (8, 13) apparently refers to the district of Bigorre, just north of the Pyrenees. A reference in Abbo Floriacensis indicates that he lived in Toulouse: Scripulus appenditur xvi granis lentis, licet Vir- gilius Tolosanus in suis opusculis asserat pensari xviii granis hordei, etc.^ And in the title of the Fragmentum Mediolanense he is called Virgilius presbyter Hispanus. Zimmer (op. cit. 1071) argues that this last indiction is not inconsistent with the others; that from 415 to the middle of the sixth century the West Goths held sway both north and south of the Pyrenees, and that from 507 on the region north of the Pyrenees might very well be regarded merely as an appanage of their Spanish kingdom. He further maintains (p. 1092 f . ) that Virgilius visited Ireland in the last quarter of the fifth century, and that he had a very considerable influence upon the language of Irish poetry for the next two hundred years. And a recent conjecture by H. A. Strong, Hihernorum for Hibonorum (26, 3), is offered as confirming this opinion. ^ A single passage may be quoted to show that Virgilius was a Christian: (108, 17) in tribus divinae unitatem sub- stantiae personis coaeternaliter subsistentem.^ He was thus a 'Homoousian' a fact which qualified him (in Zimmer's theory, p. 1094) to take part in the great exodus of Ortho- 1 Mai, Class. Auct. v, 349, n. 1. The statement here ascribed to Virgilius Tolosanus does not appear in our Virgilius' extant works, and it may come from some other writer of the same name. Still, Abbo's Virgilius Tolosanus and our Vii-gilius are now commonly regar- ded as one and the same person; cf., e. g., E. Ernault (p, 11): Tolosa- num Virgilium esse nostrum censeo, quia certum est scholam eius in Aquitania tloiuisse, ubi Hiberni, qui cum eadem regione saepe com- municabant, illam doctrinam novcrunt, susceperunt et ipsi postea Anglo-Saxones docuerunt. ■' Classical Review, xxv, (1911) 201. •' For other evidence, see 107, 7; 129, 12; 133, 20; 175, 12. [12] dox Christians from Gascony to Ireland before the perse- cution of the Arian West Goths. His writings, so far as they have survived, are all of a grammatical character. There are fifteen Epitomae: 1 De Sapientia, 2 De Litera, 3 De Syllaba, 4 De Metris, 5 De Nomine, 6 De Pronomine, 7 De Verbo, 8 De Adverbio, 9 De Participio, 10 De Coniunctione, 11 De Praepositione, 12 De Interiectione, 13 De Scinderatione Fonorum, 14 De Cognominationibus Nominum atque Verborum, 15 De Ca- talogo Grammaticorum. There are also eight Epistolae, devoted to the eight parts of speech. He speaks himself of an earlier set of fifteen Epistolae sent to his young friend Fabianus: (107, 5) in xv epistolarum Affricam mis- sarum volumine. At 25, 1 he mentions a metrical treatise which he had written: de safico autem et heroico versuum metro in quadam epistola, quam inter duodecim ad Dona- tum Romam missimus, discribsisse me suflficienter memini. And at 175, 11 he mentions a bit of religious controversy in which he had engaged: cum librum de mundi creatione commentarium adversus paganos ediderim. The very subjects of some of his Epitomae may prepare the reader for a somewhat unusual treatment of Latin grammar. The first and firteenth give some account of the twelve dialects of Latin, the thirteenth gives examples and rules for various kinds of cryptic writing.^ Such topics suggest that Virgilius is of less importance for the study of late Latin in general or of Gallic Latin as a whole than for the specific forms of Latin which were written and spoken in Southwestern Gaul in his day. His own position as a grammarian is the more puzzling, because he never quotes a classical author, or even one of the Grammatici Latini. Some of his authorities, like him- » Zimmer, op. cit., 1038 f., finds parallels in Old Irish poetry. For the twelve dialects of Latin, cf., also, Hisperica Farnina (ed. Jenkin- son, Cambridge, 1908, p. 5) : Bis senos exploro vechros qui ausonicam lacerant palatham. [13] self, had borrowed ancient and honorable names: Cato, Cicero, Donatus, Homerus, Honoratus, Horatius, Lucanus, Propetius, Quintilianus, Terrentius. Others bore such names as Balapsidus, Bregandus Lugenicus, Falanges, Fassica, Iscenus, Mitterius, Sabatinus, Siva, But these men, if they ever existed at all, may all have been his own contemporaries, or residents of his own region.^ There is a similar variety in the names of his teachers. These were, Aeneas: (34, 1) cum me Aeneas sciscientem adhuc rudem- que discipulum lucide vellet inbuere;^ Galbungus: (60, 13) ab omni scola Galbungi, quam et ego adivi; Gratianus: (63, 2) Gratiano magistro fretus, cuius in scola decem an- nos feci; Reginus Cornelius: (133, 18) qui me quoque docuit; Sulpita and Istrius: (24, 7) praeceptores nostri et praecipue Sulpita atque Istrius. Some of the theories, and some of the incidents, which Virgilius reports give one a rather discouraging picture of the grammatical studies of his country and time. For ex- ample, Galbungus and Terrentius once debated for fourteen days and as many nights over the vocative of 'ego': (123, 16) quattuordecim diebus totidemque noctibus in conten- tione mansisse reiferuntur. Regulus and Sedulus carried on a still longer and more heated controversy over the in- choative verbs: (138, 25) non minima quaestio habita est, quae usque ad gladiorum pene conflictum pervenit. quin- decim nam.que noctibus totidemque simul diebus insomes et indapes mansere, tribus militibus utrimique sumptis. And the 'catalogus grammaticorum' (87, 23) reminds one of Trimolchio's account of the origin of Corinthian bronze: • At 49, 17 he reports that he once attended a congress of not less than thirty grammarians. It is hardly safe to assume, as many schol- ars have done, that Virgilius invented all his authorities. Manitius suggests (op. cit., 121 n. 1) that foreign names of places are some- times used to deyignate places in Gaul — for example, Arabia, Endia, Africa (92, 12) — that perhaps Africa (107, 5; 1G4, 8) means Spain, Europa (109, 15; 164, 8) means France. 2 See, also, 60,14; 62,23; 92,18. [14] Primus igitur fuit quidam senex Donatus apud Troiam, quern ferunt mille vixisse annos. hie cum ad Romulum, a quo condita est Roma urbs, venisset, gratulantissime ab eodem susceptus, IIII continues ibi fecit annos, scolam construens et innumerabilia opuscula relinquens fuit itidem apud Troiam quidam Virgilius eiusdem Donati audi- tor, qui in discribendis versibus diligentissimus erat, qui LXX volumina de ratione metri scribens et epistolam ad Virgilium Assianum missam de verbi explanatione .... hunc vidi meis oculis, et puerulo mihi notas caraxavit. It may be said, however, that our author's own Latinity is much better than these passages might lead one to suspect. This statement may be made with confidence, as it rests upon a somewhat detailed examination of his usage. The following investigation attempts in a similar manner to gain a more definite idea of his ability and training as a grammarian. It proceeds upon the well-known fact that grammar in general, and antique grammar in particular, is both exact and conservative to a remarkable degree. The varying theories of individual scholars or of different schools are clearly marked in scholastic tradition. Techni- cal terms, definitions, the arrangement and division of topics, even the examples chosen for paradigms and illus- trations, are passed down from generation to generation. In this preliminary investigation, therefore, I have con- fined myself to the parts of speech and have applied these tests to Virgilius by comparing his statements with parallel material in the corpus of Latin Grammarians. In some cases, unfortunately, the chronological sequence of these grammarians cannot be certainly established. Experience, however, has shown that in this particular investigation no important point has been seriously affected by lack of definite dates. [15] The Parts of Speech ^ Virgilius suggests (5,4 ff. ) that the parts of speech should be called partes latinitatis, ^ rather than partes orationis. For latinitas, which is derived from latitude ipsius linguae, is a broader term than oratio, which properly means the ornate language of the orator: (5,10) haec autem latinitas propter oratorum ornatissimum leporem oratio nuncupatur, unde et partes orationis intellegendae sunt partes latinita- tis. This discussion is a more or less distant echo of such definitions or derivations of oratio as Scaurus ap. Diomed. (K. 1300,19); Dositheus, (K. VII 389,8); Charisius (K. I 152,11) ; Explanat. (K. IV 487,23) ; Pompeius (K. V96,19) ; Victorinus (K. VI 192,3); Audax, (K. VII 324,9) ; Priscian, (K. II 53,27). 'Partes orationis' (hi*pti toCXoyov) appears first in Varro,— see Job, p. 162,— and becomes traditional. The attempt to solve the difficulty by substituting Latinitatis for orationis has not been observed outside of Virgilius and his teacher 'Aeneas'. At 107,3 he says that there are eight parts of speech: orationis partes octo sunt. This is the number recognized by Palaemon, (see Quintilian, I 4,20), and by all the Gram- matici Latini except Priscian, who omits the interjection and gives only seven (K. II 55,6 ft'.). At 42, 13 he reports that some grammarians counted the pronomen with the nomen, the participle with the verb, the conjunction with the adverb; and so made only five parts: in divisione omnium partium orationis alii octo par- tes dixere, nonnuUi eundem numerum minuentes nomen et pronomen in eandem partem redigendum esse putave- 1 For the parallel references to the Grammatici Latini, see Jeep, p. 122. ^ So his teacher 'Aeneas' said (26,10) : ergo principalis pars latinita- tis est. [16] runt, verbum quoque et participium in unum conglome- rantes absurdum dixerunt velut a corpore membrum ita a verbo separare participium. adverbia autem et coniunc- tiones unam partem esse putaverunt sequestratis praepos- sitione et interiectione atque ita erat, ut pro octo partibus quinque annumeraverint. This division into five parts was originally due to the Stoics (Servius, K. IV 428, 13; Cledo- nius, K. V 34, 23; Pompeius, K. V 135, 26; Priscian, K. II 54, 8). For a discussion of this Stoic theory, see Jeep, p. 123. Virgilius discusses the parts of speech in the order 1 nomen, 2 pronomen, 3 verbum, 4 adverbium, 5 particip- ium, 6 coniunctio, 7 praepositio, 8 interiectio. This is the order of the grammarians, with the exception of Probus (K.V 57,18), who gives the verbum the last place; and Priscian, who gives verbum the second place, participium the third, pronomen the fourth (K. II 54,8; 55i.), and omits the interjection. He remarks that the Hiberni put the verb first: (26, 4) cum in Hibonorum^ elocutione et conpossitione primatum estimatur verbum. But his teacher Aeneas insisted that the first place should be given to the nomen: (26,5) super hoc Aeneas X libros edidit, ex quibus ego unum tantum sumere oportunum puto. quassum no- men inquit secundum sensum principium est, non secundum appellationem. cum enim nascitur homo, antequam rem ali- quam agat vel discat vel sciat, nomen illi inditur. nomen ergo principalis pars latinitatis est. The nearest parallel to this statement of Aeneas is a statement in Pompeius (K. V 96, 27) : et bene primum positum est nomen. non enim potes tractare de littera, de voce, nisi prius scias, quid est hoc ipsum. idcirco quoniam nomina primigenia sunt in rebus omnibus, haec prima debent poni. nemo enim potest trac- tare primo de aliis partibus. For the nomen and verbum 1 H. A. Strong conjectures Hibernorum, Classical Review xxv (1911) p. 201. [17] as 'principales partes' seeDonatus, K. IV 372,26; Servius, K. IV 428,8; Consentius, K. V 338,6. The Nomen^ Virgilius' conception of the nomen can be gathered only from a comparison of his own treatment of it with that of the other grammarians. He himself does not define it. Donatus (K. IV 355,4) defines it as follows: nomen est pars orationis cum casu corpus aut rem proprie communiterque significans. The definition of Priscian is the same in sub- stance (K. II 56,28). Charisius (K. I 152,16) has: cum casu sine tempore. Later grammarians omit cum casu: see, for instance, Consentius (K. V 338,10), Cledonius (K. V 34, 26). Virgilius, to judge from his treatment of the nomen, would probably say: 'nomen' is a part of speech (26.3)^ having qualitas ^26.12^ comparatio ^^a.is) , genus ^=^^'^\nu- merus (32,14)^ figura ^^•^\ casus ^=^5- ^2) ^ This outline and order is found only in Donatus (K. IV 355,6), and Consentius (K.V 338,16). Charisius (K. I 153,6) does not include comparation, and Priscian (K. II 57,8) has species instead of qualitas. In Diomedes (K.I 320,27) qualitas stands last. Nomen, Qualitas^ The Grammatici Latini divide nouns into propia (Kvpi«), and appellativa (irpoo-ri^opiKd) , or, as we say, 'proper' and 'common'. ^ The terms appear first with Julius Romanus, and are used by all the Grammatici Latini. Definition as ' See Jeep, p. 124. 2 See Jeep, p. 125. 3 So, for instance, Diomedes (K. I 320,29), qualitas nominum biper- tita eat. aut enim propria sunt nomina aut appellativa: quaedam et propria sunt et appellativa. [18] such is generally clear and true to the tradition. ^ Confu- sion and incompleteness, however, in their sphere and ap- plication is initial with the Greek Grammarians, and recurs with monotonous regularity in their Roman followers. Only Pompeius (K. V 139,25; cf. Explanat. in Donat, K. IV 490, 37) seems to realize that these divisions should be ex- clusive and cover the entire field. Others, and for that matter even Pompeius himself, discuss primitives, deriva- tives, patronymics, etc., either under appellativa or as further divisions of nouns in general. Sometimes the no- men proprium is defined simply as consisting of praenomen, nomen, cognomen, agnomen, and in no case is the discus- sion of these types omitted. In his treatment of qualitas Virgilius is unusually diffi- cult and obscure. 'The qualitas of nouns, ' he says ^26.12)^ 'jg not two- fold, as some think, but manifold; it consists of many different species. Nevertheless, in their discussions of nomina pro- pria and nomina appellativa some persons insist upon as- suming that there is such a thing in Latin as a nomen propium (i. e., a noun proper and singular to and by itself). The fact is that (no Latin noun is proprium in this sense), all Latin nouns, for that matter all Latin words, are so closely interrelated that each depends upon the other and is derived or 'named' from it. Proprius, therefore, as ap- plied to nouns, must not be given its obvious and ordinary meaning; it has a certain deeper and more subtle signifi- cance which must be divined. Now proprius has two meanings; it connotes the idea of specialitas, something , ^ So Charisiua (K. I 533,12), propria sunt nomina quae specialiter proprieque dicuntur. .. .appellativa nomina sunt quae generaliter communiterque dicuntur. Diomedes (K. 1320,39), propria sunt quae propriam et circumscriptam qualitatem specialiter significant .... appellativa nomina sunt quae generaliter communiterque dicuntur. Con- sentius (K. V 338,19), appellativa enim nomina a genere et specie manant: propria at) individuis. [19] inherent, special and peculiar (the obvious and usual mean- ing discussed above), or it means superiority in position. This is shown by the fact that proprius is applied to per- sons who, because of their rank or fortune, are ascribed to the foremost position in a state; hence, 'Cicero', writing of a certain Roman named Justinus, began thus: 'lustinus autem proprius omnium sese obdedit, that is, primus om- nium'. Nomina propria, therefore, should be called nomina primaria, as Roma, Carthago. Further, nomina communia are of inferior rank, as civitas. As regards res and corpus, many are at a loss. Res is a Hebrew letter which means 'head'. ^ Res, therefore, is the equivalent of nomen primarmm. Hence, just as an army of inferior rank (cf. what is said of nomina communia above) is lead by a primarius, a person of foremost position, so all the corpus, the body, is ruled by the head {corpus, then, is the equivalent of nomen commune) . Further, nomina appellativa have many species, etc. : ^ (-•^■^-^ Qualitas nominum non bipertita, ut quidem rentur, sed multipertita; variis siquidem et multis consistit specie- bus, nonnulli tamen de propriis et appellativis nominibus quaestiones obtendere constant, quasi propriumsit aliquod nomen latinum. omnia namque nomina latina, imm© omnia fona ita invicern conexa sunt, ut quodque sicut alteri heret ita ab altero appelletur. propria ergo nomina non secun- dum simplicem sonum sonanda sunt, sed secundum subti- liorem quandam interpretationem. proprietas quippe duas res significat aut specialitatem aut propriorem ordinem: proprii enim recte dicuntur qui in primordio ciuitatum ho- nore vel dignitate vel censu scribuntur, unde et Cicero de ' See Isaac Taylor, The Alphabet, 1883, Vol. I. p. 174: 'Resh' clearly means the head, and the Hieratic form sufficiently suggests the oval of the head, supported by the neck. ' - The remainder of our author's discussion is concerned with a miscellaneous assortment of types, rnany of which, as we saw above, are regulaily associated with the topic of api)ellativa. [20] lustino quodam Romano scribens ita infiuit lustinus autem proprius omnium sese obdedit hoc est primus omnium, pro- pria ergo nomina primaria dicenda sunt ut Roma Carthago. communia autem inferioris ordinis sunt ut civitas. de re autem et corpore multi hessitant. res hebrea litera est quae interpretatur caput, res ergo hoc est quod et primarium no- men, sicut ergo a primario quolibet ducatur exercitus infe- rior, ita et capite corpus omne regitur. appellativa autem nomina multifidas species habent, etc. ' The gist of Virgilius' argument, such as it is, seems to be about as follows: The traditional two-fold division into propria and appe- llativa ('ut quidam rentur' includes, as we have seen, all the Grammatici Latini) must be rejected. This follows from the fact that proprius in its ordinary traditional sense (again 'nonnulli constant' includes all the Grammatici La- tini) cannot be applied to nouns of any sort. No Latin noun can be inherently sole and singular. On the contrary, they are all so interdependent that 'quoque ab altero appelletur. ' The true meaning of proprius in this connection, as of other technical terms used by the great scholars of old, is, of course, hidden from the ordinary eye. In this case the deeper meaning of proprius is primus, primarium, i. e. , 'foremost'. This is shown by 'Cicero's' proprius omnium in the sense of 'the first, the foremost, man of all. ' Nomina propria, therefore, are, properly speaking, nomina prima- ria, the master-nouns, the leaders and lords in the world of nomina. It follows, then, that nomina communia are of the lower orders, they are nouns of the common herd. It is unnecessary to say that despite the quotation from some contemporary 'Cicero', this 'deeper' meaning of pro- prius has no warrant in genuine Latin usage. Nor can we point with certainty to the source from which this meaning was derived. We may assume, however, that Virgilius and his teachers were well aware of the fact that nomen pro- prium was a translation of Kvpiov ovojia. It may be, therefore, that the theory of proprius here elaborated was derived [21] ultimately from the more common meaning of Kvpios — lord or master — especially frequent and familiar since Chris- tian times in the substantive use 6 Kvpios, the Master par excellence, the Lord of the world. The 'Ciceronian' ex- ample quoted by Virgilius was doubtless itself inspired by the same theory, and as such is probably a good illustration of some one of the 'twelve kinds of Latinity' to which our author alludes in 88,22, Having once set up the theory that propria nomina were primaria, it followed of course that all others were of lower rank; in other words, nomina communia are not 'common' because communiter dicta, but 'common' because they belong to the common herd. It may be remarked in passing that the reference here to no- mina communia implies that the use of nomen commune in the sense of nomen appellativum (from which our 'common' noun was derived) was already familiar to scholars. With the exception, however, of Serg. Explanat. in Donat. (K. IV 490, 15), proprium est quod unius est, commune quod multorum est — but again a few lines below, appellativum est quod multorum est (cf. Probus, K. IV 119, 26)— the word is never so used by the Grammatici Latini. In Pris- cian (K. II 156, 10 and III 472, 20) nomina communia are nouns having a 'common' gender. Having disposed of nomina propria and communia, Vir- gilius takes up the matter of 'abstracts' and 'concretes' ; for this is what is meant by the discussion at this point of res and corpus. It is also practically certain that the ulti- mate source of the discussion was Donatus, The Gram- matici Latini divide appellativa into corporalia ( 'concre- tes'), and incorporalia ('abstracts'). Donatus alone in- cludes all nouns in this division, and designates abstracts by res, concretes by corpus (K. IV 373,1), nomen est pars orationis cum casu corpus aut rem proprie communiterve significans. His model v/as evidently Dionysius Thrax (Grammatici Graeci, P, ed. Uhlig, Leipzig, 1883, p. 24, ovojxa €