PART TWO

Part Two applies some of the conclusions reached in Part One to some descriptive topics in Linguistics. These topics are all in the areas of Anaphora and Ellipsis (broadly defined).

One of the main aims of Part Two (and of Chapter Six in particular) is to explore the extent to which it might be productive to look upon Ellipsis and Anaphora as a single phenomenon, or family of phenomena. In fact, I propose to use the term Anaphora (from time to time) in a sense broad enough to include Ellipsis (of all kinds):

Anaphora ... (consists) of the avoidance of redundancy or repetition by the use of a semantically (and, optionally, lexically and phonologically) attenuated expression in place of the full, lexical expression initially used (Cornish 1986,1).

From a Cognitive point of view, what Anaphora and Ellipsis (A & E) have in common is that they both crucially involve memory, which is the faculty of mind which allows the link to be created between antecedent and (null or non-null) anaphor. From this perspective, it seems only natural that the antecedent should precede the anaphor in the majority of cases, as the speaker/writer would want to avoid unnecessary doubt in the mind of the hearer/reader as to who or what he was referring to.

Thus one principal division within A & E is between the majority of cases, where antecedent precedes anaphor, and the minority of converse cases, i.e. Cataphora (Buehler 1934).

A second main division is between those cases where the link between anaphor and antecedent is subject to the Semantic Identity Criterion, and those where it is subject to my proposed Phonological Identity Criterion.

A third main division is between Anaphora and Ellipsis (where the "Anaphor" is null).

My final division is between the various types of relationships that can exist between mind, anaphors, and antecedents.

All these divisions cut across each other. One chapter is devoted to each of these divisions.

Cornish (op.cit.) is, apart from Hirst (1981), one of the few works I know of that comprehensively surveys and classifies the various types types of Anaphora.


There are a number of different ways of classifying anaphoric relations in natural languages: one may draw a broad distinction between lexical and non-lexical anaphora, between nominal and non-nominal anaphora, between referential and non­referential anaphora,, or between strict anaphora (where 'antecedent' - here, the controller - and anaphor are involved in an essentially grammatical relation) and discourse anaphora (where antecedent-trigger and anaphor are subject to much less stringent restrictions). Basically, it is the properties of the anaphor, together with the discourse as well as grammatical status of the segments containing the anaphor and the antecedent­trigger (where there is one), which determine the nature of the anaphoric relation established.

In somewhat broader terms, however, anaphoric relations may be defined in terms of the syntactic status of the antecedent-trigger and its anaphor; in terms of the semantic or pragmatic relation in terms of which anaphora is established between the two expressions; or in terms of some other (e.g. phonetic) parameter by means of which an identity relation may be established between the two expressions (Cornish, op.cit., 7-8).


Taking deixis into account, Cornish takes the position on Anaphora that I myself adopt in Chapter Four -- namely, that (as only the anaphor, but not the antecedent, needs necessarily be present in the text/discourse) what the anaphor actually refers back to is a discourse referent within a discourse model. How that discourse referent became manifest to the parties to the discourse (deictically or linguistically) is a separate issue.

Chapter Three addresses two of these ways of classifying Anaphora, viz. the referential/non-referential, and nominal/non-nominal distinctions.

The broader syntactic, semantic, and pragmatic factors involved in the relationship between antecedent and anaphor are discussed in Chapter Four.

The putative distinction between "strict" and discourse Anaphora is discussed in Chapter Five.

The distinction between lexical and non-lexical Anaphora is discussed in Chapter Six.

Back to Contents Page