Thread by @PhDniX

"The feminine endings *-ay and *-āy in Semitic and Berber" is finally out in BSOAS. It's an article I'm rather happy with; very technical and linguistic, but I'll try to give a summary here. #semitic
cambridge.org/core/journals/…
academia.edu/37112479/The_f…

Berber and Semitic clearly have a shared common ancestor; usually called Proto-Afro-Asiatic. A macrofamily that is difficult to reconstruct, but at the same time the parallels are so striking that it cannot be a coincidence. One of these parallels is the feminine formation.

The general feminine ending is *-t in both Berber and Semitic; Semitic also points to a by form *-at (which becomes the regular form in Arabic). It has long been recognised that Berber and Semitic share this. Besides this in both languages we find a different feminine ending.

In Semitic there is a feminine ending *-ay- which is clearly seen in Arabic, but has traces in other Semitic langs too (especially Classical Ethiopic). This is especially productive in the feminine form of the elative: Cl. Ar. ʾakbar fem. kubrā < Proto-Ar. *ʔakbaru fem. *kubrayu.

Besides this there is also a feminine ending *-āy- which shows up in Classical Arabic -āʔ (e.g. ʾabyaḍ 'white' fem. bayḍāʾ); Traces of the original pronunciation are still found in Old Arabic ṣḥry < *ṣaḥrāy 'desert' and some Yemeni dialect, e.g. bayṯ̣āy 'white (fem.)'

I argue that the Arabic *-ay- should be connected with the Classical Ethiopic *-e ending; and *-āy- with the Classical Ethiopic *-a ending.

Functions of these endings are similar: mostly verbal abstract, and does not work as a gender derivational form the masculine (unlike *-t).

There seem to be enough traces Semitic-internally that we can reconstruct it for Proto-Semitic with a couple of typical morphological features: 1. It always takes a stem CvCC or CaCaC; 2. It's regular plural replaces *y with *w: *ṣaḥrāyu pl. *ṣaḥrāwāt.

Berber also has two feminine endings besides the ubiquitous *-t, namely *-e and *-a, which mostly make deverbal abstract nouns and some concrete nouns, e.g. taɣm-a 'thigh', tayur-i < *ta-gəHr-e. I argue that these likely ultimately go back to *-ăy and *-ay.

*-ăy and *-ay are perfectly equivalent to Semitic *-ay and *-āy; and like in Semitic, the stems that take these endings have the shape CvCC; and like in Semitic there sound feminine plural alternates *y with *w: *ta-năkr-a pl. *ti-nakr-iw-en 'standing up'.

The morphology, behavior and semantics are so similar that I argue that the similarities cannot be coincidence; These must go back to a shared ancestor. In the article I call this shared ancestor Berbero-Semitic. It may go further back to other Afro-Asiartic languages.

My expertise, however, lies within Berber & Semitic, and their similarities on all kinds of levels (also the verbal) are so great that a shared 'subbranch' of Afro-Asiatic at least seems likely. I did not want to overstep my expertise, a problem one often sees within Afro-Asiatic

It is hoped that this paper will motivate others to look for similar formations in Chadic, Cushitic and Egyptian; Perhaps this formation turns out to be Proto-Afro-Asiatic proper. Very recently another paper on Berbero-Semitic came out (submitted on the same date!).

My good friend and colleague Benjamin Suchard and my mentor Maarten Kossmann wrote a brilliant article on the reconstruction of the Berbero-Semitic verbal system, here too the similarities between Berber and Semitic are striking. cambridge.org/core/journals/…
academia.edu/36081227/A_rec…

Hopefully this summer I'll be able to dust off another paper on this suddenly strangely active field of Berbero-Semitic reconstruction: An article on the Berbero-Semitic Adjectival formation -- once again many striking morphological parallels.