Issues with Lexical Comparison of Berber within Afroasiatic

I have just finished a really inspiring and interesting course at the Leiden Summer School named "Reconstructing Afro-Asiatic" together with Benjamin Suchard, Marwan Kilani, Ahmed Sosal and Maarten Mous. A big aim of the course was to introduce an audience (and, to be honest, each other) to the specific challenges we see in the reconstruction of our respective Afro-asiatic language families of specialty, Berber (Me), Semitic (Suchard), Egyptian (Kilani) and Cushitic (Sosal and Mous).

In the past years a contingent of Leiden scholars have made steady headway into pre-Protolanguage reconstructions of especially Berber and Semitic, and have found a number of really interesting and strikingly specific morphological commonalities especially between Berber and Semitic, but on occasion have also lead to notable comparisons with Egyptian (Kossmann & Suchard 2018; Van Putten 2018; Suchard & Groen 2021; Suchard 2021). This is an exciting development, and has made me much more optimistic about the prospects of Afro-Asiatic reconstruction. And there is much more than what is presented just in these papers especially with Semitic, but also ni other languages. We can reconstruct quite specific plural formations in common between Berber and Semitic (e.g. Ar. fāris pl. fursān vs. Proto-Berber *a-ǵăllid pl. *i-ǵəldan; Ar. kalb pl. kilāb versus. Proto-Berber *a-ǵălzim pl. *i-ǵəlzam), the prefix conjugation, the suffix conjugation and adjective formation (article on this forthcoming by me very soon!), the shape of the infinitive PSem. *parās³-  PBerb. *a-lămad, and of course the feminine marker *-t.

Still, we are stuck with an issue: even though we can reconstruct really significant amounts of specific morphology, but vocabulary has lagged behind. In my fairly strict estimation, I've so far found about 30 lexical comparisons between Semitic and Berber (my strongest language families of the bunch) that I actually consider compelling. Yet, Gábor Takács who by now has published 15 articles on "Some Berber Etymologies" speaks of an "immense Afro-Asiatic heritage in the Berber lexical stock" and has collected no less than 609 (!) comparisons. That's 20x more than what I have found! Militariev has found 518 entries in his records. Did I not look hard enough?

No, the issue is that I find almost none of them convincing. Militariev's dictionary is, as far as I'm concerned, completely useless. As Lameen Souag has shown years ago, Militariev really lacks the required knowledge of Berber to even be able to recognize obvious cognates within the language family. Takács' work is significantly less problematic in this regard. As a whole he typically correctly identifies cognates within Berber. But Takács has refused to engage with Berberists that have worked on Proto-Berber reconstruction, despite important treatises on the topic having existed for several decades now (they, in fact, typically go completely uncited).

Even disregarding typical methodological issues:

• Comparing isolated words in one language family with isolated words in another;
• allowing for absurd amounts of semantic leeway
• Assuming ad hoc "root extensions" without explanation whenever you have a consonant too many or too few, etc.

the etymologies simply fail to connect with anything we know about the proto-language.

Root Structure and Why it Matters

Proto-Berber, like Proto-Semitic, as a rule has triradical roots, and like Semitic a number of consonants can be thought of as "weak" consonants which leave traces through some vowel alternations in most languages, but are mostly lost. Most notable of these are *β (also sometimes reconstructed as *h) and *ʔ. Clear retention of these consonants in some form or the other are most notable in a number of archaic peripheral languages such as Ghadames (Western Libya), Zenaga (Mauritania) and Tuareg (Southern Algeria/Libya and Mali, Niger, Burkina Faso). Ignoring these and their regular reflexes that have been detected, will on occasion leave you with only one root consonant where there were actually three. Thus the verb like Central Moroccan Tamazight aḍu 'to fold' may then be reconstructed as ḍ. But comparison with Ghadames odəβ 'id.' unambiguously shows that we should reconstruct this word as *ăʔḍəβ. Of course, if you start looking for cognates assuming there is only one consonant (or reconstruct the second one as the wrong one) in this root, rather than three rather specific ones, it is much easier to find comparisons. If you don't pay close attention to established sound correspondences by Berberist colleagues, you may likely even miss that this specific root also underlies CMT t-aḍuf-t 'whool' < *ta-ʔḍuβ-t.

That's not all, it has long been recognised that a class of Berber verbs that have an initial geminate such as *ăqqəd 'to burn' seem to get that geminiate as the result of some kind of assimilation with a labial element, as this resurfaces in the causative *əssuɣəd. A Pre-Proto-Berber root *w-q-d seems perfectly plausible, and is in fact very attractive, since Semitic has a root just like that with the same meaning! When this suits Takács, he is sure to mention this kind of comparison (as is the case for this), but far more often than not, the initial geminate is ignored, as is the case in *ăṭṭəs 'to sleep', which is taken to be from an Afro-Asiatic root *ṭs and compared with Gəʿəz ṭasasa. 'to be tired, weak'. Not even a bad comparison, but it is an unresolved problem with the root structure, which is not commented on at all.

Moreover, besides 'lost consonants', in Berber, long vowels appear to be an integral part of the root structure, and verbs can have "long vowel" slots. Let us take the verb akəl 'to go by foot' as an example. This verb in terms of apophony works just like a 'triradical root' except that it has a long vowel where triradical roots have a short vowel, compare *ăβdəd 'to stand' e.g.

Aorist *akəl Perfective *ukăl Verbal Noun *t-ekle
Aorist *ăβdəd Perfective *əβdăd Verbal Noun *te-βădde

Indo-Europeanist readers might be tempted to reconstruct a lost "Laryngeal" here: *ăHkəl Perfective *əHkăl VN *te-Hăkle, and indeed they would not be the first to do so, Prasse did so already in the '70s. But of course that gets ignored whenever it is inconvenient. It does in fact occasionally yield great comparisons though, such as *aləy (< *ăHləy) "to go up" alongside the Semitic root *ʕ-l-y "to be high, go up, etc.".

This is valid, although it does not seem all stem-internal long vowels can be readily explained in such a manner. From a proto-Berber descriptive perspective, it is usually more elegant to see them as long vowels slots in the root. But we're talking pre-proto-Berber here, and it's fine to speculate about it, even if not all the facts are in yet on all the origins of stem-internal long vowels. What we cannot do however, is just ignore those long vowels, and pretend like these are just "normal" biradical roots. Berber has those too (e.g. *ăǵər 'to throw') and they behave differently.

This does not mean that all long vowels must go back to lost laryngeals, but you do have to account for them, and not pretend like they are a Berber-internal issue that we have to go and figure out once the Afro-Asiatic correspondence has been established. That is doing the comparative method the wrong way around. Now, admittedly, in better-known language families every historical linguist does the comparative method "the wrong way around" every now and then, but that's typically done 1. to resolve an issue in a lower branch, rather than just leave it; 2. when the sound correspondences have been well-established and analogies well-understood. I don't think we can claim either of those things for Proto-Berber and definitely not for Proto-Afro-Asiatic.

We still have a long ways to go within Berber, and it is unrealistic at the moment to resolve all the issues that are present in comparing Berber to other Afro-Asiatic languages. But at the very least an attempt at reconstructing the Proto-Berber or Pre-Proto-Berber form that is assumed to underlie the comparanda would much more transparently (also for the Afro-Asiaticists themselves) show the problems and questions at hand. An honorable mention goes to Václav Blažek who pays very close attention both to Maarten Kossmann's previous etymological work AND Prasse's excellent "triradicalizing" internal reconstructions of the Berber (actually Tuareg) verbal system. Those reconstructions listed there could receive a bit of updating with more specific radicals than *H (Prasse's *H subsumes *ʔ, *β, and the long vowel slot), but it exposes the questions, and in my opinion and finds a couple of compelling comparisons that would not be found if you just glossed over these problems (I quite like the Berber w-d-r 'to live' compared to the Semitic d-w-r root for example).

Comments

[David Marjanović]

*β (also sometimes reconstructed as *h)

Any chance this was a *p once?

[David Marjanović]

...or its emphatic counterpart even, given that there's a *f in the system, IIRC.

[PhoeniX]

*β almost certainly goes back to *b at an earlier point (Proto-Berber has very few compelling examples for a *b, so there's a gap in the system there). Also some good cognates with Afro-Asiatic languages seem to suggest it's a *b!

[David Marjanović]

Oh, good. What's the reason for reconstructing it as [h] then – just "we need a consonant that disappears easily"?

[PhoeniX]

Because it surfaces as /h/ in Tuareg and disappears in most Northern dialects with compensatory lengthening of preceding short vowels. Intervocalically it sometimes surfaces as /ww/ or /ggʷ/ in Northern dialects though.

In Ghadames it is /β/ and in Awjili it is /v/, and for some dissimilatory patterns it clearly behaves as a labial (patterning with /m/, /f/ and /b/), which seems difficult to explain if it were *h. Therefore all still living historical linguists of Berber I think are in agreement that *β was some kind of labial. :-)

[David Marjanović]

That's good enough for me, thanks!