Two aspects of the study of animacy and case: A contribution to the long-range history of Russian and Serbian

László Desző

[Články]

(pdf)

Dva aspekty zkoumání životnosti a pádu: příspěvek k vývoji ruštiny a srbštiny

This paper is divided into two sections: after a typological introduction, the first section examines the role of animacy in the shaping of the first argument of sentence structure in Pre-Indo-European (PreIE) of active type. Animacy was not relevant in Proto-Indo-European (PIE) of accusative type but it reemerges in Slavic in the expression of the accusative (A) which is the typical case of the direct object (DO); i.e. of the second argument. This process will be dealt with from a rather unusual aspect – case frequency – in the second section.

 

0. Introductory remarks

In a recent version of structuring typology, Lehmann (1996) distinguishes agreement (active and class) types and government (accusative and ergative) types according to the dominant role of these two major procedures. In the agreement type, structuring is concentrated on the first argument (ARG1), which can be active (animate) or inactive (inanimate); the second argument is less differentiated. The verb-subject relation expressed by agreement is dominant. Animacy has an important role in the expression of ARG1 and its agreement with the verbs, which can be active or stative. In the government type the verb-object relation becomes dominant and animacy can, but need not, be relevant. In Slavic languages with case systems, the accusative has a distinct form variant: the genitive-accusative for animate nouns. In Hungarian, animacy has no role in structuring, however, specification or definiteness has been relevant in the shaping of the forms of direct object from Proto-Uralic up to Modern Hungarian. The two categories: animacy and specification are fundamental to sentence structuring in various languages; they manifest variations within the same category and they can even appear in combination.

Traditional grammar was founded on the empirical base of accusative languages in which grammatical relations played a fundamental role. In languages with case systems, the cases are defined according to these relations. With the increase of typological research these terms were complemented by semantic roles, labelled first as deep cases in grammatical theory. In the sixties I intended to compare Serbian with other languages within a typological framework, therefore, I used also semantic roles (Dezső, 1982 (1970)). However, following the Prague tradition, I characterized them with features among which ‘active’ and ‘animate’ had a central role. At present, semantic roles and categories like ‘animate’, ‘active’ are organic components not only of typology and linguistic theory but also of long-range history dealing with changes such as from active to accusative type in Indo-European which was accompanied by changes in morphological, ordering typology and by the emergence of a single-dimensional case system according to Hjelmslev’s typology.

 

[106]1. Animacy in case system from Pre-Indo-European to Slavic

In PreIE animacy, together with activity, was a fundamental factor in sentence structuring. The case system could not exist because the relation between ARG1 and V was central, that of V and ARG2 was not yet diversified enough to give rise to a case system based on a relatively clear distinction of the major grammatical and adverbial relations and requiring synthetic word structure of agglutinative or inflective type for nouns. Let us have a look at the three earliest variants of the first argument (A, B, C) in early PreIE.

A) ARG1, the typological subject, is an inherently animate and active 1st and 2nd person pronoun or active noun, marked by a formal device (-s which later becomes one of the Nominative forms). It is in agreement with an action verb which has a full paradigm -m, -s, -t (which gave rise to Slavic verbal forms).

B) ARG1 is an inactive, inanimate noun and has either the zero form or -m (the latter will give origin to the neuter). ARG1 is in accordance with a stative verb in 3p only (-e).

C) These are the typical ARG1 and V structures. However, a number of verbs expressing physical state, involuntary emotions, uncontrolled will, desire and a possessor with the verb ‘to be’, have ARG1 in the „dative“ and the verb has a stative form in agreement with ARG2. Similar constructions were preserved in ancient IE languages and reemerged in Slavic (Rus. mne nravitsja, mne chočetsja, u menja jest’).

When PreIE of the active type changed into PIE of the accusative type, constructions A and B were able to enter into the new system as the subject complemented with a specified ARG2, expressing direct object, indirect object, adverbial of direction and place (cf. Gamkrelidze – Ivanov, 1984 (1995), p. 233–276; Lehmann, 1993, p. 187–233); PIE became a flective language with a case system uniting the various paradigms of inflection. As in a consistent nominative-accusative language the subject is not differentiated according to activity or animacy, and so these categories lost their relevance. Structure C had no place in the new system; the category of person and the unusual form of the subject deviated too much from the standard system.

In PIE the subject was expressed by the N opposed to the A of the direct object (DO) according to the standard rules of the nominative-accusative type. However, in a number of early IE dialects, there were particular constructions of DO with verbs which required the dative subject in PreIE according to the active type of structuring. Such constructions of early IE dialects were described in the standard Neogrammarian grammars (Brugmann, 1911, p. 464–642, esp. 565–616; Delbrück, 1893, p. 181–400, esp. 308–333); I have established the relationship between the semantic groups of verbs with dative subject in active languages and the genitive or partitive object in Slavic and Balto-Finnic (Dezső, 1982, p. 41–59) before the reconstruction of PreIE as an active language.

The dative subject, however, can coexist with nominative-accusative structuring. The dative subject is characteristic for Dravidian languages and could be reconstructed for Proto-Dravidian (Zvelebil, 1990, p. 23). It was reestablished in Indo-Aryan group of IE and is one of the major characteristics of the South Asian (Indian) lin[107]guistic area (Abbi, 1994, p. 57–100). The dative subject has reemerged in Slavic (cf. Rus. mne chočetsja, mne nravitsja) or in Romance group (It. mi conviene ‘to me (it) is useful’, mi piace ‘I like it’) but its use is limited. Such constructions have appeared also in Hungarian (nekem kell ‘I need it’, nekem tetszik ‘I like it’). In early Indo-European dialects constructions with dative subject were isolated residues of PreIE. In Indo-Aryan, Slavic, Romance they are secondary because they can appear also in accusative type languages according to the testimony of Hungarian.

I shall limit my comments to direct objects expressed by the G or more precisely by the genitive-ablative in Baltic and Slavic. In Baltic they have been preserved from the Baltic dialect of IE, and were probably extended. In Lithuanian they are used with verbs denoting (1) desire and other similar feelings, (2) fear, shame and the like, (3) want, loss or having enough, (4) asking or requesting and also with partitive and negated object (Ambrazas, 1997, p. 503–504). A Slavicist can easily find an overlap in Lithuanian and Slavic constructions. Under the influence of Baltic such constructions spread into the Balto-Finnic group of Uralic (cf. Larsson, 1983 for a detailed discussion).

In Lithuanian the N and the A have been well differentiated, the genitive object remained preserved, possibly extended. In Slavic the use of genitive object has been more and more limited. One of the possible parameters of the Slavic historical syntax is the degree of limitation on the genitive object and its variation (cf. Deže, 1983, for a comparison of Serbian, Russian and Ukrainian). They are relatively better preserved in Old Serbian (Dezső, 1982 (1970), p. 138–140) compared to modern language.

In this way, we can observe two parallel processes in the history of Slavic: the tendency to limit the genitive object which is typologically expressed by non-animate nouns, and the spread of the genitive-accusative from pronouns to masculine nouns in singular and plural, and to feminine nouns in plural, if they denote a person or animate noun. In both tendencies, the animacy has a major role, implicitly or explicitly, but these processes take place in the morphological system of the noun (cf. Klenin, 1983, on the animacy in Russian).

These morpho-syntactic changes have produced a variant of the A and, if we accept Jakobson’s proposal (1936), a variant of the G in Russian: the partitive expressed by -u. I add that in Balto-Finnic a particular partitive case was formed (from one of the ablatives) to render the functions mentioned above in Lithuanian. In Baltic, Slavic subjects too can have a partitive function and the genitive subject with the verb ‘to be’ seems to resist the tendency of loss of expression of partitive relations, more than the G of the object.

In various languages animate nouns can occur not only as subjects and direct objects but also as indirect objects in the semantic roles of beneficiary or addressee. Still in late PreIE these functions could be expressed by clitic -ei which was used with animate nouns, meanwhile the ablaut variant -i was attached to inanimate nouns denoting place. In this way, initially, the D was a particular form of animate nouns and the L inanimate. The tendency to have a particular form for animate masculine nouns in the D appeared in Ukrainian (-ovi, -evi). This language has case variants according to animacy (Bilodid, 1969, p. 60–62). Only animate masculine nouns have -a in the G-A; [108]in the D singular: -ovi, -evi is typical for animate masculine nouns, in the N plural only animate nouns have -ove.

Thus, we have arrived to the questions of declension. Before examining it, we will look at the Indo-Aryan languages (IA) in order to widen the typological perspective within Indo-European. In Slavic, the G is a case of possessor in noun phrases and an argument of the verb phrases. This is a typologically relevant characteristic which differentiates Slavic from many other languages, among them IA which applies the G in possessive constructions only. The problems raised in our discussion could have been irrelevant or very different if the G was only the case of possessor in noun phrases in Slavic. However, such a G must not exclude the role of animacy in a case system. In IA there are two forms expressing direct object: the zero form, identical to the N, and the D; the former is used with inanimates, the second mostly with nouns denoting a person or animate noun (for details see Zograf, 1976, p. 55–64). However, the lack of distinction between the N and the A would not have been sufficient to require two distinct forms for a direct object based on animacy without an inherent tendency to the expression of this category.

I have approached the problem of the expression of the direct object in Slavic from a typological point of view. It is to be complemented by a treatment from within Serbian proposed by Gortan-Premk (1971, p. 161–173) which I shall examine elsewhere in detail. According to her, in the old language the function of the A was confined to objects in physical contact with the verb of action which extended to the object; the G was used when these factors were lacking. In modern language, the function of the A has become more ample, also involving objects which only have a connection with the action of the verb etc.; in this way, the A limits the use of the G.

In the next chapter we shall deal with the frequency data of medieval Russian and Serbian Church Slavic which was close to Old Church Slavic and does not show the state of Old Serbian. Russian literary language was conservative: „in the field of expression of object relations, the A and the G were consistently preserved as basic forms“ (Borkovskij, 1978, p. 356). The norms valid for Modern Russian began to be shaped from the 16th–17th centuries. These facts may contribute to the explanation of the close figures of medieval Serbian and Russian data and the difference between medieval and modern Russian in the following section.

 

2. Animacy in Russian case system: a frequency analysis

Saussure’s antonymy: synchrony vs. diachrony had an important role in the history of linguistics. The descriptive study was directed at the synchronic system of language and separated from history. However, having reached this objective in concrete languages and linguistics, synchrony and diachrony must have been connected in concrete and typological linguistics. Saussure’s langue and parole antonymy have contributed to establishing a language system by separating it from secondary phenomena occuring in its functioning. However, having reached this objective, linguists should clarify how language system functions in speech or, with a more recent term, in discourse. Here, I shall compare various synchronic systems functioning in a historical-typological per[109]spective. First, I shall present certain data from Greenberg’s analysis of the Russian case system (1970), then I shall examine medieval Russian and Serbian Church Slavic and Russian literary language comparing them with the language of medieval Hungarian religious literature. The statistical data of case systems will be in the center, but they imply a number of issues of comparative and typological linguistics: the different types of case system, the varieties of expression of grammatical and adverbial relations and marking theory. The relationship between animacy and case categories will be the central issue with respect to the correlation of these cases: the nominative (N), the accusative (A) and the genitive (G) which cannot be separated from the dative (D), the instrumental (I) and the locative (L) or prepositional case (P) in Slavic. The N and the A are unmarked and are related to the other cases (cf. Greenberg, 1966, p. 37–38, 1970). They are frequently expressed by the zero form with possible differentiation of the A from the N. Unmarked status also implies higher frequency in discourse. In accusative languages the linguistic relevance of the category of person or animate is manifested in the forms of grammar and their use in discourse.

Persons or animates constitute a relatively small part of lexicon as one can observe in Proto-Indo-European or Proto-Uralic (cf. Décsy, 1991, 1990). The category of person is more significant in word formation but only 6 of 50 PIE formatives and derivational suffixes are used for the formation of this category (Décsy, 1991, p. 49–53), however, the number of nouns denoting person increases by deverbal and denominal derivation. Mel’ničuk (1986, p. 52–55) presents the Slavic suffixes of persons related to activity (-nik, -č, -ar, -ak, -c, -tel’) or characteristics, origin etc. (-ik, -ak, -ec), the suffixes can differentiate sex (-ka, -ica, -inja, -yni for females). Nonetheless person is only one of the derivational categories beside concrete objects, instruments, materials, abstract and collective nouns or those of subjective evaluation. Arriving from lexicon via word formation to morphology we find that person becomes a central category, the others are not or barely distinguished in Slavic.

The central issue of our treatment is the genitive-accusative expressing of animacy in Slavic. The first legitimate question is why it is necessary to have particular forms distinguishing animacy. A plausible answer is because the N and the A are not consistently differentiated by the case forms as in Baltic or by fixed word order as in English or Italian. The next question is why the accusative has double forms: a possible answer indicates that the A is more marked than the N. Then one can ask why the G is used for coding the category of animacy. The answers to the former questions were typologically plausible and generally valid. Here, the explanation is more specific. (1) The G was the secondary case of direct object if partitivity or negation were involved or the verbs belonged to certain lexical groups. (2) The G had the advantage that its major role was that of denoting a possessor in a noun phrase which was quite distinct from its use in a verbal phrase.

 

Modern and medieval Russian

Greenberg’s analysis of the Russian case system according to the semantic groups of nouns (1970) permits us to have a closer look at the use of noun classes in case [110]forms. I cannot present his table (ibid. 220) and his explanations, but I shall compare the percentages of all nouns, personal common individual nouns (PersN) and pronouns of the 1st and 2nd person (PersPron) with my comments.

 

	N	G	D	A	I	P
all nouns	33.6	24.6	05.1	19.5	07.8	09.4
PersN (11,5%)	54.1	22.5	06.9	07.0	08.0	01.5
PersPron (15,6%)	51.8	21.8	14.5	09.5	2.2	0.2

 

1. There is a significant difference in the figures of the N. It is far more frequent with PersN and PersPron because the N denotes subjects as agents or experiencers. The figures for big groups like concrete countable or place nouns are low (23.0, 11.8).

2. The A is far more frequent in the whole set of nouns than in PersN and PersPron because persons are less frequent as DO. The share of body parts and concrete countable nouns is high (36.5, 32.0). This means that the differentiation between the accusative and the accusative-genitive is a problem for the A case and not only for the person category; i.e. the N and the A were to be differentiated and this could have been done clearly by the category of animacy or person.

3. The D is far more frequent with PersPron because it is an optional and easily applicable constituent if expressed by PersPron.

4. The great difference between the figures of P indicates that adverbial, mostly locative relations are far more frequently expressed by inanimate nouns, especially by those denoting place and place institutions (23.8, 24.1).

Farmini’s book (1976) examines the statistical data of Russian, Bulgarian and Serbian medieval literature according to genres. However, only Russian literary language is presented by confessional-narrative literature and chronicles as well as Church Slavic religious texts, Serbian means the Serbian redaction of Church Slavic: Miroslavovo evandjelje (12th century), Bukanovo evandjelje, Šestodnev (13th century), Bucharestestkaja psaltyr, Šišatovac apostol, Sinodik (14th century). The following table presents Serbian and Russian medieval data of case frequency (in percentages) for both singular and plural nouns and similar data from Russian chronicles (Farmini, 1976, p. 196, 190). These medieval data are complemented by Modern Russian common nouns as proper nouns are rare in medieval texts (cf. Greenberg, 1970, p. 210).

 

	Serb.	Rus.	Rus. chron.	Mod. Rus.
N	27.4	24.4	24.3	28.3
A	27.8	28.6	30.7	21.8
G	17.7	19.3	18.8	26.0
D	10.7	09.7	10.2	05.1
I	06.4	06.1	08.5	08.6
L	07.6	08.4	07.0	10.3
Total	97.3	96.5	99.5	100.0

 

Before examining medieval data, we should pay attention to the significant differences between the frequencies of medieval and Modern Russian. They are manifested in the lower frequency of the A (28.6 and 30.7 vs. 21.8) in Modern Russian and [111]in the lower frequency of the G (19.3, 18.8 vs. 26.0) in medieval Russian. The tables of Farmini’s book usually show considerable divergencies in data of the same period, therefore, chance factor can account for differences within certain limits. I wish to underline the role of the semantic groups of nouns. Certain major groups like parts of the body and concrete countable nouns have a much higher frequency in Modern Russian (36.5 and 32.0) in the A, but their share is not sufficient to raise the average frequency. However, if certain semantic groups constitute the majority of the nouns in a text, they will also influence the average frequency. The lower figures of the G can be due to a stylistic factor: multiple genitive constructions, frequent in modern texts. These are factors which should be considered in text analysis. However, the medieval data are homogeneous enough, which could indicate similar lexicon and style within the same period. Despite the differences in percentages the average of the N + A is nearly the same: 53.0 in medieval and 50.1 in Modern Russian which is due to the unmarked status of these cases.

 

Medieval Russian and Hungarian

We can compare the medieval Russian data for case frequency with those of medieval Hungarian (MNyTNy II.1, p. 358). The Hungarian case system belongs to a different type and the N and also usually the G have zero forms. Therefore, we should exclude the N and the G from comparison in Russian and only count those case suffixes in Hungarian which have parallels in Russian cases and prepositions (cf. Dezső, 1982, p. 85–98). The set of Russian cases include the A, D, I, L. The Russian A corresponds to the Hungarian A (-t 39 of 52%) and to the direction-contact cases (-ba ‘in’, -ra ‘on(to)’); the Russian D to the Hungarian D and the case of direction towards (-hoz); the Russian I is to be related to the Hungarian (-val); the Russian L to the two Hungarian locative cases of place (-ban ‘within’, -n ‘on the surface’). I had to exclude the Hungarian ablative and terminative cases rendered in Russian by the G with prepositions, the sporadic use of the marked G and other scarcely used case suffixes. We have arrived at the following comparative table (in percentages): 

 

	Russian	Hungarian
A	56	52	(t. ba. ra)
D	18	25	(nak. hoz)
I	11	11	(vai)
L	15	12	(ban. n)
Total	100	100

 

When I saw the relative correspondence, I was at first surprised. Then I realized that, despite the differences in morphology and case systems, both languages are of the accusative type, their formal devices will correspond (if one knows their systems) and, last but not least, the texts are of the same genre and biblical texts have a dominant role in them. Beside the grammatical and locative relations, the types of government also show similarities. As the data of Serbian Church Slavic are close to those of Russian, the results are also valid for Serbian liturgical texts. On the basis of these [112]data one can make certain conjectures, in the methodological sense of the term, to be proved or disproved.

1. Given the similarity of genre, the distribution of morphological devices, comparable with each other, tends to be similar.

2. Such similarities in the distribution of morphological devices indicates a similar share of grammatical and locative relations expressed by them: direct object, indirect object, locative relations of direction and place.

3. If conjectures 1 and 2 are not false, one can assume that the distribution of semantic roles will also be similar.

4. Even if conjectures 1, 2, 3 are not false and they are valid for the accusative type structuring, despite their validity for the ergative type, especially the active, is to be verified even if on the „deepest“ level of semantic roles there is a similarity.

Conjectures 1 and 2 are probable, 3 is possible, 4 may be possible but it goes too far. At any rate, one could check all conjectures using texts translated into various types of language. This is possible, for instance, using translations of the Bible as a first approximation for comparison. The reader might have the impression that I have gone too far, but this is typical of conjectures.

Greenberg (1970, p. 219–20) examines the role of frequency in linguistic studies. In his view frequency is neither trivial nor an epiphenomenon, it is fundamental because the linguist’s intuition about basic and derived phenomena „seems to be guided largely by frequency knowledge without actual counting. The investigation of the connections between what is basic and what is frequent would seem to be a topic of fundamental theoretic interest“ (ibid. 220). Frequency is an important factor in stylistics and, according to me, stylistic factors are to be considered in frequency studies of grammar. In our case, frequency clearly shows that the linguists’ intuition about the fundamental role of the N and the A expressing subject and object relations is correct. The distinction between these relations should be coded by form devices or ordering. The rules regulating the application of devices denoting the same relation, in our case that of direct object, must be clear; and animacy is a category appropriate to this role. As we have seen, in Indo-European prelanguage animacy was fundamental, it then appeared in the Slavic dialect of the protolanguage. In Slavic, having entered in coding grammatical relations, animacy „acts“ within a subsystem of forms. I have used biological metaphors, but I think they are appropriate in substituting a longish linguistic exposition of well known facts.

 

Concluding remarks on frequency analysis

The first section examined the change in Russian and Serbian morpho-syntax in a long-range history in which animacy was one of the basic factors. It was a so-called qualitative systemic analysis which arrived at medieval Slavic, also considering its later development. I have described Old Serbian (Dezső, 1982 (1970)) and late medieval Ukrainian (Deže, 1967) and here I did not intend to characterize either Serbian, or Old Russian or Ukrainian well-known from historical grammars. I turned to a frequency analysis of the case system which expressed grammatical relations. Animacy [113]was not an overt, immediate factor in a case-system analysis. As statistical research is usually not appreciated by linguists, it was by chance that I found Farmini’s book, the data of which could be related to my topic even if it was not optimal. At any rate, no text-statistical analysis for PreIE, PIE, PSI is possible, the earliest available texts were those of Old and Middle Church Slavic beside Old Russian and Old Serbian.

The novelty of cross-linguistic frequency analysis was in the correspondence of Slavic and Hungarian data. The differences in morphology and case system could not influence considerably the frequency of grammatical and adverbial relations which is based on semantic roles. It turned out that frequency represented the invariant because it reflected invariant factors. Formulated in this way, it is no longer a surprise despite the general belief that statistics is something variable, not reliable and not to be accounted for. Here, I have dealt with text frequency. Earlier I examined another aspect of frequency: that of valence dictionaries. It revealed the cross-linguistic correspondence of government relations in Serbian and Hungarian (Dezső, 1997). They showed fundamental similarities in frequencies based on the same factors, despite the variety of formal devices, which also included prepositional systems.

I am aware of the gap between the first systemic and the second quantitative section of this paper. There is a number of reasons for this. The major one is general: the two approaches have been separated. This follows from the separation of langue and parole in linguistic analyses existing long before Saussure. The dominant systemic analysis has not been accompanied by statistical examination which continues to be sporadic and unrelated to the former. My article intended to show that quantitative analysis has a fundamental role in cross-linguistic comparison. This has been proved by a number of typological studies. Frequency is one of the basic factors in marking theory which permeates the typology of grammatical categories (cf. Greenberg, 1966). A task for the research program in both genetic and comparative studies should be to account for the role of quantitative aspect in linguistics which has neglected it despite the evidence of other sciences.

 

REFERENCES

 

ABBI, A.: Semantic Universals in Indian Languages. Shimla 1994.

AMBRAZAS, V. (ed.): Lithuanian Grammar. Vilnius 1997.

BILODID, I. K. (ed.): Sučasna ukrajins’ka mova. Morfolohija. Kiiv 1969.

BORKOVSKIJ, V. I. (ed.): Istoričeskaja grammatika russkogo jazyka. Sintaksis, prostoje predloženije. Moskva 1978.

BRUGMANN, K. (ed.): Vergleichende Laut- und Stammbildungslehre und Flexionslehre nebs Lehre vom Gebrauch der Wortformen der indogermanischen Sprachen. Zweiter Bearbeitung. Zweiter Band: Lehre von Wortformen und ihrem Gebrauch. Zweiter Teil. Straßburg 1911.

DÉCZY, G.: The Uralic Proto-language: A Comprehensive Reconstruction. Bloomington 1990.

DÉCZY, G.: The Indo-European Proto-language: A Computational Reconstruction. Bloomington 1991.

DELBRÜCK, B.: Vergleichende Syntax der indogermanischen Sprachen. Ertster Teil. Straßburg 1893.

DEZSŐ, L.: Typological Studies in Old-Serbo-Croatian Syntax. Budapest 1982. (English version of the Hungarian original: Ószerhorvát mondattan, 1970.)

DESZŐ, L.: Studies in Syntactic Typology and Contrastive Grammar. Budapest 1982.

[114]DESZŐ, L.: Partitività in italiano: note comparativistiche sulle lingue indoeuropee e uraliche. In: XXème Congrès International de Linguistique et Philologie Romanes, t. III, s. IV, Typologie des langues romanes. Tübingen 1993, p. 81–89.

DEZSŐ, L.: Gramatika i leksika u proučavanju valence glagola. In: Naučni Sastanak slavista u Vukove dane. Beograd 1997, p. 25–33.

DEŽE, L.: Očerki po istorii zakarpatskich govorov. Budapešt 1967.

DEŽE, L.: Ob istoriko-tipologičeskoj charakteristike ukrainskoj grammatiki v sopostavlenii s russkoj i serbochorvatskoj. In: Hungaro-Slavica. Budapešt 1983, s. 17–29.

FARMINI, L.: Il sistema della declinazione nominale nei testi slavi dei secoli XI–XIV. Un’ analisi diacronico linguistico-computazionale. Galatina 1976.

GAMKRELIDZE, T. – IVANOV, V. V.: Indo-European and Indo-Europeans. Part I. Berlin 1995 (English version of the Russian original of 1984).

GORTAN-PREMK, D.: Akuzativne sintagme bez predloga u srpskohrvatskom jeziku. Beograd 1971.

GREENBERG, J. H.: Language Universals with Special Reference to Feature Hierarchies. The Hague 1966.

GREENBERG, J. H.: The relation of frequency to semantic features in a case language (Russian). In: On Language. Selected Writings. Stanford, Cal. 1970, p. 207–226.

JAKOBSON, R.: Beitrag zur allgemeinen Kasuslehre (Gesamtbedeutungen der Russischen Kasus). TCLP, 6, p. 240–288 (SW, II, 4, p. 23–71).

KLENIN, E.: Animacy in Russian. A New Interpretation. Columbus, Ohio 1983.

LARSSON, L.-G.: Studien zur Partitivgebrauch in den ostseefinnischen Sprachen. Uppsala 1983.

LEHMANN, W. P.: Theoretical Bases of Indo-European Linguistics. London 1993.

LEHMANN, W. P.: Agreement languages. In: The Twenty-second LACUS Forum. Chapel Hill 1996, p. 19–34.

MEL’NIČUK, A. S. (ed.): Istoričeskaja tipologija slavjanskich jazykov. Fonetika, slovoobrazovanije i frazeologija. Kijev 1986.

MNyTNy = A magyar nyelv történeti nyelvtana (The Historical Grammar of Hungarian Language) II.I. Budapest 1992.

ZOGRAF, G. A.: Morfologičeskij stroj novych indoarijskich jazykov (Opyt strukturno-tipologičeskogo analiza). Moskva 1976.

ZVELEBIL, K. V.: Dravidian Linguistics. An Introduction. Pondicherry 1990.

 

R É S U M É

Dva aspekty zkoumání životnosti a pádu: příspěvek k vývoji ruštiny a srbštiny

Stať má dvě části: po typologickém úvodu autor v prvé části zkoumá úlohu životnosti při formování aktivní větné perspektivy v PreIE. V druhé části se zabývá frekvencí pádových forem.

Autor si je vědom rozdílu mezi první systémově-strukturní a druhou kvantitativní částí stati. Konstatuje, že v dosavadním jazykovědném bádání bývají oba přístupy nedostatečně propojeny, což má mimo jiné příčinu v dosud značně oddělovaném myšlení o langue a o parole. Následkem toho pak systémové analýzy nejsou vždy doplňovány zcela paralelními výzkumy kvantitativními. Stať podle autora ukazuje, že kvantitativní analýzy mají zásadní význam pro porovnávání jazyků. Frekvence je jeden ze základních faktorů, které prostupují i typologii gramatických kategorií. Zanedbávání soustavného kvantitativního výzkumu neprospívá ani vývojovým, ani komparatistickým výzkumným programům.

Department of General Linguistics
University of Padua, Italy

Slovo a slovesnost, ročník 62 (2001), číslo 2, s. 105-114

Předchozí Alena Macurová: Jazyk a hluchota

Následující Jana Hoffmannová: Odborný diskurz a vědecká komunikace v nových publikacích