Published online by Cambridge University Press: 22 April 2026
[In Venetic inscription no. 249, Ethnographic Museum of Vienna, the r in the second line goes with tola in the first and tolar is a verb, third singular present middle, and not a noun. It belongs with OIr. canar and has an indicative and personal meaning ‘he dedicates’. canar means in the Old Irish glosses ‘is sung’, not ‘there shall be singing’. The comparison of Old Irish, Latin, Oscan (fufans), Venetic, Greek (ἔτλᾱν), Lithuanian, Old Ch. Slav., and perhaps Tocharian, indicates that this ā-formation in PIE was essentially indicative. Its ‘injunctive’ value is simply a modal use of the indicative.]
1 The Prae-Italic Dialects of Italy, R. S. Conway, S. E. Johnson, J. What-mough, 1. 5 (Cambridge, Mass., 1933).
2 On the position of Venetic see P. Kretschmer, Einleitung in die Geschichte der griechischen Sprache 271 (Göttingen 1896); F. Sommer, IF 42.132.
3 Prae-Italic Dialects 1.190.
4 Ibid. 1.170.
5 Carl Pauli, Die Veneter und ihre Schriftdenkmäler, volume 3 of Altitalische Forschungen (Leipzig 1891).
6 Conway's reference (134) for the gens Vennonia should be Part II vii C. Cf. Pauli, Veneter 310; Thurneysen, WklPh. 1892.289.
7 Cf. Sommer, l. c.: ‘Was das Wort bedeutet, wissen wir nicht; ebensowenig, was in dem voraufgehenden tola·r· steckt.‘
8 Pauli, 326 f.; cf. Sommer's criticism, IF 42.116–7.
9 Prae-Italic Dialects 1.191 ff.
10 See Pauli 208, and the examples there given.
11 A good Greek parallel in IG 1.410.
12 [The author here acknowledged certain suggestions of mine. As I have found it necessary to revise them, I shall reserve their publication for another issue. GMB]
13 Cf. Conway, Outline of Venetic Grammar, Prae-Italic Dialects 1.185.
14 254; cf. 273–4 and Sommer 116–7.
15 See Conway, op. cit., 1.190; Pauli 404; Thurneysen, WklPh 1892.289.
16 Op. cit. 1.191–7. Cf. Whatmough, ClPh. 29.283 (1934). The argument of this paper is, however, in no way affected if one does not accept Conway's theory of the accentual value of the puncts, provided that they are not interpreted in such a way as to separate the r from tola.
17 Cf. Conway, op. cit. 1.197–9.
18 Meillet, BSL 19.97, says that ‘*tlā est la forme à voyelle longue finale de la racine dissyllabique *telə.‘ It is—in Dor. ἔτλᾱν; but that is not necessarily the only form of the second full grade of the base. The reduced vowel also occurs in the first syllable of such bases (Hirt's Vollstufe II a, cf. Hirt, Indog. Gram. 2.121, 113; Ablaut 76, cf. 85).
19 See E. Kieckers, Handbuch der vergleichenden gotischen Grammatik 240 (Munich 1928).
20 Cf. Walde, O-farbige Reduktionsvokale, Festschrift Streitberg 175–87 (Heidelberg 1924). Miss Ruth E. Moore of Montreal has very kindly communicated to me by letter a number of interesting examples in Latin and Greek (from a research that she has made in the reduced vowels) which she considers illustrations of or, ol developing out of the dull reduced vowel of o-color. E. g. Lat. morior, mortuus: Gk. μορτός · ἄνθρωπος, θνητός Hesych., βροτός; Lat. orbus: Gk. όρφανός, όρφ∊ύς, όρφός; Lat. orior: Skt. ṛṇóti. It is of course extremely difficult, as both Walde and Miss Moore recognize, to determine whether in a given instance or, ol represent the reduced or the normal deflected grade in PIE.
21 The u of Lat. tulam itself requires explaining, since el normally gave al in Latin. Cf. Kent, Sounds of Latin §§80, 94. For varying views as to the origin of the u of tulam, see Brugmann, Grdr.2 I.1.455–6; Meillet, BSL 19.97; cf. Reichelt, KZ 39.56; Hirt, Handbuch des Urgermanischen 1.59 (Heidelberg 1931). Meillet remarks: 'on ne recherchera pas ici si l'u de tulam est un ancien u, développé devant l, comme dans gula, cf. arm. ekul “il a avalé”, ou si, ce qui est peut-être plus probable, tulam est tiré de formes à préverbe comme attulam.' This second view, which Meillet considers plus probable, would seem to imply that the earlier form of the Latin word, without prefix, was tolam, which would provide a perfect parallel for Venetic tolar. Cf. de Saussure, Mémoire 100: 'En latin la voyelle obscurcie en u pourra généralement passer pour o.'
22 Cf. Walde-Pokorny, Vergl. Wörterb. der indog. Sprachen, s.v. tel-; Solmsen, KZ 38.445; de Saussure, Mémoire 239–42; Hirt, Indog. Gram. 2.109; Sommer, Hdb. der lat. Laut- und Formenlehre2,3 231.
23 Cf. Hirt, Indog. Gram. 2.192; Carnoy, Manuel de linguistique grecque 164 (Louvain 1924).
24 See E. Fraenkel, IF 50.226 (1932); Schulze, Sieg, Siegling, Tocharische Grammatik 448 (Göttingen 1931). Cf. Osthoff, Vom Suppletivwesen 4, 6.
25 In no. 166, with which no. 169 according to Conway ad loc. is to be classed, the verb zonasto, commonly used in dedicatory inscriptions, is preserved.
26 See my Syntax of the Boeotian Dialect Inscriptions 68.
27 Ibid 69.
28 See Conway, op. cit. 1.197, §50; and 194, §41.
29 Cf. Carnoy, op. cit. 165. Meillet (BSL 19.95 ff.) is inclined to consider this ā-formation as autonomous, but not necessarily aorist in origin. The Venetic form would tend to confirm its aoristic character, as recognized by most scholars (see, e.g., Stolz-Leumann, Lateinische grammatik 325 and the references there given). Cf. Vendryes, Sur quelques présents en -Ā- du verbe italo-celtique, MSL 16.305: ‘On aboutirait donc à reconnaître en latin trois groupes de présents en -ā- issus d'anciens aoristes.‘
30 For the probably votive character of these bronzes, see Conway, op. cit. 1.171–2, cf. 173.
31 See Conway's Resultant Text, op. cit. 1.6–18 passim.
32 Sommer, also, IF 42.126, takes vhaχ · s · θo as ‘eine Meistersignatur’ and transcribes faχsθo = fak-s-to, though without commenting on the meaning of the middle ending.
33 See G. Dumézil, Études comparatives sur les langues caucasiennes du nord-ouest 105, 117–8 (Paris 1932), but cf. Id., Introduction à la grammaire comparée des langues caucasiennes du nord 32–7 (in Bibliothèque de l' Institut français de Léningrad, Tome XIV, Paris 1933); A. Dirr, Einführung in das Studium der kaukasischen Sprachen 39, 342–3 (Leipzig 1928).
34 W. Petersen, AJP 53.209 (1932); E. H. Sturtevant, Comparative Grammar of the Hittite Language 262 (1933).
35 Thesaurus Palaeohibernicus, edited by Whitley Stokes and John Strachan, 2.252, 1.5; cf. ibid. 252, 11.12, 14, 19, (Cambridge, 1903).
36 Ibid. 253, 1.3 and 1.5; compare the use of arcanar in Glosses on the Carlsruhe Beda, Thesaur. Palaeohib. 2.18, 1.33, where it glosses Latin cantatur.
37 Cf. ‘Vermutlich’ 664 ad fin. . . . ‘Es mag hiernach’ . . . ‘vielleicht also’ 666. Brugmann here seems to be reproducing the argument of Thurneysen, KZ 37.104. On page 705 also Brugmann renders Ir. canar by ‘canatur’, but this too is purely a translation ad argumentum, since, as we have seen, the Irish themselves use canar to gloss Lat. canitur and cantatur, both indicatives.
38 The Hypothesis of the Italo-Celtic Impersonal Passive in -r, Language 5.232–50 (1929).
39 Cf. R. Thurneysen, KZ 37.104: Berar fedar dlegar, die Gr. Celt.2 474 als conjunctive aufgeführt werden, sind alle drei indicative; bei den letzteren geht das schon daraus hervor, dass diese verba nur den s-conjunctiv bilden.
I am of course aware that Thurneysen, Hdb. des Alt-Irischen 1.345, says: 'Die mangelnde Synkope erweist, dass sich-ar zunächst aus silbebildendem -ṛ entwickelt hat (§109), z. B. ·canar aus *·canṛ.' This would seem to remove canar altogether from the class of ā-forms. I agree with Pedersen, however (Kelt. gram. 2.402, ftn. 1), that this assumption is 'keineswegs notwendig'. It seems highly improbable that we should separate Ir. canar from Umbrian ferar or ferar in turn from Lat. ferātur, which closely glosses it.
40 Cf. M. Leumann, Die ital. f- und b-Tempora, IF 42.70, ftn. 1: ‘nur halte ich es [*es-ām] nicht für eine junge, sondern für eine uralte, vielleicht schon idg. Nachbildung von *bhuu̯,-ām (vgl. lit. bùvo).‘
41 Harpers' Latin Dictionary, s.v. Cf. G. Lodge, Lexicon Plautinum, s.v. Inquam—1. iterat atque confirmat: . . . 2. orationem rectam affert. (1904–1924).
42 Cf. W. A. Merrill on Lucretius 3.341: ‘inquam L. writes when he is very much in earnest.‘ (T. Lucreti Cari, De rerum natura, Libri sex, New York, etc., 1907).
43 Cf. Stolz-Leumann, Lateinische Grammatik 325: 'inquam (falls ursprünglich Praeteritum).' This seems more likely than that it was 'ein eingeschaltetes “möcht' ich sagen” ' (Sommer, Hdb. 488, cf. Brugmann, Grdr.2 2.3.540, 541, 867).
44 See W. Petersen, The Italic Imperfect, Lang. 3.175–83 (cf. Lang. 8.133–7); and, on the antiquity of the form, cf. M. Leumann, IF 42.69, ftn., and 72.
46 Cf. M. Leumann, IF 42.72: ‘fuam “ich sei” ist lat. Neubildung.‘
46 Cf. H. Koppelmann, Die Eurasische Sprachfamilie (Heidelberg 1934), Vorwort.
47 See A. Leskien, Handbuch der altbulgarischen Sprache6 18 (Heidelberg 1922).
48 Id., op. cit. 137.
49 See Carl J. S. Marstrander, Norsk Tidsskrift for Sprogvidenskap 3.243 (1929), and the references there given; W. Schulze, E. Sieg, W. Siegling, Tocharische Grammatik 325, 341 (Göttingen 1931). Cf. A. Meillet, Bull. 19.98, ‘L'emploi des désinences secondaires ne prouve donc rien’; Id., Esquisse d' une histoire de la langue latine2 25 (Paris 1931).
50 See H. Hirt, IF 12.212 (1901): ‘Das hat mich an der Existenz des Injunktivs immer ein bisschen zweifeln lassen’, and compare his recent view, Handbuch des Urgermanischen 2.133 (1932). Cf. Meillet, BSL 19.98.
51 See John Avery, Unaugmented Verb-Forms in the Veda, JAOS 11.330–1 (1885), and the statistics there given.
52 Vergleichende Syntax der indogermanischen Sprachen, Grundriss 4 (1897). 356.
53 Cf. Hirt's remark (Indog. Gram. 2.113), ‘dass das Suffix -ā eine vollständig unklare Grösse ist.‘ But see Meillet, BSL 19.98, and Vendryes, MSL 16.300–5.