In a recent OP at Uncommon Descent, Vincent Torley (vjtorley) defends a version of libertarian free will based on the notion of top-down causation. The dominant view among physicists (which I share) is that top-down causation does not exist, so Torley cites an essay by cosmologist George Ellis in defense of the concept.
Vincent is commenting here at TSZ, so I thought this would be a good opportunity to engage him in a discussion of top-down causation, with Ellis’s essay as a starting point. Here’s a key quote from Ellis’s essay to stimulate discussion:
However hardware is only causally effective because of the software which animates it: by itself hardware can do nothing. Both hardware and software are hierarchically structured, with the higher level logic driving the lower level events.
I think that’s wrong, but I’ll save my argument for the comment thread.
In your world, what does “defensible” mean? I gave you an example; you said nothing about it. Perhaps it was not attackable?
Erik,
Are you kidding?
Link, link, link
keiths,
Fool! Your pathetic “evidence” is no match for the overwhelming power of True Logic!
You mean your assertion that the brain gets a nudge from itself? I guess you are not kidding.
You don’t know what “defensible” means. If you knew, you would have been able to define it, because I asked you to, and you would not use indefensible assertions yourself.
Erik,
This is disappointing, though not surprising. I was hoping you would at least attempt to respond to my three comments, for which I provided convenient links that you merely need to click.
You don’t seem to have a lot of confidence in your position. That’s okay. When evidence is on the table, you tend to get skittish. Perhaps someone else will pick up the ball you dropped.
You mean “intellect”, or whatever immaterial stuff you think causes us to think, gets a nudge from itself?
keiths is against top-down indefensible assertions but has a bottoms up explanation handy.
Mung,
Keiths is simply saying that when it comes to understanding and explaining intentional action, the burden of proof lies on those who insist intentional actions must be caused by something that isn’t described by any of the natural sciences.
As usual in these kinds of cases, the naturalists and anti-naturalists can go all day and all night insisting that the burden of proof lies solely on the other side. Rather few philosophers have tried to show how free will is compatible with naturalism — Dennett being one of them.
I am less certain that compatiblist free will makes really good sense. It could be that the libertarian conception of free will makes better sense of the concept of free will than other conceptions do. But in that case, so much the worse for free will.
I’m willing to say that I don’t think there’s such a thing as free will and it doesn’t matter anyway. Our social practices of holding each other responsible, punishing vice and rewarding virtue, do not require any metaphysics of “free will” and are perfectly compatible with cognitive and affective neuroscience.
Making sense is not a particularly good sieve.
We have intuitions based on everyday experience and cultural learning, and we have the findings of physics and of neuroscience.
None of these address all the questions that philosophers can ask, and none of them provide perfectly self-consistent systems that lead us to the best solutions to social problems, or even to solutions of how to manage our own lives.
But if I need to place a bet on which paths will lead to progress, I will bet on physics and neuroscience.
petrushka,
I agree with that completely in spirit and mostly in letter.
I’ve opposed the idea that we can do metaphysics by logic alone because I think that logic is far too weak to constrain our choice of metaphysical systems. Logic can tell us which systems are impossible or incoherent, but it cannot tell which of the possible, consistent metaphysical systems is likely to be true. (This is why one of the reasons, I suspect, for why metaphysics is poorly regarded by scientists and other more practically-minded persons.)
In fact, the situation is even worse than that because there are many different logics, classical and non-classical. We simply do not know if there is a single logic yet to be discovered which unifies all logics or if logical pluralism is our condition. And since what is possible within one logic is not allowed within other logics, there is even less hope for grounding metaphysics within logic alone.
There are a few ways of unpacking the concept of naturalism, but one of them — one that’s important to me, certainly — is the idea that
(1) iterated testing of explanatory models offers more reliable grasp of objective phenomena than other forms of inquiry,
(2) other forms of inquiry, however valuable for living a rich and rewarding human life, run afoul of the Myth of the Given when they are taken as securing a grip on objective phenomena, and
(3) because of (1) and (2), our metaphysics and epistemology should be as tightly constrained by science as we can make them, while nevertheless recognizing
(4) in philosophy we use concepts in ways that exceed or transcend the currently available resources of science.
In other words, metaphysics is metascience — not in the sense of grouding science, but in the sense of going beyond it.
I absolutely and utterly disagree with you. keiths could care less about intentional action. keiths is a level 0 physicalist. A class 1 reductionist.
The properties of water can be reduced to and explained by the physics of hydrogen and oxygen which can be reduced to and explained by the physics of whatever it is that makes up and explains the elements which can be reduced to and explained by whatever it is that makes up and explains the whatever it is that makes up and explains the whatever it is that makes up and explains whatever it is that make up and explains which can be reduced to …
I don’t see that in keiths argument.
So?
Not making sense is better.
Keiths can speak for himself should he choose to do so. But it seems pretty clear to me that his criticism of “top-down causation” is a criticism of a specific model or conception of intentional action, and not a rejection of the very idea of intentional action.
If you want to convince us that agent causation is the only or even the best way of making sense of intentional action, then you would need to actually make an argument. And you would need to convince us that the kind of “top-down causation” that agent causation requires is either (somehow) consistent with the causal closure of the physical, or if it is not, then we have better reasons to affirm agent causation than we do to affirm the causal closure of the physical.
It’s not easy to see how mental causation — how the mind causes actions — is consistent with causal closure of the physical. Figuring that out is an immensely difficult problem. Philosophy is, in fact, not easy. Bullshitting is easy. Philosophy is not.
KN, to Mung:
That’s right. I’m not an eliminativist. I think that intentional action is real, but that it can also be described at the physical level with no reference to intentions or agents. Same phenomenon, two levels of description.
Wouldn’t it be great if Mung could actually muster such an argument instead of sniping and dodging?
Hence Mung’s preference for the former.
Gregory,
Can you offer — and defend — an example of top-down causation?
keiths,
I liked Ellis’s essay very much, but I’m not sure the examples he uses actually support his thesis.
Those examples could be interpreted to show that there are many interesting cases in which system-level properties function as parameters that constrain how much freedom can be assigned to the variables that describe properties of parts of the system. This is a system-to-subsystem constraint, not a top-to-bottom constraint. One could stipulate that system-to-subsystem constraint is “top-down,” but that presupposes that a hierarchy of descriptions is the correct metaphysics of science, and (for reasons I’ve sketched previously in this thread) I do not think it is.
That aside, system-to-subsystem constraint does not support any strongly voluntaristic theory of agency, libertarian freedom, or agent causation, because system-to-subsystem constraint does not involve any violation of the causal closure of the physical. I take “the causal closure of the physical” to mean causal efficacy can only be meaningfully attributed to structured processes that can located within a spatio-temporal relational system. (Thus, “abstract objects,” if there are any, lack causal efficacy.)
But unlike Keiths, I do not think that it makes sense to say that all causal powers can be reduced “in principle” to the structured processes described by the models used in fundamental physics.
Could this be a frame of reference issue? Was the death caused by the bullet? (yes, if you’re the coroner), the gun? (yes, if you’re an arms dealer), the shooter? (yes, if you’re a judge), or a violent armed society? (yes, if you’re a politician).
Yes, I think so. How we talk about “the cause” depends the kind of inquiry we’re engaged in, the background assumptions being made that inform that kind of inquiry, and so forth. Somewhere John Austin remarks that there are as many different kinds of cause as there are uses of “because”. (I think that’s actually a deeply Aristotelian insight. I myself find it helpful to read a lot of Aristotle through the lens of ordinary-language philosophy — esp. Ryle and Strawson.)
But, precisely of that, I don’t share Keiths view that causal relations described in our models of fundamental physics are the “base” level, and everything else “supervenes” or “emerges” from that level. Put in Dennettian terms, I don’t think that the physical stance is somehow “closer” to real patterns, or “more real”, etc. than the other stances.
I’ve consisted rejected both top-down and bottom-up causation throughout this conversation. The idea that I’m defending bottom-up causation indicates a lack of basic reading comprehension.
I don’t think causation can be coherently defined. It’s one of those intuitive/ learned concepts we take for granted, but the more we know, the less we know.
Ok, I’m putting Gregory on “ignore commenter”. The rest of you can engage with him — or not — as you wish.
Moved some comments to guano.
“Bullshitting is easy.” – Kantian Naturalist
(Will this direct quote with NO COMMENTARY escape TSZ’s censorship?)
Not accusing fellow commenters of “sophistry” (which in my book is synonymous with “lying”) would help.
Gregory, accusing someone of lying has less of a chance of ending up in Guano than does accusing someone of sophistry, unless your name is Gregory.
It used to bother me a lot when Gregory accused me of “sophistry,” because (a) it is a term of abuse and (b) it is false.
But I’ve long since stopped caring, because it gradually became clear to me that Gregory uses this term because I reject “vertical transcendence”. There is a long tradition within Western (and Eastern) philosophy of rejecting “vertical transcendence”, including Democritus, Epicurus, Lucretius, Spinoza, Nietzsche, Marx, Dewey, and Adorno.
That is indeed the tradition of anti-mysticism, anti-dualist, anti-hierarchical, anti-kyriarchy philosophical-political critique within which I operate.
If Gregory wants to call me a “philosophist” because I am on the side of justice, freedom, dignity, equality, and diversity, that says a good deal about him and nothing at all about me. His vitriol “neither picks my pocket nor breaks my leg,” as the saying goes.
I use the term ‘philosophist’ because KN is amongst the slipperiest of quasi-philosophers I’ve ever read. Pretend wisdom (to which his protégé here doesn’t even qualify for consideration). It’s sad how confused he sounds to the ears of a qualified social philosopher (and he’s denied Erik audience because, well, that’s KN’s duplicity). He’s called himself an empiricist, naturalist, ecosocialist, environmentalist, feminist, emergentist, LGBTist, anti-foundationalist, compatiblist, reductionist (sometimes), scientism proponent, sometimes ‘Darwinist,’ pragmatist and skeptic. But the key to understanding KN is that he is an atheist (without even mentioning that other BIG issue). Period.
Look at the clever devils he chooses to call his heroes!! Gollum.
And that’s what this hive of unbelief blog TSZ really promotes: atheism. Lizzie denies this weakly, but it’s quite obviously true.
Who here of the ‘skeptic’ ilk actually denies this?
No, KN, you are on the side of relativistic waffling based on ultimate atheistic despair.
Anyone with half a brain would accept Ellis’ ‘top-down’ understanding of reality.
Religious Jews, along with their brothers and sisters among Christians and Muslims are champions of those ‘concepts’ you seek to wrestle free as your own individualistic (from a socialistic family, as told here) worldview. That quickly ends the ‘champion’ claims of KN’s confused cognitive and moral shuffle.
No one genuine need pay any attention to the philosophistry of someone who has demonstrated themselves amply as TSZ’s resident ‘let’s be comfortable in our atheism’ disenchanted philosopher operating under the pseudonym ‘Kantian Naturalist’. Don’t burden people with your scientistic Sellars-out unbelief. Don’t sell-out belief for twisted self-congratulation without faith.
Gregory:
Then it will be easy for you to answer my question:
What do you mean by top-down causation?
And round and round we go.
Mung,
Your evasions are boring, Mung. I told you on January 30th:
We’ve also discussed Noble’s and Talbott’s writings on the topic. Did you understand any of them, or were they all over your head?
According to Ellis, “top-down causation happens wherever boundary conditions and initial conditions determine the results. Environmental variables (a macro scale concept) act down to determine the values of physical fields locally”. All of his examples fit this model of macro-to-micro constraint. That is, his examples show how system-level properties can constrain the behavior of subsystems.
It seems perfectly cogent and correct to say that, when we adopt the mechanistic stance, we identify some system as providing boundary conditions for the subsystems that we want to manipulate. And that means we can’t understand our own manipulations without taking into account the boundary conditions of the systems that constrain those manipulations.
All well and good!
But nothing there offers any hope for agent causation, libertarian freedom, voluntaristic conceptions of agency, and so on. There’s no violation of the causal closure of the physical, which is what agent causation seemingly requires.
All that Ellis has done here is remind us that we can’t forget about macro-to-micro constraint when we adopt the mechanistic stance. That is indeed quite useful and important. But he hasn’t done anything beyond that, and frankly I’m puzzled by why Torley thinks that Ellis is an ally here.
Since we’re discussing Ellis’s article, we’re using his definition of the term. That’s why Keiths posted the link to it.
Top-Down causation: When I move the light switch from the up position to the down position the light goes off.
Bottom-Up causation: When I move the light switch from the down position to the up position the light comes on.
And so I keep asking.
keiths is not arguing against top-down causation, he is arguing against Ellis’s notion of top-down causation which, keiths tells us, is wrong.
So assuming a dismantled Ellis notion of top-down causation, what have we really learned about what top-down causation is or whether top-down causation actually exists?
Now perhaps it’s just me, but I can’t imagine claiming in an OP that I am going to argue against some concept and then choose as my antagonist someone who is mistaken about that very concept.
Maybe I will start an OP about why evolution is false and quote young earth creationists to make my case.
keiths has been given many examples of top-down causation. From computer software to computer hardware to transistors. Perhaps a simple switch is all that is needed for the light to come on.
And yet creationists and IDists continue to quote Darwin and neo-Darwinists against evolution, even though there have been a few discoveries since 1940.
So imagine how foolish it is to start a thread entitled george ellis etc, and expect to discuss george ellis.
Watch your language please , Mung is a moronic theistic ” skeptic”
There are a couple of different issues here.
One issue is whether there is any top-down causation within the physical/material world. Ellis offers a specific conception of top-down causation and some examples of what he means by that. Keiths is taking issue with that conception. I am still unclear on Keiths’s objections to Ellis.
To my way of thinking, what Ellis (and also Talbott) offer us are good examples of macro-to-micro constraints in holistic systems. I think that they are right about that. But that still leaves the floor open for anyone to offer a different conception of top-down causation.
Thus far the only person to offer an alternative — Erik — is an alternative that looks like a version of agent causation that involves violating the causal closure of the physical. Ellis and Talbott does not commit themselves to any violations of the causal closure of the physical, whereas Erik explicitly holds that intentional agency does involve violating the causal closure of the physical. (Mung seems to agree, but who can tell with him?)
Another issue is whether the causal closure of the physical entails that all empirically detectable structures supervene on the structures modeled by fundamental physics. Keiths seems to think that it does. I don’t.
I object to Keiths here because I don’t think that we’re entitled to make a metaphysical claim unless we can indicate the epistemological basis of that claim. (“No metaphysics without epistemology” is a pragmatist mantra.) “In principle reduction” is a just metaphysical blank check if there’s no corresponding epistemological account of how intertheoretic reduction even could succeed for any possible community of inquirers.
And while there are some good examples of intertheoretic reduction — say, the reduction of Maxwell’s equations to quantum electrodynamics — I tend to think that successful reduction is going to be extremely rare. The good examples of successful reduction in the history of science are all drawn from the history of physics. I don’t even think that biology is reducible to physics, a least not in any sense that matters to successful inquiry.
But I’m happy to accept the causal closure of the physical as a working principle, as long as it is loosened from reductionist shackles. Specifically, I don’t see how we can verify the ascription of causal efficacy to something not located within a system of spatio-temporal relations. (Verification is a bad criterion of semantic content — the logical positivists were wrong about that — but it’s a good criterion of epistemic significance!) Consequently, I don’t see how any invocation of causal efficacy to something not located within a system of spatio-temporal relations can be anything more than just saying “it’s magic! Now stop asking questions!”
(This is the heart of Benacerraf’s Dilemma for mathematical platonism, by the way.)
To conclude what should have been a much shorter post:
1. The causal closure of the physical is the only alternative to saying “it’s magic!”, which dooms all of the following: agent causation, libertarian freedom, and substance dualism (the immateriality of the mind).
2. Whether there is “top-down causation” within causally closed physical systems depends on (a) whether one is persuaded by Ellis and Talbott that there is macro-to-micro constraint in holistic systems and (b) whether you want to stipulate that kind of constraint as “top-down causation”. I accept (a) but not (b).
It certainly appears to me as if you and I are in agreement here.
I asked keiths what he means by top-down causation. He claimed in his OP that he was going to argue against top-down causation.
Later I accused him of actually arguing for top-down causation, but he assured me that is not the case because Ellis’s notion of top-down causation is mistaken and he is arguing against Ellis’s notion of top-down causation. From which it in no way follows that he is arguing against top-down causation.
But it’s no doubt my fault for not grasping some missing nuance. Maybe I’ll find the secret in his drifting Weasel program. 🙂
🙂
I don’t think I’ve appealed to either minds or agents. I think a transistor provides an example of top-down causation. As does a light switch.
What ball did I drop? I say length is not a physical phenomenon, you say it is. We cannot agree on a definition of money or on the nature of memory.
You have some inexplicable opposition to top-down causation even though given your presuppositions you are completely unable to distinguish top-down from bottom-up. Since this is so, there never was a ball in the first place. Or you can prove me wrong by giving an example of both kinds of causation and how you tell the difference.
There is no way to say anything that would be uncontroversial to you, but just as an example, let’s take Wikipedia, “In geometric measurements, length is the most extended dimension of an object.” So, length is dimension. What is dimension? “In physics and mathematics, the dimension of a mathematical space (or object) is informally defined as the minimum number of coordinates needed to specify any point within it.” Already the first statement was pure math, not physics. You can tell the difference between math and physics, can’t you? Math would generally be a prime example of conceptual abstraction to people. But when you are keiths, then math is physics.
The difference is, I say math is real. Hardcore physicalists say math is “useful fiction”. But you say math is physics, which goes beyond hardcore physicalism into the realm of category error. To me, conceptual consistency (such as avoiding category errors) is important, as mental things are more important than physical. Given physicalism, logic should be irrelevant (except when others sound illogical to you, of course).
There is that.
Strange. I see length as an example of something that is physical but not material.
This seems to confuse two different meanings of “dimension”. I would tend to use “dimensionality” instead of “dimension” for the minimum number of coordinates.
I’m not any kind of physicalist, but I say that mathematics is a useful fiction. I have come across people whom I would consider hardcore physicalists, who insist that mathematics is physical and that platonism is a version of dualism (they are not sure what to make of fictionalism). And they most certainly do not take logic to be irrelevant to their physicalism.
Maybe all physicalists are not alike and some of them don’t fit your conception.
Mung,
I’m afraid this discussion is over your head. You’ll have to sit it out unless you can find someone willing to spoon-feed you. I am not that someone.
KN,
If the “macro-to-micro constraints” can be recast as “micro-to-micro constraints” with no loss of causal completeness, then the phenomenon in question isn’t an example of downward causation.
That’s not obvious to me, particularly when Ellis says things like this:
KN:
What’s an example of one that doesn’t?
The epistemological basis for my reductionism is that there is massive evidence for reducibility and little against it. My reductionism is provisional, and I’m ready to consider counterexamples — hence this thread — but I haven’t seen any yet.
Again, don’t confuse reducibility in principle with reducibility in practice.