User:Melroch/Vulgar Latin Phonology

Syllabification and stress

Stress in Classical Latin

Old Latin had fixed stress on the initial syllable of a word. In Classical Latin this had been replaced by a movable (but not free) stress. The well-known Classical Latin stress rule is that the stress falls on the penultimate (next to last) syllable of the word if that syllable is long, but on the antepenultimate (third from the end) syllable if the penultimate is short.

Syllable length and syllable boundaries in Classical Latin

In Classical Latin a syllable was long if its vowel nucleus was long or diphthongal, or if it was a short vowel followed by more than one consonant, including a geminate or double consonant, thus e.g. boniˈtātem, deˈlĕcto, ˈcompŭto. If a vowel was followed by more than one consonant the first of these belonged to the preceding syllable, which then counted as closed. A single consonant between vowels belonged to the following syllable, e.g. ˈpo-si-tus'.

A syllable with a short vowel followed by a consonant cluster is called "long by position"; the vowel is said to "be in position" and the consonant cluster is said to "make position". While the term "long by position" and "positional length" are still in common use the terms "to be in position" and "to make position" are now somewhat old-fashioned and will be avoided in the following.

Muta cum liquida

The main exception to these rules in Classical Latin was when a vowel was followed by a single consonant followed by a liquid consonant (l or r), the so-called muta cum liquida rule: in this case the entire stop + liquid combination belonged to the following syllable, and did not attract stress, thus ˈtenebrae, ˈvolucres, ˈpharetra, 'cathedra etc.

Stress in Vulgar Latin

The stress in Vulgar Latin normally remained in the position determined by the Classical Latin quantitative stress rule even after changes in vowel quality and quantity had obliterated the old quantity system. Vulgar Latin thus acquired distinctive stress.

Exceptions to the Classical stress placement in Vulgar Latin

Muta cum liquida in Vulgar Latin

In Vulgar Latin a stop + liquid usually attracted stress, i.e. the muta cum liquida rule was not operative. Thus e.g. caˈthedra. Interestingly pre-classical poets and dramatists didn't observe the muta cum liquida rule — e.g. Naevius accents inˈtegram on the penultimate syllable. Even classical poets sometimes let a muta cum liquida attract stress, so obviously there was dialectal variation in pre-classical Latin, and between dialects contemporary with Classical Latin.

Vowels in hiatus

In Vulgar Latin i, e and u followed by a vowel became a semivowel j < i, e ([j]) or v < u [w]). Since the j or v counted as a consonant the stress had to shift. For some reason the accent shifted to the following vowel in the case of j but to the preceding vowel in the case of v:

muˈlĭĕrem > mulˈjerem;

puˈtĕŏlis > putˈjolis (pvteo‍́lis is attested in an inscription);

paˈrĭĕtes > par'jetes

but

bat(t)ˈŭĕre > *ˈbattuere > ˈbattere;

conˈsŭĕre > *ˈconsuere > ˈconsere;

haˈbŭĕrunt > *ˈhabuerunt > ˈhaberunt;

teˈnŭĕram > *ˈtenueram.

Probably there was analogy with forms like ˈbattuo, ˈhabuit, ˈtenui where the stress fell on the first syllable according to Classical rules. Also note that the u > [w] was lost in these words.

Does Spanish muˈjer come from the accusative but Italian ˈmoglie from the nominative, or was there dialectal variation in accentuation of hiatus words?