To download click here
Lexical semantics and syntax: commercial transactions
reanalyzed
last changed 3 July 2008
Bibliographical information
This draft was written in June 2008, and has been submitted to Language. It is a fundamentally revised and re-oriented rewrite of 'Commercial transactions revisited', which in turn was based on 'Buying and selling in Word Grammar'.
Abstract
The paper addresses the general theoretical question of how arbitrary syntax is, and more
specifically, to what extent a word’s syntactic valency may be predicted from its
meaning. It argues that syntax is at least largely predictable given an analysis of meaning
which pays due attention to the full cognitive structure which ‘frames’ each word’s
meaning. The theoretical background includes Fillmore’s ‘frame semantics’, but the
argument is based on a new analysis of the so-called ‘commercial transaction’ verbs
BUY, SELL, PAY, CHARGE, SPEND and COST. In 1976, Fillmore claimed that these
six verbs shares a single frame, in spite of the considerable syntactic differences among
them such as their choice of prepositions and the options for passivization and omission.
After re-analysis, it turns out that all of these syntactic differences can be explained as the
regular syntactic pattern of a semantically coherent group of verbs to which the verb in
question belongs. The methodological point of the paper is that semantic analysis should
be guided by syntactic details as well as by careful structural analysis of the meaning.
The main theoretical point is that the analysis requires a very flexible network analysis in
which relations, both semantic and syntactic, can be created freely, rather than one which
assumes a small and fixed universal set of relations; a subsidiary point is the benefit of
being able to assume multiple default inheritance. |