On the 3rd person verbal mark er *-sV and the definite conjugation in Uralic languages.
1. There may be three conjugations in Uralic languages: indefinite, definite and reflexive (or indeterminative, determinative and reflexive-medial; for a long time the same conjugations have been known in Uralistics as subjective, objective and reflexive ones). Erika Kortvely has recently offered a systematic overview of Uralic languages, focusing on the number of conjugations in each one of them. Based on the use of various conjugations, E. Kortvely distinguishes four types in Uralic languages. 1) With one conjugation only. Actually, the determination of respective languages is rather problematic. It is quite common in both definite and reflexive conjugations that a language does not have a full paradigm. They have only a few and often regional verb forms which signal (at least partially) the existence of the conjugation. E. Kortvely writes that the only Uralic language with just one conjugation is Saamic. However, she will also consider the Finnish, Karelian, Mari and Permic languages as belonging to this group. 2) Distinguishing between indefinite and definite conjugations: Mordvin, Ob-Ugric (Mansi, Hanti), Southern-Samoyedic (Selkup, Kamass, Mator) languages. 3) Distinguishing between indefinite and reflexive conjugations: some Finnic languages, and more distinctly Southern Estonian and Veps. 4) Distinguishing between all three conjugations: Hungarian and Northern-Samoyedic (Nenets, Enets, Nganasan) languages (Kortvely 2005 : 27-42; cf. also Pusztay 1995 : 89-92; Keresztes 1999 : 12-14). As we see, it could not have been easy to make the distinction nor is its result distinctly univocal. Peter Hajdu has summarised the incidence of the 3P verbal marker *-sV in Uralic, pointing out that the suffix has widely spread in the languages with a definite conjugation. In the languages that lack the definite conjugation the distribution of the suffix is confined to some more limited sphere of conjugation, e.g., to some forms of transitive or reflexive verbs; besides, in that case the suffix often occurs in the imperative, optative or desiderative and conjunctive forms ([TEXT NOT REPRODUCIBLE IN ASCII] 1985 : 329).
2. Observing the incidence of the definite conjugation and the 3P verbal marker *-sV in Uralic, we can get the following picture. Finnic languages have no definite conjugation (see also Kortvely 2005 : 31-32), however, Lauri Posti has brought out that in case of these languages the verb in 3P may occur both with the (*)s-material (s ~ z ~ h ~ s ~ j) verbal inflection and without it (Posti 1980). The incidence itself is rather inconsistent and varied in various languages. We can see that in Southern Estonian such an incidence is divided between different verbs, e.g. jaga 's/he divides' : jaga-va? 'they divide' ~ kazva-s 's/he grows' : kazva-ze? 'they grow', as well as between different verbal derivatives, e.g., nage 's/he sees' ~ reflexive nau-s 's/he is seen'. From among 3P verbal inflection s-forms modern Northern Estonian generally knows only three naus-type verbal reflexive derivative forms: naik-se 'can be seen', kuuluk-se 'can be heard', tunnuk-se 'can be felt'. Traces of the (*)s-material suffix can be found also in Vote, e.g., eittii 's/he threw her/himself down' (< *heitti-hen). In Finnish eastern dialects and Veps the personal ending of the (*)s-material presents an independent paradigmatic reflexive conjugation, e.g., Finnish panee 's/he lays' ~ panek-sen 's/he lays her/himself', Veps peze-b 's/he washes (something)' ~ peze-se (pezi-he) 's/he washes her/himself' ([TEXT NOT REPRODUCIBLE IN ASCII] 1975 : 87-99; Posti 1980; Mikola 1984 : 398-400; Dahl, Koptjevskaja-Tamm 1992 : 22; cf. also Tauli 1966 : 166). L. Posti writes that in some Ingrian dialects in the 3P present the personal endings in reflexive verbs differ from those in other verbs, e.g., vatti-he 's/he dresses', l'aga-jan 's/he speaks' (Posti 1980 : 135-137; cf. Mikola 1984 : 399), cf. lukko 's/he reads' ([??]aaHeCT 1975 : 92). The same we see in Karelian, e.g., lugo-u 's/he reads' ~ haikostele-h 's/he yawns'. L. Posti also indicates that "The third person ending of the middle verbs -(k)sen, -hen is also found in Finnic as a passive ending, which in Late Proto-Finnic had the forms (pres.) to-[delta]a-k-sen, (imperf.) to-t-i-hen 'to bring'" (Posti 1980 : 113; cf. also Kortvely 2005 : 31-32, 38-39). In Finnic L. Posti sees traces of the onetime reflexive conjugation that was provided with 3P verbal inflection (Posti 1980 : 111-112).
Saamic languages have no definite conjugation either (see also Kortvely 2005 : 29), however, Mikko Korhonen indicates that a 3PSg/Du/Pl verbal inflection -s (identical with a respective possessive suffix) occurs in them only in the imperative, e.g., malesta(h) 'boil the meal' : malestu-s 'let her/ him boil the meal' (< Proto-Finnic-Saamic *-sa/sa, Korhonen 1981 : 260-263, 273, 279, 284).
We find only one language group in Finno-Permic languages, containing both the indefinite and definite conjugations--Mordvin (see also Kortvely 2005 : 36-37). A verbal form of the definite conjugation, as a rule, indicates the person and the number of an object. E.g., indet. conj. pali 's/he kisses' ~ det. conj. pala-si 's/he kisses her/him' : pala-sinze 's/he kisses us' ([TEXT NOT REPRODUCIBLE IN ASCII] 1974 : 318-319; [TEXT NOT REPRODUCIBLE IN ASCII] 1975 : 314; Pusztay 1995 : 89-91).
However, the traces of the 3P verbal marker *-sV exist also in those Volgaic and Permic languages that lack a separate definite conjugation--in Mari, Komi and Udmurt (see also Kortvely 2005 : 30-31). In Mari, in the 3PSg of the indicative present we find the variation -a ~ -es. This variation is divided among various verbs, e.g., kola 's/he dies' ~ tole-s 's/he comes' (Alhoniemi 1985 : 108-131). Generally, here the suffix -s is not considered as a trace of the 3P verbal marker *-sV but as a trace of an old preterite (or perfect participle) suffix *-sV ([TEXT NOT REPRODUCIBLE IN ASCII] 1974 : 324). The trace of the marker *-sV is observed in the Mari suffix -sV, -zV, primarily in case of the imperative and optative (desiderative), e.g., wi[delta][??]-z[??] 'let her/him lead' : wi[delta][??]ne-z[??] 's/he wants to lead'. Together, alternating with each other, the 3P verbal inflections -s and -sV, -zV occur rather generally in the whole of the Mari verbal paradigm. In most western Komi dialects some of the verbs are used in the 3P present/future singular without any inflection, some others--with the s-inflection, or with both variants, e.g., kile 's/he hears' ~ adza-s 's/he will see', mune 's/he goes' ~ muna-s 's/he will go'. Udmurt uses the 3P verbal inflection *-sV > -z not in the present tense but in the future tense, e.g., mino-z 's/he will go' (the present tense with -z is lacking in Udmurt altogether). (See e.g. [TEXT NOT REPRODUCIBLE IN ASCII] 1955 : 213-221; [TEXT NOT REPRODUCIBLE IN ASCII] 1962 : 194-219; [TEXT NOT REPRODUCIBLE IN ASCII] 1976 : 173-174; [TEXT NOT REPRODUCIBLE IN ASCII] 1985 : 245-246, 323; Hausenberg 1996 : 181; Csucs 2001).
In all Ugric and Samoyedic languages the definite conjugation does really exist (see also Kortvely 2005 : 32-36, 39-42). In Hungarian are apparently no univocally interpretable traces of the 3P verbal marker *-sV, although this suffix has been supposed to be a starting point for the 3P verbal inflections -a, -e, -ja, -je, -i that occur in the Hungarian definite conjugation, e.g. indef. conj. nez 's/he looks' ~ def. conj. nez-i id., indef. conj. lat 's/he sees' ~ def. conj. lat-ja id. ([TEXT NOT REPRODUCIBLE IN ASCII] 1976 : 394-399; [TEXT NOT REPRODUCIBLE IN ASCII] 1982 : 72; [TEXT NOT REPRODUCIBLE IN ASCII] 1985 : 327; Keresztes 1999 : 50-52; but cf. e.g. Redei 1966 : 118-127; Pusztay 1995 : 89).
In Ob-Ugric and Samoyedic the 3P verbal inflection of the definite conjugation is (*)t-material (t ~ d ~ [delta] ~ r ~ l). In Ob-Ugric (Mansi and Hanti) and Northern-Samoyedic (Nenets, Enets and Nganasan) languages the verbal form indicates also the number of the object. E.g., Mansi indef. conj. toti 's/he carries' ~ def. conj. toti-te id. (one object) : totija-[gamma][??] id. (two objects) : totija-ne id. (many objects) ([TEXT NOT REPRODUCIBLE IN ASCII] 1976 : 298-300; Kortvely 2005 : 33), Hanti indef. conj. werl 's/he makes/works' ~ def. conj. werl[??]-li id. (one object) ([TEXT NOT REPRODUCIBLE IN ASCII] 1976 : 326-328); Tundra Nenets indef. conj. mada 's/he cuts' ~ def. conj. mada-da id. (one object) : mada-xaju-da id. (two objects) : mada-j-da id. (many objects) (Castren 1854 : 388; Kortvely 2005 : 41), Forest Nenets indef. conj. d'ata[eta]a 'he shot' ~ def. conj. d'ata[eta]a-ta id. (one object) (Sammallahti 1974 : 87), Enets indef. conj. mota 's/he cuts' ~ def. conj. mota-ra id. (one object) (Castren 1854 : 595-518; Kunnap 1999a : 25-26), Nganasan indef. conj. ko[delta]ut[??]tu 's/he kills' ~ def. conj. ko[delta]ut[??]tu-[delta]u id. (one object) (Kortvely 2005 : 41); Selkup indef. conj. tu[eta]a 's/he arrives', conde 's/he covers' ~ def. conj. conde-d id. (Castren 1854 : 519-544), Kamass indef. conj. ner[degrees]elIe 's/he fears / is afraid' ~ def. conj. parg[??]l'a-t 's/he carves' (Kunnap 1999b : 21-22). Drawn from a relatively scanty material of Mator, the 3P verbal infection is lacking (Helimski 1997 : 166).
3. A fairly complicated question arises in discussing the choice of the indefinite and definite conjugations in one or another Uralic language. Despite a long-term and many-sided study of the issue no unanimous opinion has been reached. (Often a definite object is mentioned in this connection, however, the definite object has not been determined more specifically, see Gulya 1995 : 96-98). There is no common set of rules valid for the whole Uralic language group (see also Tauli 1966 : 168-172). Various researchers have admitted that definite rules about the use of the Ob-Ugric definite conjugation seem to be lacking and in different dialects one can find only certain tendencies with numerous deviations (see first of all Honti 1987 : 34). Irina Nikolaeva believes that the use of the definite conjugation in Ob-Ugric languages is related to topicalisation (focusation) and is similar to the respective use in Samoyedic (Nikolaeva 2001 : 456-471; see also below in 4).
In connection with Northern-Samoyedic languages P. Hajdu has mentioned that in Nenets the choice principles of the indefinite and definite conjugations seem to be inconsistent (see Wickman 1997 : 87). Pekka Sammallahti writes about Forest Nenets regretting that semantic differences between the two conjugations are not clear to him (Sammallahti 1974 : 67). Recently E. Kortvely stated, and obviously with a good reason as she has carried out respective tests with native-speaking informants, that the views spread in Samoyedology to the effect that the choice of the indefinite and definite conjugations in Tundra Nenets--as if the definite conjugation can be used with the transitive verbs when, first of all, the logical emphasis or the syntactical focus is on the predicate--are erroneous by nature already (Kortvely 2005 : 114-115, 118-119; cf. also Wickman 1997 : 78-83). E. Kortvely believes that the higher the transitivity of the situation in which the verb occurs, the greater the probability that it will be conjugated in the definite conjugation (Kortvely 2005 : 152; about the transitivity see Hopper, Thompson 1989; cf. also Dahl 1984).
For example, the analysis of two Tundra Nenets sentences from E. Kortvely's work: 1) t'onakomda nertevanz pedara pomna irt sal'am vajerta-da (def. conj.) 'In order to pass the fox, in the forest he roved straight over the land at the bend of the river'. In this sentence there is a high level of transitivity: the sentence reports on a change; its subject can be considered human, the situation is dynamic; the sentence is a statement, is realistic, it does not formulate a question, the verb is of perfective aspect and telic; the object is concrete, countable, identifiable and definite (Kortvely 2005 : 150, 155). 2) sarevi hal'am', pirevi hal'am menen? (indef. conj.) 'Do you like fried fish or cooked fish better?' In the sentence there is a low level of transitivity: its subject is not an agent, the situation is not dynamic; the sentence is not a statement but a question, the verb is of imperfective aspect and in the present tense; the object is not concrete or countable (Kortvely 2005 : 155).
Until further investigations provide new and more reliable findings, E. Kortvely suggests that just the high level of transitivity in Tundra Nenets be accepted as the explanation for the use of the definite conjugation in a linguistic situation (Kortvely 2005 : 157, 160). Since Natal'ja Terescenko assumes that in all Northern-Samoyedic languages the choice principles of the indefinite and definite conjugations are essentially the same (Tepe-[TEXT NOT REPRODUCIBLE IN ASCII] 1973 : 185-199), a working hypothesis could be put forth to the effect that E. Kortvely's conclusions about Tundra Nenets can be transferred also to other Northern-Samoyedic languages--Forest Nenets, Enets and Nganasan--, i.e. that the high level of transitivity evokes in them the need for use of the definite conjugation.
It is not easier in case of clarifying the application principles of the definite conjugation for the two Southern-Samoyedic languages--Selkup and Kamass (there is no sufficiently relevant evidence for the third language --Mator). E. Kortvely notes that in Selkup the two existing conjugations are connected, first of all, in accordance with the intransitivity or transitivity of the verb or the linguistic situation, e.g., intrans. verb tuna (indef. conj.) 's/he comes' ~ trans. verb w[??]ccenty-ty (def. conj.) 's/he takes' (see [TEXT NOT REPRODUCIBLE IN ASCII] 1980 : 8, 44). However, the choice principles of the conjugations have actually not been found out for the Selkup language. In Kamass, mainly the indefinite conjugation is used. The definite conjugation occurred very rarely even in case of the transitive verb. E.g., intrans. verb uzulije (indef. conj.) 's/he prays' ~ trans. verbs [gamma]ulia (indef. conj.) 's/he looks', pierije-t (def. conj.) 's/he shows' (Kunnap 1999a : 22-23, 41; Kortvely 2005 : 34-35; cf. also Kunnap 1978 : 11, 25-26, 28). If a supposition that the high level of transitivity evokes the use of the definite conjugation in them is true concerning Northern-Samoyedic languages, then it denotes a significance in the choice of conjugations of the transitive level characteristic of verbs on the one hand, and in the relationship of verbs and their objects, on the other. In that case, in a more generalised manner, the Northern-Samoyedic choice criteria are obviously also valid in Southern-Samoyedic, at that more in Selkup than in Kamass.
Temporal change in the choice principles of conjugations can be illustrated by the fact that those were not necessarily the same rules of choice of the indefinite and definite conjugations in Old Hungarian as in Modern Hungarian. Janos Gulya has given examples of two Old Hungarian sentences with their equivalents in Modern Hungarian and we can see that the present-day definite conjugation is used instead of the old indefinite one, and the present-day indefinite instead of the old definite one. J. Gulya has also mentioned that the rules of use of the Hungarian definite conjugation are no match for the Hanti language, otherwise relatively close to Hungarian (Gulya 1995 : 97-98; see also Marcantonio 1985).
One of the examples of the variety of choice principles shows that in Mordvin the definite conjugation is used mainly in the perfective aspect even when used with a definite object ([TEXT NOT REPRODUCIBLE IN ASCII] 1980 : 283; Alhoniemi 1996 : 68; Keresztes 1999 : 14).
4. Speaking about the application peculiarieties of the Uralic 3P verbal marker *-sV it is essential to note that Finnic 3P verbal inflection *-sen (? < *-sV) primarily occurs in the easternmost Finnic languages and dialects. Traces of the North-Eastern European interupted language chain have been supposed because it is in the languages locating farther east--Mari and Permic--that we can find the application of the 3P verbal marker *-sV, a rather similar one with that of the Finnic easternmost languages (see Hausenberg 1996; 2001 : 314, 317; Kunnap 2004c : 137-140). Besides, a supposed etymological equivalent of the Finnic suffix *-sen can possibly occur in Northern-Samoyedic reflexive conjugation (Mikola 1984 : 402-403; 1988 : 255; see also below in 6).
In Mordvin the object conjugation expresses an object in all three persons, in Ugric and Samoyedic this conjugation occurs, as a rule, when an object is in 3P (see e.g. Joki 1944 : 169; Tauli 1966 : 168-169).
Use of verbal forms with the Uralic 3P marker *-sV in Mordvin and Mari in the imperative / optative meaning attracts attention. No other mood marker needs to be used in the 3P verbal forms of imperative/optative because the 3P verbal inflection itself has become to express imperative/optative. In Mordvin it is one of the optative (in fact, hortative) forms, primarily used instead of the 3P forms of the imperative, spread also in the form of the z-material suffix into the 1P and 2P forms of the optative, e.g. mora-za-n 'let me sing' : mora-za-t 'let you sing' : moraza 'let her/him sing'. In Mordvin the imperative lacks a paradigmatic marker altogether, cf. the only forms with the original imperative k-marker (2PSg), e.g. mora-k 'sing', mora-ka 'come on, sing!' ([TEXT NOT REPRODUCIBLE IN ASCII] 1975 : 317-318, 320; cf. also [TEXT NOT REPRODUCIBLE IN ASCII] 1986 : 21). In Mari an analogical phenomenon can be found in the imperative that also lacks a paradigmatic marker in that language. But in Mari the 3P verbal inflection -sV, -zV has not spread into imperative verbal forms of other persons, e.g., Meadow Mari puro 'come in' : puri-zo 'let her/him come in' ([TEXT NOT REPRODUCIBLE IN ASCII] 1976 : 70-71).
In Finnic, Saamic, Northern-Samoyedic and Selkup (partly also in Ugric?) where the verbal inflection is also made use of in the imperative/ optative, the respective imperative verbal forms are, in turn, provided with specific imperative/optative markers preceding a 3P verbal inflection, e.g., Finnish tule(-') (< *tule-k) 'come' : tulkoon (< *tul-ko-hen) 'let her/him come' (k-marked imperative), Saamic le-kku-m 'let me be' : le-kku-s 'let her/him be' (k-marked imperative) (Korhonen 1981 : 260), Nenets madarawa-u 'I could cut' : mada-rawa-da 's/he could cut' (rawa-marked optative; Castren 1854 : 370, 394), Selkup qo-la-m 'I would be able to could find/see' : qo-la-ti 's/he would be able to could find/see' (la-marked optative) ([TEXT NOT REPRODUCIBLE IN ASCII] 1980 : 245-246; Helimski 1998 : 568).
Mordvin, Komi and Udmurt share a common peculiarity in that that instead of expected present meaning the 3P verbal marker *-sV expresses only or in most cases the future meaning, whereby any other future marker is lacking in Mordvin and Komi because the 3P verbal inflection itself has become to express the future tense, e.g., Mordvin kundi 's/he catches' : kunda-si 's/he will catch' ([TEXT NOT REPRODUCIBLE IN ASCII] 1974 : 318-319), Komi mune 's/he goes' : muna-s 's/he will go' ([TEXT NOT REPRODUCIBLE IN ASCII] 1976 : 175). In Udmurt, however, it is preceded by a paradigmatic future marker -o-, e.g., mine 's/he goes': min-o-z 's/he will go', cf. e.g. min-o-d 'you will go'. Supposedly, earlier the -o- designated also the present tense (the present-future, in fact). Considering the latter, at the diachronic level the Udmurt language with its z-suffixal future forms is consistent with Mordvin and Komi ([TEXT NOT REPRODUCIBLE IN ASCII] 1976 : 173-174) (in Permic languages the 3P verbal inflection occurs in other verbal forms also: in Komi only in the preterite tense but in Udmurt almost overwhelmingly).
I suppose that it would be only natural to see a definite connection between the above listed cases of peculiar uses of the 3P verbal marker *-sV as imperative-optative on the one hand, and the future tense, on the other. Namely, it is known typologically that viewed from the temporal aspect the imperative/optative is related to the future (see e.g. [TEXT NOT REPRODUCIBLE IN ASCII] 1986 : 21). In case of Uralic languages Nadeczda Kuznecova has observed the connection in Selkup. She gives examples about re-rendering Selkup optative forms as future forms, e.g., hiri[??]. nadem me-[??]ai '(we two) will make a girl from snow' (a [??]a-marked optative 1PDu form in the future meaning), likewise, adhering optative forms with the imperative paradigm, e.g., nop peldi-la 'God help' (a [??]a-marked optative 3PSg form in the imperative meaning) ([TEXT NOT REPRODUCIBLE IN ASCII] 1991 : 260-261; see also [TEXT NOT REPRODUCIBLE IN ASCII] 1980 : 245-246; cf. [TEXT NOT REPRODUCIBLE IN ASCII] 1973 : 153). Consequently, the peculiar and supposedly mutually related cases of application of the 3P verbal marker under observation in the imperative/optative and future cover the whole Volgaic-Permic linguistic area.
Boris Serebrennikov has indicated that the re-rendering of the optative as the future tense takes place also in some Turkic languages ([TEXT NOT REPRODUCIBLE IN ASCII] 1974 : 242). The Volgaic-Permic linguistic area demonstrates a strong mutual influence between Finno-Ugric and Turkic languages (see first of all [TEXT NOT REPRODUCIBLE IN ASCII] 2005; the same could be said also about the Southern-Samoyedic (Selkup, Kamass, Mator) language area). Therefore I would not exclude the Turkic influence in case of such a peculiar application of the future tense or inclusion of optative forms into the imperative paradigm in Volgaic-Permic languages. However, I would draw attention to the fact that a paradigmatic application of the primary imperative marker is lacking in Mordvin while in Komi and Udmurt the primary imperative marker is lacking altogether (see e.g. [TEXT NOT REPRODUCIBLE IN ASCII] 1974 : 312). I believe the fact might help to explain the peculiar application of the future tense under observation in those languages: as if, at the expense of missing imperative forms, there were space for this particular application of the future, semantically related to imperative or even need for such a "replacement". It should not be forgotten that the imperative is linguistically one of the oldest and most central moods (see e.g. [TEXT NOT REPRODUCIBLE IN ASCII] 1986 : 18-19) whose unmarked occurrence in Mordvin, Komi and Udmurt could have brought about marked replacements. I fail to fully open the "replacement mechanism" but apparently it exists. Deviations from the use of Selkup verbal forms proposed by N. Kuznecova may be related to the fact that consistent paradigmatic future and imperative markers are partially lacking in Selkup (Kunnap 1978 : 96-100, 133-134; [TEXT NOT REPRODUCIBLE IN ASCII] 1980 : 239-240, 247-248).
As far as the application of the 3P s-material verbal inflection is concerned only in the imperative in Saamic it could be quite a different phenomenon in comparison with the above in the Volgaic-Permic linguistic area. If namely the supposition that the imperative marker of the Saamic 3P forms originates from the imperative marker *-ko(i)- of the Finnic 3P forms, e.g., Saamic le-kku-s 'let her/him be' (Korhonen 1981 : 260, 264; see also Redei 1982 : 8-9), is valid then it is fully possible to suppose that the marker was actually included into Saamic together with the Finnic 3PSg verbal inflection *-sen that follows it, cf. Finnish liekoon 'let her/him perhaps be' (< *le-ko-hen < *-zen < *-sen). At that the proof for the change of the Finnic 3PSg *-sen > Saamic 3PSg -s would be provided by a phonetically identic equivalence of the change--the phonetic evolvement of the Saamic imperative 3PDu verbal inflection, e.g., tappo-s 'let them (two) shut' (< *-sen, where -n is a dual marker) (Korhonen 1981 : 279). According to M. Korhonen, the Saamic verbal inflection has exact equivalents in Mordvin and Mari imperative/optative, e.g., Mordvin pala-zo 'let her/him kiss' and Mari [TEXT NOT REPRODUCIBLE IN ASCII] 'let her/him lead', similarly, a close equivalent in the Finnic imperative, e.g., Finnish lukekoon 'let her/him read' (< *lukeko-hen < *-zen < *-san) (Korhonen 1981 : 273). If the 3P verbal inflection -s of the Saamic imperative has really been obtained from Finnic, it probably began to be used in Saamic accidentally and it has no relationship with application of the inflection in imperative/optative in Mordvin and Mari.
Reference to the number of an object in the verbal forms is a linking feature for Mordvin, Ob-Ugric and Northern-Samoyedic languages. Besides, in Mordvin verbal forms a reference is made to the person of an object. References to the person of an object in a verbal form can also be found in Hungarian in the form of the inclusive -lak, -lek conjugation, although here it is only confined to the object of the 2PSg/Pl, e.g., nez-lek 'I look (at) you' (see e.g. Kortvely 2005 : 39-40; see also in 2).
Observing the incidence of the 3P verbal marker *-sV in Uralic with an emphasis on the intransitivity / transitivity of verbs, we can get the following picture. One could note that Finnic languages prefer using the verbal inflection under discussion in case of intransitive and reflexive verbs, e.g., Southern Estonian pala-s 's/he burns', lovvu-s 's/he is found', Veps kumarda-se 's/he bows', Karelian pessak-cse 's/he washes her/himself' (see above in 2, but cf. Hajdu 1985 : 246). In Mordvin transitive verbs must be used instead of the finite conjugation in the indefinite one if an action is not perfective ([TEXT NOT REPRODUCIBLE IN ASCII] 1975 : 314; Alhoniemi 1996 : 68; Molnar 2001 : 68-91).
P. Hajdu refers to B. Serebrennikov's viewpoint, according to which an essential amount of Mari verbs provided in the 3P with the suffix -s are mainly transitive and the verbs lacking the suffix are largely intransitive. True, P. Hajdu regards this point of view as questionable ([TEXT NOT REPRODUCIBLE IN ASCII] 1985 : 246). As a rule, the Komi preterite uses a variant without -s with intransitive verbs and with -s--with transitive verbs, e.g., muni 'he went' ~ bosti-s 'he brought' ([TEXT NOT REPRODUCIBLE IN ASCII] 1985 : 245-246). In Udmurt the suffix -z adheres to both indefinite and definite verbs ([TEXT NOT REPRODUCIBLE IN ASCII] 1962 : 194-219).
As was mentioned above, E. Kortvely suggests that just the high level of transitivity in Tundra Nenets be accepted as the explanation for the use of the definite conjugation in a linguistic situation. I gave two Tundra Nenets sentences by way of examples, along with reasons for applying definite or indefinite conjugations, all borrowed from her work (see in 3). The reasons show that among the indicators to denote a high or low level of transitivity of the sentences there are also perfective or imperfective aspects of the verb, respectively. Against the background of the fact I suppose it is easier to understand why in Mordvin the definite conjugation occurs mainly in perfective aspect (there are numerous deviations from the latter rule, at that to a different extent and manner in Ersa Mordvin and Moksha Mordvin--see Alhoniemi 1996).
J. Gulya has paid attention to the fact that if the Nenets direct object is expressed by a personal or demonstrative pronoun, the indefinite conjugation is used instead of the definite one (Gulya 1995 : 99; see also Wickman 1997 : 218; Pusztay 2001 : 72). But this rule is valid as far as the personal pronouns for all Northern-Samoyedic languages--Nenets, Enets and Nganasan are concerned, e.g., Nenets sita (3P) kolhozanda" [eta]esin' hana" (indef. conj.) 'drive him to the kolkhoz village', Enets mod' si" (1P) mogad kada" (indef. conj.) 'drive me to the forest' ([TEXT NOT REPRODUCIBLE IN ASCII] 1973 : 191-192), Nganasan mana"ku siti (3P) kacecemi"[??]m (indef. conj.) 'I just saw him' ([TEXT NOT REPRODUCIBLE IN ASCII] 1979 : 186). And in case of the 1P and 2P personal pronouns also in Hungarian, e.g., ker (indef. conj.) engem (1P) 's/he asks me' ([TEXT NOT REPRODUCIBLE IN ASCII] 1976 : 395), in Selkup, e.g., t[??]p sinty (2P) qontyrtenta (indef. conj.) 'he will see you' ([TEXT NOT REPRODUCIBLE IN ASCII] 1980 : 234-235), and only as a tendency in Ob-Ugric languages, e.g., Hanti luw jir[??]s (indef. conj.) manem (1P) anta na[eta]en (2P) 'he tied me, not you' (Nikolaeva 2001 : 464, 470-471). I. Nikolaeva notes the situation in Ugric, "Ein direktes Objekt der 1. oder 2. Person Personalpronomen ist [---] selten ein Topic; wenn es vorhanden ist, tragt es wahrscheinlich einen Fokusstatus. Dieses widerspricht dem Verhalten des 3. Person Personalpronomens, das ein sehr haufiges (in der Tat das typischste) sekundare Topic ist. Wenn die subjektive/ objektive Konjugation dahingegehend analysiert wird, dass sie Fokus bzw. Sekundaren Topicstatus des Objekts markiert, [---] ist es klar, dass sie die 1. und 2. Person Objektpronomina dazu neigen, mit dem subjektiven Verb zusammen aufzutreten, wahrend das 3. Person Objektpronomen dazu tendiert, mit dem objektiven Verb zusammen vorkommen" (Nikolaeva 2001 : 470-471). The Northern-Samoyedic use of the indefinite conjugation in the case of the 3P personal pronoun is an exception from this tendency (see also Helimskij 1982 : 84-85).
5. As far as the origin of the Uralic definite conjugation and the 3P verbal inflection *-sV are concerned, by way of an ordinary explanation Klara Majtinskaja has offered the following. In the Common-Uralic period there were preconditions for the emergence of two conjugations whereby the main precondition consisted in the fact that there were two variants in creating the 3PSg: intransitive verbs without a verbal inflection and transitive verbs with a verbal inflection ([TEXT NOT REPRODUCIBLE IN ASCII] 1974 : 319; see also Kunnap 1978 : 80-81). 3P verbal inflection descends from a personal pronoun with the 3P possessive suffix as an intermediate stage ([TEXT NOT REPRODUCIBLE IN ASCII] 1974 : 321-322). As a result of versatile observation and from novel aspects at that, J. Gulya comes to a conclusion that there were intransitive and transitive conjugations in Proto-Uralic. The definite conjugation is a secondary (area-related?) phenomenon (Gulya 1995 : 99; but cf. Redei 2002 : 279). P. Hajdu also shares the view that Proto-Uralic had already had two possibilities of the formation of the 3P for the verbs: with and without a verbal inflection. His opinion is that then already the 3P verbal inflection indicated a definite object and apparently originated from a personal or demonstrative pronoun with an accusative meaning ([TEXT NOT REPRODUCIBLE IN ASCII] 1985 : 247).
In the Finnic material presented above (see 2), L. Posti first and foremost sees traces of the onetime Finnic reflexive conjugation that was provided with 3P verbal inflection *-sen (in sg. and *-set in pl.) (Posti 1980 : 111-112; see also Itkonen 1966 : 273-274). Tibor Mikola supposes that the inflection comes from the derivative suffix of deverbative adjectives (Mikola 1984 : 400). P. Hajdu believes that the inflection comes from the Proto-Finnic 3P personal pronoun *sen, and only thereafter obtained the meaning of a reflexive pronoun ([TEXT NOT REPRODUCIBLE IN ASCII] 1985 : 244-247). But we should not hurry with the supposition that in Finnic the personal pronoun *sen has turned into a personal suffix. The Finnic 3PSg personal suffix *-sen, due to its final component *-n, is not a common personal suffix in Uralic languages, yet its initial component *-se- does not exclude a link with the common Uralic suffix *-sV. More so, the varying application manner of the 3PSg personal suffix *-sen in Finnic verbs--partly without the suffix, partly with the suffix--is similar to the analogous application of the 3P verbal inflection *-sV in Mordvin, Mari and Komi languages.
Besides, the Northern-Samoyedic reflexive conjugation has presumably also had an n-ending 3PSg verbal inflection *-ten (see for it lower in greater detail in 6). On the other hand, however, if the Finnic 3P verbal inflection *-sen can etymologically be linked with the supposed 3P verbal inflection *-ten of the Northern-Samoyedic reflexive conjugation, then the fact that the preference use of the suffix *-sen with reflexive verbs in Finnic is noted can support the supposition of P. Hajdu and T. Mikola to the effect that both the Finnic and Northern-Samoyedic suffixes come from the 3P pronoun *sVn ([TEXT NOT REPRODUCIBLE IN ASCII] 1985 : 244-247; Mikola 1988 : 255). Besides, a tendency appears in Finnic to use the suffix -sen primarily with intransitive and r e f l e x i v e verbs, in Mordvin, Mari and Komi--the suffix *-sV with transitive verbs, in the first place.
In the case of a number of Finnic languages--Finnish, Karelian, Ingrian, Veps--we have to mention a polyfunctional particle -han, -han etc. (probably also Saamic and Mordvin -son, -han etc.) which adheres to verbs also, e.g., Finnish syo-han nyt jotakin 'now eat something, will you ("eat-will-you")' (Saamic jogo-son 'is (it) already rightaway ("already-rightaway")') (SSA 1 1992 : 192; Larsson 1998). In Finnic its source is seen in the 3P personal pronoun *sen, cf. Finnish han 's/he' (here the relationship with the Finnic 3P verbal inflection at any rate). Probably in this case Julius von Farkas sees non-personal nature of the use of the Finnish particle (and consequently also the 3P personal pronoun *sen), quite credible in every way (Farkas 1954a : 57).
Whatever has been supposed about the development of the 3P verbal inflection in the Mordvin definite conjugation, the traces of the 3P verbal suffix *-sV are clearly there, particularly emphasised by P. Hajdu ([TEXT NOT REPRODUCIBLE IN ASCII] 1985 : 327-328; see also Korhonen 1980 : 273, 284; cf. also Keresztes 1999 : 53-66, 105-106). J. Gulya (1995 : 96) thinks that the whole Mordvin system as such has nothing to do with the actual definite conjugation occurring in Uralic languages and, at the utmost, it could only be compared with the Hungarian inclusive -lak, -lek conjugation, an argument that I can hardly agree with. In Hungarian,--if the subject is 1PSg--the inclusive -lak, -lek conjugation expresses the 2PSg/Pl object, e.g. indef. conj. nE eze-k 'I look' ~ def. conj. neeze-m id. and nez-lek 'I look (at) you' (Kortvely 2005 : 39-40). Similarity between Mordvin and Hungarian does show here, yet in case of Hungarian it confines itself to a single 1PSg verbal form whose object can only be 2PSg/Pl (see also Redei 1966 : 127).
P. Hajdu is ready to consider the possibility that in Mari the -s (see about this suffiks above in 2) is a trace of an original 3P verbal inflection and not that of the preterite/perfect marker ([TEXT NOT REPRODUCIBLE IN ASCII] 1985 : 246, 320, 327). Besides, as far as the Mari language is concerned, what makes me also tend to support the view that -s is an original personal suffix and not that of preterite/perfect, is the fact that the preterite / perfect suffix *-s'V cannot have spread in Uralic languages too widely, being inherent perhaps only to the easternmost Uralic languages--Ob-Ugric and Samoyedic (see first of all Kunnap 2000 : 33-37). J. Gulya writes that the supposedly similar system to the Hungarian inclusive -lak, -lek conjugation or Mordvin definite conjugation in Komi must be secondary (Gulya 1995 : 96). Probably he must mean, in a way, an earlier assertion that the definite conjugation is or has been used in Komi (see e.g. Kortvely 2005 : 30).
According to P. Hajdu, the development at the end of the Hungarian 3P verbal inflection is as follows *s > *v > *h > *i ([TEXT NOT REPRODUCIBLE IN ASCII] 1985 : 327). Principally, a similar phonetic development has also been supposed in case of the Hungarian 3P possessive suffix (see first of all Farkas 1954b : 65-79; [TEXT NOT REPRODUCIBLE IN ASCII] 1985 : 312). Karoly Redei writes that in Mansi the 3P possessive suffix which descended from the Finno-Ugric sV-initial personal pronoun became the 3P verbal inflection (Redei 1966 : 128-129). In Hanti the l-material verbal inflection descended from 3P personal pronoun and verbal inflection -[??]t, -et from the 3P possessive suffix (129-130). Besides, the suffix -ja-, -i- indicates an object (not a subject!) in the 2PPl verbal forms of the Hungarian definite conjugation, followed by the 2PPl verbal inflection -tok, -tek (indicating a subject), e.g., indef. conj. la-tok 'you (Pl) see' ~ def. conj. lat-ja-tok id (see first of all Pusztay 1995 : 89), cf. def. conj. 3PSg verbal inflection -ja, -i, e.g., lat-ja 's/he sees'. In Mansi the 3P verbal infection *-tV has taken the form of the -lV and penetrated into the 1P and 2P verbal forms, thus becoming a marker of the definite conjugation there, e.g., rati-le-m 'I beat (it)' : rati-le-n 'you beat (it)'. The reason was the use of (identical to possessive suffixes) similar 1P and 2P verbal inflections both in the indefinite and definite conjugations in Mansi that did not allow their mutual discrimination without an additional marker for the definite conjugation (Redei 1966 : 128-129). All the above-said indicates in the direction of a possibly common origin of the 3P verbal inflection and 3P possessive suffix and testifies to a close intertwining of their use.
It should also be mentioned that the origin of the Samoyedic imperative t-material (t ~ d ~ r) 2PSg verbal inflection as adhered to the verbal stem, e.g., Tundra Nenets tola-d 'read it', Nganasan matu-da 'cut it', Kamass parge-t id. (Mikola 2004 : 127), is not clear. This suffix cannot etymologically be connected to the supposed 3P verbal inflection *-tV due to the differences both in their consonant and vowel material (at least in Northern-Samoyedic, see in greater detail Mikola 2004 : 117-118, 127). I see a verbal noun suffix as the source for the 2P verbal inflection of the imperative because as is known, a number of Samoyedic verbal noun suffixes have turned into future, conjunctive, optative, adhortative and imperative markers (Kunnap 1978 : 100-111, 193-196; cf. also Mikola 1988 : 246-249; 2004 : 111-116).
A consonant alternation of the 3P verbal marker Finno-Permic and Hungarian s ~ Ob-Ugric and Samoyedic t is usually explained in Uralistics by the change of s > t (see e.g. [TEXT NOT REPRODUCIBLE IN ASCII] 1974 : 325; [TEXT NOT REPRODUCIBLE IN ASCII] 1985 : 327-328). But typologically the change s > t is a rare case (see also [TEXT NOT REPRODUCIBLE IN ASCII] 1982 : 123-125; Kunnap 2004c : 137-140). Eugen Helimski has offered a versatile analysis to the alternation s ~ t under observation and eventually considers the development s > v > t as possible, connecting it with a similarly supposed development in one of the Western-Siberian non-Uralic language--in Yenissey Pumpokol ([TEXT NOT REPRODUCIBLE IN ASCII] 1982 : 119-125). P. Hajdu considers the proceeding from the t-initial demonstrative pronoun as a possible case of t-type 3P verbal inflections of Ob-Ugric and Samoyedic languages (Hajdu 1985 : 328). Likewise, a demonstrative pronoun has been regarded as the source for the Komi 3P verbal inflection -s when comparing it with the pronoun sije 's/he, younder' ([TEXT NOT REPRODUCIBLE IN ASCII] 1955 : 217; see also [TEXT NOT REPRODUCIBLE IN ASCII] 1974 : 319).
It could consequently be conceivable to take into account the descent from two independent old demonstrative pronouns *s'V and *tV. It should be viewed in connection with its determinative function in case of a possessive suffix and the use as a personal inflection of the finite conjugation and its *s-initiality in Finno-Permic languages (but not in Hungarian as there *s > s) and *t-initiality in Ob-Ugric and Samoyedic languages. Such a supposition leads to the determinative function of the 3P possessive suffix as a more indigenous one (see for it also Kunnap 2004c : 137-140). In the forms of *s'V and *tV those are demonstrative pronouns that Juha Janhunen has reconstructed for Proto-Uralic. For intermediate protolanguages he has reconstructed Proto-Finno-Permic ? *sa- (e.g. Finnish se) ~ Proto-Samoyedic *sV- and Proto-Finno-Permic *tV(-) (e.g. Finnish tama, tuo) ~ Proto-Samoyedic *tV(-). (Janhunen 1981 : 269). At the same time it should be mentioned that Laszlo Honti has recently considered it as possible to suppose, in case of congruity of the Finno-Permic *s ~ the Ob-Ugric and Samoyedic *t, the descent of both of those sounds from an earlier common source--an original voiceless dental fricative [??] (e.g. *[??]ulka > Finnish sulka ~ Hungarian toll 'feather') (Honti 2004). If L. Honti's idea is true then we have an option at hand that the *s- and *t-initial variants of the 3P personal suffix in Uralic languages are mutually etymologically related, and both may be descending from the original *-[??]V.
6. As we have been able to see from above, in case of the observation of the definite conjugation we often rely on the data about the reflexive conjugation. An earlier or current existence of the conjugation has been brought forth in case of Finnic (see above 2). It is interesting what T. Mikola has said in connection with the Finnic 3P verbal inflection sg. *-sen, pl. *-set. T. Mikola regarded the comparison of the Finnic and Northern-Samoyedic reflexive conjugations as relevant. He admits that in Northern-Samoyedic languages a reflexive conjugation has taken shape that is formed by means of the reflexive suffix -j-. It started in the verbal forms of 3PSg and 3PPl that are provided with verbal inflections *-tVn and *-tVt, respectively. According to T. Mikola, the latter descend from the Proto-Uralic pronouns *sVn 's/he, that' and *sVt 'they'. The verbal inflection *-tVn has yielded in Nenets -[??], e.g., Tundra Nenets toerej-[??] 'he pastures', in Forest Nenets taewa [??] 'he came', in Enets -ro[??], -do[??], in Nganasan -[??], -[delta][??]. (Mikola 1984 : 402-403; 1986; 1988 : 255; 1990; 2004 : 124-127; see also Katzschmann 1999-2000 : 117-121; in Forest Nenets also -t--Lehtisalo 1938 : 25). In the light of T. Mikola's considerations the contact between Finnic and Northern Samoyedic is conceivable in every aspect, it can truly be expected that the reflexive conjugation occurs only in Northern-Samoyedic languages which are located geographically closer to the area of Finnic than Southern-Samoyedic languages. Primarily, of course, in this case when the Finnic suffixes *-sen, *-set can etymologically be linked with the Northern-Samoyedic suffixes *-ten, *-tet (s ~ t!). But even without any etymological link there is a possibility that the use of the suffixes *-ten, *-tet in Northern Samoyedic developed on an example of the use of the Finnic suffixes *-sen, *-set (see also Kunnap 2000 : 49-50).
Yet in case of Tundra Nenets (and substantially also Forest Nenets) E. Kortvely mentions that the supposition of such an origin based barely on a glottal stop -[??] as the form of a personal ending is not unproblematic since such an extensive deletion of the material is unusual even in a word-final position. E. Kortvely considers the onetime existence of two functionally completely identical 3PSg verbal inflections--*-tVn and *-k--as possible in the Northern-Samoyedic reflexive conjugation as well as the preservation of different traces of these suffixes in Nganasan (Kortvely 2005 : 90-91). E. Helimski has reconstructed the 3PSg verbal inflection of the Samoyedic reflexive conjugation in the form of *-k, proceeding at that from a respective dialectal suffix -k, -[eta] of the Selkup indefinite conjugation ([TEXT NOT REPRODUCIBLE IN ASCII] 1982 : 81). In my turn I have reconstructed the Nganasan 3PSg verbal inflection -[??] (see e.g. [TEXT NOT REPRODUCIBLE IN ASCII] 1979 : 185) in the form of *-tV (Kunnap 1978 : 46, 60) and I cannot support the proceeding presumably from a secondary Selkup suffix -k, -[??] (see for Toivo Lehtisalo's opinion about an exceptional origin of the latter in Kunnap 1978 : 34-35; cf. also Forest Nenets -t).
An equivalent to that reconstructed Samoyedic personal inflection *-k has been attempted to find in Hungarian ([TEXT NOT REPRODUCIBLE IN ASCII] 1982 : 82-83). Namely the reflexive conjugation is represented in Hungarian by ik-conjugation (with an element -ik in the 3PSg), e.g., 1PSg kese-m 'I am late' : kes-ik 's/he is late' (see e.g. Mikola 1984 : 400-401; Kortvely 2005 : 40). In the given case there is no reason to speak about any etymological correspondence between Samoyedic and Hungarian languages if the supposition about the common Samoyedic personal inflection *-k is erroneous (see Kunnap 2007).
7. In conclusion, as far as the Uralic language group is concerned, no original traces of the occurrence of the 3P verbal marker *-sV need necessarily not be found in Finnic and Saamic (see 3 and 4), conceivably not even in Hungarian (the Hungarian supposed development *s > *v > *h > *i as well as the lengthening of the preceding vowel are not so very convincing--see 2 and 5). Among other Uralic languages without the definite conjugation besides Finnic and Saamic the marker is used in Mari and Komi, with transitive verbs in the first place, but in Udmurt such a preference is not observed (see 4). My supposition is that in all Northern-Samoyedic languages the high level of transitivity evokes the use of the definite conjugation, naturally with the 3P verbal marker *-tV. I also suppose that the same principle is valid in Southern-Samoyedic languages too, more clearly in Selkup (see 3). All these facts make the supposition according to which the primary determinator of the choice between the indefinite or definite conjugations in Uralic languages was intransitivity / transitivity only too probable: intransitive verbs were used without and transitive verbs with the 3P verbal marker.
In this connection a question is well-founded: what was the marker *-sV originally like? In Finnic and Lapp the 3P possessive suffix is used instead of the 1P and 2P possessive suffixes, but not the other way round. In Volgaic, Permic, Ob-Ugric and Samoyedic both the 3P and 2P possessive suffixes are used in the non-personal definitive function. Thereby, in Volgaic, Permic and Ob-Ugric the 3P possessive suffix is used in the non-personal definitive function more often than the 2P possessive suffix. The 3P possessive suffix occurs in the non-personal definitive function also in Turkic and Tungusic languages (Tauli 1966 : 148; Pusztay 1975 : 364). Based on the use of (Northern-)Samoyedic languages one could imagine that in a number of Uralic languages the text meant for the addressee (= the 2nd person) sounds natural when determined by means of the abovementioned 2P possessive suffix, meaning something like 'the matter heard by you' (in case of the Permic 2P possessive suffix Janos Pusztay emphasises the function of expressing a mutual contact, or intimacy of the contact, between the speaker and the listener, see Pusztay 1975 : 363). The 3P possessive suffix usage is here related to a more general definiteness (in Samoyedic languages in case of the concepts 'sun', 'moon', 'sky', 'earth' etc., since these phenomena are determined by their unique and universal nature already, not by the listener's hearing; J. Pusztay also adds Mordvin to this group, see 1975 : 363; see also Kunnap 2004a).
Thus, it may be supposed that a non-personal general definitive function has always been inherent to the Uralic 3P possessive suffix. One could also assume that the 3P possessive suffix in the mentioned general-definitive function has been used in those Finnic languages in which the traces of the 1P and 2P possessive suffixes are practically lacking--in Estonian and Livonian (see for it more closely in Kunnap 2001 : 123-128). In those two languages the 1P and 2P possessive suffixes may never have been in general use, and the use of the 3P possessive suffix may have originated from the possibly primary non-personal general-definitive function of the 3P possessive suffix. The source for the 3P verbal marker was a demonstrative suffix, identical to the 3P possessive suffix (or possibly two different demonstrative suffixes--in Finno-Permic *-sV and in Ob-Ugric and Samoyedic *-tV). Namely, it is principally possible that primarily there was no 3P possessive suffix / 3P verbal personal suffix in the Uralic language group (that is easier to conclude in case of the possessive suffix, based on the data of modern languages than in case of a verbal personal suffix). As far as the Uralic 3P possessive suffix is concerned, Julius von Farkas has pointedly noted that the 3P ownership relationship is seldom expressed by formal elements and that is why the 3P possessive suffix could have preserved its original determinative function for a very long time, up to date (Farkas 1954a : 58). I also agree with Hajdu's points of view that from the start already the Uralic 3P verbal inflection indicated a definite object ([TEXT NOT REPRODUCIBLE IN ASCII] 1974 : 321; [TEXT NOT REPRODUCIBLE IN ASCII] 1985 : 247; as the 3P possessive suffix indicated a primaril see also 5).
Discussing the choice of the indefinite and definite conjugations we need to point out that there is no common set of rules valid for the whole Uralic language group. However, the rules have never been consistently cleared up in the Uralic language subgroups and individual languages. It goes without saying that we need further in-depth research. Regretfully, even after such a research there is always a possibility of not getting a picture about how the choice was made sometime earlier. For example, in Modern Hungarian the definite conjugation may be sometimes used instead of the indefinite one in Old Hungarian and vice versa (see 3).
A peculiar use of the verbal forms provided with the 3P verbal marker *-sV can be noted in the forms of imperative / optative ~ future in the whole Volgaic-Permic linguistic area. Besides a possibility of an influence of Turkic languages we probably need to take into account the application of a defective occurrence or lacking of a paradigmatic imperative marker in the majority of Volgaic-Permic languages that could have been a consequence of "replacement" forms of imperative/ optative or the future tense, semantically close to imperative/optative (see 4).
Discrimination of indefinite and definite conjugations is a phenomenon that is inherent to the whole of Northern Siberia and, besides Uralic languages occurs in Yukaghir and in many Paleosiberian languages--Yenissey, Eskimo-Aleutic and Chukotka-Kamchatkan language groups (see Pusztay 1995 : 89-91; cf. also Klesment, Kunnap, Soosaar, Taagepera 2003 : 378). A concomitant reference to to the number of the objects in the verbal forms (in Mordvin, Ob-Ugric and Northern Samoyedic) as well as to a person of the object (in Mordvin and partly in Hungarian) occurs first of all in Northern-Siberian languages, in a number of Paleosiberian languages, among them (see Pusztay 1995 : 91-93). It is hard to imagine that reference to the number and person of an object in Uralic languages would have developed in Uralic language subgroups or single languages completely independently, without mutual or any other contacts. For this the phenomenon under observation is far too peculiar.
Undoubtedly, it is an extensive areal phenomenon, as a general rule, represented in Uralic languages the stronger the more eastern, closer ones to Paleosiberian languages we have in mind (reference to the person of an object in verbal forms occurs also in Basque, some Caucasian languages and Amerindian languages--see Tauli 1966 : 169). The discrimination of those two conjugations and reference to the number as well as person of objects in the verbal forms have not been made use of independently in Uralic languages, in isolation from non-Uralic Siberian languages. (In a number of cases, on the part of some specialists in Uralistics, one can observe a tendency to explain the origin of whatever phenomena in those languages through genuine reasons, avoiding the consideration an influence of other languages at any cost. Such "patriotism" will do no good to science).
Address: Ago Kunnap University of Tartu E-mail: firstname.lastname@example.org Phone: +372 7375423
Alhoniemi, A. 1985, Marin kielioppi, Helsinki (Apuneuvoja suomalaisugrilaisten kielten opintoja varten X).
--1996, Uber die Unterschiede bei der Verwendung der Subjekt- und Objektkonjugation im Ersa- und Mokscha-Mordwinischen.--CIFU VIII. Pars III, 68-71.
Csucs, S. 2001, A targyas ragozas es a zurjen nyelv.--Mikola-emlekkonyv, Szeged (Neprajz es Nyelvtudomany XLI/1), 69-76.
Castren, M. A. 1854, Grammatik der samojedischen Sprachen, St. Petersburg.
Dahl, O. 1984, Notes on Aspect, Case, and Transitivity.--Symposium in memoriam Janos Lotz (1913-1973). Foredrag hallna vid minnessymposiet den 27 september 1983 vid Stockholms universitet, Stockholm (Studia Hungarica Stockholmiensia), 16-21.
Dahl, O., Koptjevskaja-Tamm, M. 1992, Language Typology around the Baltic Sea. A Problem Inventory, [Stockholm] (Papers from the Institute of Linguistics. University of Stockholm 51).
Gulya, J. 1995, Transitivitat, Determiniertheit und objektive Konjugation.--CIFU VIII. Pars III, 96-99.
Farkas, J. von 1954a, Bemerkungen zu der ungar. historischen Formenlehre III. Bildung des Possessivstammes.--UAJb. 26, 37-60.
--1954b, Bemerkungen zu der ungarischen historischen Formenlehre IV. Bildung des ungar. Possessivstammes.--UAJb. 26, 60-93.
Hausenberg, A.-R. 1996, Onko komin ja itamerensuomalaisissa kielissa areaalisia yhteispiirteita?--CIFU VIII. Pars IV, 180-182.
--2001, Kadunud luli Kirde-Euroopa keeleahelas.--CIFU IX. Pars IV, 314-319.
Helimski, E. 1997, Die matorische Sprache. Worterverzeichnis--Grundzuge der Grammatik. Sprachgeschichte unter Mitarbeit von Beata Nagy, Szeged (Studia Uralo-Altaica 41).
--1998, Selkup.--The Uralic Languages, London--New York, 548-579.
Honti, L. 2004, "The Missing Link", avagy rokon-e a magyar toll es a finn sulka?--Mikola-konferencia 2002, Szeged, 75-84.
Hopper, P. J., Thompson, S. A. 1980, Transitivity in Grammar and Discourse.--Language 56, 251-299.
Itkonen, E. 1966, Kieli ja sen tutkimus, Helsinki.
Joki, A. J. 1944, Kai Donners Kamassisches Worterbuch nebst Sprachproben und Hauptzugen der Grammatik, Helsinki (LSFU VIII).
Katzschmann, M. 1999-2000, Die samojedischen Konjugationsarten und die Frage eines uralischen Ergativs.--Neprajz es Nyelvtudomany XL, 113-130.
Keresztes, L. 1999, Development of Mordvin Definite Conjugation, Helsinki (MSFOu 233).
Klesment, P., Kunnap, A., Soosaar, S.-E., Taagepera, R. 2003, Common Phonetic and Grammatical Features of Uralic Languages and Other Languages in Northern Eurasia.--Journal of Indo-European Studies 31 3 & 4, 363-390.
Korhonen, M. 1981, Johdatus lapin kielen historiaan, Helsinki (SKST 370).
Kortvely, E. 2005, Verb Conjugation in Tundra Nenets, Szeged (Studia Uralo-Altaica 46).
Kunnap, A. 1978, System und Ursprung der kamassischen Flexionssuffixe II. Verbalflexion und Verbalnomina, Helsinki (MSFOu 164).
--1999a, Enets, [Munchen--Newcastle] (LINCOM EUROPA. Languages of the World/Materials 186).
--1999b, Kamass, [Munchen--Newcastle] (LINCOM EUROPA. Languages of the World/Materials 185).
--2000, Contact-induced Perspectives in Uralic Lingustics, [Munchen--Newcastle] (LINCOM Studies in Asian Linguistics 39).
--2001, On Some Pecularities of the Estonian Language.--Estonian Typological Studies V, Tartu (Tartu Ulikooli eesti keele oppetooli toimetised 18), 123-137.
--2004a, About the Non-Personal Definite Function of the Uralic 3rd Person Possessive Suffix.--LU XL, 1-4.
--2004b, On the Evidence of the Verbal 3rd Person Suffix *-sV.--LU XL, 81-86.
--2004c, Some More Possible Traces of the Lost Uralic Language Chain of North-East Europe.--Myths and Facts in Uralistics, Tartu (FU 26), 134-150.
--2007, On the Uralic Verbal Personal (*)k-Markers.--LU XLIII, 81-89. Larsson, L.-G. 1998, Suomen -han ja saamen -son.--Oekeeta asijoo. Commentationes Fenno-Ugricae in honorem Seppo Suhonen sexagenarii 16. V. 1998, Helsinki (MSFOu 228), 276-279.
Lehtisalo, T. 1936, Uber die primaren ururalische ableitungssuffixe, Helsinki (MSFOu 72).
--1938, T. Lehtisalo's university lectures in 1938 (manuscript).
Marcantonio, A. 1985, On the Definite vs. Indefinite Conjugation in Hungarian.--ALHung. 35, 267-298.
Mikola, T. 1984, Mediale Konjugation in den uralischen Sprachen.--NyK 86, 398-403.
--1986, Beitrage zur nganasanischen Sprachgeschichte.--Festschrift fur Istvan Futaky, Hamburg (Finnisch-Ugrische Mitteilungen 10).
--1988, Geschichte der samojedischen Sprachen.--The Uralic Languages. Description, History and Influences, Leiden--New York--Kobenhavn--Koln, 219-263.
--1990, Mediale Konjugation in den uralischen Sprachen.--[TEXT NOT REPRODUCIBLE IN ASCII] VI [TEXT NOT REPRODUCIBLE IN ASCII] 2, Moskva, 341-344.
--2004, Studien zur Geschichte der samojedischen Sprachen aus dem Nachlass heraugegeben von Beata Wagner-Nagy, Szeged (Studia Uralo-Altaica 45).
Molnar, J. 2001, Zur Verwendung der Objekt- und Subjektkonjugation im Ungarischen und im Mordwinischen.--Zur (Morpho-)Syntax der uralischen Sprachen. Materialien zweier Konferenzen uber Die (Morpho-)Syntax der uralischen Sprachen (Szombathely, Oktober 1997, 1998), Savariae (Specimina Sibirica XVI), 59-92.
Nikolaeva, I. 2001, Syntaktische Analyse der objektiven Konjugation im Ob-Ugrischen.--CIFU IX. Pars V, 454-472.
Posti, L. 1980, The Origin and Development of the Reflexive Conjugation in the Finnic Languages.--CIFU V. Pars I, 111-144.
Pusztay, J. 1975, A determinalas kifejezese az urali nyelvekben.--Nyr., 356-364).
--2001, Determinaltsag a konjugacioban: magyar-jurak-szamojed osszevetes. --Kontrasztiv nyelveszeti tanulmanyok. Az 1999-es Veszprem Alkalmazott Nyelveszeti Kongreszus kontrasztiv szekciojanak eleO oadasai, Savariae (Colloquia Contrastiva VII).
Redei, K. 1966, Der Entstehung der objektiven Konjugation im Ungarischen. --ALHung. XVI 1-2, 111-133.
--1982, Die Herkunft des Imperativzeihens im Ungarischen.--FUF 44, 1-10.
--2002, Gondolatok az urali targyas igeragozasrol.--Ausgewahlte Schriften von Karoly Redei. Hergestellt fur Karoly Redei zum 70. Geburtstag. Ausgewahlt: Honti Laszlo, Budapest, 277-281.
Sammallahti, P. 1974, Material from Forest Nenets, Helsinki (Castrenianumin toimitteita 2).
Tauli, V. 1966, Structural Tendencies in Uralic Languages, The Hague (UAS 17).
Wickman, B. 1997, Uber die Verwendung der objektiven Konjugation.--B. Wickman, Selected Writings Published on the Occasion of His 80th Birthday 7. IX 1997, Uppsala (Ursus. Skriftserie utgiven av Finsk-ugriska institutionen i Uppsala 3), 78-87.
[TEXT NOT REPRODUCIBLE IN ASCII]
AGO KUNNAP (Tartu)
|Printer friendly Cite/link Email Feedback|
|Date:||Sep 1, 2008|
|Previous Article:||'Habere' in Uralic/'Habere' "Auf Uralisch".|
|Next Article:||The Mari system of locative cases/Das system der lokalen kasus im marischen.|