A SYNOPSIS OF ROLE AND REFERENCE GRAMMAR 153- There is also a right-detached position, as in sentences like / know them, those
boys. It will not be discussed further. See Lambrecht (1981) for discussion of this
construction in French. - See FVV, chapter 5, for detailed discussion and justification of these distinctions.
Directionals are given with either the core or nucleus as the unit they modify
because of cross-linguistic variation; in German and Jacaltec, for example, they
are core modifiers, i.e., they express the direction or orientation of one of the
core arguments, while in Kewa and Yimas (Papua-New Guinea) they are nuclear
modifiers, i.e. they express the orientation or direction of the action itself and not
that of any of the participants. Negation may also be a nuclear operator, but this
is relatively rare. - The operator projection is not represented. The issue of the placement of com
plementizers in the LSC will be discussed in section 6. - The terminology is taken from Bresnan (1982).
- See Jolly (1987), this volume, for a detailed presentation of the RRG analysis of
prepositions in English. - See Nunes (1990), this volume, for detailed discussion.
- See Rijkhoff (1988) for a discussion of NP operators from the perspective of Func
tional Grammar. - See Reinhart (1981) for a formal definition of pragmatic "aboutness".
- There are several other pragmatic theories which could also serve this function,
e.g. the Atlas & Levinson (1981) principle of informativeness or the Sperber &
Wilson (1986) theory of relevance. - See Lambrecht (1986, in preparation) for detailed discussion of the syntax,
semantics and pragmatics of these constructions. - A couple of points need to be made about these representations. First, si is
treated as a derivational morpheme signalling a marked subject assignment; cf.
Van Valin (1990b), Centineo (1986). Second, the pronoun mi in (12b) is a topic
expression, but this is not in conflict with the analysis of this construction as sen
tence focus. Lambrecht (1988), commenting on this example, states "it is not the
absence of any topic expression that defines a sentence focus construction but the
absence of a topic-comment relation between a lexical subject NP (or an NP that
is a subject at the underlying propositional level) and the proposition expressed by
the sentence. It is important to understand that the formal contrast between pred
icate focus and sentence focus crucially involves the category "subject NP"."(14) - I am grateful to Martin Braine for pointing out the relevance of imperatives to this
issue. - In the table, "d.n.a." means that the test does not apply to verbs of this (sub)class.