12th August 2024: digital purchasing is currently unavailable on Cambridge Core. Due to recent technical disruption affecting our publishing operation, we are experiencing some delays to publication. We are working hard to restore services as soon as possible and apologise for the inconvenience. For further updates please visit our website: https://www.cambridge.org/news-and-insights/technical-incident
We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Published online by Cambridge University Press: 02 December 2020
During the past fifty or sixty years in the history of Romance linguistics, the erratic development of vowels in hiatus has constituted an ever recurring problem. Scholars have been little preotcupied with the subject except in so far as it touches the history of the number words (CL dŭo) and the possessive adjectives and pronouns (CL měus, tŭus, sŭus). A study of the latter led d'Ovidio to give the first account of any extent of the development of vowels in hiatus in a Romance tongue; he limited his study to Italian. At about the same time, Karl Brugmann was investigating the mode of development of vowels in hiatus in Latin and Greek. Although he was able to establish a fairly clear system of development for the forms in the ancient languages, further studies in Romance linguistics failed to clarify the issue for the modern languages. In 1893, Menger attempted to answer, at least tentatively, some of the outstanding questions incidental to the problem and cleared the atmosphere somewhat; his treatment was, however, primarily a development of the possessive pronoun and was limited to Italian. The work of scholars such as Meyer-Lubke led to mere tabulation of forms; other treatments have considered at most a set group of words, like the possessive pronouns, in the various languages or have dealt only with the multitude of dialectal forms in an area geographically restricted. In short, the development of vowels in hiatus in the whole of the Romance territory has never been the object of a comprehensive study.
1 “Pronomi personali e possessivi”, Archimo Glottologico Iialiano, ix (1886), 25–101.
2 Grundriss ier Vergleichenden Grammatik der Indogerm. Sprachen (Strassburg: K. J. Triibner, 1886-1900), i, 111–151 passim. For the pertinent facts under one heading, see the translation of Joseph Wright, Elements of the Comparative Grammar of the Indo-Germanic Languages (New York, 1888), i, 105,453–458.
3 Louis Emil Menger, The Historical Development of the Possessive Pronouns in Italian (Baltimore: Modern Language Association of America, 1893).
4 “Diphthong Formation”, Archives Néerlandaises, v (1930), 31 ff.
6 Ibid., p. 31 f.
6 See Edwin Williams, From Latin to Portuguese (Philadelphia, 1938).
7 . Gorra, “Dell'epentesi di iato” in E. Monaci, Studj it Filologia Romanza, vi (1893), 502, 535.
8 Hugo Schuchardt, Der Vokalismus des Vulgàrlaieins (Leipzig: Teubner, 1866-68), ii, 468–469.
9 E. Diehl, “Vulgärlateinische Inschriften” 190, in H. Lietzmann, Kkine Texte filr Vorlesungen uni Übungen (Bonn, 1910), p. 62.
10 The vowels in a hiatus group of either of these two categories show few peculiarities of development: an e if first in the group gives almost invariably i: crěáre>criar (Prov., Port., Cat.), crier (OF). Under the same circumstances an í did not give e in VL but usually remained i or i: CL apiàrium> achier (OF); VL aviólu(m)> aïeul (Modem French). Likewise ù in these positions did not give ç as elsewhere but remained u or was reduced to y: CL idŭa>vedue (OF, Friulian). If first in the group, usually remains o, on the other hand, though examples are rare: CL bôârius>boaro (It.); as previously noted, the hiatus group is usually reduced and o>u: CL côâgulum>cuajo (Sp.), quaglia (It.); the early value of the in Port, coalho is open to question.
11 The ensuing examples are grouped roughly according to the language to which they belong, except in a few instances of such complication that grouping by word better clarifies the situation.
12 soum: Schuchardt, Der Vokalismus des Vulgdrlateins, ii, 162.
13 sovo: C.I.L., i, 1007; soveis=suis: C.I.L., I, 198, 50; sovam=souom, genitive pl.: C.I.L., i, 588.
14 svvo; tuos: C.I.L., i, 1242 and vi, 1527, d 41, respectively.
15 Cf. Plautus: piius; Oscan piihiiu=La.tin pio; Umbrian pihaz = Latin piatus, Umbrian peihaner—Latin piandl; Umbrian pihatu=Latm piato.
16 Cf. my article on “Hiatus and Vocalic Quality in Classical and Vulgar Latin”, in Classical Philology, XLI (1946), 226.
17 Though it is not the purpose of this study to clarify the formation of Fr. pieux of which only the radical vowel concerns the question at hand, I allude in passing to W. Meyer-Lübke, Historische Grammatik der Franzôsischen Sprachen (Heidelberg, 1913), p. 73, for his supposition that on the analogy of the Latin nominative vilis>OF vius>(with the introduction of a glide-vowel) vieus, Latin filius>fieus, etc., similarly Latin pius>OF pieus which was maintained through confusion with -eus VL -osus. This hypothesis agrees in any case with my contention that the word is popular, not learned.
18 That rio seems to follow the same phonetic pattern as dio and mio is a strong argument against the triphthongal stage, as there was probably no diphthongization of the ç because of the preceding r. I am inclined to consider reo (meaning “bad”) popular rather than learned because of its frequent occurrence in Old Italian and its similarity in respect of phonetic pattern to popular deo, meo, and eo; whereas these latter forms are no longer in use, reo has been strengthened probably by confusion with learned reo (meaning “guilty”).
19 The question of the shift of accent brings up the established Tuscan forms lugo=∗luogo, nuvo=nuovo, lucu, gittco, pute=puole, upere=uopere; thus ∗sóo> ∗s óo>sûo.
20 Cut is usually monosyllabic in poetic usage in CL; the later writers however treated it as dissyllabic, as it had probably always been in spoken discourse. Moreover Priscian says, v. Keil, Gr. Lat., ii, 304, 19: “Cui quoque inveniuntur quidam bisyllabe protulisse per diaeresin… ” This cumulative evidence is proof of the existence of a dissyllabic form; hence modem It. cui points to VL cui. Engadin cui (koi) with its obvious vocalic dissimilation indicates also a dissyllabic form in VL.
21 J. D. M. Ford, Old Spanish Readings (Boston: Ginn, 1906), p. xii.
22 Frederic George Mohl, Introduction à la chronologie du latin vulgaire (Paris : . Bouillon, 1899), pp. 287–288.
23 A. Ernout and A. Meillet, Dictionnaire étymologique de la langue latine (Paris, 1932), p. 736. Also for both plus and dies, see Seelmann, Die Aussprache des Latein (Heilbronn, 1885), p. 93. Dies: C.I.L., vi, 7527, 7579.
24 J. W. Ducibella, The Phonology of the Sicilian Dialects (Washington: Catholic Univ., 1934), p. 44.
25 Translated and reduced by Bartoli and Braun (1914), p. 74. Actually, this is Bartoli's rather than Meyer-Lübke's work.
28 Evidently a correction must be made on the text which in this and subsequent editions reads steadily ŭ at this point.
27 Italienische Grammatik (Leipzig), sec. 96.
28 (Paris, 1890–1906), p. 246.
29 Gröber, Grundriss (Strassburg: Triibner, 1904-6), i, 926,929.
30 Die Aussprache des Latein, p. 93.
31 RomanischeSprachwissenschafl (Berlin—Leipzig, 1921).
32 Ibid., p. 67.
33 (Milan, 1918), p. 294.
34 See footnote 1.
35 Edward Bourciez, Elements de la linguistique romane (Paris, 1930).
36 Ibid., p. 45.
37 d'Ovidio, of. cit.
38 See footnote 3.
39 “Die betont. Hiatusvokale im Vulgàrlatein”, Zr. R. P., xxv (1901), 342.
40 Pietro Galor Goidànich, “L'origine e le forme délia dittongazione romanza”, Zr. R. P., v (1907), 158–159.
41 “Vocali in iato”, Zr. R. P., xliii (1924), 215.
42 Einführung in das Sludium der Romanischen Sprachmssenschafl (Heidelberg, 1909), p. 129.
43 Ibid.
44 OF voie<VL vęa, soie<Yh sea, sue but suen; in Rum. ŭ>u, and in Sic. and Sard. ï>i and ŭ>u.
45 Any vowel except a, which ordinarily is not followed by a glide vowel, is susceptible to closure in hiatus, and û showing a marked tendency in the direction indicated.
46 Rum. zeu illustrates the point that there was no closure, even though àěu>>diêu>zeu.
47 At the same time, closure or as in this case maintenance due to metaphony before final i should be noted as widespread in Romance. The maintenance of u in the continuant of fui in Tuscan Italian, in which this metaphony is supposedly not operative, and the maintenance of cut in Tuscan also, indicate rather the type of closure under consideration, that is, of a vowel in hiatus. Furthermore, if Engadin jo, which occurs beside füc, is dismissed as an analogy to fo fubit, the Romance forms point uniformly to a VL fui. There were in Latin two roots, one with a long vowel for use before consonants, and one with a short vowel for use before vowels. Note also old Oscan FUUTREI, Emout and Meillet, Dictionnaire étymologique de la langue latin, p. 380. Fûuimus is attested in Ennius. However, further evidence for/tfj is scant and, if dismissed, imposes the form ∗fui in VL, in accord with the Romance continuants of the word. It seems to me arbitrary to invoke a metaphony which operates somewhat fitfully in various parts of the Romance territory to explain the present forms which show almost complete uniformity in the whole territory. The evidence for this kind of metaphony is clearly presented by Friedrich Schurr in “Umlaut und Diphthongierung in der Romania”, Romanische Forschungen, L (1936), 275 ff.; parts of the argument are greatly weakened in my opinion by recourse to forms which, as far as I have been able to discover, are unattested; e.g., It. ∗mieo, ∗dieo, though Schurr does not provide these forms with an asterisk. I am indebted to Professors Leo Spitzer and Julian Bonfante for valuable indications on this subject as well as on several others incident to the problem.
48 For siat: Archivvm latinitatis medii aevi, Bulletin au Cange, v (1929–30), 156.
49 These last mentioned so and to forms may well stem from the reduced VL sum and tum. 50 For duoi: Annibal Caro, Lettere dell'Autore (Padova: Comino, 1725), ii, 93; also Ben-venuto Cellini, Vita (Florence; Piatti, 1829), i, 223.
51 Also in modern Italian, Tuscan presents bove. The form is amply attested in the past; see among others: Jacopo Sannazaro, L'Arcadia, ed. Michèle Scherillo (Torino, 1888), p. 303, 1. 214.
