Short term morphosyntactic change

Suprun (1957, 1959) - STMC bibliography


Suprun A. E. 1957. K upotrebleniju roditel´nogo i imenitel´nogo padežej množestvennogo čisla prilagatel´nyx v sočetanijax s čislitel´nymi dva tri, četyre v sovremennom russkom jayzke. Učenye zapiski Kirgizskogo gosudarstvennogo pedagogičeskogo instituta 3. 72-84.

[Reprinted: Sočetanija čislitel´nyx dva tri, četyre s prilagatel´nymi, opredeljajuščimi sčitaemye predmety. In: Suprun A. E. 1959. O russkix čislitel´nyx. Frunze: Kirgiz State University. 61-75.]


The author addresses the variation between the nominative and the genitive case in modifiers in Russian quantified expressions with the numerals ‘two’, ‘three’ and ‘four’. This variation is observed when a quantified expression occurs in non-oblique cases, i.e. in the position of a subject (nominative case), as in (1), or direct object (accusative case), as in (2):

(1)   Dve   sil´n-ye / sil´n-yx   ruk-i   ottolknuli ego
  two strong-PL.NOM.PL / strong-PL.GEN   hand-SG. GEN   pushed.away   him
  ‘Two strong hands pushed him away.’
(2)   On  uvidel dve  massivn-ye /massivn-yx dveri
  he saw  two  heavy-PL.NOM / heavy-PL.GEN  door-SG. GEN
  ‘He saw two heavy doors.’

The choice of the case is analyzed with respect to the following conditions:

  • animacy and gender of the noun;
  • word class of the modifier (adjective vs. pronoun vs. participle);
  • number of the predicate;
  • word order within the quantified expression (numeral + modifier + noun or modifier + numeral + noun);
  • syntactic function of the noun phrase (subject vs. direct object);
  • semantics of the numeral (i.e. the difference between ‘two’, ‘three’ and ‘four’).

In the beginning, the author reviews various analyses of the variation between the two forms of modifiers as presented in grammars starting from the 19th century, and shows that strict prescriptive rules do not capture the variation in usage. He then proceeds to the situation in 20th century Russian and presents a number of probabilistic patterns which account for alternative choices. These patterns are based on statistics extracted mainly from 20th century Russian prose (although at least one source from the 19th century is mentioned). Effects of the factors investigated may be summarized as follows:

Gender of the noun: nominative modifiers prevail with feminine nouns and genitives are more frequent with masculine and neuter nouns.

Case of the quantified expression: if the phrase is in the accusative case, rather than in the nominative, this favours genitive modifiers.

Number of the predicate: singular predicates are found with genitive adjectives unexceptionally, while with plural predicates both forms are possible. Constructions with nominative modifiers however occur only with plural predicates.

Semantics of the numerals: the genitive is more frequent with higher numerals and vice versa. Thus, with ‘two’ the genitive is less common than with ‘three’, and with ‘three’ is less common than with ‘four’.

Word order: “numeral + modifier + noun” favour the genitive, “modifier +
numeral + noun” favour the nominative. The number of examples found for other word orders is very small.

Animacy of the noun: Suprun argues against the view that this factor plays a role in case assignment on modifiers in quantified expressions. His statistics do not reveal significant differences between animate and inanimate nouns in this respect.

Word class: pronouns (vse ‘all’, èti ‘these’, te ‘those’) take the nominative more frequently than the genitive, owing to the fact that in the majority of instances they precede the numeral. Participles take the nominative as for the most part they occur after the noun (3/4 of instances in the author’s corpus). This indicates that such peculiarities in the behaviour of pronouns and participles reflect the effect of word order rather than of the word class. Substantivized adjectives follow the pattern of adjective modifiers: feminines favour the nominative, masculines and neuters show strong preference for the genitive.

The author observes that individual lexical items have different preferences for one of the cases. Thus,poslednij ‘last’, pervyj ‘first’, drugoj ‘other’, každyj ‘each’ favour the nominative but this may be due to the fact that they normally occupy a position before the numeral (i.e. in the construction “modifier + numeral + noun”). However dobryj ‘good’ and celyj ‘whole’ usually take the genitive even in this construction.

Texts investigated

Literary works mostly between 1920 and 1980. The author does not provide a list of the texts investigated. According to references he makes, prose by the following authors constitute the corpus: A. Tolstoj, K. Paustovskij V. Kaverin, A. Kuprin V. Inber, D. Granin, M. Šaginjan, A. Avdeenko, V. Ketlinskaja, V. Panova, L. Leonov, G. Nikoleva and P. Pavlenko. One text from the 19th century was included as well (F. Dostojevskij Selo Stepančikovo i ego obitateli, 1859).


236 instances illustrating the phenomenon in question were extracted from the corpus and classified according to the conditions mentioned above. Statistics are presented separately for each of the conditions in raw figures and in percentages.

Which data from the source were used

Statistics from the source are presented in the database in their entirety.

Project members

Prof Greville G. Corbett
Dr Matthew Baerman
Dr Dunstan Brown
Dr Alexander Krasovitsky
Dr Alison Long

Period of award:

September 2004 - May 2008


Arts and Humanities Research Council (AHRC) - RG/AN4375/APN18306