Memory Conjunction Errors in Normal and Amnesic Subjects

Disproportionate deficit in associative recognition relative to item recognition in global amnesia

“…Therefore, successful discrimination between old and conjunction stimuli cannot be based on memory for individual stimulus elements, but must be based on memory for the associations formed between individual elements at study. The results from these studies demonstrate that amnesic patients are disproportionately impaired at rejecting conjunction stimuli (Kroll, Knight, Metcalfe, Wolf, & Tulving, 1996;Reinitz, Verfaellie, & Milberg, 1996; but see Stark & Squire, 2003). Although these findings provide support for the notion that amnesia reflects an impairment of memory for associative information, the results do not provide direct evidence of a differential deficit in memory for complex associations (as opposed to simple associations), because no measure of memory for simple associations (i.e., the individual elements of which the stimuli were composed) was obtained.…”

False memories for compound words: Role of working memory

“…This binding process has been the topic of much recent research and has been proposed to be a central function of the hippocampus (see, e.g., Chalfonte & Johnson, 1996;Cohen & Eichenbaum, 1993;Kroll, Knight, Metcalfe, Wolf, & Tulving, 1996;Reinitz, Verfaellie, & Milberg, 1996). More specifically, these authors have proposed that the function of the hippocampus is to encode episodic information by creating new associations between preexisting units stored in memory, rather than to encode novel units into memory.…”

“…ing recently studied STARGAZE and CATFISH; Kroll et al, 1996, Reinitz et al, 1996. In addition, there is substantial evidence that working memory processing mediated by prefrontal cortex plays a central role in binding together parts of stimuli so that they are remembered together later (for discussions, see Moscovitch, 2000;Reinitz, 2002).…”

Perceptual specificity of priming for compound words not presented

“…The authors proposed that independent access of postperceptual representations for stimulus parts (e.g., check from checklist and point from needlepoint) supports performance on word fragment completion and recognition memory tasks, but not on perceptual identification tasks. However, significant perceptual identification priming for conjunction words and feature words (half old, half new words) has since been obtained (Reinitz, Verfaellie, & Milberg, 1996). This finding appears to be at odds with the earlier theoretical proposal: If priming for conjunction words can be obtained for perceptual identification, too, then additional representations seem unnecessary.…”