Year

2013

Degree Name

Doctor of Philosophy

Department

School of Psychology

Abstract

Verbal short-term memory and long-term memory were once thought to be distinct processes, with short-term memory responsible for the maintenance and manipulation of verbal information in the immediate consciousness, and long-term memory responsible for storage of words and associated information over the lifespan. However, more recent research has suggested that the processes are intrinsically linked. The ability to learn novel items suggests that short-term memory is able to contribute to long-term memory, and recent findings also indicate that long-term memory has an influence on short-term memory. Typically, long-term memory is thought to assist the recall of items from short-term memory through a late stage redintegration process but does not influence the serial ordering of items in memory. More recently, one long-term memory variable, phonological neighbourhood density has been shown to influence memory for order in an immediate serial recall task with lists containing both large and small phonological neighbourhood words. This result has implications for the locus of the link between long-term memory and short-term memory, and is indicative of a link occurring earlier than assumed by most models of short-term memory. The aim of this thesis was to replicate evidence that phonological neighbourhood density is a variable that is sensitive to order processing and to identify the conditions under which this sensitivity appears. In a series of experiments, the effect of phonological neighbourhood density on memory for order was examined using tasks that varied in their requirement to recall order information or item information. Memory for order was better for words from large neighbourhoods than from small neighbourhoods, for pure lists but not for mixed lists in a task that reduces the requirement to remember item information, namely serial reconstruction. Further experiments using this paradigm revealed that this result was not due to differences in word lengths between large and small stimuli sets nor due to differences in reaction time to the items at output, and could not be attributed to different processes underlying immediate serial recall and serial reconstruction. The difference in memory for order between large- and small- phonological neighbourhood lists was found to disappear under conditions of articulatory suppression, suggesting that sub-vocal articulation acts as an intermediary between long-term memory and short-term memory. Finally, a second long-term memory variable, word frequency, was examined and was also shown to influence memory for order. This is indicative of a general long-term memory effect on short-term memory for order that has not been identified previously in the literature. These results are discussed in line with models of memory for order and are found to be most compatible with psycholinguistic models of memory that view order as a pattern of activation across a set of lexical units within a common system for memory and language.

Share

COinS