Time and memory.

Standard animal memory tasks require judgments of event recency: Delayed matching to sample (DMTS) requires that the animal identify the stimulus seen most recently; radial-maze-type (RM) tasks require that the animal identify the place visited least recently. Delayed-reaction tasks are intermediate. I argue that time discrimination (temporal control) and event memory call on the same processes: Proactive and retroactive effects occur in both, brief events have less effect than protracted events, and increases in event duration have smaller and smaller effects. If the "ages" of past events are represented by animals in a way consistent with Weber's and Jost's laws, and if there is a limit to the number of different recencies that can be discriminated, then the major differences between these three types of memory task can be explained. DMTS performance is poor because the animal must discriminate between two sets of recencies (memory arrays) that differ only in respect of the most recent event; RM performance is good because the recencies of places visited on the current versus earlier trials are always clearly discriminable.