A few years ago, there was an article in New Scientist by Dan Jones,entitled, Grand delusions: Why we’re determined to be free. It began:
IT IS the year 2500. Physicists have long had a Grand Unified Theory of Everything and neuroscientists now know precisely how the hardware of the brain runs the software of the mind and dictates behaviour. Lately, reports have begun to emerge that computer engineers at the Institute for Advanced Behavioural Prediction have built a quantum supercomputer that draws on these advances to predict the future, including what people will do and when. Trusted sources say that IABP researchers have secretly run thousands of predictions about citizens’ behaviour – and they have never been wrong.
Suddenly, deep philosophical questions are making headlines as commentators sound the death knell for free will. On the face of it, the consequences of proving all our actions are predetermined look bleak. Psychological experiments have shown that undermining people’s sense of free will leads them to behave more dishonestly, more selfishly and more aggressively. But perhaps there is no need to panic. Some philosophers have found that our sense of free will is less threatened by determinism than the commentators suppose – so even faced with incontrovertible evidence that behaviour is predetermined, we still see ourselves as free and responsible for our own actions. Nothing will change.
Who is correct? Will the public buy this reassuring message? Or will the manifest truth of determinism kill off belief in free will, taking down notions of moral culpability and punishment with it? Will nihilism, moral disintegration and anarchy follow?
I composed a response, which I ended up not sending, but sent to Daniel Dennett instead, from whom I received a very nice reply, in which he attached a relevant article he’d recently written, Some Observations on the Psychology of Thinking About Free Will. Here is the draft of my own response to New Scientist:
Dear New Scientist
Your cover story for 16th April 2011, “Free will: the illusion we can’t live without”, perpetuates the illusion that free will is an illusion. Fortunately, by the end of the article, Dan Jones introduces some much-needed nuance, citing the philosopher Eddy Nahmias as making the argument “that instead of rendering conscious thoughts and decision-making irrelevant fictions, neuroscience could actually illuminate the biological basis of free will”.
Yes, it could, and indeed it does. It is a gross misrepresentation of neuroscience to portray it as describing our behaviour “as the result of a chain of cause and effect, in which one physical brain state or pattern of neural activity inexorably leads to the next”. Firstly, there is nothing inexorable about the way that “one physical brain state…leads to the next” – it may hinge, like weather, on something as finely poised as the proverbial butterfly in Peking – a spot of rain, a passing car, a familiar face in a crowd, as well as internal stochastic processes which may well include quantum effects. But secondly, even if we lived in a completely deterministic universe, for which we had a complete description, it would tell us nothing about “free will”.
Dan Jones reports a study by Nahmias himself, in which he apparently posed a futuristic scenario to a group of participants in which “the computer predicts the birth of a boy, Jeremy, and also that he grows up to rob a particular bank at a given time and day when he is 30” and asks the participants whether Jeremy acted “of his own free will”.
Most of the participants, thankfully, answered “yes”. Dan Jones’ article treats this phenomenon as evidence that we are naturally resistant to the concept that we do not have free will. I suggest that we are naturally resistant to the concept of idiotic scenarios. No computer will ever predict that a child will “rob a particular bank at a given time and day when he is 30”, because the reason that we regard brain-bearing creatures as decision-makers – choosers, will-ers – is that that’s exactly what they do. For any computer to predict Jeremy’s decision to rob a given bank on a given day, that computer would not only have to emulate Jeremy’s brain, but also model every single life event, from the molecules he ingests to the ideas he engages with, in order to predict the outcome of any decision faced by Jeremy. In other words, any infallibly predictive model of Jeremy’s life would be as large as Jeremy’s life itself.
Science is about making models of reality, i.e. incomplete representations of the real thing. A complete model of the universe would not only be the size of the universe, it would have to include the model itself, rendering it intrinsically Incomplete. What we do instead, in science, is to try to infer general principles, identify causal factors, compute probabilities – in other words make model that have useful predictive power. But all models must necessarily leave countless degrees of freedom within which events can take, literally, unpredictable turns. That is why despite our sophisticated technology, weather-forecasts remain unreliable beyond a few days, or even hours.
And our brains themselves are predictive model makers; we choose our actions by weighing up simulated outcomes. Freedom rightly, is the name we give this ability, and it is powerfully advantageous, in a Darwinian sense; it’s what makes us able to navigate our environment flexibly and productively.
Neuroscience is no threat to the concept of free will; neuroscience is the science of how we exercise our freedom to choose, and how factors that affect those choices are weighed in the decision-making process. We know now that there is no inner homunculus, pulling the levers of our brains. Nonetheless, the idea persists that there is an inner subjective “consciousness” that has no access to the levers (“has no free will”) but is nonetheless is a necessary postulate to account for our experience of existing. We have shorn the homunculus of all executive power, but retain it as titular head of self-government.
But no such postulate is necessary. “I” am a not helpless passenger borne by a mechanistic decision-maker whose decisions “I” am unable, Cassandra-like to influence but doomed, regardless, to observe. I am that decision-maker, and, as a member of the species homo sapiens, I have evolved not only to take what will bring immediate benefit to myself into account, in terms of sensory reward or evasion of risk, but also long-term benefit, as learned from previous experience of links between our actions and effects. Furthermore, I can also take advantage of the huge modelling capacity our possession of language endows us with to set abstract goals that also factor into my decision-making process – moral principles, for instance, such as “the greatest good for the greatest number”, or even the well-being of members of other species. Most of all, our language capacity allows us to reify the decision-making agents that populate our world, including ourselves – and the word we give to the decision-making agent that drives our own actions is “I”.
As Daniel Dennett repeats like a Greek chorus throughout his book Freedom Evolves: If you make yourself really small, you can externalize virtually everything. Conversely, it is by the very act of accepting responsibility for my decisions that I create what I call my self – a first person subject to which I attribute moral agency. There’s nothing illusory about that. It’s what I am.
EL said:
This problem illustrated by your terminology runs throughout you response. You write as if there is a distinction between “you” and “your brain”, between “you” and “your simulations”, between “you” and “your decisions”. – as if you are an operator sitting at the control panel operating your brain, looking over your simulated data, and independently making decisions after weighing that data via some independent system of evaluating it.
Under materialism, you have no free will capacity in the libertarian sense – no independence from the machine. You are only a caused sensation of free will capacity; caused decisions accompanied by caused sensations of making decisions; a caused sensation of employing reason; etc. You are the the caused sensations; you are the computation of they physical interactions of your brain and body.
If that body computes that you run simulation X, that’s what you are; if it computes you interpret simulation X in a certain manner, that’s what you are; if it computes action Y as a result, that’s what you are; if it computes you have certain sensations or thoughts accompanying those activities, that is what you are. Even if there are random or non-deterministic chaotic influences/outcomes involved, you are still nothing but a computation under materialism. There is no ghost in the machine that can, in a true top-down or outside-in manner, intervene in the process of the computation.
You’re fee to call actions derived from the physical computation “free will”; you are also free to call black “white”, if you wish. It’s still nonsensical.
William,
You should have kept reading. Lizzie is denying that distinction, not reinforcing it:
William:
Compatibilist free will is the only kind that makes any sense. Libertarian free will is incoherent.
keiths said:
Denying one is the computational output of the system is still just the computational output of the system. There’s what materialists say, and then there is what is logically consistent under materialism. They are not often the same. EL often necessarily implies the very thing she denies exists – as do virtually all materialists. Her – and your – arguments rely on the very commodities you insist do not exist and are “incoherent”.
Physical computations output the funniest things!
William,
Slow down and think about this. Lizzie and I are denying the distinction between the self and the physical decision-making system. The self is that system.
We are physical systems, and our decisions are produced by the physical evolution of those systems as they interact with their environments.
It’s a sensible position, it’s consistent with the scientific evidence, and we’re neither explicitly nor implicitly invoking the incoherent concept of libertarian free will.
Libertarian free will is, well, non-referential to say the least. But while it might make some sense that compatibilist free will works, it’s simply not what is meant by “free will” in the minds of most people. To that extent, I’d have to agree with WJM. Yet, of course, we retain the capacity to act with freedom (anything that interacts in a complex manner with the world does), so in the etymologic sense (and while it may not dominate meaning, it tends to affect meaning), of course “free will” exists in the sense that we have “wills” that freely choose from alternatives.
In the end, it’s the words that end up being argued about, then. WJM is rather absolutist, and insists that libertarian free will is the only kind of “free will,” while not having the evidence to say that it exists, nor even a real notion of what it is or could be (indeed, what could it be?), but is correct that “free will” typically does refer to “libertarian free will.” Then others say that “free will” exists, in that there is a will and that it is free, which is an important distinction, but doesn’t really tell us that we have “free will” in the sense usually meant.
Maybe we should just claim that we have “wills that are free,” in that they can act independently.
Glen Davidson
Although I enjoy reading discussions about “free will”, they are as satisfying as discussing how many angels can fit on the head of a pin.
Whether or not “free will” truly exists is something that cannot be tested. There are so many factors affecting any choice we make that it would be impossible to control for all of them. I suspect that, when you get down to the finest details, that “free will” does not exist. But that when you step back and examine all of the possible inputs, reactions and outputs, that what we see is indistinguishable from “free will”. Much in the same way that a pixel on my computer screen is nothing but a dot of light, but when I step back, all of the pixels viewed in context form a picture.
“Free will” is one of the foundations of the Judeo-Christian belief. It, supposedly, was given to us by an omnipotent and omniscient God. I am sure that I am not the only person who sees the logical inconsistency here. If God can see everything, past, present and future, then how can we have free will?
Well, sure, but you could say the same thing of “design.” As long as it’s general enough, and unspecified by reference to anything specifically identifiable as “Mayan design” or some such thing, it could certainly be there and “obvious” within a certain sort of theology and/or ideology.
But we’re not really justified in assuming that either design or libertarian free will operates in life unless we have some sort of entailment, some explanatory detail(s) that only this “design” or “free will” explains. Lacking that, we’re not justified in assuming that it exists, unless it’s a question that we choose to address (I don’t think there’d be anything wrong with a researcher assuming it exists, so long as the researcher is willing to be “proved wrong” by experiment or observation).
Since “libertarian free will” fades into the complexity of life, and lacks any meaningful evidence and even meaningful identifying characteristics, it’s as good as non-existent, to our knowledge. Might exist anyway, but so might the Lucky Charms leprechaun.
Glen Davidson
Whether or not he exists doesn’t take away the fact that they are magically delicious.
If self is the system…and the system is all the component parts that collect sensory input etc…does that mean if I lose component parts of myself that I am no longer myself? If I lose my arms, eyesight, hearing, etc…am I somehow less of a self?
Which parts of the system are self and which parts aren’t? And why?
Glen Davidson said:
No, I’ve argued that it is the only form of free will that can serve as a proper logical foundation for how we actually behave, how we think we behave in any practical sense, and what we expect is going on when we argue wrt ourselves and others.
Asking to explain what “libertarian free will” is, and how it operates, is like asking what a force “is”, and “how it works”. It’s a nonsensical question. You don’t explain forces, you describe them and then reify that description as a causal explanation for other things.
Compatibilist free will cannot serve as a logical foundation for that which libertarian free will can.
So is that the plan from now on, just call whatever you claim exists a “force” (without any good evidence of how it is a force, of course) and claim that thus it needs no explanation?
Aside from the many explanations of forces that exist (Feynman diagram, anyone?), the point of real forces is that they have demonstrable, repeatable effects. And yes, we do say what a force is, in terms of origin (gravity comes from mass, for the Newtonian version), extent, and effect, and typically they are measurable. They are not mere labels that avoid explanation of real phenomena, they become part of the explanation of phenomena. “Free will” explains nothing, indeed, it seems not to correlate very well with how mass thinking occurs in societies.
Glen Davidson
And yet we can observe the force of gravity, perform experiments, come to conclusions about how it operates on large and small scales.
Apart from the enjoyment aspect, I totally agree with this. Why should I worry if someone else wants to pontificate about free will? As usual, theologians want to waffle and indulge in wordplay without first deciding what words mean in the context of the waffle, and I say let them get on with it. It has about as much interest for me as watching paint dry.
With all due respect to philosophy and neuroscience, free will vs determinism is one of the most worthless topics one could explore. It’s not merely unanswerable; it’s entirely the wrong kind of question to ask about behavior.
Just as nothing in biology makes sense except in the light of evolution, nothing about brains and behavior makes sense except in the light of learning.
Just as some aspects of genomes are the result of selection and others the result of drift, some aspects of behavior have been reinforced or discouraged by consequences, and others by drift.
The point is the billiard ball model of causation is simply worthless. We cannot predict the course of evolution because we cannot know what kinds of selection will impinge on populations, and we cannot predict human behavior (or even animal behavior) because we cannot foresee the range of consequences that will impinge on individuals and populations.
What gets modified in brains is the tendency to behave. and by this, I mean all kind of behavior, including perceiving, thinking, feeling, and imagining.
This is all ignored by the traditional determinism vs free will debate, but it is the only model worth discussing and researching.
petrushka,
Amen, brother!
Glen,
But more importantly, it’s incoherent, and would remain so even if physicalism were false.
What most people mean by “free will” is libertarian free will, which cannot exist. But if you look closely, the reason that people reject physicalist, compatibilist free will is that they are approaching it from an implicitly dualist perspective ( a subject on which I’ve been meaning to do an OP).
The typical thinking goes something like this: “If I’m just a physical system, then my decisions are really being made by physics, not by me. That’s not free will.” (Sounds a lot like WJM, no?)
But if physicalism is true, then each of us is a physical system, and to say that our choices are ultimately physical is to say that they come from us.
Jackson Knepp:
It’s a complex topic deserving of its own OP, but my short answer is that what makes the five-year-old Jackson Knepp the same person as the 75-year-old Jackson Knepp is not any particular set of atoms, but rather a particular kind of causal continuity between them.
Dennett, who is a fount of useful characterizations, describes the self as a “narrative center of gravity”.
William,
Libertarian free will is itself incoherent and cannot serve as a logical foundation for anything. It cannot exist, even if physicalism is false.
It is highly unlikely that in the year 2500 hardware running software will be the model of the human brain/mind relationship, and the idea that behaviour is dictated has already been refuted.
petrushka,
The interesting philosophical issue isn’t free will vs. determinism, because a) free will isn’t incompatible with determinism, and b) physics doesn’t seem to be fundamentally deterministic anyway.
Why do you think it’s unanswerable?
I’m unaware of anyone in the free will debates who suggests that learning is unimportant.
The usefulness of the billiard ball model is in thinking about causality, not in making predictions. Laplace’s Demon was avowedly fictonal.
I’m not sure where you got that idea. Whether perceptions, thoughts, feelings and imaginings are deterministic is a question that many in the free will debate have asked.
Ooh, Dennett again; Lizzie quoted Dennett a few days ago, on something else I agree with. Yeah, I agree Dennett is a “fount of useful” — and more than just useful characterizations. The Self as a Center of Narrative Gravity IF there could only be one philosopher in the world, I would be completely happy to have that one be Dennett.
Of course I’m biased; he’s my hero because he’s a non-theist who’s not a typically sexist jingoist asshole like the others nominated as the Four Horsemen of the New Atheists (Dawkins, Harris, Hitchens). But that’s a subject for another thread, or none at all … sorry.
You’re talking about Newtonian mechanics right? Particles and forces and all that stuff? I wouldn’t say it’s completely worthless, and maybe that’s not what you mean.
Yes, you’re denying it. But then you go on to act is if what you have just denied is in fact true. IOW, William’s charge against your stated position is one of self-contradiction and incoherence.
You talk as if the self is not the system.
keiths,
I am naturally interested in exploring this idea further, especially, the “particular kind of causal continuity between them” idea. And what one means by that – but perhaps this is not the thread as you mentioned.
In a any event, it seems to me that if one holds self is the physical system, then taking away/changing parts of the physical system would necessarily change the self in some way. One might consider an old home that is replaced one board at at time until it is completely new. In what sense is this the same home it always was? Is it the design or conception of the original home that retains the identity. But we know that in these situations the make-up of the home and some say the actual identity of the home changes. Which is why folks that remodel antique homes don’t like to “gut them.” Or one might consider a home where half of it was torn down. Is it the same home? The same questions are true for the physical self notion.
Ask away, but it remains a worthless question. Verbal masturbation.
If a phenomenon is unpredictable, it makes no difference if it meets some arbitrary criterion for being determined. The philosophical question has no traction in the real world..
From the standpoint of morality and criminal justice, people are accountable for their actions if they are mentally capable of learning. The law vacillates around the issue of accountability, but in general asks the right questions. Or at least relevant questions. There are times an places where the law behaves irrationally, but it is evolving in a useful direction.
From the standpoint of predictive science, behavior evolves in a manner not unlike evolution. Which is to say, even if you can make a complete physical accounting of the process, you can’t make reliable predictions.
I’ve been watching this debate for 50 years without seeing anything useful said, unless it was from the perspective of learning theory.
From the OP:
This is self-contradictory and anti-science, the essence of which is prediction.
What makes for predictability is that one state inexorably follows from the previous state and that each state is entailed by the prior state.
That one billiard ball is struck by another billiard ball thus changing it’s trajectory, does not make the states of the system any less inexorable.
Why should the physical brain be any different than any other physical system?
Suggestion: drop the usage of “brain states” since it’s use seems to mitigate against the position you appear to be trying to take.
I would say it’s the same thing that makes a species the same from generation to generation. the physical configuration of atoms rather than the specific atoms).
The configuration is maintained (with modification) by purifying selection.
If the configuration is significantly disrupted by injury, disease, or trauma, JK may not remain the same.
Elizabeth:
Gold star for quote mining.
Which was sort of my point!
As so often, Mung, you accuse me of being “self-contradictory” without, it seems, reading the post very carefully.
Try again. Do ctrl F for “predictable” (it’ll pull up “unpredictable” as well).
Mung, there are regular phenomena that cannot be predicted. the toss of a fair die, the spin of a fair roulette wheel, the deal of an honestly shuffled deck of cards.
I don’t expect you to care about or understand this distinction.
Nor the general problem of complexity and chaos.
And again: the charge of contradiction. Even though I have explicitly laid out why these things are not contradictory. Feel free to challenge my argument that they are not, but simply baldly stating that they are, without reference to the argument is sort of annoying.
Precisely.
Which is why no respectable scientist thinks Chaos Theory has any place in science? Really, Mung?
No climatologist thinks that an inexorable billiard-ball model of molecules in the atmosphere will ever (not even in principle, not even with better data and more computational resources) be a good-enough match to real weather.
This isn’t even news. The initial insights into chaos theory are more than a hundred years old.
Yes, science reaches its limits where there is no predictability. Yes, you might be correct to say that “the essence of science is predictability”. (I don’t think many scientists would fundamentally disagree with that, although I was surprised to find that google returns only 5 results for that exact phrase.”)
But for more than 100 years, science has had a more-accurate and more-nuanced view of “predictability” than your flat “state inexorably follows from previous state” view.
Exactly, good question. Why would the physical brain be different from other physical systems which we know are stochastic, chaotic, imperfectly-predictable even in principle, and perhaps subject to quantum effects even when averaged at the larger scale? Why?
Think bigger, Mung, don’t limit yourself to a billiard-ball model of the universe. It’s useful, but only in certain cases.
Forgive me, but I am breaking in a new eye, and I may scan long series of posts without reading everything.
Or I could have been unclear 🙂
Best of luck with the new eye.
Yes. And?
As far as I know, the problem of chaos and complexity applies to the billiard ball model of physics.
The business of predicting the behavior of a system that learns is somewhat different, at least in my mind.
A learning system changes in response to feedback, so in addition to the problem of the complex behaver, you have the complexity of the system supplying the feedback.
A possibly silly example would be the change in human behavior generated by the discovery of some distant astronomical object. The butterfly could be billions of light years away.
Of course I care. That’s why I took a stance against “science is prediction”
But the primary stance here at TSZ continues to be that of Newtonian mechanics, in which each state of a system is entailed. As WJM points out in the “POOF” thread, Elizabeth is asking him to accept as the basis of her argument against ID a model of the world that he does not accept.
That said, in what way is quantum mechanics substantially any different from Newtonian physics when it comes to the matter of entailment?
Lizzie, I thought you might enjoy this:
And:
– Marcello Barbieri, Code Biology: A New Science of Life
Biosemiotics. 🙂
Quantum tunnelling enables the construction of the computer you use to post.
As a minor example.
OK, let me try to make myself a little clearer.
Scientific models are predictive models. They often take the form of equations that tell us the expected value of some quantity, but they can take other forms.
They always have error terms of some sort – they do not predict perfectly.
Sometimes the error terms tend to cancel out, and so, on average, our predictions are good.
However, sometimes they propagate, as, for instance, in weather forecasting. In these models, prediction is impossible. However, what we can do is produce contingent models that predict broad patterns. For instance, our climate models can predict, within a fairly good confidence limits, how many hurricanes will form in a year, and where most of them will start. What they can’t do is predict when and where any one will start.
Same with human brains. We have a pretty good idea, now, and can predict with really very tight confidence intervals, what kind of large-scale patterns of blood flow and electrophysiological oscillations we will observe when, for example, we ask someone to distinguish between two kinds of stimuli.
However, we cannot predict at all what will happen at the very small scale – as with weather, the errors in our estimates propagate.
So yes, science is all about predictive. What scientific models do is predict.
But some phenomena, specifically chaotic phenomena, i.e. phenomena in which positive feedback loops play a major role and pften cancel out homeostatic negative feedback effects, can only be predicted at a much coarser scale. I can predict, broadly, how people will vote next week, given enough data (well, I can if I’m Nate Silver)/ I cannot, and will not ever, be able to predict how Johanna Jones will vote in 2025.
There’s too much we don’t know, and too many natural phenomena are critically depending on factors we cannot measure, because they are too tiny and too numerous, and too unknown.
Who knew that that butterfly we released out of the window rather than squash would cause the next hurricane in New York?
Thanks Mung. I’m not sure what your point is, but those passages sound reasonable.
Maybe these are not good questions… but wouldn’t a corpse have essentially the same physical configuration of atoms at least for a very short time after death? Also – what about two objects that theoretically are identical such as clones or identical twins? And would some one’s self really disappear, and then be revived, in cases where folks are revived (hearts shocked etc) back to life after say cardiac arrest. Or would they continue to exist the whole time?
And, even if a complete replacement of material, left the original design (or I guess I should say form) intact, it starts to seem to me then that the actual identify of self would reside in the form, or as you put it, configuration, rather than the particular material construction? I also am still thinking that the logical extension would also mean that if the physical components are the self then a loss of some parts of that construction necessitates a loss of self in some sense.
The Two Lizzies
Elizabeth,
I like the story it illustrates well some of the difficulties though it doesn’t resolve all of them in my mind. And it seems to illustrate that in fact identity, or self, doesn’t reside in a specific form or configuration of atoms as earlier postulated – unless one includes in that conceived form a particular place in space and time. I don’t know if it addresses my earlier questions about what happens to self when component parts go missing. Or very shortly after death…
Non-predictability doesn’t change the fact that under materialism, all thoughts are caused by happenstance interactions of matter according to law/chance. Not being able to predict a particular choice is entirely irrelevant. I cannot predict the path of a rock tumbling down a hill – that doesn’t mean the rock is making a choice in any meaningful context.
Materialist use semantics – like unpredictability, chaos – as if increasing the complexity and unpredictability of a system can magically turn an event caused by physics into a free will choice. If choice is ultimately caused by physics, it’s not a free will choice except when defined as such by the most absurd of equivocations.
hotshoe_
Oh, go for it. I’ve been agreeing with way too much of what you’ve been posting lately. 😉
William,
So? The idea that thoughts are the result of physical processes is no more problematic than the idea that computation is.
This mental block has been plaguing you for more than three years, William. It’s time to buckle down and think things through.
Thought, like computation, rests on a physical substrate. The evidence supports this thesis overwhelmingly, and there is nothing at all incoherent about it.