Imagery, Memory and Cognition (PLE: Memory)

Essays in Honor of Allan Paivio

John C. Yuille

• Publisher: Psychology Press
• Year: 2014
• eBook ISBN: 9781317685463
• Topic: Psychology
• Subtopic: Cognitive Psychology & Cognition
• Index: Psychology

Originally published in 1983, the 14 chapters in this volume are based upon presentations made to a conference held at the University of Western Ontario in June, 1981. The primary purpose of that conference was to mark the 10th anniversary of the publication of Allan Paivio's text, Imagery and Verbal Processes, and to acknowledge the continuing contribution that Paivio was making to imagery research and theory at the time.

His landmark book had been the major publication in the field of imagery, and during the decade prior to this volume Paivio's theorizing and research dominated the investigation of imaginal processes. It was felt the most appropriate way to honor his achievements and activities, was to hold a conference on current developments in imagery research and theory at the time.

1	Spatial-Imagery Ability, Sex Differences, and Hemispheric Functioning

Carole H. Ernest

The central theme of the research I am reporting evolved from a study that Al Paivio and I did when I was a graduate student in his lab. In that study (Paivio & Ernest, 1971) we asked the following question: Are high imagers more accurate than low imagers in their visual recognition of nonverbal stimuli such as pictures and geometric forms, particularly if these stimuli are presented to the right hemisphere of the brain? In other words, is high visual imagery ability associated with superior right hemispheric functioning?

The question was a theoretically important one for two reasons. First, Paivio’s (1971) dual-coding approach to memory and cognition distinguished two independent but interconnected symbolic processing systems, a verbal system and a nonverbal or imagery system. The verbal system was viewed as being specialized for dealing with relatively abstract information, such as language, whereas the specialization of the imagery system was processing concrete-perceptual information, such as nonverbal objects or events (see Paivio, 1978a, and this volume, for recent discussions of the assumptions underlying dual-coding theory). We viewed imagery ability as the individual differences counterpart of the imagery symbolic system. And we defined visual imagery ability primarily in terms of psychometric tests of spatial ability, tests such as the revised Minnesota Paper Form Board test (MPFB; Likert & Quasha, 1941) and Space Relations of the Differential Aptitute Test battery (Bennett, Seashore, & Wesman, 1947). Both of these tests are measures of spatial visualization, requiring individuals to manipulate two-dimensional figurai information mentally in three-dimensional space. Our choice of visualization tests to index imagery ability, as distinct from other forms of spatial functioning (see Barratt, 1953; Ekstrom, French, Harman, & Dermen, 1976; Guilford, 1967; McGee, 1979), was influenced by Barratt’s (1953) factor analytic study in which he concluded that “imagery is an important component in the solution of those tasks that involve the ‘mental’ manipulation of spatial relations” (pp. 160–161 ; more recently, see Carpenter & Just, in press).

A second reason why the question was important theoretically was because Paivio’s two symbolic systems seemed to have their counterparts in the neuropsychological literature on lateralization of brain function. Studies using braindamaged populations (e.g., Paterson & Zangwill, 1944; Penfield & Roberts, 1959; Sperry & Gazzaniga, 1967) as well as normal individuals (e.g., Kimura, 1961, 1964; Kimura & Durnford, 1974) suggested that the left hemisphere of the brain is specialized for verbal/linguistic/analytic functions whereas the special skills associated with the right hemisphere pertain to nonverbal/spatial information processing as well as global/holistic modes of analysis.

Thus, it seemed reasonable to expect that people who have enhanced spatial-imaginal skills should excel in a task that requires the visual recognition of nonverbal stimuli when such stimli are presented initially to the right hemisphere of the brain. No differences between high and low spatial-imagers might be anticipated when these stimuli are presented to the left hemisphere, that is, the verbal hemisphere. Nor indeed, we speculated, should the imagery groups differ in their recognition of verbal stimuli, such as single letters.

The program of research discussed in this chapter began with essentially the same question. But it also answered some questions I didn’t realize I was asking. This research is interpreted within a framework (or model) of hemispheric functioning for high and low spatial-imagers. In its simplest form, this framework suggests essentially two things. It suggests that although high spatial-imagers do excel in their perceptual processing of nonverbal information relative to low imagers, this superiority does not appear to be exclusive to the right hemisphere of the brain. Rather, high spatial-imagers seem to be “bilater-lized” for spatial processing. Secondly, individual differences in spatial-imagery ability must be examined within the context of sex differences. Male and female high spatials do not always behave similarly, nor do male and female low spatials. Instead, the sexes within these two ability groups appear to differ in an unexpected fashion in the “lateralization” of their verbal functions.

My intention is to discuss several unpublished studies which evolved from the question posed above; to propose and elaborate an organizational framework which seems to encompass most of the experimental findings reported here; and, finally, to speculate on what relevance the proposed framework may have to the more general cognitive functioning of high and low spatial-imagers, particularly in the areas of learning and memory.

But first, the present approach is placed in broader perspective by reviewing selectively relevant literature in three independent but interrelated areas of enquiry: imagery and perception; cognitive abilities from an information processing perspective; and individual differences in cerebral lateralization. These areas are viewed as independent because they evolved from different “motivations.” They are viewed as interrelated because all three postulate the existence of two coding or processing systems—a verbal/analytic system and a spatial/holistic system.

THREE PERSPECTIVES

Imagery and Perception. The notion that imagery and perception tap similar underlying processes is explicit in Paivio’s dual-coding theory (1971, 1978a, 1978b) as well as in other theoretical and empirical statements in the literature (e.g., Brooks, 1968; Hebb, 1968; Segal & Fusella, 1970). Hebb (1968), for example, postulates three levels of cell assembly activity. The actual perception of an object, he proposes, involves the activation of first-order as well as higher-order assemblies, whereas a memory image “may consist only of second-and higher-order assemblies, without the first-order ones that would give it the completeness and vividness of perception [1968, p. 473].” And Paivio (1978b) has suggested that the representational units of the image system (“imagens”) may be viewed as perceptual isomorphs or analogs, whereas those of the verbal system (“logogens”; Morton, 1969) are assumed to be discrete entities “only arbitrarily related to perceptual information [p. 379].”

Possibly the first evidence that spatial-imagery ability and perception may be functionally related emerged unexpectedly in a study on incidental learning (Ernest & Paivio, 1971a, Expt. 2). The orienting task required written identification of briefly-exposed pictures or words; this was followed by an unexpected free recall task. Surprisingly, high imagers excelled in picture identification but not in concrete or abstract word identification. This differential pattern of imagery-picture and imagery-word effects was subsequently pursued in a series of studies employing different recognition paradigms and procedures (Ernest, 1972, 1979, 1980; Paivio & Ernest, 1971). The paradigms were recognition threshold, recognition latency, and visual half-field—all of which involved a clear speed component, either with respect to stimulus presentation or response requirements. Timed, but relatively unspeeded, paper-and-pencil tests requiring the identification of fragmented pictures and fragmented words were also used. Procedural differences included the presentation of pictures and words in heterogeneous lists, as homogeneous blocks in a within-subjects design, or as homogeneous lists in a between-subjects design, or as homogeneous lists in a between-subjects design. The sought-after differential pattern of picture/word effects emerged most unambiguously using the latter experimental design, that is, homogeneous list presentation to independent groups of high and low imagers (Ernest, 1979, Expt. 4, as well as Ernest & Paivio, 1971a, Expt. 2). The only occasion when imagery ability was not clearly associated with superior picture recognition, even under these circumstances, was in the threshold paradigm. Here, imagery differences occurred only for pictures relatively low in familiarity (Ernest, 1979, Expt. 2).

From these studies it seems reasonable to conclude that the cognitive skills involved in the solution of spatial manipulation tests—skills such as encoding and storing figurai segments and mentally constructing a complete figure or gestalt from these segments—are also involved in the identification of figurai information under speeded or reduced cue conditions. Clearly, however, the conditions for observing such a relationship are not easily determined. Stimulus mode per se seems to be less the source of inconsistency than the context within which stimuli are presented. Contexts that permit the “expected” processing strategy to be primed—“expected” meaning congruent with stimulus mode— appear to yield most unambiguously the imagery ability-picture/word effects first observed, serendipitously, in the incidental learning study. In other words, imagery ability-picture effects emerge when a nonverbal/spatial processing strategy is primed, presumably because such a strategy can be used more effectively by high than by low spatial-imagers. Imagery ability-word effects do not emerge when a verbal/linguistic processing strategy is primed, presumably because high and low imagers are comparable with respect to their competence with this strategy.

The view that context can modify one’s processing strategy is not new, of course. Indeed, context effects are of interest in their own right and have attracted the attention of many investigators in recent years (e.g., Godden & Baddeley, 1980; Stanovich & West, 1979). They represent a phenomenon that must be reckoned with—at a minimum from the perspective of choice of experimental design (see also Poulton, 1975).

Also relevant to a discussion of spatial-imagery ability and perception is evidence that high spatial-imagers can generate images to words more quickly than can low imagers (Ernest & Paivio, 1971b). This finding is compatible with evidence from recent studies contrasting high and low imagers in speed of mental comparisons (e.g., Paivio, 1978c, d). In mental comparisons tasks (Moyer, 1973), individuals are presented two words, for example, and must decide which one of the pair has more, or less, of a given attribute. Typically, the more similar are the item pairs in a given attribute, such as size, the more difficult is the decision; that is, response times are longer. This has been labeled the “symbolic distance” effect by Moyer and Bayer (1976). Paivio (1975) has assumed that performance on such tasks requires the generation of long-term memory representations of the named items via the imagery system, with these perceptual analogs then mentally contrasted with respect to the relevant attribute.

Pertinent here are those studies concerned with intrinsic or defining attributes of objects. In one study (Paivio, 1978c), for example, high and low spatial-imagers were required to decide which of two digitally-presented clock times formed the smaller angle. Both ability groups demonstrated the typical symbolic distance effect; more importantly, high imagers were significantly faster in two of three “mental clocks” experiments, although the trend was the same in all three. A similar trend emerged when the defining attribute of shape was involved and the stimuli were word pairs (Paivio, 1978d).

These findings have been confirmed in a group-administered version of the “mental clocks” task as well as a size comparisons task where the dependent measure was number correct within a specified time period. Paivio and Harsh-man (see Paivio, 1980) report significant correlations between Space Relations and the accuracy measures for both tasks. Interestingly, verbal processes, as reflected in Inference Test and (to a lesser degree) Word Fluency scores, were also significantly...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Original Title Page
• Original Copyright Page
• Table of Contents
• Preface
• 1. Spatial-Imagery Ability, Sex Differences, and Hemispheric Functioning
• 2. What Does It Mean to be a High Imager?
• 3. Picture Memory
• 4. Imagery Instructions and the Organization of Memory
• 5. Pictures and Words in Semantic Decisions
• 6. Comprehension Effects in Comparative Judgments
• 7. Properties of Mental Color Codes
• 8. Schemas and Images In Self-Recognition
• 9. Paivio’s Dual-Coding Model of Meaning Revisited
• 10. Representational Memory: Paivio’s Levels of Meaning as Experiential Model and Conceptual Framework
• 11. Expectancy, Equilibration, and Memory
• 12. The Crisis in Theories of Mental Imagery
• 13. Emotion, Imagery, and Verbal Codes: A Closer Look at an Increasingly Complex Interaction
• 14. Empirical Case for Dual Coding
• Author Index
• Subject Index