The paper addresses the question of state alternation in Amazigh. It provides a detailed description of the morphological and phonological changes that target CS nouns. Then, it discusses the environmental conditions under which CS and FS occur. The paper argues that these changes go beyond the morphological and phonological system of the language. Later, it outlines the main previous studies about the phenomenon and presents an alternative complementary approach that takes a reconciliation path between Lahrouchi’s templatic analysis (2013) and El hankari’s syntactic model (2014) to account for state alternation.  It argues against Achab’s analysis (2000) of the initial vowel as a determiner since [+ definite] in Amazigh is not an inherent feature of DP and it is contextual. Moreover, a further argument against DP analysis is that in some languages demonstratives, quantifiers, possessives do not occur with (in) definite articles, unlike Amazigh. I postulate that CS is a language-specific phenomenon that is not only the result of phonological operations but also of certain syntactic operations. I argue against El hankari’s analysis of d ‘and’ as a preposition, that CS marks its complement. The same goes for bu and mu ‘the one with’. I postulate that buNouns contain an inflectional morphology inside them, which make bu a derivational affix. A phrasal affix which attaches to words that are already inflected (see also Bensoukas 2015). The affixation of bu CS marks its selected noun. I hypothesize that Amazigh does not allow the consonant cluster /tw/. In this line of analysis, the initial vowel does not change when it is part of the stem. Hence, Amazigh has not only consonantal roots but vocalic ones as well. I argue that a noun cannot be marked for CS in isolation. It is marked by the functional head that c-commands the NP. The head is not only T or P, but it can be also a coordinator; a numeral or a phrasal affix.

This article studies the derivational system of gender and number in Ait Atta variety of Amazigh language. Thus, the following claims are made: first, gender is overtly marked on feminine nouns by the prefixation of the gender morpheme t-. The paper argues that [t…t] is not a circumfix or a discontinuous morpheme, but it is an asymmetric inflection. In the derivational system of gender, large majority of nouns allow for gender opposition. However, mass nouns allow only for one lexically determined gender and number. Moreover, masculine has no overt realization in Amazigh language. Vocalic initial nouns fall into the category of masculine nouns. I argue that the initial vowel is a nominal marker. I submit that the noun, in general, consists of maximally three main parts: a prefix, a lexical base, and a suffix. Second, the majority of Berber noun stems involve, at least, one vowel in addition to the prefixal vowel (e.g. ‘a-ḍar’ foot, ‘a-funas’ bull, 'a-ɣrum’ bread). Third, when the plural noun is specified as [+feminine], it takes the gender morpheme t-. Fourth, number takes the form of a feature assigned lexically to the noun. The study provides a templatic analysis to account for internal noun change. It implements Lahrouchi and Ridouane (2016) analysis of diminutives and plurals in Moroccan Arabic and argues that sound plurals in Amazigh language are associated with standard Num projection, while id/istt-plurals are associated with lower in the structure with n projection.