Commentary on "Free Will in the Light of Neuropsychiatry"
- Commentary on “Free Will in the Light of Neuropsychiatry”
For the new generation of cognitive neuroscientists, the mind-brain problem is no longer a matter for philosophical speculation; how the mind links with the brain can be studied experimentally. The strength of this belief is demonstrated by a stream of popular science books purporting to show how consciousness emerges from the brain. In contrast, Sean Spence presents a rigorous, modest and wholly admirable discussion of the physiological underpinnings of free will. It is of particular importance that he brings to our attention various phenomena associated with psychotic illnesses. It is generally agreed that schizophrenia has a biological basis. Yet the characteristic symptoms reflect disorders of self-awareness, especially in the experience of free will (delusions of control). It follows that studies of the physiological basis of psychotic symptoms (e.g. Silbersweig et al. 1995) must also increase our understanding of the physiological basis of free will. The test for any discussion of the mind-brain problem from a scientific stance is that it suggests new possibilities for experimental studies. I think a number of the points raised in this essay provide clear starting points for the development of such experiments.
Cognition and Physiology
My understanding of the argument from Libet’s data is as follows: there is a characteristic change in brain activity several hundred msec before the subject is aware of deciding to act. Therefore the observer knows that the subject is going to act before the subject is aware of making the decision. Either the act is predetermined (because we can predict it from physiology) or the decision to act is made below the level of awareness. This argument makes the dangerous crossing from physiology to cognition, and it is critical that the correct mapping is made. For this purpose we require a thorough appraisal of the task at the cognitive level. The subject is asked to move his finger whenever he “feels like it.” But, as in all laboratory tasks, the subject’s freedom of action is severely constrained. The experimenter would be very displeased if the subject did not feel like moving his finger for the whole of the experiment. 1
The problem for the subject is to move his finger several times at “random” intervals. The only way he can demonstrate his freedom is by making the timing of his movements as unpredictable as possible. To do this he must keep track of the intervals between movements and make sure they are not too regular. A possible strategy would involve the subject thinking to himself fairly frequently, “shall I move my finger” and then deciding yes or no. Clearly there are several cognitive components to this task and we do not yet know how these map onto the underlying physiology. It should be possible to distinguish [End Page 91] between these components and identify the associated physiology be devising suitable tasks for comparison. The question remains, which is the critical cognitive component associated with will?
Our Experience of Our Own Actions
I agree with Spence’s conclusion (following Velmans 1991) that the decisions underlying the experience of free will must be made non-consciously. This is an example of the general principle that we are not aware of cognitive processes, but only the results of these processes (Nisbett and Wilson 1977). In the case of decisions, we are aware of the information on which the decision is based. Following a pause, we are then aware of the action chosen. We are not aware of how we got from the input to the output. However, being unaware of the process by which we choose how to act is not something unique to the problem of free will. It is a general feature of awareness.
Spence makes the important point that timing is a crucial in the processes underlying our awareness of making decisions. He proposes that a disruption of timing might lead to the abnormal experiences of volition associated with psychosis which I think arise from a defect of self-monitoring (Frith 1992). However, I am worried that there is a suspicion of dualism in his diagrams. There seems to be an early...
