Use of the nominative of Samoyedic substantives as instances of object and attribute/[TEXT NOT REPRODUCIBLE IN ASCII].
Bo Wickman who has thoroughly treated an object form in all Uralic languages has not found a connection between an object case form and definitiveness/indefinitiveness in Samoyedic languages (Wickman 1955 74-144).
N. M. Terescenko has asserted in her monograph on Samoyedic syntax that in three Samoyedic languages, particularly clearly just in (Tundra) Nenets and Enets, partly in Selkup--contrary to the norms valid for Finno Ugric languages--the nominative of substantives indicates the definiteness of a direct object. She claims that in such a use of nominative in those languages, the speaker's particular attention is focused namely on a direct object. N. M. Terescenko adds that in westward dialects of Tundra Nenets the overlapping of singular nominatives and accusatives in consonant-stemmed direct objects is entirely common. According to her data, in some Tundra Nenets eastward dialects in the case of consonant-stemmed direct objects, definite speakers just drop an accusative ending, e.g. Malaya Zemlja (1, 2).
(1) [TEXT NOT REPRODUCIBLE IN ASCII.]
sledge-NomSg (pro sledge-AccSg) fix-Aor1Sg
'I fixed the sledge'
(2) [TEXT NOT REPRODUCIBLE IN ASCII.]
castrated-reindeer-NomSg (pro castrated-reindeer-AccSg) harness-Aor3Sg
'he harnessed a castrated reindeer (to the sledge)'
Nominative is particularly characteristic of the Forest Nenets direct object and of the Selkup Taz dialect direct object ([TEXT NOT REPRODUCIBLE IN ASCII.] 1973 : 178-181).
E. V. Gruskina does not agree with N. M. Terescenko's view as to the latter's assertion that in Selkup the nominative of substantives denotes definiteness of a direct object--E. V. Gruskina claims that the rule cannot be established in Selkup definitively ([TEXT NOT REPRODUCIBLE IN ASCII.] 1980 : 382-385). Probably N. M. Terescenko and unquestionably E. V. Gruskina have drawn their conclusions from Selkup Taz that belongs to Selkup northern dialects. According to E. G. Bekker, in Selkup southern dialects the nominative case denotes indefiniteness of an object altogether, e.g. (3),
(3) [TEXT NOT REPRODUCIBLE IN ASCII.]
I bread-NomSg take-Pret1Sg
'I took bread',
but depending on the context, it may also express its definiteness, e.g. (4),
(4) [TEXT NOT REPRODUCIBLE IN ASCII.]
such man not see-Pret1Sg
'We did not see such a man' ([TEXT NOT REPRODUCIBLE IN ASCII.] 1995: 111-112)
Considering the views of E. V. Gruskina and E. G. Bekker, only two Samoyedic languages--North Samoyedic (Tundra) Nenets and Enets--remain that supposedly consistently denote definiteness of an object by means of nominative.
In his recent article, dedicated to the treatment of the unmarked object in Uralic languages Ferenc Havas writes, "In at least half of the Uralic languages finite indicative forms of transitive verbs may be accompanied by an unmarked object. Interestingly, this unmarkedness indicates the indefiniteness of the object in the Finno-Ugric languages, whereas in the larger part of Samoyedic, its definiteness" (Havas 2008 : 3-4). He adds that "all Samoyedic languages clearly have a category of unmarked object next to an indicative finite verb, and in three of these: (Tundra) Nenets, Enets and Selkup this form expresses the definiteness or the focus function of the object" (Havas 2008 : 5).
It should be noted that in her monograph on syntax of Samoyedic, N. M. Terescenko does not especially distinguish either definiteness of an object or focusing on it, which is not the same even without much theoretical talk. A good example about it lies in her interpretations of the following sentences in Tundra Nenets: with a nominative object (5) and with an accusative object (6) ([TEXT NOT REPRODUCIBLE IN ASCII.] 1973 : 178).
(5) [TEXT NOT REPRODUCIBLE IN ASCII.]
this reindeer-NomSg food-PredestinativePx1Pl buy-Aor1Pl
'this reindeer we bought for food (with an emphasis on this one,
not some other reindeer)'
(6) [TEXT NOT REPRODUCIBLE IN ASCII.]
this reindeer-AccSg food-PredestinativePx1Pl buy-Aor1Pl
'we bought for food this reindeer (the speaker's main attention
is focused on the fact of the purchase, not what sort of reindeer was
In either sentence a different part of speech is focused on: in the first (5), on an object group, i.e. 'this reindeer' and in the other (6), on a predicate group, i.e. 'we bought for food'. (Here F. Havas considers a possibility of having "the definiteness or the focus function of the object", see Havas 2008 : 5.) N. M. Terescenko's interpretations of those sentences could be remodelled respectively: in the first sentence it is emphasized that it was this reindeer that was bought for food, in the second one it was this reindeer that was bought for food. Apart from the differences of the focus, the form 'this reindeer' is in either sentence a definite object thanks to the added demonstrative pronoun t'uku 'this'.
The correctness of N. M. Terescenko's interpretations appears somewhat questionable also in the following cases of Tundra Nenets: with a nominative object (7) and with an accusative object (8) ([TEXT NOT REPRODUCIBLE IN ASCII.] 1973 178).
(7) [TEXT NOT REPRODUCIBLE IN ASCII.]
'pitch the tent (a definite tent or a tent in a visible place)'
(8) [TEXT NOT REPRODUCIBLE IN ASCII.]
'pitch the tent (no matter which or whose)'
Here, too, it may denote that in the first case (7) t e n t is focused on, in the second case (8) p i t c h i n g, and the definiteness/indefiniteness has nothing to do with it.
The same may be supposed in the case of N. M. Terescenko's Enets examples: with a nominative object (9) and with an accusative object (10), in which actual focus is on the stockyard and seeing, respectively ([TEXT NOT REPRODUCIBLE IN ASCII.] 1973 : 179).
(9) [TEXT NOT REPRODUCIBLE IN ASCII.]
'I can see the stockyard (a familiar one)'
(10) [TEXT NOT REPRODUCIBLE IN ASCII.]
'I can see a stockyard (first time)'
Sometimes the change of the case form of the object under discussion is accompanied by the change in the definite/indefinite (objective/subjective) conjugation type in predicate, e.g., Tundra Nenets definite conjugation (11) and indefinite conjugation (12) and Enets definite conjugation (13) and indefinite conjugation (14) ([TEXT NOT REPRODUCIBLE IN ASCII.] 1973 : 178-179).
(11) [TEXT NOT REPRODUCIBLE IN ASCII.]
'undress your child'
(12) [TEXT NOT REPRODUCIBLE IN ASCII.]
'undress the child (it can be said about somebody else's child
or when it is immaterial that the child is yours)'
(13) [TEXT NOT REPRODUCIBLE IN ASCII.]
'he crossed the rivers (namely these rivers)'
(14) [TEXT NOT REPRODUCIBLE IN ASCII.]
'he crossed rivers (no reference to which ones)'
Probably here the choice of the conjugation type is related to focusing.
In one of her articles dedicated to the observation of Samoyedic grammatical cases (nominative, genitive and accusative), N. M. Terescenko ([TEXT NOT REPRODUCIBLE IN ASCII.] 1974) speaks only about the highlighted/non-highlighted position of a direct object, in other words, about the occurrence of emphasis, not mentioning definiteness/indefiniteness of an object anywhere at all. She writes about (Tundra) Nenets, Enets and Nganasan languages as follows: "In the case of a non-highlighted position of a direct object the predicate form contains reference only to the subject of the activity [---] In the highlighted position of a direct object the predicate, on the contrary, contains reference not only to the subject but also to the direct object [---] [---] the highlighted position of the direct object marked by the nominative case is expressed in a peculiar relationship with its predicate" ([TEXT NOT REPRODUCIBLE IN ASCII.] 1974 : 238-239). Substantially it means that the definite (objective) conjugation type is related to a focused object, the indefinite (subjective) conjugation type is related to an unfocused object. But this article by N. M. Terescenko ([TEXT NOT REPRODUCIBLE IN ASCII.] 1974) was overshadowed by her monograph on Samoyedic syntax ([TEXT NOT REPRODUCIBLE IN ASCII.] 1973) (e.g. no reference in Havas 2008).
The assertion that the nominative case in some Samoyedic languages denotes definiteness of an object belongs only to N. M. Terescenko, to my knowledge. Several other researchers have only repeated this assertion of hers (e.g. Havas 2008 : 5). In sum, I cannot consider N. M. Terescenko's point of view that in some Samoyedic languages the nominative denotes definiteness of an object as correct. However, it should be emphasized that in one of her articles that drew little attention, she actually gave up the assertion about definitiveness, as could be seen above.
If the above supposedly different way of marking the Finno-Ugric--Samoyedic definitiveness/indefinitiveness still exists in reality, the reason for the difference remains obscure. Juha Janhunen has written that "There is no need to assume that the synchronic nominative "object" has always been an object. Rather, since it is in the unmarked nominative case of the subject, it is likely to have been the grammatical subject of the sentence." And F. Havas adds to it that "The unmarked accusative governed by transitive finite verbs in Mordvin and Samoyedic can hardly be explained with reference to a reanalysis of the subject...." (Havas 2008 : 15-16.) F. Havas is probably right in saying this. He emphasizes that in a human language in general "object is originally a Par excellence unmarked category. It follows that what demands a linguistic explanation is not why the object is unmarked in certain cases but just opposite: it is its markedness that must be historically explained" (Havas 2008 : 31). Likewise, N. M. Terescenko wrote, "It is quite possible that denoting a direct object by nominative (by a fundamental case) is an old phenomenon, as a remnant preserved in some territorial dialects of Nenets" ([TEXT NOT REPRODUCIBLE IN ASCII.] 1973 : 181). Therefore, in the Samoyedic cases where nominative marking is common, there is obviously no reason to speak about definiteness/indefiniteness of an object.
2. The case form of the substantival attribute in Samoyedic is either genitive provided with the case ending *-n or unmarked nominative. The case ending of substantival attributes in Samoyedic languages is expressed by a genitive suffix, as a general rule (see also Honti 2008 : 303). However, rather a large number of exceptions to that general rule of attribute genitiveness in Samoyedic are known to exist in (Tundra) Nenets, Enets and Nganasan: instead of genitive they use nominative. E.g. Tundra Nenets Gen + Nom (15), cf. e.g. Forest Nenets Nom + Nom (16).
(15) [TEXT NOT REPRODUCIBLE IN ASCII.]
(16) [TEXT NOT REPRODUCIBLE IN ASCII.]
As far as I know, Selkup has characteristically no nominative (at least a single-stem nominative) attribute, yet it is not completely unknown, e.g. Nom + Nom (17).
(17) [TEXT NOT REPRODUCIBLE IN ASCII.]
Aulis J. Joki wrote about the Kamass language that sometimes it has "genitive forms without a case ending" that, based on his examples, can also be interpreted as attributes of the nominative form (Joki 1944 : 132). It is not quite sure whether the few Kamass cases mentioned by A. J. Joki are not just spelling errors, e.g. di n e [TEXT NOT REPRODUCIBLE IN ASCII.] 'dieses Weibes Hals wieder durchschnitt er' (? pro ne-n), or resulting from phonetic changes, e.g. [TEXT NOT REPRODUCIBLE IN ASCII.] 'God's daughter' (? < in fact: [TEXT NOT REPRODUCIBLE IN ASCII.]), since the attribute case form in this language is very persistently the genitive with the suffix -n. Cf. much more definite cases of Kamass Nom + Nom, seem to confine themselves to determining persons' gender, e.g. [TEXT NOT REPRODUCIBLE IN ASCII.] 'Manns-person', n e-k'uza 'Weibs-person'. In all above Samoyed nominative cases the attribute expresses relevance of kinds, as is characteristic of Uralic languages in the case of nominative.
N. M. Terescenko has stated that in the case of Turkic izafet type of attribute constructions, in which the main word is provided with a possessive suffix, the sentence structure changes in Nganasan. For example, the sentence (18) is replaced with another sentence (19) ([TEXT NOT REPRODUCIBLE IN ASCII.] 1979 98-99).
(18) [TEXT NOT REPRODUCIBLE IN ASCII.]
brother-GenSgPx1sg garment-Nom very warm
'my brother's garment is very warm'
(19) [TEXT NOT REPRODUCIBLE IN ASCII.]
brother-NomSgPx1Sg garment-NomSgPx3Sg very warm
'my brother, his garment very warm is'
According to A. A. Kim, in Selkup southern dialects there is a phenomenon that corresponds to the use in the Nganasan language: instead of the izafet construction of an attribute group the sentence is constructed in the way that the nominative is used instead of the genitive attribute to express a possessor, e.g. Ket (20).
(20) [TEXT NOT REPRODUCIBLE IN ASCII.]
mother-NomSgPx1Sg hat-NomSgPx3Sg new
'my mother, her hat new is'
A. A. Kim regards such a use in Selkup southern dialects as a possible result of recognizable marking of the 1st person possessive form that otherwise would not be identifiable owing to phonetic developments (PxSg1 and GenSg, in both of them -n is used) ([TEXT NOT REPRODUCIBLE IN ASCII.] 1986 : 56; 1990 : 101-102). A. A. Kim's model for the interpretation is in all probability based on the respective Nganasan interpretation offered by N. M. Terescenko.
E. G. Bekker draws an example from Selkup southern dialects in which the same sentence type nominative is not provided with a possessive suffix (21).
(21) era alagut menemba
Old-man-NomSg boat-NomSgPx3Sg old-be-Narrative3Sg
'old man, his boat old is'
But she indicates the similarity between such Selkup nominative attribute cases and Turkic izafet, and more concretely II izafet in Chulym Tatar ([TEXT NOT REPRODUCIBLE IN ASCII.] 1995 : 113-114).
Both N. M. Terescenko and A. A. Kim state that in the case of the supposed change of sentence structure in Nganasan and Selkup, the izafet construction is lacking in the changed sentences (according to them also Kiinnap 2004 : 144-145). Actually, they most obviously mean the lack of the so called Turkic III izafet type, in which case the attribute is of the genitive form. Hereby it is necessary to consider the so called Turkic II izafet type, in which case the attribute is of the nominative form, occurring both in Nganasan and Selkup. The incidence makes it possible to interpret the Nganasan sentence type (19) the nominative form [TEXT NOT REPRODUCIBLE IN ASCII.] and Selkup sentence types (21) and (20) nominative forms era and [TEXT NOT REPRODUCIBLE IN ASCII.] not as a subject but an attribute of the nominative form. (See for those Turkic izafet constructions in Honti 2008 : 293-295.) The so called Turkic II izafet type in Nganasan and Selkup derives most likely from Turkic languages.
It does not seem logical to replace an attribute in attribute constructions of the observed languages with a subject, more so, in two geographically relatively distant languages. I fail to imagine that speakers of a language who regularly use attribute constructions, replace the latter in certain cases for whatever reasons (e.g., for phonetic ones) with an altogether different construction in which a subject is used instead of an attribute, turning the whole construction into something like: The old man's boat is old [right arrow] The old man. His boat is old. The reader might guess whether in their own native language the use of a subject as an equivalent to the attribute construction can be considered as logical. Hardly ever.
Based on Eugene Helimski's data, the nominative in Selkup Taz dialect may, besides genitive, replace incidentally also other case forms: accusative, illative, locative, translative, e.g., [TEXT NOT REPRODUCIBLE IN ASCII.] qumyn i m a (Nom pro Transl ima-tqo) [TEXT NOT REPRODUCIBLE IN ASCII.] 'became a rich man's wife'. In his opinion the reasons for the phenomenon may be phonetic, on the one hand, and grammatically incorrect description of situations, on the other: first, an actant of a situation is named in the nominative case, and only thereafter some "coherent" text follows. ([TEXT NOT REPRODUCIBLE IN ASCII.] 1980 : 172, see for few respective incidents in Selkup southern dialects in [TEXT NOT REPRODUCIBLE IN ASCII.] 1978 27-28.) At the same time, as related to nominative, an impression may be left about a tendency of the formation of two separate sentences, e.g. Nganasan [TEXT NOT REPRODUCIBLE IN ASCII.] 'My brother. His garment very warm'. Additionally, in Nganasan and partly in Selkup, when replacing a genitival main substantive with its nominative form, a possessive suffix is adhered to the nominatival main substantive. It is as if the possessive suffix puts the finishing touches to the second sentence: after having "unhooked" the first sentence from the second, the latter reminds one of an "undotted i" which can only be adjusted by the possessive suffix (marking the "unhooked" possessor as an expression of "full value" for the second sentence). Thus there remains a principal opportunity that respective interpretations of nominatives, offered by N. M. Terescenko and A. A. Kim are valid as they indicate exceptions: interpretations are based on grammatically incorrect expressions.
AGO KUNNAP (Tartu)
University of Tartu
* This article is supported by the Estonian Science Foundation, grant No. 7724.
|Printer friendly Cite/link Email Feedback|
|Date:||Jun 1, 2009|
|Previous Article:||Attributive possession in Udmurt language/[TEXT NOT REPRODUCIBLE IN ASCII].|
|Next Article:||Language maintenance and Finno-Ugric world/[TEXT NOT REPRODUCIBLE IN ASCII].|