On the Function of the Infinitive and the Expression of the Grammatical Subject in [Pacc/Pdat] + Finite Verb + Infinitive Constructions (With Evidence from Russian)

Dagmar Divjak, Katholieke Universiteit Leuven

0. In this this paper I deal with the problem of the syntactic role of the infinitive in Russian in constructions consisting of [case-paradigm]+third person singular finite verb+infinitive. Here I introduce the Pronominal Approach, (PA), a linguistic method, situated in the field of tension between theoretical and descriptive linguistics.

I. The PA is a constructivist method for the description of syntax, which was drafted by K. van den Eynde and further developed by different teams in Belgium (Leuven) for French, Dutch, Chinese, Russian and Japanese, in France (Aix-en-Provence) for French and in Denmark (Copenhagen and Odense) for Danish (see References).

The PA provides linguists with an economic tool for analyzing syntactic data in a formal way by focusing on the pronominal entities, that are selected as valency elements by a predicator or construction kernel, instead of on the lexicalized items. Since the number of the minimal referentials of a language—pronouns and other proforms—is limited, they are highly suited for an exhaustive analysis.

The PA calls on the constant relation that exists between a pronominal construction and the lexicalized construction to which it refers in order to examine syntactic constructions without explicit semantic paraphrase. The pronouns, grouped into paradigms, cover the basic syntactico–semantic skeleton of a language in a language immanent way. This relation of proportionality to pronominal paradigms reveals dependency as well as substitution possibilities; this helps to recognize the constituents of a construction, to define their syntactico-semantic status and to examine the implicational relations between constructions. Consistent sense distinctions are based on these syntactico-semantic features (see: http://www.cbs.dk/departments/dl/ska/ovd2.htm).

II. Traditionally, the above-mentioned constructions are classed under impersonal one-compound sentences. However, the Russian Grammar (AN SSSR) endorsed the viewpoint of these constructions being impersonal two-compound sentences, giving rise to a discussion on whether the infinitive can or cannot accomplish the function of grammatical subject.

Two problems have to be solved in a constructivist way in order to answer the question of the syntactic role of the infinitive within the framework of the PA: the characteristics of grammatical subjects have to be distinguished and defined and a test has to be developed by which the different syntactic functions of the infinitive can be distinguished from one another.

The nominative paradigm (Pnom) is characterized by a set of suspensive pronouns kto ‘who’ or čto ‘what’, by their assertive counterparts and all possible instantiations (see Soldatjenkova, On the Valency of Russian Verbs: A Constructivist Approach, Leuven: Peeters, to appear). This Pnom fulfills the function of grammatical subject. This fact is the base for dividing the studied constructions in those, which reserve a place for the Pnom and those, which don’t.

I will argue that in Pnom-constructions infinitives can fill up the Pnom slot as long as the pro-infinitive form čto (s)delat′ ‘to do what’/èto (s)delat′ ‘to do this’ is proportional to both čto ‘what’ and èto ‘that’. The neuter/third person singular ending of the predicator is triggered by the grammatical subject, an infinitive,which is not specified as to gender or number. For example,

Rebenkunadoelaèta kniga
The kid is bored with this book.
komu ‘to whom’čto ‘what’
emu ‘to him’èto ‘this’

The kid is bored with reading
komu ‘to whom’čto ‘what’
čto delat′ ‘to do what’
emu ‘to him’èto ‘this’
èto delat′ ‘to do this’

Apart from the proportional Pnom–constructions there are others, in which only the proform čto (s)delat′ ‘to do what’/èto (s)delat′ ‘to do this’ maintains the relation of proportionality with the infinitive. I will argue that the roots of this divergence lie in the nature of the finite (third person singular) verb: the fact that an infinitive is only proportional to čto (s)delat′ ‘to do what’/èto (s)delat′ ‘to do this’ indicates that the finite verb is not a predicator but a modificator, with no valency scheme of its own. For example,

Namponadobitsjavaša pomošč′
We will need your help.
komu ‘to whom’čto ‘what’
emu ‘to him’èto ‘this’

Namponadobitsjapokazat′sja vraču
We will have to go and see a doctor
komu ‘to whom’*čto ‘what’
čto delat′ ‘to do what’
emu ‘to him’*èto ‘this’
èto delat′ ‘to do this’

Each finite verb can be assigned to one of the categories of modificators, that were established after research on the approximately four hundred verbs that can be combined with an infinitive . In order to do this the verbs have to be tested on their ability to add different rectional (verb general) elements of time and/or location to finite and infinite verb. Turning the construction into a subordinate (explicative) one with čto/ čtoby ‘that’ can provide useful subcategorizations.

An important characteristic of the non-proportional constructions is the absence of a position for the Pnom. They share this peculiarity with constructions, built around finite verbs that never select a Pnom position, not even when having predicator status. The neuter/third person singular is triggered by the lack of a Pnom that could initiate agreement. For example,

The patient wants a bit of water.
komu ‘to whom’čego ‘of what’
emu ‘to him’ètogo ‘of this’

Starikuxočetsjavernut′sja na rodinu.
The old man wants to return to his home country.
komu ‘to whom’*čego ‘of what’
čto delat′ ‘to do what’
emu ‘to him’*ètogo ‘of this’
èto delat′ ‘to do this’

It is interesting to study the role of the Pdat in Pnom-less sentences containing an infinitive-predicator. Different functions, varying from semantic or logical to oblique-case grammatical subject, have been assigned to this Pdat. If we take into account the valency schemes and formulas of these infinitives-predicators, it becomes possible to define the Pdat formally as a discontinuous grammatical subject or as fulfilling the function of a grammatical subject without having a one to one relationship with a morphological entity, nor occupying the usual paradigm.


Eynde van den, K. and C. Blanche-Benveniste. 1978. “Syntaxe et mécanismes descriptifs: présentation de l’approche pronominal.” Cahiers de Lexicologie vol. 32, 3–27.

Blanche–Benveniste C. Eynde van den K. e.a. J. 1987. Pronom et syntaxe: L’approche pronominal et son application au français. Paris.

Melis, L. and C. Eggermont (eds.) 1994. “A Pronominal Approach to Valency Dictionaries.” International Journal of Lexicography vol. 7/2 (summer), special issue.