The Position of Clitics in Udi

Abstract It has been claimed that endoclitics do not exist (e.g. Klavans 1985). On the other hand, it has been suggested that Pashto is a language in which clitics occur inside the verb. Kaisse (1981, 1985: 132 – 43) showed that in Pashto, the clitics at issue always follow a morpheme; they do not occur inside a morpheme. Roberts (1997) has argued recently that the morpheme at issue is, in fact, part of the verb; thus on his analysis the clitics are endoclitic, of the intermorphemic variety. In this chapter it is shown that in Udi, too, intermorphemic clitics occur. It is shown in addition that in Udi clitics occur inside monomorphemic verb roots. As discussed in Chapter 1, the position of the PM in Udi has never been fully described. The most complete statement is that of Schulze (1982: 168 – 9), cited in §1.4.1, and repeated as Table 6.1 below. (Recall that PZ = my PM, V = verb, T = TAM marker, Vn = the nominal incorporated into the verb, and HV = light verb.) Examples I have elicited and collected in texts require us to recognize additional patterns; 6’ is only a variant of Schulze’s pattern 6.