The Dissociation between Tacit and Reflective Measures of Memory

Abstract

<p>RE-cognition - "to know again" - implies an element of awareness for remembering to occur. But is awareness a necessary condition for remembering? Research with amnesics shows that it is not. Amnesics, by definition, have a memory deficit in that they are unable to report recent prior experiences. None the less, they still display the effects of those prior encounters on their behaviour if tested in appropriate ways (e.g. Corkin, 1968; Warrington and Weiskrantz, 1970; Brooks and Baddeley, 1976). Although the amnesic does not recognize having participated in a particular task before (for example, reading inverted text - Cohen and Squire, 1980), he/she will show savings or improvement in the learning of that task on subsequent occasions when measures are employed that do not require a conscious report of the earlier event. In other words, amnesics do have memory, but are unaware of its influence.</p> <p>Persons with normal memory abilities can also show evidence of memory in their actions or behaviour without coincident awareness of remembering. For example, imagine the execution of well-practiced skills, such as driving a car or typing a manuscript. These tasks require a great amount of prior knowledge in order to be performed properly, but do not seem to tax one's memory or rely on one's being aware of remembering the sequence of skills needed to be carried out.</p> <p>The implication of these findings is that there can be a "dissociation" in memory between the awareness of remembering and having one's ongoing behaviour influenced by remembering. It may be that many effects of memory remain undetected, given that traditional memory research measures require the expression of deliberate interrogation and conscious retrieval of prior memory events; and consequently, ignore the investigation of tacit memory forms.</p> <p>The main purpose of this dissertation is to experimentally identify, investigate and evaluate this dissociation as revealed by the memory measures. The generality of the effect and possible theoretical accounts of the effect will be explored. The first three experiments demonstrate that the dissociation can be obtained with a variety of populations and tasks. One class of measures will demonstrate with amnesics the existence of a memory influence, while another set simultaneously will deny its presence. With normal subjects, each of these measures may consistently demonstrate an effect of memory, but they will produce behavioural results which are stochastically and phenomenologically independent of one another. In Experiments Four and Five, further specification of this dissociation will be obtained by demonstrating that under certain conditions the dissociative result can be disguised or eliminated through manipulations of particular experimental variables. Finally, the results from Experiment Six provide converging evidence to suggest that the dissociation is related to the type of information that is available for processing. The relationship between the measures reflects the cognitive processes of remembering which, in fact, are dissociable.</p>