In this erudite and well-documented study, the author questions the linguistic unity of the variegated languages spoken in the mountainous regions comprised between the basin of the River Oxus, and that of the Indus, which are usually thought to draw the linguistic divide between the Indian languages to the East, and the Iranian languages to the West.  This analysis is supported by a joint appeal to the methods of comparative linguistics and typology, as well as by a historical outlook on the past evolutions of the local idioms.  Though they have grown very distinct elsewhere on the continent, the Dardic, Pamir and Kafiri languages to be found in this area have developed a number of linguistic features in common, three of which the author examines in the following order: 1) the evolution of transitive verbal clauses in the past tense, 2) the structure of the determinative complement, 3) the double stem of personal pronouns.  Point (1) concentrates on the local outcome of two major historical shifts which followed one another.  The first was carried out in almost all of the old Indo-Iranian languages, which developed a specific construction for transitive verbs in the past perfect, namely, the “ergative” construction.  Among the geographically marginal languages in question, however, a later shift was yet to follow:  Verbal clauses tended to be remodelled in line with the paradigm of the present tense, and the “agent” (or logical subject) consequently regained the status of a real grammatical subject, all of which meant the loss of ergativity.  Whereas this second shift was carried out as early as the Middle Ages in some languages (like Soghdian), in other languages it is just only in progress today.  Now, if one wishes to know these languages better, F. Jacquesson argues, one is bound to investigate “the logic of forms in the past tense”.  Another specific feature is shown by point (2).  Whereas the izafa construction came to dominate the Iranian expression of the determinative complement early on, it never imposed itself in the Pamir.  Isolated as they were, local languages resisted the expansion of this “late” innovation of the Occidental Iranian languages, and kept favouring the alternative construction of compound expressions which, in the absence of a linker between determiner and determined, entails the inversion of word order.  Lastly, the issue of etymologically divergent stems for the “short” and “long” forms of personal pronouns raised in point (3) questions the particularities of the Dardic and Nuristani languages.  As a conclusion, the author underlines the liveliness of the Pamir languages.  According to him, their peculiarities testify not so much to a conservative or even archaic feature, as to the vivacity of an independent development.  Indeed, showing the linguistic unity of theses languages makes a strong case for the recognition of the human coherence of the peoples in the Pamir highlands, and belies the political fragmentation they are subject to.

Justine Landau, New Sorbonne University, Paris
CER: I-6.2.C-558