Representation of intentions: Persisting activation in memory.

In 4 experiments we investigated dynamic properties of representations of intentions. After Ss had memorized 2 texts describing simple activities, they were instructed that they would have to later execute one of the scripts. On an intervening recognition test, words from the to-be-executed script produced faster latencies than did words from a second to-be-memorized script. This intention-superiority effect was obtained even when (a) selective encoding and poststudy imagery or rehearsal of the to-be-executed script was prohibited and (b) subjects expected a final free-recall test for both scripts. In a control condition in which subjects had to observe someone else executing a script, latencies for words from the to-be-observed script did not differ from neutral words