Free Will and Compatibilism

7. Free Will and 'Could Have Done Otherwise'

7.2 Analyzing Free Acts

Book cover: Free: Why Science Hasn't Disproved Free Will by Alfred R. Mele

In the previous section, I offered a compatibilist analysis of an unfree act in which the agent felt they could not have done otherwise than what they did. I want to move now to considering the opposite kind of act; that of a free act. The person on the street feels that when they act freely, they could have done otherwise. This common belief forms the basis of a key incompatibilist argument against the compatibilists. The argument states that in a deterministic world, a person could not have done otherwise because the initial conditions (the person's brain states and their physical environment), coupled with deterministic physical laws, logically entail the person's act. The incompatibilist concludes that it was not open for the person to have chosen other than the way they in fact did.

I want to show that this argument is mistaken. Let me start by contrasting two senses of 'could have done otherwise' used in ordinary language. I will call these the PLIC and the PLUC sense. We use both senses when talking about inanimate objects. The PLIC sense refers to Physical Laws + Initial Conditions and is more typically the sense used by scientists when discussing scientific theories and making predictions from those theories. To illustrate this sense, think about this scenario. John is at his local hotel playing billiards with his friends. He hits the billiard ball precisely the right way for it to land in the top pocket. It is natural for us to say, 'The billiard ball could not have done otherwise than to go into the top pocket.' Of course! Because cues, billiard balls and billiard tables obey physical laws. In this case, expanding 'could not have done otherwise' as a PLIC counterfactual conditional means:

Given the laws of physics, the characteristics of the billiard ball and the exact same initial conditions, the billiard ball would have landed in the top pocket.

By 'initial conditions' is meant the state of the system under consideration at the start of a process. In this case, the initial conditions include the position, size, weight, velocity, shape, smoothness and hardness of the billiard balls, cue and billiard table at the time of John striking the billiard ball. For this PLIC sense, both the characteristics of the billiard ball (that is, its intrinsic properties) and the other initial conditions are made invariant.

Now imagine that on his next shot, John miscues seriously with the result that the billiard ball begins hurtling up into the air in the direction of some hotel patrons standing at the bar. Fortunately, the billiard ball does not hit anyone as no one was standing in the position where the billiard ball came hurtling through. It's natural for us to say, 'The billiard ball could have knocked someone out'. Now, it's the same billiard ball obeying the same physical laws in the same deterministic universe. Yet we are saying that the billiard ball could have behaved differently.

Here, we employ the PLUC sense to understand what we mean when we say that the billiard ball 'could have done otherwise'. The PLUC sense refers to Physical Laws + Unchanged Character and is more typically employed when discussing events of significant concern to humans. So, in this PLUC sense, when we say that the billiard ball could have harmed patrons, we mean:

Given the laws of physics, the characteristics of the billiard ball and some specified different initial conditions, the billiard ball would have knocked someone out.

As in the PLIC sense, we fix the characteristics of the billiard ball by making invariant its intrinsic properties (such as its size, weight, shape and hardnes). As before, we also fix the laws of physics. Otherwise, we would be talking about what is logically possible and not just what is physically possible. In contrast with the PLIC sense, however, our conditional now ranges over different initial conditions while keeping the context of the situation the same. Using the terminology introduced in the previous section, the conditional ranges over different external circumstances within the given scenario. Our scenario consists in John playing billiards at his local hotel and sending the billiard ball flying towards the bar at high speed and at head height. To determine whether the billiard ball could have harmed someone, we ask: Are there any external circumstances that would have led to patrons being knocked out? Well, if Rudy had stood two metres to the left of where he was standing or Jane had stood one metre to her right, they would have been knocked out by the flying billiard ball. So, in this case, we had good reason to say that the billiard ball could have knocked someone out.

In their critiques of compatibilists, libertarians and hard determinists ignore this second sense, the PLUC sense, of 'could have done otherwise' when we speak of free human acts. By proposing that we are exclusively using the PLIC sense when we assert our free will, libertarians are forced to construct an untenable metaphysical story about how intentional human acts break the laws of physics. Scientifically-minded hard determinists, in also acquiescing to the PLIC sense, are thus compelled to follow where their reasoning takes them and declare the counterintuitive conclusion that we are not free to act. Once we recognize the PLUC sense of 'could have done otherwise' and how employing it in ordinary discourse does not imbue billiard balls and the like with contra-causality, we are in a position to do the same with talk of free agents. By applying the right semantic analysis, it becomes clear how human beings can act freely without contravening the laws of physics.

I want now to apply the PLUC sense of 'could have done otherwise' to a case involving a free act. The basic schema I am suggesting here is as follows. An agent could have done otherwise if:

Given the agent's character, the agent would have done otherwise in this situation if certain external circumstances were different.

Consider the Officer from the scenario pictured in the previous section waiting in line in the police canteen the morning of the shooting. Suppose that he likes both tea and coffee and has drunk them on different occasions. That morning in the canteen, say he chooses coffee at the dispenser to drink with his breakfast. Being a free act, he could have selected tea. On a compatibilist analysis employing the PLUC sense, his selecting coffee freely means that, given the Officer's character, the Officer would have done otherwise in this situation if certain external circumstances were different.

Book cover: Free Will by Sam Harris

As per my previous analysis of unfree acts given in the previous section, what I mean by 'different circumstances' is possible circumstances that are external to the agent's character and that both preserve the opportunity for the agent to perform the act and the situational context of the question at hand. In this case, if the canteen was out of coffee that morning, then the Officer would not have had the opportunity to have selected coffee and the question of his choosing coffee freely or not would not have arisen.

Given that the Officer had the opportunity to select coffee, were there any external circumstances that would have led the Officer to choose tea instead? Well, he would have selected tea with his morning breakfast if it had been offered at a discount price or if he had had coffee for the last five days or the canteen was trialling a new brand of tea leaves. We do not even need to know the exact circumstances that would have led to the Officer choosing tea that morning. Our assent to the counterfactual conditional only requires that we have reasonable confidence that given the Officer's character trait of liking tea and coffee, there are some circumstances in which he will choose tea.

Note how in employing a counterfactual conditional analysis using the PLUC sense, the concept of 'characteristics' of an inanimate object is commensurable with the concept of 'character' of a human agent. The size, weight, shape and hardness of the billiard ball that give it the ability to knock out a patron is commensurable with the police officer's beverage preferences, price sensitivity and propensity for experimentation that give him the ability to choose tea over coffee.

It should be noted at this point that the truth of the compatibilist's thesis in general and of my 4C theory of free acts in particular does not necessarily require that a free act is one for which the agent could have done otherwise. Some compatibilists have resolutely rejected this latter thesis. Frankfurt [1969], for example, has constructed counterexamples purportedly demonstrating the possibility of agents acting freely and for which the agent is morally responsible and yet could not have acted other than they did. In my Allan [2016], I try to show how the type of counterfactual conditional analysis offered in this essay survives Frankfurt counterexamples.

The distinction I drew above between PLIC and PLUC senses makes it clear that the type of analysis using counterfactual conditionals that I am proposing here is not some kind of special technical or philosophical translation of what ordinary folk mean by 'had no choice' and 'could have done otherwise'. I'm suggesting that this is how common folk understand these phrases when talking about free and unfree acts. This analysis is a more natural rendering of 'free will' talk compared with the burdensome metaphysical presuppositions of both the hard determinist and libertarian analysis.

That the PLUC sense of 'could have done otherwise' is the most natural sense is further supported by a review of how people use the term in common discourse when referring to inanimate things and processes. Consider this commonplace observation by Lakhani [2014]: 'Hurricane Katrina could have killed more people than the Galvaston hurricane did.' According to Lakhani, Katrina didn't kill more people because it struck considerably later when the resident population was much less. Given Katrina's characteristics, it would have killed more people if it had struck at a different time, in different external circumstances. Examples like this abound in the common literature. The same mundane counterfactual conditional sense in terms of fixed characteristics and different circumstances is given to the number of people that could have been killed by a falling branch [Kisiel 2013], the number of animals that could have been saved from a flood [Barber 2016] and the number of people that could have been killed by a bushfire [CBS News 2015].

As these ordinary-language examples illustrate, the kind of analysis using counterfactuals that helps us make sense of the application of the modal verb, 'could have', to inanimate objects likewise helps us make sense when applied to human agents. In either case, we have no need to import additional metaphysical presumptions about uncaused causes to carry through the analysis.

Copyright © 2016

You will be interested in

Share This

  • twitter
  • facebook
  • linkedin
  • googleplus
  • gmail
  • delicious
  • reddit
  • digg
  • newsvine
  • posterous
  • friendfeed
  • googlebookmarks
  • yahoobookmarks
  • yahoobuzz
  • orkut
  • stumbleupon
  • diigo
  • mixx
  • technorati
  • netvibes
  • myspace
  • slashdot
  • blogger
  • tumblr
  • email
Short URL:


PDF Download Free Will and Compatibilism

Download this essay