Deixis, gesture, and cognition in spatial Frame of Reference typology

The three Frames of Reference recognized in the current inventory of spatial-language types are differentiated by their placement of the Anchor from which the vector of search space from Ground to Figure is calculated (Levinson 1996). In certain well-recognized examples, Anchor merges with Ground. The existing analysis treats this merged component as analytically Ground rather than Anchor; its location in or out of the speech situation is therefore taken to be independent of the Frame of Reference typology. Instead, I treat this component as analytically Anchor, making its speech-situation status criterial to the typology. Four, not three, Frames of Reference now appear. The fourth, “Direct” frame, distinguishes binary locutions with a speech participant as Ground/Anchor (e.g. ‘in front of you’) from “Object-Centered” binary locutions in which Ground/Anchor is not a speech participant ( e.g. ‘in front of the kettle’). This four-frame analysis corresponds better than does the three-frame one to the logic of rotation sensitivity which has been used to show Whorfian parallels between language and conceptualization across cultures. I close by discussing the application of Frame of Reference typology to pointing gestures, and show how recognition of the fourth frame of reference allows us to bring discussion of these, and of the linguistic demonstratives and locatives with which they so frequently co-occur, fully within the Frame of Reference typology.