feminine ācris was restored in Latin (though not in North Oscan) by
the analogy of other adjectives, like tristis, while the masculine ācer
was protected by the parallel masculine forms of the -o- declension,
like tener, niger [from *teneros, *nigros]).
21. Long vowels generally remained unchanged, as in compāgo, condōno.
22. Of the diphthongs, ai and oi both sank to ei, and with original ei further to ī, in unaccented syllables, as in Achivi from Gr. Ἀχαιϝοί, olīivom, earlier *oleivom (borrowed into Gothic and there becoming alēv) from Gr. ἔλαιϝον. This gives us interesting chronological data, since the el- must have changed to ol- (§ 16. 3) before the change of -ai- to -ei-, and that before the change of the accent from the first syllable to the penultimate (§ 9); and the borrowing took place after -ai- had become -ei-, but before -eivom had become -eum, as it regularly did before the time of Plautus.
But cases of ai, ae, which arose later than the change to ei, ī, were unaffected by it; thus the nom. plur. of the first declension originally ended in -ās (as in Oscan), but was changed at some period before Plautus to -ae by the influence of the pronominal nom. plur. ending -ae in quae? hae, &c., which was accented in these monosyllables and had therefore been preserved. The history of the -ae of the dative, genitive and locative is hardly yet clear (see Exon, Hermathena (1905), xiii. 555; K. Brugmann, Grundriss, 1st ed. ii. 571, 601).
The diphthongs au, ou in unaccented syllables sank to -u-, as in inclūdō beside claudō; the form clūdō, taken from the compounds, superseded claudo altogether after Cicero’s time. So cūdō, taken from incūdō, excūdō, banished the older *caudō, “I cut, strike,” with which is probably connected cauda, “the striking member, tail,” and from which comes caussa, “a cutting, decision, legal case,” whose -ss- shows that it is derived from a root ending in a dental (see §25 (b) below and Conway, Verner’s Law in Italy, p. 72).
Consonants.—Passing now to the chief changes of the consonants we may notice the following points:—
23. Consonant i (wrongly written j; there is no g-sound in the letter), conveniently written i̭ by phoneticians,
(i.) was lost between vowels, as in trēs for *trei̭es, &c. (§ 17. 6);
(ii.) in combination: -mi̭- became -ni-, as in veniö, from Ind.-Eur. *Ƨṷ mi̭o, “I come,” Sans. gam-, Eng. come; -ni̭- probably (under certain conditions at least) became -nd-, as in tendō beside Gr. τείνω, fendō = Gr. θείνω, and in the gerundive stem -endus, -undus, probably for -eni̭os, -oni̭os; cf. the Sanskrit gerundive in -an-īya-s; -gi̭-, -di̭- became -i̭- as in māior from *mag-ior, pēior from *ped-ior;
(iii.) otherwise -i̭- after a consonant became generally syllabic (-ii̭-), as in capiō (trisyllabic) beside Goth. hafya.
24. Consonant u (formerly represented by English v), conveniently written ṷ,
(i.) was lost between similar vowels when the first was accented, as in audīui, which became audiī (§ 17 [6]), but not in amāuī, nor in avārus.
(ii.) in combination: dṷ became b, as in bonus, bellum, O. Lat. dṷonus, *dṷellum (though the poets finding this written form in old literary sources treated it as trisyllabic); pṷ-, fṷ-, bṷ-, lost the ṷ, as in ap-erio, op-erio beside Lith. -veriu, “I open,” Osc. veru, “gate,” and in the verbal endings -bam, -bō, from -bhṷ-ām, -bhṷō (with the root of Lat. fui), and fīo, du-bius, super-bus, vasta-bundus, &c., from the same; -sṷ- between vowels (at least when the second was accented) disappeared (see below § 25 (a), iv.), as in pruīna for prusuīna, cf. Eng. fros-t, Sans, pruṣvā, “hoar-frost.” Contrast Minérva from an earlier *menes-ṷā, sṷe-, sṷo-, both became so-, as in sorōor(em) beside Sans. svasār-am, Ger. schwes-t-er, Eng. sister, sordēs, beside O. Ger. swart-s, mod. schwarz. -ṷo- in final syllables became -u-, as in cum from quom, parum from parṷom; but in the declensional forms -ṷu- was commonly restored by the analogy of the other cases, thus (a) serṷos, serṷom, serṷī became (b) *serus, *serum, *serṷi, but finally (c) serṷus, serṷum, serṷi.
(iii.) In the 2nd century A.D., Lat. v (i.e. ṷ) had become a voiced labio-dental fricative, like Eng. v; and the voiced labial plosive b had broken down (at least in certain positions) into the same sound; hence they are frequently confused as in spellings like vene for bene, Bictorinus for Victorinus.
25. (a) Latin s
(i.) became r between vowels between 450 and 350 B.C. (for the date see R. S. Conway, Verner’s Law in Italy, pp. 61-64), as āra, beside O. Lat. āsa, generis from *geneses, Gr. γένεος; eram, erō for *esām, *esō, and so in the verbal endings -erām, -erō, -erim. But a considerable number of words came into Latin, partly from neighbouring dialects, with -s- between vowels, after 350 B.C., when the change ceased, and so show -s-, as rosa (probably from S. Oscan for *rodi̭a “rose-bush” cf. Gr. ῥόδον), cāseus, “cheese,” miser, a term of abuse, beside Gr. μυσαρός (probably also borrowed from south Italy), and many more, especially the participles in -sus (fūsus), where the -s- was -ss- at the time of the change of -s- to -r- (so in causa, see above). All attempts to explain the retention of the -s- otherwise must be said to have failed (e.g. the theory of accentual difference in Verner’s Law in Italy, or that of dissimilation, given by Brugmann, Kurze vergl. Gram. p. 242).
(ii.) sr became þr ( = Eng. thr in throw) in pro-ethnic Italic, and this became initially fr- as in frīgus, Gr. ῥῖγος (Ind.-Eur. *srīgos), but medially -br-, as in funebris, from funus, stem funes-.
(iii.) -rs-, ls- became -rr-, -ll-, as in ferre, velle, for *fer-se, *vel-se (cf. es-se).
(iv.) Before m, n, l, and v, -s- vanished, having previously caused the loss of any preceding plosive or -n-, and the preceding vowel, if short, was lengthened as in
prīmus from *prismos, Paelig. prismu, “prima,” beside pris-cus.
iūmentum from O. Lat. iouxmentum, older *ieugsmentom; cf. Gr. ζεῦγμα, ζύγον, Lat. iugum, iungo.
lūna from *leucsnā-, Praenest, losna, Zend raoχsna-; cf. Gr. λεῦκος, “white-ness” neut. e.g. λευκός, “white,” Lat. lūceō.
tēlum from *tēns-lom or *tends-lom, trānāre from *trāns-nāre.
sēvirī from *sex-virī, ēvehō from *ex-vehō, and so ē-mittō, ē-līdō, ē-numerō, and from these forms arose the proposition ē instead of ex.
(v.) Similarly -sd- became -d-, as in īdem from is-dem.
(vi.) Before n-, m-, l-, initially s- disappeared, as in nūbo beside Old Church Slavonic snubiti, “to love, pay court to”; mīror beside Sans, smáyatē, “laughs,” Eng. smi-le; lūbricus beside Goth, sliupan, Eng. slip.
(b) Latin -ss- arose from an original -t + t-, -d + t-, -dh + t- (except before -r), as in missus, earlier *mit-tos; tōnsus, earlier *tond-tos, but tonstrīx from *tond-trīx. After long vowels this -ss- became a single -s- some time before Cicero (who wrote caussa [see above], divissio, &c., but probably only pronounced them with -s-, since the -ss- came to be written single directly after his time).
26. Of the Indo-European velars the breathed q was usually preserved in Latin with a labial addition of -ṷ- (as in sequor, Gr. ἕπομαι, Goth, saihvan, Eng. see; quod, Gr. ποδ-(απός), Eng. what); but the voiced Ƨṷ remained (as -gu-) only after -n- (unguo beside Ir. imb, “butter”) and (as g) before r, l, and u (as in gravis, Gr. βαρύς; glans, Gr. βάλανος; legūmen, Gr. λοβός, λεβίνθος). Elsewhere it became v, as in veniō (see § 23, ii.), nūdus from *novedos, Eng. naked. Hence bōs (Sans. gāus, Eng. cow) must be regarded as a farmer’s word borrowed from one of the country dialects (e.g. Sabine); the pure Latin would be *vōs, and its oblique cases, e.g. acc. *vovem, would be inconveniently close in sound to the word for sheep ovem.
27. The treatment of the Indo-European voiced aspirates (bh, dh, ĝh Ƨh) in Latin is one of the most marked characteristics of the language, which separates it from all the other Italic dialects, since the fricative sounds, which represented the Indo-European aspirates in pro-ethnic Italic, remained fricatives medially if they remained at all in that position in Oscan and Umbrian, whereas in Latin they were nearly always changed into voiced explosives. Thus—
| Ind.-Eur. bh: | initially Lat. f- (ferō; Gr. φέρω). |
| medially Lat. -b- (tibi; Umb. tefe; Sans, tubhy-(am), “to thee”; the same suffix in Gr. βίη-φι, &c.). | |
| Ind.-Eur. dh: | initially Lat. f- (fa-c-ere, fē-c-ī; Gr. θετός (instead of *θατός), ἔθη-κα). |
| medially -d- (medius; Osc. mefio-; Gr. μέσσος, μέσος from *μεθιος); except after u (iubēre beside iussus for *i̭udh-tos; Sans. yṓdhati, “rouses to battle”); before l (stabulum, but Umb. staflo-, with the suffix of Gr. οτέργηθρον, &c.); before or after r (verbum: Umb. verfale: Eng. word. Lat. glaber [v. inf].: Ger. glatt: Eng. glad). | |
| Ind.-Eur. ĝh: | initially h- (humī: Gr. χαμαί); except before -u- (fundo: Gr. χέ(ϝ)ω, χύτρα). |
| medially -h- (veho: Gr. ἔχω, ὄχος; cf. Eng. wagon); except after -n- (fingere: Osc. feiho-, “wall”: Gr. θιγγάνω: Ind.-Eur. dheiĝh-, dhinĝh-); and before l (fīg(u)lus, from the same root). | |
| Ind.-Eur gṷh: | initially f- (formus and furnus, “oven”, Gr. θερμός, θέρμη, cf. Ligurian Bormiō, “a place with hot springs,” Bormanus, “a god of hot springs”; fendō: Gr. θείνω, φόνος, πρόσ-φατος). |
| medially v, -gu- or -g- just as Ind.-Eur. Ƨṷ (ninguere, nivem beside Gr. νίφα, νείφει; frāgrāre beside Gr. ὀσφραίνομαι [ὀσ- for ods-, cf. Lat. odor], a reduplicated verb from a root Ƨṷhra-). |
For the “non-labializing velars” (Hostis, conGius, Glaber) reference must be made to the fuller accounts in the handbooks.
28. Authorities.—This summary account of the chief points in Latin phonology may serve as an introduction to its principles, and give some insight into the phonetic character of the language. For systematic study reference must be made to the standard books, Karl Brugmann, Grundriss der vergleichenden Grammatik der Indo-Germanischen Sprachen (vol. i., Lautlehre, 2nd ed. Strassburg, 1897; Eng. trans. of ed. 1 by Joseph Wright, Strassburg, 1888) and his Kurze vergleichende Grammatik (Strassburg, 1902); these contain still by far the best accounts of Latin; Max Niederman, Précis de phonétique du Latin (Paris, 1906), a very convenient handbook, excellently planned; F. Sommer, Lateinische Laut- und Flexionslehre (Heidelberg, 1902), containing many new conjectures; W. M. Lindsay, The Latin Language (Oxford, 1894), translated into German (with corrections) by Nohl (Leipzig, 1897), a most valuable collection of material, especially from the ancient grammarians, but not always accurate in phonology; F. Stolz, vol. i. of a joint Historische Grammatik d. lat. Sprache by Blase, Landgraf, Stolz and others (Leipzig, 1894); Neue-Wagener, Formenlehre d. lat. Sprache (3 vols., 3rd ed.
