During the seventeenth century, Sir Issac Newton performed an experiment in which he passed white light through a prism causing the different wavelength composition of light contained within to refract at distinct angles producing an array of colors which has since been described as the visible spectrum. He noted that long-wave length light produced a sensation of red, middle-wave length light a sensation of green, and short-wave length light a sensation of blue, and concluded that "every body reflects the rays of it's own color more copiously than the rest and from their excess and predominance in the reflected light has it's color" (Ladd-Franklin, 1973). It can, however, be argued that color perception is a totally psychological experience produced by the effect that reflected light of specific wavelengths has on both the nervous system and brain of many species. A great deal of research concerning the physical properties of light and the physiological responses of organisms to the distinct properties of light have been conducted in order to acquire a more complete understanding of color and it's perception. One basic conclusion of such research is that "color sensations are related in consistent and measurable ways to the physical features of light" (Arnheim, 1954).

There are three major physical characteristics of light which affect color sensation: wavelength, intensity, and spectral purity. The visible spectrum ranges from approximately 380 nanometres(nm) which is a short-wave length light, to 770 nm which is a long-wave length light. The corresponding color sensation of wave length is hue, and is determined by the predominant wavelength reflected from a given surface, such that a surface will have a blue hue provided the predominant wave length reflected is roughly 473 nm. It should be noted that only under strict laboratory conditions can a surface be made to reflect only one wave length, in which case it can be said to be spectrally pure (Schiffman 1990).

Schiffman (1990) maintains that the spectral purity of a light bears a simple relationship to the psychological color sensation of saturation, such that a saturated color contains only one wave length and a desaturated color contains more than one wave length with the hue corresponding to the dominant wavelength. To illustrate, a surface which reflects ten percent long-wave light, five percent middle-wave light, and eighty-five percent short-wave light is a desaturated light with a bluish hue.

The third major characteristic of light is intensity which is related to the sensation of brightness. Essentially, with an increase in intensity, there is a uniform increase in all the wave lengths reflected, causing the hue to become brighter. In contrast, with a decrease in intensity, each wavelength is decreased proportionally and the appearance of the hue will be darker until the intensity level is reduced to zero in which case no wave length is reflected and the resulting color sensation will be the absence of any wavelength composition....black. Again it is necessary to reiterate that a uniform increase in intensity across all wave lengths is only possible under laboratory conditions (Schiffman 1990).

Under normal viewing conditions, the wavelength composition of light reflected from a given surface varies continuously depending on the illuminant used. If one were to measure the wavelength composition of the light reflected from an object (i.e. an apple) either during the day or at night; or under a high pressure sodium, fluorescent, or metal halide lamp; one would find different amounts of long, middle, and short-wave light reflected in each viewing condition. Although the wavelength composition reflected back in each condition is unique, the color of the apple remains red, even though the shade of red is subject to change. Were it not for this phenomenon, known as color constancy, color vision would lose it's significance as a biological signalling mechanism (Pickford, 1951). To illustrate, if a tiger and it's surrounding environment were to change color depending upon the illuminant in which they were viewed, neither would be recognizable by it's color, nor would one be able to identify or distinguish between the two solely based on color. The only physical constant available to the nervous system that allows it to construct the constant hue of a surface from the continually changing illuminant conditions is reflectance.

Reflectance refers to the percent of light of any given wave length that a surface reflects. For example, a particular red surface may have a reflectance of seventy-five percent long-wave light, fifteen percent middle-wave light, and ten percent short-wave light. If one were to shine one hundred watts each of long-, middle-, and short-wave light on that surface, it would reflect back seventy-five, fifteen, and ten watts respectively of long-, middle-, and short-wave light. If, on the other hand, one were to shine two hundred watts each of long-, middle-, and short-wave light on that surface, it would reflect back one hundred and fifty, thirty, and ten watts respectively of long-, middle-, and short-wave light. This illustrates that the reflectance (the percentage of each wave length reflected) of a particular surface does not change even though the amounts of each wave length reflected are different in the two conditions. Therefore the human nervous system "must make itself as independent as possible of the fluctuations in the amounts reaching it and approximate its constructs as nearly as possible to the true physical constants of surfaces in nature, i.e. their reflectance" (Zeki, 1990).

Land's experiments suggest that the nervous system accomplishes this independence by comparing the intensity of the reflected light of one wavelength from one surface with the intensity of the reflected light of the same wavelength from the surrounding surfaces. This process occurs independently for each wavelength and are followed by a comparison of these comparisons.

Zeki (1990) maintains that the responses of cells at various stages along the pathways from the retina to the primary visual cortex suggest that the comparisons that are necessary to create color occur in the higher visual areas. This is because the cells along the retino-geniculo-cortical pathways are responsive to only a very small part of the field of view and are not affected by what occurs outside their receptive field. Before examining the neural pathways involved in the construction of color, it is necessary to discuss other theories of color perception.

According to Schiffman (1990), the trichomatic receptor theory, also called the Young-Helmholtz theory, suggests that there are three types of receptors within the retina each maximally sensitive to the wavelengths that correspond to the hues of red, green and blue. More specifically, there are three types of photopigments that are segregated in three types of cones, and each of these photopigments absorb light of several different wavelengths but absorbs more of one specific wavelength than the others. Studies such as those conducted by Marks, Dobelle, and MacNichol (1964) indicate that there are three major groups of cones whose photopigments have a maximum absorption at approximately 447nm, 530nm, and 563nm. The more the photopigment of a cone absorbs a given wavelength, the more sensitive that cone is to light of that wavelength. For example, as one views a red surface, long-wave light is reflected back in greater amounts than middle-wave or short-wave light producing a chromatic experience of red due to the strong excitation of the photopigment in the cones that are maximally sensitive to long wavelength light combined with a weaker excitation of the photopigments in the two cones that are maximally sensitive to middle-wave and short-wave light. The Young-Helmholtz theory is able to account for all the hues in the visible spectrum by the appropriate proportional contribution of these three receptors, yet does not account for the physiological structures beyond the retina such as the opponent cells of the retino-geniculo-cortical pathways.

Cells along the retino-geniculo-cortical pathway tend to have small receptive fields with a simple structure. The most common type of receptive field structure is the centre-surround, in which the stimulation of one part of the receptive field (the excitatory part) results in an increase in cell firing (centre-ON) whereas stimulation in another part of the receptive field (the inhibitory part) results in electrical activity only when the stimulus is turned off (surround-OFF). There are a number of different kinds of centre-surround cells with respect to input they receive from the three cone types (Fletcher, Volke, 1985).

The first type, known as an antagonistic centre-surround receives input from all three types of cone in both the centre and surround, therefore if one were to illuminate the entire field the excitation would be cancelled out by the inhibition and there would be no response from the cell. Because these cells receive input from all three cones they are unable to signal differences in the intensity of a given wavelength between one part of the receptive field and another. They may however, be able to explain why the after image of black is white, or the reverse.

Other cells in the retino-geniculo-cortical pathway that are not organized in the centre-surround fashion give a response of one kind (i.e. ON) to, say, long-wave light and the opposite response (i.e. OFF) to, say, middle-wave light, to both parts of the receptive field. These cells could account for why the after image of red is green or why blue and yellow cancel each other out.

The remaining centre-surround cells receive input from just one kind of cone in their centre, and from another in their surround. These cells are closely tied to the Opponent-Process theory which implies that each receptor is capable of two kinds of responses that are mutually opposite to each other. The three mutually opposite response types of cells suggested are: red-green, blue-yellow, and black-white. This theory postulates that a receptor can respond in only one of two ways; that is, red or green, and yellow or blue is the chromatic experience, not red and green, or yellow and blue. Once again, this is an example of a theory that enlightens one to a degree about the mechanisms involved in color perception yet does not go deep enough to develop a full understanding of how the human mind constructs color.

Zeki (1990) states that when one records from single cells in the different areas in the prestriate cortex, one finds that the cells of different areas respond best to different kinds of stimuli including form, color and motion. As it turns out, area V5 is specialized for the detection of motion in the field of view whereas the color of the moving stimulus is not important for them, nor is form. By contrast, cells in V3 and V3A are orientation selective, responding to lines of a certain orientation but not to others. Here again, the color of the oriented line is not important. More central to the topic, in the V4 areas, there are heavy concentrations of cells which are responsive to wavelength showing that area V4 is specialized for processing information relating to color.

Zeki (1990) also suggests that "different prestriate visual areas receive their inputs from subregions of V1 which are specialized for the corresponding visual function. Thus V1, in addition to its other functions, acts as a segregator and distributer of different visual areas for separate, and parallel, processing."(Zeki, 1990). The segregation of which Zeki speaks can be demonstrated by staining the striate cortex with cytochrome oxidase which highlights areas of high metabolic activity when the receptive fields of a cells are exited. It was discovered that wavelength stimulation produced high metabolic activity in layers two and three of the striate cortex. Cells in these layers have since been named 'Blobs'. Cells inside the blobs are not orientation selective, and about half are wavelength selective. Cells outside the blobs are orientation selective but the large majority are not wavelength selective. This evidence would suggest that the pathways for color and form are segregated within V1 (Zeki,1990). From the blobs of V1 (the striate cortex) the neural connection travel to the thin stripes of area V2 of the prestriate cortex.

Using the same method as above, sections of area V2 are stained for metabolic activity, and the area is found to be characterized by a set of alternate thick and thin stripes which are separated from each other by interstripes. The studies show that the majority of thin stripes are wavelength selective and are not concerned with the orientation of the stimulus. Some compartments of V4 (our color centre) receive their predominant input from the thin stripes, and the majority of cells in them are likely to be wavelength selective (Zeki 1990).

Zeki (1990) points out that it is not surprising that there is a good form related input into area V4, because every form has a color and every color, being confined in visual space, has a form. In addition, contours and Mach bands are critical in generating colors. While the projections from area V4 are predominantly to the temporal lobe and the cortex of the superior temporal sulcus, it also has a consistent projection to the parietal cortex.

Finally, the theory of functional specialization, which supposes that different attributes of the visual world are processed separately into the separate visual areas of the prestriate cortex suggests that these areas should be regarded as the pivotal areas for processing information on form and color and distributing this information to yet higher visual areas. A great deal more research will have to be carried out on these and other areas to develop the full understanding of how the human brain interprets its environment.

REFERENCES

Arnheim, R. (1974). Art and Visual Perception: A Psychology of the Creative Eye. University of California Press: Los Angeles.

Fletcher, R., and Volke, J., (1985). Defective Color Vision. Adam Hilger Ltd.: England.

Ladd-Franklin, Christine, (1973). Color and Color Theories. Arno Press: New York.

Marks, W.B., Dobelle, W.H., & MacNichol, E.F. Visual Pigments of Single Primat Cones. Science, 1964, 143, 1181-1183.

Pickford, R. W., (1951). Individual differences in Color vision. Routledge and Kegan Ltd.: London.

Schiffman, H.R., (1990). Sensation and Perception: An Integrated Approach. Wiley & Sons: New York.

Zeki, S., Functional Secialization in the Visual Cortex. Disscussions in Neuroscience, 1990, 8.

COLOR PERCEPTION

K HOGBEN