The Hippocampal and Parietal Foundations of Spatial Cognition

The Hippocampal and Parietal Foundations of Spatial Cognition

Burgess N, Jeffery KJ, O’Keefe J, editors

Oxford: Oxford University Press; 1999. 490 pp. with index

ISBN 0-19-852452-8 (paper)

This well-organized volume has much of interest to basic researchers. The 3 roughly equal sections discuss, first, the parietal cortex, then the hippocampal formation and, finally, the interaction of the two in spatial learning and memory. Studies of rats, monkeys and humans are included. Among the techniques reviewed are behavioural, neuropsychological, electrophysiological, brain imaging and neural net computational modeling. Reading this book from cover to cover reveals a lot of repetition of similar material, such as anatomical details. It is often surprising to read about the same ideas in different chapters that do not cross-reference each other. However, for the reader who uses this book as a source of specific information about cortical regional specialization of function, the chapters provide independent and complete reviews of the latest relevant work.

The sophisticated studies of Milner and Goodale, identifying 2 streams of visual information processing with different functions, provide a useful basis for integrating much of the work presented in the first section on the parietal cortex and some of the studies presented in the third section on integration of parietal and hippocampal function. Goodale et al have argued that the dorsal stream of visual processing into the superior parietal lobe is concerned with the use of that information for the generation of motor actions; damage to this area leads to optic ataxia. The ventral stream projecting into the inferior parietal lobe and dorsal temporal cortex is concerned with what is being seen; damage here leads to spatial neglect. The chapter by Karnath, for example, presents results of studies showing that patients with damage to the right parietal cortex made exploratory eye movements consistent with an ipsilesional deviation of egocentric space representation. The same patients showed no deficit in goal-directed arm movements to targets around them. These findings are consistent with the idea that there are 2 streams of processing and that the damage in the patients studied affected the ventral but not the dorsal stream.

Colby presents fascinating electrophysiological data recorded in the ventral intraparietal area in monkeys. Cells were found to be responsive to both visual and so-matosensory stimuli; neurons with foveal visual receptive fields had somatosensory receptive fields on or around the muzzle. It was as if the mouth was the “fovea” of the facial somatosensory system! Furthermore, visual receptive fields moved across the retina in order to maintain spatial correspondence with somatosensory fields, suggesting that stimuli are coded in a head-centred reference frame. Patients with parietal cortical damage can be seen to suffer from a deficit in updating spatial representations for use by the motor system. Colby suggests that the remapping of visual fields observed in parietal cortical neurons provides the substrate for this updating.

In coaching students preparing for comprehensive examinations, one of my colleagues often counsels them to identify landmark papers that open whole new areas of investigation. Two such works in Spatial Cognition are Scoville and Milner’s paper and O’Keefe and Nadal’s 1978 book The Hippocampus as a Cognitive Map. The former is the first report of a role for the hippocampus in recent memory, and the latter identifies the place specificity of hippocampal cells. Juxtaposition of the ideas from these 2 classic works influences much of the thinking in Spatial Cognition. Some interesting ideas linking recent memory and place specificity can be found.

Mishkin et al, for example, discuss episodic and semantic memory. When a new item, association or fact is being encoded into memory, the relevant sensory information arrives as an episode that includes spatial information, as well as temporal cues and information about emotional and mental states. The amount of contextual information that is retained determines the nature of the stored memory: con-textually rich memories include spatial and other information and are, therefore, episodic; contextually poor memories record only the facts, and therefore constitute semantic memory. From this point of view, semantic memory is lower in a hierarchy of mnemonic sophistication than episodic memory. Mishkin et al argue that the hippocampus is necessary for episodic but not for semantic memory. Spatial information would be intrinsic to episodic memory. Rolls, Gaffan and Hornak make similar arguments in later chapters.

One distinction that guides some of the discussion of the interactions between parietal and hippocampal systems in space and memory is that between allocentric and egocentric frames of reference. Egocentric reference is putatively mediated by parietal circuits, and allocentric reference by hippocam-pal circuits. Rolls reported that 46% of the spatial cells of the hippocampus represented space in allocentric coordinates, versus 10% that, by comparison, were egocentric. Maguire reported that positron-emission tomographic (PET) images of people who walked mentally along a recently learned spatial route showed right hippocampal activation, suggesting that this region provides an allocentric representation of space. Parietal cortical regions seemed to play a role in egocentric movements through environments.

Another work that strongly influences many authors writing in Spatial Cognition and could be added to the comprehensive reading list that I mentioned earlier is The Visual Brain in Action by Milner and Goodale. This is the source of the idea, mentioned above, that the dorsal stream guides visuomotor actions and the ventral stream identification of what is seen. In the final chapter, Milner et al suggest that, if a participant was required to perform a delayed motor act, accurate performance would depend on the ventral stream because the egocentric coordinates that are tracked by the dorsal stream will have changed during the delay (assuming the participant moves). Thus, visually guided motor acts like pointing should be impaired after a delay in people with damage to the ventral stream. Results supported this conclusion.

One interesting contrast that I found in Spatial Cognition was between Rolls’ and Maguire’s view of imaging studies. Rolls reported the results of electrophysiological studies in rats and monkeys showing that hippocampal cells in rats were place cells, responding when the rat was in that place; in monkeys, hippocampal cells fired when the monkey looked to a particular place, even if it didn’t go there. Rolls argues that imaging studies could not make this distinction; they did not provide a full description of what was being represented in the brain. Maguire reports differences in regional activation assessed by PET imaging, as described above, and concludes that PET offers a means to pursue many outstanding questions in understanding neuronal control of spatial cognition. I suspect that they are both right. The breadth of techniques reported in this book and the emerging clarity of the knowledge about the cognitive functions of these brain regions attests to the power of multiple empirical approaches to the study of the brain for discovering the mechanisms underlying the amazing abilities of this structure.