I will be skipping ahead for this one to work on the chapter on free will, given my interest in the topic.
"The metaphysical issue then concerns how to reconcile the sense that we are free agents with the existence of [the laws of nature]." (pg 330)
Philosophers often say that we all have this sense of being free. I don't have that sense. Free from what? From my body? I am not free from that. From my genetics, or epigenetics, or upbringing, or education, or health, or media influences, or culture, or the rules that govern the universe, or human nature, or the psychological mechanisms that explain my behavior, or the onslaught of time, or the evils of the world, or my limitations of knowledge and understanding, or the limitations of those around me, or...?
No, I'm not free from any of that. I do not feel free at all. Yes, I make choices, like the choice to write this summary. But where did those choices come from? From me? And what am I? Where did I come from? Am I making choices in a vacuum, or are there obvious influences that explain my choices? If you knew all the factors of my biology and my life, would it be surprising at all that I made the choices that I did, or would you be able to trace my every choice to events that happened to me?
". . . the main issue . . . is whether in any important sense any of our . . . actions are up to us. This is what it means to say an agent's action . . . is free - it is up to that agent. It comes from her and not from . . . something else." (pg 330)
I agree with this way of putting it. Are my actions from me, or from my circumstances?
Robert Kane makes a distinction between surface freedom (we can act in such a way that fulfills our desires) and ultimate freedom (we have surface freedom plus we are the ultimate source of our desires).
(So we have two definitions of free will on offer: Free will 1 = You have free will when your choices come from you and not from something outside of you. Free will 2 = You have free will when you are the ultimate source of your desires.)
(I don't want to mix up 'freedom' with 'free will.' I would define freedom as having options to choose from. The more free you are, the more options you have. When you have options to choose from, you must select one option and decline the rest. A choice is a selection from among options. To have free will is to have control over our choices such that we, not our circumstances, are responsible for them.)
Determinism = Given the past and the laws of nature, all future events are predictable by someone who knows everything.
The tension between free will and determinism is that if all your choices were determined before you were born, then it wasn't you that caused your choices, as you didn't exist yet to make them.
(Indeterminism = Given the past and the laws of nature, all future events are NOT predictable by someone who knows everything. The outcome of chancy events, or events that depend on free choice, cannot be known until the present moment.)
(This probably comes up later but I will mention it here: some folks argue that free will requires determinism because your choices must be self-determined to count as yours. If your choices are indetermined, then your choices are caused by chancy processes, which are external to you. But this assumes that indeterminism implies that choices are caused by chancy processes. But indeterminism doesn’t imply this; if your choices are caused by you (i.e., are self-determined), then the moment you make your choice contains the information needed to predict the choice you will make. But the moment you make your choice does not occur until the present moment,, at which point it is no longer a prediction. So, on indeterminism, our choices cannot be known until they are made. At least, this is a way of fitting self-determination and indeterminism together.)
Alyssa Ney shares Peter van Inwagen’s famous Consequence Argument, which is meant to formally spell out the tension between free will and determinism.
Np should be read as: “No one had any choice about whether p.”
Axiom: α = □p implies Np. (If p is necessary, then no one had any choice about whether p.)
Axiom: β = Np and N(p⟶q) implies Nq. (If p is necessary, and p entails q, then q is necessary. So: If no one had any choice about p, and p entails q, then no one had any choice about q.)
Now here are the steps:
1 - □((P∧L)⟶A) (Necessarily, the past plus the laws of nature entail your actions.) [Definition of determinism.]
2 - Therefore, □((P⟶(L⟶A)) (Necessarily, given the past, the laws of nature entail your actions.) [Equivalent to 1.]
3 - Therefore, N((P⟶(L⟶A)) (No one had any choice about the fact that given the past, the laws of nature entail your actions.) [From 2 & α.]
4 - NP (No one had any choice about the past; no one had any choice about what occurred before they were born.) [Premise.]
5 - Therefore, N(L⟶A) (No one had any choice about the laws of nature entailing your actions.) [From 3, 4, and β.]
6 - NL (No one had any choice about the laws of nature.) [Premise.]
7 - Therefore, NA (No one had any choice about whether your actions took place.) [From 5, 6, and β.]
(The conclusion of the consequence argument is strange. It says we have no choices over our choices, but doesn’t that sound like a contradiction? When we experience ourselves making a choice, what’s going on if it’s not us making the choice? I see two options. One is that we truly are not making any choices. Rather, we are witnessing the end of a deliberation process that takes place in our brains. Our brain, acting like a calculator, is forced by constraints of time, intelligence, access to evidence, prior belief commitments, and so forth, and by these constraints engages in a calculation process. At some point, that process ends. Our experience of “making a choice” or “choosing” or “selecting from options” is really just us witnessing the end of a deliberation process, a process that is not up to us but is up to our circumstances. On this option, the experience of choice is purely illusory. The second option is that we really do make our choices; choice is not illusory at all. We introspect and see ourselves making choices; we see that our choices are coming from us, not from our brains. However, every choice we make is not made in a vacuum; it’s made in a context. We do not choose that context. We do not choose those factors that make up our deliberation process; those factors are not up to us. If we cannot be held responsible for that which is not up to us, then we cannot be held responsible for those factors that make up our deliberation process. If our choices depend entirely on that deliberation process, then we cannot be held responsible for our choices. This is a principle of the transferability of responsibility: If you are not responsible for x, and x causes y, then you are not responsible for x-causing-y. If y depends on x-causing-y, then you are not responsible for y. So on this option, we really do make our choices, but we are not responsible for them.)
Incompatibilism = the view that free will is incompatible with determinism.
Compatibilism = the view that free will is compatible with determinism.
(Free will skepticism = the view that we do not have free will.)
Hard determinism = a form of free will skepticism that accepts both incompatibilism and determinism.
Libertarianism = the view that accepts both incompatibilism and free will, and thus rejects determinism.
Is determinism true?
Laplace’s Demon = the ultimate predictor; a being that knows everything about the past and the laws of nature and can therefore, on determinism, predict every future event.
One objection against determinism comes from quantum indeterminacy. However, equations associated with quantum mechanics, such as Schrodinger’s equation, are deterministic. Some people speculate that there must be an indeterministic law describing the collapse of the wave function. But this is not a settled opinion. The many worlds interpretation, a popular interpretation of quantum mechanics, is deterministic.
Some have argued that quantum indeterminacy is irrelevant to free will, because the probabilities are so tiny so as to be irrelevant at the macroscopic scale.
(I’ve heard others also argue that if your decisions are the result of quantum fluctuations, then your choices aren’t yours.)
Humeanism = a view of the laws of nature that says the laws of nature are explained in terms of what happens in the world.
Anti-Humeanism = the view that what happens in the world is explained in terms of the laws of nature.
AJ Ayer argues that what matters is whether our actions were constrained or not. If there is no gun to your head, then your action is yours. (But this is just surface freedom again. Even if your action aligns with your desires, where did your desires come from?)
Hobbes advocated for this surface freedom view (and so did David Hume). Ayer (and David Hume) believed that determinism was necessary for free will. This is because they defined free will in terms of surface freedom; if your actions match your desires, then those actions were free. But if indeterminism is true, then there is no strong connection between desires and actions. If your character, motives, and desires do not determine your actions, then your actions do not necessarily reflect your character. Evil actions could arise out of good characters, and good actions out of evil characters.
(None of this makes sense to me but it’s apparently what David Hume and others believed back in the day. It's worth noting this account of free will has fallen out of favor.)
We have sourcehood freedom when we are the causal source of our actions.
We have leeway freedom when we have the ability to do otherwise.
Harry Frankfurt attempts to fix the desire view by saying there are first-order and second-order desires. You can desire something, and yet desire to not desire it. A free action is one that aligns with both our first- and second-order desires.
(I think this just pushes the problem back a step: Why do you desire to desire what you desire to desire?)
John Martin Fischer has developed a reasons-responsive compatibilism that says actions are free when the agent used their own reasoning when choosing the action.
Susan Wolf has argued that free actions require the agent to act according to moral standards.
The key insight of compatibilism, apparently, is that there is a difference between those actions that are caused by our (desires, second-order desires, reasons, or moral standards) and those that are not.
(I guess the idea is that those actions caused by our desires or reasons say something about our character in a way that actions caused by our biology or past trauma only say something about our circumstances. But surely, if our desires and responsiveness to reasons are determined, then they too only say something about our circumstances. More precisely, some of our actions really do demonstrate our character, whether good or bad, but then whether we are good or bad is up to our circumstances and not up to us.)
A major debate in free will is whether free will requires leeway freedom, sourcehood freedom, or both.
Incompatibilists argue that free will requires leeway freedom, and yet determinism is incompatible with leeway freedom.
Compatibilists have responded by coming up with thought experiments that aim to show that leeway freedom is not needed for free will; we only need sourcehood freedom.
One example comes from John Locke. We imagine a man who wakes up in a locked room. He decides to stay in the room. So his choice to stay in the room is free even though he has no ability to do otherwise.
Harry Frankfurt, in a 1969 paper, shares the main kind of thought experiment used today for this purpose, now called a Frankfurt case.
Frankfurt case = a thought experiment where someone seems to have moral responsibility (and thus free will) even though they did not have the ability to do otherwise.
(Side note: Another definition for free will I’ve heard is that to have free will is to have the control necessary for moral responsibility; for fair moral praise and blame. If moral responsibility entails free will, then arguments on behalf of moral responsibility are arguments on behalf of free will, and arguments against moral responsibility are arguments against free will. As far as I can tell, moral responsibility does entail free will, and free will entails moral responsibility. However, I don’t see Ney, or the Free Will SEP entry, equating free will with the control necessary for fair moral praise and blame. Rather, free will is a more general kind of control, and moral responsibility falls out of that general control. We perform morally neutral actions all the time (like choosing a burrito bowl for lunch instead of a sandwich), and we can wonder about the nature of those choices.)
A Frankfurt case could go like this: Imagine someone steps into a voting booth to vote for a candidate. In an ironic twist, Trump had microchips placed in COVID vaccines, which allows Trump to control the votes of those who were vaccinated. If someone forms the intention to vote for Kamala / Tim Walz, the chip activates and causes the intention to vote for Trump instead. Certainly, we would say that if someone’s vote is forced in this way, then it’s not free, and the voter cannot be held responsible for such a vote. However, imagine the voter, on their own accord, votes for Trump. Then the vote was not forced, and the voter is responsible for their vote even though they did not have the ability to do otherwise. So leeway freedom is not needed for responsibility.
(I’m not impressed, at this point, by this thought experiment. Think about option 2 from above, as well as the distinction between freedom and free will. On one interpretation of incompatibilism, we choose our choices, but we are not responsible for them, because we are not responsible for that which causes us to choose as we do, and therefore, given a transferability of responsibility, we are not responsible for what we choose. The chip prevents our freedom to vote for the opposing candidate, which prevents our control in one sense, but does nothing to change the control of the choice we do make. We still have to decide what it meant to “freely” vote when we voted without constraint. In the case where someone votes by force, it’s clear that they are not blameworthy for that vote; voting by force doesn’t say anything about their character, only about their being a victim of manipulation. But if someone votes on their own accord, while this might say something about their character, it doesn’t prove that this person is responsible for their character. If someone’s biology, psychology, trauma, etc., causes them to have a poor character, then this person is not to blame for their poor character; their circumstances are to blame.)
Manipulation arguments are used by incompatibilists to show how compatibilist accounts of freedom fail. Derk Pereboom has developed some of the most famous manipulation arguments.
Imagine the following. Plum murders White. Plum desired to murder White, and approved of this desire. So the first- and second-order desire accounts of freedom are satisfied. Plum also acts according to his determination of what is right, so his act satisfies Susan Wolf’s account of freedom. And Plum acts according to his own reasons, satisfying John Martin Fischer’s account of freedom. Pereboom has us imagine the following four cases:
Case 1: Neuroscientists manipulate Plum’s brain, causing him to have the desires, reasons, beliefs, etc., needed to satisfy all the above accounts of freedom at the time he kills White.
Case 2: Plum was programmed by scientists at the beginning of his life to eventually have the desires, reasons, beliefs, etc., needed to satisfy all the above accounts of freedom at the time he kills White.
Case 3: Plum lives in a community of assassins who brainwash Plum into having the desires, reasons, beliefs, etc., needed to satisfy all the above accounts of freedom at the time he kills White.
Case 4: The laws of nature cause Plum to have the desires, reasons, beliefs, etc., needed to satisfy all the above accounts of freedom at the time he kills White.
We can summarize the manipulation argument as follows:
P1: If an agent is manipulated into having the desires, reasons, beliefs, etc., needed to satisfy all compatibilist accounts of freedom, then that agent is not free despite satisfying those accounts.
P2: There is no relevant difference between being manipulated by scientists or communities and by the laws of nature.
C: Therefore, an agent is not free if determinism is true.
Compatibilists have argued against both P1 and P2. Michael McKenna has argued against P1, saying that if the agent truly is acting according to their desires, reasons, beliefs, etc., then they really are free despite their manipulation. McKenna gives examples of real-world “manipulations”, such as unexpected pregnancy or religious conversions. And yet we don’t consider these events to rob us of our free will.
John Martin Fischer has argued against P2, saying that there is a relevant difference between being manipulated by neuroscientists and being “manipulated” by the laws of nature. When your actions are manipulated by a person, your choices are really the choices of the manipulator. But when your actions are determined by the laws of nature, your choices are your own; your actions are preceded by a causal history involving your own desires, reasons, beliefs, etc.
(Assuming I’ve characterized Fischer correctly, then this is really not convincing. Both cases involve a causal history where your desires, reasons, beliefs, etc., are in fact yours, but the fact that they are yours is caused by something external to you, such as the laws of nature or a manipulator. Your actions can say something about what you are without you choosing to be what you are. Compare: a dog’s actions can say something about that dog’s aggression without the dog choosing to be aggressive.)
Alfred Mele has a zygote argument which aims to show that one can have a causal history involving their own desires, reasons, beliefs, etc., and yet not have free will because they were manipulated. In this scenario, a goddess curses a zygote, determining that the person born from it will perform a specific action 30 years into the future. The person’s desires, reasons, beliefs, etc., are determined by the curse, because the person must make those choices that will lead them to make the ultimate choice they are fated to make at 30 years old. All of this person’s actions are preceded by a causal history involving their own desires, reasons, beliefs, etc., and yet they are not blameworthy for their actions, because the curse explains where these desires, reasons, beliefs, etc., came from.
Moving onto libertarianism, the main challenge for libertarianism is to make sense of how indeterminism allows us the control free will requires. If determinism is false, then the state of the world at one time does not guarantee the next state. So when one intends to perform an action, there is no guarantee that the intended effect will occur. Whether your choices have the intended effect will be a matter, to some degree, of coincidence.
(This leads to two questions. One is: How can you tell whether an action really reflects a person’s intentions? If someone stabs someone intending to kill them, but the laws of nature change at the last moment and stabbing someone helps them instead, then intentions and effects are disconnected. Good people could accidentally perform horrific actions and bad people could accidentally perform good actions. The second question is: If your actions are caused by random processes, then how are you responsible for them?)
Roderick Chisholm attempted to solve this problem through agent causal libertarianism. On this view, agents cause their actions. Your actions are determined in the sense they are determined by you; they are indetermined in the sense that they are not caused by anything other than you.
A challenge for this view is that it seems to commit you to substance dualism, an unpopular view with many objections. Plus, it’s mysterious how the agent who determines their own actions is able to do so. They must, somehow, be able to rise above all the physical, material events that shaped them. Why do agents get to stand outside of the causal order?
An alternative to agent causal libertarianism is event causal libertarianism. Instead of agents as causes, only events are causes. This is the view of Robert Kane, who posits self-forming actions. Kane concedes that most of the actions we take every day are relatively subconscious and automatic. However, in life we occasionally face significant, crossroad-type decisions. These are the actions that form who we are.
Jean-Paul Sartre gives what could be taken as an example of a self-forming action in the case of a French student who was debating whether to stay home or to join the French Resistance in WW2. Sartre’s famous idea of existentialism, which says existence precedes essence, is very much a doctrine of free will, as it means that who we are, our essences, is not determined before our existence. Rather, we first exist, and then our essence, what it means to be us, is formed by our choices.
But how are self-forming actions not determined? And if they are indetermined, then how are they free and not caused by chancy processes?
Kane says chaos theory might apply to brain activity, and thus brain activity, at times, is unpredictable. When we make a self-forming action, we hold two opposing decisions very strongly in our minds. One of these decisions must win out, and which one wins out might be a matter of quantum indeterminacy in the brain. If someone chooses self-forming action A, they might choose self-forming action B if you rewound the clock. But if our actions are up to chance, then how are they up to us?
Finally, we arrive at free will skepticism, the view that we do not have free will.
Illusionism = the view that we do not have free will, but we must live as if we do (either because we can’t help it, or because of the disastrous consequences of dispensing with free will).
(I disagree with illusionism for three reasons. One, I reject “noble lie” arguments; they’re explicitly anti-truth. Two, I do not need to see the world through the lens of free will. I can easily imagine the world without it. Three, there are no disastrous consequences at all from dispensing with free will. In fact, to the contrary, the no free will view is the least judgmental, least blaming, least accusing, least hateful, most understanding view you can have toward human behavior. Someone who believes in free will might argue that the no free will view robs us of meaning. But that doesn’t work; our experiences are meaningful. What happens to us, whether we suffer or flourish, is highly meaningful.)
Free will skepticism faces the challenge of explaining how our lives and relationships are meaningful when our choices only reflect our circumstances and not who we are (or: reflect who we are, but who we are reflects only our circumstances). And how do we make sense of punishing criminals?
(I would add: free will skepticism faces the challenge of explaining our intuitions around ‘should’ statements, moral facts, and our common sense attitudes about who deserves what, and about praise and blame.)
The famous Benjamin Libet experiment showed patient brain activity (‘readiness potential’) lighting up 550 milliseconds before an intention was formed to press a button. Some have interpreted this to mean that our internal sense of forming an intention occurs after our brain has already “decided” to perform the action. So all our “choices” are really just us bearing witness to what our brain is doing. But interpretations over the results of the experiment are debated. One response is that there is a lag between when we form an intention and when we recognize that we have formed an intention. And so the intention really is caused by us, not our brain, but it takes a moment for us to establish metacognition over our intentions.
A few of my notes:
So we end up with the following options, though there are many variations:
Libertarianism = we have free will, but free will is incompatible with determinism, so determinism is false.
Agent causal libertarianism = Choices are caused directly by agents.
Roderick Chisholm, Timothy O’Connor
Event causal libertarianism = Choices are caused by events.
Robert Kane, Mark Balaguer
Compatibilism = free will is compatible with determinism.
Classical compatibilism = Actions are free if they are unconstrained; we are free when we act according to our desires.
Thomas Hobbes, David Hume
Semi-compatibilism = We do not have leeway freedom on determinism, but we do have sourcehood freedom. The latter is what we need for moral responsibility.
John Martin Fischer, Michael McKenna, Taylor Cyr
Free will skepticism = the view that we have no free will.
Hard determinism = Free will is incompatible with determinism, but determinism is true, so we don’t have free will.
Hard incompatibilism = Free will is incompatible with determinism and indeterminism, so free will is impossible.
Derk Pereboom
I suppose in theory you could have someone who is convinced both of indeterminism and the incompatibility of free will on indeterminism, and would therefore be another type of hard determinist. You could also have another kind of semi-compatibilist who believes we don’t have sourcehood freedom, but we do have leeway freedom, and its leeway freedom that’s necessary for morality. I’m not sure how coherent that position is. And I can imagine a compatibilist who believes we need both leeway freedom and source freedom, and both are compatible with determinism. Looking at the SEP article for compatibilism by McKenna and Justin Coates, there are many variations of just compatibilism. https://plato.stanford.edu/entries/compatibilism/.
Further notable libertarians: Michael Huemer, Justin Capes.
Alfred Mele is agnostic about free will, but has defended event causal libertarianism (https://ndpr.nd.edu/reviews/aspects-of-agency-decisions-abilities-explanations-and-free-will).
Further notable compatibilists: P.F. Strawson, Harry Frankfurt, Carolina Sartorio.
Further notable free will skeptics: Galen Strawson, Bruce Waller.
Manuel Vargas has defended ‘free will revisionism.’
Peter van Inwagen has been called a ‘mysterian’ about free will; he doesn’t see how we could have it, but he doesn’t see how we couldn’t have it either.
No comments:
Post a Comment