Element theory

[malloc]

Has anyone else heard of element theory in phonology? What I have read about it so far seems interesting, but many things still confuse me about it. If elements are defined by acoustic properties, then phonemes with widely differing allophones would seem to pose some problems. Consider the /h/ of Japanese whose allophones include [h ~ ç ~ ɸ], each with rather distinct acoustics. Furthermore, it seems that the theory proposes for some languages, coronal consonants contain the element |A| and thus pattern somehow with low vowels, much like labials pattern with round vowels and palatals with front vowels. Yet apart from that claim, I have never come across anything to suggest that phonemes like /t n/ pattern with /a/ the way /p m/ can pattern with /u/ and so forth.

[mèþru]

I looked it up to respond but got bored in the introduction. How does this differ from autosegmental theory?

Usually I see allophony where post-velar positions, like uvulars or glottals, patterning with low vowels (especially back low voewls) rather than a coronal pattern.

[WhiteWolf]

I could see coronals patterning with central vowels considering different analyses of Mandarin syllabic fricatives (i.e. [ʂʐ̩] vs [ʂɨ]) and the Index Diachronica lists PIE to Proto-Tocharian with n̩ → ə → ∅ / C_#. And I guess /a/ can be central, but I don't see it with other low vowels. Where did you read that coronal consonants contain |A|? So far, I've only found references to |I|.

[Nortaneous]

don't do theory, kids

[Curlyjimsam]

[h ~ ç ~ ɸ] are arguably more acoustically similar than they are articulatorily similar. I think the sound changes that led to this split (and similar changes in other languages, which are quite frequent) happened for acoustic reasons.