Looking at modern French in light of vulgar Latin, or Chinese compared with Proto-Sino-Tibetan (if that can even be reconstructed), there seems to be quite a few contexts in which phonemes are subsumed, vowels are nasalized to indicate the loss of a nasal, or tones are introduced to retain distinctions formerly maintained by distinct consonant clusters. Modern Celtic languages certainly seem to have done a lot of this too, and I have heard that a similar thing happened in Sumerian.

I was wondering if anyone has any theories or reference to theories about

(1) What makes a language "prepared" for such a phenomenon to occur (e.g., are there some languages that this would be much less likely to happen to?)

(2) Whether there are any aspects of a culture speaking a language that make people more likely to gravitate towards doing something like this across whatever number of generations.

(I would also be interested in hearing whether anyone believes there was the strong presence of a Celtic substratum in French that influenced it in this way, but that's kind of tangential.)

  • 2
    maybe related Which phenomena compensate for sound losses in languages?
    – Louis Rhys
    Commented Jan 4, 2012 at 1:43
  • 1
    Re: your tangent, there are various musings about the possible influence of Celtic phonology on French phonology, but in the end it is very difficult to separate that out from other influences (e.g. Germanic superstratum), from natural processes of sound change, and also from the possibility that related languages underwent similar changes at the same time. Attributing the phonemic nasalization of French vowels to a Celtic influence, for example, has not been considered especially plausible, partly because this happens all over the world and also in other Romance languages (e.g. Portuguese). Commented Jan 4, 2012 at 11:56

1 Answer 1


It looks like you are interested in the transphonologization of phonemic contrasts. There is a nice downloadable paper (Hyman 2008) dealing with the topic in an accessible way which includes multiple examples and plenty of references for further reading.

  • @Daniel Briggs, I think this reference is exactly the sort of thing you are looking for, and hopefully it will make clear that these sorts of processes (the enhancement of phonetic detail into phonological patterns, or the conversion of substance to form) can occur in any language in which natural processes of language change take place (i.e. all languages). The specific changes that occur in each language will depend on both universal influences of the vocal tract and language-specific characteristics and pressures, but every language (and culture) will exhibit some form of these processes. Commented Jan 4, 2012 at 11:12
  • Working through the paper. I'll eventually accept, but I wanted to see if anyone had theories on specific pressures that cause the phonologization to entail largely of dropping consonants or, if that's always the trend in phonologization (as i guess it may be, on account of chains like p > f > h > zero), then simply what may cause it to happen (or even correlate with it happening) at a fast rate. For example, French has had a lot more of it than Spanish. One can deduce that quite a bit happened in Sumerian, but Finnish has had very little in 3000 years. Commented Jan 8, 2012 at 3:05
  • 1
    The development of phonological prominence often entails the weakening of consonants in non-prominent positions, so a language that develops word-accent should be more likely to undergo the kind of weakening changes you refer to. I would like to see a good reference on diachronic aspects of word-accentual systems.
    – user483
    Commented Jan 8, 2012 at 5:40
  • 1
    There’s some stuff about this in the prosody literature, since lenition is expected to happen in non-prominent areas of words. This is because non-prominent areas of words tend to be rushed. The typical non-prominent part of a word is the end, and that is indeed where you see nasalization very frequently. I don’t have any references on hand for this though.
    – James C.
    Commented Jan 11, 2012 at 1:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.