Hostname: page-component-76d6cb85b7-92wsb Total loading time: 0 Render date: 2026-07-13T20:06:33.099Z Has data issue: false hasContentIssue false
  • English
  • Français

Dialectal variation in republican Latin, with special reference to Praenestine

Published online by Cambridge University Press:  28 February 2013

Robert Coleman
Robert Coleman
Affiliation:
Emmanuel College, Cambridge
Get access Rights & Permissions [Opens in a new window]

Extract

1.1 Dialectical diversity is common to all languages. It comes about when groups of speakers of the same language become isolated from one another and from the conformist pressures of an administrative or cultural centre. The tendency to local change can be accelerated by the presence nearby of another language and the growth of bilingual interaction in settled conditions over several generations. Like linguistic change in general, dialectal diversity tends to be retarded by improved communications and a universal homogeneous education system.

1.2 Modern dialect geographers go about their work with a questionnaire, a tape-recorder and a note-book, listing items of pronunciation, vocabulary, morphology and syntax that distinguish particular localities from one another. If their aim is not the purely linguistic one of establishing an underlying diasystem and the range of variation within it, but is more concerned with the sociological aspects of the linguistic data, then they will plot the dialectical variations against the ‘standard’ language, viz. the dialect which because of its association with an administrative or cultural capital has in an elaborated and artificial form acquired prestige and dominance throughout the whole speech community. If researchers find that their information is incomplete, they can simply return to the field with further questions.

Information

Type
Research Article
Information
Proceedings of the Cambridge Philological Society (Second Series) , Volume 36 , 1990 , pp. 1 - 25
Copyright
Copyright © The Author(s). Published online by Cambridge University Press 1990

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.)

Article purchase

Temporarily unavailable

References

NOTES

1. See §II.A.3 below. The two categories are conflated in Ernout, A., Les éléments dialectaux du vocabulaire latin (1909)Google Scholar. His extensive catalogue and full discussion, though dated now, are still indispensible and have been constantly consulted for the present study.

2. See Coleman, R., ‘The central Italic languages in the period of Roman expansion.’ TPS (1986) 100–31, esp. 111–14Google Scholar.

3. See Goidanich, P. G., ‘Varietà etniche e idiomatiche in Roma’, Atti del Primo Congresso Nazionale di Studi Romani (1929) 2.396–414, esp. 398–400Google Scholar. The precise location of Italic forms cited in this paper can be found in the indexes to Vetter, E., Handbuch der Italischen Dialekte (1953)Google Scholar and Poccetti, P., Nuovi Documenti Italici (1979)Google Scholar. References to CIL are merely by number, e.g. 12155, 5.2960); those to ILLR are given in full, those to Vetter by V and the number.

4. Their cases suggest that they were preserved in a context, but with the original spellings, e.g. *denguam, *dakrumeis, classicised.

5. For AŘFERTUR (Lat. *adfertor) cf. Lat. arfuere, showing loss of occlusion as the midway stage in the assimilation of adf- to aff-;cf. adw- > arw-in aruorsum on the same inscription (12581, from 186 BC). peres for pedes, cited by Consentius, 5.392.15K, as a recent aberration of the plebs Romana, is too late to be relevant.

6. Perhaps influenced by lingere: cf. Arm. lezu ‘tongue’ and lizanem ‘I lick’.

7. For the present purpose it is of no matter whether lacrima was cognate with or borrowed from the Greek, as Leumann, M. thought (Lateínische Laut- und Formenlehre (1963) 128)Google Scholar.

8. The equation of lympha with Osc. DIUMPAÍS is too problematic to merit inclusion.

9. Probably the Sabine language, of which admittedly we know very little, not, as R. S. Conway thought (IF 2.157ff.) the Sabine dialect of Latin, which may however have been affected by the phenomenon; cf: §3.1.

10. The Faliscan family name Hirmia, Firmia (V. 269c, 338a) cannot be evaluated.

11. The reverse process occurred with e.g. bos, lupus and perhaps leuir (§1.3). See §3.3.

12. We know tantalizingly little of the Volscian language; but if it shared gen.sg. -osio with Faliscan (see Coleman (n.2) 120–1), then it may have shared some of its phonology as well.

13. Cf. the lists in Conway, R. S., The Italic dialects (1897) vol. 1Google Scholar, who appends them to the relevant inscriptions, and Vetter (n.3) 362–78. See also §I.2.3 and n.9.

14. Both the root and its suffix, however, show great variation in IE: > OE wolfs beside kwo- > Ved. , *lu-kwo > Gk lúkos, etc. So p may not after all reflect *kw.

15. Casa and miser, without secure etymologies, cannot be placed.

16. Augustus on the other hand preferred the older spellings, at any rate in mancupium, manubiae (Vel. Long. 7.67).

17. Cf. 42: illud quod loquitur priscum uisum iri putat, si plane fuerit rusticanum; also Br. 137 and 259; Quint. 11.3.10.

18. trebibos (12398) is uncertain as to date and provenance, though praifectos alongside it suggests the 3C, which would make it an isolated precursor, aidiles and Tempestatebus, also 3C (128, 9), are false archaisms in forms where -is and -ibus were original. Although celeber, integer do not show weakening to *celiber, *intiger, no Latin dialect is known to have been immune to the change, as apparently Oscan and Umbrian were, and the retention of e is probably analogical with celebrem, integrum etc. See Ernout(n.l) 38–9.

19. Fr. voisin, Sp. vecino indicate however that uēcinus and uēcus must have become widespread in Vulgar Latin, ę represents a higher, closer articulation of the vowel, as in Fr. fée, ę a lower, more open one, as in Fr. fait.

20. Praenestine ministrei, Praenestinei (121446) are less conclusive. Archaic spellings with -ei survive even in Roman till the end of the Republic. For full lists of -e forms and their distribution see Blümel, W., Untersuchungen zu Lautsystem und Morphologie des Vorklassischen Lateins (1972)Google Scholar.

21. In dialects where ei > ī, the e is more likely to be short and due to lowering of the -ǐ that had been produced by iambic shortening.

22. Cf. Asconius's otherwise unmotivated adjective in the phrase Liuius noster (Comm. Cic. Corn. 68).

23. So Whatmough, J., HSCP (1933) 95f.Google Scholar

24. Too early to be cited, as it often is, for dialectally significant o against Roman luna.

25. As nōnus (< *nounus < neun- cf. dat.sg. neuna on ILLR 11: Lavinium) is due to nouem.

26. The length is of course not specified. From the MS variants domos … pro domuos and domos … pro domus Ritschl and others have favoured domuos … pro domus, which would be archaic for Rome but might well have been current still at Vellitrae; cf. n. 28.

27. parenteis occurs at Rome (121214) with erodita and infista, all perhaps pointing to an Oscan native speaker.

28. It is no coincidence that the -os alternant in fourth-declension forms, also inherited, is attested with the word senatus: zenatuo (12365: Falerii) and senatuos on official inscriptions from Lucania (I2581) and Aquileia (I22197). For a full list of third-declension -us forms see Blümel (n.20) 63–4.

29. See Lazzeroni, R., ‘Il dativo sabellico’, SSL 4 (1964) 65–86Google Scholar; Coleman (n.2) 106.

30. Even more if one follows Blümel (n.20) 57 in thinking that the Latin dat. -ō reflects the original instrumental -ō; not in itself a very plausible hypothesis, given the inherited status of the dat.sg. -ōi and the general pattern of abl.-instr. syncretism in Italic.

31. Where the very few instances of -ā, like Mijnerua and Flaca (12460, 477) may be attributed to immigrant influence along with the remarkable dat. pl. Deuas Corniscas (12975).

32. Which may originally have been Faliscan-speaking; we know nothing of the extent of this language beyond Old Falerii and its immediate environs.

33. Nominatives like Asiane (6.17408) are by analogy with the Gk nom. Eutychē beside gen. Eutychēs.

34. Gen.sg. VALAIMAIS twice (V6: Capua) may be a significant lapse in the Oscan text. For the Greek source see especially Väänänen, V., Le Latin vulgaire des inscriptions pompeiennes (1966) 83Google Scholar.

35. Cf. Osc. SAKRAFÍR, Umbr. pihafi: Lat. *sacraueritur, *piaueritur (sacratum sit, piatum sit).

36. Assuming that the verb is fundare; see §4.4. From fundere the relevant forms would be fundito (d), fundat, fuderit, *fusit.

37. -ād was presumably the original form of the Latin subjunctive also, but it must have been levelled out long before vowel weakening, which apart from -iec- this inscription shows in full. Degrassi's dating (ILLR 504) paulo post a.315 uel 314 is too early, perhaps by as much as a century.

38. proiecitatid, suggested by Ribezzo, (RIGI 6 (1922) 311Google Scholar) requires a less probable verb suffix; proiēcatid, proposed by Pisani, (Testi Latini Arcaici e Volgari (1950) 18Google Scholar) requires an impossible combination of lengthened root vowel with ā- suffix.

39. First noted by Brugmann, in Miscell. Ascoli (1901) 1ff.Google Scholar

40. There is also [in] ium = in eum (cf. Osc. ionc) later in the text, if this and not [ceiu] ium is the correct restoration.

41. Lucilius' objection to Vettius Philocomus' sermo Praenestinus seems from the context of Quintilian's citation (1.5.56) to refer to vocabulary. See §111.1 and Ramage, E. S., ‘Early Roman urbanity’, AJP 81 (1960) 65–72, esp. 71Google Scholar.

42. The relationship here may be phonological, if Praenestine spellings like Dcumius, Ptronio and Gemna, Numtoriai are not syllabic but represent syncope.

43. Already attested in uinias (121853: Amiternum region), aenia (4.64: Pompeii).

44. propiteas, Septimea etc. at Pompeii probably represent the convergence of ē and ĭ, on which see §E.1.

45. Some caution is needed here and in other instances of confusion between e and i, given the possibility of error resulting from the use of II for E in cursive script.

46. Which could however be analogical: nom. -īs: ace. -īs (in i-stems) ← nom. ēs: acc. ēs (in cons, stems).

47. See in general Sturtevant, E. H., The pronunciation of Greek and Latin ed. 2 (1940) 131–3Google Scholar.

48. For the two e-vowels seen n. 19.

49. The testimony has been somewhat obscured by manuscript corruption: cod. F has faedus; hedus; aedus. The text printed is Kent's; Collart prefers fedus, edits, aedus. For occasional loss of h- in dialects see §4.2.

50. See Coleman (n.2) 114–17.

51. gen.sg. moltai (ibid.) = moltdi; cf. Umbr. gen.sg. MUTAR ( < *moltās).

52. The authenticity of the diphthong is guaranteed by etymology in laeuir (see §A.3.3), and in faenus by the fact that, though Cato preferred fenus (Gell. 16.12.7), the connection with fetus assumed by Varro and others (ibid.) would have protected an original ē. Quite distinct are the Greek loanwords scaeplrum, scaena, in which the variation against Gk ē may be due to Etruscan intermediaries.

53. For the details see Coleman, R., ‘The monophthongization of ae and the Vulgar Latin vowel system’, TPS (1971), 175–91Google Scholar; ‘The monophthongization of ae: a reply’, TPS (1974) 86–92Google Scholar.

54. For other examples, some of which allow other explanations, see Coleman (n.53) 190.

55. The qualitative distinction between and is exemplified in the vowels of Fr. bonne and Beaune.

56. 4.2360. There are no examples at Pompeii of au for . See V. Väänänen (n. 34) 30–2.

57. plostrum on the Lex lulia (1.206.57) indicates that this form was not confined to the Sabine dialect.

58. Monophthongisation occurs subsequently in French and Spanish; e.g. chose ( < *chausa) and cosa (beside Portug. cousa). Gmn kaufen indicates caupo not copo.

59. Cf. Diomed. 1.382K: frequens est apud ueteres ‘plodo’.

60. See Sturtevant (n.47) 130–2.

61. Cornutus (7.153K) reports Cicero's use of uemens for uehemens.

62. Incidentally, Etruscan has h regularly in initial position but rarely elsewhere.

63. Cf. Ahenobarbus but no *aherius, where ae is also dissyllabic.

64. See further Allen, W. S., Vox Latina ed. 2 (1978) 30–1CrossRefGoogle Scholar.

65. The final E in Cato's diee for diem (Quint. 9.4.39) may in fact have been Σ, a tilted M. On one possible interpretation of Velius Longus' account of -m (7.54K: peruenisse fere ad aures peregrinam litteram inuenies) it could have been a reversed digamma ; cf. Claudius's for consonantal u (Priscian 1.18 = 2.15K). We cannot tell what Verrius Flaccus' ‘partial’ m looked like or even whether the reason that Velius gives for it, ut appareret exprimi non debere (7.80K), is reported or merely conjectured.

66. Whether dant > *dan lies behind danunt at Sora (121531) is doubtful. The form occurs in Plautus (Most. 129) and other early poets and may reflect a verb *də3-n°/e- beside the noun *deə3-n°/e- (> dōnum).

67. JHS 69 (1950) 2Google Scholar.

68. See Conway (n.13) 1.321.

69. Most of which are printed in Ernout, A., Le Parler de Préneste d'après les inscriptions (1905)Google Scholar, which is still basic for the study of this dialect.

70. Cf. Plaut., Trin. 609Google Scholar, True. 691; Quint. 1.5.56 and Festus (see §III. C.5.1 above).

71. Dissentient voices were heard now and again; e.g. that of Pisani, V. in RIGI 16 (1932) 93Google Scholar and AGI 34 (1942) 98Google Scholar.

72. But not exclusively. The opening salvoes were fired by Gordon, A. E., The inscribed Fibula Praenestina, problems of authenticity (Univ. Cal. PCS 16 (1975)Google Scholar) and Ridgway, D., ‘Manios Faked?’ BICS 24 (1977) 17–30Google Scholar.

73. See Guarducci, M., ‘La cosiddetta Fibula Prenestina’, Mem. Linc. ser. 8.24.4 (1980) 413–574Google Scholar; ‘La cosiddetta Fibula Prenestina: elementi nuovi’, Mem. Linc. ser. 8.28.2 (1984)Google Scholar. Gordon, A. E., Illustrated introduction to Latin epigraphy (1983)Google Scholar, includes the text as ‘a salutary exemplum’.

74. The fibula was first displayed by Helbig at a public meeting of the German Archaeological Institute on 7 January 1887 and published by him in Wochenschrift f.Kl.Phil. 4 (January 26, 1887)Google Scholar and by him and Dümmler, F. in Röm. Mitt. 2 (1887) 37–43Google Scholar.

75. Guarducci (n.73) (1980) 543–561.

76. Guarducci (n.73) (1980) 557, 558. The stylometric assessments, being less objective, belong in the latter category also.

77. Guarducci (n.73) (1980) 554.

78. The possibility was suggested by Gordon's remark (n.73 (1975) 30) that ‘we cannot have a bonafide ancient inscription on a modern fibula’.

79. For photographs see Guarducci (n.73) (1980) tav. 1, 2, 5–11; Gordon (n.73) (1983) pl. 1.

80. This purely practical consideration accounts for the word order, and there is no justification for regarding this as a precursor to the Romance word-order patterns with verb in non-final position.

81. First published respectively in Not. Scav. (1899) 151f.Google Scholar, Gamurrini, F., Mon. Ant. 4 (1895) 339Google Scholar, some years after the fibula appeared.

82. Guarducci (n.73) (1980) 447, 534–5. A supplementary but not wholly compatible accusation, for which an expert graphologist is cited (Guarducci (n.73) 1984) is that the letter-forms closely resemble those found in Helbig's manuscript transcriptions of epigraphic texts. This is truly remarkable, given the differences in the instruments and writing techniques employed.

83. See Campanile, E., ‘Riflessioni sui più antichi testi epigrafici latini’, ΑΙΩΝ 7 (1985) 89–99, esp. 90f.Google Scholar, for doubts about the need to equate forgery of an inscribed object with linguistic inauthenticity of its text, and misgivings about the security of some of Guarducci's other more specific conclusions.

84. 6.11–12K. populoi in his text is impossible and we must read poploi or less likely *popeloi or *popoloi.

85. Already published in CI It (1880) app. 934; so a canny forger might have played safe with Numisioi.

86. fepacid, presumably for fefacid (pf. subj.), also occurs on the Tabula Bantina.

87. For the contrary view, that fefac- was an independent innovation in Latin and Oscan, see Hamp, E. P., AJP (1981) 150–3Google Scholar.

88. Fabretti, A., CIIt (1867) 455Google Scholar, proposed feci < *fefici in his discussion of the Oscan Tabula Bantina, which he was publishing for the first time.

89. The Duenos inscription (124: ? 5C) was first published by Dressel, H. in Ann. dell. Inst. di Corr. Arch. (1880) 158ff.Google Scholar

90. The Etruscan inscription with Fhalmus (read as Fhelmks) was published by Von Duhn, , Bull. Inst. Arch. Rom. (1879) 157Google Scholar. The corresponding digraph in Faliscan was explained by Deecke, W., Die Falisker (1888) 276Google Scholar, and in Venetic by Pauli, C., Die Veneter und ihre Schriftdenkmäler (1891) 99Google Scholar.