2 Variation in Working Memory Capacity

If, as many psychologists seem to believe, im­ mediate memory represents a distinct system or set of processes from long-term memory (L TM), then what might· it be for? This fundamental, functional question was surprisingly unanswer­ able in the 1970s, given the volume of research that had explored short-term memory (STM), and given the ostensible role that STM was thought to play in cognitive control (Atkinson & Shiffrin, 1971 ). Indeed, failed attempts to link STM to complex cognitive· functions, such as reading comprehension, loomed large in Crow­ der's (1982) obituary for the concept. Baddeley and Hitch ( 197 4) tried to validate immediate memory's functions by testing sub­ jects in reasoning, comprehension, and list­ learning tasks at the same time their memory was occupied by irrelevant material. Generally, small memory loads (i.e., three or fewer items) were retained with virtually no effect on the primary tasks, whereas memory loads of six items consistently impaired reasoning, compre­ hension, and learning. Baddeley and Hitch therefore argued that "working memory" (WM)