Psyche 9(11): 'Looking and Seeing with the Mind's I, and its Brain

1 view
Skip to first unread message

patrick wilken

Apr 26, 2003, 3:40:55 PM4/26/03


Tony Dickinson
Dept. of Anatomy & Neurobiology
Washington University School of Medicine
660 South Euclid Avenue
Saint Louis, MO 63110

Copyright (c) Anthony Dickinson 2003

PSYCHE, 9(11), April 2003

KEYWORDS: Visual processing, Attention, Frontal cortex, Visual
cortex, Sensory Systems, Systems neuroscience.

REVIEW OF: Braun, J., Koch, C. & Davis, J.L. (Eds.) (2001). *Visual
attention and cortical circuits.* ISBN: 0262024934. $60 hbk. 344 pp.
MIT Press: Cambridge, MA.

ABSTRACT: Contributions to this edited volume argue for the existence
of top-down, context- and task-dependant modulating mechanisms of
attention occurring in the mammalian brain. Such positions support the
view that areas of the brain traditionally thought to be involved in
relatively 'late' stages of visual processing activity (typically
frontal cortex) can, and do affect the response properties of 'early'
visual processing neurons, including primary visual cortex (V1).
Neural circuitries concerned with the processing of visual information
should now be viewed less as involving unidirectional mappings from
sensory input to motor output for the purposes of planning
visually-guided movements. Instead, the attentional processes required
to support the co-ordination of sensorimotor transformation functions
involve a variety of widely distributed parallel and reciprocally
connected neural pathways, including the visual, parietal and frontal

Aimed principally at the student and researcher working in cognitive
neuroscience, this book reports the findings of the "Visual attention
and cortical circuits" workshop held on Catalina Island, USA, in 1999.
Its primary focus was to bring together a variety of interdisciplinary
approaches to inform a better understanding of visual attention
processing in the cerebral cortex. Some fourteen contributions are
collected here, addressing the long standing implications of dorsal
versus ventral stream visual processing, and, of more recent interest,
the functional significance of the (often reciprocal) connections now
known to exist between temporal, parietal and frontal cortical neurons.
Appropriately illustrated throughout with task paradigms and
experimental data presentations, it is perhaps surprising that this
volume contains only one putative cortical circuit diagram [Tsotsos et
al.] in an attempt to show the ways in which variously proposed cortical
areas might be critically connected in support of their role in visual
attention and/or its modulation.

Although it has been clear since the demonstration of Yarbus (1967) that
we employ quite different brain circuitry when 'looking for', as opposed
to 'looking at' something in our visual field, I have never felt that
the physiological significance of this observation has really been
considered prior to single-cell recording from the awake, behaving
monkey. At least nine of the contributions to this volume (both imagers
and electrophysiologists, from human and monkey labs) explicitly argue
for the existence of top-down, context-dependant, task effects of
attention. One way or another, this amounts to claiming that at least
some 'late' (typically frontal cortex) visual processing activity can be
shown to be affecting the response properties of 'early' visual
processing neurons, including primary visual cortex (V1), V2-V4, and
extrastriate areas MT & MST. One clearly emerging story to be taken away
from this book, is that traditional claims for the visuomotor system
operating largely via unidirectional, monolithic 'Retina > RGN > V1 >
V2... > frontal cortex' pathway models must be discarded. Recurrent,
massively parallel cortical circuits are the order of the day here.

Throughout its middle six chapters, the results of visual attention
experiments using monkey single-cell recording is variously interpreted
as providing evidence for the biasing of response competition amongst
early visual input neurons [Duncan; Reynolds & Desimone], their possible
gating [Heeger et al; Tsotsos et al] and other modulations of their
activity [Ito et al; Maunsell & MacAdams]. Several authors have now
expressed support for Schall's notion of the frontal eye-field area
potentially serving as a task-based "saliency map" for the purpose of
supporting both the selection and preparation of visually-guided action
[Thompson et al] and a similar model is proposed to underlay
visually-guided search [Sperling et al]. Others rightly encourage our
caution lest we forget the necessarily constraining architectures of
bottom-up processing, upon top-down operating pathways [Braun et al;
Pouget et al].

The problem space for attentional research has always been one largely
concerned with determining how the cortex selects and locates targets
from a potentially infinite candidate array of such targets for focused
consideration. Furthermore, attention experiments must be conducted in
the face of limited processing capacity and with reference to one's
ontogenetic life-history of experience and learning with any number and
variety of tasks. There are no surprises amongst the results presented
in this book emerging from the studies of (visual) attention as
reported, but it does provide for a useful review as to some of the
current thinking 'outside the box' of the old monolithic pathways.
However, and more importantly, it also warns that we must continue to
explore the neural bases of behavior bearing in mind that the subject's
task understanding, and the context(s) in which their tasks are
presented, will necessarily affect the very cortical circuitries we are
attempting to characterize. Fortunately, this holds true as much for the
researcher's brains, as it does for the brains of their research
subjects, and for the fine details and anatomy of the neural circuits
themselves, we must be content to await the reports of future workshops.

Reply all
Reply to author
0 new messages