On the modal markers -a and -ka in Moksha Ilya Yegorov Moscow State University Department of Theoretical and Applied Linguistics

About Mordvinic Languages Uralic Finno-Ugric Finno-Mordvinic Mordvinic Erzya Moksha Agglutination with some cases of cumulation In contact with Russian since 9 th century

-a and -ka in Mordvinic: State-of-art “A Grammar of Mordvinic Languages” (ГМЯ 1962: 279): there is an -a particle in Mordvinic which “is easily affixed to the Imperative forms” and gives them “a shade of tenderness and softening” “A Historical-comparative Grammar of Finno-Ugric Languages” (Майтинская 1979: 15): a -ka particle for the softening of the Imperative forms has been borrowed from Russian, and “the imperative -k formant and the -k part of the -ka particle merged” What is the case with these particles and what is the relation between them (if there is any) ?

The Russian -ka One of the main functions is the non-categorical imperative formation:

Field Study Two Moksha villages: Lesnoe Tsybaevo and Lesnoe Ardaševo, Temnikovsky district, Mordovia, summer 2015 Both -a and -ka exist in the investigated dialect and can modify Moksha Imperative forms A number of peculiar morphosyntactic features of these formants has been revealed

1. Distribution The following system of distribution has been drawn out: Form ending in a vowel is affixed with a -ka: Form ending in a consonant is affixed with an -a (-ɛ after palatal): (2) (3)

Furthermore, a form already equipped with an -a can be overaffixed with a -ka, if it doesn’t result in a -kaka sequence: Plain imperativesingle particledouble particle (a in mora-s’t’-a-ka is regularly restituted from ɛ before non-palatal k) According to the informants, there is no semantic difference between the single- and double-particle variants

Hypothesis The non-categorical imperative marker -ka is borrowed from Russian to Moksha -kk- is simplified to -k- at the morpheme boundaries (as the imperative affixes of IMP.SG and several of the objective conjugation often end in -k ) -a which is a variant of -ka emerges -a becomes a postconsonantal allomorph of -ka The -a-ka structure is pleonastic, probably because -a and -ka are reinterpreted as different markers One more -ka cannot be added to mora-k-a as it already exists in the phonological form of this word Unlike mora-s’t’-ɛ !

Moksha Imperative paradigm Only second person subject (and non-second person object) Subjective conjugation: Objective conjugation: Object1SG1PL3SG3PL SuSG-(ə)mak-(ə)mas’t’-k-it’ PL-(ə)mas’t’ -(ə)s’t’ SG-t/-t’ after consonants -k after vowels PL-(ə)də -a -ka -a -ka -a -ka

2. Combinatory Power It has turned out that -ka can be added not only to all the members of the Imperative paradigm, but also to some other verbal forms Optative: third person forms can be affixed with -ka (4) The semantics of the ka-equipped optative can exhibit some peculiarities which will be shown later

Note that the Optative is a somewhat marginal category in the contemporary Moksha, especially its first and second person forms The Conditional forms are very common to be used instead Maybe it is due to the influence of the Russian language which exhibits no special optative but has a multifunctional Conditional mood

Non-past + ka: forming a hortative mood? Also -ka can modify the first person forms of the Non-past tense of the Indicative mood (5) This case tends to be the most interesting There are three synonymous forms which exhibit inflectional affix doubling (6b) and its externalization (6c): (6) a. mol’a-n-ka b.mol’a-n-ka-n c.mol’a-ka-n ‘what if I go’ (hortative)

All the variants (6a-c) seem to be equivalent, though the acceptability depends on a certain verb and is fluctuating among the informants The combination has a hortative meaning, similar to its Russian counterpart

Externalization Haspelmath 1993: The diachronic externalization of inflection [Linguistics 31, ] A diachronical change in the order of affixes generally motivated by the Inflection-outside-derivation principle: The typical situation is when inflectional affixes are trapped in an internal position as the result of the grammaticalization and affixation of an uninflected element, often a clitic particle A morphologically complex word is preferred if its inflectional affixes are further away from the root than its derivational affixes.

Externalization R-Infl=Post >…> R-Aff-Infl A feature of such changes is intermediate hybrid forms with the inflection both in internal and in external position: R-Infl=Post > R-Infl=Aff=Infl The existence of these pleonastically affixed hybrid forms can be explained, according to [Haspelmath 1993], by the principle of conservatism: the innovations cannot be too radical, and speakers prefer a more conservative form that is closer to the earlier well-known form FEET FEET-S FOOT-S

Externalization One of the common examples is the externalization of the case inflection of the pronouns with postfixed indefiniteness particles ra ‘what’  ra-me ‘anything’ dat. ra-s-me > ra-s-me-s > ra-me-s adv.ra-d-me > ra-d-me-d > ra-me-d Russian exhibits rather rare case of antefix internalization, cf. кое у кого vs. у кое-кого GEORGIAN

Hypothesis Moksha language, which had no specific means to express a hortative meaning, adopts Russian NPST.1 + -ka construction -ka is grammaticalized from a clitic to a Hortative marker This is probably rather recent process since we can see the stages of the grammaticalization synchronically, different informants preferring different variants in a particular situation INFL=KA INFL=KA=INFLHORT-INFL

Semantic feature As it was mentioned, forms with -ka sometimes exhibit some peculiarity in semantics The Optative forms modified with it have a strong connotation of warning or even threatening Most informants explain the difference somehow like this: (7) vs.

This meaning can also be found in the Imperative+ka forms, compare (8) and (9): (8) (9) Maybe this is some kind of grammaticalization of the ironical use of this construction (Certainly the problem has to be elaborated in more detail)

Conclusion The -ka particle, having been adopted from Russian by Moksha language, has had further development in an original way It has got some features which it didn’t use to have in the donor language

Acknowledgements The author is very grateful to V.A. Plungian, M.A. Kholodilova, A.A. Kozlov and his other colleagues for their help during the course of the study.

References Haspelmath 1993 — M. Haspelmath. The Diachronic Externalization of Inflection. // Linguistics 31: ГМЯ 1962 — М. Н. Коляденков, Р. А. Заводова (ред.). Грамматика мордовских (мокшанского и эрзянского) языков. Ч. I. Фонетика и морфология. Саранск: Мордовское книжное издательство, Майтинская 1979 ­­­­­— К. Е. Майтинская. Историко- сопоставительная морфология финно-угорских языков. М.: Наука, Плунгян 2014 — В. А. Плунгян. Экстернализованные аффиксы: материалы к типологии. Доклад на Чтениях, посвященных 50-летию Лаборатории автоматизированных лексикографических систем НИВЦ МГУ им. М. В. Ломоносова. Москва,