Francesco Ferretti University of Rome 3

Blindness, Visual Content and Neuroscience Abstract This essay addresses a specific issue in the attempt of responding to a more general problem. The specific issue concerns an aspect of the nature of mental content; the more general problem is the age-old problematic relationship between philosophy and the empirical sciences. Despite the fact that the cognitive sciences are a firmly consolidated reality in the contemporary theoretical landscape (even in Italy), several authors obstinately persist in considering philosophy a separate and distinct discipline, independent from the methodologies used and the results produced by the empirical sciences. In opposition to isolationistic positions of this kind, this essay intends to demonstrate that philosophy (at least the philosophy that investigates the nature of the mind and of language) is inextricably connected to the results and methodologies of those sciences that have the very same kinds of inquiries as object of their study.

sciences that have the very same kinds of inquiries as object of their study. The connection between science and philosophy can be interpreted according to at least two different definitions. The first is the following: The philosopher constructs (in complete autonomy) interpretative models of a particular phenomenon (such as language comprehension or the nature of consciousness, for instance) while the scientist, on the other hand, has the subsequent or secondary responsibility of putting such models to the test in order to discover whether they are true or false. If we instead see things according to the second definition, philosophy and science are in a relationship of mutual exchange: the empirical data produced by science constrains from the bottom-up the construction of interpretative models. The idea this essay is based on, is that in order to criticize the very roots of the isolationistic assertions of philosophy, one must analyze its relationship with the empirical sciences according to the second definition. In order to demonstrate the explanatory efficacy of an approach based on the idea of a relationship of mutual exchange between philosophy and the empirical sciences, we will analyze a textbook case: the relationship between thought and language. This will be done through the analysis of a very specific problem: the distinction between «knowledge by acquaintance» and «knowledge by descriptions» proposed by Bertrand Russell in The Problems of Philosophy (1912). A big help in clarifying the problem, as we will see in the conclusive section of this essay, comes from the contribution of the neurosciences: an emblematic case of a fruitful exchange and relationship between philosophy and empirical science.

2. What is lost in blindness? This essay addresses a specific issue in the attempt of responding to a more general problem. The specific issue concerns an aspect of the nature of mental content; the more general problem is the age-old problematic relationship between philosophy and the empirical sciences. Despite the fact that the cognitive sciences are a firmly consolidated reality in the contemporary theoretical landscape (even in Italy), several authors obstinately persist in considering philosophy a separate and distinct discipline, independent from the methodologies used and the results produced by the empirical sciences. In opposition to isolationistic positions of this kind, this essay intends to demonstrate that philosophy (at least the philosophy that investigates the nature of the mind and of language), is inextricably connected to the results and methodologies of those

Does it still make sense to keep using Russell’s distinction today? A good way of asking oneself if the direct experience of things leaves a trace in the conceptual system is to analyze the extreme case represented by congenital blindness. Bryan Magee (a sighted philosopher) and Martin Milligan (a blind philosopher) have given rise to a lively debate on the topic. Magee’s basic idea is that the conceptual system of blind people is devoid of an essential part of informative content (the content related to the visual experience of reality). Instead, according to Milligan, the concepts of the blind can be perfectly superimposed to those of people who can see; His idea is that in spite of the fact that blind people cannot do all of the things that sighted people do, they have knowledge of the things sighted people have knowledge of. Every form of knowledge, according to him, is founded on knowledge by description. What proof and arguments do these authors provide to support their hypotheses?

118

119

1. Introduction

Magee uses the distinction proposed by Russell between «knowledge by acquaintance» and «knowledge by descriptions» (propositional knowledge). Russell’s opinion is that due to the blind’s lack of access to appropriate sensory information, their representation of reality must be radically different from that of the sighted. He thinks that, in fact, two different types of knowledge exist: propositional (or by description) knowledge that avails itself of language and which can exist independently from direct experience (as in the case of when someone describes an object to us that we have never seen), and the knowledge that instead depends on the direct experience we acquire through the sensory perception of the things and events of the world. According to Russell, a blind man can have only propositional knowledge, of that of which, he can never have any direct experience of. Light, is an emblematic example: It is sometimes said that “light is a form of wave motion”, but this is misleading, for the light which we immediately see, which we know directly by means of our senses, is not a form of wave-motion, but something quite different – something which we all know if we are not blind, though we cannot describe it so as to convey our knowledge to a man who is blind. A wave-motion, on the contrary, could quite well be described to a blind man, since he can acquire a knowledge of space by the sense of touch; and he can experience a wave-motion by a sea voyage almost as well as we can. But this, which a blind man can never understand, is not what we mean by light: we mean by light just that which a blind man can never understand, and which we can never describe to him (Russell, 1912, pp. 28-9).

The question concerning whether the lack of direct access to reality also entails deficiencies on a cognitive level, or whether language can serve as a substitute for perceptual experience is an open question, one that is still the focus of heated debate. Magee sustains that because the phenomenology of visual experience continues to be denied to Milligan, the only knowledge he can have of the things and events that regard visual experience, is propositional, and thus, of a highly indirect nature. Knowledge by description, however, while it works fine when used for other purposes, is, in and of itself, too poor a means for it to supply the correct concept of any given visual experience, to one who is deprived of the sense of sight. Magee, referring to Milligan, writes: You disregard the fact that since vision-related content in any empirical concept must derive form direct experience, and since you yourself cannot apprehend such content in terms of your own experience but have to acquire understanding of it

120

indirectly by behavioural observation and inference, not only are you altogether incapable of understanding some of those concepts but you are almost certainly apprehending only part of the content of the others (…). By pitching your claims as high as you do, you talk as if the balance between propositional knowledge and experiential knowledge were different from what it is, indeed as if it were almost the opposite of what it is. Propositional knowledge, knowledge by description, is pale, grey, thin, second-hand stuff compared with the knowledge by acquaintance from which it is abstracted (Magee & Milligan, 1995, pp. 24-5).

Magee’s thesis is that direct knowledge (that is supplied in an intentionally conceived way and without a mediating deductive process) cannot be reduced to knowledge by description: it then follows that, because the blind cannot perceive the visual appearance of objects (as the case of light demonstrates), the sense of the terms the blind use to refer to this particular type of reality, is strongly impaired. Milligan’s response, basing itself on the fundamental power language has on cognition, finds its theoretical point of reference in the philosophy of Wittgenstein: language renders the thoughts of the blind completely superimposable to those of the sighted because language is the basic tool used in the construction of all forms of knowledge. There are obvious differences between those who can see and those who instead cannot, but these differences do not interfere with conceptual content for the simple reason that the structure of the conceptual system is of a propositional nature (and propositions are shared by the blind and the sighted). Milligan writes: All knowing (as the term is commonly used today) is knowing about, or knowing that (knowing about, and knowing how, too, in my view, can quite properly be treated as forms of knowing that); (…) all knowledge is propositional knowledge; and (…) knowledge by acquaintance is just one kind of propositional knowledge. (…). Knowledge by acquaintance differs from other kinds of propositional knowledge not by being non-propositional, but only by having been generated by direct contact with, by sense-experience of, its objects (ivi, pp. 61-2).

As a result of these considerations, Milligan’s idea is that the «difference between blind and sighted people is not, as such, a difference in the kind or amount of knowledge possessed about light, but only a difference in the way it has come to be possessed» (ivi, p.67). That which distinguishes the sighted from the blind is the way in which they construct their conceptual system,

121

and not the nature of the concepts they possess: when you take into consideration the level of knowledge there is no significant difference to be found between the sighted and the blind. At this point, the terms under discussion appear to be sufficiently clear: on one hand, by basing his argument on the intrinsic propositional nature of knowledge, Milligan sustains that as a result of language the conceptual competences of the sighted and the non-sighted are completely homogeneous; on the other hand, by making use of Russell’s distinction between «knowledge by acquaintance» and «knowledge by description», Magee sustains, instead, that part of the concepts the blind possess are impaired in content, and as a result the meaning of the terms used by the blind to describe certain aspects of visual experience, cannot be the same as that used by the sighted. This might seem a problem of little importance, but by means of this circumscribed and limited dispute, we enter into the very center of one of the most crucial reflections concerning the philosophy of language: the question concerning the relationship between thought and language. Which of these two hypotheses is the most convincing? Can empirical science help us resolve this problem?

3. Thought and Language John Locke’s An Essay Concerning Human Understanding is the natural and necessary starting point for the argument that language is based on experience. Two aspects of what he proposed are particularly important to our discussion: the fact that names refer to ideas; the priority given in learning to the naming of ideas or of simple substances. Now, because the names of simple ideas are not susceptible to definitions (their meaning cannot be expressed by means of terms other than their synonyms), Locke’s idea is that although the blind may use the same terms used by the sighted, when these terms are directly connected to the impressions of simple ideas, their meaning cannot ever be completely the same as the meaning attributed to them by the sighted. In these kinds of situations, in fact, it is not possible to explain with words an experience one has not had: Simple ideas, as has been shown, are only to be got by those impressions objects themselves make on our minds, by the proper inlets appointed to each sort. If they are not received this way, all the words in the world, made use of to explain or define any of their names, will never be able to produce in us the idea it stands for. For words, being sounds, can produce in us no other simple ideas than of those

122

very sounds (…). He that thinks otherwise, let him try if any words can give him the taste of a pineapple, and make him have the true idea of the relish of that celebrated delicious fruit. (…). In light and colours, and all other simple ideas, it is the same thing: for the signification of sounds is not natural, but only imposed and arbitrary. (…). And no definition of light or redness is more fitted or able to produce either of those ideas in us than the sound light or red by itself. (…) And therefore he that has not before received into his mind, by the proper inlet, the simple idea which any word stands for can never come to know the signification of that word by any other words or sounds, whatsoever put together according to any rules of definition. The only way is by applying to his senses the proper object, and so producing that idea in him for which he has learned the name already. A studious blind man, who had mightily beat his head about visible objects and made use of the explication of his books and friends to understand those names of light and colours which often came in his way, bragged one day that he now understood what scarlat signified. Upon which, his friend demanding what scarlat was, the blind man answered it was like the sound of a trumpet. Just such an understanding of the name of any other simple idea will he have who hopes to get it only from a definition or other words made use of to explain it (Locke, 1690, III, 4, 11).

Starting off from considerations very close to those of Locke, Anne Dunlea (1989) put to experimental test the hypothesized role of visual perception in language, by studying the case of the congenitally blind. Empirical results confirm the reality of such a role at various levels of analysis: lexical, syntactical and pragmatical. According to her, two questions deserve attention: from a cognitive point of view, what renders visual information specific? What is the role vision plays in the acquisition and use of language? Both questions are closely related: only by attributing peculiar properties to vision (by demonstrating that visual information cannot be reproduced by other sensory modalities, for example) is it possible to sustain that the terms used by the blind in reference to visual experience do not have the same meaning as those used by the sighted. Dunlea insists on the difference between sight and touch: Consider for a moment the nature of vision. It is the form of information input that allows the easy summation of simultaneous spatial reality independent of time. Vision, then, enables us to establish and maintain a coherent concept of the environment and our existence in it without struggling with memory and information retrieval. Quite simply, the blind must remember what the sighted can effortlessly reconstruct with a single look. Above all, vision has the outstanding and

123

unique quality of simultaneity. In the absence of vision, all the other modalities put a tremendous burden on the mind’s ability to synthesize a coherent sense of the objective environment and one’s position in it» (Dunlea, 1989, p. 10). Tactual perception requires direct contact with the object to be “observed”. Certain objects are therefore inevitably inaccessible – distant objects (sun, moon, stars); large objects which cannot be perceived in full (mountains, rivers, large buildings); fragile objects (soap bubbles, smoke, fog); and minute objects (fleas, dust). Moreover, certain objects under certain circumstances are inaccessible to tactile exploration (burning fuel, boiling water). Thus a great many experiences taken for granted by sighted children are difficult or impossible for blind children. Another difference between sight and touch is that visual perception goes on throughout the waking state, whether or not one fully attends to it, whereas hands as touch organs need to be actively applied and the scope of their application is limited to one arm’s length without moving (ibidem).

According to Dunlea, visual experience affects the conceptual system in a very particular way. The abstraction and generalization capacities allowed by vision are not comparable to those allowed by other perceptive modalities. When the conceptualization processes at the basis of lexicon acquisition are studied, the specificity of this perceptual modality strongly comes to the foreground. Now, if vision has traits of specificity that weigh heavily on the processes of categorization, Dunlea’s hypothesis is that «even if blind children use the same lexical forms as sighted children, it is probable that the referents to which they extend the original word will differ from sighted children’s» (ivi, p. 42). One of the many cases proposed by Dunlea to prove this hypothesis is the role of extensions and overextensions at the basis of word acquisition. The extension of the meaning of a word (i.e. dog for animal) is well known in the initial phases of language acquisition; from Dunlea’s experimental results it emerged that while the sighted extended a lexeme to a heterogeneous class of referents (showing that the typical principles of generalization and abstraction were operating correctly), the “extensions” observed in the blind children were far more limited (…). As was the case with overextension, those lexemes that were actually extended to new referents on the basis of common attributes were applied to only one or two additional referents, rather than to a large set of items. This suggests that while some of the forms were applied to new referents, they remained underextended in terms of standard usage (ivi, pp. 58-9).

124

Another phenomenon at the basis of lexical development is «decentralization», the capacity children have of shifting attention from the objects they come in contact with or act upon, to the identification of the actions of others and the objects they act upon; the capacity to decentralize also seems highly impaired in blind children. The analysis of the effects of visual deprivation on categorization processes, leads Dunlea to conclude that the kinds of referents for which blind and sighted children have lexemes are remarkably similar in that object words and word for actions constitute the majority of terms. But, the way in which the children understand these lexemes is vastly different (ivi, p. 59).

The idea underlying Dunlea’s argument is that if the conceptual system is impaired, so is the comprehension-production of language. Following this line of reasoning, her conclusions are that blind children’s conceptions of events are at times different from sighted children’s conceptions of them, and this is necessarily reflected in what blind children express linguistically (ivi, p. 103).

In order to completely understand the theoretical importance of what Dunlea proposes it is useful to extend her argument to its most extreme consequence. The empirical data supplied by the author seem convincing; as is the idea that language (its acquisition and its effective use in the processes of production and comprehension) grounds its roots in a conceptual system largely constituted – at least in its initial phases – by visual experience (Mandler, 2004a, 2004b). Arguments of this nature are also important to other critical issues (underlined by Magee in reference, once again, to Russell’s theses): the opportunity of clearly distinguishing between the comprehension of a linguistic expression according to the terms of others linguistic expressions and the comprehension of a linguistic expression also according to the terms of nonlinguistic experiences. Magee writes: If everything we could say were fully explicable in terms of other things we could say then the entire system of language would be circular. At no point would it be earthed to anything non-linguistic. Therefore this cannot be the case. It has to be that some elements of what is said are understood not in terms of other things that are said but in terms of something non-linguistic, for if that were not so the-

125

re would be no point at which language made contact with reality (whether this be “the world” or “experience”) (Magge & Milligan, 1995, p. 116)

The topic of how to anchor language to the world is a really crucial topic in order to understand the processes of comprehension at the basis of semantic competence. Having said this, to insist too much, however, on the aspects that exalt the dependency of language on the perceptual system, can lead to a conception of language, called «verbalism», according to which the meaning of some of the terms used by the blind would be «empty» in respect to the common use of these terms. Thomas Cutsforth’s book (1951) is unanimously considered the most extreme example of this theoretical position. His idea, in fact, is that the name of a thing that is seen has a completely different meaning to that of a name of a thing that is felt (or heard); the consequence of this is that, through language, the blind person comes in contact with a world that is totally unreal: A congenitally blind person’s discussion of the theory of color mixture is not a whit more verbal than a college student’s opinion on the labor question when he has never worked a day in his life. The factual content in each case is derived from the experience of others. The adolescent college student knows no more about the unrealistic aspects of love and marriage than a man born blind knows about the hue, tint, and chroma of colors. There is, however, a vital difference between the verbal college student and the verbal blind. At the end of his college course, If he does not enter the academic field, the college student will be thrust out into a world which demands verification and readjustment to reality. The blind have no such avenue of escape. They are compelled to continue in a world of unreality. They remain under the necessity of dealing with the unreal as though it actually existed for them (Custforth, 1951, p. 49).

Much as the conceptual system cannot exclusively depend on language, one need not forget that language is a potent instrument at the service of conceptualization. Barbara Landau and Lila Gleitman (1985) have written a book that represents a turning point in the study of the relationship between vision and language. Their thesis is a strong one and it openly contrasts Locke’s concept, which was taken up by Dunlea and then taken to an extreme by Cutsforth. What is really being discussed here? The point is not establishing whether the experience of blind children is different from that of sighted children: it obviously is. The point to be established is how blind children are able to acquire the relationship between lexical forms and their meanings, considering that although the blind two-year-old has difficulties in figuring out what is going on and therefore in conversing competently, this does not necessarily mean that she therefore is unable to learn the meaning of words and sentences (ivi, pp. 19-20).

The thesis of these authors is that, after overcoming an initial delay, from the third year of age onward, blind children have language that is virtually indistinguishable from that of sighted children (in syntax, thematic relationships and vocabulary). Seen from this prospective, language becomes the key element of convergence of the representation of the world of both the blind and the sighted. In support of this thesis the two authors show how the congenitally blind have an accurate knowledge of the significance of expressions relative to vision. To this end, it is very interesting to note the experimental proof related to the difference between the verbs «to see» and «to look»: Landau and Gleitman have demonstrated how sophisticated the competence even blind children have of the subtle difference between these two verbs. The obviously interesting question is understanding exactly what makes such a competence possible. Landau and Gleitman’s response – in reference to Kelli, a congenitally blind child – is clear:

With an argument very similar to that used by Russell to highlight the circularity of a language that is founded solely upon itself, Cutsforth sustains that the acquisition of words by the blind «serves both the objectivity and to socialize the life of the blind child and at the same time to isolate him still further from the seeing world in which he lives. This is the beginning of verbal unreality» (ivi, p. 11). How can we deal with these objections? How can we avoid the solipsistic drifting of language in the blind? Our response foresees two different moves: the first move to make involves recognizing the role language plays in processes of conceptualization; the second regards the revision of the concept of visual experience.

These authors’ hypothesis is that «the verbs as used by Kelli’s mother would be distinctive in their syntactic environments, providing a format within which their meaning could be individuated» (ivi, p. 111). The following are

126

127

Extralinguistic information in the conversational settings does seem to account for part of what was acquired. But further aspects of the learning suggest strongly that Kelli was also sensitive to the syntactic structure associated with individual verbs, and that this linguistic context is – along with the extralinguistic context – important for the discovery of verb meanings (ivi, p. 98).

a series of syntactic restrictions that allow the individuation of the diverse meanings conveyed by «to see» and «to look»:

Landau and Gleitman’s idea is that «to look» and «to see» occur in different syntactic contexts and the analysis of the information conveyed by the two different types of syntactic structures is decisive in establishing the difference in meaning between the two verbs. From their point of view, in other words, the conceptual content critical to the comprehension of the terms related to sight does not arrive by way of the eyes. Landau and Gleitman also apply the same model of interpretation to the (extreme) case of terms related to color. What the blind know about these terms depends on the language elaboration system, just as in the case of the sighted. In the case of color, however, a problem arises. If we take Landau and Gleitman’s thesis to its extreme, we arrive at the point of supporting the hypothesis in which the comprehension of the meanings produced by both the blind and the sighted coincide perfectly. This is a thesis that is probably false (just like verbalism, if although for the opposite reasons). The two authors also refer to only one «partial coincidence» in the meaning of color terms. The perplexities expressed by Landau and Gleitman in regards to color, highlight an important issue and that is that language, by itself, cannot «exhaustively» communicate all perceptual information. To sustain that some of the properties of color remain inaccessible to the blind, takes us back to the Magee-Milligan debate. The residue of information that language does not exhaustively communicate weighs heavily on the conceptual system and thus on the different comprehension of meaning the

sighted and the blind have. Even if we were to accept the compensating role language may play in the construction of the blind’s conceptual system, what is left over in the comparison of both conceptual systems, appears to be a difficultly contested fact. On the other hand, as we have already seen, to insist too much on the differences that exist between these two systems leads us to consider the language of the blind as an instrument made up of (at least some) terms that are empty of meaning. How are we to avoid the difficulties presented by these opposing views of the problem? The answer to this question passes by way of an argument that is capable of keeping together that which both conceptual systems have in common, and that which instead makes them different. The key lies in the revision of perceptual experience as a concept. The clear distinction between «direct knowledge» and «knowledge by description» is, in fact, based on a naïve conception of what perception is. According to this conception, the act of perceiving is a typically passive act: when we open our eyes, the world enters in us. Though this idea continues to exist in our common sense thinking, Hermann von Helmholtz’s theory of «unconscious inferences» questioned it. According to this theory, which represents a point of no return in research on the topic, visual perception does not only depend on sensory access, it is an active process that crucially depends on internal elaboration processes as well. Using the title of a collection of essays edited by Dario Galati (1992), we could say that to perceive means above all, to see with the mind, rather than with the eyes. An important result of Helmholtz’s thesis is the idea that making the distinction between sensory and linguistic does not address the problem in an exhaustive manner. In order to fully explain the process of perception, a third level of analysis seems necessary: the level of the representation and elaboration of information. The appearance of objects does not solely depend on their perceivable properties (it does not depend only on the sensory apparatus whose job it is to gather the stimuli presented by the environment); it also depends on the internal information elaboration system. Seen in this way, blindness is often just a peripheral disturbance. From this point of view, one could end up supporting a thesis that is in a way, paradoxical: the thesis is that, the blind, though not having visual access to the world, may very well have access to at least part of the knowledge regarding the visual appearance of objects. This would lead us not only to a new concept of what blindness is, it would, above all, lead us to a new idea of what «visual experience» is. In order to support a thesis like this, one must demonstrate that the blind’s representation of the world makes use of some kind of visual content.

128

129

1. only see appears in deictic interjective queries (“See? and “See?, that’s a circle”) 2. with a single exception, only see appears with sentential complements (“let’s see if Granny’s home”) 3. only look appears in deictic interjective commands (“Look!” and “Look, That’s a boot”. 4. only look appears deictically with a free how relative (“Look how I do it”) 5. only look appears in constructions with like (“You look like a kangaroo in those overalls”). 6. Look and see allow a maximum of two associated NPs, Ds, and PPs, while other verbs may (get, hold) or must (give, put) take three of these: English allows Mary gives John the book, Mary holds the book out, and Mary puts the book o

4. Visual Content What types of representation of reality do the blind possess? An interesting response to this question is provided by the study of the visual imagery of the congenitally blind. Is attributing a visual content to mental images of the blind possible? One way of approaching this problem is to study their drawings (Kennedy, 1993). The careful analysis of these drawings demonstrates that visual perception depends on a series of elaboration structures that go well beyond those strictly responsible for sensory access. Kennedy and Juricevic (2006) describe the case of the Turkish artist blind from birth, Esref Armagan, who was asked to draw a cube taking into consideration the different perspectives a viewer could take. His drawings are remarkable: they respect perspective and occlusion, two properties that depend on being able to see an object from a distance (fig.1).

say that blind subjects did generate visual mental images, we support the idea that, as in sighted subjects, mental imagery mechanisms undergo perceptual influence in blind people» (ivi. p. 10). Some of the most interesting data regarding the problem of whether visual experience is necessary in order to produce visual images comes from studies done on dream activity in the congenitally blind. Since dreaming, in sighted people, is associated to vision, the prevalent assumption is that the blind are incapable of dreaming. Bértolo (2005) in opposition to this idea, proposes an alternative path in investigating the issue. Dreams with visual content are expressions of visual imagery. Therefore if dreams with visual content could be demonstrated in congenitally blind persons, this would imply that visual imagery is possible in subjects who have been prevented from having visual experiences. Furthermore, this would allow one to infer that visual imagery does not depend on specific visual perception, but can emerge from activation of visual cortex by non visual inputs (ivi, p. 180).

Now, how are the blind able to represent the appearance of the world in perspective, an ability typical of seeing objects from a distance? A possible answer is that the blind use cerebral areas involved in vision and visual imagination. Lambert, Sampaio, Mauss Y. & Scheiber (2004) in an fMRI study have supplied data that supports the idea that the primary visual areas (PVA) of congenitally blind test subjects are activated when they perform visual imagination tasks. These researchers, conclude that «without going quite as far as to

In the first place, different studies have shown that the blind have dream experiences and that these dreams «contain mostly sounds, touch sensations or emotional experiences» (Bértolo, Paiva, Pessoa, Mestre, Marques & Santos, 2003, p. 277). In any case, the idea that the dreams of the congenitally blind or of those who became blind before the age of 5-7, are void of any visual content, remains predominant (Kerr, 2000). The issue of visual content is extremely controversial because it concerns subjective experience and as such, is difficult to confirm empirically. Today, however, we have objective means of testing this hypothesis, which confirm the presence of visual elaborations in dream activity. Bèrtolo and his collaborators have used EEG and measurement of alpha power as indexes of visualization activity in the brain. The alpha rhythm is, in fact, an index of the visual activity the brain is involved in: studies done on sighted subjects have shown that alpha activity attenuation or blocking is a good indicator of visual imagery capacity. Lopes da Silva (2003) was the first to provide empirical evidence (as a result of a study done using EEG) of the decrease of alpha strength recorded from the central and occipital regions of blind people engaged in visual imagery tasks. He thus, provided proof to sustain the idea that «subjects who have never had visual experiences can have dreams with virtual images that are probably mediated by the activation of the cortical areas responsible for visual representations» (ivi, p. 328). Bértolo and his collaborators (2003), using these correlations as

130

131

Fig. 1. Esref Armagan (congenitally blind) draws the correct number of sides of a cube in respect to perspective and eliminates the lines hidden by the effect of occlusion of the sides in the frontal view. Figure taken from Kennedy e Juricevic (2006). URL: http://www.utsc.utoronto.ca/~kennedy/one-point.pdf

basis for their work, studied the dream activity of the blind, focusing in particular, on the visual content of their dreams. The most important result of this study regards «a significant negative correlation between the visual content of the dreams and the alpha power» (Bértolo et al. 2003, p. 282).

nitally blind subjects, for what concerns the debate on the relationship between vision and imagery, regards the fact that the visual nature of the representations involved in the tasks performed, in no way depends on having had an actual visual experience, it instead depends on the shared use of a specific elaboration system. Farah (1988), in fact, emphasizes that: Imagery is not visual in the sense of necessarily representing information acquired through visual sensory channels. Rather, it is visual in the sense of using some of the same neural representational machinery as vision. That representational machinery places certain constraints on what can be represented in images and on the relative ease of accessing different kinds of information in images (ivi, p. 315).

Results such as these, lead us to seriously consider the thesis in which visual experience does not exclusively depend on sensory access. Even when such access is denied or obstructed, some of the properties of visual experience can be efficiently represented, by way of the activation of specific internal systems of elaboration.

5. Conclusions

The most compelling evidence that emerges from the studies done on conge-

What conclusions are we to draw from this data, in regards to the theoretical question from which our discussion began? The first consideration to make is that the study of the congenitally blind offers interesting ideas to reflect upon concerning the relationship between thought and language. The study of the conceptualization and signification processes of the blind, in fact, represents a privileged perspective on the analysis between «direct knowledge» (based on sensory perception) and «knowledge by description» (based on language). The first result that emerges from the empirical literature is that Russell’s distinction is a very effective explanatory device and as such deserves to be maintained. Dunlea’s experiments have shown, in fact, the causal role the perceptual system plays in conceptualization and the weight the conceptual system has on processes of signification (linguistic comprehension and production): when a part of the informative content (the part concerning direct experience) is missing, the significance of certain linguistic terms, as a result, is also impoverished. Upon further reflection, however, Russell’s distinction appears still to be too restrictive: the idea that one could distinguish vision from blindness in terms of a net distinction between those who can see and those who cannot, is, though intuitive, a much too weak conception from an explanatory perspective. The fact

132

133

Fig. 2. Graphic representation of a dream scene of a blind subject. Taken from Bértolo et al. (2003), p. 181.

In fig. 2 we have the drawing of a blind subject asked to graphically reproduce a dream he had previously described verbally. No statistically noticeable difference was found between blind and sighted subjects: not only were the blind able to verbally describe the visual content of their dreams, they were also able to draw a graphic representation of this content. Bértolo concluded that: According to these results, the congenitally blind, who have never experienced sight, are able to visualise (…). The observation of alpha attenuation/visual content correlation along with the no differences in the graphical representations leads us to hypothesize that blind subjects can produce virtual images, that is, that their dreams correspond to the activation of visual cortical regions (Bértolo, 2005, p. 183).

that certain properties typical of visual perception also characterize the representational capacities of the blind, leads us to sustain that the conceptual system the blind have access to, uses an important part of the perceptual information the sighted make use of. The understanding of the type of knowledge that is involved in cases such as these (a task that must be legitimately recognized as being one of the many activities of philosophy) is still an open and controversial question. Moving beyond the theoretical interpretations that today are made possible of phenomena such as these, another issue appears to be important from a methodological point of view: the results coming from research in neuroscience and experimental psychology constrain the construction of interpretative models, whether we like it or not. This of course, with no offense intended to isolationist philosophers.

Lopes da Silva, F.H. (2003). Visual Dreams in the Congenitally Blind? Trends in Cognitive Sciences, 7, pp. 328-330. Magee, B., & Milligan, M. (1995). On Blindness. Oxford University Press: Oxford. Mandler, J.M. (2004a). Thought before Language. Trends in Cognitive Sciences, 8, pp. 508-513. Mandler, J.M. (2004b). The Foundations of Mind. Oxford University Press: Oxford. Russell, B. (1912). The Problems of Philosophy. Williams and Norgate: London.

References Bértolo, H. (2005). Visual Imagery without Visual Perception? Psicològica, 26, pp. 173-88. Bértolo, H., Paiva, T., Pessoa, L., Mestre, T., Marques, R. & Santos, R. (2003). Visual Dream Content, Graphical Representation and EEG Alpha Activity in Congenitally Blind Subjects. Cognitive Brain Research, 15, pp. 277-84. Cutsforth, T.D. (1951). The Blind in School and Society. A Psychological Study. American Foundation for the Blind: New York. Dunlea, A. (1989). Vision and the Experience of Meaning. Blind and Sighted Children’s Early Language. Cambridge University Press: Cambridge. Farah, M.J. (1988). Is Visual Imagery really Visual? Overlooked Evidence from Neuropsychology. Psychological Review, 95, pp. 307-17. Galati, D. (1992). Vedere con la mente. Conoscenza, affettività, adattamento nei non vedenti. (F. Angeli, Ed.) Milano. Kennedy, J.M. (1993). Drawing and the Blind, Yale University Press: New Haven. Kennedy, J.M., & Juricevic, I. (2006). Foreshortening, Convergence and Drawings from a Blind Adult. Perception, 35, pp. 847-51. Kerr, N.H. (2000). Dreaming, Imagery and Perception. In Kryger M.H. et al. (Eds.), Principles and Practice of Sleep Medicine (pp. 482-90). W.B. Saunders Company. Lambert, S., Sampaio, E., Mauss, Y. & Scheiber, C. (2004). Blindness and Brain Plasticity: Contribution of Mental Imagery? An fMRI Study. Cognitive Brain Research, 20, pp. 1-11. Landau, B., & Gleitman, L.R. (1985). Language and Experience. Evidence from the Blind Child. Harvard University Press: Cambridge (MA).

134

135