Attention in Spatial Language: Bridging Geometry and Function

This chapter argues that spatial language is determined in part by attentional deployment over objects. This attentional deployment may be driven by both geometry and habitual object function. Thus, geometric and functional influences in spatial language, while separate in origin, may act through the same psychological mechanism: attention. This argument is made concrete through a computational model of projective spatial terms — the Attentional Vector Sum (AVS) model. The chapter begins by reviewing the success of this model in accounting for some purely geometric aspects of spatial...

This chapter argues that spatial language is determined in part by attentional deployment over objects. This attentional deployment may be driven by both geometry and habitual object function. Thus, geometric and functional influences in spatial language, while separate in origin, may act through the same psychological mechanism: attention. This argument is made concrete through a computational model of projective spatial terms — the Attentional Vector Sum (AVS) model. The chapter begins by reviewing the success of this model in accounting for some purely geometric aspects of spatial language, in terms of attention. It then presents a natural extension to the model, one that allows an influence of object function on attentional deployment, and thus on predicted spatial language use. It shows that this extended version of the model can account for data that indicate an influence of object function on spatial language. These results suggest that attention may serve as a unifying level of description in spatial language, bridging geometry and function.