Latin nasalization of the final -m


Senior Member
Did the final "-m" always nasalize the preceding vowel from the very beginning of classical latin or was it an [m] initially and then early on became a marker of nasalization? (like in the first century AD or some other time for example)
  • Sardokan1.0

    Senior Member
    Sardu / Italianu
    I think that the nasalization it's only a thing of English speakers trying to pronounce Latin. There is no trace of this nasalization in Ecclesiastical Latin or in Romance languages and even in Sardinian, the most conservative language of the Romance family.


    Senior Member
    Final -m at least sometimes disappears in writing already in Old Latin, for example in this inscription from 298. B. C. All the accusative singulars in that inscription miss final -m!

    This is obviously a feature that appeared early in the spoken language. As Sardokan mentioned, not a single Romance language preserves final -m (except in some monosyllables, like lat. quem > sp. quien or lat. rem > fr. rien.). And there wasn't probably much of a difference between the classical and the vulgar language when it comes to that feature: I barely know anything about Latin poetry, but even in the classical language, words ending in -m could have their final vowel elided like words ending in a vowel. So, whatever be the value of final -m in Classical Latin, it wasn't the consonant /m/, otherwise elision couldn't happen.

    The question is, if final -m disappeared rather early, even in writing (as the inscription I liked to, and other inscriptions, prove), how could it be perfectly restored in writing in the classical period? This is where the ideal of nasalisation comes from: that final -m was lost as a consonant, but preserved as nasalisation for a time, so that it could still be heard and restored in writing, but still allow elision in poetry. There's probably also a statement by a classical author somewhere which alludes to such a thing, but note that their "phonetics" weren't nearly as precise as ours so there's always space for debate. For example, a Renaissance French grammarian wrote that s before a consonant is barely pronounced... but we know for sure that what he meant was that it isn't pronounced at all, but it's presence indicates that the preceding vowel is long!

    While these deductions about Latin pronunciation aren't likely to be confirmed once and for all, there's precedent in living languages for them. For example, in my native language, Croatian, there's a similar process of nasalisation to what is reconstructed for Classical latin: before fricative sounds, n isn't pronounced as a consonant, but as nasalisation of the preceding vowel, and in some dialects the same happens to final -m! So this comes naturally to me and there's not theoretical objection for it to have been the case in some stage of Latin (and the situation with vowel length and the loss thereof is also rather similar, except that we're likely to go Sardinian way and keep the 5 vowel system... systems where there's open and closed e's and o's are considered "rural" and "uncultivated").


    Senior Member
    English - South-East England
    Yes, pre-Classical inscriptions without <m> show that it had ceased to be a consonant already by then. However, it apparently was something: W. Sidney Allen mentions that Verrius Flaccus (d. 20 CE) proposed writing it with a half M, that is an inverted V. There is also the fact that it assimilated like a consonant in groups such as etiam nunc and tam durum. But it seems to have disappeared entirely not long after that time.
    < Previous | Next >