One crucial characteristic of DGS and other sign languages is that the referent associated with the locus of the signer can be shifted, that is, it can be associated with a referent other than the signer. This possibility implies that the rest of the referential framework is shifted in relation to this locus (Morgan & Woll 2003: 303). The notions of referential framework or frame of reference are used to designate a set of referential loci used in a particular discourse situation (cf. Lillo-Martin & Klima 1990: 193; Bellugi et al 1990: 18). Two types of referential frameworks are distinguished in the literature, namely, (a) the fixed referential framework (FRF) and (b) the shifted referential framework (SRF) (Morgan & Woll 2003: 303; Morgan 2006).
Fixed referential framework. In the FRF (cf. (57a)), the sign space in front of the signer serves to
- - pick up referential loci previously established;
- - mark subject and object agreement through the movement between the relevant referential loci;
- - point to referential loci for pronominal reference;
- - position and move classifiers.
Shifted referential framework. In the SRF (cf. (57b)) the sign space incorporates the signer’s own body. SRF is used to
- - encode morphosyntactic information through location on the signer’s own body and locations related to the orientation of the signer’s body;
- - indicate agreement with subject and object through the movement away or toward the signer’s body.
Referential shift, as will become apparent in the following sections, is a grammatically and lexically constrained phenomenon that can serve various pragmatic functions. It is thus a phenomenon that involves the syntax-discourse interface.