|
COPYRIGHT 2004 Spanish Association for Anglo-American Studies (AEDEAN)
This paper accounts for the semantic features of induced motion verbs. Following the approach of Componential Analysis, and inscribed within the framework of Role and Reference Grammar, these verbs are arranged into two groups, according to a number of semantic variables related to their argument structures and to the types of locative arguments they govern. New logical structures that account for this distinction are proposed. This typology is in line with a functional view of language, and it highlights the fact that meaning is reflected on the grammatical structure of clauses.
Key words: active, accomplishment, motion, predicate, argument, argument-adjunct, adjunct, goal, source, path.
1. Introduction
Verbs of movement constitute an interesting source of semantic analysis, since they are attached to the spatial dimension, which is, together with the temporal dimension, the basis of cognition. (1) One important kind is those of induced movement. In this work, a typology is proposed in terms of their semantic structure and of the types of locative expressions they take. This work is inscribed within the framework of functional grammars, more specifically within Role and Reference Grammar--hereafter RRG--(Jolly 1991; Van Valin 1993; Van Valin and LaPolla 1997) and Functional Grammar (Hengeveld 1992, 1997; Dik 1997a, b; Samuelsdorff 1998; Mackenzie 2001). The verbs under analysis are classified according to their Aktionsart. The notion of Aktionsart is adopted from Vendler (1967 [1957]), and it is used as a basic criterion to identify argument structure and predicate relations. Besides, only transitive verbs that imply an induced motion of the undergoer have been included in the corpus of analysis. Thus, verbs such as push, which may be transitive or intransitive, have been discarded.
This study is also based on Componential Analysis (Pinker 1989; Saeed 2003) according to which lexical decomposition is used as a basic device to describe the meaning components of words. Such components allow verbs to be organized in groups and enable their argument structure to be dealt with. Thus, the view is held that the different semantic classes of verbs reflect different syntactic and semantic argument structures. This explains the direct relation of verbal predicates with the kind of prepositional phrases they take. On this point, we refer the reader to Gropen, et al. (1991), Levin (1993), Grimshaw (1994) and Levin and Rappaport (1995), on the interaction between syntax and semantics.
2. Review of the literature and research hypothesis
The verbal phrase is a universal category. It is the nucleus of the clause, and the other elements depend on its meaning components and on its syntactic behavior. Thus, a central issue in linguistic research is the linking between semantics and syntax. This can be approached either from a semantic or from a syntactic perspective. The semantic approach defends the view that an analysis of the semantic properties of verbs reflects their grammatical behavior. This is followed by the Lexico-Grammar Model--hereafter LGM--(Faber y Mairal 1999; Mairal 2001; Mairal y Cortés 2000-01), and it is taken as a starting point here. The syntactic approach, on the other hand, assumes that those verbs that share a number of syntactic properties belong to the same lexical class. Thus, lexical classes are grouped according to their syntactic behavior. This idea is followed by Levin (1993), Levin and Rappaport (1995, 1999) and Demonte (2002), among others. Both approaches are inscribed, in a very general way, within the functional paradigm.
Within the cognitive paradigm, Jackendoff (1983, 1990, 1992) has developed what he calls conceptual semantics, based on decomposing meaning through the description of mental representations. His attempts at generalization through such semantic decomposition are also taken into account here, especially in relation to his definition of the localist hypothesis. Nevertheless, with the exception of Talmy (1975, 1983, 1985), who studies the semantic components of motion verbs at the syntax-semantics interface and analyzes how they affect the other elements in the clause, few studies have been made within the functional paradigm about this topic. Talmy, however, does not pay attention to causative verbs and their correlation with spatial constructions. Accordingly, we analyze this and extract the semantic factors of verbs that exert an influence on locative constructions. For our explanations, we have selected RRG's system of semantic representation, where the logical structure used for induced motion verbs is the following:
(1) [do' (x, [conjunto vacío]] CAUSE [BECOME be-LOC' (z, y)]
We have analyzed the semantic components of 6,500 verb samples, extracted from the British National Corpus. According to the results, induced motion verbs are organized in two groups: causative accomplishment movement verbs, such as place or remove, and causative active accomplishment movement verbs, such as carry or take. Important differences can be found between both. In this sense, the latter group has been ignored in RRG until now. Therefore, demonstrates that such verbs should be recognized as an independent mode of action (Aktionsart), since the logical structure given in (1) qualifies as insufficient. As a consequence, a different one is proposed in section 4, which serves to distinguish these verbs from the former. The need for two different semantic representations becomes more evident when we focus on Aktionsart: active accomplishments are derived from an activity predicate, while non-active accomplishments come from a state predicate. In the following section, we account for (non-active) causative accomplishments, and we establish a typology for them. After that, in section 4 we turn to the analysis of causative active accomplishments. Finally, in section 5 we summarize the data presented and draw some conclusions in the light of the results obtained.
3. Causative accomplishment verbs of motion
Causative accomplishment verbs are the result of a process of change. In this case, since we focus on motion verbs, it is a change of location. These verbs are telic, and their basic predicate is a state. This factor determines their logical structure and the kind of locative arguments they take. Thus, they do...
Read the full article for free courtesy of your local library.
|