What is a picture? We may be inclined to say that it is a visual representation of part of the visible world. From this we might conclude that a blind person could not understand or create a picture. Not so. Raised outline drawings -- created simply and cheaply by pressing a ball point pen on plastic sheets -- can be perceived both visually and haptically. (Following Gibson, haptics is defined as "the sensibility of the individual to the world adjacent to his body, by the use of the body." (8)) Pictures are not essentially visual.
Kennedy's explorations and experiments with the understanding and production of raised outline drawings by the blind have practical implications for these persons, many of whom have been told that they could never understand or make pictures. In fact they do have a natural capacity to understand and produce such drawings. Thus with a little practice they can engage in useful communication by means of pictures, diagrams and maps. Kennedy's compassion and sensitivity are evident throughout the book. One suspects these aided him in the difficult task of soliciting, at statistically significant levels, the cooperation of the blind in his experiments.
Our concern here, however, will be with the deep theoretical issues that Kennedy raises about our perception and representation of objects, especially their spatial features and relations, as well as, more generally, our ability to represent and communicate ideas. "My task in this work has been to understand space, depiction, direction, and communication -- especially tactual space, outline, perspective, and metaphor." (293) Let us begin with two examples to illustrate the importance of the project.
Do we live and move and have our being in utterly different, unconnected spaces -- the visual, auditory, olfactory, haptic, perhaps even gustatory, spaces? Or is our lebenswelt that of one space of objects accessed through different perceptual modalities with a common structure? The debate goes back to Molyneux, Locke, Reed, Berkeley, Condillac, Lotze and Kant. Kennedy's studies support the view of Diderot that the blind can recognize things from drawings, and that the blind can haptically perceive and represent such phenomena as occlusions and perspective, which are often thought of as exclusively visual. He suggests that at some level of perceptual processing there arises the concept of one common space that has a specific structure, that is accessible through different modalities, and that corresponds to a real space of objects. Furthermore, he suggests that this real space can be represented in a variety of ways, including two-dimensional outline drawings that are accessible both visually and haptically.
Take a second example, which would be of interest to semioticians concerned with tropes such as prolepsis, synecdoche, hyperbole, metonymy, and so on. It appears that one can create haptic metaphors, just as one can create verbal -- "He has a heart of stone" -- and visual ones -- a cartoon of a British bulldog glowering at a jackbooted stormtrooper. Thus haptic representations can go beyond literal here and now depictions to have some of the most important general powers involved in thought and communication.
Kennedy practises psychology from within a particular epistemological and metaphysical framework. It is essentially that of his mentor, J.J. Gibson. The hypothesis is that our perceptual systems evolved as detectors of useful properties of our surroundings. Our perceptions enable us to determine the geometry of flat and curved surfaces -- their corners, vertices, edges and boundaries -- and thus to come to know a world of objects that exist independently of perceivers. Although cultural and technological changes bring an abundance of new kinds of images to the senses and new modes of representation for understanding and communication, they "capitalize on what evolution has required of observers for eons." (2) The framework might be characterized as common sense realism in an evolutionary setting.
It is with the details of basic perceptual and representational capacities and their ontogenetic unfolding that Kennedy is concerned. While rejecting theories that emphasize or make central sensationalism (with its pure Given), formal Gestalt properties, social conventions, and the extreme relativism of deconstructionism, Kennedy is not illiberal, and gives them their due when appropriate. He is interested in the physiology of the sense organs and cerebral mechanisms, but the book is full of phenomenological descriptions -- how things appear to us -- and intentionality is amply recognized: "...human beings are semantic engines, not automata." (8) Initial encoding, pace the work of Tulving, does not wholly determine meaning; we have the capacity to extract, create, and manipulate meanings.
Kennedy is suitably cautious in stating when a result has been firmly established, when further work needs to be done, when a suggestion is speculative, and when alternative theories have useful qualifications to offer. As a good scientist he meticulously records his experimental procedures, the nature of his subjects (e.g., categorizing them as, for instance, sighted but blindfolded, early blind, late blind, etc.), and the statistical significance of his results. On the whole I found this caution and documentation reassuring, and I shall pass over the details. The line of argumentation and experimentation is quite complex, consisting of an interplay of theories and experiments that confirm or refute them, an interweaving of theories of perception and theories of representation, and a consideration of the similarities of the visual and the haptic. In summarizing his work, I shall not follow this tortuous path, but rather will focus on its results and some important controversies.
First a metaphysical issue. Within Kennedy's realistic framework, perception consists in a series of events linking: object, medium, receptor system, the nerves, brain reception, cognition and motor areas, and the actions that change the perceiver's vantage point. (8) He speculates that, after initial processing in specialized sensory areas of the brain, signals are sent to the parietal area.
There the set of principles about shape in general, rather than inherently visual shape, could be readily accessed by touch and to some extent by audition, pain, smell and taste as these attempt to discover the spatial arrangement of sensory sources in the environment and the body. (49)The chain of events, thus described, does not include experience or conscious awareness. He might therefore seem to be a physicalist. Yet the bulk of his book and his experiments are not about the physics and physiology of perception. And he cautions us:
There are three distinctions to bear in mind: events at different energy levels, events making different kinds of patterns, and events that are or are not guided by intention to communicate. Mixing them up can cause havoc in a theory of perception. ...Pattern perception aims to detect invariant shape despite changes in energy level....it detects what is there, eliminating any irrelevant changes in energy levels....Many different kinds of principles can govern [pattern recognition]: the principles of geometry, botany, perspective, geology, astronomy and so on. The list involves major categories of existing things, each having its own rules for form -- rules that allow certain kinds of shape to exist, to continue, to be stable, and to change. Perception involves getting to know these rules...there are events that inherently involve a mind. (18-19, italics mine)It would appear that Kennedy implicitly subscribes to the criterion of the mental commonly assumed in cognitive psychology: intentionality (in the sense of representation -- x is "of" or "about" y -- and/or in the sense of goal or purpose) is the mark of the mental. I would not dispute this as one important criterion. However, there is another, submerged, denied or denigrated during the decades of the ascendency of behaviourism and physicalism, but now coming to the fore again: conscious awareness. Oliver Sacks, the doyen of the deficit in brain damaged patients, had some deep things to say about the remarkable contractions and expansions of subjective space and time experienced by his patients in his discussion of "Parkinsonian Space and Time" in Awakenings. And, in his discussion of the "Case of the Colorblind Painter" in An Anthropologist On Mars, he puts his finger on a fundamental, unresolved issue:
On the ultimate question -- the question of qualia: why a particular sensation may be perceived as red -- the case of Jonathan I. may not be able to help us at all. After describing "the celebrated phaenomenon of colours," Newton drew back from all speculation about sensation and would hazard no hypothesis as to "by what modes or actions light produceth in our minds the phantasms of colours." Three centuries later, we still have no hypothesis, and perhaps such questions can never be answered at all. (41)At the risk of being labelled a "new mysterian," to use the phrase of Owen Flanagan (Flanagan, 1991), I suggest that Sacks may be right: after tracing the series of events involved in perception to whatever levels in the brain, we are left with a feature that is different from the precedingevents: our conscious experience of space and time, which normally forms a unified manifold. It is puzzling that aspects of that unity -- edges, movement, colour, for instance -- are processed in different parts of the brain (V1, V2 and so on) at different objective times, and that there seems to be no hard evidence of an objective site or a time where or when they are all brought together. As Daniel Dennett, who denies the existence of any "Cartesian Theatre" of consciousness, says in Consciousness Explained, "The representation of space in the brain does not always use space-in-the-brain to represent space, and the representation of time in the brain does not always use time-in-the-brain." (131) It does seem wrong to expect complete isomorphisms between brain events and phenomenal experience. Perhaps the best we can do is to discover neural and other events that are necessary conditions for experience. It may be that we are left with a form of emergentism that is compatible with a form of physicalism. (Seager, 1991) I shall now leave these philosophical issues to turn to Kennedy's experiments in empirical psychology.
It is important to note that Kennedy works with outline drawings. The pictures make no use of colour or chiaroscuro (shading or coloured areas) -- features not detectable by touch. Sighted persons can recognize chiaroscuro pictures, say pure light and shadow pictures of a face. Interestingly, they cannot do so when the boundaries between light and shade are replaced by lines following the same contours. It seems that shape from shading is determined by a process different from that used in shape from outline. Outline drawings are sketchy and schematic, which makes them useful for generic depictions. In any case, outline drawings put the sighted and the blind more or less on the same footing.
In Chapter Three Kennedy presents evidence from his work and the work of others that blind people can recognize pictures haptically. The pictures range from a simple drawing of a hand to fairly complex scenes and illustrated stories. Often the performance of blindfolded sighted subjects is used as a benchmark in judging the performance of various groups of the blind. The conclusion is "...that the blind have a capacity to understand pictures, but there is still much to explore that may improve recognition rates and indicate the basis for the capacity." (73)
A key issue is whether this capacity is independent of training in a convention. Although practice and training can bring about recognitional improvements, it seems that there is some native capacity for picture recognition, whether the modality be visual or haptic. Kennedy and Ross found that the Songe people of Papua New Guinea, who have no pictures in their indigenous culture, could quickly recognize, visually, a highly schematic outline drawing of a person. Similarly, the blind can recognize, haptically, outline drawings without specific instruction or appeal to learned conventions.
In Chapter Four Kennedy turns to the issue of whether the blind can produce pictures. He demonstrates that they can, using the same outline system that governs their recognition of haptic pictures. Furthermore, he shows "...that the patterns of lines produced by the blind are appropriate rough copies of the shapes of objects, demonstrating how the parts of objects are laid out with respect to each other and the observer's vantage point." (95)
The drawing skills shown by blind adults attempting drawings for the first time are midway between the unrecognizable marks of the sighted child making drawings for the first time and the unmistakable accurate renderings of an adult draftsman aiming to be unambiguous and precise. (125)Even if there is some innate capacity to understand and create drawings, the question arises, does the capacity develop or unfold according to a specific pattern or timetable? For example, does the development of haptic representations among blind children follow roughly the same path as the development of visual representations among sighted children? The issue is a complex one, and there are at least two sorts of theories that have been brought to bear on it.
The skills they show are products of understandings they possess as a result of abilities with line elements and general spatial principles that arise from the constitution of the perceptual systems and the general appreciation of the space around them available to several senses. (126)
First, a representation such as a line drawing may, but need not be, iconic. (This is my terminology, not Kennedy's.) "The cat is on the mat" is composed of lines and represents a cat on a mat, but it is non-iconic: there is no resemblance between the representation and that which is represented. On the other hand, the representation could have been a drawing of a cat (eyes, ears, whiskers and all) superimposed on a rhomboid shape. This is iconic: there is at the perceptual level a resemblance between the representation and the represented. All this is obvious. The complication is that an outline drawing need not be entirely iconic and literal; it may be quasi-iconic. Consider a drawing by Kathy, an early blind girl. It shows a side view of a crib, with its characteristic side bars. Is this an iconic drawing? Well, surrounding the crib she drew a stylized heart. It takes little imagination to realize that the drawing is "of" or "about" neither cribs nor hearts. She was depicting a loved child, using iconic metaphor. Note, however, that unless she intended, and we understood, that the set of vertical and horizontal lines was an iconic representation of a crib, we would not get the metaphorical extension to a child and thus would miss the whole point. To be sure, she and we also have to understand the curved line as a representation of a heart (which it does not closely resemble) and that it is a conventional metaphor for love. These complexities make it difficult to decide between two sorts of theories about development.
According to one theory very young children start out drawing as a purely expressive motoric activity.
Accidental discoveries are made repeatedly by the child, and in time mastery of a dictionary of schemas develops. Some of the schemas are copied deliberately from pictures made by other people, especially other children. With time, the child invents more abstract principles, including aspects of perspective such as convergence. (132)According to another theory, which Kennedy thinks is "probably closer to the truth," the stress is on the abstraction of features considered relevant and the will to intentional representations right from the outset. This is not to deny that copying and learning conventions also enter into the matter, but they are secondary.
From a survey and classification of almost 2 000 drawings of a cube by sighted persons of various ages, Kennedy, Nicolls et al. concluded that there are some general principles involved in drawing development. The youngest children chiefly drew a single aspect of the cube following the principle of matching a feature. Somewhat older children added more features by drawing the cube in fold out style. Still older children (mean age 10) and adults restricted themselves to the features that could be observed from a specific vantage point and showed connections, proportions and angles for perspective. It is true that many sighted children are taught formulae and are given rules and expectations, although that is not the whole or most important part of the story. What of blind children? The evidence indicates that they too have a basic ability that gradually manifests itself in a similar fashion.
Blind children are not taught to draw. They are not given formulae by their teachers (or schoolmates) and do not aquire them as prototypes, like Snoopy in the Peanuts strip, as sighted children do. Rather, they seem to be driven by some developing ability that grows and burgeons unknown and unsuspected in all blind children. It only needs tapping to make itself evident. (p.179)I now turn to two intriguing matters to which Kennedy devotes a chapter each: haptic perspective and haptic metaphor.
It is obvious, but never to be forgotten, that outline drawings are two-dimensional representations of a three-dimensional world. They are not replicas, as say a sculpture is. Save for exceptions such as "exploded" drawings, which we have noted feature-bound children using at a certain stage, drawings are of objects from a vantage point and can depict depth and perspective. How do they make these representations? Do they depend on arbitrary conventions, or do they have a basis in fundamental perceptual processes that deal with vantage points and perspective?
Leaving aside chiaroscuro and texture, since we are dealing with outline drawings, Kennedy says:
...conditions for observation always entail a vantage point lying in a terrain at a given direction from the object, as well as the source of the energy. Visual perception takes the outcome of three factors -- illumination, shape, and direction -- and must solve the outcome as an equation with three roots. (181)Visual depth perception depends on the geometry of perspective, to within a reasonable approximation. For instance, if we look at a cube so as to see a front surface and a top surface,the angle subtended at the eye by the top front edge is larger than the angle subtended by the top rear edge. This is why we perceive the top rear edge as farther away from us than the top front edge. (There is no suggestion that we consciously calculate these angles.) I shall leave aside the other details of our dealings with visual perspective to go to the heart of the matter.
It is a principle of perspective that an object subtends a smaller angle when it is far away than it does when it is near. The decrease in angle with distance is familiar to sighted people, but there is an unfortunate tendency to think of the decrease as purely visual. (192)I confess that I was prey to this tendency, and was sceptical about this claim:
It is in the nature of direction and perspective that the principle of convergence is not restricted to vision. Rather, it is a general principle that governs hearing, smell and vision -- any perceptual system that encounters objects and locations. (193)The following thought experiment that I put together made me a convert. Imagine having before you, on the left, a rose bush and, on the right, a nicotine plant. On the rose bush is a singing warbler and on the nicotine plant a bee. Now, as you move away from the flowers the visual angle subtended between them decreases, so that you know from vision that you are moving away. However the same holds true of auditory, olfactory and haptic perception. The angle through which you have to move your nose to obtain the most intense smell from each flower in turn decreases as you move away. Similarly with the angle you have to turn your head to hear the bird sing, then the bee buzz; similarly with the angle through which you have to turn your arm to touch the plants. So there is no reason to reject the general claims about vantage points, direction and perspective that Kennedy discusses. Of course his experiments, and the many experiments of others that he reports, were empirical ones with various subjects, rather than gedanken experiments such as mine.
One such was a variation of Jean Piaget's three mountains test. Three objects, a cube a ball and a cone are arranged on a table. How would the three appear (visually or haptically) when the vantage point is changed to each of the four sides of the table? For instance, when viewed from one side the cube will appear between the cone and the ball; from another side the cone will appear between the cube and the ball. The task can be done through the identification of, or through the production of, outline drawings. It turns out that there were no statistically significant differences between the early blind, the late blind and sighted volunteers in performing these tasks. The differences that do occur are between age groups. So... "blind people have the same ability as sighted people to envisage the pictorial role of a vantage point on a simple scene." (208)
A difficult and contentious issue concerns how spatial knowledge is encoded, represented and manipulated internally. For the blind, it is claimed by some, the encoding would not be in the map or iconic form, but rather verbal--even though early blind, late blind and the sighted can do equally well on map-like tasks such as walking a triangular route. The problem in resolving the dispute lies in the fact that space can be expressed in many different ways: in words, by a map, by coordinates from a given point, as geographic directions, in pictures or by flow charts, each of which can be translated into the others. Moreover, there is nothing to stop a person from coding in one way and decoding in another in order to pass a test. Thus "In principle there is no single test that can tell what representational sysytem is being used by someone who knows how to translate from one system to another." (211) What is tolerably clear is that the blind can have a pretty good understanding of space, including vantage point principles, relevant directions, proportional distances and convergence, and somehow make use of spatial "imagery". Kennedy admits there are unsolved problems, and concludes, modestly, with a general point: "Key principles that create perspective effects are understood by the blind because perspective is a science of direction, and direction is as relevant to touch as it is to vision." (215)
It is clear, for example from cartoons, that there can be pictorial metaphors. Indeed Kennedy has argued that for each figure of speech there is a pictorial metaphor. (Kennedy and Simpson, 1982) Since outline drawings can be understood and created haptically, there should be haptic metaphors. It is, however, notoriously difficult to frame a precise and uncontroversial definition of "metaphor". Nevertheless, to do empirical investigations it is necessary to have some working definition of the term, i.e., to make a distinction between literal and metaphoric expression. For Kennedy, a metaphor is an intentionally apt error. (218). "Ergo, to have a pictorial trope the only essential matter is that some part of the drawing be an intentionally apt error." (220)
The two key questions are: When is a picture in error? and When is an error apt? I could quibble about the answers to these questions, for any answer is fraught with difficulty. And it would be intriguing to follow up on a suggestion by John Vervaeke, who pointed out that metaphors can give rise to failures in deductive implication: "All surgeons are butchers; all butchers are shopkeepers; therefore all surgeons are shopkeepers" seems to have a valid form, but the conclusion does not follow. (219) Nevertheless, it is more productive to see what useful work can be gotten out of Kennedy's definition.
Consider an experiment he used to show that pictorial metaphors can be judged similarly by the blind and sighted children and adults. There were five outline drawings of a wheel having differently shaped spokes: curved, bent, wavy, dashed and one with spokes extending beyond the circumference of the wheel. The task was to rank them as best to worst for showing a spinning wheel. All three groups agreed strongly on the best (curved spokes) and worst (extended spokes) for depicting motion; there was also pretty good agreement on the intermediate cases.
In a further experiment, with different groups, the task was to match the pictures with kinds of motion: spinning, jerky, wobbly, too fast, and brakes on. Again there was a remarkable amount of agreement. (See Table 7.7, p. 247 for numerical details.) We see here both error and aptness at work. The spokes of a real wheel would usually be straight. The various distortions in the depictions, which can be perceived haptically as well as visually, were, with considerable agreement, judged apt for depicting specific kinds of motion. The fact that the blind had had no experience or training in the pictorial representation of motion strongly suggests that the aptness is "...perceptual and intuitive rather than thought-out in a manner easy to make explicit." (249) What does seem to be required for an understanding of metaphor "...is some method of comprehension ... that tries to find the right level of generality...[and] a good general knowledge of a culture and its roles and artifacts." (251) There will be conventions involved, to be sure, but there will also be a grounding in basic perceptual capacities that are "hard-wired".
The blind can not only understand haptic metaphors but can produce them. They can, for instance, add lines to a drawing to indicate motion, pain, noise, and smells. It also turns out that the blind can understand stick figures used to portray emotions, such as anger, sadness and pride. (263-267) Given these abilities, I see no reason in principle why there could not be blind cartoonists telling haptic jokes and creating caricatures, both in pictorial form. In a different vein, why could we not develop for the blind such projective tests, normally visual in form, as the Thematic Apperception Test or the Rorschach Ink Blot Test?
Chapter Eight raises a budget of questions that have promising leads but leave us, at present, in quandries:
What does it mean to get a perceptual impression from a representation? What is an axis? What can a picture show? How may outline pictures have originated? What can be amodal? What can a line picture reveal about a psychological state or a property of a surface other than shape? How can pictorial style be captured in nonvisual media?Kennedy has some intriguing and suggestive things to say about pictures (whether visual or haptic) with figure-ground ambiguities, about the use of stick figures to represent emotions, about the universality of recognizable pictures in the rocks and caves of prehistory, about the role of octaves in vibratory touch and hence the possibility of haptic representations of music, about the depiction of surface features such as texture, hardness or softness, and about the question of pictorial style. His main claim, the leitmotif of his book, is "...that there are universals of perception and representation, and that it is important to determine what can be built from the universal elements and what rules are spontaneously followed by those who have not been taught them." (289)
There are two issues about which I wish to quarrel a little, one concerning evolution, the other Kennedy's theory of axes.
It may be that "Evolution has produced a species that uniquely can make and interpret outline pictures." (269) It may be, as he says, that there is no record of gradual improvements in this ability. But Kennedy goes on to make some strong claims that I think do not follow and are questionable:
The making of pictures did not confer an advantage that selected one species over another. Rather, the capacity to understand depiction is an evolutionary spin-off. It is an example of a complex skill universally present in contemporary humanity that has had no role in evolutionary selection. (269-270)The absence of evidence does not mean that something has not occurred -- there are many gaps in other evolutionary stories that only gradually get filled. It is possible that the making of pictures gave no evolutionary advantage, but it is also at least plausible that it did. Picture-making, like language in general, makes possible sophisticated communications that could aid in group survival. If the ability is "hard-wired," it is plausible that it has a genetic component. Of course, maybe none of this is true. But I would like to see a stronger argument for its contrary.
Kennedy holds an amodal theory of spatial perception: both vision and touch, for instance, can stimulate a kind of "apprehension...that is not primarily rooted in one sensory modality." (16) Note the use of a very abstract term "apprehension". I doubt that we can have an experience of spatial features or relations that is amodal, one that is neither visual, nor auditory nor haptic and so on for all the modalities. It may be -- and I find that Kennedy's work makes this plausible -- that we can have an understanding of space that is not tied essentially or primarily to one modality. Now, in explaining his amodal theory he makes use of a special concept of an "axis". He first introduces and partially defines this term in a discussion, too long to recount here, of neurological events in perception. Suffice it to say that when cells fire to two changes close together in the visual field "...they take the two changes to act as one and to stand for a single feature. Let us call the single feature the axis of the lines or contour. The axis is a location. It shows the location of the feature depicted by the line." (36, italics mine) I found this confused and confusing. It makes certain brain cells agents that are capable of intentional states such as belief and representation. There is no need to bring brain cells into the definition. An axis is an abstract feature of space in the dictionary sense of "one of the reference lines of a coordinate system." We can then ask the question, What goes on in the brain when we perceive (or at least understand) an axis in an object? Kennedy's answer to the question may be correct, but it is highly speculative.
Those who are "challenged," who are afflicted with the absence or loss of some ability, such as the lack of sightedness, have afforded science with a wealth of useful information that throws light on our normal capacities. They also show, through their use of capacities that they do have, enormous talent, strength and flexibility. Through his fascinating studies of blind drawing Kennedy has done both the blind and the sighted a signal service.
Dennett, Daniel C. 1991. Consciousness Explained. Boston: Little, Brown and Company.
Flanagan, Owen. 1991. The Science of Mind, second edition. Cambridge: The MIT Press.
Kennedy, J.M. 1997. How the blind draw. Scientific American, Vol. 276, No. 1.
--- and Simpson, W. 1982. For each kind of figure of speech there is a pictorial metaphor: A figure of depiction. Visual Arts Research 16:1-11
Sacks, Oliver. 1996. An Anthropologist on Mars. Toronto: Random House of Canada Ltd.
---. 1990. Awakenings. New York: HarperCollins Publishers.
Seager, William.1991. Metaphysics of Consciousness. London: Routledge.
J.T.Stevenson is a Professor Emeritus of Philosophy at University College, University of Toronto. He maintains his interest in epistemology and philosophy of mind by teaching a graduate course in epistemology. Among his other interests are the philosophy of economics ("The Monetary Power", forthcoming in Philosophy of Law: An Encyclopdia. New York: Garland Publishing), engineering ethics (Engineering Ethics: Practices and Principles. Toronto: Canadian Scholars Press, 1990), and political philosophy ("Aboriginal Land Rights in Northern Canada" in Contemporary Moral Issues ed. by Wesley Cragg. Toronto: McGraw-Hill Ryerson Ltd., 1992). His involvement in medical ethics currently takes the form of work as a volunteer in palliative care with the Dorothy Ley Hospice.