PhDniX's blog

A fetus in the belly!

Currently I'm working on my master Thesis which focuses on biradical nouns of the shape C(e)C¹ and CVC. One of the words I've looked into is the Kabyle word for fetus.

Kab. tadist (ta) pl. tidusin (td), tidusa (td)

The second plural formation is probably the original one, it's quite rare to see it. The first plural is very productive.

Now I've gone looking for cognates, and I have found some exclusively in Libyan Berber languages, but the meaning of these words struck me as odd.

Ghadamès tadist, tadiss pl. tədisēn  'belly', Aujila tedûšt 'id.', ElFoqaha tadîst 'belly'

I'm somewhat troubled by this difference in meaning. You'd have to assume that since a fetus is unborn, and still in your belly, you could refer to it as 'your belly'. That's a very playful way with words, so playful in fact, that I find it hard to believe that this would become the only meaning of the word.

It's a shame that it's specifically the Libyan dialects that have this word, especially Aujila and ElFoqaha are very badly described for obvious reasons. So badly in fact, that we're not even sure what Paradisi, the main source of both these languages, meant exactly with his rather exotic transcriptions. He has about 15 different ways to denote vowels, which is odd in a group of languages where, generally, there are vowel systems that range from usually 3/4 to about 7 in the most extreme cases.

It would be interesting to see if maybe the words in those languages have a double meaning. That would mean the development is particularly ancient which would give a bit more space for it to take place.

I'm curious if any of my readers have come in contact with such a development before in other languages.

 

---

[Edit] I may not have access to Libyan situated Berber speakers, but I do have access to Kabyle speakers, so I contacted a friend with experience in Kabyle berber who asked an informant what they thought tadist meant.

The speaker says tadist means 'pregnancy' and he gave the following meanings example sentences.

aqlin s tadist 'I am pregant' (litt. I am with pregnancy)

tameṭṭut attsen s tadist 'The woman is pregnant' (She is with pregnancy)

This gives a nice parallel in Tamazight where adis (the masculine form) can mean both 'belly' and 'pregnancy'. That it's found in Tamazight (and Tarifit as Lameen points out in the comments) shows that it's not only Libyan. I wonder why adis slipped past me when I worked through Taifi when collecting the data. I did find it in Amaniss' digital dictionary though.

In fact the idiomatic nature of the sentences mentioned above would still work if the meaning was 'belly'

I am with a belly, the woman is with a belly = pregnant. Then, only when the word lost its use as belly (because it was replaced by taεbuṭṭ) the word came to mean pregnancy.

A semantic shift from 'having a belly' to 'pregnancy' is something I find absolutely acceptable. So did Dallet really mistake 'pregnancy' with 'fetus', I wonder how that happened.

Maarten Kossmann (p.c.):

Well, maybe it was a taboo term for foetus. I would in principle believe Dallet, who worked, among others, with Kabyle doctors.

---

 

¹ I generally use e to transcribe the schwa, which often can be a dummy vowel.

 

The thing with Causa

As you may know, a considerable amount Latin loanwords have snuck into the Berber language over the years. Some forms you can still find the masculine -us  from Latin masculine singular nominative ending from the classical language as in Tamazight tafullust 'chicken' from Latin pullus 'little animal' but 'chicken' in Vulgar, while other words show a stage of Vulgar Latin where this final s is lost such as Tashelhiyt urtu 'garden' from Latin hortus 'id.'

All in all, these words are old and stem from the time that the Romans came to resided in North Africa around the year 0. One of the words the Berbers have loaned from the Romans is the word causa 'legal case, trial; cause, purpose' (But in VLat. 'thing' cf. Fr. chose)  an example of a reflex of this word is Tamazight taγawsa 'thing' (with feminine prefix ta-).

The loan works great and we see the correspondence Latin c to Berber γ consistently in front of a in other words too eg. Nefusa γasru 'fortress' from lat. castrum 'id.' We find Berber k in front of i and u as in Tashelhit ikikr 'chickpea' from Latin cicer 'id.' and Kabyle afalku 'falcon' from falko 'id.' But now that we have this correspondence, the question is why do we have this correspondence.

γ (voiced uvular fricative) isn't exactly an obvious candidate to represent a sound that was probably a voiceless velar stop in Latin, why not just use k or γ's long counterpart qq? Well, I think some internal reconstruction can help us further. In Berber every consonant has a long counterpart, thus we find pairs such as:

• k ~ kk
• s ~ ss
• ḍ ~ ṭṭ
• w ~ ggʷ
• γ ~ qq

It seems quite obvious that γ was originall a *q which was later lenited to the form it has now. Then the distirbution becomes a lot clearer. k for high vowels and q for low vowels (q often has a relation with low vowels, cf. Arabic where a is generally pronounce [æ] but [ɑ] in the vincinity of q).

As clear-cut as this seems, it actually presents a problem. If we look at the reflexes of Berber γ we find γ in almost all Berber languages, exceptions have to be made for Ghadamès Berber and Zénaga Berber which reflect this consonant as ε (voiced pharyngeal fricative) and Ɂ respectively.

So although at an earlier stage γ was certainly q, by the comparative method one has to reconstruct Berber γ for Proto-Berber as *γ. So does that mean Berber was still one language at the time the Romans? Surely that is unlikely, there is enough indication that the Berbers had spread over a large part of the current places they live already, and being spread over such a large piece of land makes it quite unlikely that their language was uniform.

I personally think that, much like how many developments in Germanic remained to developped accross languages due to their close relatedness, one has to imagine such a development for Berber as well. At one point after the arrival of the Romans, despite an already existing differentiation between languages, a sound law *q > *γ must have occurred accross all Berber dialects at the time.

One Berber Language is a candidate to have not undergone this change of *q > *γ. And that is the isolated and in many ways very unique Berber language of Mauritania: Zénaga Berber.

Zénaga Berber is the only Berber language that, as far as we can tell, clearly retained the Proto-Berber *Ɂ . There is some proof for this consonant in all Berber languages, especially in Biradical verbal roots of the type CC, and in the vocalisation of Ghadamès, but nowhere this consonant is actually retained as a consonant. Zénaga also retains, though less clearly, the Proto-Berber *β.

This archaicity in the phonology (and  also quite different morphology) seems to indicate that Zénaga was isolated from the other Berber languages at quite an early stage. A shift from *γ to Ɂ is by no means impossible (I can imagine *γ > *ε > Ɂ) but *q > Ɂ provides a much more elegant and direct solution (this is a very common sound shift, as can be seen in Egyptian Arabic for example). So could it be that indeed Zénaga was isolated from the other Berber Languages at such an early stage that this Pan-Berber shift went straight past it? Hard to prove, but pretty attractive.

As a last side note a little methodological question for you to ponder about: If the comparative method tells us that we are to reconstruct *γ for Proto-Berber, while there is clear proof in Latin that this can't have been the case at a stage of the language that doesn't require internal reconstruction what do you do? Do you reconstruct *q or do you stick with the methodologically correct *γ that represents a hypothetical construct?