Human Spatial Memory

Remembering Where

Gary L. Allen, Gary L. Allen

• Publisher: Psychology Press
• Year: 2004
• ISBN:
  9781135635121
• Edition: 1
• Topic: Psicología
• Subtopic: Historia y teoría en psicología

The chapters in Human Spatial Memory: Remembering Where present a fascinating picture of an everyday aspect of mental life that is as intriguing to people outside of academia as it is to scientists studying human cognition and behavior. The questions are as old as the study of mind itself: How do we remember where objects are located? How do we remember where we are in relation to other places? What is the origin and developmental course of spatial memory? What neural structures are involved in remembering where? How do we come to understand scaled-down versions of places as symbolic representations of actual places? Although the questions are old, some of the answers-in-progress are new, thanks to some innovative theorizing, solid experimental work, and revealing applications of new technologies, such as virtual environments and brain imaging techniques. This volume includes a variety of theoretical, empirical, and methodological advances that invite readers to make their own novel connections between theory and research. Scholars who study spatial cognition can benefit from examining the latest from well-established experts, as well as milestone contributions from early-career researchers. This combination provides the reader with a sense of past, present, and future in terms of spatial memory research. Just as important, however, is the value of the volume as a touchstone resource for researchers who study perception, memory, or cognition but who are not concerned primarily with the spatial domain. All readers may find the fact that this volume violates the trend toward an ever-narrowing specialization refreshing. Chapters from cognitive psychologists are alongside chapters by developmentalists and neuroscientists; results from field studies are just pages away from those based on fMRI during observation of virtual displays. Thus, the book invites integrative examination across disciplines, research areas, and methodological approaches.

II
The Task of Remembering “Where Is It?”

The four chapters in this section each have unique aspects but also share some important themes. In chapter 4 (this volume), Jeanne Sholl and Stephanie Fraone first provide a comprehensive review of the concept of spatial working memory and its application to small-scale space, that is, to the type of situation typically involving a stationary observer and a small, often two-dimensional figure, object, or object array. Then, they expand their consideration to the role of spatial working memory in large-scale space, that is, in the type of situation typically involving a mobile observer surrounded by large, stationary environmental objects. Thus, chapter 4 foreshadows the task of answering “Where am I?”, which is the concern of the next section. In chapter 5, Ruth Schumann-Hengsteler, Martin Strobl, and Christof Zoelch present findings from a research program focused on the development of spatiotemporal memory in children. There is good conceptual continuity with the preceding chapter 4 in terms of invoking a working memory theme and good continuity with chapter 6 that follows in terms of examining memory for sequences versus memory for patterns of locations. In chapter 6, David Uttal and Cynthia Chiong are concerned with the multifaceted cognitive challenges and benefits of conceiving of patterns of locations, a phenomenon they refer to as “seeing space in more than one way.” Again, the emphasis is on developmental analysis. In chapter 7, Albert Postma, Roy Kessels, and Marieke van Asselen examine the literature dealing with the neuropsychology of object-location memory and report extensive findings from their laboratory. The connections with other chapters in this section with respect to working memory mechanisms and spatiotemporal versus pattern memory are clear and compelling.

4
Visuospatial Working Memory for Different Scales of Space:

Weighing the Evidence

M. Jeanne Sholl and Stephanie K. Fraone
Boston College

It is commonly agreed that working memory (WM) entails the online maintenance and processing of information necessary for higher level cognitive functioning. Functional definitions of WM about which there is consensus include the “moment-to-moment monitoring, processing, and maintenance of information” in everyday cognition (Baddeley & Logie, 1999, p. 28) and the system underlying the maintenance of information in the service of complex cognition (Miyake & Shah, 1999). There is less consensus about the cognitive architecture that underlies WM. Baddeley’s (e.g., Baddeley, 1986; Baddeley & Hitch, 1974) fractionation of short-term store (Atkinson & Shriffrin, 1971) into a multicomponent WM system with a domain- general central executive and two domain-specific subsidiary storage systems—verbal and visuospatial—has framed much of the theoretical debate and empirical research in this area. Although not all subscribe to this model of WM, the psychological reality of domain-specific verbal and visuospatial WM subcomponents is largely beyond dispute and is not reviewed here (see Shah & Miyake, 1996, for a brief review).

The primary objective of this chapter is to explore the construct of visuospatial WM and to raise the question of whether a single WM system operates across different levels of spatiotemporal extendedness or whether there is a different WM system for each behaviorally relevant scale of space. We do not presume that we are able to answer this question here. Our more modest aim is to review the research on visuospatial WM at different spatial scales for the purpose of exploring underlying similarities and differences.

Montello (1993) identified three behaviorally relevant scales of space and has suggested that at each scale a different functional system may operate for processing visuospatial information. On a continuum of spatiotemporal extendedness, Montello differentiated between figural space (object-sized spaces), vista space (room-sized spaces), and environmental space (spaces that cannot be seen in their entirety from one vantage point). Montello made the case for functional differentiation outside the domain of WM, and in this chapter we apply his taxonomy to the construct of WM. To emphasize differences in scale, we restrict our review to figural and environmental scales of space, in the latter case focusing on regions of space small enough to be explored comfortably on foot. At both scales of space, we explore the construct of visuospatial WM from both a cognitive and a neuroscientific perspective. In some instances, the research we review on environmental space was actually conducted in room-sized space. However, in these instances vision was either obstructed or restricted so that the interobject relations defining the space were not simultaneously visible but instead were revealed over an extended trajectory, thus simulating, albeit at a reduced spatiotemporal scale, the properties of environmentally scaled space.

Perhaps because the construct of human WM developed from a chronometric analysis of cognitive functioning, the study of visuospatial WM has been largely limited to cognitive tasks using visuospatial displays small enough to fit onto a computer screen or a piece of paper. Others have provided comprehensive reviews of how WM operates at this figural scale of space (Logie, 1995; see also Schumann-Hengsteler, Strobl, & Zoelch, chap. 5, this volume), and we do not reiterate those efforts. Instead, we provide a limited review of some of the theoretical issues that have driven research on figural WM and from which evidence of its characteristic properties has emerged. We then explore the research on WM for larger scales of space.

Ideally, our review would include a comparison between the properties of figural WM and the properties of environmental WM. Such a comparison would allow a preliminary assessment of the extent to which the visuospatial WM system is unitary or divided because the greater the similarity in functional properties, the more likely there is to be a similar underlying architectural structure. However, although there has been extensive theoretically driven research on the properties of figural WM, there has been little research on the properties of environmental WM; therefore, a direct comparison of properties is not possible at this time.

Our review is organized into three sections. In the first section, we review WM for figural space from both a cognitive and neuroscientific perspective; in the second section, we do the same for environmental space; and in the last section, we explore the issue of their connectedness. Because the cognitive research on figural and environmental WM is at different stages of development, these sections are organized differently from one another. The cognitive review of WM for object-size spaces is organized around theoretical issues, whereas the cognitive review of WM for environment- size spaces is organized around tasks likely to require WM. The organizational structure mirrors the current state of cognitive research in each area, and the disparity in their organizational structure illuminates their current theoretical and empirical gaps. By laying out this disparity, our objective is to draw attention to some issues that may previously have been overlooked, to identify some of the gaps in our knowledge that would need to be filled before the question of spatiotemporal scale can be answered, and to suggest some possible lines of future inquiry.

THE VISUOSPATIAL WM SYSTEM(S) FOR A FIGURAL SCALE OF SPACE

In the first half of this section, we review some of the properties of WM for figural space and the theoretical issues that have driven cognitive research on these properties. Much of the research on visuospatial WM has been framed by Baddeley and Hitch’s (1974) multicomponent model of WM. According to this model, a visuospatial store or “sketchpad” acts as a medium for the temporary storage of analogue cognitive representations, formed either as a result of the perceptual analysis of visual input or the retrieval of visuospatial knowledge from long-term memory (LTM). Analogue representations are “depictive representations” (Kosslyn, 1983, p. 33) that preserve the visuospatial properties of the physical stimulus. Visuospatial information is stored in the sketchpad when temporary retention is required to solve a spatial problem. Types of everyday WM problems thought to require figural visuospatial WM include anticipating the outcome of a spatial transformation (“Will the oversized chair fit through the narrow door opening if it is rotated this way?”), mental rearrangement of a group of objects (“Will the suitcase fit into the overhead compartment if the stuff already in it is rearranged?”), anticipating how the parts of a whole will fit together (a skill required to assemble anything that comes in pieces), and so on. The control processes recruited to operate on the stored visuospatial information draw processing resources from a domain-general central executive. This general framework has motivated a series of empirical questions that have helped to articulate further the properties of visuospatial WM. These questions include whether the visuospatial store can be further subdivided into separate visual and spatial stores, the extent to which the visuospatial store is separable from executive functioning, and the role of attention in visuospatial store.

In the second half of this section, we review some of the neuroscientific research on WM for figural space in both nonhuman primates and humans. In large part, the neuroscientific research on visuospatial WM has different scientific origins from the cognitive behavioral research, yet there are interesting parallels between the two domains of inquiry.

Visuospatial WM From a Cognitive Perspective

Visual Versus Spatial Temporary Store. Interest in the fractionation of visuospatial store into separable visual and spatial subcomponents followed from empirical findings suggesting that its verbal counterpart, the phonological store, could be functionally subdivided into a phonological loop, which stores speech-based codes, and an articulatory mechanism, which refreshes the phonological codes to keep them activated. This led to the conjecture that analogous subdivisions may exist in the visuospatial store, with Logie (1995) having proposed a visual “cache” for the storage of visual information and an “inner scribe” that refreshes visual codes and is involved in the planning and control of movement. Logie equated space with movement; hence, he proposed a functional division between visual and spatial WM subcomponents.

Without subscribing to Logie’s (1995) subdivision of visuospatial store and with the adoption of a more conventional, location-based definition of space, we review behavioral evidence related to the separability of visual and spatial storage systems. Arguably the most convincing behavioral evidence is the double dissociation observed in interference paradigms (Della Sala, Gray, Baddeley, Allamano, & Wilson, 1999; Logie & Marchetti, 1991; Tresch, Sinnamon, & Seamon, 1993). In this paradigm, two primary tasks are each paired with the same two secondary tasks. If one secondary task interferes only with one primary task and the other secondary task interferes only with the other primary task, then each pair of primary/secondary tasks is likely to rely on a common cognitive substrate that is different from the substrate common to the other pair. The properties of the shared cognitive substrate underlying performance on the primary and secondary tasks are inferred from an analysis of the tasks’ commonalties.

The primary tasks used by Logie and Marchetti (1991) were short-term retention tasks in which participants held an inspection sequence in WM for 10 s prior to judging whether or not it matched a test sequence. In the visual primary task, the sequences consisted of four square patches in different shades (hues) of the same color, and in the spatial primary task, the sequences consisted of six square patches all of the same color shade but displayed at different locations. During a 10-s delay interval, participants performed a secondary task. They either engaged in a hand-movement task, which consisted of a nonsighted hand movement following a zigzag () trajectory, or they passively observed black and white line drawings of common objects and animals in an irrelevant-pictures task. The irrelevant- pictures task interfered with memory for hues but not locations, and the hand-movement task interfered with memory for location but not hues. These findings are consistent with a visual WM system that maintains information about color and shape over short delays and a spatial WM system that maintains information about spatial location in the service of action. Based on the latter finding, Logie and Marchetti argued that the primary function of a spatial WM system is to plan and execute spatially directed actions.

Using a similar experimental design but a very different set of primary and secondary tasks, Tresch et al. (1993) also found a double dissociation in visual and spatial short-term retention. Their spatial primary task tested retention of a single dot location for 10 s, and their visual primary task tested retention of a simple geometric form. Because of the simplicity of the primary tasks, stimulus duration was adjusted to produce an 80% to 90% accuracy rate. One secondary task was a movement-detection task—find the stationary character in a field of moving characters—and the other was color-discrimination task: judge whether a square is more blue than red or more red than blue. The movement-detection task interfered with the spatial but not the visual primary task, and the color-discrimination task interfered with the visual but not the spatial primary task.

Additional behavioral evidence for a functional dissociation between visual and spatial processing systems is provided by selective interference and developmental studies. In general, selective interference paradigms show that spatial-motor secondary tasks, such as nonsighted movement of the hand/arm to track a swinging pendulum and nonsighted tapping of the four corners of a square grid, interfere with spatial but not visual temporary storage tasks (Baddeley & Lieberman, 1980). In contrast, visual secondary tasks, such as viewing flickering dot patterns or abstract paintings, interfere with visual but not spatial temporary storage tasks (Quinn & McConnell, 1996). Using tasks designed to test the capacity of WM and testing children between the ages of 5 and 12 years, one developmental study (Logie & Pearson, 1997) showed the capacity of a visual short-term storage system developed at a faster rate than the capacity of a spatial system. This finding was replicated by Pickering and colleagues (Pickering, Gathercole, Hall, & Lloyd, 2001) in experiments comparing children between the ages of 5 and 10. Interestingly, as measured by a spatial delayed response task, memory for spatial location continues to develop until age 20 (Zald & Iacono, 1998).

Although the interference paradigms provide convincing evidence for the separability of visual and spatial short-term retention systems, the developmental results are more difficult to interpret. In the developmental studies that measured WM span (Logie & Pearson, 1997; Pickering et al., 2001), the spatial span task differed from the visual span task on a temporal/ dynamic dimension. In the visual task, participants had to recognize a static grid of filled and unfilled squares after a short retention interval, and span was measured by incrementally increasing the number of squares in the grid. The spatial task was similar, but instead of a simultaneous display of squares, filled squares were presented one at a time in different locations in the grid, and the participants remembered the locations in the order of their occurrence. Thus, spatial span has a strong temporal component, and it is unclear whether the delay in its development relative to visual span is due to difficulty remembering a set of locations or difficulty remembering the temporal order in which those locations occurred. Smyth and Scholey (1996) demonstrated that spatial span tasks show serial position effects similar to those shown by verbal span tasks, suggesting retention of temporal order may be subject to domain-general constraints.

Although visual and spatial components of WM can be dissociated in the laboratory, in natural cognition it is likely that the two are inextricably intertwined and work in concert with one another, unless damage to one system or the other disrupts their joint function. This conclusion is consistent with the extensive interconnectivity between the ventral and dorsal visual systems of the primate brain, which in turn indicates a high degree of cross talk and interactivity consistent with significant functional interdependence (e.g., DeYoe & Van Essen, 1988). We continue our review of visuospatial WM at a level of analysis that does not differentiate spatial from visual processing. This is largely because much of the research treats visuospatial WM as a unified system, but it is also based on the premise that the two subsystems normally work together as a single, highly integrated system. We only distinguish visual from spatial processing in subsequent sections when it is empirically or theoretically important to do so.

Separability of Storage and Executive Processes. In the verbal domain, complex memory span tasks that test both storage capacity and central executive functioning, such as Daneman and Carpenter’s (1980) reading span task, predict higher level verbal functions such as reading comprehension better than simple span tasks that test only storage capacity (Daneman & Merikle, 1996). The separability of on-line storage and executive functions in the verbal domain is illustrated by the differential predictive power of simple and complex span tasks, despite their high degree of correlation. Applying the terminology used by Miyake, Friedman, Rettinger, Shah, and Hegarty (2001), we call simple span tasks, which have only a storage requirement, short-term memory (STM)-span tasks and complex memory span tasks, which involve both storage and processing, WM span...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• Contributors
• Preface: Routes of Human Spatial Memory Research
• I Theoretical Issues in Remembering Where
• II The Task of Remembering “Where Is It?”
• III The Task of Remembering “Where Am I?”
• IV Remembering Where in Artificial Media and from Alternative Perspectives