1. STIMULUS AND RESPONSE

IF the series that is connected in cascade were infinitely long there would be no first device. Then it would, of course, be meaningless to ask any questions at all about a primary relay. There would be nothing that could justify the term.

It is, therefore, attractive to consider whether one could follow the series of devices that we have so far traced only up to synapses in the craneman's brain further and further back along a never ending path. Can one go on saying without stopping: This device is controlled by a further one and this, again, by yet a further one? Can one say this, I am asking, not when one attaches any recondite or abstract meaning to the word "control", but when one uses the word in the sense in which One can say that the cranemotors are controlled by contactors, the contactors by switches, the switches by the craneman's hands?

Many would, I think, like to believe that it is so. Of course they would not be so foolish as to say that an infinite number of devices is located in the brain. They would say that the series of devices that control each other in cascade passes through the brain. They would remind us that the brain is connected to the outer world by two distinct nervous systems, these being, respectively, the afferent and the efferent system to which I have referred already. Just as some members of the series, such as contactors, switches and nerves, can be observed, on the efferent side, between the craneman's brain and the casting, so other members, they would tell us, can be observed on the afferent side, disposed along endless paths into the outer world.

The justification of such an interpretation of the facts would be that the craneman acts on what he perceives, on what he sees and hears and feels. The sum of his sense perceptions constitutes the stimulus that reaches his brain through the afferent nervous system. What he does is his response to the stimulus. A well defined path (called a reflex arc when control is not conscious) passes through afferent nerves, brain and efferent nerves.

From such considerations one might argue that the series of cascaded devices that I have called relays is by no means limited to contactors, switches, muscle fibres, endplates, nerves, synapses, but that it can be pursued further back past the brain and through afferent nerves to the organs of sensory perception, on from there to all the objects that the craneman sees and hears and feels, and yet further back to the events that are the causes of the events perceived by him, on and on as far as causation extends into infinite space. The argument would find support from philosophers of the behaviourist school. Can it be sustained?

If my question had been a different one, if it had asked about causes and not about control, then this argument would be relevant. It could perhaps be proved true. For all I know there might then be no flaw in it. For it is perfectly true that a stimulus received by the afferent system is the cause of a response by the efferent system. One really can trace a path of causation right through the craneman's brain. In the sense in which it is correct to say that the performance of the motors is caused by the performance of the contactors, the performance of these by that of the switches, and so on, so it is also correct to say that the performance of the craneman's hands is caused by the things that he has perceived, that these things are caused by other more remote events and that in fact, the recital in which one says that one event is caused by a previous one can go on without limit. Had my question been: "What causes the casting to move as it does?" a correct answer would have been: "An infinite series of successive events, of which some occur within the craneman's body and most outside it." But the word "cause" does not form part of my question and the word "control" does.

2. UNCONTROLLED STIMULI

If one word were as good as another one could say correctly that an object controls the light that it reflects, the sound that it emits, the force that it exerts. One could say that the wind controls the waves, the moon the tides, the storm the leaves that are tossed before it. One could describe the wind, the moon, the storm as controlling devices. One could say that a stimulus controls the craneman's behaviour. And, as any observable object may be the cause of a stimulus, one could say that everything observable is a controlling device. And then one could justify the argument that the series of devices in cascade by means of-which movement of the casting is controlled extends indefinitely beyond the afferent side of the craneman's nervous system to and through all the things that combine to form a stimulus to his behaviour. Any random set of objects that he might be seeing and hearing and feeling at the moment could be described as a set of controlling devices. One would not need to distinguish them in any essential feature from contactors, switches, nerves and other devices that are, by common use of language, called controlling devices.

However, a scientist cannot afford to be as complacent as that about the use of words. One word is not as good as another. Cause and control are not synonyms. A conclusion in which it is pretended that they are, cannot be scientifically sound. To cause an event is not necessarily to control it. One may sometimes observe causation with control and sometimes causation without control. And it is important to distinguish between them. There are some objects that can correctly be called controlling devices and some that cannot be so called without improper use of words. And it is important to know which is which.

3. THE EFFERENT AND AFFERENT SIDES CONTRASTED

Between the craneman's brain and the casting, that is on the efferent side, one can find objects that are controlling devices. And they are connected in cascade. They are called controlling devices because each of them is so constructed that it can be in alternative states, such that when it is in the one state energy flows, inevitably, to the next device in the series and when it is in the other state energy is, equally inevitably, prevented from so flowing.

None of this can be said of the objects from which, on the afferent side, the craneman receives his sensory data. Instead of specific devices connected to each other in a systematic manner by copper wires, nerves, semi-permeable membranes and whatever else may serve to transmit impulses at controlled moments of time, we observe on the afferent side beyond the craneman's body an uncoordinated conglomeration of objects each and all of which are transmitting sensory data to the craneman, not at controlled, but at uncontrolled moments of time. It would be absurd to say that each and everyone of them was a controlling device. The craneman sees the top of the crane as it lumbers past his cabin. He sees the crane-hook with the casting suspended from it, swaying a little as the crane accelerates. He notices a bright shaft of light, dust speckled, that enters through the cabin window and the printed instructions nailed to the wall above his indicating instruments. His ears are assailed by all sorts of sounds, by the various shouts and clangs and clashes of a busy foundry. He smells hot oil and the bacon that his mate is frying for breakfast. He is touching a switch handle and senses a slight looseness in it. He feels on his thigh the sharp edge of a frame against which he is leaning. One could go on and on with a catalogue of the sense data perceived by the craneman at the moment when he is performing one of the operations needed for the controlled movement of the casting. And the various items in this catalogue are not coordinated as the items are that one can enumerate on the efferent side. They are not systematically connected to each other in cascade or any other definable arrangement. They may all be causes of some aspects of his behaviour and some of them (though not all) are among the causes of his behaviour as he controls the movement of the casting. But the sensory data do not do the controlling; they provide causation without control.

So one may concede everything that the behaviourist school has to tell us about the relation between stimulus and response, about causes and effects, and yet distinguish quite clearly between what is to be found respectively on the efferent and the afferent side. It is on the former side between the craneman's brain and the casting, that one may observe a set of devices connected in cascade and each serving the function of control. On the other side there are no such devices. All the comforting "isms", I have said already, depend for their plausibility on ignoring something that is real and relevant. The plausibility of the behaviourist and mechanist "isms" depends among other things on ignoring the distinction between causation and control.

And so it is in the brain that we must seek the beginning of the path along which there is causation with control or, as I prefer to put it, diathesis. This path, unlike the path of causation without control, is of finite extent, and contains a finite number of controlling devices. The first of these is the one to which I have given the name primary relay. One may dispute the appropriateness of the name, but one cannot dispute that a finite series of objects through which control is transmitted has a beginning and that the first of the objects is to be found in the brain.

The imp of complacency has been defeated a second time. At his first frustrated attempt he tried to obliterate the problem by confusing it with discussion about definitions and choice of approach to the theme of our investigation. And at the second attempt he tried to obliterate it by confusing it with discussion about the causal relations between stimulus and response. Has he any other arguments why it would be better not to ask how and by what a primary relay is controlled? If he has any that could sound in the least convincing I do not know what they are.

Top of Page