220 María de los Ángeles Gómez-González
basis of this relationship, an expression evokes a certain body of concep-
tual content, its ground or base. This directs attention to some particular
substructure – the profile – construed as the relationship (subsuming the
entity or entities) that the expression designates or refers to. By way of
illustration, (1a) and (1b) above profile an abstract presentational
relationship, (1b), (2c), and (2d) one of spatial attribution, and (2a) one of
‘existence’ (see Sections 2.3 and 2.4 for further reference to these alternate
relationships).
A D-Topic therefore refers schematically to the relationship (or thing)
profiled. It has a prospective, or forward-looking potential by virtue of sig-
nalling that the profiled relationship or entity will function as a conceptual
reference point for purposes of interpreting a subsequent proposition or es-
tablishing mental contact with another reference point (its target) within –
as is most likely – or outside its cognitive schema, or dominion, that is to
say:
... any system of concepts related in such a way that to understand any of
them you have to understand the whole structure in which it fits; when one
of the things in such a structure is introduced into a text, or into a conversa-
tion, all of the others are automatically made available. (Fillmore 1982:
111)In addition, D-Topics have a retrospective or backward-looking dimen-
sion, in the sense that they normally refer to information that is already
accessible in the discourse co(n)text. This is implied by the very notion of
building a coherent or cohesive structure for, when structure is added, it is
usually anchored to what has already been built, by virtue of the conceptual
overlap. It follows that – as implied in FG – D-Topics tend to be endowed
with the features of identifiability and activation. Identifiability accounts
for the difference between referents for which the speaker assumes a file
has already been opened in the discourse register and those for which such
a file does not yet exist (Chafe 1976). What counts for the linguistic ex-
pression of this cognitive distinction is whether or not the speaker is able to
pick out a referent from among all those which can be designated by a par-
ticular linguistic expression and identify it as the one which the speaker has
in mind, provided that, given the appropriate discourse context, a referent
is more or less permanently stored in the long-term memory of S and A,
and can be retrieved without difficulty at any particular time. A referent
may be identifiable either because it has been mentioned in a discourse (i.e.
through anaphoric reference) or because it is either visible or otherwise sa-