Kabyle dialect geography and the Kutama-Zwawa divide

Recently I came across A. Basset's Etudes de Geographie Linguistique en Kabylie (1929) - an interesting if incomplete work mapping the distribution of different body part terms across the Kabyle Berber-speaking region of eastern Algeria. The variation is significant, but I noticed a persistent trend: if there was any variation at all, the small Kabyle-speaking area east of Bejaia very often seemed to have a different term, or terms, than the rest of Kabylie. For example, the whole rest of the area has either aqerru or aqerruy as the normal word for head; this small eastern area instead has both ixf and akerkur. Almost the whole area has allen for "eyes"; the far east has taTTiwin. For "ear", everywhere has amezzugh, except the far east, which has imejj. For "knee", variants of tageshrirt (or Arabic borrowings) are nearly everywhere except the far east, which has afud. What's up with that?

A quick look at Ibn Khaldun suggests an explanation. In his History, he outlines the locations and notional genealogies of the principal Berber tribal confederations of his time. He describes the Zwawa - a name more generally associated with Kabyles - as extending through the mountains from Dellys to Bejaia, and the much larger Kutama group as extending throughout a wide area (the northern half of which is now Arabic-speaking) stretching from the Aures Mountains to the coast between Bejaia and Buna (modern Annaba), as well as including scattered groups outside this range, around Dellys and in Morocco (modern Ketama in the Rif.) (He personally inclined to the view that the Zwawa were in origin a subgroup of Kutama, but notes that this was not generally believed.) In other words, the division between Kutama proper and Zwawa lay around about modern Bejaia - exactly where the suspicious isoglosses I noticed seem to be. The next question: where these far eastern dialects diverge from the rest of Kabyle, do they resemble Chaoui?

(See الخبرعن كتامة من بطون البرانس and الخبر عن بني ثابت for the Ibn Khaldun quotes.)