In the first installment, wherein I bashed William Lane Craig (as I do at every opportunity – he’s a terrible philosopher and I’m not sure he is even a good person, what with thinking genocide is okay if it is ordered by god) and presented evidence which suggests that the classical version of free will, which relies heavily on the discredited concept of mind-brain duality, should be discarded. Instead, the only model which makes sense of these data is the computational model of the mind. (For those interested in getting into the nitty-gritty of this way of looking at consciousness, I highly recommend Stephen Pinker’s How the Mind Works.) Perhaps I should clarify my position on philosophy a bit, since I ragged on it in my last post. I don’t think philosophy useless. But a philosophical argument in a vacuum is not evidence (listening, Craig?). The argument must have a context within it is couched. Indeed, philosophy in the absence of at least some form of verification simply underscores its impotence. Craig is an extreme form of this, making the claim that “verificationalism” (is that even a word?) has been discarded.
If by “verificationalism” he means we demand that claims be verified, then, my response is this: “Says who, Brylcreem-boy?” Without verification, one can believe anything, like some dude turning water into wine or such. Has he never heard the phrase, “Extraordinary claims require extraordinary evidence”? According to Craig, historical claims can’t be verified. Well, a little evidence would be nice, wouldn’t it? Craig’s view of science is that we use only our senses to find explanations for phenomena. Horse manure! We accept historical evidence all the time! But the quality of historical evidence matters to a skeptic. Francis Collins defends this same position in Religulous, where he makes the claim that nonbelievers set up an impossible standard of evidence for Christianity to meet. I don’t think so at all. I think my standard of evidence is quite reasonable. That the evidence for Christianity fails to meet this standard is not my problem. Craig and Collins seem to think that we should lower our standards that have worked well enough for everything else, but that is just so much special pleading. Do that and you will accept anything at all, like homeopathy or iridology. I don’t play that, and both of these guys should be ashamed of themselves for trying to make excuses for belief in the face of a paucity of evidence. If they want to believe, believe! But don’t try to snow me….
Well, now that I got out of my system, let’s get back to free will and the idea of using the computational model of the mind to make decisions. When viewed in this manner, it is possible to construct a testable model of volition. Volition is modeled “as a set of decision processes that each specify details of an action.”1 The final decision of whether to perform an action (called the ‘whether decision’) is a combination of early and motivational components, with a final predictive check. Each process specifies particular kinds of information that determines our actions, and thus volition can be looked at as a form of decision making. We shall look at what each of these processes supplies in turn.
The early ‘whether decision’ processing determines whether an action is to be taken at all. This is determined by needs and desires, for instance. Voluntary actions may be initiated for three distinct reasons:
- routine processing of external stimuli fails to remove choice between two or more possible actions;
- new reasons for making an action can pop up and the decision to make an action must be reassessed;
- a drive to perform occasional voluntary actions allows the exploration of the ‘behavioral landscape’.
The ‘what decisions’ processing determines what voluntary action or actions to perform relating to the motivation involved in the early ‘whether decision’. It involves selecting between goals and selecting between possible actions to achieve these goals. This is a difficult area in which to study. In the experimental paradigm given in the last post, what action to take in Libet’s study was prescribed. The subjects were not required to make any choice. But there is some evidence from the medical literature. There is something called dysexectutive syndrome, sufferers of which are disorganized in their ability to schedule voluntary actions (voluntary actions generally occur in series).
‘What decisions’ seem to involve the frontal cortex, as evidenced by brain injury victims. Damage to the preSMA can exhibit a compulsion to grasp and use any object within reach, even if the object is not drawn to their attention. Another disorder, anarchic hand syndrome, the hand opposite the side of the brain injury moves in reaction to stimuli even though the patient self-reports that they will perform no hand movement. These and other data show that the preSMA is involved in suppressing automatic responses to external stimuli. It seems that the frontal cortex (and the preSMA in particular) is key in keeping volition focused and orderly.
For movement selection in ‘what-decisions’, the parietal and premotor areas are thought to be used in processing to generate the final choice of which from amongst several possible actions to achieve a goal will be utilized. Monkeys autonomously deciding whether to indicate a target location by reaching for it or making a saccade (movement of the head or eye to directly view the object) were found to activate neurons in regions identified as being responsible for reaching and saccade actions prior to motor region activation. Thus, the choice of reaching or saccade could be predicted before the action was taken.
Late ‘whether-decisions’ involve a final predictive check on the action and holds a veto over whether it is executed or not. The ‘what decisions’ provide information activating specific motor outputs, the the details of the outputs may not be predictable at the point where early ‘whether decision’ is made. This veto power is important. For instance, the cost of the action may be too great, the choice of action may be deemed inappropriate to accomplish the goal, or the situation may have changed in the interim, requiring a new task selection to accomplish the goal.
‘When decisions’ are probably the most studied of each of these steps in the decision making process. When brain activity is compared between self-paced actions (the subject decides when to make an action) and the same action made in response to a stimulus, the preSMA and dorsolateral prefrontal cortex are activated in making ‘when decisions’.
But these models are situation-specific. For instance, ‘whether decisions’ and ‘what decisions’ are generally iteratively looping in the real world and the timing of voluntary actions will generally depend on external factors and internal motivations. Thus, the ‘when decision’ in this case might be better described as a change in the evaluation of the ‘whether decision’ or ‘what decision’. Some decisions can bypass the ‘when decision’ altogether, such as when the cost of an action gets to high and the action is intentionally inhibited. Such intentional inhibition and decisions to quit occur in the frontomedian cortex, as measured using fMRI.
In the last installment, I will be dissing William Lame Craig once again and discussing volition in relation to consciousness.
- Haggard P. “Human volition: towards a neuroscience of will.” Nat Rev Neurosci 9:934-46 (2008)
- Brass M, Haggard P. “To do or not to do: the neural signature of self-control.” J Neurosci 27:9141–9145 (2004)