Is the hat on the table? Language and Spatial Relations in and out of Context

Language and space are closely inter-related as they reflect, and are constrained by, fundamen- tal properties of the human conceptual system , and also because of the mapping of conceptual structure and spatial representation into language (Jacken- doff, 1996). In addition, they can be instrumental in sheding light on each other, as several study in memory and description of places have shown (e.g. Linde & Labov, 1975; Ullmer-Ehrich, 1982; Mainardi Peron, Baroni, Job, & Salmaso, 1990). Numerous analysis and empirical studies have shown that in the comprehension of spatial prepo- sitions factors related to the functions and contex- tually-defined roles of the elements, as well as to the linguistic meaning (Bennett, 1975), affect par- ticipants’ performance (e.g. Garrod and Sanford, 1989; Coventry, Carmichael, & Garrod, 1994; Cars- lon-Radvansky and Radvansky, 1996). In the present study, we investigated the com- prehension of spatial preposition using a sentence/ picture verification task with the aim of elucidating the role of real-world knowledge in the processing of sentences. On the assumption that the compre- hension of expressions indicating spatial relations among elements relies crucially on contextual infor- mation we manipulated the type of pictorial informa- tion participants were presented with factors. On half of the trials, we provided an environment that could act as a frame for the interpretation of the spatial relation referred to by the verbal expression. On the remaining trial, the geometrical information remai- ned the same but no environmental information was provided. The predictions were the following: If environ- mental information is processed as fast as, or fas- ter than, geometrical information, a facilitation effect may be detected for the conditions in which a frame is provided. However, when environmental informa- tion and geometrical information mismatch, then providing a frame should hinder processing.