Search Results

This chapter discusses two distinct and quasi-separate brain systems, the dorsal and ventral visual streams. First, it uses the metaphor of two people singing from the same hymn sheet, and argues that the two visual streams necessarily sing from different, though in large part mutually consistent ones. It then considers evidence derived from experimental strategies to inquire into the nature of the hymn sheets from which each visual processing stream is singing. Clearly the hymn sheets have to be different. In order to provide direct control of one's movements, the dorsal stream has to see the world egocentrically, that is to code visual information in direct relation to the observer's body coordinates. In contrast, the ventral stream specifically needs to encode the world in a way that will be useful not only in the short but also in the long term, for which purpose egocentric coding would be useless.

This chapter reviews the recent literature on the organization of the visual system. New insights into the ventral and dorsal streams are explored. How the development and now wide use of fMRI help in recent research studies are also presented. The authors furthermore clarify their position on the theory about the function of the two main incoming streams of visual processing in the primate cortex. The role of the ventral streams and the role of attention in integrating the two streams in adaptive behaviour are reviewed. Further, the issue of the role of the dorsal stream and other structures in the acquisition of new motor skills is assessed here. The chapter ends with suggestions for future research on this area of study.

This chapter discusses the Atkinson and Braddick neurobiological models of early visual development in typically developing infants. The first model is based on the idea developed by Atkinson from Bronson that newborn vision depends largely on subcortical networks which come under the control of the visual cortex, which starts to function in the first six months of life. This model is developed further, describing human development of the dorsal cortical stream responsible for spatial processing, visual motion, and visual control of action; and ventral stream, subserving recognition of objects and faces. In the fuller model, the development of cortical modules for discrimination of orientation (slant), colour, motion, and stereo disparity (3D vision) is linked to developing systems for visual attention. These attention systems determine the information delivered to dorsal stream visuo-motor modules controlling gaze, manual actions, and locomotion at different stages during the first year of life.

This chapter provides an overview of the syndrome of unilateral spatial neglect (USN) as a multicomponent deficit, focusing on the distinction between its ‘perceptual’ and ‘premotor’ impairments, as well as on related evidence from the processing of visual illusions. It compares the USN syndrome, and its neural correlates with the two disorders representing the neuropsychological counterparts of vision-for-perception (the ‘ventral’ stream), and of vision-for-action (the ‘dorsal’ stream). From the neuropsychological vantage-point of USN, the chapter takes the view that the two visual streams dichotomy — both in the original version of Ungerleider and Mishkin (1982), and in the development by Milner and Goodale — captures only partially the neural loops concerned with perception and action in the visual domain. The syndrome of USN suggests the existence of a neural system supporting perceptual awareness in spatial reference frames, for vision, and for other sensory modalities, as well as goal-directed, intentional action in the space surrounding us. A third, dorsal-ventral, stream, including the inferior parietal lobule, and the ventral premotor cortex, may constitute the neural underpinnings of spatial awareness for perception and action.

The ‘two-visual’ systems hypothesis has recently come under attack regarding its proposed functional dichotomy between vision-for-action and vision-for-perception as well as for the limited interaction it allows between visual awareness and processing in the dorsal stream. Schenk (2006) questions the rigid functional dichotomy between vision-for-perception and vision-for-action, arguing that the dual model of vision is best accounted for in terms of a dissociation between egocentric and allocentric spatial coordinate systems. Wallhagen (2007) argues that there is no evidence to claim that the processing in the dorsal stream cannot underlie visual awareness. This chapter offers a response to both challenges and disentangles the contribution of two separable factors to the two-visual systems model, namely, how spatial information is coded and the relation between consciousness and processing in the ventral and dorsal streams respectively.

Theories differ in how they relate consciousness to action. At one extreme, enactivists argue that perceptual consciousness essential involves motor responses or a representation of motor action. At the other extreme, inactivists argue that systems associated with visually guided action are entirely unconscious. This chapter critically reviews the evidence for both extremes. It argues that motor responses are not necessary for consciousness or even consciously experience. But it also argues against the claim that perception divides into two streams, a conscious stream for recognition and an unconscious stream for action. In place of these views, the chapter argues that action is related to the function of consciousness, rather than the content of consciousness. Conscious states present a menu for action: intermediate-level representations present the world from an action-relevant point of view, and attention allows us to bring such representations into working memory for practical decision making.

The seventh chapter shows how the empirical premise can accommodate evidence for the dual visual systems in cognitive neuroscience. The main claim of this chapter is that the crucial difference between the two cortical streams is in their spatiotemporal processing, rather than their functional output: the dorsal stream processes peripheral retinal input with a high temporal resolution, and the ventral stream specializes in foveal input with less temporal resolution.

This chapter argues that the separation of the cortical visual processing into two streams is insufficient and, in the version where perception and action are kept separated, leads to a misunderstanding of the true nature of perceptual processes. It shows that the processing carried out in the inferior parietal lobule is different from that performed in the inferior parietal lobe, and that the so-called dorsal stream is in fact formed by two streams: the dorsodorsal stream (D-D) and the ventrodorsal stream (V-D). The chapter also discusses the relation between action and perception as it emerges from neurophysiological data on the V-D stream. It proposes that both action perception and space perception derive from a preceding motor knowledge based on self-generated actions.

Outlining possible suggestions in addressing questions related to visual and dorsal streams of visual processing is the main concern of this chapter. Specifically, it attempts to examine the internal functional organization of each of these streams and how they interact in the production of integrated patterns of behaviour. In addition, how the phenomena of visual attention and consciousness map onto the two streams and how the puzzling syndrome of hemispatial neglect fits into this account of cortical visual processing are also assessed. The chapter begins with a discussion on attention and consciousness. Subsequently, physiological studies of visual attention in the dorsal and ventral streams are presented and it is shown that networks of cells in both ventral and dorsal systems appear to participate in spatial attention. Two well-known deficits commonly thought to reflect disorders of spatial attention are also explored: hemispatial neglect and visual extinction.

The thesis of the cognitive penetration of vision asserts a specific type of informational exchange between cognition and vision. It remains unclear, however, what this amounts to, whether cognition affects vision in this way, and even what counts as evidence for it. This chapter asks: Is visually guided action cognitively penetrated? Specifically, it focuses on the possible penetration of dorsal visual stream computations by semantic/conceptual representations of the function and purpose of the use of objects. On this view, conceptual representations are used in the computations needed for visual guidance of movement. The chapter provides a clear account of cognitive penetration of dorsal visual stream computations that is conceptually defensible and plausible in the context of the anatomy of inputs to regions of the brain associated with the dorsal visual pathway. The chapter concludes by suggesting a specific hypothesis about how concepts, hence cognition, penetrate dorsal visual stream computations for action.

The fronto-occipital fasciculus (FOF), also known as the occipitofrontal fasciculus, is one of the long association systems of the dorsal visual stream. The subcallosal fasciculus of Muratoff that links the cerebral cortex with the caudate nucleus was mistaken for the FOF, and this conceptual and terminological confusion continues to the present day. This chapter begins with historical accounts of the FOF and Muratoff bundles. It then presents the results of the investigation of the FOF of rhesus monkey brain. Observations confirm the existence of the FOF where Dejerine located it in the human, and provide compelling evidence that it is a true association fasciculus linking parieto-occipital regions with the dorsolateral premotor and prefrontal areas. The present study also adds detail to the understanding of its location and to the origin and termination of its fibers.

This chapter summarizes the neuroscientific model of visual development that emerges from the work discussed in the book, including the role of dorsal and ventral cortical streams, and considers whether there is differential vulnerability of these visual brain systems. The book suggests new ideas and areas of future study in asking whether we can understand what visual experience is like for the young infant, and the role of ‘conscious’ experience and control in infant visual development. The chapter considers variation and anomalies in visual development, and asks how we can define ‘visual disability’ for the developing child. The need to consider multiple levels of analysis in understanding visual development and its disorders is discussed. The need is highlighted for future interdisciplinary studies, combining expertise across many disciplines and levels of analysis, to improve our understanding of the visual world of the child and the brain mechanisms underpinning this development.

In this chapter the authors argue that the relevant computations for the control of spatially directed movements are carried out by neural systems that are quite independent of those underlying much of what is called ‘spatial perception’. They propose that many visuomotor transformations rely on algorithms embodied in circuitry within the dorsal stream, whereas circuitry in the ventral stream may support the visual processing of spatial information to be used for purposes of perception and cognitive manipulation. Since actions such as grasping are clearly directed at objects rather than at disembodied locations, additional information about the objects, such as their size, shape, and local orientation must also form part of the required set of visuomotor transformations underlying the control of the constituent movements. This chapter examines the behavioural and neuropsychological evidence for such proposals and attempts to draw functional limits on the two visual processing streams.

This chapter provides background on the genetic profile of people with WS, followed by a review of what is known about the possible structural and functional brain bases for the spatial deficit. Current findings indicate regions of abnormal structure and function in the WS brain that include parts of the dorsal stream, specifically, areas of the parietal cortex, as well as the hippocampus and adjacent regions. This review is followed by a detailed analysis of the hallmark block construction task, emphasizing the many cognitive capacities that are engaged when people carry out such tasks. We close the chapter by discussing the hypothesis that the WS spatial deficit may be characterized by strength in those spatial functions normally engaging the ventral stream of the visual system, and weakness in those functions normally engaging the dorsal stream. This hypothesis is evaluated and progressively modified in Chapters 3, 4, and 5.

This chapter reviews the development of the ability to segment the visual field and group elements into perceived objects. Infants can be shown to segment on the basis of ‘texton’ patterns, texture orientation, and motion. They respond to aligned terminators forming illusory contours, another grouping process. Integrative processes are measured by coherence sensitivity to patterns of global motion or form, where coherence levels are varied using patterns containing a controlled proportion of randomly rather than coherently arranged elements. Adult fMRI studies from the Visual Development Unit show distinct cortical areas activated by form and motion coherence. Results from the use of novel Visual Development Unit computer games, measuring relative sensitivity to form and motion coherence, show the differential development throughout childhood of dorsal and ventral stream global processing. Children's developing ability on block construction copying tasks is also discussed, revealing the integration of local and global information under the control of visual attention.

The visual guidance of bodily motion is conducted by a system, here entitled ‘motion-guiding vision‘, that is separate from the system that furnishes us with visual qualia. The latter is called ‘descriptive vision‘. Vision scientists commonly hold that motion-guiding vision makes no contribution to sensory consciousness. Here it is argued that by giving us the wherewithal physically to make contact with external objects, motion-guiding vision supports perceptual demonstratives and accounts for the ‘feeling of presence‘ that distinguishes seen physical objects from those that are merely pictured or imagined. In this way, motion-guiding vision contributes to our experience of seeing something, though sensory qualia such as colour or shape cannot be traced to it. The visual representation of space is assembled from descriptive and motion-guiding vision.

Damage to the primary visual cortex can cause a profound loss of awareness of the visible world. However, surviving pathways appear to provide sufficient visual information to allow the brain to mediate a range of visual tasks in the absence of such awareness. In this chapter, the authors discuss how ‘blindsight’ should not be characterized as ‘unconscious perception’. Instead, it should be more correctly seen as a collection of residual visuomotor responses that may depend on a variety of relatively independent circuits in the superion colliculus and dorsal stream. Further, this chapter contends that ‘blindsight’ only becomes paradoxical if one regards vision as a unitary process. If it is accepted that the mechanisms capable of providing perceptual experience are separate from those underlying the visual control of action, the paradox disappears.

The evidence for the coding of visual location in multiple ways within the dorsal stream, for different motor purposes, is becoming increasingly clear. But all of these forms of spatial coding, of course, would only be of value over rather short time spans, since every time the animal moved its body, the usefulness of the coding would be lost. In other words, any form of egocentric spatial coding would be useful for guiding action in the present, but not for storing spatial information for use very far in the future. It would provide useful information about object location for calibrating the amplitude and direction of an immediate movement of the eye, head, or limb, but not for the long-term storage of information about the relative location of that object vis-a-vis other objects in the world.

This chapter describes how visual information is transformed into motor acts. It examines the concept of two visual systems and illustrates that the dorsal stream is actually formed by two basically independent pathways. It also considers the issue of the role of different visual streams in perception. This chapter shows that V6 is an extrastriate area at the origin of a relatively direct visuomotor pathway, which conveys visual information to caudal superior parietal lobule (SPL) areas. It suggests that the two dorsal streams play a differential role in the organization of visuomotor behavior. It also demonstrates that the actions and space organization are the building blocks on which perception is built.

This chapter discusses the origins of vision. In evolutionary terms, vision began as a system for the sensory control of movement. Vision as conscious sight is a relative newcomer on the evolutionary landscape. Work on the visual systems of different animals from frogs to monkeys leads to the conclusion that there are many independent visuomotor ‘modules’ and these modules are separate from the visual pathways that provide our perceptual representations or ‘models’ of the world. The two-visual-streams proposal is described.