Access to prior spatial information

In six experiments, reading times and probe naming times were measured in order to examine the conditions under which spatial information became accessible and/or reactivated. In Experiments 1-4, reading times were measured for target sentences containing spatial inconsistencies. Spatial inconsistencies did not disrupt processing (Experiment 1) unless there were increases in task demands (Experiment 2), elaboration of the protagonist's location (Experiment 3), or both (Experiment 4). In Experiments 5 and 6, naming times were measured to directly assess the activation of spatial information, specifically objects associated with a protagonist. Spatial information was highly active in memory immediately after being read and less active after four intervening sentences (Experiment 5), but explicit cues (e.g., location or object) as well as references to the current situation model were effective in reactivating previously mentioned spatial information (Experiment 6). The combined results of six experiments are discussed within the context of the RI-Val model.