Ekaterina Rakhilina & Olga Lyashevskaya - Categorial structure and construction grammar: containers

Container can be said to represent one of the key notions in the cognitive linguistic paradigm (cf. the work by Herskovits, Vandeloise, Bruce Hawkins et al.). It is one of the basic “topological categories” (L. Talmy) with an undoubted cross-linguistic relevance. These categories are commonly represented as having a radial structure, which presupposes, on the one hand, a clearly defined categorical center (as boxes or bowls), and, on the other hand, a periphery, assembling various names of objects (like hats or spoons) which do not readily fit into what a prototypical container is supposed to be. This view of container as a category is much in demand, whenever particular linguistic problems arise: in each case, properties of both prototypical and non-prototypical representatives of containers are compared and contrasted. Cf. also studies of basically non-cognitive orientation1, with the same statement of problem. Still, it is not a priori clear, whether containers can be represented as an integral notional category with a common structure – so that it would display the same type of behaviour in different linguistic conditions.   Our investigation2 has shown that at least in the domain of containers what we are dealing with is rather an opposite situation: it is not a notional category which determines the meaning of a given construction, but, on the contrary, each construction sets its own way of how the world’s properties are to be interpreted and, so to say, builds its own categorization.   The paper examines four Russian constructions which are semantically related to the type of containers: attributive construction with the adjective glubokij ‘deep’ (deep X), comitative construction of the type ‘X with Y’, locative construction of the type ‘X in Y’, and genitive construction of the type ‘X of Y’. All these constructions involve nouns with a container-like topology, cf. examples such as deep bowl, bowl with porridge, porridge in the bowl, a bowl of porridge. A more detailed analysis, however, shows that each construction prefers its own classes of containers – and that this choice is motivated by the semantics of the construction. This is entirely in keeping with the theory of Construction Grammar as suggested in Fillmore’s and Goldberg’s work. The data in Table 1 can serve as illustration to this.

[1] As Apresjan Ju. Semantic motivation of non-semanticproperties of lexemes. In: B. To­šo­vi (ed.), Die grammatischen Korrelationen. Graz:I nstitut für Slawistik, 96-116; or Borschev V. & Partee B. Semantics of genitive construction : different approaches to formalization. In: E. Rakhilina & Ja. Testelets (eds.). Typology and the theory of language. Moscow, 1999, 159-172.

2 Cf. also:  Rakhilina, E. Container and content in Russian: a naive topology. In: Linguistic meanings. Moscow, 2004, 233-257.