Causative alternation licensing in Urdu: An event structure account

Abstract

Given the central role of the verb in clause structure, it is vital to understand the properties of the SEMANTIC ROOT and the EVENT SCHEMA, two constituent aspects of verb meaning, in order to understand how lexical semantic categories relate to syntactic categories. The nature of the interface between these components can, in turn, reveal the overall design of language. However, the main challenge is to make precise the nature of the semantic root and event schema, and their interactive role in argument realization options. To address this challenge, empirical evidence from diverse languages is required to determine how argument realization can be universally accounted for in terms of semantic root and event schema-based lexical semantic representation. The primary purpose of this study is to explicate the roles of semantic root and event schema in Urdu change-of-state (COS) verbs’ causative alternation, formulating licensing conditions on the lexical semantics-syntax interface involved in the phenomenon. On the semantic side of the interface, the argumentation is framed within Rappaport-Hovav and Levin’s (1998a) event structure account, and on the syntactic side, the study assumes Culicover and Jackendoff’s (2005) Simpler Syntax which accounts for an alternation in terms of constraint-based interface principles.

Given that the adequacy of theory is bound up with the reliability of empirical evidence, this study is based on data from multiple sources (lexical translation, Urdu WordNet, Urdu Lughat, individual and dialogical introspection, and speaker survey), conducts extensive analysis of morphosemantic as well as morphosyntactic aspects of 112 Urdu COS verbs, and shows that the causative alternation results from an interaction of multiple licensing factors.

The study reaches the following conclusions: (a) The anticausative form of a COS verb is basic and causative forms are derived. (b) The causative derivation shows gradient and dynamic productivity, and an interaction between lexical schemas and morphological operations, marking the CAUSE relation which reflects causal responsibility between the event participants. (c) An anticausative lexicalizes both manner and result, with a [BECOME<MANNER> [Y <STATE>]] event structure. (d) An anticausative’s event schema and root license only the patient argument; any additional argument is licensed by the root. The cause arguments in causatives are introduced by causative operations, and are obligatorily event schema participants. The syntactic realization of semantic arguments is sensitive to the causal responsibility relation which is reflected in the predicate’s event structure through the primitive predicate CAUSE and its relation with ACT and BECOME. (e) The various aspects of Urdu COS verbs’ causative alternation lead us to the linking rules which show that the argument structure reflects the semantics it inherits from its semantic sources of roots and event schema.

Overall the study shows that the event structure account of Urdu COS verbs’ causative alternation supports the decomposition of the grammar into independent generative components that interact through interface rules. The bottom line is that such a syntax-semantics interface formulation of alternation avoids syntactic complexity.