Causal Determinism and Going with the Flow

I have been suggesting that people don't think that mental states can just straightforwardly cause human action, but surely people do recognize that mental states play some role in our actions. So how exactly do people understand the relationship between our beliefs, desires, emotions, etc. and the actions we choose to perform?

To begin with, let's imagine an agent whose beliefs, desires and emotions are drawing her to perform some specific action. I want to suggest that people think it would be possible for this agent to proceed in either of two ways. 

 

One option would be just to passively allow herself to be controlled by her mental states. Drawing on previous work by Queens of the Stone Age (2003), we might refer to this first option as going with the flow. 

 

Another option would be to respond more actively. Instead of just allowing herself to be controlled by her mental states, she could resist the pull of these mental states and freely decide to do something else instead. 

 

Now, you might be thinking: "I can certainly imagine cases in which it feels like an agent is choosing that second option, but in such cases, the agent's actions are still caused by mental states. Basically, what is happening is that certain mental states are causing the agent to act in a way that seems to go against other mental states." Fair enough. But what I want to suggest is that this is not the way people ordinarily make sense of such cases. On the contrary, people's ordinary conception is that such cases involve an element that is fundamentally different on a metaphysical level. The agent, people believe, has a contracausal power to reject the direction her mental states are proposing and simply do something else instead. 

 

To test this hypothesis, Adam Bear and I ran a series of studies. Participants were introduced to a deterministic universe and were asked whether it would be possible in such a universe to perform certain actions. Note that we are not asking here about whether it would be possible to be morally responsible for those actions or to perform them freely. We are asking whether it would even be possible to perform them at all.

 

People thought that many things would still be possible if the universe were deterministic. They thought it would be possible to ride bicycles, fall in love, compose complex pieces of music, appreciate the beauty of a sunset.

 

However, people actually thought that certain actions would be impossible in a deterministic universe. They thought it would be impossible to do things that involved using self-control to go against one's desires or appetites (e.g., resisting the urge to eat a steak). In addition, they thought it would be impossible to do things involved breaking off spontaneously from everyday rules or habits (e.g., impulsively shoplifting from a convenience store).

 

To figure out what was driving these responses, we recruited another group of participants and asked them to rate the degree to which each of these actions involved 'going with the flow.'

 

Sometimes people just passively go with the flow, while at other times, they seem to be actively breaking off to do something different.

 

For each of the following behaviors, we want to know whether you agree that people can only do it by actively breaking off from the flow of things to do something different.

 

Sure enough, there was a significant correlation. The more participants in this second group said it was possible to perform an action by 'going with the flow,' the more participants in the first group thought it would be possible to perform this action in a deterministic universe. 

 

All in all, these results suggest that people have a complex understanding of the relationship between mental states and human actions. An agent's mental states can draw her to perform certain actions, but people do not seem to think that mental states simply cause those actions. Instead, they think that the agent is always faced with a chose between (a) just going with the flow and (b) using her contracausal power to resist and do something else instead. 

Comments

[Matt King]

Joshua,

I'm trying to pair these results with the previous post's. If no mental states were relevant to what someone did (that is -- they had "no effect") then on what basis would anyone act?

Do you have some sense of what sort of explanation rationalizations could provide if not causal ones? I don't see any better options. We point to mental states as reasons not just to explain past behavior, but to predict future action, mind-read, etc. I find it hard to make sense of all these ordinary features without taking rationalizations to be some species of causal explanation.

Perhaps I'm just too embedded in orthodoxy, but I worry that on this interpretation, we get contracausal-ness without action.

[Chandra Sripada]

Hi Josh, I've been following all these fascinating posts, but because of travel, haven't had a good chance to chime in. But just wanted to let you know I am really enjoying seeing your research distilled into these juicy pieces so we can see the connections.

Let's suppose that you are right and people think the self stands outside the causal order -- it gets to ratify or resist motives without itself being subject to determination.

But surely you also agree that this idea of a contra-causal "willing" self is likely just plain wrong, or at least the folk impression that it exists can be readily explained away. Our best cog sci tells us that when a person resists wayward motives, there are in fact a slew of motivational/decisional factors that cause and fully explain why they resisted. These factors, however, operate largely outside conscious awareness. So it is not at all surprising that ordinary people "mythologize" a bit -- they think there is some spooky contra-causal source of their willings when there is none.

I am curious to know what you think about the tension between the folk view you are uncovering through your experiments and the actual findings from cognitive science. Do you think, for example,: 1) the folk view should be abandoned; 2) cog sci might yet vindicate the folk view; 3) it doesn't matter because probing the folk view is interesting in its own right; 4) something else??

[Joshua Knobe]

Hi Matt,

I'm not sure if you will find this helpful, but I think that these results point to a particular picture of the way people ordinarily understand the way mental states explain human action. Consider a case in which an agent eats a steak because she has an urge to do so. The thought is that the agent's mental state is drawing her to perform a particular action. She could have used her capacity for contracausal freedom to go against what this mental state is drawing her to do, but in this case, she does not use that capacity, simply allowing herself to be controlled by her mental states.

In other words, in the cases where the agent's mental state explains her action, people do not seem to be thinking that there is any necessary role for a capacity for contracausal freedom. However, even in those cases, people do not see the mental state as straightforwardly *causing* the action. The reason for this, it seems, is that they think the agent could have used her capacity for contracausal freedom to go against what her mental states were drawing her to do and just to do something else instead.

[Joshua Knobe]

Hi Chandra,

Wonderful to see you here! Your work on the self has obviously been a huge influence on all of the recent papers discussing these issues.

I completely agree with the point you are making here. So if the claim we make in these papers is correct, people ordinarily understand free human action as involving an agent that stands outside the causal order, but if the point you make in this recent comment is correct, contemporary research in cognitive science suggests that nothing like that actually exists in real human decision-making.

The philosophical upshot of all of this lies in the tension it suggests between our ordinary understanding of free will and the picture coming out of cognitive science. Of course, there are difficult philosophical questions about what to infer from that tension, but just as you suggest in your comment, it seems like this is the right place to be looking if we want to know what is philosophically important about these results.

[Thomas Nadelhoffer]

Josh,

Thanks for this post (as well as all of the others). As always, you are doing fascinating work. Unsurprisingly, I like these latest studies as they suggest that folk intuitions are in line with their theories--that is, something like folk incompatibilism (of some sort of another) is the default view. That doesn't mean it's the right view, normatively speaking, but you have provided further evidence that people have a conception of agency that requires a kind of contra-causal power to do otherwise (which in turn requires indeterminism). Notice this is embedded in the way we talk about mental states. My mental states are mine (at least the one's I "claim"). They are things I experience, resist, give in to, explore, think about, act upon, etc. But I stand above them nonetheless. I am the ultimate arbiter. As such, I can't be a mental state without being my own mental state. The dualism between me and my mental states is hard to eliminate from how we think and talk about mental states (or so it seems to me).

To answer Chandra's questions:

1) Whether one thinks we should abandon the folk view depends on (a) whether we think it's the right view, (b) what we take the consequences of abandonment to be, (c) whether the view can be abandoned "in the real world" rather than "in the armchair," and (d) whether one thinks cognitive science could *establish* that then kind of folk incompatibilism and dualism is false (i.e., whether this is an empirical issue). I don't see why the onus falls on Josh to answer these questions rather than those who are committed to revisionist theories, folk compatibilism, etc.

2) What would it look like for cognitive science to show that I do not stand above my mental states? A related question: What would it look like for cognitive science to show that I do not have an immaterial soul? I am not saying we have an immaterial soul, of course, I am simply highlighting the problematic assumption that seems embedded in this question.

3) Let's assume, for the sake of argument, that the gathering evidence on folk intuitions (e.g., the work by Josh) and the work on folk theories (e.g., our work on the scale), shows that most people are operating with some kind of latent indeterminism and contra-causal conception of the soul. You ask whether this is interesting in it's own right. I take it you mean to ask whether it is *philosophically* interesting (as it seems to be quite obviously interesting from the standpoint of psychology). This will depend on one's philosophical projects. For those who have claimed that incompatibilism is the folk view, these findings are philosophically probative. For those who think we ought to analyzing ordinary concepts, these findings are relevant. To those who are interesting in applied and public policy issues--e.g., how conceptions of free will and the self operate in the law--these findings will be probative. For those who are not interested in any of these approaches or views, the findings may not be relevant. And so it goes with the relevance of empirical evidence to philosophy more generally. Its relevance will depend on one's philosophical predilections.

[Joshua Knobe]

Hi Thomas,

Thanks for chiming in! These thoughts really help a lot in connecting all the nitty-gritty details of the experimental studies back to the larger philosophical issues.

There is definitely a lot to chew on in this comment (and maybe Chandra will have some thoughts on it), but I wanted to especially highlight your point 1c. Just as you suggest, it is really important to ask whether it actually would be possible for human beings to abandon this folk conception and, if so, what the result would be. I don't have any answer to this question, but it is certainly an exciting avenue for further research!

[Angra Mainyu]

Hi Joshua and Thomas,

Thanks for posting, Joshua.
I have a comment/question about Thomas's point that "For those who have claimed that incompatibilism is the folk view, these findings are philosophically probative.", and one about how to interpret what most people are saying.
While your experiments provide significant support for the view that most of the folk believe in some sort of contracausal free will, human action, etc., I wonder whether that's also significant support for the view that the folk, or most of them, have concepts of free will, human action, etc. (not just beliefs about them), that are not compatible with determinism.
For example, the experiments suggest most people believe in an deterministic universe, things would be far different from what they are in the actual universe (i.e., people would behave very differently), but their beliefs might not be significant evidence about the metaphysical commitments of the concepts that they use. For instance, most people probably believe that time is absolute, but I don't think experimental evidence of that would constitute significant support for the view that folk concepts of time are incompatible with relativity (and so, that an error theory of folk talk involving time, past and present events is true if relativity is true). It would constitute a bit of evidence, but I think not much. I'd like to ask what you think about the strength of the evidence with regard to concepts of human action, freedom, etc.

With regard to the replies that it's not possible in a deterministic universe to, say, suddenly change their mind about what they're planning to do (and similar ones), I think the most probable interpretation of those replies is that such behavior would not happen, and that as a consequence, such a universe would be very different from ours even with regard to human observable behavior, human beliefs, etc.
If that interpretation of the answers is correct, those who give those answers are clearly mistaken: as long as we accept (even if just for the sake of the argument) that deterministic universes are possible, of course human behavior could be just the same as in our universe: one just needs to pick the right set of deterministic rules. It might be very complicated, but it seems clearly possible (and regardless of whether most people have compatibilist or incompatibilist concepts).
An alternative interpretation of their replies (which I think is less probable, but haven't ruled out) would be that in a deterministic universe everyone might feel, behave, etc., as in our universe, but an instance of apparently suddenly changing one's mind (even believing one has suddenly changed one's mind) would not be an instance of actually suddenly changing one's mind (because somehow it's not sudden if it was already caused long ago). Or maybe different people meant different things when they replied that those actions weren't possible. I'd like to ask whether you know of some experimental result that rules out such alternatives.

Regarding point 1.c., it seems to me it's very probably psychologically possible to abandon the most common folk beliefs about human action, given that there is a non-negligible percentage of people who do not appear to have those beliefs, going by the results of the experiments (I can just point out that I don't have such beliefs). Maybe it's not doable for some (many) specific persons to abandon such beliefs, but I don't see an obstacle for (perhaps distant) future generations to grow up with beliefs more like those of the minority. Or do you think that the minority who responds differently have some of those beliefs about contracausal stuff nonetheless?

[Joshua Knobe]

Hi Angra,

These are all great points. I'll come back later to the more empirical questions you raise. For now, I just want to focus on the more purely philosophical question at the beginning of your comment.

Just as you say, these results provide some information about what ordinary folks think is going on in free human action, but they don't tell us anything about whether people think that free action would be possible without some of the properties it presently has. An interesting philosophical question then arises as to whether these results are relevant to discussions about the compatibility of free will and determinism.

I don't know whether you would find this helpful, but one might consider an analogy with other kinds of cases, i.e., ones that arise outside the context of the free will debate. Suppose that physicists posit the existence of a particle they call a 'bosin.' Now suppose that subsequent work shows that there is no particle that has all of the properties originally believed to hold of bosins but that there is a particle with most of those properties. Philosophers then begin wondering whether this actual particle counts as a bosin.

One possible view would be that a central aspect of this task would be to look in depth at precisely what physicists thought bosins would be like and then ask about the degree to which the thing that actually exists has a relatively large number of those properties. (This is, for example, Lewis's view.) If you think that approach is probably on the right track, you might think that the present studies shed at least some light on these issues.

[Eddy Nahmias]

So much interesting stuff to talk about here (sorry I've been preoccupied with less interesting stuff), but can I just pick up on a question Angra is pointing towards (one that is usually discussed 'in the other direction')--namely, what is the error theory for seemingly compatibilist intuitions or responses? If it's being suggested that the contracausal understanding of (some) human actions is a deep feature of our psychology (is it?), then are people who don't respond accordingly misunderstanding the implications of determinism? (This seems to be what some people say about the results I've reported, e.g., suggesting that most people do not see a conflict between free will and perfect prediction of decisions based on prior brain activity.)

And just as I get asked whether my error theories for incompatibilist intuitions can explain what incompatibilist philosophers think, what is the explanation for compatibilists who report no experience of, or beliefs about, having contracausal powers?

If the intuitions (or experiences) of contracausal power are not so deep, then it seems more plausible that people who have them would 'flip flop' to compatibilism if convinced of determinism (or naturalism, or lack of contracausal powers) and could be led to do so without too much gnashing of teeth.

[Joshua Knobe]

Hi Eddy,

Awesome questions! You are totally right to say that the questions you pose here seem to be kind of flipping things around and looking at them from the opposite perspective, but that strikes me as exactly the right way to go.

I wonder what you would think of the following hypothesis. Maybe people have an implicit conception of agency that they ordinarily use to make sense of human actions, but some people also have an explicit view that they have developed through more conscious reflection. Then it could be that people's implicit conception has certain libertarian elements but that some people nonetheless reject libertarianism at an explicit level.

Also, I just wanted to say that Angra makes a helpful point above that might be good to keep in mind here. These studies suggest that people think of human beings as having a contracausal power, but the results don't specifically suggest that people think it would not be possible to have free will without that power.

[Chandra Sripada]

Thanks Thomas for these elaborations. I especially agree with your point 1 that lays out some important considerations when considering a full on error theory.

You asked what would it look like for cognitive science to show that I do not stand above my mental states or have an immaterial soul. I agree it would be hard to definitively show *that*. But I had in mind something weaker. We have a pretty good handle of what happens when someone resists an urge in terms of specific computational/psychological processes: reasoning processes, evaluative inputs, inhibitory control circuits, etc. And these seem quite sufficient to fully explain why a person resists when they do. Given we have the makings of a complete and detailed materialist explanation that fully captures the data to be explained, the postulation of something immaterial that stands above and intervenes seems otiose. So that is the sense in which our best view from cog sci points to falsity of the folk contracausal view.

[Paul Torek]

Matt, Angra, Joshua, Eddy, Chandra -

Your intriguing discussion brings me back to something I wrote in comments on the last post, about mental causes' insufficiency. So suppose I, like Joshua's steak-hungry woman, also crave steak. This mental state makes it likely that I will eat steak. But I resist, and my resistance cannot be explained by another mental state in any reasonably narrow sense of "mental state" ("state that explains later actions" does not count as reasonably narrow).

Is all this consistent with the denial of contra-causal agency? Yes, and moreover, it's probably at least sometimes true.

Peter Tse gave us plenty of evidence that the brain relies on chaotic processes, wherein subtle differences in timing and frequency of neural spikes can lead to dramatically different decisions and actions. It would be implausible to suppose that introspective and intuitive classifications of mental states ("hungry for steak", "not hungry for steak") always line up with these subtle physical differences. So, probably, mental states are insufficient to cause specific behaviors, at least some of the time. The full cause requires specification of physical states. Now, Peter holds that these physical states behave indeterministically, and lots of physicists agree; but that part is dispensable (and disputed). If we remain agnostic about determinism (as any non-physics-expert, at least, should) then we have to leave open the possibility of determinism without psychological-determinism.

But when I resist, is it really *I* that resist? Arguably, yes. The physical events involved are part of the normal, proper functioning of the decision-making brain. I have good reasons why I resisted the urge, even though someone with information only about those reasons and other (mental) motives would rationally bet against my resistance.

So, I claim, the evidence *favors* the man on the Clapham omnibus's claim that mental states "flow" in a certain direction but that he, himself, can and sometimes does resist. I now posit that most folks have the *theory* that the "I" that resists is a dualistic, contra-causal intervener. The theory is simple and elegant, at least if one doesn't know much neurology. It explains why mental states aren't sufficient (but as I've just argued, it's not the only explanation possible). It's taught in Sunday school, and implicit in movies like All Of Me, Heaven Can Wait, etc., etc. They may hold this in a Theory-Lite fashion a la Eddy, but regardless, it can explain why they think certain actions couldn't happen in a deterministic world.

[Angra Mainyu]

Joshua,

That's an interesting analogy. Thanks.
Before I address it, I would say that the results are relevant at least to some extent, because if most people had said that in a deterministic world, there would be free will, or the actions in question would be possible, etc., that would provide some evidence supporting the view that the concepts of free will, human action, etc., that they intuitively grasp is compatible with determinism. So, the fact that most of them say otherwise workds against a potential argument for compatibilism (an argument that might be available if the results had been different), and provides a little evidence in support of the view that the concepts of free will, human action, etc., that they intuitively grasp is not compatible with determinism. But still, it seems to me be that the evidence they provide for that incompatibility is very small, absent other factors.
Regarding the bosins, I think that approach (i.e., the number of predicted properties it has) might work in some cases, but it depends on other features of their concepts, and more precisely on how ontologically heavy the concept is, so to speak.
For example, let's say that some physicists see a "bump" in their data from the LHC that is not predicted by the standard model, and they decide whatever is causing the bump a 'bisin'.
They then go on to make predictions based on alternative models that are available. So, different physicists make different predictions, depending on the model they find more plausible, though all of them agree on at least a few of the predicted properties.
It turns out that the new physics required to account for the bump would be quite different from what the physicists knew before, and the particle that caused it does not have any of the properties the scientists predicted (except for the ones already measured in the LHC). In that case, the particle is a bisin, it seems to me.
One question here is whether the case of, say, suddenly changing one's mind is more like a bosin or a bisin, in the relevant respect, and at least in the way most people use the expression. Or maybe it's neither.
At any rate, those results are very interesting, so I look forward to more experiments.

[Chandra Sripada]

Thanks Paul for reminding us that cog sci might very well vindicate the folk view (as Josh has characterized it). This was option 2 on my list and I kind of did assume that this option was deeply implausible. I tend to take a dim view view of the idea that stochasticity during decision might be harnessed to deliver genuine libertarion freedom (for reasons that many on this blog are familiar with). But then again I am not aware of the details of all the views out here (such as Peter Tse's). Would be delicious if the pointy heads all got it wrong and the ordinary shmoe was all along right on!

[Paul Torek]

Chandra,

Just to clarify, I'm with Dickinson Miller (aka Hobart) - the real one, not Wikipedia's version - on the undesirability of chance. A little bit of chance, however, is tolerable. And the chance in question is chaos, not necessarily true indeterminism. Psychological determinism is hard to reconcile with neurology, but physical determinism seems untouched, given a suitable interpretation of QM.