Ricardo Mairal & Francisco Ruiz de Mendoza
National University of Distance Education, Madrid & University of La Rioja,Spain
Grounding lexical representation in a functional-cognitive theory of sentence meaning
This presentation draws insights from functional models of language (especially, Van Valin’s Role and Reference Grammar or RRG; Van Valin, 2005) and Cognitive Linguistics (especially, Lakoff’s cognitive model theory; Lakoff, 1987; Lakoff & Johnson, 1999; and Goldberg’s constructional approach, Goldberg, 1995, 2005) in order to investigate the way lexical and constructional representations interact. In this regard, we formulate a new framework called the Lexical Constructional Model with a view to determining the relationship between lexical and syntactic meaning (see Figure 1 below) (cf. Mairal and Faber, 2005; Ruiz de Mendoza and Mairal, 2006a, b). The initial claim is that a theory of semantic interpretation should be constructed on the basis of two representational mechanisms, i.e. a lexical and a constructional template, and a set of cognitive constraints that regulate the unification between the two constructs. This presentation will focus on the format of lexical and constructional templates, while the type of cognitive constraints will be dealt with in another presentation.
RRG uses a decompositional system (which combines insights from Vendler’s Aktionsart and Dowty’s decompositional system) for representing the semantic and argument structure of verbs and other predicates (their Logical Structure, LS). The LS inventory has a number of shortcomings: (i) there is no way to account for arguments that are not strictly derivable from the meaning of a predicate; (ii) the nature of the primitives involved is unclear and inconsistent; (iii) there is no place in the account for linguistic expressions based on metaphor, metonymy, and image schemas; (iv) some important meaning generalizations pertaining to the cognitive grounding of grammatical structure are missing. Some of these shortcomings have been sorted out by the Lexical Constructional Model, which proposes to substitute the notion of lexical template for logical structures. Lexical templates are enriched semantic representations of the logical structures proposed in RRG, and combine low-level and high-level semantic components (see figure 1 below): high-level elements (e.g. know) are shared by items belonging to a number of lexical classes (e.g. some of the verbs of the lexical domain of Cognition); low-level elements are specific to (and therefore definitional of) the item in question (e.g. the lexical function MAGNOBSTR, which refers to the difficulty involved in carrying out an action, as is the case with fathom, which involves great difficulty, a low-level semantic parameter). One of the advantages of positing lexical templates is that the metalanguage they encode is very similar and methodologically closer to constructional templates as envisaged in Goldberg’s constructional approach (e.g. Goldberg, 2005). Constructional templates, on the other hand, are based on combinations of high-level idealized cognitive models, including high-level propositional models (primitive and/or primary frame-like structures), grammatical metaphor, grammatical metonymy, and image schemas. We claim that constructional templates “coerce” lexical templates as a consequence of a more general cognitive principle whereby higher-level structures invariably take in lower-level structures. Thus, when a low-level frame and an image schema interact, the frame elements become part of the image schema. Lexical templates are in fact lower-level (or lexical) constructions that can be fused into higher-level characterizations such as the caused-motion, the resultative, or the benefactive constructions (see figure 2 below for a simplified representation of lexical-constructional subsumption in She loved me back into life). Since the formal apparatus of lexical templates shares with higher-level constructions all elements excepting those that are specific to a lower-level class, absorption of a lexical template by a construction becomes a straightforward, redundancy-free process. This kind of formulation captures relevant features that lexical template representations share with constructional representations, which makes our description fully at home with the idea of a lexical-constructional continuum.
References
Goldberg, A.1995 A Construction Grammar Approach to Argument Structure, Chicago: University of Chicago Press.
Goldberg, A. 2005. Constructions at work: the nature of generalization in language. Oxford: Oxford University Press.
Lakoff, G. & Johnson, M. 1999. Philosophy in the Flesh. New York: Basic Books.
Lakoff, G. 1987. Women, Fire, and Dangerous Things: What Categories Reveal About the Mind. Chicago: University of Chicago Press.
Mairal, R. and P. Faber 2005. ‘Decomposing semantic decomposition’. Proceedings of the 2005 International Conference on Role and Reference Grammar. Academia Sinica: Taiwan, pp. 279-308.
Michaelis, L. 2003. “Word meaning, sentence meaning, and syntactic meaning”, in Cuykens, H., Sirven, R. & Taylor, J. R. (eds): Cognitive Approaches to Lexical Semantics. Berlin: Mouton de Gruyter; 93-122.
Ruiz de Mendoza, F. & Mairal, R. 2006a. “Lexical representation and constructions: bridging the gap between the constructional and process models of grammar”. Universidad Complutense de Madrid; forthcoming.
Ruiz de Mendoza, F & Mairal, R. 2006b. High-level metaphor and metonymy in meaning construction. In Radden et al. (eds.) Constructing Meaning. John Benjamins: in preparation.
Van Valin, R. D. Jr. 2005. The Syntax-Semantics-Pragmatics Interface: An Introduction to Role and Reference Grammar. Cambridge: Cambridge University Press.
