|Title:||The Encoding of Complex Concepts in Hindi and English||Add Dissertation|
|Author:||Bhuvana Narasinham||Update Dissertation|
|Email:||click here to access email|
|Institution:||Boston University, Linguistics Department|
|Abstract:||As originally pointed out by Talmy (1985), Germanic languages have constructions which encode two events within a minimal clause: 'The boy crawled to the door' encodes the event of directed motion as well as the causal/manner event of crawling which precedes or accompanies the motion event. However, other languages (e.g. Spanish, French, and Hindi) cannot combine a verb ('crawl') with a goal phrase ('to the door') to conflate two events in a minimal clause, rendering ungrammatical 'The boy crawled to the door.' The causal/manner event must be expressed separately as an adjunct, e.g. 'The boy went to the door, crawling.' This typological differences poses problems for any theory of grammar which presupposes that, crosslinguistically, verbs with similar lexical semantics will project their arguments onto similar syntactic configurations.
A number of analyses have been proposed to explain this phenomenon. I investigate additional complex constructions in the domains of motion, change of state ('He cooked the stove black'), creation ('She kicked a hole in the wall'), and communication ('She smiled her gratitude') that reflect similar differences between Hindi and English. I show that previous analyses suffer from empirical inadequacies and moreover cannot explain these additional phenomena. I propose that the encoding of complex constructions in languages like English is not mediated by the verb at all: rather, complex constructions are inserted as a fixed Verb Phrase unit in the syntax. The VP-constructions are listed in the English 'phrasicon' and compose the meanings of the verb and its constructionally licensed complements, producing a complex event interpretation. Hndi does not allow this 'extralexical' meaning composition with phonologically unspecified VP templates of the type described here for English. The 'Extralexical Hypothesis' also makes testable empirical predictions:
(1) The extralexical mapping strategy is a marked option for certain languages--this predicts that the unmarked option of the lexical mapping strategy should be available for all languages which allow extralexical mapping; (2) An implicational hierarchy of verbs determines which verb types are accessible for the phrasal encoding of complex events across languages; (3) Other constructions that require an extralexical analysis are predicted to be absent in Hindi. These predictions are partially confirmed and implications are discussed.