DEVELOPMENTAL MEDICINE & CHILD NEUROLOGY

REVIEW

Separate visual systems for perception and action: a framework

for understanding cortical visual impairment
MELVYN A GOODALE
The Brain and Mind Institute, The University of Western Ontario, London, ON, Canada.
Correspondence to Professor Melvyn A Goodale, Room 205, Natural Sciences Centre, The University of Western Ontario, London, ON N6A 5B7, Canada. E-mail: mgoodale@uwo.ca

PUBLICATION DATA

Accepted for publication 4th April 2013.

ABBREVIATION

CVI

Cortical visual impairment

The visual control of skilled goal-directed movements requires transformations of incoming

visual information that are quite different from those required for visual perception. These
differences in the required computations have led to the emergence of dedicated visuomotor
modules in the dorsal visual stream of the cerebral cortex that are quite separate from the
networks in the ventral visual stream that mediate our conscious perception of the world.
Although the identification and selection of goal objects and an appropriate course of action
depends on the perceptual machinery of the ventral stream and associated cognitive modules
in the temporal and frontal lobes, the execution of the subsequent goal-directed action is
mediated by dedicated online control systems in the dorsal stream and associated motor
areas. This functional distinction can provide a useful framework for interpreting the range of
perceptual and visuomotor deficits observed in children with cortical visual impairment.

Traditional accounts of vision, from Helmholtz to Marr, have

assumed that the function of vision is to construct an internal
model of the external world, which then serves as a perceptual
foundation for all visually driven thought and action. However,
this monolithic account of visual function is almost certainly
incorrect. Indeed, a strong argument can be made that vision
evolved in vertebrates and other organisms, not to provide perception of the world per se, but to provide distal sensory control
of the movements that these organisms make in living their oftenprecarious lives. A long history of work across all vertebrate classes has shown that different patterns of visually guided behaviour
from prey-catching to obstacle avoidance are independently mediated by separate pathways that arise from the visual receptors in
the retina and project eventually to the motor nuclei in the
brainstem and spinal cord, each module processing a particular
constellation of inputs and each evoking a particular combination
of effector outputs.1,2
The brain mechanisms mediating visual perception were late
players on the evolutionary stage. With the emergence of cognitive systems and complex social behaviour, a good deal of motor
output had to become quite arbitrary with respect to sensory
input.3,4 Indeed, many vertebrates, particularly humans and other
primates, behave as though their actions are driven by some sort
of internal model of the world in which they live. The representational systems that use vision to generate such models or percepts
of the world have to perform different transformations on visual
input from those performed by the visuomotor modules described
earlier. Moreover, these perceptual systems are not linked directly
to specific motor outputs but access these outputs through cognitive systems involving memory and semantics. The term perception is being used here, not in the broad sense of any processing
of sensory input, but in the more restricted sense of assigning
meaning and significance to external objects and events. This use,
which is most common in philosophical and psychological writings, often (although not always) carries experiential connotations,

and tends to be identified with ones phenomenological experience of the world. It is fair to say, however, that the line of
demarcation between cognition and perception, and the role of
consciousness in both these processes, are matters of some debate
and can vary depending on the task. Nevertheless, it is perception that allows us to model the world beyond our bodies, to
share that experience with other members of our species, and to
plan a vast range of different actions with respect to objects and
events that we have identified. However, even though perception
allows us to choose a goal, the actual programming and control
of the execution of an action that is performed to fulfil that goal
is mediated by dedicated visuomotor modules that are not
dissimilar in principle from those found in lower vertebrates. In
other words, vision in humans and other primates (and perhaps
other animals) has two distinct but interacting functions: (1) the
perception of objects and their relations, which provides a
foundation for the organisms cognitive life, and (2) the control of
actions directed at (or with respect to) those objects. As we
shall see, the organization of the cerebral visual pathways in
higher primates reflects this fundamental distinction in visual
processing.3
The distinction between vision-for-perception and vision-foraction provides a useful framework for unpacking the myriad of
visual deficits that can be observed in children who have been
diagnosed with cortical visual impairment (CVI).57 Moreover, it
can help to explain why some visual abilities are more likely to be
affected by preterm birth than others.

TWO CORTICAL VISUAL SYSTEMS

Beyond V1 in the primate brain, the ascending visual pathways
project to a complex mosaic of higher-order visual areas.8,9
Despite the complexity of the interconnections between these different areas, two broad streams of projections from early visual
areas have been identified in the macaque monkey brain: a ventral
stream projecting eventually to the inferotemporal cortex and a

The Author. Developmental Medicine & Child Neurology 2013 Mac Keith Press, 55 (Suppl. 4): 912

DOI: 10.1111/dmcn.12299 9

dorsal stream projecting to the posterior parietal cortex.10 More

than a decade of neuroimaging studies in humans has made it
clear that the visual projections in the cerebral cortex of the
human brain are also separated into a ventral and dorsal stream,
in a manner that appears to be homologous with that seen in the
monkey.1113
Over 20 years ago, David Milner and I proposed that the distinction between vision-for-perception and vision-for-action could
be made onto these two pathways.4 We argued that the ventral
stream plays the major role in constructing the perceptual representation of the visual world and the objects within it, whereas
the dorsal stream mediates the visual control of actions directed
at those objects. Note that this is not the distinction between
what and where (object vision and spatial vision) that was originally put forward by Ungerleider and Mishkin, who proposed
that the ventral stream processes incoming information about the
intrinsic visual properties of an object whereas the dorsal stream
processes information about its spatial location.10 Our perspective
on the division of labour between the dorsal and ventral streams
is to place less emphasis on input distinctions (e.g. intrinsic
object qualities versus object location) and to take more account
of the output characteristics of the two cortical systems. We propose that the intrinsic properties of an object and its spatial location are processed by both streams, but for different purposes: in
the case of the ventral stream, for constructing a perceptual representation of the world, and in the case of the dorsal stream, for
the visual control of action. This is not to say that the distribution of visual inputs does not differ between the two streams, but
rather that the main difference lies in the nature of the transformations that each stream performs on those inputs. Of course,
the two streams are not hermetically sealed from one another.
They work together in controlling our behaviour as we live our
complex livesbut they play separate and complementary roles
in the production of adaptive behaviour.12 But what was our evidence for proposing this division of labour between the dorsal
and ventral streams?

NEUROPSYCHOLOGICAL EVIDENCE
It has been known for a long time that patients with lesions in
the dorsal stream, particularly in the intraparietal sulcus and the
superior occipito-parietal area, can have problems using vision to
direct a grasp or aiming movement towards the correct location
of a visual target placed in different positions in the contralesional
visual field, especially in the periphery.14,15 This deficit is often
called optic ataxia. But the failure to locate an object with the
hand should not be construed as a problem in spatial vision; many
of these patients, for example, can describe the relative position of
the object in space quite accurately, even though they cannot
direct their hand towards it.16 Moreover, sometimes the deficit
will be seen in one hand but not the other.14,15,17 It should be
pointed out, of course, that these patients typically have no difficulty using input from other sensory systems, such as proprioception or audition, to guide their movements. Some of these
patients are unable to use visual information to rotate their hand,
scale their grip, or configure their fingers properly when reaching
out to pick up an object, even though they have no difficulty
describing the orientation, size, or shape of objects in that part of
the visual field.1719 Clearly, a disorder of spatial vision fails to
capture this range of visuomotor impairments.10 Instead, this pattern of deficits is consistent with our proposal that the posterior
parietal cortex plays a critical role in the visual control of skilled
actions.4
10 Developmental Medicine & Child Neurology 2013, 55 (Suppl. 4): 912

What this paper adds

This paper describes a new framework for understanding the complex array of deficits associated with cortical visual impairment.

The opposite pattern of deficits and spared abilities can be

seen in patients with visual agnosia. Take the case of patient DF,
who developed a profound visual form agnosia after carbon monoxide poisoning.20 Although magnetic resonance imaging (MRI)
showed evidence of diffuse damage consistent with hypoxia, there
was clear evidence of bilateral focal lesions in the lateral occipital
cortex, an area in the ventral stream that has been implicated in
object recognition.21 Even though DFs low-level visual abilities
were reasonably intact, she could no longer recognize everyday
objects or the faces of her friends and relatives; nor could she
identify even the simplest of geometric shapes. If an object was
placed in her hand, she had no trouble identifying it by touch;
i.e. her deficit is visual in nature. Remarkably, however, DF
showed strikingly accurate guidance of her hand movements
when she attempted to pick up the very objects she could not
identify. Thus, when she reached out to grasp objects of different sizes, her hand opened wider mid-flight for larger objects
than it did for smaller ones, just like it does in people with normal vision.20 It has been recently suggested that DFs spared
ability to scale her hand to the width of the target object
depended more on haptic feedback than it did on visual processing.22 This account, however, cannot account for why she continued to scale well even when objects of different sizes were
randomly interleaved during testing.23
DF also rotated her hand and wrist quite normally when she
reached out to grasp objects placed in different orientations,20,24
and she configured her hand appropriately when picking up
objects with different shapes.18 At the same time, she was quite
unable to distinguish between any of these objects when they
were presented to her in simple discrimination tests. She even
failed in manual matching tasks in which she was asked to show
how wide an object was by opening her index finger and thumb a
corresponding amount.20 DFs spared visuomotor skills were not
limited to grasping. She could step over obstacles during locomotion as well as controls, even though her perceptual judgements
about the height these obstacles were far from normal.25 Contrary
to expectations based upon the ideas of Ungerleider and Mishkin,10 then, a profound loss of form perception coexisted in DF
with a preserved ability to use form in guiding a broad range of
actions. Such a dissociation, of course, is consistent with our proposal that there are separate neural pathways for transforming
incoming visual information for the perceptual representation of
the world and for the control of action.4
The pattern of visual deficits and spared abilities in DF and
other patients with visual form agnosia is in many ways the mirror image of that observed in the patients with dorsal-stream
damage described earlier. In other words, patients with damage
to the ventral stream can reach out and grasp objects whose
form and orientation they cannot perceive, whereas patients with
damage to their dorsal stream are unable to use vision to guide
their reaching and/or grasping movements to objects whose form
and orientation they perceive. Although this double dissociation
provided the initial, and perhaps most compelling, evidence for
the perception-action model that David Milner and I proposed,
the model is also supported by a wealth of anatomical, electrophysiological, and lesion studies in the monkey and a growing
body of human neuroimaging studies both in patients and in
neurotypical participants.12,26

IMPLICATIONS FOR UNDERSTANDING VISUAL

DEFICITS IN PRETERM INFANTS
CVI is often reported in children born extremely preterm, and
can be observed even when those children score within the normal range in intelligence tests.6 The nature of the deficits associated with CVI can vary from difficulties in visual perception to
problems with visuomotor control.57 In short, CVI is an
umbrella term that includes a broad range of possible visual deficits. The dissociation between vision-for-perception and visionfor-action that David Milner and I have identified can provide a
useful framework for interpreting the nature of the deficits associated with different kinds of brain damage in children with CVI.
Thus, children with damage in the occipito-temporal cortex
might be expected to have deficits perceptual processing, manifesting as deficits in object recognition, reading, and even form or
colour discrimination. In contrast, children with damage in the
occipito-parietal region might present with visuomotor dysfunction and/or deficits in the deployment of attention.
Consider the case of damage to the periventricular white matter.
This is one of the most common causes of visual deficits in the
developing brain, particularly in children born preterm.5 Moreover, damage typically occurs more often in the white matter adjacent to the posterior parietal cortex than in the occipito-temporal
cortex.27,28 As a consequence, the dorsal stream of visual processing tends to be more compromised than the ventral stream. Not
surprisingly, then, children who are born preterm and diagnosed
with white matter damage (from periventricular leukomalacia, for
example) commonly present with visuomotor deficitswhile their
ability to recognize objects and faces and to read can be relatively
intact.2932 Nevertheless, even though damage to the periventricular white matter serving the ventral stream is less common,5 some
children with periventricular leukomalacia do show typical ventralstream deficits, such as disturbances in reading and/or in the ability
to recognize objects, patterns, and faces.7 But although the

visuomotor and perceptual deficits are often dissociated in periventricular leukomalacia, it has been difficult to correlate the observed
pattern of deficits with morphometric data from MRI, probably
because damage to the periventricular white matter leads to abnormal functional development of visual areas in the dorsal and/or
ventral streams, changes that are difficult to track with structural
MRI.33 For this reason, techniques such as functional MRI, functional connectivity analyses, and tractography that can detect these
changes are much more likely to uncover the relation between
abnormal brain organization and behavioural dysfunction.6,3436
It is also worth noting that CVI is observed in children with
genetic developmental disorders, such as Williams and fragile X
syndromes. In both these syndromes, the visual and attentional
deficits would appear to be associated with dorsal-stream damage.37,38 The MRI morphometric studies certainly support this
conclusion,39,40 but, as is the case with children with periventricular leukomalacia, there have been few functional MRI or connectivity studies of these genetic disorders.38

CONCLUSION
In this brief review, I have tried to show how the distinction
between vision-for-perception and vision-for-action and the differences in their underlying neural substrates can help to unpack
the complex array of visual deficits that characterize CVI. Vision
is not a monolithic entity and quite different visual deficits can
arise depending on the locus and nature of the abnormalities in
brain development that can arise from preterm birth and various
genetic disorders. Understanding this point is a critical first step
in the construction of useful diagnostic criteria and effective pediatric rehabilitation.

CONFLICTS OF INTEREST
The author has no conflicts of interest.

REFERENCES
1. Goodale MA. Vision as a sensorimotor system. In: Robinson TE, editor. Behavioral Approaches to Brain Research.
New York, NY: Oxford University Press, 1983: 4161.

11. Culham JC, Valyear KF. Human parietal cortex in

action. Curr Opin Neurobiol 2006; 16: 20512.
12. Goodale MA. Transforming vision into action. Vision

20. Goodale MA, Milner AD, Jakobson LS, Carey DP. A

neurological dissociation between perceiving objects and
grasping them. Nature 1991; 349: 1546.

2. Goodale MA. Visuomotor modules in the vertebrate

brain. Can J Physiol Pharmacol 1996; 74: 390400.
3. Milner AD, Goodale MA. The Visual Brain in Action.
2nd edn. Oxford: Oxford University Press, 2006.
4. Goodale MA, Milner AD. Separate visual pathways for
perception and action. Trends Neurosci 1992; 15: 205.

Res 2011; 51: 156787.

13. Tootell RBH, Tsao D, Vanduffel W. Neuroimaging
weighs in: humans meet macaques in primate visual
cortex. J Neurosci 2003; 23: 39819.
14. B
alint R. Seelenlahmung des Schauens, optische Ataxie,
raumliche St
orung der Aufmerksamkeit. Monatsschr Psy-

21. James TW, Culham J, Humphrey GK, Milner AD,

Goodale MA. Ventral occipital lesions impair object recognition but not object-directed grasping: an fMRI
study. Brain 2003; 126: 246375.
22. Schenk T. No dissociation between perception and
action in patient DF when haptic feedback is withdrawn.

5. Jacobson LK, Dutton GN. Periventricular leukomalacia:

an important cause of visual and ocular motility dysfunction in children. Surv Ophthalmol 2000; 45: 113.
6. Lehman SS. Cortical visual impairment in children:
identification, evaluation and diagnosis. Curr Opin Ophthalmol 2012; 23: 3847.
7. Roman-Lantzy C. Cortical Visual Impairment: An

chiatr Neurol 1909; 25: 5181.

15. B
alint R, Harvey M. Psychic paralysis of gaze, optic
ataxia, and spatial disorder of attention. Cogn Neuropsychol 1995; 12: 26581.
16. Perenin M-T, Vighetto A. Optic ataxia: A specific disorder in visuomotor coordination. In: Hein A, Jeannerod
M, editors. Spatially Oriented Behavior. New York, NY:

J Neurosci 2012; 32: 20137.

23. Milner AD, Ganel T, Goodale MA. Does grasping in
patient D.F. depend on vision? Trends Cogn Sci 2012;
16: 2567.
24. Carey DP, Harvey M, Milner AD. Visuomotor sensitivity for shape and orientation in a patient with visual
form agnosia. Neuropsychologia 1996; 34: 32937.

Approach to Assessment and Intervention. New York,

NY: AFB Press, 2007.
8. Van Essen DC, Lewis JW, Drury HA, et al. Mapping
visual cortex in monkeys and humans using surfacebased atlases. Vision Res 2001; 41: 135978.
9. Goodale MA. The functional organization of the central

Springer, 1983: 30526.

17. Perenin MT, Vighetto A. Optic ataxia: a specific disruption in visuomotor mechanisms. I. Different aspects of
the deficit in reaching for objects. Brain 1988; 111: 643
74.
18. Goodale MA, Meenan JP, B
ulthoff HH, Nicolle DA,

25. Patla AE, Goodale MA. Obstacle avoidance during locomotion is unaffected in a patient with visual form agnosia. NeuroReport 1996; 8: 1658.
26. Kravitz DJ, Saleem KS, Baker CI, Mishkin M. A new
neural framework for visuospatial processing. Nat Rev
Neurosci 2011; 12: 21730.

visual pathways. In: Dutton GN, Bax M, editors. Visual

Impairment in Children Due to Damage to the Brain.
London: Mac Keith Press, 2010: 519.
10. Ungerleider LG, Mishkin M. Two cortical visual
systems. In: Ingle DJ, Goodale MA, Mansfield RJW,
editors. Analysis of Visual Behavior. Cambridge, MA:
MIT Press, 1982: 54986.

Murphy KS, Racicot CI. Separate neural pathways for

the visual analysis of object shape in perception and prehension. Curr Biol 1994; 4: 60410.
19. Jakobson LS, Archibald YM, Carey DP, Goodale MA. A
kinematic analysis of reaching and grasping movements
in a patient recovering from optic ataxia. Neuropsychologia
1991; 29: 8039.

27. Gronqvist S, Flodmark O, Tornqvist K, Edlund G,

Hellstrom A. Association between visual impairment and
functional and morphological cerebral abnormalities in
full-term children. Acta Ophthalmol Scand 2001; 79: 140
6.
28. Dutton GN, Jacobson LK. Cerebral visual impairment
in children. Semin Neonatol 2001; 6: 47785.

Separate Visual Systems for Perception and Action Melvyn A Goodale 11

29. Atkinson J, Braddick O. Visual and visuocognitive development in children born very prematurely. Prog Brain
Res 2007; 164: 12349.
30. Jakobson LS, Frisk V, Knight RM, Downie ALS, Whyte
H. The relationship between periventricular brain injury
and deficits in visual processing among extremely-low-

children with periventricular leukomalacia. Dev Med

Child Neurol 2009; 51: 97481.
34. Nagae LM, Hoon AH Jr, Stashinko E, et al. Diffusion
tensor imaging in children with periventricular leukomalacia: variability of injuries to white matter tracts.
AJNR Am J Neuroradiol 2007; 28: 121322.

characterize diffuse perinatal white matter injury: a case

report. J Child Neurol 2009; 24: 795800.
38. Atkinson J, Braddick O. From genes to brain development to phenotypic behavior: dorsal-stream vulnerability in relation to spatial cognition, attention, and
planning of actions in Williams syndrome (WS) and

birthweight (<1000 g) children. J Pediatr Psychol 2001;

26: 50312.
31. Santos A, Duret M, Mancini J, Gire C, Deruelle C. Preterm birth affects dorsal-stream functioning even after
age 6. Brain Cogn 2009; 69: 4904.
32. Seitz J, Jenni OG, Molinari L, Caflisch J, Largo RH.

35. Ortibus EL, De Cock PP, Lagae LG. Visual perception

in preterm children: what are we currently measuring?
Pediatr Neurol 2011; 45: 110.
36. Sie LTL, Rombouts SA, Valk IJ, Hart AA, Scheltens P,
van der Knaap MS. Functional MRI of visual cortex in
sedated 18 month-old infants with or without periven-

other developmental disorders. Prog Brain Res 2011;

189: 26183.
39. Hagerman RJ, Hagerman PJ, editors. Fragile X
Syndrome: Diagnosis, Treatment, and Research. 3rd edn.
Baltimore, MD: Johns Hopkins University Press, 2002.
40. Meyer-Lindenberg A, Mervis CB, Berman KF. Neural

Latal Hajnal B. Correlations between motor performance and cognitive functions in children born <1250 g
at school age. Neuropediatrics 2006; 37: 612.
33. Fazzi E, Bova S, Giovenzana A, Signorini S, Uggetti C,
Bianchi P. Cognitive visual dysfunctions in preterm

tricular leukomalacia. Dev Med Child Neurol 2001; 43:

48690.
37. Yeatman JD, Ben Shachar M, Bammer R, Feldman HM.
Using diffusion tensor imaging and fiber tracking to

mechanisms in Williams syndrome: a unique window to

genetic influences on cognition and behaviour. Nat Rev
Neurosci 2006; 7: 38093.

12 Developmental Medicine & Child Neurology 2013, 55 (Suppl. 4): 912