Diffuse responsibility

I recently watched an excellent talk given by Sam Harris in which he argued that free will is an illusion (http://www.youtube.com/watch?v=pCofmZlC72g). I don't want to go into all the arguments here so check it out on YouTube if you haven't already seen it. I think the most convincing argument, for me, however, was that if you frame the discussion properly, the idea of free will even becomes nonsensical. It goes something like this: If you know that your actions are determined by your thoughts, and if you know that thoughts simply occur by some unselected means, what could it even mean to suggest that you had the free will to make a decision? Even if you argue that free will is the ability to decide not to act upon a thought, you are forced to ask 'how did you make the decision not to act?'. You made it at the behest of another, competing, thought. If you can never really choose your thoughts, what does it mean to say you had free will to choose your actions?

More recently I watched a talk given by Daniel Dennett (http://www.youtube.com/watch?v=zwbnGqOrAEM&list=LP1VcCvcWr-nM&index=3&feature=plcp) in which he argued that free will does exist, but not as the 'ghost in the machine', not as a soul, not as an ultimate freedom which he considered to be obviously impossible. He seemed to be arguing, although he never clearly redefined the term, that free will is the title we give to people we consider competent to be held accountable for their actions. He goes on to discuss the idea of considering 'practical' free will over free will 'in principle'. It is slightly ironic that he complains about reviewers who have misunderstood his previous works by claiming that he simply redefined the terms in question, because I must find myself levelling the same criticism here. Saying that kind of free will is obviously untrue doesn't provide a justification for redefining the term since that kind of free will is what most people mean when they say 'free will'. But, okay, a discussion about 'practical' free will, or moral competence, is still worth having.

Punishment seemed to be a primary theme of the remainder of his talk. Somewhat curiously he intimated that those who attack free will might be doing so simply in order to reach the end conclusion that 'no one is responsible for their actions' (a conclusion of determinism). He seemed particularly alarmed at experimental results which showed subjects cheating more at a game after having read that free will is an illusion. It seems to me, however, that the same charge could equally be brought against Dennett; that he is desperate for us to not abandon punishment, which he sees as a useful tool for managing society. I think the whole argument is really trying to address the issue of responsibility, since 'how can you be held accountable without free will?'. I think this is a failure to think clearly about the issues, however. We can hold an earthquake accountable for the destruction it causes, not in the sense that we should express blame or malice, but in the sense that it was the very real cause of the destruction. The over zealous determinist might then argue that we are obliged to follow the never-ending chain of causes such that responsibility is never attributed to anything (or anyone). Dennett (a compatibilist), on the other hand, seems to want to ascribe responsibility to the first 'agent' involved in the chain; i.e. if I shoot someone the gun is not at fault but I am, having the moral competence to decide not to do it. Here I think the illusory nature of free will as described by Harris undercuts the compatibilist argument; if we know my decision is just a function of my upbringing, my circumstances, my genetics, and some other random factors, then in what sense was I responsible for the decision? what about the causes which led to the manafacture of the gun? This of course, does not change the fact that I pulled the trigger and, if not stopped, might again. We can hold agents accountable for their actions, we can incarcerate them when they are a danger in the same way we protect ourselves from harmful non-agents such as the natural elements, but must also consider the chain of events.

I argue, therefore, for a diffusion of responsibility. Like the roots of a tree, many causes contrived to produce a result, and the most directly involved agents must bear the brunt of the responsibility. However it is absurd to consider that any one agent is the sole author of his actions, and so lesser responsibility must be ascribed to other agents within the chain of causes.

I think there are two common criticisms of this position:

The first is a reflexive fear of the removal of blame, as if that somehow has a logical bearing upon responsibility. You could, for example, blame a serial killer for the murders he commits, but for a determinist this makes little sense. His mind was just the product of his experiences, genetics, upbringing etc. so to what extent can you truly ascribe blame? And aren't the parents partially responsible for the deaths if their parenting was atrocious? how much are they to blame? I would argue that blame is an irrelevant and often harmful framework for thinking about this problem. Instead of focusing upon blame, punishment, or retribution, we should focus on minimising and repairing the damage. We don't need to blame the serial killer in order to know that it makes sense to incarcerate him to protect ourselves. If there is something we can do to repair his broken mind as well, all the better.  The same reasoning applies to the parents; if they were in some way negligent to their child, some way of preventing further negligence seems necessary. This might entail compulsory education on parenting, medical intervention (preventing them from having further children), or physical imprisonment. Perhaps one, or all, of these solutions seem distasteful, but the question of whether they are more harmful than negligent parenting, or whether there are better solutions, will have a definite answer (whether we know that answer or not). Thus the most appropriate decision will be the one which, to the best of our knowledge, most significantly minimises the harm (where harm includes psychological as well as physical states).

The second is a practical complaint. One could argue that it is difficult, if not impossible, to consider every step in the chain. Okay, so a serial killer may be responsible for his murder, but then to what extent was he influenced by advertising, friends, teachers? Surely there are a myriad of other causes of which we are unaware? This may be true. The fact that we often can't perfectly model the entire chain of events doesn't undermine the logical consistency of the argument, and isn't the same thing as saying that we know nothing about the chain of causes. From a pragmatic standpoint we can, and sometimes do, implement the approach of diffuse responsibility, when we refer to 'mitigating circumstances'. Just because we don't know the true nature of every step in the chain of causes, however, does not vitiate the argument, or unburden us of our responsibility to minimise the possibility of such harm recurring. Where we are not confident of a specific cause in the chain of events, and cannot reasonably investigate further, we can only note down the events (and our suspicions), keen with the intent of further inquiry in the light of similar recurrent events.