This statement annoys me because, if one actually bothers to assign a meaning to the words “free will”, it’s clearly not true, but it’s claiming a veneer of truth entirely based on the use of a deliberately undefined word.
If a free-willed agent “decides” to do A in response to B, there are two questions:
Why did they decide this?
Why haven’t they changed their mind since then?
Until the agent changes their mind, they become a simple link in the causal chain - they saw A, and because they have a standing decision in place to do B when they see A, they do B. This is not a disproof of causality; it’s an example of it. And this is true regardless of why they decided to act this way, because once the decision is made it’s a rule that’s in place regardless of its origins.
But let’s pretend for a moment that why they decided to act this way matters to causality and look at those two questions again. If you define “free will” in the stupid way, where it means “unpredictability”, then the answer to 1 is “it just randomly happened” and the answer to 2 is “it hasn’t randomly changed (yet)”. Under more sensible ideas of free will, people with free will make decisions based on their internal state -memories, moods, preferences- possibly with some random perturbation influencing the process but not enough to make you randomly decide to eat dogshit one time out of ten. Under these ideas of free will, a person who chooses to follow A with B will have done so for a reason, and this reason will remain in place until the variables that caused it change enough to cause a different decision to be made instead. (Or when the reasons for and against doing it are so perfectly balanced that the person really doesn’t care, at which point it will be in the range for random perturbation push it one way or the other.)
Essentially, if you don’t choose to define free will (and human brains) as a big black box with a giant question mark on it, free willed agents become part of causality, not an exception to it.
You don’t have to prove to be able to do otherwise to meaningfully act freely. My decision to eat peanut butter toast for breakfast isn’t only free if I choose jelly at least once. For one, it doesn’t actually do anything to establish my freedom; furthermore, it would mean I’m actually less free: there’s a choice that’s not open to me, the choice of doing the same thing time and again.
The notion of causality here is just the same black box you (rightly) claim free will to be. Nobody has an account of how free will could work; but nobody has an account of what it is about A that makes it cause B. You can appeal to natural law; but then, the black box just is the mechanism of how natural law makes the world work in accordance with is.
All we have is generalizations from observed rehularities. You observe that A leads to B, and claim that this is due to a causal link; but that’s just a flat stipulation. You can claim it’s due to natural law; but that’s the same. Equally, you can claim it’s because B was freely chosen; that’s once again a flat assertion. I don’t claim to know how this free choice could conceivably work—we’re in agreement there. I’m just also aware that I have equally as little of an idea of how causation works.
Um, I TOTALLY have an idea of how free will works. To me it’s not a black box. It’s not cryptic at all. It’s absurdly simple, in fact.
Sentience, now, that is pretty freaking elusive. (Probably.) But free will? Trivial. All you have to do is define the damn thing, preferably in a way that’s not self-contradictory or nonsensical.
And I think it’s reasonable to draw a distinction between not knowing why the laws of physics are as they are and the subsequent actions that we know will happen because the laws are in place. We may not know why electrons repel each other, but taking the fact that they do as a given we can explain why a row of dominos fall.
I’ve put the point to you, several times, that it is not the same because causality makes verified predictions. But let’s imagine for a moment that you were right and they were the same.
The best that your argument could possibly be saying is that free will and causality are equally baseless. So what if this is true? It would tell us precisely nothing.
No, because causality has a definition.
Your separate issue of why causes cause effects is completely besides the point in terms of empirical models. Yes it’s an assumption but it’s an assumption that underpins basically all rational thought including any free will model.
And I responded by pointing out it’s not causation that makes these predictions, it’s the regularities inferred from observation. Causation is then postulated to account for these regularities, but really, that’s just saying ‘there’s some regularity-creating principle at work’. The regularities themselves could also exist without causation—they could be due to the free choices of some agent, or they could be random, and we’ve just gotten lucky, or they could be due to Leibnizian pre-arranged harmony, or due to a block universe sort of view. Neither of which explains anything: we don’t know how things happen randomly, and we don’t know how they happen according to volition, but neither do we know how they happen according to causation.
The point is that this means that attacks on free will that assume causation are ungrounded, and don’t actually succeed in making any headway on denying free will. They just betray the predilection of those making these arguments to accept causality as a primitive notion. Note, that’s entirely a valid move: there’s no reason to expect everything in the world to be amenable to explanation; it’s honestly a bit of a shock we’ve gotten as far as we did. But then, accepting free will on the same grounds should be just as valid.
Well, so does free will in that sense; but neither tells us anything about how their objects act. You require of those adhering to free will to be in possession of such an account, but you can’t put one forward for causation.
I’m not asking why, I’m merely asking how—what is it about causes that make them cause effects. As pointed out, physics has just gotten round this problem by calling all change-causing power ‘energy’, but this obviously doesn’t explain anything—we could just have gone with ‘causal power’ instead, just putting a label on it. Which is completely fine, of course: it’s simply not the job of science to explain how the regularities it describes come about; it works completely fine without it—hence, the notion of causality does not in any way underpin ‘basically all rational thought’. The notion of inference from observed regularity to (in principle fallible) rules does, but whether there’s any causality behind it, or these rules hold just by whim and will of a free agent, or whether we’re just experiencing a spacelike slice of a timeless four-dimensional block universe does not influence whether that inference is valid.
Perhaps to illustrate my point that causality isn’t needed to do science, take a look at Schmidhuber’s Algorithmic Theories of Everything. In this, a ‘history’ of the world is just a process of random sampling from a probability distribution P(x)—that is, any event occurs randomly. Now he proves, given some formal requirements on the probability distribution, that almost all universes given rise to by this process will be highly compressible—that is, their history is much more ‘orderly’ than just a random sequence of events. In fact, for these universes, a formal notion of induction—Solomonoff induction—holds, and it can be shown that given the history of the universe to a given point, one can make a prediction likely to be true with a high probability.
There is no causality at all in these universes. Every event occurs completely independently of those coming before them. How these events occur is, of course, again left open—it exchanges the black box of causation for that of random sampling. But this gives a demonstration of the claim that it doesn’t need causation to explain the success of science, and hence, this success doesn’t allow an inference to the reality of causation.
No. Causality makes those predictions, the “regularities inferred” is a starting assumption for all rational reasoning, including any potential model for free will.
Sure, but such hypotheses need to be verified.
Let me explain what is meant by verification, because I think this might be part of the sticking point. Consider the following two attempts at reasoning:
A: If magic pixies exist, I expect to find pixie dust footprints in my kitchen. I found such footprints, therefore I have grounds to think they exit.
B: If magic pixies exist, the sun will rise tomorrow. I find that the sun did rise, therefore I have grounds to think that magic pixies exist.
(A) is valid empirical reasoning. We might want to see a lot more to gain confidence in our pixie model, but it’s a start. (B) meanwhile is not good reasoning, as we expected the sun to rise anyway, plus we haven’t suggested any mechanism linking pixies with planetary rotations.
Now, when it comes to causality, we have plenty of (A)-type models. Indeed countless billions of implicit predictions every day are verified true in agreement with causality.
Meanwhile the best you have suggested for free will so far is a (B)-type rationalization.
Again, I don’t think this is even a well-formed question.
Let’s take for example an electron absorbing a photon and raising its energy level. We have detailed mathematical models explaining the mechanism of this in terms of the properties of electrons and photons. We use devices that rely on this happening to high precision countless quadrillions of times.
Now, the question of why these particles have the exact properties they do is a valid one and a subject of study and debate.
However the question of “how” a well understood mechanism causes effects is one that would leave most scientists scratching their heads as to what you mean. You may as well ask the same question of truth by definition. “How” is it, that given that X is true, that X is true?
(I’m putting “how” in scare quotes because I think you’re really asking a why question, but with how substituted to make it sound like an empirical claim).
No. The best that you would be doing here is showing that an argument against free will applies equally to causality.
I think you are wrong, for the reasons given above, but even if right, your point would not be showing anything wrong with that argument, let alone anything to support free will.
Again, that’s just not true. I’ve given several examples now where you have just the same inferred regularities, yet no causality. Schmidhuber’s view above, but also a plain old block universe, or something like Julian Barbour’s view, where really only three-dimensional slices of the universe are real, which we perceive as having a causal connection due to the fact that each present moment encodes past moments via memories, or ‘records’, of them—but where it’s not true that one three-dimensional slice ‘causes’ another. (Of course, the occasionalist picture, where each next event is due to the free choice of an agent, works just as well.)
The problem is that the block universe, Schmidhuber’s algorithmic universes, and Barbour’s view have all the same predictions as one in which causality, whatever it may mean, exists—because it’s just not an empirical issue.
Which is, really, my only point. It’s just that in the case of free will, the absence of an explicit account is grounds, to you, to dismiss it, whereas in the case of causation, it just means we shouldn’t ask. This doesn’t tell us anything but where your metaphysical commitments lie—which is fine, but you shouldn’t pretend that these are somehow the final word on the issue, or ‘a starting assumption for all rational reasoning’. Rational reasoning works just fine in a universe without causation.
I’m not trying to make anything sound like an empirical claim. Indeed, I’m basically arguing that it’s not an empirical matter!
Not if that argument hinges on the reality of causation. Which e. g. any argument that alleges that a decision must be ‘either random or based on prior data and brain states’ does.
I would agree with this, with the additional detail of “…in relation to that particular choice” i.e. the focus of a disqualifying compulsion must be the particular choice under consideration. Which is, I think, what you meant but I just wanted it stated clearer.
Of course I am a compatabilist, Dennett fanboy that I am.
Yeah - to abrogate free will you have to have an external force ‘reaching in’ and manipulating the agent’s decision-making process, possibly to the degree of outright determining the outcome, with the intent of achieving this end. “God hardened pharoah’s heart”-type of things.
I guess I’ll leave it as an open question how direct and ‘non-traditional’ the interference has to be to count as an abrogation of free will. I mean, there’s sort of a continuum:
Seeing a billboard that says “eat Wheaties.”
Being strapped down and having “eat Wheaties” piped into your ear on a continuous loop.
Being indoctrinated into a cult, brainwashed by them, and told to eat Wheaties.
Having a chip implanted into your brain which physically interferes with your thought processes and makes you eat Wheaties.
4 is clearly an abrogation of free will, and 1 clearly isn’t, but I’m not entirely certain whether 3 is, since it does deliberately alter how you think, just not with tools.
I’m putting together two of your responses here to show the incoherency.
I said that causality makes those predictions. You say “that’s just not true” but then go on to explain how other descriptions make the same predictions, implicitly agreeing that causality does indeed make those predictions.
Now, I would say that the reason that these alternative descriptions make the same predictions is because the latter two actually assume causality (e.g. even in a block universe, the events in time = 2 are necessarily linked, causally connected, to the events in time = 1. The only difference is we are not saying that there is an arrow of time by which time flowed from 1 to 2). But this is besides the point, the important thing to note is that your objection here is completely flawed.
Maybe, but that’s not what I said. I said that the “regularities inferred”, as you put it, which is alluding to inductive reasoning, is necessary for rational reasoning. So, your argument that causality relies on such an inference is a red herring since any model relies on that, including any model of free will.
Forget causality. If you can talk about any alternative to these options, I’d love to hear it. Heck, that’s what I asked in my very first post in this thread.
If the response is “free will…but we can say nothing about what that actually means and how free will makes decisions” then I rest my case. The concept itself is garbage.
Hence, any other principle giving rise to the same regularities will allow the same predictions, with anything else being metaphysical speculation (which isn’t a bad thing, but should not be confused with empirically accessible data). Thus, it simply isn’t true that the ability to make predictions necessitates causation.
No. There is in particular no influence that goes from t_1 to t_2, and hence, we are, when constructing a block universe, completely free what we put at t_2. Hence, in no sense has what is at t_2 been caused by what is at t_1.
No. You claim that because we can successfully make predictions, we have reason to believe in causation (which is essentially affirming the consequent). But that’s false exactly if there are ways to make the same predictions without appealing to causation. As there are such ways, your argument, besides being formally invalid, also demonstrably yields a false conclusion.
Essentially, you’re saying, whoever is wet must’ve taken a shower. Thus, since there are wet people, we have reason to believe showers must exist. I then point out that it’s not showering that makes people wet, it’s coming into contact with water. As other ways to come into contact with water exist, like going outside in the rain, your inference doesn’t hold (and formally, it wouldn’t even hold without counterexamples, as it’s logically fallacious). Of course, then you go on to claim that there’s really no such thing as rain, it’s just a really big shower…
I still don’t know what you think that argument does. You can do perfectly sound reasoning without relying on any sort of inferred regularities. Can you perhaps give some sort of argument for why you think we couldn’t?
I’ve given several, but for ease of reference:
A universe following physical laws:
It’s sometimes posed as an answer to the problem of causality that there really isn’t anything to A, as such, that enables it to produce B, but that rather, there is a physical law to the effect that ‘A leads to B’
Black Box: How is it that physical laws make stuff do things? (‘What breathes fire into the equations?’)
Schmidhuber’s algorithmic universes:
Every event is sampled from a probability distribution which is subject to a constraint of formal describability; this can be shown to imply that the sequence of events will typically be far from random, and instead, highly regular, thus allowing us to make sensible predictions (without needing a notion of causation)
Black Box: How are events chosen from the probability distribution?
A mathematical universe/radical Platonism:
Whatever happens, does so because1 of constraints of mathematical entailment
Black Box: How does mathematical entailment make anything happen?
A timeless universe:
in the most straightforward formulation of quantum gravity, you end up with an equation that’s just a constraint, and doesn’t contain any sort of time evolution; a sort of time evolution can be constructed by the Page-Wootters mechanism, which uses the entanglement between systems to provide a kind of ‘clock’ for one of the subsystems—any state of the subsystem is then relative to a particular state of the clock system, but it’s not the case that ‘prior’ states of the subsystem cause ‘later’ ones—they’re just particular decompositions of the total system
Black Box: Why those decompositions?
A block universe
A block universe is just like a line of text: each following character isn’t caused by the preceding one, and neither is each ‘later’ three-dimensional slice caused by the preceding one; in particular, each regularity could stop at any monaoeginbogb
Black Box: Well, why is the block universe laid out as it seems to be?
Barbour’s view of time as a succession of ‘moments with memory’
Picture each instant as a polaroid, which contains (among other things) another within it: you can use these to assemble a succession of moments, by looking at which one is contains within which other, but that doesn’t entail that one caused the other—indeed, it might be that all possible polaroids exist, which would give rise to all possible successions
Black Box: Where to the polaroids come from?
Malebranche’s occasionalism
At each point in time, god (or a demiurge, or whatever) intervenes to create the next one
Black Box: Well, god?
Leibniz’ pre-arranged harmony
The universe is, like a film, set up in such a way to play out in a particular sequence; as you wouldn’t say one frame of a film causes the next, you can’t really say one instant of time causes the next
Black Box: Who did the setting up?
Dependent origination
Buddhist and other nondual traditions have always been skeptical of causation, as it implies a dualism of ‘cause’ and ‘effect’; in Buddhist views, this burden is born by the notion of dependent origination, which holds (roughly) that things emerge in pairs, supporting each other—neither of which being cause or effect
Black Box: Well, how does that work, though?
The no-information universe
It is sometimes noted that ‘everything’ contains just the same information as ‘nothing’, namely, none; but once you take things away from ‘everything’, a finite amount of information emerges, paralleling a ‘creation from nothing’ from the other direction; changing the way you split ‘everything’ gives rise to different ‘somethings’
Black Box: Again, how does that work?
Lewisian Humeanism
The world is just a patchwork of local matters of fact, occurring independently
Black Box: But where does that patchwork come from?
There’s probably loads of others I’m forgetting, but I think that’s a start. If you have questions about any, I can provide some pointers to further material.
Yeah, that’s really the extent of your argumentation, isn’t it. Have no account of causation? Well, don’t worry, since causation is true, that just means it’s probably not a sensible question, causation works in mysterious ways, we were not meant to know, etc etc, all is fine. Have no account of free will? A-HA, see, that proves that the concept is garbage, contains nothing but sophistry and illusion, commit it to the flames!
1
Summary
By the way, not that I think you’re silly or desparate enough to try, but I think there are some that might: using ‘because’ does not in any way entail anything about causation; it’s a relation of logical entailment—there is no finite representation of the square root of two because it is irrational, but it’s not like its irrationality ‘causes’ the absence of such a relation
Forgive me, but this doesn’t seem like what you are claiming. He is claiming that people who are wet are wet because they came in contact with water, while you are claiming that the fact that people are wet tells us nothing about a past state that caused this and in fact people may be wet for no reason at all.
Arguing against causation is arguing that just because we have witnessed people come in contact with water in the past and get wet is not evidence that they are wet BECAUSE of that contact, and in fact it could simply be a coincidence that we have observed wet people following observing people coming into contact with water, and some people may be wet for no reason at all or may come in contact with water but not get wet.
No, you’re misunderstanding the analogy. He’s proposing (as I’m understanding it) that our ability to make predictions is due to the existence of causality, and that hence, without causality, our ability to make predictions would be mysterious. I’m saying that, actually, it’s the consistent regularities inferred from observation that allow us to make predictions, and hence, everything that leads to such consistent regularities would be sufficient grounds for our ability to make predictions.
So the analogy is:
A: being wet ↔ ability to make predictions
B: taking a shower ↔ causation
C: coming into contact with water ↔ consistent regularities generalized from observation
Basically, he’s saying that B → A: because of causation, we can make predictions about the world (possibly with an added, unstated premise that nothing else could explain our ability to make predictions about the world, as otherwise, the argument would just boil down to fallaciously affirming the consequent). I point out that no, it’s in fact C → A: because of persistent regularities inferable from observation, we can make predictions. Hence, the fact that we can make predictions doesn’t tell us anything about the reality of causation.
So the thing about getting wet and coming into contact with water wasn’t intended as an argument against causation, it was just intended to highlight the structure of the argument I take @Mijin to be making, and how pointing to an alternative explanation of these regularities (such as the ones given in my previous post) shows it to be ineffective.
