Tononi's Defense of Free Will

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view

Tononi's Defense of Free Will

Craig Weinberg
"Seeing conscious deliberations as maximally irreducible has some relevance for the issue of free will. Consider for example the requirement for autonomy: to be free, one must certainly be independent from constraints outside one’s deliberating consciousness. These include both environmental constraints, such as limitations that force us to a particular choice or that impede our own choice, and unconscious, ‘alien’ constraints that, while generated somewhere within our brain, affect our actions largely outside the control of the conscious self. Given the definition of a complex, a conscious choice is necessarily autonomous, as it is made intrinsically.

The requirement for understanding implies that, to be free, a choice must be based on a concept of what is at stake – for example, one can freely choose between right and wrong only if one has a notion of which actions are right and which are wrong under some circumstances. According to IIT, a complex can be held responsible for a certain choice only if it has a mechanism implementing the corresponding causal concept, in this case the backward component of the concept. For example, I must have a concept corresponding to the distinction between right and wrong (IF certain sets of past states occur, THEN certain sets of future actions/omissions are right/wrong) to be responsible for that choice – that concept is a maximally irreducible cause for my action. Similar considerations apply to the requirements for self-control, since the forward component of concepts within a quale ensures control.

The requirement for irreducibility implies that a choice can only be free if it cannot be ascribed to anything less than myself – I am the only entity that can be said to be responsible for my choice. That is, when asking who is responsible for the choice, the answer should be ‘me’, meaning all the circuits underlying my present conscious experience, and nothing less than that. IIT indicates that each experience is a maximally integrated conceptual structure generated by a complex, and therefore what it will choose given a particular present state cannot be ascribed to anything less than the full structure, with all its concepts (recall the light-blink example of a previous note). This structure is supremely causal to account for the choice in that it is maximally irreducible – anything less won’t do, anything more won’t matter. Furthermore, the choice happens at the macro-level at which ΦMIP is maximized, meaning that our conscious choices are not an illusion supervening upon micro-level events that are the true causes, as is often assumed. Indeed, the macro-level exists only if it has more causal power than the micro-level, which it then supersedes. Thus, each choice is a choice of the whole complex, not reducible to a number of choices made within nearly independent modules, each in a limited context, or to choices made by micro-elements. Therefore a choice is the freer, the more it is conscious: more consciousness, more freedom. Moreover, a bit paradoxically, a choice is the freer, the more it is determined (intrinsically). This is one fundamental sense in which the key notion of alternative possibilities – the feeling that one could have acted otherwise, which is essential to the feeling if being responsible for one’s action, is captured by a large integrated conceptual structure: such a structure implies a very large number of counterfactuals (alternative possibilities) that are under the control of the agent (they are part of his consciousness). In other words, a conscious choice is one in which a large number of highly informative concepts that make up my perceptions, thoughts, beliefs, desires, feelings, memories, and character, all concur in determining a choice in the integrated ‘tribunal’ of consciousness. Note however that, even though every conscious choice involves a large number of counterfactuals, it is still useful to distinguish between ‘deep’ and ‘shallow’ conscious decisions, based on how many concepts are directly involved in determining the choice here and now. At one extreme, the decision to request a divorce or not is likely to involve simultaneously many different concepts within the complex, so it is deep. At the other extreme, the decision to flex one’s finger or not during an experiment on free will depends on just a small number of concepts (do I feel the urge or not), so it is shallow. This is because the previous conscious decision to participate in an experiment on free will has had the consequence of fixing most variables within the main complex, so the only variable that is left free to vary is the ‘urge’ to act.

In this view, freedom requires first and foremost irreducibility, meaning that a choice cannot be ascribed to anything external, or anything less, than the agent. However, indeterminism also plays a role, though not the usual role of reducing responsibility by substituting it with chance. Recall that if a complex generates maximal integrated information at a macro-scale in space or time (say neurons instead of subatomic particles, and over hundreds of milliseconds), this means that: i) the system is most determined, in an informational/causal sense, at that macro-scale than at any micro-scale; but ii) it is also necessarily under-determined, because the macro-level can be more informative/causal than the micro-level only if there is some indeterminacy. Given that our own consciousness appears to flow at a macro spatio-temporal level, some degree of indeterminism is a given (in line with both physical sources of indeterminacy and the simple fact that the environment is unpredictable). But IIT does not consider indeterminism as a drop of randomness that instills some arbitrariness into a preordained cascade of mechanisms, thus decreasing their causative powers. Rather, in this view indeterminism provides a backdrop of ultimate unpredictability against which macro-level, integrated mechanisms fight to increase understanding and control – a fight for increasing the causative powers of consciousness, and the more these increase, the more freedom increases. But since this is a battle against a backdrop of indeterminism, its results are never completely predictable. In other words, freedom of will is a fight in which order (integrated information) tries to minimize disorder (lack of constraints) by taking into account as many constraints (knowledge) as possible. A bit like building a society or a civilization out of relative chaos, or a bit like evolution creating macro-order out of micro-level disorder, thus increasing complexity. But as with societies, civilizations, and evolution, what will actually occur can never be predicted exactly before it happens, and micro-fluctuations – a queen and a squire falling in love, two lizards separated from the mainland after a flood – may initiate an extraordinary turn of events that nobody could predict, not even the universe itself."

You received this message because you are subscribed to the Google Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email].
To post to this group, send email to [hidden email].
Visit this group at
For more options, visit