8,622
edits
Line 32: | Line 32: | ||
Long vowels /iː uː/ were diphthongized to /ai̯ au̯/ when in open syllables, while they shortened and merged into /i u/ in other cases: <small>LV̄NA</small> > ''rauna'' "moon", <small>VICĪNVM</small> > ''uiçain'' "neighbour; close", <small>MAURĪTĀNIAM</small> > ''Muridonea'' "Mauritania", <small>NV̄LLIFICĀRE</small> > ''nuiifigòi'' "to cancel, revoke". Latin /au̯/ had probably merged with /uː/ before, and shifted back to /au̯/ for the same reason, cf. <small>AVRA</small> > *ūra > aura "gold" <small>(plurale tantum)</small>. | Long vowels /iː uː/ were diphthongized to /ai̯ au̯/ when in open syllables, while they shortened and merged into /i u/ in other cases: <small>LV̄NA</small> > ''rauna'' "moon", <small>VICĪNVM</small> > ''uiçain'' "neighbour; close", <small>MAURĪTĀNIAM</small> > ''Muridonea'' "Mauritania", <small>NV̄LLIFICĀRE</small> > ''nuiifigòi'' "to cancel, revoke". Latin /au̯/ had probably merged with /uː/ before, and shifted back to /au̯/ for the same reason, cf. <small>AVRA</small> > *ūra > aura "gold" <small>(plurale tantum)</small>. | ||
Short word-final vowels except for /a/ were lost as in Gallo-Romance, but in some words short vowels, usually /i/ or /u/, were later added again in order to break clusters; typically it was inserted into a -Cr cluster (<small>CASTRVM</small> > *castr > ''caistur'' "city", cf. <small>CASTRA</small> > ''caistra'' "cities") but after a -NC cluster (<small>QVĪNQVE</small> > *quingui > *pimb > ''pimbi'' "five"). /-ts/ (> /s/ today in most dialects) and /s/ + stop clusters were not changed, cf. <small>LACTEM</small> > *nasti > ''nast'' "milk". | Short word-final vowels except for /a/ were lost as in Gallo-Romance, but in some words short vowels, usually /i/ or /u/, were later added again in order to break clusters; typically it was inserted into a -Cr cluster (<small>CASTRVM</small> > *castr > ''caistur'' "city", cf. <small>CASTRA</small> > ''caistra'' "cities") but after a -NC cluster (<small>QVĪNQVE</small> > *quingui > *pimb > ''pimbi'' "five"). /-ts/ (> /s/ today in most dialects) and /s/ + stop clusters were not changed, cf. <small>LACTEM</small> > *nasti > ''nast'' "milk". First-person singular verbs typically added /a/ to avoid difficult clusters, as in *<small>ADOP(E)RŌ</small> *dubr > ''dubra'' "I use" | ||
/l/ completely disappeared from the Vulgar Latin dialects that became Atlantic, often by dissimilation to /r/ before other consonants (cf. <small>(PRŌVINCIĀS) ĀTLANTICĀS</small> > *Ordanticot > ''Ondàrtigot''), and usually by turning into /j/ (after front vowels or /a/) or /w/ (after back vowels), cf. <small>CABALLVM</small> > ''cauài'' "horse", <small>MĪLLE</small> > ''miu'' "thousand".<br/>Word-initially, it most commonly turned into /n/ – cf. <small>LINGVAM</small> > ''nimba'' "language" – but if the word contained an onset nasal, then it turned into /r/, as in <small>LV̄NA</small> > ''rauna'' "moon". This is the origin of the two sets of definite articles used in Atlantic, with the usual forms being ''in'' and ''na'', but with ''ir'' and ''ra'' being used if the word they attach to has a nasal in its first syllable (cf. ''na rauna'' "the moon" but ''ra nimba'' "the language"). Originally this only applied to the feminine article, and if the word had a nasal but in coda it didn't apply, but analogy has extended this to all cases (cf. ''ir ondartigòr'' "the Atlantic man").<br/>/l/ changed this way also in the <small>FL-</small> initial cluster, where (just as in <small>FR-</small>) the initial /f/ became first an approximant /w/ and then got fully vocalized to /u/, giving ''ur-'' or ''un-'' in such words, like <small>FLŌREM</small> > */wnor/ > ''unòr'' /uˈnor/ "flower" or <small>FLV̄MEN</small> > */wrau̯m/ > ''uraum'' /uˈrau̯m/ "river". <small>PL-</small>, however, consistently became ''pr-'', which means that /l/ got rhotacized before it got lost in other places (cf. <small>PLVVIAM</small> > ''pruia'' "rain"). | /l/ completely disappeared from the Vulgar Latin dialects that became Atlantic, often by dissimilation to /r/ before other consonants (cf. <small>(PRŌVINCIĀS) ĀTLANTICĀS</small> > *Ordanticot > ''Ondàrtigot''), and usually by turning into /j/ (after front vowels or /a/) or /w/ (after back vowels), cf. <small>CABALLVM</small> > ''cauài'' "horse", <small>MĪLLE</small> > ''miu'' "thousand".<br/>Word-initially, it most commonly turned into /n/ – cf. <small>LINGVAM</small> > ''nimba'' "language" – but if the word contained an onset nasal, then it turned into /r/, as in <small>LV̄NA</small> > ''rauna'' "moon". This is the origin of the two sets of definite articles used in Atlantic, with the usual forms being ''in'' and ''na'', but with ''ir'' and ''ra'' being used if the word they attach to has a nasal in its first syllable (cf. ''na rauna'' "the moon" but ''ra nimba'' "the language"). Originally this only applied to the feminine article, and if the word had a nasal but in coda it didn't apply, but analogy has extended this to all cases (cf. ''ir ondartigòr'' "the Atlantic man").<br/>/l/ changed this way also in the <small>FL-</small> initial cluster, where (just as in <small>FR-</small>) the initial /f/ became first an approximant /w/ and then got fully vocalized to /u/, giving ''ur-'' or ''un-'' in such words, like <small>FLŌREM</small> > */wnor/ > ''unòr'' /uˈnor/ "flower" or <small>FLV̄MEN</small> > */wrau̯m/ > ''uraum'' /uˈrau̯m/ "river". <small>PL-</small>, however, consistently became ''pr-'', which means that /l/ got rhotacized before it got lost in other places (cf. <small>PLVVIAM</small> > ''pruia'' "rain"). |
edits