Hostname: page-component-6b989bf9dc-jks4b Total loading time: 0 Render date: 2024-04-14T13:52:09.787Z Has data issue: false hasContentIssue false
  • English
  • Français

O egregie grammatice: the vocative problems of Latin words ending in -ius

Published online by Cambridge University Press:  11 February 2009

Eleanor Dickey
Eleanor Dickey
Affiliation:
Columbia University, ed202@columbia.edu
Get access Rights & Permissions [Opens in a new window]

Extract

A long-lasting and sometimes acrimonious debate over the correct vocative form of second-declension Latin words in -ius began more than 800 years ago. For the past century most classicists have considered the matter to be settled, and little discussion on the subject has taken place. Yet the century-old conclusions we now so unthinkingly accept are based on very little evidence and are internally inconsistent in some of their details. The past hundred years have provided us not only with more Latin to work with, better tools for search and analysis, and a more complete knowledge of the history of the Latin language, but also with a new understanding and respect for the ancient grammarians and their views on the structure of their language. It is time to re-examine the ancient and modern views on the vocative of -ius words, to see whether any viable conclusions can be drawn and whether the ancient grammarians may have more to contribute than our predecessors believed.

Type
Research Article
Information
The Classical Quarterly , Volume 50 , Issue 2 , December 2000 , pp. 548 - 562
Copyright
Copyright © The Classical Association 2000

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 This article grew out of a paper given at the Princeton Colloquium on Latin Linguistics in April 1999. I am grateful to Philomen Probert, Joshua Katz, Michael Weiss, Karla Pollman, Glen Bowersock, Steven Pinker, John Penney, Alan Nussbaum, David Langslow, Anna Morpurgo Davies, and Leofranc Holford-Strevens for their encouragement, corrections, information, and advice (though they are not responsible for any errors which remain), and to the Institute for Advanced Study for providing ideal working conditions.

2 For example, Tati: Enn. Ann. 104 (Skutsch).

3 Goetz, G. (ed.), Corpus Glossariorum Latinorum (Amsterdam, 1965), 647–8.Google Scholar

4 For example, Marce fili, Cato, fr. 1 (p. 77 Jordan).

5 In approximate chronological order by first appearance of each word. These lists should be reasonably complete for Latin literature until c. a.d. 200, but no claim for completeness is made with regard to later material or to inscriptional evidence.

6 Both Darie and Dari appear in manuscripts, and the latter is found in many editions, but the most accurate critical edition (Biblia Sacra Iuxta Latinam Vulgatam Versionem ad Codicum Fidem, vol. 16 [Vatican, 1981]) takes Darie as the more authentic reading.

7 Neue, F. and Wagener, C., Formenlehre der lateinischen Sprache (Leipzig, 1892–1902), vol. 1, 127–33Google Scholar; vol. 2,42–4.

8 For example,Kühner, R. and Holzweissig, F., Ausführliche Grammatik der lateinischen Sprache, vol. 1 (Hannover, 1912 2), 446–7Google Scholar; Gildersleeve, B. L. and Lodge, G., Gildersleeve's Latin Grammar (Boston, 1895 3), 16Google Scholar, 37; Greenough, J. B. et al., Allen and Greenough's New Latin Grammar (Boston, 1931), 22Google Scholar; Holford-Strevens, L., ‘More notes on Aulus Gellius’, Liverpool Classical Monthly 9 (1984), 149–50.Google Scholar Note that Wackernagel, J., Über einige antike Anredeformen (Göttingen, 1912), 17Google Scholar, appears to endorse this view but in fact introduces important qualifications of it; Wackernagel's formulation of the rule for the vocatives of -ius words, which he mentions only in passing, is actually closer to the one presented in this paper than to that given in Neue and Wagener.

9 Cf. A. Mau, index to CIL 4, p. 761. Interpretation of the graffito containing this word is difficult if one takes noxsi as a vocative, but even more difficult if one does not.

10 L. Håkanson (Stuttgart, 1982); G. Lehnert (Leipzig, 1905).

11 See Leumann, M., Lateinische Laut- und Formenlehre, vol. 1 (Munich, 1977), 424–5Google Scholar; cf. also Lachmann, C., In T. LucreHi Cari De Rerum Natura Libros Commentarius (Berlin, 1882 4), 325–9.Google Scholar

12 For example, Orpheu (Vergil, G. 4.494), Theseu (Catullus 64.133).

13 For example, Pari (Ovid, Ep. 16.83), Iri (Vergil, A. 9.18).

14 For example, Neue and Wagener (n. 7), vol. 1, 127–8.

15 Penee (Ovid, Am. 3.6.31), Alphee (Stat. Theb. 4.239).

16 Skutsch, O., The Annals of Q Ennius (Oxford, 1985), 210–11.Google Scholar The vocative δῖε is not uncommon in Greek poetry, eg. Iliad 8.185, 11.608, 12.343, 14.3, 24.618.

17 Ovid, Am. 2.6.3, Tr. 5.3.47; Statius, Silv. 5.3.284.

18 Throughout this paper, statements about the non-occurrence of words are based on the following evidence: (i) A corpus of 15,319 vocatives collected by hand from literary and non-literary sources from the Republican period and the first two centuries of the Empire. Most works in Latin literature before a.d. 200 are included in this corpus, but not all. (ii) Where practical, electronic searches of the PHI Latin database using the Pandora program. This method could not be used in the case of vocatives having the same form as very common adverbs, genitives, or other common words (e.g. proprie, Jili, die), but it has been used in all other cases. Most words have also been checked on the Patrologia Latina database, though this fact is not usually relevant for claims of non-occurrence, (iii) Perusal of the standard reference works, including where available the Thesaurus Linguae Latinae.

19 Enn., Ann. 4 (Skutsch) and Skutsch (n. 16), 157.

20 Propertius 2.9.20,2.17.3; Ovid, Met. 10.345, 14.736; Seneca, Ag. 953; Lucan 5.158, 9.71.

21 Seneca, Ag. 234, Med. 568, Her. F. 309,900, Phaed. 864, Her. O. 880.

22 I have found twenty-nine examples, e.g. Lucilius 1323; Vergil, A. 1.198; Horace, Carm. 1.25; Livy 21.21.3; Valerius Maximus 3.1.2; Luean 2.483.

23 Ovid, Met. 11.585; Valerius Flaccus 1.794.

24 Ovid, Met. 7.21, 13.483, 13.523, Pont. 2.9.1; Seneca, Ag. 341.

25 Vergil, A. 7.212; Statius, Theb. 10.240.

26 Conscia: Propertius 1.12.2; Ovid, Ep. 7.191; Martial 6.10.9;temeraria: Ovid, Met. 1.514; Lucan 8.579, 8.795; Statius, Theb. 12.366; noxia: Lucan 8.823; seria: [Tibullus] 3.6.52; Fronto, p. 9.15 van den Hout; anxia: Valerius Flaccus 2.113; funebria: Ovid, Am. 1.12.7; hostilia: Statius, Theb. 12.256.

27 Statistics taken from the glossary of E. Dickey, Latin Forms of Address (forthcoming).

28 Cf.Leumann, M., Lateinische Laut- und Formenlehre, vol. 1 (Munich, 1977), 424Google Scholar; Sommer, F., Handbuch der lateinischen Laut- und Formenlehre (Heidelberg, 1948), 343.Google Scholar An early date for this change was often argued from the antepenultimate accentuation of vocative forms like Vdleri, though this evidence is difficult to evaluate; see Gellius (13.26); Priscian (Grammatici Latini, ed. Keil, II 302.13–18); Servius (on Aeneid 1.451); Leumann (above), 425;Bernardi Perini, G., L'accento latino (Bologna, 1964), 43–5Google Scholar; Sommer (above), 344.

29 Michael Weiss, Alan Nussbaum, and John Penney, personal communications; for the antecedents of this idea, cf. Sommer (n. 28), 100.

30 Cf.Bybee, J., Morphology (Amsterdam, 1985), 119–21.CrossRefGoogle Scholar

31 Cf.Pinker, S., Words and Rules: The Ingredients of Language (New York, 1999), 125–31.Google Scholar

32 For precise statistics, see Thesaurus Linguae Latinae, s.v. filius, 752–3.

33 Catullus 33.2, 33.8, 37.18; Pomponius (p. 108 Blansdorf; this reading is perhaps to be emended to filia; cf.Courtney, E., The Fragmentary Latin Poets [Oxford, 1993], 109)Google Scholar; Phaedrus, Fabulae Novae 5.5; Terentianus Maurus, De Syllabis 283 (p. 52 Beck). There are some more exceptions in the late empire, e.g. Commodianus Apol. (=Duob. Pop.) 379, 449, Instructiones 2.17.1 (Martin).

34 Plautus, Men. 822, 844; Propertius 4.11.67; Horace, Carm. 1.16.1; Ovid, Met. 1.481,10.467, Fasti 4.456, 4.483.

35 For the expected ratio of males to females in this type of address, compare the figures on (g)nate and (g)nata in poetry, at 179 to 46, or fili and filia in prose, at 61 to 9.

36 See my article ‘O dee ree pie: th e vocative problems o f Latin words ending in -eus’, Glotta (forthcoming). One should note that the two cases of avoidance are not identical, since in comedy fili seems to be avoided while mi is common; in other types of poetry, however, both are equally avoided.

37 Carm. 1.2.43 and perhaps Sat. 1.6.38; for a different interpretation, see Svennung, J., Anredeformen (Uppsala, 1958), 269.Google Scholar

38 Additional motivations for Andronicus' use of filie can also be found. J. H. W. Penney suggests (personal communication) that trisyllabic forms of filius and filia prefer the position after the caesura of a Saturnian (cf. Andronicus frs. 12,13,19,21 Blänsdorf), providing formulaic pressure for filie. L. Holford-Strevens suggests (personal communication) that ‘Livius Andronicus was not a native speaker, hence (like the early Irish Latin writers and their Anglo-Saxon pupils) not inhibited by a sense of what is simply not done; moreover he was constructing from scratch a Latin epic diction that should stand to the Latin he heard spoken as Greek epic diction did to the Greek he spoke. We may therefore expect not only archaisms, or borrowings from the Italic dialects, but straightforward inventions.’

39 In opposition to the above it could be noted that filius is also attested at Pompeii as a nominative for vocative (CIL 4.5213), and that a number of other -ius words also occur there in the nominative for vocative construction: Διονυσιος (3885), Pumidius (4338), Latimius (4844), (H)[e]rmorius (10174). It is sometimes implied that these words occur in the nominative at Pompeii because of their problematic -ius ending (cf. Svennung [n. 37], 271, 278), but this is not the case. The use of nominatives for vocatives is a general feature of the language of the Pompeiian graffiti (cf. V Väänänen, Le latin vulgaire des inscriptions pompéiennes [Helsinki, 1937], 195, 217) and is in fact less frequent with words in -ius than with other words in -its: when used as addresses, the former appear in the nominative 6 per cent of the time (total 82 examples) and the latter 8 per cent of the time (total 172 examples).

40 Holford-Strevens, L., Aulus Gellius (London, 1988), 131Google Scholar, and id., ‘More notes on Aulus Gellius’, Liverpool Classical Monthly 9 (1984), 149–50.

41 For example, Keil IV 132.1, 11, 17, 24, 31, 133.22, 38. The ‘vocative’ forms are the same as the nominatives.

42 Keil III 22.13–23.10, 166.3–8, 207.14–19.

43 David Langslow, personal communication; see also e.g.Löfstedt, B., Studien über die Sprache der langobardischen Gesetze (Stockhom, 1961), 37.Google Scholar

44 Barwick 23.16–24, 203.2–5 = Keil 124.26–32, 159.19–21.

45 This phrase is in fact an exclamation, not an address, and therefore it does not illustrate the nominative for vocative construction, but rather the nominative of exclamation. The point the grammarians were making is, however, unaffected by their choice of examples.

46 Keil IV 9.1–3, 19.6–10, 105.15, 194.35–7; cf. 4.14–18.

47 Keil II 300.17–305.21, HI 447.24–448.4, 487.7–19, 511.20–512.3; cf. also III 205.20–3.

48 Keil IV 496.34–497.7,498.17–20, VIII 104.15–34.

49 Keil V 556.7–16, VI 474.1–4, VII 541.15–18.

50 For example, Keil VIII 104.15–34.

51 Barwick 55.19,99.15–17 = Keil 145.15,79.3–5.

52 Cf, also in connection with the genitive, Barwick 21.22–4, = Keil 123.19–20.