Aspects of English anaphora
View/ Open
Date
1979Author
Borsley, Robert D.
Metadata
Abstract
This thesis is concerned with various aspects of English anaphora
and a number of related phenomena. Roughly two thirds is devoted to
nominal anaphora. The remainder considers some of the ways in which
constituents other than HP's enter into anaphoric relations.
The discussion of nominal anaphora begins with a consideration of
two quite widely accepted theories of pronouns and shows that they are
fundamentally inadequate* Evidence is then presented for a 'mixed
theory', which recognises more than one kind of pronoun. The two main
kinds of pronoun that must be recognised are bound variables and
'referential pronouns'. The former are much like bound variables in
logic. The latter are a kind of definite description. In their
anaphoric use, they can be termed 'pronouns of laziness', but their
anaphoric use is not fundamentally different from their non-anaphoric
use. There is evidence that so-called 'sentential pronouns' are
ordinary pronouns of laziness. It appears, however, that what are
termed 'intensional pronouns' are a third kind of pronoun.
The discussion of non-nominal anaphora emphasizes the importance
of definite descriptions in English anaphora. It is argued that so
(in its central use), such, then and there derive from expressions
involving definite descriptions. In its prosententlal use, so appears
to be an idiomatic realization of a sentential pronoun. Certain uses
of so, that and which appear to be idiomatic realizations of and, and
hence only pseudo-anaphora.
Three general conclusions are drawn: firstly that definite
descriptions are central to English anaphora secondly that English
anaphora generally do not derive from copies of their antecedents, and thirdly that, while
NP's enter into amaphoric relations directly, adjectives and adverbs only do so
indirectly through inferences. These conclusions may well apply universally.