zompist bboard

Hi all,

Do we know of any languages that have changed a previous *h to j (for example before front vowels)?

English? (huge, human are pronounced yuge, yuman by some people.)

Very interesting and a good point. Although I guess more commonly the analysis would be that the vowel became a falling diphthong and the h was dropped?

Given that those words do have /hj/ for people who don't drop the h, yes, I'd wager that a direct change h > j isn't the case here.

h > j was reconstructed as a change in the development of some Khoisan languages, IIRC.

Livonian: *h > j between a front vowel and any other vowel, e.g. lejā 'meat' ~ Finnish & Estonian liha.

Coptic apparently has x > š unconditionally, or at least as best we know given Egyptian phonology. It may have even been h>š. That's t least halfway by itself, and I think in uncodnitional
Š>j is reasonable.

Tropylium wrote:Livonian: *h > j between a front vowel and any other vowel, e.g. lejā 'meat' ~ Finnish & Estonian liha.

How likely is it that that was actually lehā > leā > lejā?

Travis B. wrote:

Tropylium wrote:Livonian: *h > j between a front vowel and any other vowel, e.g. lejā 'meat' ~ Finnish & Estonian liha.

How likely is it that that was actually lehā > leā > lejā?

Not strictly ruled out, but that would make it the only case in Livonian where something is either lost or epenthesized intervocalically.

I was thinking about phonemic alternations in Korean and Spanish the other day and musing that it would be fun to have a language where /ð/ (whether from original /d/ or not) devoices to [θ] finally (cf. Castilian Madriz) and is either lost medially or becomes [j] to avoid hiatus (cf. Finnish). Final [θ] could debuccalise and you'd end up with a [j] ~ [h] alternation.

Aside from Khoisan:
27.5 Proto-Mayan to Kaqchikel-Tz’utujil h → j / _#
29.1.1.1.27 Proto-Western Southern Athabaskan to Navajo {ɣ,h} → j / in prefixes related to word derivation