A Consideration of Humphrey's "Cerebral Sentient Loop" Explanation of Consciousness from "A History of the Mind" by Nicholas Humphrey

MAX BENNETT

The evolution of the nervous system may have started a thousand millions years ago with the sponges or just six hundred and fifty million years ago with polyps like jelly-fish and corals. Sponges are extremely simple multicellular organisms (Figure 1A). A section through a sponge, when stained with silver, shows some cells that connect one side of the body wall of the animal to the other (Figure 1B); it has been claimed that these may be primitive nerve cells. With the evolution of the Colenterates, such as hydra, jelly-fish and corals, the identification of nerve cells and muscles is unequivocal: jelly fish have two layers of cells with a jelly-like substance seperating the two, giving the animal some rigidity (Figure 2A). Nerve nets for the control of swimming, tentacle position and feeding are composed of either bipolar or multipolar neurones (Figure 2B). These nerve nets come together in integrating centres, where a mixture of both neurone types may be found (Figures 2C and 2D); these centres are known as ganglia.

Figure 1. [Click image to enlarge] The Porifera or sponges are very simple multicellular parasites. They do not possess a nervous system and it is controversial as to whether they have any neurones. In A is shown the simple motor reactions of the fresh water sponge Ephydrata, with the mouth chimney changing its' form as water is drawn into the body through small pores and passed out through the mouth. In B is shown some of the cell types that stain with silver in the sponge Sycon Raphanus; the outer surface is connected to the inner surface by two cells with long and thin processes that may be nerve cells; the cells on the inner surface are collar cells ( called choanocytes).

Figure 2. [Click image to enlarge] The Colenterates or polyps like hydra, jelly-fish, sea-anemones and corals have a sack-like body with tentacles as shown in A. In B are shown the two different kinds of neurone networks present in Aurells Aurata: one is composed of neurones that possess two axons (bipolar neurones ) and these are promenent in relation to the radial and circular muscles that are exposed in this drawing; the other is composed of neurones with more than two processes (multipolar neurones) and these are shown in relation to the gastric cavity. Both bipolar and multipolar neurones from each nerve net are apposed to each other in collections on neurones called ganglia, as shown in C; here the input to the bipolar cells a2sociated with the muscle is transferred to the multipolar neurones associated with the gastric cavity. This collection of neurones into a ganglion for the purposes of neural integration occurs for the first time in the Coelenterates.

A very large increase in the complexity of the nervous system occurs with the appearance of the flatworms (Figure 3). In this case the ganglia are fewer in number and concentrated at one end of the animal, which may be distinguished as the head for primitive eyes and mouth are found there (Figure 3) This is the first sign of the head ganglia which are eventually destined to evolve in their most complex form into the brain of Homo Sapiens (Figure 4), which receives sensory information from nerve receptors in different areas of the skin called dermatomes (numbered in Figure 4) as well as from the distance receptors such as the eyes, ears and nose.

Figure 3. [Click image to enlarge] The grouping of large numbers of neurones into an integrating centre or ganglion. which is not symmetrically placed in the animal, first occurs in the Platyhelminthes or flatworms. Shown here is the dorsal nerve plexus of Notoplana Atomata( Polycladida) converging on the head ganglion. The labels refer to the 'gpl' (genital nerve plexus), hdn' (posterior dorsal nerve) and 'tau' (tentacle eyes).

Figure 4. [Click image to enlarge] The grouping of neurones into a head ganglion that provides an integrating centre for the nervous system reaches its' most complex level of evolution in the brain of Homo Sapiens. Shown are the individual areas of skin, each subserving a different set of neurones, that bring information to the brain concerning such sensations as touch, pressure, temperature and pain. The individual areas are labelled C2 to C5 (cervical spinal cord levels 1 to 5), TI to T12 (thoracic spinal cord levels 1 to 12), Ll to L5 (lumbar spinal cord levels 1 to 5) and S I to S4 (sacral spinal cord levels 1 to 4). Nerves enter the spinal cord at each of these levels C2 to S4.

Nicholas Humphrey, until recently Director of the Unit of Animal Behaviour at Cambridge University, has written a book called " A History of the Mind". Humphrey argues that in the most primitive animals, such as jelly fish, the nerve nets convey information from the body wall concerning sensory phenomenon such as touch towards the ganglia which then issue an outgoing signal for the muscles of the body wall to respond, for example with a contraction giving a wriggle (Figure 5A). Humphrey goes on to suggest that with further evolution of the nervous system the outgoing signal to the muscles of the body wall in response to an incoming sensory signal became modified so as to actually alter the incoming signal as well as to contract the muscles (Figure 5B); this collateral effect, of the outgoing motor nerves altering the signal arriving along the sensory nerves, may even be present in the early evolved nervous system of flatworms. Indeed an even further degree of collateralization can occur in which the motor collateral can give rise to sensory experiences independent of any incoming sensory signals (Figure 5C). Another form of collateralization involves the outgoing motor signal in response to a sensory input modifying the incoming sensory signal, without actually contracting muscles at all (Figure 5D); these collaterals can be used to sustain the sensory experience well after the actual event that gave rise to the initial sensation has passed. This ability to maintain a sensation at will by using the collateralization effect is called 'sustaining the sentient loop'. The final evolution of this process, according to Humphrey, probably only occurs in the higher mammals. It involves the motor output that has been modified to only change an incoming sensory signal now gedrating and sustaining sensory signals itself within the brain in the absence of any sensory input (Figure 5E). The nervous system is in this way able to voluntarily generate sensations and maintain them at will. It is this ability to use the sentient loop that is the highest form of consciousness.

Figure 5. [Click image to enlarge] Evolution of the 'sentient' loop according to Humphrey. A: during evolution a most elemental form of nervous system consists of sensory neurones bringing in information, concerning for example touch and pressure, to an integrating centre consisting of a large number of interconnected neurones constituting a head ganglion; this then issues motor command to contract an appropriate muscle given the type of sensory information received by the head ganglion. B: the next level of sophistication was reached with the appearance of collateral nerve branches emanating from the outgoing motor nerves and ending in relation to the incoming sensory nerves; in this way the motor command was able for the first time to modify the sensory input to the head ganglion (see Figure 6 for an example of this process). C: these collaterals then became modified in two important ways, one of which is shown here; on issuing a motor command the collateral is able to induce a sensory experience independent of any input to the head ganglion along the sensory nerves themselves (see Figure 7 for an example of this process). D: the other important way in which the collaterals became modified is that they could be used to modify incoming sensory signals independently of any motor signals at all (see Figures 8B and 11 for examples of this process). E: the final level of sophistication involves the appearance during evolution of the 'sentient loop', in which the collateral acts on its own without any motor command being issued or sensory information about the environment being received; the head ganglion or brain can in this way generate its own sensory experiences, and it is this process that constitutes consciousness ( for an example of the brain generating activity in a voluntary way, without motor or sensory activity, see Figure 11).

Figure 6. [Click image to enlarge] The corollary discharge. Diagrammatic representation of the brain and spinal cord showing the possible levels of corollary discharge by which motor output from the cortex acts on incoming kinesthetic signals arising from the sensory neurones. Corollary discharges are obtained from motor commands and they can influence perception either by modifying incoming sensory signals (in this case at the level of the thalamus) or by acting independently of the incoming sensory signals. Kinesthesia is the sensation by which body weight, position, muscle tension and movement are perceived. Corollary discharges can alter the way in which such kinesthetic signals arising from sensory endings in muscles are interpreted. Such sensory endings in muscle spindles may send signals relating to the length and velocity of movement of a particular set of muscles; these spindles will also send signals arising from their being activated by a certain class of motoneurones in the spinal cord called gamma motoneurones. ]be signals due to the gamma activation of the spindles are removed by a corollary discharge, which at the same time allows the signals from the spindles due to the length and velocity changes to be perceived. In the example shown gamma motoneurones are activated from the motor cortex giving rise to spindle receptor discharges; these discharges together with the additional discharges due to the contraction of the muscles are received by the sensory neurones and transmitted through the group of neurones constituting the gracilis and cuneatus to the thalarnus and thence to the somatosensory cortex; here they give rise to the sense of movement of the muscles. However the initial motor discharge of impulses gates out the sensory discharge relating to gamma motoneurones exciting the spindle receptors; this gating may occur at the many regions of interaction between motor and sensory pathways in the brain and are shown here as occurring in the thalamus for definiteness only.

In order to make these ideas of Humphreys clear it is necessary to look in detail at the functioning of the human nervous system. First of all what are collateral effects and can they operate in such a way as to modify sensory signals? One of the simplest motor acts that engages the brain and the spinal cord is shown in Figure 6: here a sensory stimulus, such as that arising from sensory spindle receptors in muscle cells concerned with indicating the length and velocity of shortening of the muscle, is relayed through the sensory neurones just outside the spinal cord to the nerve cells just inside the cord within an area called the substantia gelatinosa; these signals are then sent to the main relay station for sensory activity propagating between the spinal cord and the brain, in the group of neurones called the nucleus gracilis and nucleus cuneatus; from there the signals are sent to the thalamus in the brain, which is the receiving area for nearly all the sensory input to the overlying mantle of the brain or cortex; finally the thalamus projects the information to that part of the cortex which is called the somatosensory area, concerned with the analysis of information derived from sensory receptors in the limbs. The kind of information in the signals processed by the somatosensory cortex may require that the muscles that gave rise to the sensory input in the first place be contracted. In this case neurones in the part of the cortex concerned with contracting muscles, namely the motor cortex, project a signal to the appropriate motoneurones in the spinal cord connected to these muscles. These are of two different kinds, namely alpha motoneurones that are attached to cells in the muscle in question that produce the force; the other kind are the gamma motoneurones that are connected with cells that contract in the sensory receptor apparatus itself; contraction of these cells changes the characteristics of the sensory receptors so that they rapidly send signals to the sensory neurones outside the spinal cord and from there to the alpha motoneurones, leading to the contraction of the bulk of the muscle cells; they also send signals to the sornatosensory cortex via the thalamus along the pathway already described. Two kinds of information are then sent via the sensory neurones to the brain: one of these relates to the signal arising from the sensory receptors in the muscle concerned with the position, tension and movement of the muscle, known collectively as kinesthesia; the other relates to the signal arising from the receptors as a consequence of their being contracted by the gamma motoneurones. This latter signal is gated out before it reaches consciousness by a collateral signal from the motor pathway as shown in Figure 6; the sensory signal concerned with the state of kinesthesia of the muscle is not gated out, but is allowed to reach consciousness. The level in the brain or spinal cord at which this gating procedure is carried out is not known; it is shown to occur at the level of the thalamus in Figure 6 simply for the sake of definiteness. Humphrey is therefore correct in his assertion that modification of sensory signals can occur before they reach consciousness as a consequence of a collateral effect from the motor pathway.

Figure 7. [Click image to enlarge] The corollary discharge. This figure gives a diagrammatic representation of the brain and spinal cord illustrating a possible output from the motor cortex responsible for the perception of heaviness of a held object. For definiteness this corollary discharge is shown at the level of the basal ganglia. Such discharges can give rise to the sensation of muscular effort as occurs when lifting and supporting an object. In the example shown neurones in the motor cortex (called Betz cells) that project to the motoneurones in the spinal cord are illustrated; Betz cells may be activated to contract muscles involved in lifting the limbs or an object such as a suitcase; when they do this a corollary discharge is sent (at the level of the basal ganglia?) which gives rise to the perception of heaviness of the limb or suitcase, and this is simply related to the extent of the motor discharge that occurs. Subjects that experience a stroke may have to send a larger than normal discharge down the remaining functional Betz cells to achieve the aim of lifting their arms and so experience them as as an enormous burden. The graph shows the results of an experiment in which the subject has to support a 9 lb. weight with one arm ( the experimental arm) while being asked at intervals to choose what they thought were equal weights to be supported in the same way by the other arm (the control arm). When the experimental arm was allowed to rest between the trials the subject choose weights with the control arm close to the 9 lb weight held by the experimental arm (see 'rest curve'). If however the experimental arm had to support the 9 lb weight continuously, then the subject choose weights with the control arm that were successively greater (see 'fatigued ' curve) than the 9 lb weight indicating the increased sense of heaviness. This arises from the increase in corollary discharge with time as muscles have to receive a greater discharge to support the weight continuously. The perception of heaviness does not arise from sensory signals in the muscle being relayed back to the brain. This graph is due to experimental work of McCloskey, Ebeling and Goodwin carried out in 1974.

Corollary discharges from the motor pathway can also be used to generate a sensation independent of any incoming sensory signals. They can generate sensations of muscular force or heaviness although they cannot generate sensations of movement. Figure 7A shows how the sensation of the heaviness of an object held in the hand is generated by a collateral effect. The pathway from the motor cortex to the alpha motoneurones is shown to give off a collateral branch at the level of the basal ganglia; this, it is hypothesized, can generate a sensation in the cortex of the degree of heaiviness of the object by firing impulses in proportion to those that are being propagated down the motor pathway. It follows that when a muscle is weakened by fatigue, such as when holding a heavy suitcase, a greater number of impulses are required by the non-fatigued component of the muscle in order for the muscle to continue lifting the suitcase; the collateral then receives a greater number of impulses and so a greater sensation of heaviness is experienced. An experimental example of this is given in Figure 7B[;] here a comparison is made between the extent to which a subject perceives the heaviness of a suitcase held continually in one hand by comparing it with a known weight held in the other for a short period of time. The graphs show that in this matching experiment the known weight chosen to be equivalent to the suitcase gradually increases over time, indicating the increased sensation of heaviness. This sensation is due to the collateral effect.

Humphrey is correct then in his suggestion that collateral effects can modify both the kinds of sensations that enter consciousness as well as generate sensations that did not arise from the workings of our sensory receptors. Figure 7 summarizes the situation. Humphrey speculates that early during evolution nerve pathways were layed [sic] down that allowed an animal to respond to say a noxious stimulus to the skin by 'wriggleing' away; in higher vertebrates this simplest pathway might consist of the primary sensory neurones just outside the spinal cord that receive information concerning noxious stimulation projecting to upper motoneurones in the reticular formation wbich then project down to the lower motoneurones and from there to the muscles which are to be contracted to produce "wriggleing" (Figure 8A). At a later stage of evolution mechanisms were put in place that allowed the nervous system to 'gate' out sensory information, using projections from the brain to the sensory gate-way to the cortex , the thalamus, as shown in Figures 8B and 8C. We have already seen how information gathered by primary sensory neurones concerned with muscle receptors can be gated out before it reaches the somatosensory cortex by means of a collateral feedback from the motor cortex at the level of the thalamus (Figure 8B). Such a feedback could occur via the well known pathway from motor cortex to basal ganglia and from there to the reticular nucleus that lies just outside the thalamus; this then projects to the somatosensory cortex (Figure 8B). Sensory information that is gathered by the retina is also 1 gated' as it passes through the thaIamus on the way to the visual cortex, as shown in Figure 8C. The primary visual pathway is from the retina to the thalamus and from there to the visual cortex; neurones exist in the cortex that project back to the thalamus where they can gate the incoming visual information (Figure 8C). There is then evidence for both the modulation of signals arising from the primary sensory neurones as well as from the visual sensory neurones at the level of the thalamus. In this way the brain can determine the sensory information which reaches it.

Figure 8. [Click image to enlarge] Diagrammatic representation of the brain and spinal cord showing different kinds of interactions between sensory pathways and motor pathways. A, illustrates the simplest pathway involving the brain in a motor pathway. Primary sensory neurons relay information concerning kinesthesia, temperature and touch via the spinothalamic tract to the reticular formation of the hindbrain. Here a reflex act is initiated by exciting motorneurons to contract muscle in relation to the sensory stimulus. B, illustrates how kinesthetic gating, referred to in relation to Figure 6, may occur. The motor cortex activates neurons in the basal ganglia which in turn inhibit neurons in the reticular nucleus of the thalamus which normally inhibit neurones in the thalamus that are responsible for conducting the kinesthetic discharge to the somatosensory cortex. Primary motor cortex can then modulate the sensory information that can enter perception through this pathway. C, illustrates how visual information passes from the retina to the thalamus and from there to the visual cortex. This cortex itself contains neurones that project back to the thalamus; these neurones in the cortex can gate the information allowed to pass through the thalamus to the cortex. Both B and C show how the brain itself can modulate the perceptions of the world which it might allow to reach consciousness.

Figure 9. [Click image to enlarge] The sense of time in the brain as illustrated by the 'cutaneous rabbit' perceptual illusion. Shown are diagrams of arms in which the following experiments were carried out to illustrate the subjective nature of the space-time extent of experience. A: taps were delivered in the sequence shown (1 to 10 on the left) on the arm at one-tenth of a second apart so that the final tap was given at 1.8 seconds; the first five taps occur at the wrist, the next two on the forearm near the elbow and the last three at the shoulder region (the subject was not allowed to observe these procedures). Surprisingly the subject experienced the second tap as displaced from the wrist and the rest of the taps at equal distances along the length of the arm (at 1 to lO on the right). It is in this sense that the brain interprets the taps as if an animal (rabbit?) had run up the arm. B: five taps were then delivered in the sequence as shown, namely only on the wrist (on the left 1 to 5) and these were experienced as all occurring at the wrist (on the right 1 to 5). The original experiments were performed by Geldard and Sherrick in 1972.

[1] [2] 下一页