Geoffrey O'Shea

Chapter 2 – Spreading Activation

Contiguity links, or those positing associations between environmental stimuli and memory content, are examined as an historical foundation for empiricist views of memory and as an impetus for developing theoretical models of the memory search process. Anderson and Bower’s (1973) spreading activation model, FRAN (free recall in an associative network), is critiqued in terms of its significance in shifting theorists’ understanding of the nature of memory search from a contextual basis, influenced by the traditional empiricist view, to one in line with a constructivist view positing a network of abstract cognitive representations generated during learning. Access to memory content is discussed in terms of the relationship between specification cues and descriptive headers developed through abstract cognitive codes. Anderson’s further contributions toward conceptualizing the organization of memory and the role of specification cues in memory search are explored in his ACT (adaptive control of thought) and ACT-R models. Models of sequential recall, such as chaining models and complex chaining models, are reexamined with a constructivist view using examples from everyday memory.

Chapter 3 – Processing Structures

What are the underlying processes used for recalling a series of items in order? Theories for how memory codes for serial order are examined beginning with Ebbinghaus’s early associative model and proceeding to modern chaining models. Issues concerning the inability of chaining models to account for retention errors of sequential information, such as intrusions and transpositions, are discussed as the basis for proposing alternative models. The notion of a processing structure operating on a time-tracking function is introduced to explain how memory codes for the temporal positioning of items in a sequence with the implication that a context is established for the learning episode that goes beyond direct item-to-item associations. A wide range of models accounting for the internal operation of temporal processing structures, and the error patterns generated by these structures, are reviewed, including Conrad’s (1965) boxes model, Estes’s perturbation model (1972, 1985, 1997), and modern oscillator models such as Burgess and Hitch’s (2006) connectionist and Brown, Preece, and Hulme’s (2000) multioscillator model.

Dissociative Memory, Variables that Influence Reconstruction, and Propositional Coding

This chapter explores the dissociable nature of human memory in terms of the relationship between the information encoded into memory and the complex cognitive processes used to organize and interpret this information. Inferential reconstruction in recall is reviewed through how we use higher-order processes to verify the logic and plausibility of remembered events, as well as through neurological studies demonstrating the effects of frontal lobe damage on inferential processes. Furthermore, issues involving the accuracy of inference are discussed such as errors due to forced recollection; inferential effects at the various hierarchical levels of autobiographical memory; Wagenaar and Groeneweg’s (1990) work on reconstructive errors in remote personal memories; and distinctions between memory replacement and inferential reconstruction. The role of abstraction in recollection, as well as its function in providing extended content in long-term memory, is considered through a discussion of how propositional and analog codes create syntactic links with memory content.

Chapter 4 – Constructivism

The roots of constructivism in memory research are traced from its Kantian origins to modern approaches, with emphasis given to the distinctions between constructivism and empiricism in interpreting memory phenomena. The works of Frederick Bartlett and Jean Piaget are reviewed as representing a hard-line constructivist position characterized by viewing memory encoding and retrieval processes as largely interpretive and serving the function of providing coherence to our understanding of the events we experience. Bartlett’s seminal work using his repeated testing paradigm to study the nature of higher-order schemas is critiqued and discussed in support of the notion that schemas are a fundamental part of the organization of human memory and aid retention by providing memory strength. The legacy of Bartlett’s repeated testing paradigm is briefly reviewed through its contribution to research on hypermnesia, reminiscence, and output interference, as well as its role in shifting memory researchers’ mode of inquiry from empiricism to constructivism.

Piaget’s Model

The Genevan view, representing a hard-line constructivist position, is discussed with regard to the differing natures of episodic and semantic memory. Schemas, which provide the background organization to long-term memory, and thus a structure in which to fit new information, are considered to play a central role in memory strengthening under the Genevan view, which posits the non-divorce hypothesis. or the idea that activated schemas continue to make contact with long-term memory content rather than terminating at a specific point during retrieval. The Genevan view, by emphasizing semantic coding through automatic processes, is contrasted with Baddeley’s theory of memory, with its emphasis on perceptual coding and conscious control processes to retain information. Theoretical questions, such as the different degrees of schematic integration present in episodic and semantic memory, are considered within critical reviews of research on amnesia, category formation in long-term memory, and Piaget’s studies on seriation development evident in the drawings of children.

Chapter 7 – Altered Memories

Research examining alterations in memory content during retrieval is used to explore the distinctions between the traditional/empiricist and alternative/constructivist views of memory. Specifically, the debate centers on the mechanisms by which retrieved content differs from stored long-term memory content. Under the empiricist view, episodic memory content for one event is independent of content for other events. As a result, changes to content that occur during retrieval are considered to derive from interference between the two memory stores such that the retrieval of one event blocks the retrieval of the other event. Conversely, the constructivist view posits interaction among all episodic memory content such that retrieval of one event can potentially activate the retrieval of related events. Schank’s (1982) model, based on the interaction between headers and links to subset content, is discussed as a mechanism by which long-term memory content can be altered during retrieval. The implications of the empiricist and constructivist positions are discussed with regard to Elizabeth Loftus’s replacement hypothesis for eyewitness memory change.

Links and Cues

Chapter 1 reviews the underlying processes by which retrieval operates through links that provide movement between experiences in awareness and content stored in memory. Various types of links are identified such as identity, temporal, contiguity, opposition, similarity, and causality, as well as examples of how these links function in everyday memory. The historical role of links in shaping our understanding of retrieval processes in memory is also examined, with insights from Aristotle to present-day investigators. Research on the significance of cues in providing specific links between target material and memory content is also discussed through Tulving’s cue-dependent theory of recall and forgetting. Two specific models for cued retrieval are reviewed: the cyclical retrieval model, based largely on Gillund and Shiffrin’s (1984) work, is contrasted with Anderson and Bower’s (1973) spreading activation model. Differences between key components of these models, such as how links operate in working memory and long-term memory, as well as how these models account for the retrieval of autobiographical memory content, are discussed and illustrated with real-world examples.