Quantcast
Viewing all articles
Browse latest Browse all 104

Pre-Proto-Berber lenition

I tend to not work too much on really deep reconstruction of Pre-Proto-Berber, because arguments tend to become very circular very quickly. Nevertheless something has been on my mind, and I doubt it will become solid enough that I could publish it, so instead I'll share it on this blog. It may come in useful for me sometime, it may come in useful for someone else.

Proto-Berber has a fairly solid rule that changes *ḱ to y intervocalically. Examples are especially found in Verbal noun formations paired with the verb it is derived from

*ăḱrəz'to plow', VN ta-yərz-a< *ta-ḱərz-a

In post-vocalic position Berber languages have a y which usually corresponds to k in the more isolated Berber languages of Ghadames and Awjila:

*ărwəy'to knead', General Berber ărwəy; Ghd. ărwək; Awj. ərwək

But similar alternations are also found in the 2nd person direct object suffixes whether they follow a vowel or a consonant, e.g. Tuareg 2sm ki ~ ik; 2sf kăm ~ im; 2mp kăwăn ~ iwăn

Proto-Berber *b might not have existed. becomes b in all dialects but Awjila and Ghadames pre-consonantally, while it becomes and later *h in other positions.

*ăβdəd, VN *ta-βădd-e, e.g. Tuareg əbdəd VN tehădde

Proto-Berber and alternate sometimes in post-vocalic position:

Most clearly in:

*a-məẓẓuɣ pl. *i-məzǵ-an 'ear',
*a-sămăɣ pl. *i-səmǵ-an'slave'

But perhaps one should also think of the 1sg. suffix of verbs *-əɣ which would correspond to Proto-Semitic *-ku. (Although of course even *ǵ is not te same as *k).

Proto-Berber *d alternates sometimes with in intervocalic position. The only obvious and well attested example I can think of is:

*a-ydi pl. *i-yəḍ-an'dog'

But the 2sg. marker of the verb vacillates across dialects. t-...-ət is found but most commonly t-...-əd, and occasionally, and with quite a random spread across the dialects we find: t-...-əḍ

Proto-Berber *t alternates with "vowel length" (= *H?) in many places . Verbs with final *t, esp. common in  Tuareg 3sm i-bbəkă-t 3pm əbbəken< *ăbbəkăH-ăn 'be spotted'.

But also the direct object clitic 3sm and 3sf show alternation: =t, =tăt after consonants =e, =et after vowels (Also outside of Tuareg).

Proto-Berber *f which ones probably was *p has lenited in all positions and shows no variation of this sort anymore.

This all points (although incredibly strenuously with many counter-examples and corrections due to analogy) to the following correspondences:

Ct ~ VH < Vṯ? (cf. Irish)
Cd ~ Vḍ
Cḱ ~ Vy < Vḵ?
Cǵ ~ Vɣ
p > f
Cb ~ Vβ (acts as a 'guttural' in many dialects)

The voiceless stops may have become voiceless fricatives while the voiced stops, while the voiced stops became 'emphatic'/pharyngealized consonants. I'm not sure how to understand the voiced series phonetically. Does not seem like a natural development.

*b is a little bit of an odd duck was it originally an emphatic ? This might explain its reflex as ggʷ~ww sometime (which has a velar(ized) component)

===

If anyone still reads this blog, let me know if you find this 'train of thought'-type blogposts interesting. If so, I might do them a bit more often.


Viewing all articles
Browse latest Browse all 104

Trending Articles