A cognitive neuropsychological approach to false memory: Korsakoff patients and the DRM paradigm

Human memory can be surprisingly inaccurate. Not only do people forget thi ngs that happened, they also remember things that never happened. When the latter occurs, they are said to produce false memories . Despi te a long research tradition within the field of cognitive psychology, t he role of cognitive neuropsychology in false memory research has bee n rather limited. Nevertheless, neuropsychological studies about amnesia have had a tremendous impact on research and theorizing about more trad itional aspects of memory, such as the distinction between implicit and explicit retrieval. The present dissertation was aimed at integrating bo th lines of research. The objective was twofold, as we both wanted to lo ok into the memory abilities of Korsakoff patients, and to obtain a bett er picture of how false memories are created in healthy individuals. In th e DRM paradigm, participants study lists of words (e.g., bed, rest, aw ake, tired, dream, ) which are all semantically related to a critica l, but non-presented lure word (e.g., sleep). False memory is measu red as the tendency to recall and/or recognize such lure words, and has been shown to occur with a probability similar to the probability of rec alling/recognizing actually presented list words. Pioneering work of Sch acter and colleagues revealed that amnesic patients are less susceptible to false recognition and show different patterns of false recall than m emory-intact controls. There are two alternative explanations to account for these findings. First, in line with theories of amnesia referring t o deficient relational or contextual processing, diminished false memory might be due to defective encoding of the relationships between the lis t words. Alternatively, diminished false memory might also be due to def ective retrieval of properly encoded information: DRM memory tests typic ally comprise explicit retrieval instructions, and amnesic patients are characterized by impaired intentional recollection. The f irst three studies were focused on Korsakoff patients memory abilities. Whereas both immediate and delayed recall and recognition tests were pr ovided to patients and memory-intact controls in Experiment 1, implicit stem completion tests were administered in Experiments 2 and 3. Implicit memory differs from explicit memory in that it reflects automatic rathe r than controlled retrieval of information from the study phase. Both en coding instructions and encoding duration were manipulated, in order to support the patients relational processing. First, Korsakoff patients false recall/recognition was only significantly lower than controls whe n testing was postponed until the end of the experiment, but not when te sting immediately followed each study list. In the latter situation, Kor sakoff patients even showed more remember judgments of critical lures than control participants did. This could be attributed to fluency-based retrieval, unopposed by explicit recollection. Second, Korsakoff patien ts level of implicit false memory was similar to controls . Presentatio n duration of the list words only affected explicit memory: Memory score s increased when given more time to study. Hence, results of all three e xperiments emphasized the distinction between automatic and controlled r etrieval, and strongly pointed to defective explicit recollection being the main cause of earlier found diminished false memory in amnesia. The g oal of the next two studies was twofold: They were designed to investiga te a discrepancy between our own and other results reported in the liter ature, as well as to shed a light on the involvement of consciously cont rolled and automatic encoding processes in the formation of false memori es. DRM word lists were studied either intentionally or incidentally, an d participants were asked to think out loud during study. Memory was tes ted both implicitly and explicitly. Again, no group differences in impli cit false memory were obtained. Korsakoff patients did verbalize less cr itical lures than controls during intentional encoding, and showed impai red explicit recognition. Performance was lowest in a condition with nor mal study time and incidental encoding, and increased both with addition al study time and when asked to intentionally remember the list words. L ure verbalization was shown to contribute to explicit false memory, but had no clear-cut effect on false priming. Hence, evidence was obtained t hat conscious lure activation is not required to obtain significant leve ls of implicit false memory. Based on preliminary evidence that encoding difficulties might be (partly) re sponsible for the patients recollection deficit, Experiments 6 and 7 ex amined whether enhancing the saliency of the relational information woul d influence Korsakoff patients explicit false memory. First, embedding DRM list words in short stories did not affect delayed recognition: Pati ents performed consistently worse than controls. Second, however, immedi ate false recognition was affected by the personally-relevant nature of the lists: Both Korsakoff patients and memory-intact alcoholics showed l ess false recognition of alcohol-related than of neutral critical lures. Importantly, however, stimulus content did not improve Korsakoff patien ts recognition discriminability, but merely changed the response criter ia used and hence the patients tendency to accept an item as being old. In a final study, the association between clinical and experimental false mem ory was investigated. Confabulatory behavior was shown to be positively associated with false recognition of unrelated distractor words, but neg atively associated with (delayed) false recognition of critical lures. T he latter could be attributed to Korsakoff patients impaired strategic retrieval. In co nclusion, in eight experiments, the type of retrieval instructions used was shown to be the most important factor underlying both past and prese nt findings of diminished false memory in amnesia. Korsakoff patients co nsistently showed normal implicit and impaired explicit veridical and false memory. Nonetheless, deficient encoding does appear to be partly responsible for the patients recollection deficit, with diminished lure generation during study being a possible mediating factor determining e xplicit false memory. More generally, evidence was obtained that false m emories can be induced through both non-intentional, automatic encodi ng and non-intentional, automatic retrieval. Elaborative processing and conscious lure generation did contribute to explicit false memory, but h ad no effect whatsoever on implicit false memory.status: publishe