Francesco Gori

Practice and Theory of Visual Representation 1. Drawing as a Tool for Displaying the Gestalt Perception Process

Using modern technologies for investigation of the brain, such as PET and fMRi, neuroscientists have been able to observe the initial stages of perception, demonstrating how, at the beginning of the visual process, the brain divides the physical image into its essential attributes. Small edges, complex shapes, colours and movements are initially detected by specialised modules, located in different areas of the brain, and are only later recomposed into the final image. The abundance of data provided by brain-imaging techniques has revealed a great deal about “where” the individual cognitive phenomena take place, but little about “how” they are integrated into a unitary visual experience. How are a series of adjacent dots organised into lines and then objects and concepts that describe increasingly abstract aspects of reality? If philosophers of mind and neuroscientists have not yet reached an agreement on how the visual brain “puts things together”, it is probably because neither of them have been able to use an instrument that permits them to observe the working of the perceptive process as a whole. Rational and conscious thinking has no direct access to the unconscious process of perception while neuro-imaging instruments provide only partial results on the activation of individual areas and neurons in response to specific stimuli. An original approach to the study of visual perception comes from Neuroaesthetics1, a new branch of research that uses artistic experience, both in the sense of enjoyment and creation of works of art, to study the workings of the visual brain. According to neuro-aesthetics researchers, artists’ work is an extension of perception and as such a way to study the process of vision from a privileged point of view.

Lauring, J. O. (2014): An introduction to neuroaesthetics: The neuroscientific approach to aesthetic experience, artistic creativity, and arts appreciation. Museum Tusculanum Press. 1

GESTALT THEORY

© 2016 (ISSN 0170-057 X) Vol. 38, No.1, 17-40

GESTALT THEORY, Vol. 38, No.1

The development of modern neuro-aesthetics is rooted in the work of Rudolph Arnheim, a disciple of Max Wertheimer and Wolfgang Köhler, who applied the principles of Gestalt to the development of a real psychology of art. “Gestalt theory says that the factual world is not simply understood through perception as a random collection of sensory data, but rather as a structured whole. Perception itself is structured, is ordered. This also concerns art. The work of art was a prime example of a Gestalt for my psychology teachers” […] “I consider art to be a means of perception, a means of cognition. Perception makes it possible to structure reality and thus to attain knowledge. Art reveals to us the essence of things, the essence of our existence; that is its function”2.

Following Arnheim, it is no longer possible to consider the artistic process as self-enclosed or inspired from above, but this activity is now considered as a progression of that more humble activity done by the eyes in everyday life. This position has been updated by the founder of modern neuro-aesthetics, the neurobiologist Semir Zeki, according to whom artists and visual brains use two different methods to perform the same task, that of selecting the constants of continuously changing reality to provide us with a more simple and stable representation of the world. As Zeki explains: “All visual art must obey the laws of the visual system”. 3

Fig. 1 Detail of a work by Piet Mondrian.

Semir Zeki and his team used brain-imaging techniques to observe the brain’s reactions to a number of abstract paintings, providing answers regarding the neurophysiological meaning of the works of great artists like Kandinsky or Mondrian. The fact that individual neurons specialising in recognition of edges are more 2

Rudolf Arnheim: “Die Intelligenz des Sehens” in Neue Bildende Kunst (August-September, 1998), pp. 56-62.

3

Zeki, S (1994): The neurology of kinetic art. Brain, 117, 607-636.

18

Gori, Practice and Theory of Visual Representation

powerfully excited by Mondrian’s works shows that the artist, long before neuroscientists, accomplished the feat of understanding that straight lines provide the neurophysiological basis of form. At the same time, this research throws light on the neural and universal bases of aesthetical experience: “we like” the pure linearity of Mondrian’s work because it stimulates neurons sensitive to edges which are located in the primary stages of visual perception, especially in the V1 cortical area. These conclusions nevertheless seem to me to be partial. If brain-imaging techniques such as PET and fMRi allow us to observe the brain’s reactions to a finished work of art, they do not however allow us to see the process with which it has been created. Indeed, it is no accident that while neuroaesthetical research has found important analogies between the purpose of the perceptive and artistic processes, it has not produced such effective descriptions of the way in which both processes occur. If, as neuroaesthetics itself states, the creative process is the extension of the perceptive process, observing the brain when it enjoys a work of art is not sufficient to understand how perception works. We need to observe the process that led to its creation. The creative process, however, is accessible only to the artist, who has not rationalised the creative procedure from which the work originates, nor is he/she able to explain it in words.4 While it is true that an artist is not consciously a neuroscientist, the opposite is also true. Even if neuroaesthetic scientists have a thorough knowledge of art history, they have no direct access to the artistic process and so they have not yet produced effective descriptions of how the perceptive process works. In order to observe the mechanism whereby the visual brain integrates the individual attributes of reality into a unitary representation, a researcher and “artist” must be combined into a new specialist, able to study the creative process from within. Before bringing about this change of outlook, however, we must decide which aspect of visual perception we wish to study through artistic experience: visual attributes such as form, colour, movement and depth are processed and created in different areas of the brain and must be studied through differing artistic techniques.

In the words of the painter Marcel Duchamp: “If we give the attributes of a medium to the artist, we must then deny him the state of consciousness on the aesthetic plane about what he is doing or why he is doing it - all his decisions in the artistic execution of the work rest with pure intuition and cannot be translated into a self-analysis, spoken or written, or even thought out.” Marcel Duchamp, Il processo creativo, in “Art News”, 56, 4, (1957). 4

19

GESTALT THEORY, Vol. 38, No.1

Neuroscientists have partially understood the mechanism used by the brain to process and represent the constancy of colour but they have not yet understood that of forms, as Semir Zeki himself acknowledges: “Physiologists think that the cells that detect edges are the bricks that build the neural processing of each form, even though none of us knows how complex forms are built neurologically from cells that react to what they consider to be the components of each form”.5 The process governing organisation of form is particularly important. We can use it to isolate figures from the background, to know and recognise the identity of objects. We know that, during the primary stages of the visual process, the brain breaks up the form of external reality into millions of small edges, which are recognised individually by specialised neurons, each of which detects small portions of lines. To understand how these small sections merge into lines and then objects, we must choose an artistic expression whose explicit objective is to represent the edges of subjects as complex as real ones. Which form of artistic expression therefore allows researchers to study the process of creation of form? When analysing most figurative works, the difficulty of separating analysis of form from that of colour6 becomes apparent, as does the fact that the very aspect of form is influenced by the cultural goals of the artist. This is clearly illustrated by the different ways in which great artists like Michelangelo, De Chirico or Botero represent the form of the human figure.

Fig. 2 Giorgio De Chirico, “Ettore e Andromaca” (1917), Michelangelo Buonarotti “Sacra famiglia”, (1508). 5

Zeki, S. (2001, p. 136).

From a neurophysiological point of view, form, colour (and motion) seem parallel processes, which start and develop simultaneously along independent paths. However, they produce slightly asynchronous results: colour processing is 20 ms faster than form, that is 40 ms faster than motion. Moutoussis K., and Zeki S. (1997): A Direct Demonstration of Perceptual Asynchrony in Vision. Proceedings of the Royal Society of London, 264, 393-399. 6

20

Gori, Practice and Theory of Visual Representation

An artist is considered an artist precisely because of his/her unique way of seeing and interpreting reality, under the influence of cultural factors that make the process of representation extremely subjective. On the contrary, the process of perception of form is one of objective representation, shared by all individuals, and which, in order to be observed and studied, requires an artistic method that is not only specific with regard to the form, but also as “universal” as possible. Is there then a procedure linked purely to representation of form that is also “constant”: that is, on which the various expressions of figurative art are based? Before becoming abstract, impressionist, cubist or an illustrator, every artist has had to learn to draw the outlines of objects by copying them from life, using a process of simplification on which the technique of drawing from life is based. Every artist knows that in order to represent the form of a complex object he/ she must simplify it, using a process and a hierarchy of construction that has not changed over the past few centuries. If we observe preparatory sketches by Leonardo da Vinci or Raphael, we realise that the process of construction of objects and figures is the same as that taught today in drawing classes.

Fig. 3

The universal nature of drawing suggests that the human brain has no other way of representing the form of reality and that therefore the technique of drawing from life is a conscious extension of the process by which visual perception reconstructs and recognises the outlines of objects. If we stand behind an artist, we can observe the result of the various stages with which he/she first deconstructs and then reconstructs the form but, in this way, we have no direct access to his/ her mind while it makes the effort of simplifying in order to create and represent the outlines of reality. We can only do this by learning and using the drawing technique while making our own observations and describing step by step the process used by the brain to draw the form of reality, a process which perception achieves instantaneously thanks to its parallel architecture. 21

GESTALT THEORY, Vol. 38, No.1

I believe that through the practice of drawing it is possible to extend the current methodological boundaries of neuro-aesthetics research, to study not only the effects of the work on the brain, but the dynamic process that leads to representation. However, since the practice of drawing is not used as a tool to study perception, there is no real literature in support of this theory. It is no surprise that the first person to sense the relation between an artist’s sketch and perception was not a psychologist but the art critic Ernst Gombrich. In “The sense of the order,” he writes that the process by which the artist creates the first sketch can offer an analogy for the way we learn about our environment, not in a flash but according to the double principle of meaning and simplicity7. Gombrich himself was an excellent art critic but not an artist. He was therefore able to sense but not develop a real parallel between the stages of drawing and perception. Moreover, he was unable to verify it with data obtained from modern brain scanning. Therefore, I mean to submit my practical and theoretical experience in representation and my conclusions to the attention of scholars and readers of “Gestalt Theory”. To do so, I would like to introduce a parallelism between the figurative drawing technique and form perception process. 2. Practice and Theory of Visual Representation

Knowing (even before recognising) the stable structure of objects is the purpose both of the form perception process and of an artist who draws from real. Reproducing reality he retraces the stages of visual knowledge8 on a more conscious level. Perception uses millions of specialised neurons to deconstruct and rebuild the physical form of reality in a stable and simplified representation, usable by the conscious unit. Where the visual brain’s work ends, that of the artist begins: he/she starts to represent an object already separated from the background, organised by perception, and recognised by the memory. To do so, the artist carries out a process of decon7

Ernst Gombrich. The Sense of Order. A Study in the Psychology of Decorative Art. Oxford: Phaidon 1979.

Physiology of early stages of visual perception (in which many neurons with small receptive fields converge on a few neurons with larger and larger receptive fields) suggests that the process of organisation of the form proceeds in a single direction, from bottom upwards. This process seems to be characterized by a certain progressivity and linearity. It becomes “bidirectional” if we include inferences with which memory completes the picture from above: perception builds some clues from the bottom, the clues are completed by memory from above. Memory previews can be confirmed or denied by the new clues built up by perception. 8

22

Gori, Practice and Theory of Visual Representation

struction and organisation of the selected object, similar to that of perception of the whole visual field (probably using two parallel and interdependent channels to build figure and background9). Just as perception breaks down the visual field into a thousand specific particles of reality, so the artist breaks down an organised and recognised object into more little and elemental forms. These are then organised by the artist into larger objects through procedures that simulate the perceptive process at a conscious level. One of the most important rules of life drawing is precisely to “forget” the work of organisation and the meaning of memory10 to focus on the deconstruction and reconstruction of the pure form of the selected object: a head, a body, the whole scene.

Fig. 4 The artist looks at a 3D world that is already organised. He/she selects the object of representation, distributing his attention over the whole scene, or focusing on an object or part of it.

Another important unwritten rule of life drawing is that the larger and more complex is the object that the artist intends to represent, the more he/she must simplify its parts, making the task “compatible” with the limited processing resources of the conscious unit. According to the practice and theory of visual representation, our perception uses the same procedure to keep the complexity of representation low and constant: Goodale, MA, Milner, AD (1992): Separate visual pathways for perception and action. Trends Neurosci. 15, 20–5. 9

Betty Edwards, a teacher of drawing and the author of the bestselling book “The New Drawing with the Right Side of the Brain”, suggested turning the model to better see the shape and forget the meaning, limiting the action and memory conditioning. This method has enabled many thousands of people to quickly learn to draw. 10

23

GESTALT THEORY, Vol. 38, No.1

the larger and more complex is the selected object, the more perception detects only the most important variations, standardising the less important ones. In this way, the artist and perception maintain a low and constant complexity in every step of visual representation. • Creation of the linear units. • Relationships between the linear units. • Closing of the parts and their relationship. • Gestalt representation of whole. Creation of the linear units

Fig. 5

Practice The ability to organise adjacent stimuli with similar features (colour, orientation, direction and speed of motion) is the basis of the first stages of visual perception and also of the pratice of representation. To represent an object’s border, an artist drawing from life has to identify the main elements of continuity in the real scene. Joining adjacent segments with similar orientation in longer lines, artists enact a conscious extension of the Gestalt grouping laws of similarity, proximity and good continuity. There is no other way to create a line that does not exist in reality. Theory The way in which the artist draws a line is mirrored at the neurophysiological level11 in the early stages of perception where a series of cells detects adjacent Agreeing with W. Köhler’s isomorphism principle, whereby each event occurring in the phenomenal field of an individual must correspond to another event structurally identical to this first event at the physiological level in the same individual. 11

24

Gori, Practice and Theory of Visual Representation

dashes with similar orientation and sends their results to a more complex cell that organises a longer portion of border, in agreement with the Gestalt grouping laws of similarity, proximity and good continuity. Relationships between the linear units

Fig. 6

Practice To understand and process the relationships between two linear units, artists must necessarily transform the two “visual” units into two units of measurement, which make it possible, for example, to calculate reciprocal sizes and orientation. When drawing, we become aware of the fact that a “language of geometrical calculation” is required to describe the relationship between two visual units. An example is the orientation between the base and height of the fountain in the image 3.2, a relationship which is described in a simple and stable way as an angle of a given size. Like each single visual neuron, the conscious unit of the artist can create only one relationship at a time, giving a more economical result than the sum of its parts. Theory Two simple cells in the primary visual cortex detect two portions of edge and the respective orientations, sending both results to a complex cortical cell that processes an angle of a given size. Using its “parallel architecture”, visual perception constructs increasingly larger relationships between the linear units using a processing language that represents the natural matrix of the geometrical calculation used by the artist.

25

GESTALT THEORY, Vol. 38, No.1

Closing of the parts and their relationship

Fig. 7

Practice Applying a conscious extension of the Gestalt law of closure, the artist completes linear units by enclosing them in simple shapes, such as circles or rectangles representing the main parts of the real subject. By eliminating unnecessary details, he/she can better calculate, describe and correct the object’s proportions, as well as their relationship to each other in space. Theory Like the artist, perception also tends to complete the linear units, creating closed shapes, such as circles, squares and rectangles. It has been observed that some cells of the cortical V212 respond to (complex combinations of orientation and) illusory borders, probably helping to complete the items that present occlusions. The purpose of the brain is always to represent and process reality through simple combinations of a few elementary forms (lines, rectangles, circles), to describe single forms and their whole relationship in an economical and stable format13. Von der Heydt, R., Peterhans, E. & Baumgartner, G. (1984): “Illusory contours and cortical neuron responses”, Science 244 (4654): 1260–1262 12

To perform these operations, the visual brain uses a neuronal elaboration code that becomes more developed as it approaches the conscious unit, creating more and more abstract descriptions that allow it to represent and process wider and wider and more complex aspects of reality. If each neuron is intrinsically able to describe, process and communicate aspects of reality through a code, the human ability to use symbols and words cannot in my opinion be strictly localized in specific brain areas, but should be considered an integral part of a unique process that leads us to produce images and actions. 13

26

Gori, Practice and Theory of Visual Representation

Representation of gestalt.

Fig. 8

Practice Once he/she has sketched the shape and proportions of the chosen subject, for example, the entire fountain, the artist repeats the process of representation of the inner parts. The latter are represented as independent units that can be in turn broken down into sub-units until the level of detail required for the representation is reached. The overall unity is represented firstly because it is a guide, helping the artist to draw the individual parts in the correct relationship to each other, without altering the representation of the whole. If, instead of beginning with the whole, the artist starts from the parts, he/she will be able to represent a series of details that are convincing in themselves, but wrong for the overall drawing. By starting from the overall picture and then focussing on the details, the artist can make a more detailed representation, treating each scale of the scene at the same level of detail.

Fig. 9 Different scales of the scene are representend at the same level of detail. When we represent the head it is an object, but if the object of representation is the entire figure the same head becomes a part, and, as such, it should be drafted with a lower level of detail, making every step of representation “compatible” with the limited resources of the conscious processing unit (F.G. 2015). 27

GESTALT THEORY, Vol. 38, No.1

Theory Like the artist, the visual brain represents one “object” at a time (the entire composition, the figure of Neptune, its head) using the same level of detail, regardless of the scale and complexity of the physical object. The result of perceptual processing of form describes an object as a short relationship between few simple/ linear units14. According to the practice and theory of visual representation, this description is formulated through specific neural codes, which represent the basis of our cultural computing and communication languages. Neurons would be prepared to draw and describe the spatial, temporal or causal relationship between things through natural geometric and syntax abilities: “a circle, above a vertical parallelepiped with a smaller horizontal parallelepiped beneath it”, would represent a cultural and approximated translation of how the brain describes the shape of a statue within our memory. When a part or the whole of this description stored in memory corresponds to a result produced by perception, it generates a simplified mental image. The last one is constructed and centered on the external object and transparently overlaps it. This phenomenon occours in a clearly visible way when the real borders are missing, as in the illusory contours in Kanizsa’s triangle.

Fig. 10 According to the theory and practice of representation, the illusory image is the product of all our perceptions, but we notice it with greater or less evidence based on the stimulus features. In Figure a, we perceive “V” with different orientations as a whole triangle, whose borders are somehow drawn by the brain over the image. My opinion is that there is an analogy between the lines “imagined” for a triangle (Figure a) and the virtual line of the Kanizsa triangle (Figure c), in which the effect is simply more pronounced because the actual outline is missing. In Figures b and ”A short description of the relationship of few simple units” is the result of the “only borders” process, a process that builds up a specific output from below, functional for the final recognition. When the memory recognises a description (even “partial”) of the complete form, it completes it from above, adding features and meanings that do not have to do only with the shape of the object, but also with its colour, its material, its movement. That shape is therefore an accomplished but partial result that does not emerge in a regular Gestalt experiment. If you ask the subject: what do you see? He will reply: “the naked man”, which is the name of the whole set of experiences and features relating to the object. I suppose that to have access to the description of pure visual form it is necessary to experience the process leading to its creation. This process can be experienced in life drawing, which requires you to “forget” the overall meaning of the object in order to reconstruct the short relationship between linear units that represents its identity. 14

28

Gori, Practice and Theory of Visual Representation

c the effect of illusory contours appears to decrease. In the overlapping pictures (Figure c) the effect is even more marked because the brain tries to create a depth effect, a sort of shadow, which “lifts” one object from the other.

3. Why a Researcher Studying Perception Should Start to Draw from Life

One of the difficulties in studying visual perception really lies in isolating the pure process of organisation of form (bottom-up) from the top-down process of memory that instantly completes the object, recognising a partial result built by perception. Unlike “perceptive drawing” that is not fully constructed, artistic drawing must be entirely rebuilt one step at a time. For this reason, life drawing may be a tool for tracing the whole process of form perception, just because it requires you to focus on the bare understanding and form building, eliminating, as far as possible, other aspects such as colour or meanings projected on the object from memory15. In this sense, the practice of representation helps the researcher to create an overall awareness upon which the partial data originating from the study of the working of neurons and visual areas may be organised. The gradual increase in data related to neurophysiology of the brain helps and at the same time impedes the processes of organisation of an overall theory of perception. The quantitative problem (too much data = no theory) is fuelled by a second problem related to the quality of communication between the various areas of research into the mind and the brain. The absence of a “common language” makes communication between the various areas of study of the brain difficult, preventing true interdisciplinary cooperation. Can such a fragmented neuroscience, which above all observes local phenomena, understand how perception combines the parts into a “Gestalt”, and then how the perception works as a whole?16 One of the greatest difficulties faced by the artist who represents reality is limiting the influence of memory, focusing on the pure process of perception and form building. 15

Only the practice of drawing from life shows how it is impossible for the conscious brain to represent the whole relation, for example a church, focusing each time on one detail. By doing so, the single parts will be drawn well, but  the overall relation will always be  incongruous with the real model. The parts of an object should be organised beforehand into a few essential units, in this way allowing the artist’s consciousness to understand the whole relation. Only after doing this can the parts be represented with more precision without changing the whole structure. If we assume that visual representation is a knowing and creative process similar to the creation of a theory, we can find analogies between the work of an artist and that of a scientist. While the scientist needs to organise the results of many local observations into a coherent whole, unlike the artist, the scientist cannot observe a model of the real whole into which he/she can organise these parts. Researchers thus have several possible ways to organise congruent combinations with the available theoretical data, but no certainty of which is the right one. The different computational models show how it is possible for a computer to perform visual tasks in several ways. Thus far, none of these artificial systems is as efficient as human vision. 16

29

GESTALT THEORY, Vol. 38, No.1

It is to be hoped that it will be possible to understand how perception functions, merely by observing local phenomena, pending an increase in the partial results and their organisation into a few macro areas of knowledge and finally into a whole theory. However, the latter would emerge faster if the local neurophysiological process were driven by an observation of the perceptive process from above, even if “imprecise” and mediated by the consciousness. The aim of this paper is to introduce the experience of drawing as a means for visualising the perception process, encouraging researchers to ask new questions and form new hypotheses, starting from the practice of representation. For example, the fact that an artist must perform a geometrical calculation to determine the stable relationship between two lines poses a question: do single neurons, even in the initial stages of visual processing, use an elementary language to process and describe the stable ratio between two luminous stimuli? Let us take a centre-on ganglion cell specialised in detecting small dot-like light variations: the receptive fields of these neurons are made up of a centre and a circumference, both able to detect the luminosity relating to a small area of the visual field. When the intensity of the light detected is equal at the centre and around the circumference, the latter is not activated, since there is no variation in the light.

Fig. 11 Diffuse light: the cell is not activated. Different lights: the cell sends a signal.

When the intensity of the light hitting the centre is different from that hitting the circumference, the cell detects the difference and is excited, transmitting an action potential that contains the information pertaining to the stable relationship between the two light intensities. A description based on a stable relationship has two enormous advantages. The first is that it makes it possible to recognise the identity of an object or a relationship between the notes of a melody, even in conditions different from those in which we previously perceived these stimuli. The second advantage is that a stable relationship is always “more economic than the sum of its parts”, for example permitting us to describe two perpendicular lines as a right angle. According to the practice and theory of visual representation, the form and therefore the identity of any object is recorded in our memory as a short and stable description of something, like the sketch of an artist. The existence of these 30

Gori, Practice and Theory of Visual Representation

mental schematic representations is suggested by the effectiveness of caricatures or stylisations. Reducing each object to a few essential monochromatic traits, the artist helps perception to reduce the object to our innate format of mental representation, which looks just like a drawing description. My hypothesis is that it is precisely the coherence between the external stimulus and the internal representation that generates the aesthetic experience related to the form; the more coherent is the physical stimulus with the receiving system, the better will be the relationship between processing cost and aesthetic benefit, that at the level of each single neuron results in a higher discharge frequency. The picture of a known subject may have too much changing information that hinders recognition (light, shadow, angle) and at the same time too little stable information useful for recognition (the relationship between the nose and eyes may be altered or hidden). For this reason, the picture of a person may require a great deal of elaboration by the brain and several corrections to activate the corresponding internal representation, generating a low aesthetic benefit. On the contrary, the stylization of the same subject, achieved with a few strokes by a skilled portrait painter, will be more coherent with the internal representation, and in general with our system of shape representation designed to detect borders.

Fig. 12 Perception of an angle, according to the physiological model of progressive convergence: several neighbouring ganglion cells detect a line as a series of dots and send these results to a single cortical neuron that is organised in detecting a border portion on a special inclination. Several neighbouring parts of border “converge” on their own on a single complex cell that detects a bigger line and so on, to build bigger and bigger parts of the form. F.G. (2014).

31

GESTALT THEORY, Vol. 38, No.1

The perceptive effectiveness of outline drawings is due above all to the use of lines that assist the outline detection mechanism, upon which the whole process of perception of forms relies. The first layers of ganglion cells break down the image into millions of dot-like light variations that are organised into small edges, then into larger and more complex outlines in the upper layers of the visual cortex. 4. Increasing Differentiation

According to the practice and theory of representation (from now on p.t.r.), different scales of a subject are elaborated and represented at the same level of detail (see Figure 9), making the task compatible with the limited processing resources of the conscious unit. This assumes that the visual brain, like the artist, is able to scale and change the processing precision of the same stimulus, depending on whether it is the object of the representation or a part of a larger object. My conjecture is that the processes of organisation in the early stages of form visual perception do not respond only to the intensity and layout of the external stimuli, but are influenced by the level of attention that makes the response of the cells involved more or less accurate. Depending on the quantity of attention that they receive, a specific neuron will provide more or less precise descriptions regarding the same stimulus, which for this reason will be perceived as more or less different from another stimulus17. What happens if you consider the attention variable within the Gestalt similarity law, according to which close elements are grouped according to mutual resemblance? These elements will not only be grouped according to their shape and their physical arrangement in space, but also based on the level of attention that makes us perceive the internal parts of the configuration and their relationship as more or less accurate, more or less similar to each other and, as such, more or less subjected to the law of similarity. The role of attention doesn’t emerge in the presence of very regular artificial configurations, such as those used to show the law of similarity, since, as we focus, the differences in distance or shape between individual stimuli remain imperceptible.

For example: with a low attention level, the output of the neuron (A) could be: “edge with an orientation of about 10 degrees”. This description could be equal to the result expressed by the neighbouring neuron (B). With a higher attention level, the output of the neuron (A) could become “edge with an orientation of 13 degrees”. This description becomes different from the result of the neighbouring neuron (B) that describes an edge with an orientation of 11 degrees. 17

32

Gori, Practice and Theory of Visual Representation

Fig. 13

Using figures such as top left Gestalt Psychology has shown that similarity, equality and continuity are at the basis of the processes of perceptual organisation. However the natural environment has often a low level of regularity (Figire top right), so it may be more problematic to organise complex and irregular configurations in an economic, invariant and understandable description of the reality. My opinion is that the artist and the visual brain had to develop a flexible mechanism that uses attentional variables to augment similarity, equality and continuity, standardising minor irregularities. In this way it becomes more simpler to arrange every stimulus or configuration in a description, which is always compatible with available cognitive resources. According to the principle that I have called increasing differentation the level of attention modulates our accuracy to detect and describe differences and discontinuities, among internal parts of a stimulus and between two stimuli. If we look at the two configurations below with little attention, we can say that we see two irregular lines, formulating a generic description that makes the two lines equal to each other.18

Fig. 14

Looking more carefully, we realize that the first configuration shows major breaks at the beginning, while the second becomes gradually irregular towards the end, giving rise to two descriptions that make the two lines different. According to the principle that I have called increasing differentiation,19 the atAccording to p.t.r. the brain deteminates the level of similarity between the shape of two stimuli by detecting the amount degree of correspondence between their propositional descriptions. 18

My opinion is that the principle of increasing differentiation governs the interaction between the two main perceptual skills of brain organisation: the first creates relationships based on the difference between two stimuli, the second combines the stimuli based on their correspondence, reconstructing the entire performance starting from the evidence of matches. This last quickly completes the partial results of the perception based on correspondences with existing representations in the memory. If, on the one hand, heuristic shortcuts allow us to save processing resources at all levels, on the other, they may lead to a mistake, for example, making us perceive 19

33

GESTALT THEORY, Vol. 38, No.1

tention level determines the precision with which we perceive dashes, lines and objects, influencing the way we create categories, principles and concepts. At all levels of processing, with more attention we perceive the differences more and the similarities between the inside parts of the object less strongly, creating a more detailed whole that is less likely to correspond to other ones within our short- and long-term memory. On the contrary, by decreasing attention, we activate more heuristic strategies based on correspondence, which require a lower level of processing at all levels20. In general, the perception of inconsistency, an intrinsic discontinuity between parts of the object or between the whole object and memory, bears a higher processing cost than a coherence stimulus. A familiar face consistent with memory, or a form in itself coherent, requires less energy in order for it to be perceived and activated. For this reason they produce a greater aesthetic benefit that at the level of the single neuron results in a higher discharge frequency for the same attentional available energy.21 According to p.t.r., the level of aesthetic experience is inversely proportional to the amount of energy used by the entire cognitive system to perceive a shape, or more generally to perform a task22. My general hypothesis is that the attention level modulates the accuracy of all processes involved in the perceptual task (including the activity of single cells from the cortical area V1), and in this way it differs from the main theories on attention23. While the effect of attention on conscious human activities has been widely studied, there is little evidence for the relationship between attention and the early stages of visual perception, such as V1 and V2, in scientific literature. The prevailing approach among researchers is not to consider the role of attention in the early stages of human visual perception two different stimuli as equal. I recently happened to go out for a coffee taking my remote control instead of my wallet, both dark objects similar in size and shape, because my focus was elsewhere. Another point in favour of the principle of increasing differentation comes from the most famous test on selective attention (www.theinvisiblegorilla.com/videos.html) in which we are asked to count the steps between the basketball players. While we are engaged in this task we don’t notice the gorilla that passes through to the players! My interpretation of this result is that actually we “see” the gorilla, but we process it only partially, making general features emerge similar to other dark human figures. We would have noticed the gorilla if this had been of a different colour from the other players. When the difference between perception and prediction exceeds a certain threshold, the brain realizes the “mistake” and asks for more attention, allowing us to detect more internal differences to the stimulus, which will be recognised more accurately or recorded in memory. At the more local levels, specialised neurons unite equal stimuli, permitting us to make forecasts based on probable continuity, creating virtual lines to complete outlines with gaps or obstructions. 20

Activating a representation by making a prediction still has a cost that lowers the aesthetic benefit: the process of completing a circle requires processing costs higher than a stimulus of a full circle, that generates a better aesthetic experience. 21

Paradoxically, also the action itself to lower the energy, for example to reduce the irregularities of a stimulus, constitutes a change of state that requires energy, contributing to worsen the whole aesthetic. 22

The main theories on attention state that a certain part of the vsual process is not affected by the attentional variable: Broadbent, D. (1958), Perception and Communication. Deutsch J A & Deutsch D. (1963), Attention: some theoretical considerations. Treisman A.M. (1964), Selective attention in man. 23

34

Gori, Practice and Theory of Visual Representation

because, as an unconscious process, it is thought to be governed mainly by personal automatic reactions. However, while it is true that attention allows us to select certain stimuli or aspects of reality and to inhibit others, it is likely that it also affects the activity of individual cortical cells that produce conscious perception. The following image shows how the distribuition of attentional resources influences the process of perception of the form, modulating the precision with which individual neurons respond to specific stimuli.

Fig. 15 Propositional cathedral (F.G. 2014).

If we decide to observe the entire cathedral, the level of available attention is distributed over the whole and consequently decreases on the individual parts, which are therefore processed to a lesser degree, causing the words with which the cathedral is built to serve the function of simple irregular linear units. If, on the contrary, we decide to concentrate our available attention on one part of the cathedral, we are able to better process the relationships within the individ35

GESTALT THEORY, Vol. 38, No.1

ual words, so that the rest of the cathedral is partially obscured and is processed at a lower attention level. (All neurons which process the area outside the main attention focus build a less precise and less aware level of representation that is perceived as background). A recent study shows how attention influences firstly the neurons in the upper visual areas such as V4 and V3 and only afterwards the V1 cells24. The fact that attention follows a path  that is the reverse of that of perception suggests that the conscious unit looks at the world through a 3D screen that is already organised by perception, shifting the focus of attention to an object and enhancing the processing ability of the neurons involved. Is there any neurophysical evidence to support the conjecture that attention plays an active role from the first stages of perceptive organisation? Studies on the physiology of the visual cortex of the macaque have shown that attention has little or no influence on the cells of the first visual areas (V1 and V2). These results suggest that the main task of the primary visual cortex is to automatically detect and organise the elements of continuity within visual configurations. However, a growing number of research projects conducted with fMRI techniques have subsequently shown that the activity of the human visual cortex is strongly influenced by the level of attention, which changes the levels of excitation of the individual cells.25 If the level of attention influences the activity of a specific neuron, this neuron will probably give a more or less precise response to a specific stimulus. The more accurately something is described, the more dissimilar it will become to something else, for example to the response of the neuron near to it. Different responses are thus processed as a stable relationship. On the contrary, if two similar stimuli are described without much accuracy, they will prove to be “more equal” to each other than they actually are, thus yielding to the “power” that makes continuities uniform, grouped more by the Gestalt law of similarity. Thanks to this dynamic process, millions of stimuli are organised by perception into descriptions and then into stable and simple representations, compatible with the limited resources of the conscious unit. It can use the results of perception according to our purposes, to understand and represent aspects and concepts larger than those described by perception. The ability of scientists to identify a principle common to many similar events Buffalo, E.A., Fries, P., Landman, R., Liang, H., and Desimone, R. (2010): A backward progression of attentional effects in the ventral stream. Proceedings for the National Academy of Sciences. 107(1), 361-365. 24

Lee, J (1 August 2010): The Effect of Attention on Neuronal Responses to High and Low Contrast Stimuli. J. Neurophysiol. 1 August 2010, 960-971. 25

36

Gori, Practice and Theory of Visual Representation

and that of a creative person to arrange a new relationship between different pieces of knowledge are examples of how “equality” and “difference” act at a high level, allowing us to extend the work of organising perception. Generalising, we could say that during the course of cultural evolution, human consciousness has identified and developed some innate procedures that perception uses at an unconscious level. In this sense, grammar, syntax, geometry, drawing and perspective may be considered as cultural extensions of the processes with which our perceptive system elaborates, describes and represents the space-time relationship between different items of knowledge. And thanks to this isomorphism between conscious and unconscious activities, we can study the visual perception of form through life drawing. Summary Modern neuroimaging technologies allow us to observe the responses of individual neurons to specific stimuli, but they cannot show the “creative” process whereby the brain integrates this infinity of results into a “Gestalt”. According to the author, there is only one way to “observe” the process used by the brain to detect points, organize outline units and finally recreate a simple and stable representation of form: by learning and practicing the technique of drawing from life, visualising and describing, one step at a time, the process whereby the brain recreates the form of reality, a process that perception performs instantaneously thanks to its parallel architecture. With the aid of the practice and theory of visual representation, the author succeeded in identifying the “increasing differentiation” principle, according to which attention modulates the precision of the response of individual neurons, influencing the whole process of form organisation. Keywords: Gestalt theory, neuroaesthetic, art, drawing, visual perception, neural coding, theory of attention, Semir Zeki. Zusammenfassung Moderne, neuro-bildgebende Technologien ermöglichen zwar die Beobachtung der Reaktionen einzelner Neuronen auf bestimmte Reize, können aber den „kreativen“ Prozess, durch den das Gehirn diese unendlich vielen Ergebnisse in eine „Gestalt“ integriert, nicht zeigen. Nach Auffassung des Autors gibt es nur eine Methode, den Prozess, mit dem das Gehirn Punkte erfasst, umrissene Einheiten organisiert und schließlich eine einfache und stabile Darstellung der Form der Realität nachbildet, zu „beobachten“: Die natürliche Technik des Ab-Zeichnens, nämlich Visualisierung und Beschreibung, Schritt für Schritt zu Erlernen und zu Praktizieren, einen Prozess, durch den das Gehirn die Form der äußeren Realität nachbildet und den die Wahrnehmung dank ihrer parallelen Architektur unmittelbar vollzieht. Mit Hilfe von Praxis und Theorie der visuellen Darstellung gelang es dem Autor, das Prinzip der „zunehmenden Differenzierung“ zu identifizieren, nach dem die 37

GESTALT THEORY, Vol. 38, No.1

Genauigkeit der Reaktion einzelner Neuronen durch Aufmerksamkeit reguliert wird, die so den gesamten Prozess der Organisation von Form beeinflusst. Schlüsselwörter: Gestalttheorie, Neuroästhetik, Kunst, Zeichnung, visuelle Wahrnehmung, Neuro-Codierung, Aufmerksamkeitstheorie, Semir Zeki.

References Arnheim, R. (1954): Art and visual perceptìon: a psychology of the creative eye. Berkeley and Los Angeles: University of California Press. Arnheim R. (1966): Toward a Psychology of Art. Berkely and Los Angeles: University of California Press. Arnheim, R. (1969): Visual Thinking. Berkeley and Los Angeles: University of California Press. Arnheim, R. (1998): Die Intelligenz des Sehens. Neue Bildende Kunst, 56-62. Black, H., Gombrich E. & Hochberg, J. (1972): Art, Perception and Reality. Baltimore: Johns Hopkins University Press. Broadbent, D. (1958): Perception and Communication. London: Pergamon Press. Buffalo, E.A., Fries, P., Landman, R., Liang, H. & Desimone, R. (2010): A backward progression of attentional effects in the ventral stream. Proceedings for the National Academy of Sciences. 107(1), 361-365. Chabris, C. & Simons, D. (2010): The invisible gorilla: and other ways our intuitions deceive us. New York: Crown. Chatterjee, A. (2014): The Aesthetic Brain: How We Evolved to Desire Beauty and Enjoy Art. New York: Oxford University Press. http://www.theinvisiblegorilla.com/ Deutsch, J.A. & Deutsch, D. (1963): Attention: some theoretical considerations. Psychol Rev. 70, 80-90, Stanford University, Stanford, CA. Duchamp, M. (1957): Il processo creativo. Art News, 56 (4). Edwards, B. (1999). The New Drawing on The Right Side of the Brain. Putnam: Penguin. Fodor, J.A. (1983): The Modularity of Mind: An Essay on Faculty Psychology. Cambridge, MA: MIT Press. Gardner, H. (1973): The arts and human development. New York: John Wiley & Sons. Gombrich, E. (1961): Art and illusion. Princeton, N.Y.: Princeton University Press. Gombrich E. (1979): The Sense of Order. A Study in the Psychology of Decorative Art. Oxford: Phaidon. Goodale, M.A. & Milner, A.D. (1992): Separate visual pathways for perception and action. Trends Neurosci. 15, 20–5. Gregory, R.L. (19975): Eye and Brain: The Psychology of Seeing. Oxford: Oxford University Press. (First publication 1966, London: Weidenfeld & Nicholson). Huston, J.P., Nadal, M., Mora, F., Agnati, L.F. & Cela-Conde, C.J. (eds)(2015): Art, Aesthetics and the Brain. Oxford: Oxford University Press. Kandel, E. (2012): The Age of Insight. The Quest to Understand the Unconscious in Art, Mind, and Brain, from Vienna 1900 to the Present. London: Random House. Kanizsa, G. (1991): Vedere e pensare, Bologna: Il Mulino. Kanizsa, G. (1997): Grammatica del vedere, Bologna: Il Mulino. Köhler, W. (1969): The Task of Gestalt Psychology. Princeton, NY.: Princeton University Press. Koffka, K. (1922): Perception: An introduction to the Gestalt-Theorie. Psychological Bulletin, 19, 531-585. Koffka, K. (1935): Principles of Gestalt psychology. New York: Harcourt, Brace, & World. Kosslyn, S. (1986): Image and Mind. Cambridge, MA: Harvard University Press. Lauring, J.O. (2014): An introduction to the neuroaesthetics: The neuroscientific approach to aesthetic experience, artistic creativity, and arts appreciation. Copenhagen: Museum Tusculanum Press. Lee, J. & Maunsell, J.H.R. (2010): The Effect of Attention on Neuronal Responses to High and Low Contrast Stimuli. Journal of Neurophysiology 104(2), 960-971. Leonardo Da Vinci: Trattato della pittura, Brancato editore, Catania (1990). Livingstone, M. (2002): Vision and Art: The Biology of Seeing. New York: Harry N. Abrams. Marr, D. (1969). “A theory of cerebellar cortex”. Journal of Physiology 202 (2), 437–70. Marr, D. (1982): Vision. A Computational Investigation into the Human Representation and Processing of Visual Information. Cambridge, MA: The MIT Press. Massey, I. (2009): The neural imagination. Aesthetic and neuroscientific approaches to the arts. Austin: University of Texas Press.

38

Gori, Practice and Theory of Visual Representation Mather, G. (2013): The Psychology of Visual Art. Eye, Brain and Art. Cambridge, UK: Cambridge University Press. Minissale, G. (2013): The Psychology of Contemporary Art. Cambridge, UK: Cambridge University Press. Minsky, M. (1988): The Society of Mind. New York: Simon and Schuster. Moutoussis K. & Zeki S. (1997): A Direct Demonstration of Perceptual Asynchrony in Vision. Proceedings of the Royal Society of London, 264, 393-399. Paltoglou, E. & Neri, P. (2012): Attentional control of sensory tuning in human visual perception. Journal of Neurophysiology, 1260-1274. Pizzo Russo, L. (1988): Il disegno infantile. Palermo: Aesthetica edizioni. Solso, R.L. (1994): Cognition and the visual arts. Cambridge, MA: A Bradford Book. Treisman A.M. (1964): Selective attention in man. British Medical Bulletin, 20, 12-16. Von der Heydt, R., Peterhans, E. & Baumgartner, G. (1984): “Illusory contours and cortical neuron responses”, Science 244 (4654), 1260–1262. Wertheimer, M. (1982): Productive Thinking. Chicago: University of Chicago Press. Zeki, S (1994): The neurology of kinetic art. Brain, 117, 607-636. Zeki, S. (1995): Balthus ou La quête de l’essentiel. Paris: le Belles lettres, Archimbaud. Zeki, S. (1999): Inner Vision. An Exploration of Art and the Brain, Oxford: Oxford University Press. Zeki, S. (2001): Artistic creativity and the brain. Science 6, Vol. 293, no. 5527, 51-52. Francesco Gori (b.1970), researcher and visual artist, has worked as Creative Director for several international agencies, realizing communication projects for major brands. He uses creativity work and drawing experience to study the creative process of perception from a different point of view. Francesco Gori is also teaching theory and practice of simplification at the New Academy of Beauty Arts in Milan. Address: 5hort srl, via Cappuccio 16 – 20123 Milano, Italy. E-mail: [email protected]

39