Processing Views of Remembering

The chapter describes and discusses previous accounts that viewed human memory as an activity of mind. These include members of the “Act Psychology School” and other early psychologists described by Boring (1950). The theoretical ideas of James (1890) and Bartlett (1932) are described and discussed, especially as emphasized in Bartlett’s 1932 classic book, Remembering: A Study in Experimental and Social Psychology. The notions associated with “activity theory” in Soviet psychology are outlined, and the studies in educational psychology deriving from these theories are described. The relevance of Hebb’s theory of cell assemblies is pointed out, as is the congenial work of James Jenkins and his students in the 1960s and 1970s. These latter studies are a clear forerunner of later experiments in the levels of processing tradition. Finally, Robert Crowder’s views on proceduralism are summarized and discussed.

Encoding and Encoding–Retrieval Interactions

Topics covered include the notion that there are no dedicated memory encoding processes as such; rather, encoding is equated with perception and comprehension. Studies by Craik, Moscovitch, and McDowd (1994), and by Challis, Velichkovsky, and Craik (1996) are described and their implications discussed. The differential sensitivity of retrieval tests (e.g., implicit/explicit) to different encoding operations is discussed in terms of test requirements for different types of information rather than as reflecting different memory systems. General principles of encoding are proposed; these include depth, elaboration, congruity, and distinctiveness. Memory-boosting procedures such as rehearsal, organization, retrieval as encoding, the generation effect, and subject-performed tasks are discussed and evaluated. Other topics described, illustrated, and discussed include the picture superiority effect, the self-reference effect, the effect of task difficulty on later memory, differential effects of divided attention at encoding and retrieval, and the concept of consolidation.

Neuroimaging Studies

This short chapter first describes studies using PET scanning and fMRI imaging carried out by the author in collaboration with colleagues over the past 25 years. The main purpose of the chapter is to assess the extent to which current work on neuroimaging is compatible with the findings and ideas derived from the cognitive experiments described in previous chapters. The questions asked include: What are the neural correlates of deeper processing, and does the neuroimaging evidence illuminate the reasons for the strong relation between semantic processing and good memory? Is there evidence to support the proposal that retrieval processes recapitulate encoding processes? Is the similarity between perception and memory borne out at the neural level? How does novelty affect memory, and is there a conflict between the claims that both novel and familiar experiences are associated with good levels of recollection? What exactly are processing resources at the neural level? And, finally, how does the author’s emphasis on remembering as an activity square with the evidence from neuroimaging studies?

Aging I

Memory performance declines in the course of healthy aging, and this chapter discusses some reasons why this may be so. The author suggests that there is an age-related decline in both processing resources and in cognitive control, and that these deficiencies underlie less efficient encoding and retrieval processes. Age-related memory losses are greater in some tasks than in others, however, and the case is made that losses are relatively slight in situations that involve substantial amounts of environmental support and therefore require small amounts of self-initiated activity. In turn, the inefficiency of self-initiated activities is attributed to age-related deficiencies in frontal lobe functions. Age-related deficits in recall performance (which is heavily reliant on self-initiation) are reduced in a recognition test, which embodies greater environmental support. Deficits were also reduced by the use of pictures as materials, and there were no age differences in the ability to hold high-valued words in working memory. These effects are illustrated by experiments carried out by the author and collaborators.

Final Reflections and Future Trends

In this final chapter some big-picture topics are described and discussed. These include the concept of hierarchies in memory theory and an assessment of their validity in the levels-of-processing (LOP) framework (e.g., does the construct of “LOP” connote a continuum of depth or a series of qualitative stages?). A further topic is the hypothesized organization of representations from episodic-specific to abstract-general. Other issues in encoding, retrieval, and their interactions are considered, including some recent findings on the effects of divided attention (DA) at the time of retrieval. Hintzman’s ideas on reminding and recurrence, and also Jacoby’s concept of source-constrained retrieval, are discussed and evaluated. The author’s perspective on working memory is described, including the view that there is no need to invoke discrete stores or memory buffers. Some further issues in cognitive aging are discussed, including a proximal-distal hypothesis of when deficits are found. The similarities and differences between perception and memory are assessed, and the author’s perspective on the concept of memory systems is described and discussed. The final conclusion is that remembering should be viewed as an activity of mind and brain.

Remembering

This chapter makes the case for understanding memory in terms of qualitatively different codes represented by active processes, as opposed to such memory systems as episodic and semantic memory. A distinction is made between primary memory (PM), viewed as active conscious processing, and secondary memory (SM), viewed as the long-term representations of events and knowledge. The notion that PM involves attention paid to the information held in mind is discussed in light of current views of working memory. SM is described in terms of a hierarchically organized set of analytic representations running from specific episodes to context-free knowledge, as an alternative description to Tulving’s account in terms of separate episodic and semantic systems. The role of the external context in supporting retrieval is emphasized, and also the role of executive processes in enabling self-initiated activities when such environmental support is absent. Following previous researchers, the chapter endorses the notion of remembering as a set of active analytic operations, and stresses the similarity between the processes of perceiving and remembering. These ideas are illustrated by empirical examples.

From Short-term Memory to Working Memory

The concept of primary memory (PM) in the levels-of-processing (LOP) framework is described and illustrated. The history of the short-term/long-term memory (STM/LTM) distinction is reviewed. Early studies of STM in the Craik laboratory are described, including work on dichotic listening, the negative recency effect, and the surprising finding of long-lasting auditory information in STM. The distinction between PM and secondary memory is reviewed, and their respective roles in short-term retention discussed. Some experiments on release from proactive interference are described. The theoretical evolution of the construct of PM to that of working memory (WM) is described. The proposal that WM is equivalent to “attention paid to information in conscious awareness” is evaluated. It is argued that PM and WM are not separate entities, but are on a descriptive continuum. Further topics include the role of secondary memory in WM, and the possibility of LOP effects in WM (explored in experiments carried out by Craik and Nathan Rose). Further topics include a consideration of WM as a set of separable abilities, hierarchical views of WM, and a final integrated view of short-term retention.

Aging II

Chapter 8 focuses on empirical studies of age-related differences in memory performance. In accordance with the concepts of environmental support and self-initiated activities it is shown that age decrements are greater in cued recall than in recognition, also that cued recall is more resource-demanding than recognition. A further study found that age-related deficits were reduced by the use of semantically related materials but that older adults showed greater dual-task costs, especially at the time of retrieval. Age-related impairments were reduced by the use of pictorial materials as opposed to words; and reinstatement of the encoding context at retrieval was particularly helpful to older adults. A study carried out in California showed how individual differences in verbal intelligence and in daily activity levels modified the pattern of findings. The point that older adults have difficulty retrieving highly specific information is discussed and illustrated. Experiments are described that investigate memory for the source of learned information, and for age differences in prospective memory; younger adults outperformed their older counterparts in both situations. Finally, some interesting patterns of age differences in performance of implicit and explicit memory tasks are described and discussed.

Levels of Processing

This chapter focuses on the background and development of the levels of processing (LOP) ideas proposed by Craik and Lockhart (1972) and on the empirical support for the ideas provided by Craik and Tulving (1975). The chapter describes how the concept arose from the British work on models of attention by Donald Broadbent and Anne Treisman in the 1960s; specifically on how the concept of depth of processing grafts a framework for memory research on to Treisman’s hierarchical model of selective attention. After a brief survey of empirical work, the chapter deals with criticisms and rebuttals of the LOP ideas and findings, its relations to other ideas such as transfer–appropriate processing and to some more recent extensions.

Retrieval Processes

Endel Tulving’s views of synergistic ecphory and cue-dependent forgetting are discussed and endorsed, in particular the view that external stimulation (or self-initiated internal stimulation) necessarily interacts with encoded records to yield retrieval. Paul Kolers’ view of retrieval as repetition of processing operations is also evaluated. Other topics include retrieval as recapitulation of encoding, transfer-appropriate processing, environmental and schematic support, and self-initiated activities. It is concluded that the concepts of levels of processing and transfer-appropriate processing are both necessary to describe observed patterns of retrieval. Two postulated bases for recognition memory—familiarity and recollection—are described and evaluated, as are the ideas of processing fluency and attribution proposed by Larry Jacoby. Finally, studies of involuntary retrieval, mind-wandering, and prospective memory are described and their implications assessed.