How our senses really work is still a deep mystery
A knowledge of perception not only allows us to understand ourselves better, but is essential to the design of systems that interact with humans. Elsewhere on this site, you will find papers on visual perception and colour, and on a comparison of the auditory and visual senses. In this paper I want to review the other senses, taste, smell and touch, which are really the fundamental and ancient ones, and draw some inferences from the similarities and differences between all the senses. A good review can be found in Reference 2.
Seeing and hearing are 'remote' senses that tell us about distant parts of our environment by receiving waves. The other 'contact' senses all involve physical contact with the things that are sensed. They are chemical and mechanical. The chemical senses are the olfactory (smell) and gustatory (taste) sensations. The olfactory sensors are located in yellow pigmented areas on each side of the inner nose. These areas are about 2.5 cm2 in area each, and contain chemoreceptors, which are nerve cells responding to certain chemicals that are carried to the sensors as gases. The detailed functioning of these cells does not appear to be known. The axons of these nerves apparently are received in olfactory bulbs under the front part of the brain, on both sides. Here there are glomeruli, each receiving signals from some 26 000 receptors. The olfactory bulbs on either side are cross-connected. Finally nerve fibres reach the olfactory areas in the anterior lobes of the brain. The types of olfactory sensations are given as fruity, flowery, resinous, spicy, foul, and burned. The olfactory sense is some 10 000 times as sensitive as taste, and is primarily responsible for the flavours of food. There is strong adaptation, in which one soon becomes accustomed to an odour and unaware of it, as well as masking of one odour by another, the theory of perfume. An increase of some 20% in concentration is necessary to cause a perceptible increase in the strength of the perception.
In insects, the olfactory sense seems to be located on the antennae. Snakes and lizards possess a Jacobson's organ in the front of the mouth that is directly connected to the olfactory centre in the brain. The flicking tongue transfers scents to this organ for analysis. Scents seem to have a strong influence on the social interactions of all kinds of animals, from bees and ants to the sexual pheromones of mammals. Birds have a well-developed olfactory sense, which was not appreciated until recently. The scents and colours of flowers are probably not accidental, nor designed for our pleasure, but for very practical purposes of identification by co-operating insects.
Dr Le Fanu, in the Sunday Telegraph (21 May 2000) comments on how vigorously an odour calls up memory. Smell is the only sense with direct access to the amygdala, the 'emotional centre' of the brain. It is a double sense, like seeing and hearing, and the two nostrils control the relative rate of air flow between them to receive slightly different chemical signatures, which allows finer discrimination of odours. The receptors also saturate or accommodate, and this is traded side for side, so that a sensitive receptor is always available.
The gustatory sense is mediated by taste buds, small onion-shaped bags on the papillae of the tongue and elsewhere that contain 50 to 75 sensitive cells each. Liquids can pass through a small pore to reach the sensitive cells. Remarkably, the taste-sensitive cells have a limited lifetime, and are constantly being replaced. The kinds of taste sensation are usually termed sweet and sour, located on the tongue, and sour and bitter, located on the roof and back of the mouth. The "tongue map" so often seen is erroneous, a result of the passion in physiologists for creating spatial specialization where there is none ("brain maps" are a similar delusion). The lingual papillae are quite a various gang, including filliform that are touch sensors, and fungiform, foliate and circumvallate, active in taste. The name of the latter, which are at the rear of the tongue, is misinterpreted in Reference 1 as "wall-like" when really it means "surrounded by a ditch," which they so obviously are.
The taste fibres proceed along several pathways to the medulla oblongata or brain stem, then to the thalamus, and finally to the taste area on the anterior cortex. There are synaptic connections between neighbouring cells, as in the case of vision and hearing. The taste sense exhibits adaptation and masking, like the other senses. Taste and smell are not reliable guides to poisons, only to identification of known substances. Some quite innocuous substances taste terrible, while some poisons can taste delightful. Lead acetate, or sugar of lead, tastes pleasantly sweet, but is a powerful cumulative poison. The aromatic compounds benzene and toluene are fragrant, but benzene is dangerously carcinogenic, while toluene is relatively safe. The chemical senses are sometimes associated with vivid mental images and recollections, showing an unexpected connection to higher mental processes.
Reference 1 mentions that it has been found that taste perception is a result of differences in neural stimulation, and also, with some evident surprise, that different perceptions can arise from the same cells. The idea that salt excites a salty receptor that passes a salty signal through a salty nerve to a salty center in the brain recalls the primitive idea of the retinal cell that sends a signal through its nerve to paint its pixel in the brain. It would be thought that these principles, so evident in the visual and auditory senses, could hardly be surprising here. There is even a protein involved in taste that is analogous to the transducin of the visual mechanism. It should not be surprising that all senses have a common origin.
Touch is one aspect of the important and varied mechanoreceptive senses. Touch, posture or kinesthetic sense, the vestibular or equilibrium sense, and sound all involve sensitive cells that react to a mechanical stimulus. Deformation of the cell causes a change in electric potentials and the initiation of a nerve impulse. Many of these cells have tactile hairs, such as the hair cells of the semicircular canals and the cochlea. Mammals, notably cats, have vibrissae, 'whiskers', that are very sensitive. Subcutaneous receptors, another kind of sensor, seem to possess sensitive nerve endings. Clusters of such receptors make up the pain spots that differ in density over the skin. Two prongs 2 mm apart can be separately detected on the fingertip, but not on the back. The tip of the tongue has some 200 such points per square centimetre, and can detect two points only 1 mm apart. Fishes have a lateral line down their sides that is a very sensitive touch organ, capable of detecting pressure pulses before actual contact takes place. Other receptors are sensitive to heat. Many touch organs communicate with ganglia in the spinal cord, and may be part of a reflex arc that does not involve processing by the brain.
Although touch may seem to involve less mental processing than the other senses, large volumes of the brain are associated with parts of the body, and touch may play a large role, especially in learning and memory. The sensation of pain is closely related to touch, but is obviously a subjective perception like those of sound and vision, involving higher mental processes and consciousness. There is an important sense of equilibrium that uses the semicircular canals in the ear to detect motions of the head, that partly has an involuntary output, as well as an effect on conciousness. The contact senses are important in the development of the infant, especially the visual sense. I believe that motions of the hand and fingers aid learning and the memory. Touch is also subject to 'illusions' that show mental processing is involved. The best-known is probably the pencil between crossed fingers, that is sensed as two pencils. Temperature is often wrongly judged, the sensation depending on contrasts and comparisons. Objects of different materials but at the same temperature may feel variously cooler or warmer.
Our knowledge of all the senses is very incomplete and unsatisfactory, especially with regard to the neural and mental processes that are an essential, perhaps the major, keys to understanding consciousness. Anatomical and physiological knowledge of the structures of the nervous system is detailed and rather complete, but furnishes only the slightest clues to the operation of the senses. Empirical knowledge of how the senses behave is extensive, but it only describes and does not explain. There seem to be few areas of modern science so important and interesting to us in which the fundamental knowledge is so incomplete. The senses should not be studied in isolation from one another, since there are surprising connections as the result of the mental processes of consciousness. The senses do not interact solely with consciousness, but also with subconscious and involuntary responses to the environment.
Most of the sensory cells seem to be descended from ciliated primitive cells that would have been unusually active and became included in associations to take advantage of their responsiveness. The rod and cone cells of the retina have lost all apparent characteristics of these primitive sensory cells except perhaps the overall shape, while the chemical and mechanical sensors retain cilia and hairs because of their functionality. The nerve impulses from these cells do not reach the brain directly, but only through many synaptic connections involving cross connections, coding and processing, that result in complex messages carried by far fewer fibres. These trunk nerves enter intermediate bodies, with connections to both hemispheres of the brain, and pass on their signals to further bundles of fibres to distribute the signals to the cortex and the centres of consciousness, wherever they may be. Sensory perception will not be understood until all these pathways are elucidated. There are no simple senses that directly interact with consciousness. The two hemispheres of the brain appear to share sensory information equally and impartially. In vision and hearing, both halves of the brain are essential to the complete sensation. There is no support whatsoever for the view that there are two brains with different characteristics as far as the senses are involved. The senses also involve the central and old parts of the brain, the brain stem and its associated regions. Consciousness is probably located here, not in the peripheral cortex that seems chiefly devoted to information storage. Simpler psychologists so want the large cortex to be what confers humanity and consciousness that conflicting evidence is overlooked. The functioning of the brain will never be revealed by the scalpel, balance and microelectrode.
The senses cannot be understood except by careful separation of the physical and objective stimulus from the mental and subjective perception. We cannot be directly aware of the properties and qualities of external objects, though our language and thinking often identifies an object with its perception. An object cannot, of itself, be red, nor a solution of sugar sweet: these are essentialy perceptions within ourselves, not properties of matter.
All the senses appear to depend more or less on differences between the states of neighbouring sensors. This is strongest in the visual and aural senses, perhaps weakest in touch, but even here the co-operation of several neighbouring cells is probably necessary to launch a sensation. All senses have the widest dynamic range possible, which is greatest in hearing, and least in touch or taste, made possible by a logarithmic response. All senses communicate only by electrical pulses travelling down nerve axons, and are subject to noise. All senses exhibit adaptation, in which a continued steady stimulus has an effect decreasing with time, as well as masking, in which one stimulus increases the threshold for the detection of another. There is no straightforward, universal connection between the intensity of a stimulus and the strength of its perception. Sensation can judge equality with some precision, but ratios cannot be accurately estimated, even approximately.
Another common property of the senses is shown by Fechner's Law, first recognized in vision. The psychophysical quantity brightness is related to the physical quantity energy flux or intensity logarithmically, not linearly. If you cast a shadow on a piece of white paper in the moonlight, then turn on a bright electric light, the shadow will disappear, even though the difference in illumination between the shadow and its surroundings does not change. Fechner observed that the delicate shadows and contrasts of a cloud were unchanged when the cloud was observed through a dark glass. Experiment showed that a minimum fractional change in intensity was observed as a just detectable change in brightness over a wide range, though failing for illumination that was too strong or too feeble. The old system of classifying stars by magnitude in a uniform series of equal steps of brightness was found to be logarithmic, the ratio of intensities being about 2.5 in each step. The loudness of a sound has a similar relation to the intensity, and remarkably, so does pitch. The even-tempered scale of frequency ratios of 21/12 gives equal steps of pitch, as well as frequency ratios close to those that are found harmonic. Fechner's Law is, of course, approximate and inexact, but still expresses a remarkable property of the senses.
The subjective experiences of the consciousness are created by dynamic mental activity based on sensory information combined with memory, and all of these three factors are necessary. This includes things like colours, smells, tones, flavours, recognition of faces, and pain, and indeed the world we see. All the senses are simultaneously in action, and our consciousness is dominated by perceptions. How these perceptions arise is completely unknown, a mystery. If it were known, we would be able to describe what red is, how salt tastes, and what a major chord sounds like. All we can do now is assume we all have the same sensations and describe things in terms of them. This is probably a pretty good assumption, and we can even detect slight anomalies between individuals when they occur, but it leaves a nagging doubt and a thirst for knowledge.
Composed by J. B. Calvert
Created 30 March 2000
Last revised 17 February 2001