Hard Determinism and Punishment

Determinists argue that fault and blame have no place in criminal “justice”. Neuroscientist David Eagleman, for example, made this argument recently in The Atlantic:

The crux of the problem is that it no longer makes sense to ask, “To what extent was it his biology, and to what extent was it him?,” because we now understand that there is no meaningful distinction between a person’s biology and his decision-making. They are inseparable.

While our current style of punishment rests on a bedrock of personal volition and blame, our modern understanding of the brain suggests a different approach. Blameworthiness should be removed from the legal argot. It is a backward-looking concept that demands the impossible task of untangling the hopelessly complex web of genetics and environment that constructs the trajectory of a human life.

Eagleman and other determinists are against punishment but they recognize that incarceration still has a role to play because the public has a right to be safe. Philosopher Saul Smilansky now pounces with a timely paper on determinism and punishment.

It is surely wrong to punish people for something that is not their fault or under their control. (Hard determinists agree with this premise.) But incarceration is a type of punishment so under the hard determinist view, justice
requires that when we incarcerate criminals we must also compensate them to make up for the unjust punishment. Smilansky has a bit of a silly name for punishment with compensation, funishment.

Funishment, however, is very likely to cause a big increase in crime and that is also unjust. Smilansky concludes, therefore, that hard determinists have a problem:

[B]y its nature funishment is a practical reductio of hard determinism: it makes implementing hard determinism impossible to contemplate.

Smilansky has put hard determinists into a corner but I fear that they have a type of escape at least in practice if not in theory–it is the one used by many humanitarians in the past–punish people under the guise or belief that you are really doing them good. The inquisitors surely recognized that it was unjust to torture someone who was controlled by the devil. Nevertheless, if torture is what it takes to get the devil out, then torture is not punishment but treatment…and vice-versa. The record of the psychiatric profession and non-punishing treatment of criminals (and others) is not without blemish.

Eagleman goes to some lengths to distance himself from such conclusions. He says, for example, that

 To help a citizen reintegrate into society, the ethical goal is to change him as little as possible while bringing his behavior into line with society’s needs.

The tension, as I see it, is that if free will is a myth then it’s not clear why we should have an ethical goal of changing people as little as possible.


Surely any ethical system that loosely fits intuitions about morality and applies in a hard deterministic world would yield a goal of changing people as little as possible? Otherwise any kind of coercion that leads to net utility increase would also be ethical.

Well, a hard determinist probably should think that any kind of coercion that leads to net utility increase is ethical. This may be a problem, but I suspect you overestimate how often coercion would lead to net utility increase; coercion is costly in utility terms, in many ways that people who aren't used to thinking in terms of utility aren't used to considering. Mill and his ilk think coercion is justified on grounds of utility in some cases to prevent serious wrongdoing, but that more widespread coercion is to be avoided, also on grounds of utility; they may be wrong, of course, but it is far from obvious that they are.

Daniel Dennett is worth reading on questions of freewill and determinism.

Our actions are predetermined by our biology and environment, but people are complex enough to take on responsibility for their actions.

Looking at the correctional institutions in the USA, I think we could probably come up with some improvements.

"Our actions are predetermined by our biology and environment, but people are complex enough to take on responsibility for their actions."

I don't really understand how 'complexity' makes people responsible for their actions. If their actions are caused by the biochemistry of their neurons what difference does it make if that system is intricate and complicated?

People ARE the biochemistry of their neurons so there's nothing wrong with holding them responsible for what their neurons do. Your confusion stems from assuming that people are something else than their bodies, which leads you to ask yourself "why would you punish the PERSON, if it's the BRAIN that did it?". Well, if the person is the same as the brain that did it, you're punishing the right entity.

My bigger concern is that if fault and blame have no place in criminal justice, if using a "backwards-looking concept" is bad, then there's no moral distinction between punishing someone after the fact for something that is his fault, and punishing someone before the fact for something that he is likely to do, according to his genetics and our behavioral models.

The determinists seem to be moving us towards a Gattaca or Minority Report type future.

Are they really okay with punishing and incarcerating someone from a poor background, who was abused, and who has poor genetics, before he commits any crime? And punishing him more heavily than someone from a good family who has committed a crime but the models say is unlikely to commit another? Of course the criminal justice system takes some of this into account already, but that bias tends to be deplored, and we at least wait until the poor person has actually done something we consider blameworthy.

Under this deterministic logic, there is absolutely no reason to distinguish between having committed a crime and having a 20% chance of recidivism, and having not yet committed the crime but having a 20% chance of doing so.


Please start blogging. You are consistently one of the best reasoned and most interesting commenters (commentators?).





A determinist can still create incentives. If the threat of punishment reduces crime, I don't see the problem. This does require a consequentialist ethics but I suspect a lot of determinists are consequentialists.

If the threat of punishment reduces crime, I don’t see the problem.

So you would be okay with throwing someone from a poor, abusive family in prison if the models say that's he's almost surely going to grow up to be a criminal, while letting off the rich, educated, connected scion of a good family because the models say he'll grow out of it?

I know the criminal justice system does some of this now (though I thought people thought it unfair), but to now punish the poor person before they do anything?

I don't think punishing someone before they commit a crime is deterrence. You can do things that they might consider a punishment ("treatment", loss of certain freedoms, etc.) The only permissible punishments are those that deter future crime . I suspect a utilitarian would require a cost benefit analysis where the more certain the future crime the more restrictions the state can impose on the future criminal's liberty. I'm personally in favor of inalienable rights but I'm not a utilitarian. My feeling is that to the extent science can predict future crime the more things move towards Gattaca. This isn't the fault of the determinists. It's just the nature of technology and people. If we know for sure someone is going to commit a crime we will take action. The philosophical debate is going to be about what action to take.

edit: I should add that for punishments to be permissible they must be necessary to deter future crime. Obviously you can't euthanize future jaywalkers.

If all you are is a couple of pounds of grey matter (and IMO that is all people are) then how that brain functions, how it was treated in development, etc. are important factors when it comes to punishment. The problem is that we have some rough ideas what all of this means, but we are far from anything more than that.

The problem is that we have some rough ideas what all of this means, but we are far from anything more than that.

So if that's the problem, simply one of technical feasibility, then you're okay saying that once the science is better we'll lock up these people with bad genes, bad brains, abuse during their development and so on before they commit a crime?

You say "important factors when it comes to punishment," but the determinists quoted all want to get away from "backwards-looking" ideas like guilt and blame. Without guilt and blame, there is no difference between having committed something and being 20% likely to do it again, and being 20% likely to do the same crime even though you haven't done it yet.

So are you willing to say that if the science were better, you'd be perfectly fine with punishment before the crime?

Well, yes. If we had perfect knowledge that you were going to kill somebody, then it totally would make sense to stop you from doing it. Isn't that obvious?

It would make perfect sense to stop you.

Whether it would make sense to punish you is something else. If the intention of punishment is to deter a behavior, it is absolutely useless to punish the behavior before the behavior happens.

I knew an experimental psychologist who tested this with fish. When he punished them with an electric shock after they did a particular behavior, they quickly stopped doing it. But when he shocked them before they did the behavior, they didn't learn at all to stop doing it.

Or maybe the intention is vengeance for their crime? But they did not actually do it.

Preventive punishment does not work. But preventive execution works. If we kill everybody that we find out is capable of murder, then the murder rate will go way down. ;)

No, actually, I would assume that if you had knowledge enough to determine such things you could also likely treat the underlying problem. Perhaps that is a bad assumption. However, a broken brain is something to be fixed, just like a broken heart or liver. It is just an organ, and learning how to fix that organ is the goal I would say. That obviously involves some normative claims, but no more normative than the claims associated with justice, guilt, innocence, etc. we see in the traditional system of courts and the like.

For you perhaps free will is a concept that makes a lot of sense; for me it doesn't. Up until now it was a convenient and useful kind of abstraction or rule of thumb; but as we understand the brain better the more we realize that free will is not as important as a whole host of other factors regarding the way our brains work. At best our exercise of free will works with in a significantly deterministic context - a context that involves epigenetics, brain damage, early childhood development, etc.

I can give you this anecdote (and no, I'm not going to throw Phineas Gage at you); a few years ago a man became obsessed with child pornography (so much so that his wife left him for fear for her daughter), at the same time he started having very serious headaches. Turns out that he had a tumor; once the tumor was removed, his previous sexual appetites re-asserted themselves (he was no longer interested in child pornography in other words). Some time later his deviant sexual interests returned and not long after that it was determined that the tumor had returned; once it was removed again he lost said abnormal sexual desires. We have lots and lots of examples like this where brain damage, or tumors, or whatever (where nature is essentially providing us empirical evidence of the nature of brain function) effect personality, work ethic, criminality, ethical behavior, and on down the line. Similarly for adults who were sexually molested as children there are differences in brain chemistry apparently; as best as I can tell what happens to children in situations like this is their brains are soaked in the chemistry associated with fight or flight, and what happens is that tends to lead to sub-optimal outcomes in many of these survivors because it short circuits reasoning regarding long and short term decision making (or at least this is how I understand the literature on the subject).

In other words, I'm really not that interested in punishment as you would probably define the term.

BTW, I realize that a lot of what I have said depends on how one views the brain; again, for me it is an organ (a very complex one admittedly) that weighs a couple of pounds.

how about instead of punishing before the fact, identify the genetic and environmental triggers and applying corrective action.

"then there’s no moral distinction between punishing someone after the fact for something that is his fault, and punishing someone before the fact for something that he is likely to do, according to his genetics and our behavioral models"

The notion of 'punishment' implies some kind of vengeance. I think firstly that when one eliminates fault and blame we don't 'punish' anyone. What we might do is to keep society reasonably safe and try to socialize people not to commit what we define to be 'crimes'. One aspect of such a system would be to incarcerate people. However the committing of a crime would make it much more likely that such a person was dangerous. The lack thereof would show the opposite. So we would have a strong bias towards locking people up who already have committed a crime.

There may be times where people who have not yet committed crimes would be locked up in the same way as a person who is danger to themselves (but has not yet done anything) is also 'locked up'. They are dangerous. The rest of us need to protect ourselves. What is wrong with that?

(BTW: We already do that with sex offenders when we civilly commit sex offenders after the end of very long prison sentences.)

You are describing the Stepford society.
I think minimal intervention is better because we are not certain that our interventions are correct. As you note, the history of social science is littered with good intentions.
And society's norms and needs change. In a healthy society they should change. Often that change comes from the actions of a fringe group. Cohabitation, for example, is far more accepted today than a generation ago. That would not have happened if all society's actors had been brought to the same norm 40 years ago.
There are practical reasons to consider, too.
The people intervening are imperfect and corruptible. The more we intervene, the more opportunity we have to corrupt others.
And minimal intervention costs less.

We still segregate the mentally ill, even though we attribute the danger they present to mental illness, and no-one thinks it's unjust. Unfortunate yes, but not unjust. Why is this any different?

Actually, involuntary commitment of the mentally ill is much rarer than it used to be, precisely because it was thought unjust.

In this country, the mentally ill who are segregated and committed are those who either commit themselves, or those who have committed crimes but who get segregation instead of prison, often because they prefer it.

We don't commit the mentally ill against their will before they commit crimes. We used to, but contra your claim, it was thought to be unjust.

Even if we accept your point (and I'm happy to defer to you on that) it remains the fact that we are instinctively fine confining people that we think it inappropriate to punish. There are intuitive rationale's for confinement that do not involve punishment.

The criminals are not morally at fault for their behavior, they can't change or be changed, yet we need to protect society from them. A long winded way of saying "lock 'em up and throw away the key." A conclusion already reached in some jurisdictions, where laws have been passed to keep child molesters in custody in psychiatric hospitals after they have served their prison sentences. How long before this applies to all serious crimes?

An insightful and useful theory of free will was discovered by Drew McDermott in his AI research. To make a long story short it has to do with an entity having an internal model of the world to make predictions and choose actions (and in particular how it models itself in the world model). The bottom line for this discussion is, however, that any entity that has free will in McDermott's sense, i.e. its architecture has the features his theory describes, will modify its actions in the presence of credible threats of punishment.
This seems to me to be the most reasonable criterion for punishing those with free will, and not those without: because it works in the former case and not the latter. And it gives us an objective and useful definition of free will in the face of this foolish neurophysiological nihilism.

So, your handle is actually your name, not a pseudonym! Thanks for posting, very insightful. Looking forward to reading more from you.

If all our decision-making is based on biology and there's no free will, then how does Eagleman expect to persuade anyone of anything?

He believes he is the only one with free will!

"It is surely wrong to punish people for something that is not their fault or under their control. (Hard determinists agree with this premise.) "

I don't see how any hard determinist could agree with the premise. It explicitly requires a counterfactual in which that person is not punished, which is then compared to the punishment and is found to be "wrong".

But they're hard determinists. The alternative is simply impossible. The punishment is just as unavoidable as the crime. Just like we cannot judge the criminal for the crime, we cannot judge the punishers for the punishment. To do otherwise would be to have double standards.

Bingo. If everything's deterministic, then weighing the better moral choice for criminal justice systems is ridiculous.

Also, we discipline toddlers in order to bring their behavior under their control for example.

The justice system is to keep them from getting strung up at the nearest tree. The compensation is not getting punished again

Hume was the only one who got this right--punishment makes sense precisely because we live in a world that is deterministic, at least to some extent.

If we had a will the way free will advocates claim, which was not at all influenced by external stimuli, then there would be no point to punishment because it wouldn't have any impact. The fact that punishment provides an incentive against doing a particular activity is both evidence for determinism and an argument in favor of punishing behavior we don't want people to engage in!

"If we had a will the way free will advocates claim, which was not at all influenced by external stimuli..."
Name me one credible philosopher who has ever held that position on free will.

Even free will Libertarianism does not imply that there is no influence between the will and the external world. It does not claim that people with will consistently walk into trees because the fact there is a tree in front on them does not influence their decision. Instead, Libertarianism claims that there is still a free decision being made, but that it is being made with external stimuli being taken into account. Surely, someone can freely choose to walk into a tree and deny being influenced by external stimuli in their use of their free will or someone can accept the influence of external stimuli, restrict their choices based on that, and decide to avoid the tree.

You guys are all tying yourselves into knots to avoid utilitarian thinking. But this is a good example of why you can't avoid Utilitarian Calculus if you want to run a real society: The gain to everyone else from keeping prisoners from committing crimes is huge compared to the loss in utility from a criminal's loss of freedom, so it makes sense to keep criminals from committing crimes via state coercion.

The main policy implications of determinism though, is that we should try to find ways of crime prevention that hurt potential criminals as little as possible per unit of crime prevention, while moralists tend to push in the other direction. In practice, this means we need to explore stuff like GPS tracking bracelets as replacements for Prison terms when possible.

But hold on, if we're doing a utilitarian calculus, shouldn't we be taking into consideration the desires of people to see at least some criminals suffering?

You make it seem straight-forward but it really isn't. Let's run with the utilitarian calculus for a minute: laws are only justifiable if they increase utility. Not all cases are as clear as putting a mass-murderer in prison.

There are places where a bit pot will get you locked up for quite a long time. Or what about vigilantism? The problem isn't the corner cases, it's the ones right on the line that are the problem, they are the most important questions, and where the utilitarian approach can't offer solutions.

"You make it seem straight-forward but it really isn’t. Let’s run with the utilitarian calculus for a minute: laws are only justifiable if they increase utility. Not all cases are as clear as putting a mass-murderer in prison."

Sure, Utilitarian calculus is hard sometimes. But that doesn't change the fact that the right answer is the one that utilitarian calculus justifies. Some decisions are hard and complicated, but they still have to be made, and you can't really get through them without a utilitarian framework.

"There are places where a bit pot will get you locked up for quite a long time. Or what about vigilantism?"

All things told, those laws probably are a bad idea from a utilitarian standpoint, and it's not hard to come up with standard utilitarian justifications as to why. Vigilantism is actually pretty easy to handle, since societies without strong rules of law have all sorts of problems relating to property rights.

Some decisions are hard and complicated, but they still have to be made, and you can’t really get through them without a utilitarian framework.

Yes, but you can define your utilitarian framework in such a way that you're essentially begging the question.

That's like saying that you can't get through decisions without logic, but ignoring the fact that it's the premises that matter-- premises that at some point are adopted by fiat with justification to things like "obviousness." People arguing using the concept of the veil of ignorance or utility without considering whether fetuses or dolphins or great apes also get values of utility or a possibility for what we could become through the veil-- and many of the arguments that exclude one of these groups (whom many people would like to be able to kill without consequence) are arguments that could end up applying to the mentally disabled.

The post can also be read as showing the limitations of a utilitarian approach. In addition to the reductio ad absurdum presented by 'funishment', I can add another: the state should promote the idea of prisons as bestial and intolerable environments while actually maintaining a pleasant prison environment. Thus deterrence and prisoner welfare can both be achieved without much disutility. An actual, unpleasant system can be maintained for re-offenders (since they know the scam). Or maybe we'll just tell them such a system exists. Further, actual discussion of prison conditions has to be banned as a condition of release from prison.

" I can add another: the state should promote the idea of prisons as bestial and intolerable environments while actually maintaining a pleasant prison environment. Thus deterrence and prisoner welfare can both be achieved without much disutility. An actual, unpleasant system can be maintained for re-offenders (since they know the scam). Or maybe we’ll just tell them such a system exists. Further, actual discussion of prison conditions has to be banned as a condition of release from prison."

If such a system were actually feasible, it'd be a great idea! We do this thing all the time, for example, public health campaigns usually exaggerate the risk of pregnancy and of contracting HIV by at least an order of magnitude. The downside is that it makes people trust public health agencies less. But if we could avoid it, why not?

So the utilitarian supposes more imaginative premises, the avoid the need to actually calculate the consequences of the utilitarian calculus. He is by necessity a deontologist at heart.

If reducing unwanted births is a utilitarian good, people who may be anyone, why not reduce criminals by death penalty, removing them from the utilitarian calculus? This will be as effective at deterring those humans with their bestial concerns and mortal cares, without the cost of maintaining a prison environment at all.

It is surely wrong to punish people for something that is not their fault or under their control.

But how can it be wrong to punish someone, if it's not our fault or under our control?
If criminals have no free will, then people who punish criminals have no free will (be we talking about those people directly involved, as in the prison guards, or only ordering it, as judges, or only voting for it, as in citizens). A prison guard is no more at fault for locking up a criminal, or a politician for voting for tougher sentencing laws, than the criminal was in breaking the law. Saying that anything society does to punish criminals is morally wrong is about as pointless as saying that a cold virus is morally wrong - a cold virus can't stop infecting people, and if the determinists are right, society is incapable of not punishing criminals just as much as criminals are incapable of not committing crimes.

A determinist can only get past this problem if they produce some reason to think that some criminals have no free will, while the rest of people do. This seems implausible and I've never seen a determinist even try.

This is wrong. Determinism does not mean you cannot influence someone's choice, it just means that given the person and the environment, they will make the same choice every time. By changing the environment you can change the choice, obviously.

Well, yes by changing someone's environment I can change their choice. But you haven't thought determinism through correctly. Determinism also means that I have no choice about how I change someone's environment. I could perhaps change someone's environment by voting for tougher sentencing laws, or voting for gentler sentencing laws, but if the determinists are right, then I can't chose how I vote.

Tracy W, I think it's you who haven't thought it through correctly. "You" (i.e., the organism that you are, through your actions) can of course influence others substantially. But to a determinist, when you say something "I can't choose how I vote" -- well, we'll ask, what is that first "I" there, as distinct from the second? Disembodied spirit? Influence from another dimension? It seems upon examination to be mystical nonsense smuggled in through the limitations of natural language.

Good question. Why is the hard determinist importing mystical nonsense about respect for persons?

And to a consistent determinist, "justice" is also mystical nonsense. Things cannot be other than they are, for the universe and everything in it is as it is, and causality rules all. To call something "unjust" is meaningless; the unjust action is merely as the unbroken causal chain of the universe demanded, and the declaration that it is unjust is equally merely as the unbroken causal chain of the universe demanded. All value judgments are equally valid as all other value judgments, for no choice exists, merely causal chains. There is no moral difference between letting prisoners go free and torturing them to death; both behaviors were set in motion by the laws of physics. Sometimes the laws of physics will cause a society to do one, or the other, or something in-between, but a choice never intervenes; everything always is exactly as it must be. Claims otherwise do not illuminate truth; the claims are merely compelled by the same physics as the actions themselves.

And to a consistent determinist, “justice” is also mystical nonsense.

But there's no point arguing about it, because the people who believe in it are required to believe in it. On the other hand, if you are required to argue about it then there's no point in trying to stop yourself. Unless you are required to be undecided. Perhaps you will be required to passionately debate with yourself about what to do.

All value judgments are equally valid as all other value judgments, for no choice exists, merely causal chains.

Yes, and some people have no choice but to make value judgements and try to enforce them on other people. A determinist might have no choice but to do that also.

I find determinism kind of boring. It can explain anything after the fact but it doesn't do me any good.

So the rule I use is, when I want to predict what somebody else will do, I pretend that it's a deterministic world and they have to follow their programming. And when I want to negotiate with people I pretend that they have souls and minds and free will. It works for me.

The existence, or extent, of free will vs. determinism will probably never be answered definitively.

If criminals have free will, it's their fault; they're evil and they deserve to be caged/ punished. If they don't, it's not their fault, but it's still our problem. We cage, or destroy, wild animals that pose a threat to human existence. We must do the same for "wild" people.

Either way: the cage, the key, the separation.

"The existence, or extent, of free will vs. determinism will probably never be answered definitively."

What do you mean? Unless you have weird mystical beliefs about quantum mechanics, determinism wins pretty conclusively.

"Unless you have weird mystical beliefs about quantum mechanics, determinism wins pretty conclusively."

Assuming it has, then we are way overdue for scrapping a ton of government programs premised on the opposite belief.

"Either way: the cage, the key, the separation."

Yes, but we could have Scandinavian style cages.

Hayek addresses this very question in the Constitution of Liberty ch. 5 "Responsibility and Freedom". We punish people to change their future behavior, not trying to assess what they could help and what they couldn't.

Exactly. For an economics blog, the whole point of incentives is pretty much ignored...

We punish people to change their future behavior, not trying to assess what they could help and what they couldn’t.

And we assume that the punishment *will* change their future behavior.

We don't test whether the punishment actually does change their behavior. If we tested out what works, and do more of what works and less of what doesn't work, then we'd do things very differently.

That tells me that perhaps the purpose of the punishment is not actually to change behavior but something else. What else could it be? Tracy points out that people want other people to be punished. Maybe that's it.

Or maybe it's that the people who do the punishing themselves can't help it.

Just because somebody is who he is, through no fault of his own, doesn't mean he can't be irredeemable scum.

In such a case, I'm not sure why you'd have a goal of changing them as little as possible. I'd think you'd want to keep that which is worthwhile and change or discard the scummy parts. In some cases you might be able to save little more than a kidney.

"To help a citizen reintegrate into society, the ethical goal is to change him as little as possible while bringing his behavior into line with society’s needs."

I think it's to change their brain biology to bring them in line with ONE need- to change other people as little as possible, i.e. no killing them, robbing them, oppressing them, etc.

Strange. Why does the determinist care about preventing humans from interfering with other humans? Hell, why shouldn't a free-will believer start interfering with these human automatons?

> It is surely wrong to punish people for something that is not their fault or under their control.

But perhaps we are not CAPABLE of refraining from punishing them - after all, are not we (the citizens, the judiciary, the jury) just as much automata as the felons?

If so, then the discussion is pointless - we will do what we do.

If not, then the discussion is racist / classist: ** WE ** educated intellectual types have free will, but the proles don't.

Or, if we BOTH have free will, then punish away.

This is one of those topics where I scratch my head and wonder how smart people can be so effing stupid. The biology of a criminal's brain made him do it? So what? It doesn't matter because incentives are inputs that change the calculations and behaviors that are the outputs of that putatively deterministic neural machine, and experiences change the very wiring of that machine (all experiences and all learning changes neural connections -- that's how it works).

So the prospect of punishment is intended to provide an incentive (an additional input to the neural system) that will generate different results. And, if that fails, the experience of actual punishment is needed A) to make the incentive credible and effective in the first place, and B) to rewire that deterministic brain in to make it less likely to generate criminal behavior (Yes, so far we're pretty lousy at B, but that means we need different and better forms of punishment -- not because punishment itself makes no sense).

Furthermore, even if 'right' and 'wrong' and 'blame' and 'punishment' are fictions of a sort, they are fictions that come very naturally to the human brain (which has evolved in a highly social context). People are more effective in solving problems when they are posed in terms of social obligations rather than abstract rules (see Tooby & Cosmides and the Wason card problem). And people take a natural delight in punishing cheaters:

"The true pleasure is in seeing rule-breakers get their comeuppance. This is what psychologists call 'altruistic punishment' – 'altruistic' because the punisher gets no direct benefit, and may even risk a confrontation, in order to enforce a social norm. It's not rational, in strict economic terms, but it recurs repeatedly in experiments..."


So, human brains being built as they are, there's really no chance of talking people out of thinking in terms of 'blame' and 'punishment' -- and there's really no good reason to try.

I agree.

Plus, the destruction of folk theories of "free will" often come with the realization that people are more rather than less susceptible to the environment. Would you torture someone with an electric shock because an authority figure tells you to? Signs point to yes.

There are a lot of practical consequences to admitting that the biology of a criminal’s brain made him do it.

There may be cases where only a short incarceration is needed to keep society safe and to provide a good enough deterrent. For example we may find that some people are so shocked by being jailed for 30 days that they will never commit another crime and we may be able to use neural imaging to be reasonably certain of it.

Second, there is a lot of evidence that the generally longer prison sentences that are now imposed do not provide any more deterrence or keep us any more safe than the lesser sentences of 50 years ago. So the length of sentences can be cut now that people don't need to be classically punished.

Third, many non custodial sentences can be imposed instead of incarceration. We can use GPS devices, probation, etc.

And American prisons can be made a lot more pleasant as opposed to overcrowded violent places, with little access to work, leisure or high quality healthcare. And we could do this without rising crime rates.

And one question for Slocum. Are you saying that we pretend to believe something that we know from science not be true because we think we can't handle the truth? Surely not? We are quite capable creatures. We socialize little boys to sit still in a classroom when aged 5 - quite a feat, we socialized ourselves to adopt agriculture despite the fact that we evolved to be hunter gatherers, we don't torture people anymore despite that it comes quite 'naturally' and we give women equal rights despite the fact that we are evolved to 'grab one from the next village' and take her as a mate.

We make kids sit down in class, and we grow our food, and civil society, exactly because we do not believe we are walking machines.

You seem to believe that there is some inkling of spirit and worth in the criminal whose biology made him do it. The hard determinist says he does not.

I don't see how both of the following can be true: (1) crime is out of the criminal's power to control & (2) Funishment would increase crime. If the criminal isn't the one responding to the incentive in (2), then who is? What causes the crime increase if not the criminal's volition?

If one could show that criminals didn't respond to incentives, then I might be pursuaded that they don't have volition over their crimes.

Responding to incentives does not mean you have free will, any more than a snooker ball responding to a hit from a cue does.

It's all about the incentives.

I'm confused. Even hard determinism should acknowledge that a person's biology would respond to the stimuli in its environment, responding to incentives and threats. Per the hard determinist, these responses wouldn't be something a person can explicitly choose, but that doesn't mean we can't change their choices by changing their environment. If we change the environment such that certain actions (crimes) have certain negative consequences (incarceration), we get less crime. In fact, per the hard determinist, our biology drives us to create this environment.

I just fail to see the real problem between determinism and having a justice system with punishment.

Uh oh, another biologist jumping using lab results to jump into the messy, non-linear world of morality and social policy again.

An Introduction to Philosophy student can easily find absurd implications of hard determinism.

The fact (and it is a fact) that free will does not exist has exactly no implications for ethics, because we all feel and act as if we had free will anyway. If punishment is just in a hypothetical alternate universe that has free will, then it is just in our universe too.

Alternate punchline: given determinism, it seems awfully silly to argue over whether we should punish people. Whether we punish people has already been decided for us, as have our opinions on the matter.

Determinism doesn't mean outcomes are set in stone without reference to initial conditions. It just means there's an unbroken causal chain to those conditions. As such, our opinions can be deterministically determined and affected by arguments.

Eh, isn't there an unbroken causal chain to the arguments affecting the opinions and the decision to make them, then? Isn't it an unbroken causal chain all the way down, at least in "hard determinism?"

There are softer forms of determinism, but Alex specifically addresses the hard form here.

It is surely wrong to punish people for something that is not their fault or under their control.

If determinism is true, then the fact we punish is not under our control either.

Bah, you beat me to it!

What evil really is:

Jeremiah 24
King James Version
1The LORD shewed me, and, behold, two baskets of figs were set before the temple of the LORD, after that Nebuchadrezzar king of Babylon had carried away captive Jeconiah the son of Jehoiakim king of Judah, and the princes of Judah, with the carpenters and smiths, from Jerusalem, and had brought them to Babylon. 2One basket had very good figs, even like the figs that are first ripe: and the other basket had very naughty figs, which could not be eaten, they were so bad. 3Then said the LORD unto me, What seest thou, Jeremiah? And I said, Figs; the good figs, very good; and the evil, very evil, that cannot be eaten, they are so evil.

Is being flawed, rottenness like the figs. It does not matter if it is due to environment free will or genetics.

Indeed. This also fits with Kantian morality. A principle of action is immoral if it is self-contradictory (under the premise of universal rational agents).

Allright. The criminal has no choice, so she commits a crime. But neither has the judge, so she condemns her.

The logical implication of the determinist point of view is to institute a Darwin solution for all anti-social behaviors, i.e., first segregate these people from society, and then remove them from the gene pool. That would surely be the Final (and only real) Solution to the problem of deviancy. At least until we can detect the predictive DNA indicators of future social deviance, and deal with it prospectively in utero -- either by surgical gene replacement or simple abortion. Right?

I thought economists had an economic theory for everything from Roman Christianity to lemonade stands. Why not use the usual suspects to solve the problem of crime and punishment. Marginal utility, opportunity costs, the free market, entrepreneurship, rational preference, profit maximization.

This determinist argument seems internally self-contradictory. If criminals cannot be said to be responsible for their actions because they are determined by some combination of genetics and the environment, then politicians/voters/prosecutors cannot be said to be responsible for their actions, because they are *also* determined by some combination of genetics and the environment. If it is useless to talk about crimes in a moral framework, how could it conceivably be useful to talk about criminal policy within a moral framework?

If you are using determinism to discuss incarceration policies, that presupposes that your policy decision is not predetermined. But if your policy decision is not predetermined, how can you say that all criminal behavior is predetermined? What makes the decision to commit a crime so fundamentally different from the decision to punish a criminal?

On the other hand, if your policy decision is predetermined, it's useless talking about what's moral, 'cause whatever you were going to do, you were going to do anyway. Then again, there's no use complaining, because you were predetermined to talk about the issue in the first place.

You seem to be playing the reductio ad absurdum game. We already hold people less accountable for crimes we deem less volitional (although this is changing what with the end of twinkie defenses everywhere over the last few decades). What are the implications of learning that human action is even less volitional than we now think it is? It seems to me that our present approach is to keep raising the bar so that we hold the vast majority of people to the same standard. The law is useful and we'll keep it thank you. That said, if we really know that people are these clockwork orange automata, how long before our criminal justice system changes to accommodate that philosophical position. Right now, hard determinism is just an out there game that clever people play but what happens when science confirms it empirically? Things will change.

Ah yes, the god of Science which says the World is all there is, and ever was, and ought to be, will prove its own premise. Have no doubt.

The moral of the story is that you should be a utilitarian and not care about "rights."

Then the position that Slocum explained above makes perfect sense. Punishment hurts some people but no one cares because the distribution of utility is better with punishment institutionalized.

But the moral of utilitarianism is total nonsense...

I consider myself a hard determainist and I think there isn't much of a problem once you move past looking at a single individual. Once you have to weigh the benefits to multiple individuals, perpetrators and victims, it's clear the guiding light of justice should be the prevention of future crimes. You can't change the past, but you can prevent or mitigate future occurances. Incarceration still makes a lot of sense if you think that a single act of crime causes someone to be more likely to perpetrate other crimes (which is most likely true). However, it doesn't make sense if the incarceration itself actually increases the chance that that person will commit a future crime (which might very well be true in some instances). Direct punishment also has a perfectly valid justification as a deterrent, and threats of punishment must be carried through if the deterrent is to be credible. Even then you have to weigh the effectiveness of the deterrent against its cost.

You can't say "hard determinism says this" or "hard determinism says that". Hard determinism, by itself, does not say anything about incarceration. Hard determinism PLUS a whole bunch of other premises does.

Hard determinism is simply one premise, and by itself does not lead to any conclusions about incarceration.

Similarly, hard determinism, by itself, does not say that people should not be punished. I am a hard determinist and I certainly believe that people should be punished even if it is not their "fault".

How could hard determinism support another premise? What can be added to the premise that the world and all its people are purely mechanical?

That's what makes it *hard* determinism.

It must be nice living inside your head. This is an interesting thread, but seriously. The real answer is: it does not matter whether or not there is or is not free will.

I consulted Denis Diderot and he says that a society's criminal justice system is every bit as determined as criminal behavior.

Thank you.

I'm late to the conversation, but it seems that Norway's criminal justice system comes pretty close to not being harsh punishment, while fulfilling the requirements for incarceration. of course, it's not perfect, and wouldn't completely satisfy the hard determinists (because, well, you can't just leave when you want), but still, Norway has remarkably low crime and recidivism rates.

"Funishment, however, is very likely to cause a big increase in crime and that is also unjust. Smilansky concludes, therefore, that hard determinists have a problem:"
If people can make changes to their actions based on what the consequences are, then there is no such thing as determinism. Determinism says that people will act the way they act because of biology, not because of society norms and practices. FAIL.

If people can make changes to their actions based on what the consequences are, then there is no such thing as determinism.

It depends on how you think about it.

If people make changes to their actions based on a "soul" which is somehow outside of time and space, then clearly their "soul" has free will. No question about it, unless somehow the soul is like an angel who does not have free will after all.

If people make changes based on brain mechanism, does that involve free will?

You could write a computer program that makes choices based on a model of its environment. As it collects data about the environment it improves its model. The model says to perform some action which it computes will have a result it is programmed to maximize. Does the program have free will?

Whether or not the program has free will, if you don't want it to take that action you can find ways to give it results it is programmed to minimize. Do that adequately and it will stop performing that action.

If it is a sophisticated program you might persuade it that increasing your "good will" is a path toward maximizing its goals, and avoiding the behavior you punish is a step toward getting your good will.

If it is very sophisticated you might persuade it that its behavior affects others, and that its effect on the social consensus is important, maybe more important the the benefits it gets from the action you dislike.

If it is extremely sophisticated it might find ways to punish you for punishing it, to influence your behavior into letting it do what it wants.

And even beyond that, it might find ways to persuade you to seek its good will....

Does the computer program have free will?

We pre-fine people for their risk of not repaying loans already.
We pre-fine people for their risk of crashing cars and causing other damage.
All-insurance no-punishment system for dealing with crime might work along similar lines.

If you're deemed high risk, you must purchase crime insurance.
At certain risk levels the only way anybody would sell you insurance
is if you were under constant monitoring and other measures to make
sure you don't do anything criminal.
Your rates would get far higher if you committed any crimes and were
expected to commit more, even if you weren't technically punished for anything.

It would probably be saner and fairer than what Americans currently have,
but then every single country has a system saner and fairer than Americans,
so that's not telling much.

More seriously, this would totally work with drunk driving and a few other narrow
problems we don't really have good ways of dealing with now. Just force insurers
to pay big money for anyone caught DUI and all high risk groups will install
monitoring equipment in their cars to save insurance money, resulting in a lot
less drunk driving.

Comments for this post are closed