This is the fourth of six posts in my essay on free will.
A common objection to determinism is that it undercuts the foundation of responsibility, and therefore morality: if our actions are predetermined, then how can we be held accountable for them?
Let us consider what responsibility means. I think there are two components to responsibility:
1. At a very basic level, whoever performs an act is responsible for it. Assigning responsibility for an act, then, starts by determining who caused the act to be performed.
2. A person is commonly held responsible for an action if she made an uncoerced decision to perform that action. If the action is coerced, we cannot conclude that the actor was fully responsible for it.
For instance, if a bank thief presses a gun to my back and tells me to demand money from the bank teller, I cannot be held responsible for the theft of that money. Instead, the act of theft was ultimately determined by the thief himself, since I essentially had no say in the matter. The thief, then, should be held responsible.
My definitions of responsibility are incomplete, though. They consider only the question of which individual in a group of people can be held accountable for a particular action. Other possible sources of coercion are not mentioned. And so we must ask what else produces coercion.
Let’s start by looking at the laws of physics themselves – do they coerce us into performing all of our actions? Our bodies are governed, by and large, by Newtonian mechanics and biochemistry. Biochemistry, furthermore, is essentially a description of the physics of intermolecular interactions, involving mostly electrostatic forces. So does it make sense to say that any given atom or molecule, on its own, causes me to perform a particular action? Put differently, is it possible that my decisions would be any different if I removed any single molecule from my body? I think the answer to both of these questions is clearly “no”. Single molecules do not have the wherewithal to coerce complex organisms into making particular decisions.
So, if any individual molecule cannot be held responsible for decision-making, what can? The answer is that complex arrangements of interconnected molecules are needed to produce a decision-making system. A single transistor may not be able to make the decisions that a personal computer typically makes, but a carefully interconnected arrangement of many such transistors certainly can. The proper locus of decision-making, then, is the neural system as a whole, not any of its constituent molecules, or even its individual neurons.
There are, however, certain phenomena that can influence the decision-making capacity of our brains. While single molecules or neurons have no direct influence, large collections of molecules, or widespread damage to many neurons certainly can. In other words, there are processes that occur at the scale of the neural network, rather than the scale of individual molecules or neurons, that do have coercive influence. Consider alcohol, for instance, or brain injury and disease. It is clear that these things affect our ability to make decisions. If we learn that a murder suspect harbors a large tumor in his brain, we are more likely to be lenient in punishing him, because we understand that he might not have been acting freely.
Let us take stock, then. Thus far, we have identified three possible causes of coercion: other people, mind-altering substances, and brain injuries and diseases. Perhaps we could add certain other psychological sources such as social pressure and certain religious and cultural beliefs. What we cannot add to the list, though, are the laws of physics themselves, since they do not prevent free decision-making in the sense that I have defined it in this essay, i.e. decision-making that weighs the inputs objectively according to their merits. And since responsibility entails making a deliberate decision to perform an action based only on the merits of the available inputs, we cannot say that the laws of nature diminish our responsibility for our actions.
The idea that we do not have a choice in what we do because a rerunning of the clock would produce the same actions, is mistaken. Yes, it is true that under determinism, we are bound to make a particular choice in a given situation, but it is a choice nonetheless. We have to confront the available options and pick one. We have to cross that bridge from undecided to decided. And this choice wouldn’t be made unless we were there to make it.
At this point, it serves us to take a step back. The whole issue of responsibility is part of the practical framework of morality (see my essay on morality on this blog). Systems of punishment (or reward) have the function of strengthening the effectiveness of the moral system. So when looking at questions of responsibility, we need to ask what the desired consequences are for the moral system and the society that practices it.
As I argue in my essay on morality, I believe that the only sensible goals of morality are to enhance happiness and decrease suffering as much as possible among members of society. We needn’t ask, then, whether the concept of responsibility or accountability makes sense in a deterministic framework. Instead, we need to ask what approach to immoral actions will further the goals of the moral system, within a deterministic framework.
To start, we note that identification of the guilty party in a particular crime is not incompatible with a deterministic framework, whether we believe that person to be fundamentally responsible for choosing her actions or not. Once we have determined who the guilty party is, what sort of treatment should that person be subjected to in order to most effectively further the goal of the moral system, namely to enhance happiness and reduce suffering? When put this way, it is clear that applying punitive measures out of spite or some simplistic sense of tit-for-tat, is not necessarily going to get us anywhere.
There are two extremes we can consider. First, we could subject guilty parties to no special treatment at all, and simply allow them to resume their normal lives. But doing this will provide no system of moral reinforcement: if there are no consequences for immoral actions, then people are less likely to care about being moral, and the moral system will be in danger of collapsing.
The other extreme is to select some sort of grossly disproportionate punishment in order to produce a strong disincentive for the rest of the population to commit similar crimes (not as a retaliation to the offender). It is not at all clear, though, that this sort of extreme deterrent approach actually works. Furthermore, this approach would not readily be agreed to by most people, putting in jeopardy the widespread adoption of the moral system of which it is a part.
It seems, then, that we need to treat moral offenders using some middle ground between the two extremes: some sort of punitive measures may be appropriate if they produce an effective deterrent effect.
There are also other considerations to be made. First, the aim of increasing happiness and reducing suffering refers to everyone in society, including the guilty party. It therefore behooves us not only to protect society from a potentially dangerous person (a justification for incarceration) but to rehabilitate the offender so that she, too, may achieve a greater level of happiness, and be less likely to re-offend (and therefore reduce the happiness of those around her). Rehabilitation, in the context of this essay, can be described as a modification of the decision-making engine in the brain so that it is more likely to make moral decisions than immoral ones.
In America today, too much emphasis is put on the tit-for-tat punitive mentality, leading to high prison populations and high rates of recidivism. This approach clearly isn’t very effective at maximizing happiness and minimizing suffering. It is too focused on the idea that people deserve punishment because they are fully accountable for their actions.
As we learn more about psychology and neuroscience, we bring into stark relief the reality, initially demonstrated by determinism, that accountability is not a straightforward concept to be taken for granted. And as the discussion here tries to demonstrate, a reasonable model of justice can be built without having to resolve this issue of accountability, but simply by considering what actions would produce the greatest benefits to society.