Figure 8.

Inferential expression of odor–odor associations. (A) Schematic diagram of paired associate training and probe testing. Letters represent odor stimulus items; arrows without question marks indicate trained pairings, whereas arrows with question marks indicate expected transitive and symmetrical choices. Rats are first trained on two overlapping sets of paired associates (Left). Then (Right) they are tested for inferential expression in two ways. In the test for transitivity, they are presented with one of two sample cues from the first training set and the required to select between the choice cues from the second set, based on the shared associates of these items. In the test for symmetry or “reversibility” of the associations, they are presented with one of two choice cues from the second set and required to select the appropriate sample cue from that set. (B) Errors to criterion on acquisition of the two sets of paired associates for sham-operated and hippocampal subjects. (C) Preference indices on the test for transitive inference. For these probe trials a preference score was calculated as (X − Y)/(X + Y), where X and Y were the digging times in the transitive and alternate choices, respectively. (D) Preference indices on the test for symmetrical expression.