Nyelvtudományi Közlemények 109. kötet (2013)
Tanulmányok - Simoncsics, Péter: Linguistic gestures: On negation, with special reference to the Permian languages 151
On negation, with special reference to the Permian languages 161 setéj ’let us not give; may we not give’ and - with a slight difference - also Udmurt medaz mini ’let him/her not go; may (s)he not go’, meda-z mi'nele ’let them not go; may they not go’ are not strictly speaking imperatives. These are syntactic structures (imperative VXSG2 of med- ’to wish, to want, to plan’ + indicative VXSG/PL3) expressing wish or desire and as regards their modality they are optatives rather than imperatives. From a strictly morphological point of view the cores of these structures are indicative SG/PL3 forms. The Udmurt forms of the imperative SG/PL3 are exceptions due to the stress (and possibly also length) on the final syllable where instead of an expected *meduz we have meda-z. The explanation is that stressed / long low vowels do not usually rise, according to the Donegan - Tálos - Abondolo hypothesis, cf. Abondolo (1996), Donegan (1978/85), Tálos (1983). On the other hand, ё-п set ’don’t give’, ё-пё set-ej ’you (pi) don’t give’ are true SG2/PL2 imperatives of formed with the “past” stem (i.e. the original) variant of the negative auxiliary ё-. Present vs. past and imperative vs. other moods are basic coordinates in verbal orientation and, partly, also complementary, at least in negation in Uralic: when present and past are distinguished by initial labialization as in Permian, then imperative is left unmarked, when imperative is distinguished by labialization as in Nenets, then we have the opposite case where present remains unmarked. The labialization of negative e- ~ ё- stems in Permian is an example of how a linguistic change can be triggered by extralingual gestures, where a facial grimace can become internalized and its phonic representation become part of the paradigm. In a broader context, taking Mari (and even also Mordvinian) into consideration, the nucleus of the labialization process must have been in Udmurt where the first step from e- ~ ё- to o- was taken by the alteration of only one feature in the vocalic stem, i. e. „labialization”. Udmurt, that is to say the centre, developed further by raising the tongue-level from mid to high. The periphery, Komi and Mari and, indeed, Mordvinian remained at the first stage, retaining the о-stem in the present tense. The case of Komi and Mari is, in this respect, simple and straightforward. Mordvinian is problematic where negation in the present is formed in a totally different manner, using negative particles with an a- stem, whereas in the past it retained the illabial (e- ~ i-) stem (except in the use of the alternative as- stem in Moksa). The archaic nature of Mordvinian is apparent if one takes a look at the testimonial past tense forms of the negative paradigm: Moksa Sgl i-z-әп ~ assn soda Sg 2 i-zs-t ’ ~ asst' ’... ’ Sg 3 i-s ~ asss~assz ’. Erza e-z-irt soda ’I didn’t know’ e-z-it’ ’... ’ e-s ’... ’
