5626463547 | San Jose, CA 95126

Agency and Bottom-Up Determinative Causation

I am reading through a book on theories of the nature of the mind and I wanted to comment on an argument discussed in a chapter by William Hasker. For one thing, Hasker is a very good philosopher worth reading and reflecting upon. For another, his argument is worth discussing because it is interesting and not obviously wrong. This will provide an occasion, as well, to see how arguments against determinism can often be tied to other metaphysical theses without that being made explicit. And once so explicit, we can see that arguments against naturalistic determinism do not as easily slide over against theologically deterministic views like Calvinism.

Nancy Murphy distinguishes between ontological and causal reductionism. The former is a thesis about the the ontology of different levels (physics, chemistry, biology, sociology) and the latter is a thesis about how the fundamental parts together determine the whole. Murphy accepts the former but rejects the latter because it threatens our conception of freedom. In a comment on Murphy’s distinction, William Hasker spends a little time explaining in more detail Murphy’s worry about causal reductionism. I quote in full:

It may be helpful to spell out these objections a bit more fully. If causal reductionism is true, then everything human beings do and say is simply the consequence of the actions and reactions of the elementary particles of matter, operating according to the fundamental laws of physics. What we do is simply the result of what those particles do — there is nothing there but the particles to “do” anything. It is clear, then, we have no real choice about what we do; even if we have the “experience of choosing,” how that choice comes out is wholly determined by the actions and reactions of the fundamental particles. Under those circumstances, it does indeed seem that free will would be an illusion, as Murphy says. And if we have no free will, how can we be responsible? In particular, how can we be responsible before God, as Scripture says we are? There is also the question as to how, if this account is true, we can be rationally justified in any of the reasoning that we do. On this view, our acceptance of conclusions is not determined by the fact that those conclusions follow from known truths according to correct principles of logical inference. Rather, we accept those conclusions, and only those conclusions, that we are determined to accept by the particles that make up our brains, acting and reacting according to the fundamental laws of physics. And those physical laws do not in any way have in view the goal of leading us to true conclusions about the world — rather they are nonteological; they can be represented by mathematical formulas that have no reference whatever to any human goals or objectives. But to accept that all of our reasoning is determined in this way by nonrational factors is devastating to any reliance we might otherwise have had on our theories and belief systems — including, as Murphy points out, our theories and beliefs concerning God and our relationship with God.

–“On Behalf of Emergent Dualism”, In Search of the Soul, ed. Green & Palmer. pp. 84-85

This is a really nice argument against causal determinism given a commitment to common sense, the manifest image, the life-world, whatever. For one thing, it ties together a number of related concerns about reducing causation down to rock bottom: rational inference, a world governed by entities that have no concern for truth, what we do and believe is nothing beyond the result of the fundamental entities’ causal determinations. I see the force of that argument. But for another thing, this is a much more plausible argument than some of the very junior attempts to construct an argument for the claim that determinism threatens freedom and rationality that sometimes show up in Google Search results. One such attempt I’ve read contends that rationality itself would be impossible merely by the fact what is concluded was guaranteed by the past and the laws of nature, if it were so guaranteed. But clearly, some beliefs can be rational even if no decision was involved in forming the belief – look and you’ll see cases from perception that are relevant here. And a moment ago I described coming to hold a belief on the basis of an argument as being moved by a force. One would not say that had it been more forceful to me, my belief would have been less rational, if true. Hasker’s argument, by contrast, is at the least remotely plausible.

Freedom and Responsibility Argument

The first concern has to do with determination from the parts, or particles, to the bigger stuff like us, and how that undermines belief in responsibility and free choice. The way in which Hasker describes the implications of causal reductionism with respect to those phenomena, I believe, gives the argument its force. It also produces its weakness. Hasker uses limiting and extremity words (“wholly”, “none”, “simply”) in his case, so it appears that agents are causally superfluous, which is precisely what agents should not be. Let’s look at the first case of this:

Hasker opens, “If causal reductionism is true, then everything human beings do and say is simply the consequence of the actions and reactions of the elementary particles of matter, operating according to the fundamental laws of physics. What we do is simply the result of what those particles do — there is nothing there but the particles to ‘do’ anything.” (bolds mine) This is a nice statement of the Epicurean cosmology. But what is causal reductionism (CR) again? CR is the “view that the behavior of the parts of a system (ultimately, the parts studied by subatomic particles) is determinative of the behavior of all higher-level entities” (Murphy). CR does not say anything about whether there may or may not be causes from outside a system. It is a thesis about causal relations between wholes or collections and their parts or elements. Thus it is incorrect to infer that everything human beings do and say is simply the consequence of those particles.

This point about language is significant. It is not that there was merely one word too many. Which words are used in an argument makes a difference to the apparent plausibility of the conclusion. The force of the following arguments are not equal:

1. Everything human beings do is simply the consequence of actions and reactions of elementary particles.
∴  2. Human beings are not responsible.

*1. Everything human beings do is the consequence of actions and reactions of elementary particles.
∴  2. Human beings are not responsible.

Both arguments share the same conclusion; their premises differ only by the presence of the word “simply”. The inference is more convincing in the first argument. Aside, perhaps, from their order in text, the only reason for the difference in force is the presence of that word. It turns out that is so for good reason. It is part of our common conception of the world that we are engaging with our surroundings, forming beliefs from our interaction, making things happen in the world. So the common sense picture is that there is causation at the higher levels. This is consistent with CR; on CR it follows that which causal relations exist at the higher levels, what human beings do, is a function of the causal determination of the particles. This is consistent with the second but not the first argument because the word “simply” suggests that there is only bottom-up causation and no causal relations between the entities in the high levels.

There is an available move from 1* to 1, and hence from 1* to 2. The argument is that the determination by the elementary particles makes the higher level causes superfluous, and so they should not be admitted into our ontology. The causes of the particles fully explain the motion of the composing whole, and so the arm’s motion is fully explained by the particles that compose it. This argument circumscribes an argument for nihilism about composition, such that there are no arms.  But set that aside. What does it mean to say that the part’s causes fully explain the reaction of the whole? If it means that there are no other causes, that would be blatantly question-begging as a premise in the move from 1* to 1. More plausibly, it means that for any region of space to be described, every truth can in principle be derived from a collection of truths about the parts. However, even that claim may be false. Facts about particles in a region may be contingently causally related to the whole in that region rather than to a whole in another region. And so facts about the particles themselves would not be sufficient if that is true. And even if that were false, there is a sense in which the facts would not fully explain that relation between parts and wholes.

Note as well that even if CR is false, there may still be determination at the higher levels (neurophysiological determination, for example).

I have been focusing on Hasker’s first argument related to responsibility and freedom. What I’ve argued so far is that Hasker’s argument get its appeal from some loaded language. Notice how this continues in what he infers:

“It is clear, then, we have no real choice about what we do; even if we have the experience of choosing, how that choice comes out is wholly determined by the actions and reactions of the fundamental particles.”

That might be right: that if our choices are “simply the consequence of the actions and reactions of the elementary particles of matter, operating according to the fundamental laws of physics,” then we have no real choice. And if no real choice then not free. And if not free then not responsible.  The only potential flaw here is about what counts as “real choice”, and experimental studies about the concept of choice are mixed about the need for alternatives. But to say all this follows from CR, that the behavior of the parts of a system is determinative of the behavior of all higher-level entities, is implausible for the reasons sketched. (By the way, I do not accept CR. More on that shortly.)

Argument from Reason

The second argument Hasker raises to causal reductionism is a version of the argument from reason. You can see this sort of argument showing up in the work of Victor Reppert, who, I believe, would be pretty happy with Hasker’s arguments against causal reductionism. If the fundamental causes of all that we do and think are nonrational, not in any way aimed at truth, then one may be skeptical about whether our beliefs are justified and whether the inferences are rational. It is not that the inferences are irrational because they are determined – that was the point of the more junior version of the argument dismissed above – but rather because they are determined by things that are not, in virtue of their nature, aimed at truth.

Suppose this is a convincing argument against causal reductionism. On the face of it, the argument is pretty good in my opinion. If so, what follows is that we might lack justification for believing causal reductionism, for that belief is inferred from other beliefs. It wouldn’t follow that no belief is rational since basic beliefs might be properly basic and so rational.

The argument is flawed. Recall that Hasker writes, “we are determined to accept by the particles that make up our brains, acting and reacting according to the fundamental laws of physics. And those physical laws do not in any way have in view the goal of leading us to true conclusions about the world — rather they are nonteological; they can be represented by mathematical formulas that have no reference whatever to any human goals or objectives.” Is this a description of causal reductionism or is it a description of particles, namely that their nomological properties and and behaviors can be described nonteleologically? The latter, of course. Causal reductionism is consistent with a Newtonian designer who set things up – he so arranged and moved the particles – so that we would form rational inferences and true beliefs. This is consistent with parts of a system determining higher levels because the higher levels are part of the system, whereas the designer is not, or need not be, part of the system. A case where a designer is part of the system may be the demiurge of the Timaeus.

What Hasker has to have in mind is something like, how we make inferences and form true beliefs is wholly a matter of how the particles behave. But for that to be plausible, I think, requires (maybe) ontological reduction and also no deity, both of which Hasker and I reject; neither of us is a reductionist and we are both theists. So it seems that this argument against causal reductionism is not convincing because it ties metaphysical baggage to reductionism that it is no part of causal reducitonism. An interesting consequence here is that Hasker may have a plausible argument to be raised against a naturalist, but he has an unsuccessful argument against a theistic determinist.

Final Comments

Although I have defended causal reductionism here, I am not a causal reductionist. Causal reductionism seems to leave no room for miracles being injected into the system where the entry point of divine causality does not start at the particles. I know of no good reason God could not do this, and without very compelling reasons to think God does not limit himself to injecting causation starting at the bottom, it is better to not rule that out without evidence. So for that reason I do not endorse causal reductionism (although this leaves open its truth).

The other reason I am not a causal reductionist suggests that causal reductionism is false. Causal reductionism entails at least that physical determinism is true. But whether physical determinism is true is partly a scientific question. As best as I can tell, the science does not indicate that physical determinism is true; it is probably false. The laws do not narrow down a single possible future.

Although I think physical determinism is false, I don’t think that all kinds of determinism are the same. Physical determinism does not entail theological determinism and vice versa; but they are compatible with each other. And when compatible with each other, there may be an order of dependence. For theism in the case where both obtain, physical determinism depends on theological determinism: God causes the world to be physically deterministic. In the case that God’s activity depends on physical determinism: the world causes God to be a certain way, although it is not clear what causal role God would have. You might think that Hasker’s argument is a powerful argument against deterministic naturalists. It turns out to be weak against Calvinists.

One would not be successful, then, in arguing against determinism in general from these arguments. For either the arguments commit causal reductionists (determinists) to more than they are in fact committed, or there are versions of determinism that are not committed to causal reductionism.


2018-06-07T21:30:20+00:00 March 10th, 2018|11 Comments


  1. Finney April 26, 2018 at 4:34 pm


    “It turns out to be weak against Calvinists.”

    I was hoping you’d flesh this conclusion a bit more. Does God “have in view the goal of leading us to true conclusions about the world?”

    I think there’s a philosophical argument regarding Calvinism responsibility and rationality that’s somewhat related to Hasker’s point. Supposing God decrees every event individually and all events together. Every event then supervenes upon one or more of God’s decrees, much like mental states upon brain states (in a non-reductive version of physicalism). Following Jaegwon Kim’s argument about mental supervenience applied to events and God, all events are epiphenomena.

  2. James A. Gibson April 26, 2018 at 5:19 pm

    What I probably had in mind when I said that – trying to recall my thoughts from March based on what I said here – is that Hasker’s argument works against as kind of determinist that does not think that the nature of things at the bottom level, and their relations to each other, have anything to do with being guided at the truth. The theistic determinist (e.g., the Calvinist) does not have to grant that. The Calvinist (or what is not the same, the deist) can say that what happens at the bottom is setup to bring about agents with faculties which are aimed the truth, at least more often than not. It is probably wrong to say that particles are aiming or are aimed at the truth. That sound too Aristotelian about physics to me, but I’m open to being convinced otherwise. (Last I saw, Rob Koons was part of some collection about neo-aristotelian perspectives about science. I suppose I should that.. after the million other items on my backlog, sadly.) I guess if it is in the nature of particles to be such that they interact so that agents form true beliefs, then I don’t think Hasker’s argument has any bite. But because I doubt that is so about the nature of the things at the bottom, then either they are setup to interact to have that result (to some extent) or they are not. If there is nothing that can so set them up that way (as is so for the naturalist), enter Hasker’s objection. If there is something to so set them up that way (as is for the Calvinist), then Hasker’s argument doesn’t transfer over.

    I would have to see the details of the argument sketched from Kim to say too much. But I would be strongly disinclined to think that the events that result from the decree are epiphenomenal. What account of the metaphysics of causation from the divine decree can one use to make that case? Calvinists, e.g., Helm, think that the nature of God’s causal role is mysterious. It also seems that Heath White’s Calvinist view doesn’t have that implication. But again, that is probably saying too much because the details of that argument you are sketching will matter.

    Hope that clears up my thought.

  3. Finney April 26, 2018 at 6:06 pm

    I agree there’s a significant difference between the theological determinist’s world and materialists’s worldview. I’m going to table the challenge for the theological determinist’s view for the moment. It’s been a few years since I’ve read the Emergent Self.

    Re: Calvinism. Kim argues that if brain state X sufficiently causes mental state X’, and brain state Y sufficiently causes mental state Y’, then X’ cannot be a cause of Y’. Imagine two cars passing a traffic light, followed by their two shadows. The first car passes the traffic light, followed by the second. One might say that had the first car’s shadow not passed the traffic light, then the second car would not have followed the first car. But the first car’s shadow isn’t a cause of the second car’s passing the traffic light. The shadow supervenes upon the car. Likewise, each mental state is wholly determined by the content and qualities of the brain state. No mental state causes another mental state. This is Kim’s problem with non-reductive physicalism.

    Putting aside the mysterious nature of God’s causal decree, it’s at least causal in the counter-factual sense, right? God’s decree is a cause of X, Y or Z in the sense that had God not decreed X, Y or Z, they would not have occurred. That shouldn’t be controversial. Suppose God decrees that X at T1, then decrees that Y happen at T2, very soon afterward. For X to have caused or led to Y in the counter-factual sense, we would have to assert that had X not happened, Y would not have happened. But Y would have happened, since God decreed Y. Thus, X could not have caused Y.

    So basically, if every event is decreed by God, and his decree is a counter-factual cause of that event, then no event could cause any other event.

    Al-Ghazali most famously argued basically along these lines in the Middle Ages to reject Aristotelian causation in all its forms.

  4. James A. Gibson April 26, 2018 at 6:39 pm

    Nice. Can you tell me where Al-Ghazali gives that argument? Of course I know of him, but I don’t think that I’ve ever read him.

    Take this with a grain of salt because I’ve given this only a few minutes of thought. I’d object to a counterfactual analysis of causation. I think it is physically possible that there could be overdetermining causes – think, two snipers aiming at a politician pull the triggers so that the bullets hit simultaneously such that… you get the idea. So X causes Z and Y causes Z. X causes Z (by elimination). But it is false that had X not occurred, Z would not have occurred. And yet X causes Z.

    Maybe his argument doesn’t depend on that analysis of causation. Or does it? I shouldn’t say too much about arguments I haven’t given much thought to. Thanks for the pointer!

  5. James A. Gibson April 26, 2018 at 7:34 pm

    One more thing that just now occurs to me. It isn’t clear to me how the argument from Al-Ghazali is related to Kim’s argument. What’s the relationship if any between those two? (I also don’t think Kim’s concerns are exactly Hasker’s. Maybe you don’t agree? I was thinking of my post as only dealing with Hasker’s argument rather than trying to draw some grand conclusion.)

    I still am curious about the Al-Ghazali reference if you got it.

  6. Finney April 27, 2018 at 6:26 am

    Oh, I was just offering my version of a causation argument with respect to Calvinism, since it seems Hasker’s argument wouldn’t work.

    Perhaps I wasn’t clear about Al-Ghazali – he made an argument demonstrating that God is the cause of each successive event, and therefore there’s no causal link directly connecting events themselves. Al-Ghazali wrote this in the Incoherence of the Philosophers (http://www.muslimphilosophy.com/ir/tt/tt-ns.htm#d1). He didn’t say anything about mind but I think his argument could be basically upgraded and applied to that context.

    My understanding is that overdetermination shouldn’t happen all the time. Kim: “It is at best extremely odd to think that each and every bit of action we perform is overdetermined in virtue of having two distinct sufficient causes.” Also, when applied to Calvinism, what’s the analogue to the two snipers? That one sniper is shooting something, and God is shooting something at the same time? But wouldn’t the one sniper shooting be itself an event decreed by God? In that case, there’s no overdetermination. If not, overdetermination would mean there’s something beyond God’s decretive grasp. Putting it differently, assuming there are other causes of events in addition to God’s decreeing those events, then if God hadn’t decreed X, X may still happen. That seems to be contrary to the teaching of God’s sovereignty in Calvinist thought.

    This is my first time actually trying this argument out on paper. I might just post a blog.

  7. Finney April 27, 2018 at 6:29 am

    Now that I think of it though, wasn’t Hasker’s argument more about causation than the AFR and rationality? I thought his whole point was, even assuming brain states are mental states, causation could only occur under the description the mental states have qua brain states, not qua mental states, and therefore they’re excluded from causality. Maybe that’s how I jumped to this argument.

  8. James A. Gibson April 27, 2018 at 6:44 am

    You were pretty clear about Al-Ghazali’s argument being about the causal relation between two events. It still isn’t clear to me whether you think his argument depends on a counterfactual analysis of causation. The point about the snipers is merely a counterexample strategy. One person puts forward an argument that assumes as a premise something. The counterexample shows that premise is false. Argument fails. That’s all I was doing there. If you do write this up in a blog, shoot me an email with the link! I’m always interested in neat historical arguments like that.

    Anyway, think of this post is the very narrow context of defeating Hasker’s argument rather than establishing there could be no other successful objection. On that I guess you agree with me. The more general conclusion wouldn’t be warranted by the arguments I gave. The best one can do to justify that, I suppose, is put forward a theory of divine causation. I’m nowhere close to trying that one.

  9. James A. Gibson April 27, 2018 at 6:50 am

    It seems to me that the argument which I quote is a version of AFR. I would expect The Emergent Self to have all sorts of other arguments. But I have only read the parts related to moral responsibility and agency. I’m sure J Kim makes an appearance elsewhere, but he seems unimportant to the argument I discuss. So it seems to me. Cheers.

  10. Finney April 30, 2018 at 2:35 pm

    So I finally wrote a blog on it. Here’s the link to the argument. https://thoughtwordbreath.blogspot.com Let me know if you find holes. Re: Al-Ghazali’s argument, I’m not sure he meant to argue in terms of counter-factual causation but I’m not sure how else to conceptualize it – it seems since he’s reasoning that since the bush would have still burned even absent the flame, the flame was not the cause of the burn, so that sounds like counter-factual causation.

  11. James A. Gibson April 30, 2018 at 2:37 pm

    Thanks, I’ll try to take a look this week.

Comments are closed.