Linespace
Libet's short delay
Linespace

2 June 2005

home 

Timer

The experiments carried out by Benjamin Libet into the timing of conscious awareness (briefly described here ) have provoked, and go on provoking, a vast amount of discussion. His own theory of consciousness as a kind of field has received somewhat less attention; and the strange brain-cutting experiment he proposed to test it seems likely to remain unperformed for the foreseeable future. A large number of papers and discussions have been published: in 2004, Libet finally summarised his own account in the book 'Mind Time'.

Libet's early research was actually intended to explore what the minimum stimulus giving rise to a conscious sensation might be. He had an enviable opportunity to study the response of the brain to direct stimulation (using trains of electrical pulses) through the help of a friendly neurosurgeon and the co-operation of a series of patients, who remained conscious and able to report their sensations throughout the experiment. There are several ways of varying electrical stimuli, of course, but a curious fact emerged: whatever the voltage or frequency of the pulses, the stimulus had to persist for about 500 milliseconds before the subject became consciously aware of it. Actually, this is not quite true: above a certain level of voltage, the interval decreased, but the current involved was by then well above anything likely to occur in the brain normally. 

The result was unexpected, because it had already been established that stimuli applied to the skin, rather than directly to the brain, could be detected consciously even if they were much shorter than 500 milliseconds. Libet was able to demonstrate that although the stimulus to the skin might be brief, it was still the case that the resultant brain activity had to persist for 500 milliseconds before the subject became consciously aware of it. Where a patient was anaesthetised, the initial brain response to a stimulus was the same as for a fully conscious subject, but it failed to continue for the required period: moreover, a stimulus applied directly to the brain 500 milliseconds after one applied to the skin could cancel (or in some circumstances, enhance) it.

Linespace

Now, it isn't surprising that there should be some delay between an event, and our becoming aware of it: indeed, if the normal process of cause and effect is to be sustained, the event has to precede the awareness it causes. If we were passive spectators of the world, simply watching the way we might watch a film, the constant delay would be irrelevant -we should never notice that we were half a second behind reality. But we also respond to events, and here a delay is highly relevant and noticeable. The really surprising thing, therefore, was the length of the delay which seemed to be involved. 500 milliseconds - half a second - is a noticeable period of time, and it is evident that human beings often respond to events far more quickly than that. If we had to wait half a second before responding to events, we should never be able to play a good game of tennis, and we should be dangerous (or extremely cautious) drivers. 

The answer appeared to be that our unconscious responses are far quicker than our conscious ones. A stimulus applied to the skin produces an 'evoked potential' or EP in the brain within tens of milliseconds, and that seems to be enough for it to register unconsciously but effectively. A series of experiments have shown that we register unconsciously a whole host of things which may influence our response to events but which never cross the threshold into consciousness. Among other evidence, Libet quotes experiments which show that a conditioned response - a blink - can be created to events which the subject is never actually conscious of. The remarkable phenomenon of blindsight might perhaps be seen as a related case.

Linespace

That still leaves us with a considerable problem. If the foregoing is true, we ought to be aware of it, surely? On the tennis court we would find to our surprise that we returned serve competently before we actually saw the ball, and certainly without thinking about where in the opposite court we might want to put it. Our conscious and unconscious behaviour would be strangely unsynchronised. Libet's hypothesis was that conscious awareness is subjectively referred backwards in time. We consciously perceive the stimulus as occuring at the same moment it registers unconsciously, even though it doesn't in fact enter our awareness until it has persisted for half a second. Subjectively we backdate it to match the EP at the beginning rather than the end of the 500 millisecond span.

Libet was able to provide some direct evidence through experiments which, instead of comparing skin stimuli with direct stimulation of the cortex, instead made a comparison with stimulation of the medial lemniscus, part of the incoming neural pathway. Each pulse delivered here generates its own EP, but the sensation is nevertheless referred back to the time of the first in the train. Backdating remains controversial, however. Perhaps the sensations actually enter conscious awareness immediately, and the half-second delay merely allows time for them to become reportable, or fixed in short term memory? Perhaps we are merely dealing with the difference between being aware of the stimulus, and being aware that we are aware of it? Libet contends that awareness and memory, especially declarative, explicit memory, are different and independent phenomena. 

Linespace

There is a further problem to solve, in any case. The idea of back-reference allows Libet to eliminate the chronological inconsistencies which threatened to develop in perception, but that leaves our perceptions at odds with our decision-making. We now seem to perceive the ball before we perceive the answering stroke of our tennis racquet, but since both seem to occur half a second before the actual moment of conscious awareness, we should, bizarrely, be seeing ourselves begin to play the appropriate stroke just before we have consciously decided which it is.

Confronted with this problem, Libet decided that one more step in the argument was necessary: the perceived time at which we make a decision must also be subjectively referred back by 500 milliseconds. Unlikely as it seems, and contrary to our own impression, we must have made our decisions slightly before we actually become aware of them.

Linespace

Devising a further experiment to test this hypothesis was challenging on two counts. First, how can you tell when a decision has been made? In the case of perception we know when the stimulus occurs, and can track when the initial EP appears in the brain, but the only symptom of a decision seems to be the resulting action.

However, research carried out in 1965 by Kornhuber and Deecke showed that when subjects were asked to move their wrist of fingers at a moment of their choosing, the act was preceded by a measurable electrical change in the brain. This 'Readiness Potential' or RP appeared about 800 milliseconds in advance of the act, and seemed to be a clear indication that the intention to act had formed. It therefore seemed that the RP might be used experimentally as a marker for the decision.

The second problem lay in how to measure the moment at which the subject became aware of having made the decision. If the subject had to press a button or give some other sign, the careful timing would be obscured by the time taken in doing so - the experiment would be measuring two hand movements and their associated decisions! Libet's solution was to arrange an oscilloscope so that a bright dot circled around a kind of clock face every 2.56 seconds (no doubt these days a PC would be used and  the circuit time set to a more rounded figure - but this was 1977). The subjects were simply asked to note the position of the dot at the moment when they became aware of having decided to move their wrists. This allowed the timing to be reported accurately while taking the time required for the report out of the equation.

Of course, the results confirmed the hypothesis: the RP appeared some 500 milliseconds before the reported awareness of a decision to move.

Linespace

That sorted out Libet's account in purely empirical terms: in philosophical terms the problems were only beginning. The research (subsequently repeated and corroborated by others) seemed to provide a scientific proof that free will was a delusion. How could we consider ourselves responsible for decisions we were not even aware of until after they had been made? Some would have been happy to see the demise of free will, but Libet himself was not ready to let it go so easily. Although the subject's decision to move occurred too early for it to have been initiated by conscious thought, there was still - just - a window of opportunity in which conscious awareness might conceivably veto the move. This window lasts, on Libet's account, no more than about 100 milliseconds. Experimental proof is difficult. Libet has conducted experiments in which the subjects were asked to form an intention to move and then veto it at the last moment: apparently an RP appeared and then dissipated, but the weirdness of the mental gymnastics required of the subject seem to leave an element of doubt about the process. Is it possible to decide to move at a random moment while simultaneously holding on to the belief that you will not, in fact, execute the movement?

Interestingly, Libet has a moral argument here. He rightly points out that free will is important partly because it underpins the idea of moral responsibility: but morality, he suggests, is mainly a matter of vetoing things we have a built-in tendency to do. This is a faintly depressing prospect, and I think Libet underestimates the difficulty of discriminating between doing and not-doing. If I veto my momentary desire to kill you, that surely is morally good: but what if I repress the automatic tendency to grab your hand when you are about to fall off the cliff?

Linespace

There are several avenues of attack against Libet's other conclusions, of course. Is the RP really a signal that a decision has been made? If I make a decision about my insurance policy, does an RP appear, or is it just wrist movements that cause RPs? The circumstances of both Libet's experiments and the earlier ones by Kornhuber and Deecke are rather strange: they require the subject to get into a frame of mind where they are ready to make a decision any moment. Might not the RP merely signal a quickening of attention, rather than a moment of decision?

Libet believes that by timing the moment of awareness through his oscilloscope arrangement, he eliminated the need for the subject to spend any time on reporting the moment of awareness: but isn't it possible that we need a certain amount of time just in order to report the awareness to ourselves? Awareness of the decision you have made is one thing, being aware of that awareness is another - which might well be thought to require some further time to develop. 

Personally I also doubt whether it is necessary to reduce free will to a veto system  - 'free won't' as it has been described. Libet often seems to take it for granted that every free act is preceded by a specific act of will: but that isn't really the case. Often the conscious mind sets a general plan, on which we then act more or less automatically. A tennis player has thought in general terms about how to play the next stroke long before the need for actual action: drivers have a kind of running rule in the back of their mind to the effect that if something suddenly appears in front of them, they hit the brake. Free will operates at this higher level, with all our actions being managed in detail by unconscious processes. I don't have to think about where I want to hit the ball at the very moment of decision in order to control my game of tennis any more than I have to think separately about each of the individual muscles I am implicitly proposing to contract. 

Linespace

Libet has another string to his bow, however, inasmuch as his general theory of consciousness is also designed to offer a foothold for free will, along with the subjective experience which in his view sustains it: free decisions are conscious decisions, and conscious decisions require subjective awareness. He proposes a conscious mental field (CMF) intended to account for the various mental phenomena which don't appear to be natural consequences of the firing of neurons. Like other proponents of field theories, Libet believes that the prescence of such a field helps explain how the diverse activity of the brain is bound together into a single, unified conscious experience. Unlike Susan Pockett or Johnjoe McFadden, however, he does not see this field as a straightforward physical phenomenon. Libet, who says that when young he was convinced of the truth of determinist materialism, no longer believes that conscious mental activity is explainable by or reducible to, neuronal activity, although it certainly requires it. He is not, at the same time, advocating any kind of dualism or spiritual theory: instead (and rather obscurely, I think) it seems that his proposed CMF is an emergent phenomenon: something that arises from the combination of active neurons but amounts to something distinctively more than, and different from, the sum of brain activity.

Linespace

In order to verify the theory, Libet has proposed a rather alarming experiment. If a small slab of living brain in a suitable area of the cortex, could be cut off from its neural connections while being left with a suitable blood supply, he believes it would remain capable of producing subjective experience, mediated by the CMF, when appropriately stimulated. (Actually the slab would have to have some minimal neural input to keep it 'awake', but this could apparently be managed without spoiling the experiment.) The existence and role of the CMF would thereby be demonstrated - though I imagine there would still be considerable resistance to the idea and a host of objections to the experiment.

The immediate theoretical objection which springs to mind is to ask why the brain goes to such trouble to make neural connections if it can do its normal job without them: Libet's answer is that it can't do the 'normal jobs' of information processing, memory, emotional response and all the rest: it's just the subjective aspect which arises from the CMF.

Of course, there are also practical difficulties: you cannot cut out slabs of living human brain purely for experimental purposes. Libet has identified a method and a category of patients in whom the experiment could theoretically be performed innocuously, namely those who are already going to have an epileptic focus surgically removed. Very few patients of this kind exist, and Libet has had no success in interesting the surgeons in these cases in facilitating his experiment.

Linespace

Update: subsequent fMRI research confirms and goes beyond Libet's findings.

Further Update: new research (2009) suggests that Libet was wrong after all...

Earlier

home

later