Arguments for Determinism

| | Comments (17)

This is the the thirty-eighth post in my Theories of Knowledge and Reality series. Follow the link for more on the series and for links to other entries as they appear. In the last post, I defined 'determinism' and distinguished it from a view that is sometimes called fatalism. In this post I'll look at some positive arguments for thinking determinism is true and some positive arguments for thinking we are free.

Keep in mind the definition of determinism: "given the laws of nature and the state of the universe at any one time, there is only one possible state of the universe at any later time." Why might someone think this is true? I can think of two sorts of reasons why someone might think something like this. One is theological, and the other is scientific.

Some people who believe in God think of God as the sort of being who has absolute control over all of creation. Those who take such a strong view of God's sovereignty will often end up saying things very much like what determinists say. Depending on the details of the view, it may or may not be technically the same thing as determinism. To be determinism, God's means of controlling every event would have to be by setting up the laws of nature in a way that they guarantee every outcome God desires. If the laws of nature are not deterministic in that way, and God just constantly intervenes in the natural order, then it isn't strictly speaking determinism as I've defined it. But it's a sort of theological determinism, with God's interventions causing everything in a way similar to how laws-of-nature determinism would.

One thing to keep in mind about this reason for holding determinism (or really something near enough to determinism) is that it's not really an argument. It's a motivation based on other views. Perhaps there are good arguments for those other views, but it won't convince people who don't hold those other views. Since these views include the controversial claim that God even exists, that limits its force right there, and many theists do not hold such a strong view of God's sovereignty. So this isn't going to motivate everyone to believe something like determinism.

The scientific motivation for determinism is much stronger. In science, we explain things. We seek to explain events in terms of what caused them to happen. We assume that something wouldn't happen unless something caused it to happen. Otherwise why bother to find such explanations for everything that takes place? Now it's possible that most events have such explanations, and a few events do not. But a simpler explanation that treats nature as uniform would not expect such vastly different characteristics of events. If we can make sense of reality with a coherent explanation that applies to everything, we should do so. So the question is really whether there are good reasons to believe of any particular events that they are uncaused.

It turns out that there is scientific evidence for thinking some events are uncaused. In quantum mechanics, some puzzling equations lead a number of scientists to this conclusion. Technically speaking, there are several interpretations of the import of those equations. A philosopher of physics who has published work on the subject once told me that he doesn't think any of the interpretations are actually consistent with the equations, but they're the best we have at the moment. That leads me to want to take what follows with a grain of salt. But it does seem that the dominant interpretation of the equations is that some events on the tiniest level of reality we know about are not caused. That is, right now there is nothing about the world at that micro-level that, given the laws of nature, determines how things at that micro-level will be in the future.

But even if you accept this dominant interpretation at the micro-level, there are differences of opinion in the philosophy of physics about what that means for the macro-level that we can directly observe. Another philosopher of physics I know thinks the micro-level indeterminism is perfectly consistent with macro-level determinism. That is, however these micro-indeterminacies work themselves out, the same result will happen at the macro-level. Since these discussions are all extremely technical, I'm simply taking the word of philosophers of physics I know and respect, but it seems to me that it isn't clear at all that quantum mechanical observations have shown determinism to be false in any way that it matters. Perhaps they have undermined some of the motivation for determinism, but I don't take any of this to be rock-solid proof one way or the other. It will also be clear as the series goes on that those who think determinism is incompatible with freedom don't end up getting much help from indeterminism at the quantum level either, but I'll save that argument for a future post.

What should we conclude from all this? Well, there is a strong argument from scientific practice for determinism, one that is somewhat undermined but not necessarily in any helpful way by quantum mechanics. There is a strong argument from quantum theory that indeterminism is true at the micro-level, but it's also not clear that such indeterminism will make a difference in terms of what we regularly experience at the macro-level. In the end, I don't think we should place much stock in the scientific arguments. Determinism may or may not be true, as far as science shows us, and science has changed enough that current science may not be exactly right anyway. If we come to understand things at a deeper level than quantum mechanics, who knows where things will go? In the end, I don't take these arguments to be all that strong. That leaves me open to either determinism or indeterminism, and we're left looking at what follows from either view without being sure which is true.

In the next post, I'll look at some positive arguments for thinking we are free.


Another philosopher of physics I know thinks the micro-level indeterminism is perfectly consistent with macro-level determinism.

I don't see how this philosopher can possibly be right given the well known phenomenon of the butterfly effect. This effect is such that under the right conditions of instability even the tiniest of movements can be amplified to macroscopic proportions, for example in weather systems. Brownian motion, the observable random motion of dust particles in gas, is clearly controlled by micro-level quantum fluctuations; so is for example the way in which water drops condense from steam etc, a technique used to detect subatomic particles. These can certainly have macro-level effects. Probably the behaviour of a thunderstorm is controlled in many different ways (the original formation of the storm, its movements, its lightning strikes) by such effects. If the micro-level phenomena are truly random, then the macro-level effects of the storm, such as who is struck dead by lightning, cannot be deterministic.

I can't remember how it went, but it had something to do with how the wave function collapses. It was something like an indeterminacy in micro-events that would all lead to exactly the same macro-events. It's true that the macro-events would be very different if certain micro-events were even slightly different, but that doesn't mean there couldn't be several micro-structures with the same macro-structure. This view had it that the macro-structure doesn't guarantee a certain micro-structure but just a certain set of them, and it had deterministic laws at the macro-level despite minor variations at the micro-level while not violating the macro-determinism. I don't know if this is a popular view, and I can't remember where I read it, but it was by a philosopher of physics whom I know personally and who is certainly not some fringe crank on the sidelines of philosophy of physics. He's in the thick of the most central debates in the discipline, and I didn't get the sense that he thought it was inconsistent with complexity/chaos theory.

My general point was really that the issue of determinism doesn't seem to be settled by quantum theory. A number of important physicists and philosophers of physics think quantum mechanical phenomena can be interpreted in a deterministic way.

Thanks, Jeremy. I have looked into this wave function collapse issue, and it is a real phenomenon, although a controversial one. Quantum mechanical processes can be divided into two separate kinds, wave function development, which is considered fully deterministic according to Schrodinger's equation, and wave function collapse, which is apparently random.

Indeed the randomness of the latter, in the specific example of the randomness of radioactive decay, is used in certain random number generators. ERNIE has been generating random numbers for the UK premium bond lottery for 50 years now, on the basis of random noise which is essentially quantum fluctuations. These fluctuations have certainly had macro-event consequences, that a few people have won fortunes, and most others have won little or nothing. Government actuaries regularly test the output for genuine randomness.

So, if this professor is really claiming that there is only "an indeterminacy in micro-events that would all lead to exactly the same macro-events", he has to explain how this is consistent not just with Schrodinger's theoretical cat, which lives or dies on the basis of how a wave function collapses, but real practical and repeatable cases of quantum randomness being deliberately used to have macro-event consequences.

I wouldn't go quite so far as to say that determinism has been experimentally disproved. But it can only be rescued by finding some determinism in wave function collapse at a much deeper level. The only way to rescue determinism that I know of is the parallel universes model, according to which every possible outcome is true in some universe. On this basis there are presumably many universes out there in which I won the top premium bond prize as a child (when I had some bonds for a time) and have lived in luxury ever since! But I don't think I can take seriously a theory in which the number of universes expands at a rate for which "astronomical" is a gross understatement. And this model is deterministic only if all universes are taken as equal, not if the development of any individual universe is followed. Sadly, I can't choose which universe to live in (such as one in which I am a millionaire), but am stuck in this one, so in practice the universes are not all equal.

There are also hidden variable theories. According to these, QM appears to be random but is not. I do find "uncaused events" to be problematic, even at the quantum level. But don't take this as belief in determinism. I don't believe in that.

I don't think the random-number generators show that it's really random. All that it shows is that you can't predict it. I don't how it could even in principle be possibly to show that it's really random, because there's always the possibility that some factor you're unaware of caused it.

I'm not sure how they deal with Schrodiner's Cat in the collapse model, but it's not an issue unknown to the people working in this area. I'm sure they've got something to say about that.

Ultimately, whether determinism is true isn't going to be the main issue anyway. It's whether something enough like determinism to be problematic for certain views of freedom is true. That sort of thing, as we'll see, can be true even if determinism is false.

I don't think you have to take every universe to be equally real to have the collapse. I think you just have to have them in possibility space, with the actual one being the one that collapsed. This isn't something I'm up on, but that's how I thought it had been presented to me.

All I can say to this one is that the British government has been paying auditors since 1970 to try to prove any non-randomness in ERNIE, and they have failed. They can easily detect anything generated by an algorithmic random number generator. Randomness checking is important for many applications, not just lotteries, and so has certainly been studied in detail. So I am sure that any lack of real randomness in the results would have been spotted.

Nevertheless, I would suggest that God may be able to work out his purposes within this real randomness. He could change things within ERNIE so that I would win the million pound top prize (well, he could if I had any premium bonds), and without anything seeming to be non-random. So if you mean to allow for this kind of non-randomness, I would accept it. But I don't accept non-randomness of quantum processes as part of physical law, as would be required by your definition of determinism. So I would come close to claiming that our government auditors have disproved determinism.

I can see how algorithmic random number generators could be detected while much more complex deterministic processes would still appear random. If it's chaotic at all, then it wouldn't be easy to detect it with anything we can measure now, and it's pretty clear at this point that chaotic behavior is consistent with determinism, just not predictable by us.

OK, yes, you may be right about pseudo-randomness resulting from chaotic but ultimately deterministic behaviour being undetectable. Nevertheless, there is no mechanism according to physical law as currently understood for such chaotic behaviour. So, if God controls events by "by setting up the laws of nature in a way that they guarantee every outcome God desires", he has done this in a way which is quite outside the scope of physical law as currently understood by scientists. But then there may be as much new physical law to be discovered in the 21st century as there was in the 20th.

I'm not completely sure you are saying what you are saying Jeremy.

You seem to be saying that an effect could be uncaused. If this is the case, then I would say you are dead wrong. The notion that some effect is caused by nothing is self-contradictory. It just isn't possible.

When you say that "It turns out that there is scientific evidence for thinking some events are uncaused." this is also obviously wrong, and indeed unscientific. It matters not that scientists hold the opinion, but more that it is an argument from ignorance. The only way to show something is uncaused would be to rule out every possible cause, which would require omniscience. (indeterminism of the gaps?)

This is ironically brought to light in your example of micro indeterminism being consistent with macro determinism. The micro/macro divide HAS to be irrelevant to the issue. Why? Because in trying to restrict indeterminism to the micro level you are placing constraints on what can happen when something uncaused happens (Much like they put probabilities on certain quantum states happening). Essentially, the claim is trying to force indeterministic effects to follow rules (constraints). But if they are not determined, then logically there can be no rules to govern their behaviour.

I agree that nothing could be uncaused. I don't think it's contradictory to say it, but I think it contradicts a metaphysical principle that is necessarily true, and that's the Principle of Sufficient Reason. It's not self-contradictory, though. But I don't accept the conclusion that there are uncaused events. That's why I think either determinism is true or some other causal mechanism must be taking place. Fortunately, theists have something to appeal to besides the laws of nature determining things from previous states of affairs. God can cause things directly.

Just because the results of science don't prove without a doubt that things are uncaused, doesn't it count as evidence that they're uncaused? I think we've got loads of evidence that there's no such thing as phlogiston. The evidence for that is the evidence of what heat really is. Similarly, the evidence that electrons might do things without being caused by earlier states of the world is not absolute proof that there are uncaused events, but it is evidence for that claim.

As for the micro/macro issue, why can't you have constraints on something without their being determined by prior states of the world? Suppose I've got the kind of freedom some people think I have, and I can choose between options A and B without being predetermined to choose either of them. Maybe they're two different flavors of ice cream. But one thing I can't do is make a third flavor of ice cream just appear there because I want it more than A and B. I also can't fly around the world instantaneously to find the flavor I want. It's perfectly consistent for there to be constraints on something indeterministic. Indeterminism just means the particular outcome isn't determined. That doesn't mean it's so wide open that there are no constraints. I just means the constraints are so strong as to limit it down to one option.

Jeremy, maybe you have answered this before, but what is your evidence that your Principle of Sufficient Reason is necessarily true? Is this in fact just an axiom, a part of your world view which you are not prepared to question?

Anyway, it seems to me that your principle is violated by the existence of the universe, or for that matter of our own consciousness if you want to allow for a solipsistic position. There are logically two possibilities here: either the universe (or you own consciousness) exists without a cause (basically the atheist position), or there exists some cause or chain of causes outside the universe (or outside your consciousness) which caused its existence. But such a chain of causes must have an end. As Christians we call that end God - although the logic does not require the end to be personal in any sense. But we are then left with God who has no cause. If God can exist without a cause and so breach your metaphysical principle, then how can we be so sure that there are not other beings or events which exist without a cause?

Of course you might say that these other events must also be considered gods, leading to some kind of polytheism I suppose. And a universe which had more than one separate source of causation might not be philosophically very attractive - although I guess that the dualism of some Zoroastrianism is of this kind. I am certainly not arguing for that kind of universe; my position rather would be that these apparently random and uncaused events are caused directly by God. Nevertheless, there does seem to be a logical hole in your argument here.

The Principle of Sufficient Reason seems obviously true. There's got to be some explanation why things would go the way they go rather than some other way. Otherwise they wouldn't be that way. That has always seemed absolutely compelling to me, and I've never understood how anyone might deny it, although I know some really smart people have done so.

I do think it's also necessary for having a decent response to David Hume's skepticism about scientific laws. See the post on skepticism about science and the subsequent post on responses to skepticism about science, both from earlier in this Theories of Knowledge and Reality series.

As for the existence of the universe, I actually think the argument goes the other way. The Principle of Sufficient Reason is a good reason to believe the universe's existence has some explanation, and thus you get the cosmological argument. I've also discussed that issue at length earlier in this series, starting here.

I don't think it's a problem if God has no cause, as long as God has an explanation. Traditional theism takes that explanation to be within God's own nature. If the cosmological argument shows anything, it's that there's something that exists necessarily and by its own nature, and it seems to me that God is a much better candidate for such a thing than the universe is.

These are all issues that get a much more careful and comprehensive treatment earlier in the series, but those are my general lines of thought on them.

An effect is defined as that which has an antecedent cause. Hence a causeless effect is contradictory.

Or, put another way. If nothing could do something, then it would be something, not nothing.

Either the definition of 'effect', or 'nothing' make it clear that a causeless effect ..i.e. caused by nothing, is contradictory.

Like Peter, I think the principle of sufficient reason (PSR) begs the question. That is, requiring things to have a reason at all (sufficient or not), necessarily denies a causeless effect. Although I agree with the PSR, you must appeal to the evidence for the PSR in order to deny causeless effects.

I would disagree with Peter that the end of the causal chain doesn't have to be personal. An impersonal cause is not the alternate explanation you alluded to in your response to me, i.e. transcendant intelligence. What I am saying is there only seems to be two possible categories of causes, those bound by law or transcendent intelligence.

Regarding science and evidence for lack of causation, I think your example of phlogiston was a good one, for my case. It was only that we gained knowledge of what heat was, that we discared phlogiston. It was positive evidence that rejected a previous explanation, not absence of evidence.

To put it another way, an effect can have a near infinite number of possible explanations, only one (at most for this discussion) of which can be true. Ruling out 5 of those near infinite explanations is poor evidence for the lack of a cause. It is an argument from ignorance. The only evidence it has really provided was that 5 possible causes do not cause the effect, not that the effect is uncaused.

As the possible causes are near (if not) infinite, it is not reasonable to claim it is possible to have shown that something has no cause. You may show that we cannot currently predict it's behaviour or know it's cause, but this is not the same thing.

Regarding micro/macro. I was not referring to transcendent intelligence, but to 'nothing' as a causative agent. I do not dispute that a micro/macro level distinction can apply to indeterminism. I do dispute that it can apply to non-intelligent indeterminism. Your example shows that the ability of a transcendent intelligence to choose is limited (based on the interplay of power within the environment and will), but this is different to claiming that the 'nothing cause' can be limited. There are clear reasons WHY you can't create icecream C out of thin air (because you lack the power within the environment). The very nothing of nothing causing something however, necessitates that power within the environment is irrelevant (as nothing has no power). If power is irrelevant, then there can be no limit to what can be done by nothing. Any attempt to limit what nothing can do by placing micro/macro constraints or even probability inherently assumes that power is relevant and thus means that 'nothing' cannot be the cause in question.

To put the micro/macro thing another way...
If you think something can be uncaused, then the principle of sufficient reason (PSR) doesn't apply. To keep the micro/macro distinction however, you need to have PSR apply to only the macro cases. There really isn't grounds for making this distinction however as the difference between micro/macro is quantitative, not qualitative.

An effect is defined as that which has an antecedent cause. Hence a causeless effect is contradictory.

Causation is one of thorniest issues in metaphysics today. This definition is certainly not going to be universally accepted. It precludes backward causation as contradictory, for one thing, and since physics seems to indicate that backward causation might well take place (and philosophy provides all sorts of additional cases that might occur), we probably shouldn't define the word in terms that automatically rule out certain things that might be real possibilities or even actualities.

As for "caused by nothing", you're trading on an ambiguity. That expression can mean either of the following:

1. There is something that is not anything, and that caused it.
2. It wasn't caused.

The first is clearly contradictory. The second is not. It just conflicts with a metaphysical principle that I think is a necessary truth. But the form "P and not-P" the way the first sentence does.

I don't think PSR begs the question, because I think there are independent motivations to believe PSR, as you must also think since you mention that there are some. But that's reason enough to say that it's not question-begging. It's possible that those reasons are insufficient, and thus it would turn out that PSR is unsupported. If so, then determinism is unsupported. But it's not question-begging unless it assumes what it's trying to prove, and it doesn't do that unless you believe PSR because you are a determinist. The argument isn't circular like that.

In the phlogiston case, I don't think we have any positive evidence in the relevant sense. After all, our account of heat isn't an account of some alternative substance besides phlogiston. It's a reductive account, which takes heat not to be some new substance that we didn't have positive evidence for but now do. Heat is just what we already believed in moving faster. In lieu of any evidence of a substance called heat, we just take it to be properties of stuff we already believed in. Similarly, in lieu of evidence of causes of this strange behavior of electrons, we just take it to be that there is no cause. The analogy isn't as close as I'd like, because one is about a reductive definition of a term, and the other is purely about whether something occurs, but they both seem to me to be negative conclusions derived from negative evidence.

That sort of thing happens all the time in police contexts. The lack of evidence of a large conspiratorial plot and the lack of evidence of anyone's being involved besides the one suspect we have counts, together with the fact that one person provides a sufficient explanation, counts as evidence that only one person was involved, even if it's not proof. In this case, since you're right to notice that maybe you're ruled out a few kinds of explanation when loads of others might exist, it doesn't count as strong evidence. I don't dispute that. But doesn't it count as some evidence? That's all I'm saying here. It's some evidence, and we have no evidence to the contrary, together with this (according to some at least, and I'm in no position to evaluate that) being the simplest explanation. Arguments from ignorance aren't strong, but in this case it's all there is to go on, together with theoretical simplicity.

On the last issue, I think you're setting out two opposing views without allowing any middle ground, whereas the majority of philosophers who are indeterminists are in that middle ground. The two extremes are (1) that every event is caused in terms of causal laws and (2) that no event is ever caused or governed by causal laws. But why couldn't there be causal laws that restrict what might happen without determining fully everything that might happen (aside from PSR, which these philosophers reject)? Then there would be some limit on what can happen, but which possibilities within those limits would be completely undetermined. Furthermore, why couldn't there be probabilistic laws that don't determine what happens absolutely? Then there might be a 75% likelihood of an electron's going one way and a 25% likelihood of its going a different way but no chance of its going a third way.

Jeremy, thanks for your response. I knew I needed to read more of your past posts to understand PSR, and I have now read the ones which you pointed out. I now understand how the self-existence of God is defined such that it does not contradict PSR, which implies certain things about God.

But I remain unconvinced about PSR. You yourself wrote here of PSR "it's perfectly fine to believe things that we have no evidence for, under certain kinds of conditions, and this may well be one of them". Yes, it is fine to believe them, but they then become faith positions rather than proper scholarly deductions.

Anyway, your argument doesn't really seem to be any more than "Most people instinctively believe in PSR, therefore I believe in PSR". You might as well say "Most people believe in God, therefore I believe in God". But surely philosophy can't be done by majority vote?

You are saying that you believe in PSR, and apparently therefore (by the cosmological argument) you believe in God. My position would be that I believe in God, who is after all more worthy of being believed in than a mere principle! Anyway, my reasons for believing in God are based on my personal experience, not on philosophical arguments.

And I remain agnostic on PSR unless you can prove that the existence of God implies PSR, and not just vice versa. Of course you may well be able to do that. Indeed probably the orthodox Christian position implies PSR in some sense; it certainly implies that nothing is outside God's control. But that is not quite the same thing, not least because God could have deliberately chosen to allow certain uncaused events to take place within certain limits, just as humans deliberately allow events which they do not cause to occur while keeping their consequences under control - it is called game playing and gambling. I'm not really suggesting contra Einstein that God plays dice, but it is surely a philosophical possibility. And it is also a good explanation of what is observed in quantum mechanical processes, whereas there is no known explanation of them in terms of causes except I suppose to attribute their causation directly to God.

Yes, it is fine to believe them, but they then become faith positions rather than proper scholarly deductions.

I don't think those are the only two options, actually, depending on what you mean by faith. Scholarly deductions are things you derive from other things that you know. But that doesn't give you much if there are only things you believe by faith and things you deduce from things you believe by faith. There has to be something you start with that is more than faith, or everything ends up faith. Philosophers call these things a priori knowledge, the kind of thing we can know just by thinking about it. Truths of mathematics can only be derived from previously known truths, but eventually we get to axioms. Do you think we believe axioms by faith? I think we know them. Unless you think we know things by faith (in which case I'd be happy to suggest that we can know PSR by faith), what makes it knowledge is going to be stronger than faith, or else we don't know anything. Why couldn't PSR be one of these propositions that we can know just by thinking about it?

I don't believe in God because of the cosmological argument. My primary reasons have nothing to do with philosophy, but I do think the philosophical arguments taken together are not a bad collective case, since I do think several of the arguments are on the whole good enough to combine with others as a collective case. The cosmological argument by itself doesn't even get you full theism, though, never mind Christianity, so I would hope that my belief in God doesn't rest on that.

Leave a comment


    The Parablemen are: , , and .

    Twitter: @TheParableMan



Fiction I've Finished Recently

Non-Fiction I've Finished Recently