On Agent Causation

Among the band of philosophers who hold that free will is supernatural, one of the reigning ideas is called agent causation. This hypothesis states that volitional acts are a special category of event, one that is caused not by any other event but – in some deeply mysterious way – by the agent itself. Philosopher Roderick Chisholm describes this as follows:

If we are responsible… then we have a prerogative which some would attribute only to God: each of us, when we act, is a prime mover unmoved. In doing what we do, we cause certain events to happen, and nothing—or no one—causes us to cause those events to happen.

The consequence of agent causation is that free will is not a process but some sort of irreducible substance: one that spontaneously originates acts and decisions, unconnected to the causal chain that binds together all other causes and effects. The usual apologetic corollary is that even God cannot intervene in or influence this process short of destroying free will altogether. It’s plain that this is just the religious doctrine of the soul, the supernatural “ghost in the machine”, portrayed in technical philosophers’ language.

Agent causation depicts human free will as a binary state – a quantity which can either be present or not present, but which has no internal structure and cannot be subdivided. However, this is obviously false, which makes this entire view unsustainable. Free will is not a mathematical point; free will is a complex bundle of contingent desires, habits, and predispositions, which can be added to, altered or removed.

You can determine this by empirical studies of human behavior. There are countless things that human beings could do that we do not do and do not feel any desire to do. On the other hand, the vast majority of us do experience desires to have sex with an attractive partner, to consume foods high in fat and sugar, or to form tight emotional bonds with parents and relatives. Human free will, then, is not just an irreducible point source that bubbles up actions at random; it operates within a defined set of parameters, giving rise to a predictable variety of behaviors (anthropologists call them cultural universals).

You can also determine it by the evidence of the human brain: it’s well known that certain, specific kinds of brain damage alter desires and behavior in predictable ways. Dementias such as Alzheimer’s disease often cause loss of interest in religion, while epileptic seizures in the temporal lobes can induce religious experiences. Damage to the frontal lobes leaves people unable to control their behavior or ignore sudden impulses; injuries localized to the left hemisphere often cause depression, while injuries to the right can leave people constantly and inappropriately euphoric. And of course, drugs and intoxicants also have reliable, predictable effects on behavior.

If free will was an irreducible, nonphysical substance, producing actions free from external causation, then we should not see brain damage affect desires or behavior – much less change them in the predictable ways that neurologists observe. That we do see this shows beyond a reasonable doubt that it is false that “nothing… causes us” to make the decisions we do. Our decisions manifestly are caused.

Of course, there are other varieties of supernatural dualism that are not as clumsy as agent causation. But what these other varieties have in common is that they must give up the line in the sand. They cannot declare, as agent causation does, that we are purely supernatural beings whose decisions ultimately arise from the soul and nowhere else. Instead, these other dualisms must acknowledge that we are, at least in part, material beings, and that changes to the physical composition of body and brain can affect and alter our selves. Whether they realize it or not, advocates of these beliefs are drawing closer to atheism, as they implicitly grant that we are not spirits whose choices arrive from outside the world, but physical beings whose acts are an inextricable part of the fabric of cause and effect.

About Adam Lee

Adam Lee is an atheist writer and speaker living in New York City. His new novel, Broken Ring, is available in paperback and e-book. Read his full bio, or follow him on Twitter.

  • Mrnaglfar

    I’ve had this debate with too many philosophers. There’s another problem that is overlooked however, which I never did get a straight answer on:
    Option 1 – The ‘mind’ (or consciousness) is a purely physical thing (as represented by the net activity of the brain/nervous system), and we can trace cause and effect from inputs interacting with neural wiring. This option leaves us with determinism.

    Option 2 – The brain alone is not sufficient to explain consciousness, and as such, we require a separate mind that isn’t the brain. They like to say “brain states are correlated to mental states, but the two are not the same thing,” without seeming to have too much of a clue as to what that means. However, this extra mind leaves us the problem of conservation of energy. Either this non-physical mind can only observe our brain state and then experience something like ‘red’ or ‘anger’ (which they like to call qualia – something they are sure physical things can’t experience), in which case it would simply be around for the ride passively observing. This also leaves determinism.

    In the other case, the mind-stuff, whatever it is, can actually cause the body to act in a certain way making it an active consciousness rather than a passive one. However, in order to make the body act, this mind would need to add energy to the system to manipulate it, meaning energy is being created, or moved, with no physical cause. This gives us free will, but violates the conservation of energy.

  • Christopher

    AS a determinist, I see the very concept of “free will” as being absurd because it implies the ability to void causality itself – after all, a truly “free” choice can only be made in a complete physiological, psychological and sociological vacuum (otherwise, such factors will determine what “choice” is made). But seeing as to how these conditions do not exist, no “choice” ever made by anyone can be truly “free” as there are predisposing factors influencing every level of thought that the indivual experiences.

    Even if you were to invoke some kind of supernatural agent (like a “soul” – whatever the fuck that is…), you would still have to demonstrate how it would interact with the environment that the individual is in – thus subjecting it to causallity, as something would be necissary to cause this hypothetical “soul” to act upon the circumstance in question.

    When all is said and done, causallity has the last laugh…

  • http://thewarfareismental.typepad.com cl

    I don’t think volitional events are caused by other events, although they are most certainly influenced by them.

    The consequence of agent causation is that free will is not a process but some sort of irreducible substance: one that spontaneously originates acts and decisions, unconnected to the causal chain that binds together all other causes and effects.

    Why does it have to be EITHER/OR? I would agree that free will is not a substance, and is better described as an ability. I don’t think agent causation is necessarily spontaneous, either. It is influenced by prior events and feelings, for example. When a volitional act cannot be said to be influenced by prior events and feelings, then that act might stem from instinct. A baby’s laughter may seem quite spontaneous, but can actually proceed from instinct or influence. I also have other ideas that you’ll unfortunately discount a priori because of your worldview, so I’ll have to leave those well alone.

    …free will is a complex bundle of contingent desires, habits, and predispositions, which can be added to, altered or removed… Human free will, then, is not just an irreducible point source that bubbles up actions at random; it operates within a defined set of parameters, giving rise to a predictable variety of behaviors…

    I like those. I think they are reasonable paraphrases of free will that I mostly agree with, and none of this detracts from any of my ideas concerning spirit or soul.

    If free will was an irreducible, nonphysical substance, producing actions free from external causation, then we should not see brain damage affect desires or behavior – much less change them in the predictable ways that neurologists observe. That we do see this shows beyond a reasonable doubt that it is false that “nothing… causes us” to make the decisions we do. Our decisions manifestly are caused.

    Well, to you that’s what it shows. Incidentally, if free will produces actions, how is it free? I’ve also got plenty to say about AGITM, and I suggest opening a comment thread for it and your other essays on Ebon Musings, BTW.

    At any rate, let’s grant that we have a spirit, and that it’s some sort of irreducible substance. Free will would be an ability of that substance, not necessarily the substance itself. And I say that our decisions are manifestly influenced, not caused. To say they are caused is to potentially minimize or even reject the role of agent decision. If acts of murder are caused, how can we truly hold someone accountable for such crimes? Murder would no longer be a choice if it is caused. On what grounds might we punish the murderer if his or her decision is caused, not chosen?

    Instead, these other dualisms must acknowledge that we are, at least in part, material beings, and that changes to the physical composition of body and brain can affect and alter our selves.

    How anyone could deny that at least in part we are material beings seems strange to me. That changes to body and brain wouldn’t affect and alter our selves seems even stranger.

    Whether they realize it or not, advocates of these beliefs are drawing closer to atheism, as they implicitly grant that we are not spirits whose choices arrive from outside the world, but physical beings whose acts are an inextricable part of the fabric of cause and effect. (ital. mine)

    Again, false dichotomy. Why must I be one or the other? Why does it have to be EITHER/OR? You can alter the output of a light several ways, including changes directly to the physical scaffolding, as well as changes in the flow of electrical current going to the physical scaffolding. It would be equally absurd to force an observer into a decision as to whether a working light is EITHER the bulb OR the electricity.

  • http://whyihatejesus.blogspot.com OMGF

    Welcome back Mrnaglfar! I’m glad to see you back around these parts.

  • Sue Denim

    I have heard the dualist argument that the spirit/mind/free will is affected by the brain but that it is also more. The problem I have with that is, if the dualistic entity is indistinguishable from the naturalistic one then naturalism must necessarily be true due to Occam razor and parsimony. So I would say that every time a new area of consciousness is successfully mapped to the physical brain it is another nail in the coffin of dualism.

  • http://www.myspace.com/driftwoodduo Steve Bowen

    each of us, when we act, is a prime mover unmoved. In doing what we do, we cause certain events to happen, and nothing—or no one—causes us to cause those events to happen.

    Might as well descend all the way into solipsism in that case.

  • Cerus

    cl said:

    If acts of murder are caused, how can we truly hold someone accountable for such crimes? Murder would no longer be a choice if it is caused. On what grounds might we punish the murderer if his or her decision is caused, not chosen?

    Causality being the sum of its parts (even if knowing all the parts with perfect precision is an unfathomable task), the reality of punishment is a part that can act in concert with other parts (neurological predisposition for violence, education, moral base, etc.) to effect fewer incidents of murder being caused.

  • http://anexerciseinfutility.blogspot.com Tommykey

    Dementias such as Alzheimer’s disease often cause loss of interest in religion

    That brings up a question I have asked. What if a man is a professing Christian, but after being afflicted with Alzheimer’s, he forgets his religion, spends money on a prostitute, and dies in the act. He died in a state of sin, right? So, does he go to hell, or does God give him a pass because of his affliction?

  • http://thewarfareismental.typepad.com cl

    Cerus,

    …the reality of punishment is a part that can act in concert with other parts to effect fewer incidents of murder being caused.

    So are you saying the person who murders is ultimately not responsible for his or her actions, and that punishment exists merely as a deterrent with no basis in justice whatsoever? Am I understanding that correctly?

  • Mrnaglfar

    cl,

    So are you saying the person who murders is ultimately not responsible for his or her actions, and that punishment exists merely as a deterrent with no basis in justice whatsoever? Am I understanding that correctly?

    Deterrent? Most definitely.
    Justice? Depends how you want to define justice.

    Also, by extension, if the person who murders is ultimately acting deterministically (not responsible, or a choice if you want to call it that) so is the person who writes the law, enforces the law, and is victimized by the crime.

    Whether free will or determinism is the case seems to be a largely irrelevant question without any actual practical implications.

  • Anon

    It’s plainly apparent our wills are not “unmoved” or wholly separated from the influences of the wold. It’s boggling someone can seriously argue as such.

    Even if you were to invoke some kind of supernatural agent (like a “soul” – whatever the fuck that is…), you would still have to demonstrate how it would interact with the environment that the individual is in – thus subjecting it to causallity, as something would be necissary to cause this hypothetical “soul” to act upon the circumstance in question.

    I don’t understand how that implication works. X interacts with Y, therefore Y causes the acts of X? I’ve heard mention elsewhere of a theory by Roger Penrose and Stuart Hameroff which hypothesizes that information from (as far as we know, *uncaused*) wavefunction collapses at the quantum level (which can be thought of as the “choices” particles make) get carried over as intrinsic variables in the brain’s computational architecture. This is somewhat outlandish, but it is plausible, and could allow for some room of physically-uncaused “freedom” in our choices. Also, if you want macroscopic verification of the random nature of quantum mechanics as if manifests in the universe, merely look to the night sky to see the asymmetric assortment of stars in the sky: a distant and large-scale memory of the super-small chance (?) happenings close to the time of the Big Bang.

    If acts of murder are caused, how can we truly hold someone accountable for such crimes? Murder would no longer be a choice if it is caused. On what grounds might we punish the murderer if his or her decision is caused, not chosen?

    I think it should be known these questions/statement are independent of each other. I’ll take a crack at responding to each: [1] We can “hold” people accountable simply by theorizing an account by which a person or persons caused or influenced a situation, also including descriptions of their internal state (mood, mental disorder, intentions, etc) for the convenience of deciding how we wish to react to these hypothetical accounts. And, what exactly is “true accountability”? How is accountability any more “true” if a will has an uncaused component to it? [2] Does “choice” necessarily entail absence of fully-determining cause(s)? How so? [3] The answer is on whatever ground we wish to tether our choice to punish to. If you want *morally justifiable* ground, then merely look it up in whatever dictionary of ought/ought-nots you wish to live by…

    One more thing to cl, in the same vein as Mrnaglfar’s latest statement about “actual practical implications”: On what ground can we hold people morally accountable if we live in a world where the nature of choice is fundamentally uncertain? I hope that one’s a toughy.

    I have desire to offer an OPEN CHALLENGE to other commenters. Here it is: give an objective definition of (1) responsibility, (2) deserve/blame, (3) justice/fairness, and (4) freedom. That last one may take up books because it so loaded with potential meanings, and may possibly fit better in Ebonmuse’s Free Will series. The first three ideas are the moralistic ones, and so it should be instructive to finally get at any useful, clear, objective meanings behind them, preferably ones that could be put into use in the Real World.

  • http://thewarfareismental.typepad.com cl

    Anon,

    …what exactly is “true accountability”?

    Accountability as I use it refers to the idea that one should be held responsible for one’s choices, which entails that they accept the consequences of those choices. The concept assumes that certain actions are unacceptable and deserving of punishment, and admittedly, the consequences are often decided arbitrarily by others. Whether the underlying reasoning for this assumption is because the actions are morally wrong or simply because they disturb the harmonious flow of society doesn’t have much bearing on where I’m going with this, so let’s not get too bogged down there. True accountability as a phrase has exactly the same meaning to me as accountability, so I don’t know what else to say to you there.

    How is accountability any more “true” if a will has an uncaused component to it?

    Bear with the false dichotomy for a second; If our actions are ultimately caused and not chosen, then outside factors determine our actions. In such a case, to punish someone for murder becomes tantamount to punishing them for outside factors over which they did not have any control. We can argue that such is pragmatic in the sense that yes, these punishments serve as deterrents, and further causes (influences) against murder.

    Does “choice” necessarily entail absence of fully-determining cause(s)? How so?

    As much as false dichotomies annoy me, I don’t have a canned answer for that, and this is one example of an area where to me, middle ground seems elusive. I suppose one could posit that any particular choice is merely the result of several fully-determing causes, but such doesn’t seem very helpful.

    On what ground can we hold people morally accountable if we live in a world where the nature of choice is fundamentally uncertain?

    You essentially echoed my question, and I’ll object to the loading of accountability in a context of morality. The justification of accountability can be moral or pragmatic. Even if the nature of choice is fundamentally uncertain, the vast majority of people are fundamentally certain that particular behaviors cause genuine privation for other people and in society at large. On those grounds, and on the grounds that the nature of choice is currently unknown, hence choice may be actual and not illusory, we have sufficient reason to prosecute criminals on either grounds, in my opinion.

  • http://she-who-chatters.blogspot.com/ D

    Man, I get kinda sick of hearing people say that we need free will in order to believe in morality. Says friggin’ who? First off, if there are moral facts at all (and if you subscribe to a moral system, then there are “facts under that system” at the very least), then they apply no matter whether there’s free will or not – unless you simply conflate free will with moral responsibility, in which case you’re begging the question and need to start over. “Unjustified killing” is still “unjustified killing,” theft is still theft, and rape is still rape, no matter whether the perpetrator had a choice about it. Second, whatever chain of moral responsibility you want to put in place still has to stick to a causal chain (since one cannot be morally responsible for something one had no role in causing), and if anything, determinism makes this fact more explicit – it also clarifies the need for a distinction between being causally relevant and being morally blameworthy. And third, if you want free will to be true so you can justify punishment, then you want someone to suffer for wrongdoing, which involves wanting someone to suffer, which in turn is sadistic.

    Sure, it’s a lot harder to justify punishing someone if they don’t have “ultimate control” over their actions, but retributivism is primitive and dumb. A determinism-based justice system would instead seek to rehabilitate those who murder/steal/rape/harm others from being “the kind of people who do those things” into “the kind of people who can live in peace.” By itself, the suffering of a criminal accomplishes nothing (it does nothing to negate the victim’s suffering which has already occurred), and the central goal of any corrective system ought to be to rehabilitate (i.e. “correct”) its charges. Doing this without causing suffering to said charges would probably look more like psychiatric care than prison, though – which makes sense, in light of the common-sense notion that “if you think it’s OK to hurt someone else, then there’s something wrong with you.”

    Now, the real world is a lot messier, and due to epistemological and logistical constraints, we more or less have to rely on brute force deterrence. On top of that, we’re not yet culturally developed enough to let go of our desire to see wrongdoing punished (and thus to see wrongdoers suffer). Furthermore, there are many would-be criminals who simply need the example that “this shit won’t fly” in order to not try [crime du jour] in the first place, which makes retributive justice contingently useful for the time being. But that doesn’t mean we can’t recognize that the system is flawed (instead of enshrining it, as those who would enjoin moral responsibility with free will apparently wish to do) and work for the day when we can profess justice more humanely.

    OK, I guess I smuggled in the fact that I’m a consequentialist. The point is, determinism at once saps the sadism from any theory of justice, while at the same time making it glaringly obvious that the problem with morally bad people is not the person per se, but rather whatever it is that causes the person to behave in a morally bad fashion. Someone please remind me of what the problem is?

  • http://she-who-chatters.blogspot.com/ D

    Dammit… the last sentence should read, “Someone please remind me of what the conflict is between determinism and morality?” My kingdom for an “edit” button!

    Also, “vengeance” is a complete defeater to the idea that nothing and no one causes our actions. Vengeance cannot but be caused. I’m not familiar with Chisholm’s work, but in this case, I think he went off the deep end. This “agent causation” noise isn’t even coherent.

  • Leum

    Anon,

    I have desire to offer an OPEN CHALLENGE to other commenters. Here it is: give an objective definition of (1) responsibility, (2) deserve/blame, (3) justice/fairness, and (4) freedom. That last one may take up books because it so loaded with potential meanings, and may possibly fit better in Ebonmuse’s Free Will series.

    What do you mean by an objective definition? Responsibility, deserve, blame, justice, fairness, and freedom are human concepts. We made them up, and they only exist as long as we believe in them. Anything related to human ideas and conceptions is inherently subjective. In the words of Death:

    Then take the universe and grind it down to the finest powder and sieve it through the finest sieve and then show me one atom of justice, one molecule of mercy.

    D, I agree with you completely. A consequentialist or utilitarian ethic demands justice while forbidding vengeance and works to correct problems much better than the retributive model of justice ever has.

    There is a pitfall in a utilitarian justice system, though. Technically speaking, it doesn’t entirely matter if all the people punished are guilty, as long as the deterrent effect is strong enough. This can be corrected, of course, by taking the view that knowing the truth is inherently a better condition than believing a lie. I believe Ebon touched on that point in his essay on Universal Utilitarianism, but I can’t find the essay.

  • http://thewarfareismental.typepad.com cl

    D,

    “Unjustified killing” is still “unjustified killing,” theft is still theft, and rape is still rape, no matter whether the perpetrator had a choice about it.

    Certainly. But are they still immoral if the perpetrator had no choice? You yourself answered thusly:

    …since one cannot be morally responsible for something one had no role in causing…

    As for,

    …if you want free will to be true so you can justify punishment, then you want someone to suffer for wrongdoing, which involves wanting someone to suffer, which in turn is sadistic.

    I, personally, don’t want free will to be true or not true. I would just like to know whether it is true or not true. Secondly, I don’t think justification of punishment entails that one wants another to suffer, and further, I don’t necessarily think that wanting one to suffer for murder or any other crime is inherently sadistic. Although I agree that a criminal’s suffering does not negate a victim’s privation, a criminal’s suffering might just influence or cause this rehabilitative change-of-heart you seem to be advocating. Not incidentally, I can think of quite a few lessons I learned through suffering for things I did wrong. How? After experiencing the privation, I realized, “Wow, this is the exact same feeling my actions produced in so-and-so, and that’s not right.” So very clearly, the desire to see wrongdoers punished can stem from several other positive and useful motivations than the sadistic, primitive motivation to inflict pain.

    Also, “vengeance” is a complete defeater to the idea that nothing and no one causes our actions. Vengeance cannot but be caused.

    I’m not persuaded. I would say that vengeance defeats the idea that nobody influences our actions. Vengeance is an emotion, a feeling, a motivator – and not an action itself. One might be inspired towards a murder that is wholly an act of vengeance, but vengeance has not caused the murder – merely influenced it. At some point, the person had to either act on or work through their feelings of vengeance. IOW, at some point, a choice was either made or deferred.

  • http://prinzler@calpoly.edu Paul

    Echoing others:

    Deniers of determinism object to punishing someone if they had no control over their actions, but this is an empty argument. If, under determinism, we are merely meat machines, then consider the following example of a (regular, metallic) machine that goes haywire for some reason and is harming property or people. In that case, we fix the machine. So if we are meat machines, then our goal when one of us does something that harms others or their property is to make it such that they don’t do it again. The only difference is that we’re pretty complicated machines, and our efforts aren’t always %100 effective because of this complexity (we’re not sure of which lever to pull, so to speak, to fix the machine). Punishment for crimes then becomes not a moral act but a utilitarian one (and might change accordingly from our current system of punishment).

  • Mathew Wilder

    I don’t think the notion of free will has any cognitive content. It seems to me a term used to somehow deny our non-uniqueness from the rest of nature. “Meat machines” is a crude, but I think accurate way of describing ourselves.

    As Nietzsche said quite rightly, “A thought comes when it wills, not when I will.” I think psychology and neurology are more and more casting doubt on our notions of agency and willing.

    Recommended reading includes Leiter and Knobe’s “The Case for a Nietzschean Moral Psychology” and Leiter’s “Nietzsche’s Theory of the Will” (I think the second title is correct). Both were available online last I checked. Knobe also has done some interesting work on his own too.

  • http://she-who-chatters.blogspot.com D

    @ cl:
    You asked, “But are [murder, rape, and theft] still immoral if the perpetrator had no choice?” Of course they are! What makes bad things bad is that the moral demand of the system in place has been violated. For a consequentialist, harm has occurred; for a deontologist, rules have been broken; for a virtue ethicist, good people don’t behave that way. Even a commonly accepted definition of evil, “Knowing better but doing worse,” makes no mention whatsoever of “choice” – one can both “have knowledge” and “take actions” without making metaphysically free choices. We can use our moral language to describe situations just fine without invoking free will.

    Also, when I said that “one cannot be morally responsible for what one had no role in causing,” note that “to cause something” does not require free will. For any object X and event Y, the propostion “X caused Y” means nothing more than that X was antecedent to Y in the causal chain, and so X is causally responsible for Y even if X had no choice in producing Y. The chain of moral responsibility is a bit fuzzier; for example, we tend not to blame “inanimate objects” (which term I use for convenience alone, I’m aware that nits could be picked here). Blaming people, on the other hand, can be productive: as Paul points out, the point is (or at least ought to be) to fix the “malfunctioning people,” while blaming them is not an end in itself. I maintain that the fuzziness in chains of moral responsibility is mostly due to incongruities between our ad-hoc “common sense” ideas of morality and the emergent finer points of our proposed moral systems with which, on examination, we turn out to be uncomfortable.

    In answer to your bit on privation, I’ll continue Paul’s machine metaphor: when a machine malfunctions, you don’t punish it, you take it off the floor, find out what’s wrong, fix it, and reintegrate it into normal operations. But note that interfering with the machine’s normal operation (removing it, working on it, etc.) is by no means an end in itself, but simply a means to the end of fixing the machine. With people, similarly, the goal ought to be to improve them, to make them better persons, and if we can do so without harming them, we ought to because harming people is bad (if you’re a consequentialist like I am). Imagine that another option was available in those times of privation you were thinking of: what if exactly the same result could be achieved, but without causing any further harm at all? Would the “same result, less harm in the process” option not be the better of the two? The fact that harming people can be productive is a contingently pragmatic point of fortune for us imperfect mortals, but otherwise merely incidental; the point is that harm is bad and ought to be minimized, including the fixing of our malfunctioning “meat machines.” In other words: punishment has pragmatic use (for now), but is not in principle necessary; therefore the desire to see someone punished is to at least a small extent a desire for unnecessary harm to come to a person – and therefore sadism. QED.

    As for the vengeance bit, vengeance means “punishment inflicted in retaliation for an injury or offense,” which makes it an action. And even if it weren’t, then I’m talking about “the act of taking vengeance,” which is an action. Vengeance is, by definition, a response to (i.e. because of) something else, which means that on Chisholm’s view, the idea of vengeance is perforce incoherent (which conclusion is silly, ergo his idea is silly). An act of vengeance cannot be such without the initial injury/offense to trigger it – that initial act may not be the sole cause, but it is a cause, which makes vengeance at least in part caused – and makes the idea that “nothing or no one causes our actions” patently false.

    @ Leum:
    The idea that believing the truth is inherently better than believing a lie creates both moral and epistemological headaches, and these are easy to cook up just by putting “truth” at odds with “good” or “knowable.” (As a side note, it’s more desirable to believe the truth to most people, but heroin can be desirable under the right circumstances, so whatever that’s worth…) A far cleaner way to avoid the problem is simply to play the odds: punishing innocent people in order to deter innocent people will necessitate a cover-up, because if people find out that punishment has nothing to do with guilt, they will be less motivated to avoid being guilty of wrongdoing (and thus the benefit of the lie would be lost). Cover-ups are unlikely to last forever, and the longer such a thing goes on, the worse the public reaction will be when it is uncovered. The efforts required to sustain such a lie will eventually outstrip one’s ability to meet them – to say nothing of the drain on utility caused by the unproductive suffering (it’s not fixing actual criminals) and the cognitive dissonance required of the people maintaining said lie. Therefore, it is better by far to play it straight, because the odds are simply stacked against maintaining the beneficent lie in the long run. And engaging in such deceit for immediate benefit is short-sighted and irresponsible.

    The pitfalls of consequentialism are easy to imagine, but what I don’t understand is why people don’t simply recognize that, hey, “These nightmare scenarios are themselves not conducive to utility, and so probably ought to be avoided in the first place.” This extends even to encouraging efforts to make the world such a place that these nightmare scenarios are unlikely to come about.

  • prase

    When discussing free will and determinism (which is in my opinion a false dichotomy), it is good to define these terms. This is not easy even for determinism. In the first approximation one may say that determinism is a belief that we can in principle predict any future state of the world from the knowledge of its history until today and some set of natural laws. Problems may come when one has to consider in greater detail 1) what in principle means and 2) what rules are worthy being called natural laws.

    For 1) we have to realise that even in the reductionist universe we are unable to practically calculate the future states of the universe, if only because we would need to have second universe at least as big as the first one only for storing the exact data about the initial state. (Not mentioning complications due to quantum physics.)

    For 2) one can be a determinist without being a reductionist – there can surely be laws which are neither simple nor local in time and space. Imagine a situation where one has to compute the position of a particle after some specified time t. Of course we take the initial position and velocity and use the Newton’s laws to get x(t)=x(0)+vt. But suppose we instead had a large book called “The Law” full of tables where for each initial position and velocity there was the whole trajectory described point after point? It wouldn’t be a parsimonious law, but still wholly deterministic. As a generalisation, can we in principle write down the entire history of the universe (future included) and declare it a law? If so, determinism is true by definition.

    Therefore it seems be more useful to speak about reductionism than about determinism. Reductionism holds that the universe can be separated into smaller sections whose behaviour we can practically predict with precision that can be made sufficiently high when the size of the section is made accordingly small and the time interval we are predicting for enough short; the predictions are made using natural laws that can be inductively discovered. The laws can be deterministic (roughly speaking when the precision of the predictions can be made arbitrarily high) or non-deterministic. In the latter case there are events that are principially unpredictable, in other words random.

    It may be also useful to remember that the concept of causation has sense only under some set of natural laws. We may say that event X is a cause of event Y when X sets up conditions from which we can predict Y. The more the set of laws is indeterministic the less we have way to decide whether the events are in causal relation or not.

    Given this, it seems that the counterpart of determinism is randomness rather than free will, which is only a fancy term with empty content (I think Ebonmuse already pointed this out in one of his earlier posts). Since the concepts of accountability, responsibility, justice and all other related things are based on causality, determinism is needed rather than excluded in order to maintain them.

    By the way, Overcoming bias had a nice analysis about when we feel we have free will, unfortunately I was unable to find the exact post.

  • Chet

    I don’t see how a lack of free will causes any problem for justifying the justice system. If, indeed, murderers are just “meat machines” following a program that says “kill people”, then so too must police, prosecutors, judges, and bailiffs be meat machines following a programming that says “capture, convict, and punish murderers.” And the rest of us are just following a programming that says “consider it justice when they do so.”

    We can’t simultaneously assume free will for the police and judges but no free will for murderers. If they’re following a programming, so is everyone else – a programming that says “capture and punish murderers, and call it ‘justice’ when doing so.” Or, on the other hand, we’re all free-choice people making choices.

    Either way, the result is our perception of justice being done. It’s not arbitrary or capricious to punish murderers if murderers are simply choiceless drones; if our programming is to punish murderers, than we have no choice but to do so.

  • http://whyihatejesus.blogspot.com OMGF

    D and others, you may be interested in this.

  • Polly

    What we call “justice” mainly serves the purpose of either deterrence or revenge. There’s no real need for “justice as moral principle” except perhaps as it relates to giving the families of victims “closure.” But, if you can distinguish between that psychological need and revenge have at it.
    I can’t.

  • Polly

    That’s not to say that revenge itself doesn’t have social utility! It may prevent extra judicial retaliation – and anarchy. If you’re one of those that thinks anarchy is a problem.
    I, myself, am undecided.

  • http://she-who-chatters.blogspot.com D

    Woo-hoo! Looks like Dr. Dawkins has finally come over to the dark side! I must confess, I was a bit concerned that his rant against genetic determinism (with which I agree) would also extend to determinism in general, but now I can continue my hero worship of him unabated. Thank you, Chet, for facilitating my willful stagnation!

    @ prase:
    A rough-and-ready example of determinism (definition or explanation) is that “propositions have time-independent truth value.” For instance, if I say, “I shall punch a baby on January 22nd,” I may or may not do so (as far as you know). However, it is the case that tomorrow, I either will punch a baby or I will not punch a baby. If I do, then tomorrow, my statement from today will turn out to be true (rather, we’ll know it’s true). Furthermore, on the determinist view it was true when I said it today; it also would have been true if someone said it 500 years ago – and this goes for all possible propositions, stated or unstated. Moreover, what’s fueling the truthfulness of these propositions is the fact of the matter which cannot be altered by anyone or anything (i.e., because it is the case that I will punch a baby tomorrow, I cannot actually refrain from doing so, and therefore the proposition is true). Note that we cannot know with certainty beforehand what the truth value is of the proposition in question – determinism is epistemologically useless, but that’s beside the point. The point is whether propositions about the future either have truth value or do not have truth value – if free will is real, then tomorrow’s baby-punching schedule is not just epistemologically uncertain, it is metaphysically indeterminate.

    Either that, or future events create causal chains that flow backwards in time to make propositions true or false. But that’s just silly.

  • Leum

    You don’t need free will for metaphysical indeterminacy. You can also believe that some events are random and lack causality, or are probabilistic so that a cause can create multiple effects only some of which are realized.

  • http://she-who-chatters.blogspot.com D

    Free will is a sufficient condition for metaphysical indeterminacy, even though it’s not a necessary one. Similarly, you don’t need a car to have a vehicle, but if you have a car, then you have a vehicle.

  • prase

    @ D,

    thanks for your comment, you helped me discovering the problem of future contingents which I was unaware of. I was always thinking the sort of determinism you have just described is self-evident since it is clear that any proposition, whether about future or past, is either true or false. Well, it still seems quite self-evident to me, but at least I see that not everybody accepts that.

    Nevertheless, I don’t see how exactly this weak version of determinism rules out the free will. At least until somebody can explain to me what the free will actually is.

  • http://she-who-chatters.blogspot.com/ D

    @ prase:
    Yeah, it’s self-evident to me, too. Unless, of course, metaphysical indeterminacy actually is the case, whether by free will or by quantum weirdness or what-have-you. That jams up all sorts of strict determinism.

    This isn’t really a “weak” version of determinism, though: it’s still the same idea that the whole Universe, human beings included, is on causal rails. It’s just presented differently: instead of arriving at strict determinism from a pervasive causality, we instead infer that if propositional truth value is time-independent, then events cannot possibly happen in any other fashion than that in which they actually did happen (in other words, pervasive causality necessarily entails time-independent truth value). The reason it rules out free will is because, if it is the case that the future is determined entirely by antecedent causes, all the antecedent causes that can ever exist are already in place and thus we are powerless to change it. Sure, we can talk about “available options” and “making decisions,” but these are merely descriptions of events – we do in fact engage in deliberative thought and arrive at decisions, whether or not we are capable of acting otherwise at the end.

    Free will is the idea that, for any such deliberative process or decision leading to an action X, there is a logically possible world in which everything is identical down to the minutest detail except that action Y was taken instead. As an example, this morning I ate a bowl of Lucky Charms. There was also a box of Total in the pantry, though. I thought long and hard over which to eat, but ultimately decided on Lucky Charms (and then ate them). Determinism maintains that, due to all the causal factors leading up to that point, I could not possibly have chosen Total instead (though I did in fact make a choice, I could not have acted otherwise). If I had genuine free will, then it is possible that, regardless of all the causal factors leading up to that point, I could have chosen Total instead. How? Damned if I know.

  • Donald

    Man, I get kinda sick of hearing people say that we need free will in order to believe in morality. Says friggin’ who? First off, if there are moral facts at all (and if you subscribe to a moral system, then there are “facts under that system” at the very least), then they apply no matter whether there’s free will or not – unless you simply conflate free will with moral responsibility, in which case you’re begging the question and need to start over.

    No, begging the question is assuming what has to be proved. But there is nothing to be proved here–no concept of moral responsibility exists that doesn’t include free will. That’s why utilitarians don’t morally object to cancer, though harm has occurred: no agent has caused the cancer. That’s why deontologists don’t morally object to tornadoes that kill people, though the rule against killing has been broken: no agent has killed those people. And that’s why virtue ethicists don’t morally object to homicidal maniacs’ failure to possess virtues, because the maniac is not an agent and can’t do anything about it.

    “Unjustified killing” is still “unjustified killing,” theft is still theft, and rape is still rape, no matter whether the perpetrator had a choice about it.

    Not really. How can something be “unjustified” if it was inevitable? That’s why nobody talks about “unjustified death,” because you can’t control when your prostate cancer kills you. Or an “unjustified” hurricane, or bolt of lightning. And a dog that runs away with someone’s shoe hasn’t committed “theft,” because dogs aren’t agents.

  • http://www.patheos.com/blogs/daylightatheism Ebonmuse

    Free will is the idea that, for any such deliberative process or decision leading to an action X, there is a logically possible world in which everything is identical down to the minutest detail except that action Y was taken instead.

    I disagree. That’s the libertarian version of free will, but it’s not the compatibilist view. My view is that an action is free if it is chosen by a rational agent as the result of a process of deliberation in accordance with that agent’s desires and free of coercion by other agents.

    I deny that “could have done otherwise” is a necessary component of free will. Given all the causal factors acting on me from the beginning of my life up until now (including some that I set in motion myself, as the results of decisions arising from yet earlier causal chains), I could not have chosen otherwise than I did. All this is really saying is that I made the choice I did because that’s the kind of person I am. And in that realization, I believe, lies true moral responsibility: far truer and more genuine than we could ever obtain from a philosophy that treats free acts as random and inexplicable lightning bolts without cause or motivation.

  • Leum

    Is there any use in calling that sort of choice free will, though? It’s a good way of expressing moral culpability, but there seem to me too many strings attached for the term free will to be a good one. Uncoerced choice or rational choice seem like better terms that avoid the libertarian connotations.

  • http://thewarfareismental.typepad.com cl

    prase,

    I’m taking notes from you. Unlike me, you rarely say anything, but when you do, it’s usually golden.

    D,

    …when a machine malfunctions, you don’t punish it, you take it off the floor, find out what’s wrong, fix it, and reintegrate it into normal operations.

    And when a people malfunction, does not a roughly analogous process occur? We take them out of the general population, attempt to find out what’s wrong with them, attempt to fix them, and reintegrate them into society, typically without harming them, no?

    Would the “same result, less harm in the process” option not be the better of the two?

    Certainly.

    …the point is that harm is bad…

    Tell that to an angry bear about to rip your head off!

    …therefore the desire to see someone punished is to at least a small extent a desire for unnecessary harm to come to a person – and therefore sadism.

    I find this highly unpersuasive. Why the a priori correlation of punishment and harm? Surely, the vast majority of punishment inflicted upon criminals in America does not cause physical harm to them, correct? And even if we grant that all punishment entails physical harm, surely more than one motivation can exist for seeing punishment carried through, correct? I, personally, want criminals to be punished for two reasons: First, so that they cannot cause privation and problems for the rest of us; Second, in hopes that through their punishment, they’ll come to realize the destructive and harmful nature of their actions. So I’m all for punishment, and I’m not in the least bit a sadist.

    **side note: your thanks for the Dawkins link rightfully belong to OMGF, not Chet.

    Donald,

    How can something be “unjustified” if it was inevitable? That’s why nobody talks about “unjustified death,” because you can’t control when your prostate cancer kills you. Or an “unjustified” hurricane, or bolt of lightning.

    You’ve got my vote, for now at least.

    Ebonmuse,

    …an action is free if it is chosen by a rational agent as the result of a process of deliberation in accordance with that agent’s desires and free of coercion by other agents.

    What if the action is chosen by an irrational agent, or a rational agent acting irrationally? Also, define process of deliberation. And since all other causal factors influence our present actions, are not all of our present actions to some degree coerced? I see a potential for contradiction here.

    Given all the causal factors acting on me from the beginning of my life up until now (including some that I set in motion myself, as the results of decisions arising from yet earlier causal chains), I could not have chosen otherwise than I did.

    So if you murder me, in a rational state, after a process of deliberation, free of coercion by other agents (if such is possible), did you have a choice in the matter?

    And in that realization, I believe, lies true moral responsibility…

    How so? Allow my paraphrase here, but it seems to me that you’re saying that past causal factors determine present actions. Where’s the allowance for moral responsibility?

  • prase

    Free will is the idea that, for any such deliberative process or decision leading to an action X, there is a logically possible world in which everything is identical down to the minutest detail except that action Y was taken instead.

    There remains a question of what exactly is the meaning of “logically possible”. Many people would say that a necessary condition of logical possibility is the fulfillment of physical laws and that, for example, people floating free in the air without any specialised equipment are logically impossible. Or you can make some distinction between the laws of logic and physical laws and say that different natural laws are logically possible… In common language people say that an alternative outcome is possible if they cannot see any universal law it contradicts. But clearly there is a huge difference between “can’t be seen by people” and “doesn’t exist”. The size of the set of logical possibilities, in the ordinary usage of such term, is proportional to our ignorance. If we had perfect knowledge we may see that this world, all details included, is the only possible one.

    I just don’t believe that speaking about logical possibility, possible worlds and all this stuff on metaphysical level is fruitful.

  • http://she-who-chatters.blogspot.com D

    OMGF, my apologies for the misattribution! And cl, thank you for the correction. Additionally, you’re correct on my unnecessary correlation of punishment and harm – in hindsight, it appears that I’ve been using a private definition of punishment, which I shall have to correct. I think I’ll be shifting my line of attack from “a sadistic desire for punishment” to “a primitive desire for vengeance” in future discussions. Nevertheless, I believe that my general point (focus should be on “fixing,” not on “punishing”) still holds. Prase, I agree that “logical possibility” can also muddle the discussion – for my part, I go with “permitted by the laws of physics,” which clarifies the disagreement with philosophical libertarianism because whether or not the laws of physics permit free will is precisely what is at issue (libertarians say yea, determinists say nay).

    Ebonmuse, in all honesty, I’m kind of unclear on what exactly compatibilism is. By my understanding, the free will/determinism debate arises from the following inconsistent triad:

    1. People have free will.
    2. Causality governs the Universe (no exceptions).
    3. The idea of free will somehow defies causality (i.e. 1 & 2 are mutually exclusive).

    Rejection of 1 is determinism, rejection of 2 is libertarianism, and rejection of 3 is compatibilism. Here’s where I get confused on compatibilism, though: in order to reject 3, compatibilists typically redefine free will, which does in fact resolve the inconsistency. But there’s a problem with this method, because in redefining free will, the libertarian meaning of 1 is covertly rejected.

    From your post above, Ebonmuse, it looks like we do not disagree on our descriptions of the world: we agree that causality governs the entire Universe, we agree that the libertarian meaning of free will is without referent, we agree that deliberative processes and decisions occur (regardless of whether they could have turned out otherwise), we even agree that moral responsibility still exists and that justice is still possible (I think). The only disagreement we have seems to be on how we’re defining “free will.” But there is no objective meaning of language, our terms only have meaning intersubjectively. So, due to our differing usages, it appears that we now have two phrases, both called “free will,” but distinct in their meanings: free willm is the metaphysical liberty of “could have done otherwise” posited by libertarians, and free willc is the compatibilist revision you propose which rejects the “could have done otherwise” clause. We also need this distinction to clarify our mutual disagreement with libertarianism, which re-writes our inconsistent triad as follows::

    1′. People have free willm.
    2. Causality governs the Universe (no exceptions).
    3′. The idea of free willm somehow defies causality (i.e. 1′ & 2 are mutually exclusive).

    Free willc is not part of the inconsistent triad because it is consistent with determinism but also rejected by libertarians. I agree that free willc is the only kind of free will that has a referent (I am, after all, a determinist), but we must at least take libertarians at their word when they posit free willm so that we may meaningfully disagree with them.

    Can you please clarify where it is that we disagree?

  • http://whyihatejesus.blogspot.com OMGF

    No worries D.

  • http://www.thewarfareismental.info cl

    D,

    Additionally, you’re correct on my unnecessary correlation of punishment and harm – in hindsight, it appears that I’ve been using a private definition of punishment, which I shall have to correct. I think I’ll be shifting my line of attack from “a sadistic desire for punishment” to “a primitive desire for vengeance” in future discussions.

    Well, I honestly can’t tell if this is sarcastic or genuine, and I don’t want to assume, but either way my response would be the same: I don’t argue that actual harm ought befall a meat machine when it malfunctions. My sole concerns are that the meat machine remains pulled from production until it stops malfunctioning. As you noted, this is much more complex with a meat machine than a machine. This all essentially distills to,

    I believe that my general point (focus should be on “fixing,” not on “punishing”) still holds.

    I agree, and all I’m saying is one can be an advocate for punishment (correctional action) without also being a sadist. The last thing I want to see is more people suffering in the world. Hence the importance of removing malfunctioning machines and attempting to fix them.

    However, without saying it’s moral or not, the only type of desire for vengeance or the sadistic infliction of suffering I can say that I can genuinely sympathize with is that of a father who’s only daughter has been savagely raped and murdered, as one example.

  • http://she-who-chatters.blogspot.com D

    cl,
    I was being genuine. I looked up the definition of punishment, and it looks like I was using 2a: suffering, pain, or loss that serves as retribution; whereas it could also mean 2b: a penalty inflicted on an offender through judicial procedure. The former was what I meant to speak against; the latter does not necessarily entail harm. Looks like we agree on a lot of stuff, too; I can also understand and sympathize with such a father’s desire for revenge. In fact, it’s tremendously hard for me to avoid actively endorsing such a thing. The frustrating thing is that no amount of vengeance can bring the daughter back. Nothing can. That kind of permanent loss is, to my mind, one of the worst things a person can do to another person, because it can never be set right.

  • http://www.thewarfareismental.info cl

    D,

    Right on. Thanks for your sincerity, I opine that such can move mountains in debate, and excuse me for my uncertainty regarding your motive – it was rooted in several past experiences I’ve had here where the person engaging me chooses sarcasm and snark over sincerity and logic. I just didn’t want to assume anything about you or your motive, if that makes sense.

    And I agree with the frustrating nature of the crime we reference.

  • MS Quixote

    far truer and more genuine than we could ever obtain from a philosophy that treats free acts as random and inexplicable lightning bolts without cause or motivation.

    Well said, both here and in the OP, EM–a toss of the theist hat in the air for you. And what better definition of free will could we hope for (nice grounding comment prase) than always choosing exactly what we want, as compatibilism implies.

  • http://www.patheos.com/blogs/daylightatheism Ebonmuse

    Thanks, Quixote! I can’t take full credit for that phrase – I think I originally read it in Dennett, either Elbow Room or Freedom Evolves (both of which I highly recommend, by the way), although I don’t have the exact citation to hand or else I would have given it.

  • MS Quixote

    As a side note, I just purchased both books–one in hardback, one in good the other in excellent condition–plus “Consciousness Explained” and only spent pennies over $11, not including shipping. Tough dilemma here: I don’t know whether to be happy I can get premier works so cheap, or sad that demand is so low it drives the price down.

  • Terry

    No matter how absurd and seemingly random the behaviour of man, when viewed from a satelite in time series, patterns emerge that strongly suggest that all events are in lock step with a much larger, global dynamic. 6 degrees of separation is actually less than 2. What you say and do affects the middle east in less than two weeks just through your network of friends. Free will is impossible to test even as far out as the deep space probes. The many worlds hypothesis reflects all possible choices none of which is a proffered or “correct” path. Corroborative lab findings indicate quantum entanglement is real and a valid argument can be made against an entity making a choice contrary to all prior experiences.

  • andrew

    What about people who dont believe our bodies and souls are seperate at all, but are a distinctly univied whole.

    What would you say to a Christian who doesnt believe that our ‘soul’ flies off after death to heaven/hell leaving our bodies behind, but waits(sleeps or dies) for the end of times, when our BODIES will be ressurected?

  • Mathew Wilder

    @ andrew: I would say that idea is just as crazy as the dualist view. It still reeks of vitalism, and it also shows a great ignorance of basic physics. There are more people alive now than at any other time in history. Many of the molecules that make up the living have previously made up the now-dead. So if bodies are re-constituted at the ressurection, who gets the molecules?

  • andrew

    I’m sure the God who created teh universe can create the molecules that are needed to reconstruct our bodies.

    Oddly the idea of a physical ressurection is one that has only recently been lost to most Christians. It was one reason the Church opposed cremation for so long. It was believed that it would somehow hinder the ressurection. While I dont think it will(again, it should be easy for God to create whatever he needs to re-build our bodies) It does show that they believed in a physical ressurection.

  • Wedge

    andrew,

    So, instead of belief in a resurrected consciousness/soul which operates magically without a body, it’s a belief in a magically reconstructed body? It doesn’t strike me as any more plausible.