ABSTRACT
This chapter will give an overview of the typologically relevant features of Proto-IndoEuropean. By “typologically relevant” we mean those features that are cross-linguistically rare, or those that are correlated with other features, thus defining a linguistic “type”, a set of languages sharing a number of logically independent features. Moreover, we will consider those features of PIE that show genetic and/or areal patterning, in the sense that they statistically tend to be more common in certain linguistic families and/or language areas than in others. Particularly relevant are diachronically stable features that tend to be shared by most genetically related languages of a single family. Such features are important for questions about the deep genetic relatedness of Indo-European and about early language contacts between PIE and its neighbours. It has been long suggested that Indo-European languages and Uralic languages are related, forming a putative Indo-Uralic family (e.g. Collinder 1960), and Vladimir Illič-Svityč (1971-84) included both Indo-European and Uralic into a wider family, “Nostratic”, which would also include Altaic, Dravidian, Kartvelian and Afro-Asiatic languages. Several variants of this hypothesis are currently in circulation, including the daring proposal by Joseph Greenberg (2000), who also includes Gilyak, Ainu, Chukchi-Kamchatkan and Eskimo-Aleut (but excludes Afro-Asiatic, Kartvelian and Dravidian) in his “Eurasian” language family. None of these hypotheses can be considered as proven, but many linguists believe that Indo-European is indeed related to Uralic, and that it is more likely to be related to Altaic and Kartvelian than, say, to NW Caucasian or Sino-Tibetan. As this chapter will show, PIE shares a number of typologically relevant features with Uralic and, to a lesser extent, with Altaic and other language families of Northern Eurasia. However, PIE also shares a considerable number of features with languages of the Caucasus, including not only Kartvelian – which is included in the putative Nostratic macrofamily – but also NE and NW Caucasian languages, to which it is probably unrelated. These shared features can be understood only as results of prehistoric language contacts, so they are very relevant for the question of the homeland of the speakers of PIE (Mallory 1989, Anthony 2007).
