Philosopher and author Julian Baggini, a nonbeliever who wrote the Very Short Introduction to Atheism for Oxford University Press (and about 20 other books), has a nice review in the Wall Street Journal of a new book on cognitive and behavioral autonomy, with a thesis that touches on issues of free will. The book is Freely Determined: What the New Psychology of Self Teaches Us About How to Live, by Kennon M. Sheldon.
You can read Baggini’s review by clicking on the screenshots below; if it’s paywalled, I’ll quote enough to give you the tenor of the piece:
Sheldon himself is a professor of psychological sciences at the University of Missouri in Columbia, and uses the book to push his own pet theory: Self-Determination Theory (SDT). SDT has been around for about fifty years, and is about people’s motivations and self-determination. Sheldon asserts that your wellbeing is higher if you think you are the agent who produces your own actions, and, apparently, that you have the ability to freely will your actions, or to will one of several possible actions. The latter, of course, bears on contracausal (“I-could-have-done-otherwise”) free will.
As Baggini notes:
Mr. Sheldon’s interest in free will is rooted in his work in Self-Determination Theory, which he calls “the world’s most comprehensive and best-supported theory of human motivation.” A core tenet of SDT is that “people need to experience themselves as the causal source and origin of their own behavior rather than feeling controlled and determined by external forces.” When people feel autonomous, they are more content and successful. When they feel they have no control, they become morally cynical. After all, if we’re not in control of what we do, how can we be blamed for wrongdoing?
Most of Mr. Sheldon’s 10 chapters constitute a compelling and clear introduction to what SDT teaches us about nurturing a sense of autonomy. The theory gives us a rich and powerful understanding of motivation—how to harness it and avoid undermining it. Most notably, the theory points to the importance of intrinsic motivation: the desire to do something for its own reward, not for any instrumental benefit.
And indeed, Sheldon may be right: we may do best if we feel that we are deciding our own actions rather than being compelled by the desires of other people or, ultimately, by forces beyond our control. If you’re a hard determinist, like me (and I think Baggini, though I’m not sure), you realize that we aren’t really able to decide one course of action versus another: that is decided for us by the laws of physics. Still, I have no beef with the idea that we feel better entertaining the notion that we can indeed choose one course rather than another. Indeed, natural selection may have favored, for various reasons that lack the will, time, and space to discuss, the feeling that we are making free choices. But what makes us feel good isn’t necessarily true; we can have our feeling of agency and feel better for it, even if that agency is illusory.
The problem, which Baggini homes in on, is that author Sheldon seems to think that SDT and contracausal free will are incompatible. That is, if you feel that you have agency, then it must be true that you have agency:
When it comes to the metaphysical realm, Mr. Sheldon’s mistake is to think that SDT and the philosophical denial of ultimate free will are incompatible. That is only true of the most popular, if simplistic, threat to his model of human freedom: the extreme reductionism claiming that reality can be completely described in the language of physics; that consciousness is just the humming of the neural machine; and that everything is strictly predetermined.
Mr. Sheldon sees off this crude challenge with skill and clarity. A key to his argument is the idea of a “grand hierarchy of human reality”—a scale of human understanding and its modes, from micro to macro. Physics sits at the bottom, with the sciences of chemistry, microbiology and neuroscience stacked above it. Every time you ascend a level you encounter reality at a different order of organization. As Mr. Sheldon writes: “There is a kind of ‘functional autonomy’ at each new level, which builds upon what is given below. This means that each new level affects the world in a way that is partially independent of the levels below.”
Ascend the hierarchy further and you get to the social sciences: varieties of psychology, sociology and anthropology. (He might have added philosophy.) You can’t understand human societies, he observes, without reference to their values, or human actions without desires and intentions. The reductionist assumption—that thoughts and feelings are causally inert—is invalid.
The issue is that evoking “reductionism” doesn’t touch the issue of libertarian free will. A “grand hierarchy” must still show that each level is compatible with the one below it, even if it couldn’t be predicted from the one below it. And at the bottom sit the hard laws of physics, which ramify upwards to produce psychology and anthropology. Just because you can’t predict how human societies work from the laws of physics doesn’t mean that those societies aren’t the ineluctable results of the laws of physics. It’s a fundamental error to deny reductionism just because we can’t predict how phenomena ramify. What would overturn reductionism is the observation that new phenomena arise at higher levels that aren’t COMPATIBLE with what’s going on at lower levels. And to a determinist, this just doesn’t happen.
And so, argues Baggini—and I agree—this palaver about the benefits of feeling empowered to decide (which is a real feeling and may be beneficial), combined with the denial of reductionism, leads Baggini not just to reject libertarian free will, but to ignore it completely.
This is all persuasive, but it leaves the deeper metaphysical problem of free will untouched. Human beings may make choices that are not predictable or even completely determined. [JAC: I presume that Baggini’s referring here to a fundamental indeterminacy, as far as we know, of quantum mechanics.] The hard question of free will is whether, at the time of making a choice, we could have done otherwise (leaving aside randomness or chance). The most popular position in philosophy today is compatibilism: It says that, although we can’t do other than what we do, we still have a valuable form of free will that allows us to maintain ideas of autonomy, control, responsibility and blame. In short, we may not be as free as we think we are, but we are free enough.
Note that Baggini explains that compatibilism accepts the fact that we lack contracausal free will. What is “compatible” in compatibilism is the absence of “true” (what I call “contracausal”) free will with another definition of free will that’s confected by whichever philosopher is pushing compatibilism. But Sheldon doesn’t even mention compatibilism, though he alludes to a form of it—a form that, to me, seems to deny contracausal free will entirely:
In “Freely Determined,” compatibilism doesn’t get a single mention. Instead Mr. Sheldon leans heavily on a recent book by Christian List (“Why Free Will Is Real”) in which it is argued that free will requires three capacities: considering the possibilities for action; forming an intention; and acting on a chosen possibility. But whereas Mr. List delves into the complexities behind these seemingly simple check-boxes, Mr. Sheldon merely helps himself to the comforting conclusion that, since human beings possess all three capacities, we are free.
In the end, Baggini recommends the book but criticizes the author for eluding a truth that really bothers people: our inability to decide or behave other than the way we do:
Fortunately, what Mr. Sheldon has to tell us about motivation and human action remains valuable, however we resolve the philosophical problem of free will. Readers will get a lot out of his book—as long as they recognize that it doesn’t so much solve the problem as deftly swerve around it.
Ironically, what first got me thinking about determism, and ultimately rejecting contracausal free will, was a paper by biochemist Anthony Cashmore in PNAS, written as a freebie when he was elected to the National Academy of Sciences. I think it’s well worth reading, and it has the word “swerve” in it, referring to the Lucretian swerve. Read “The Lucretian swerve: The biological basis of human behavior and the criminal justice system.” (It’s free online.)
I read the Cashmore article, which is very interesting, thanks for the link. I have a question regarding neural events leading to behaviors. Do you make a distinction between involuntary and voluntary neural systems? Fundamentally they play by the same physical/causal rules, and the behavior can even look the same. But under the hood there are differences, sometimes profound, in the neural systems implementing the behavior. Two examples are agonal respiration and respiration during sleep vs. voluntary respiration, and voluntary/involuntary smiling. I am curious to know how your categorization of these would relate to the notion of free will. I suppose you would say their difference is an illusion, fundamentally? I have trouble reconciling this with a notion of free will that applies to moral culpability. I would say the distinction is readily observable at the neural level (e.g., agonal breathing and breathing during sleep does not require cortical participation or consciousness), possibly at the behavioral level (e.g., “forced smiles”), irrelevant to experiencing them as conscious events (see breathing while asleep, which is involuntary, versus breath control during speech, which is voluntary but not necessarily conscious), and irrelevant to whether the fundamental units are voluntary or not (single neuron events are not voluntary). If you allow a distinction, isn’t this distinction sufficient for moral culpability? Cashmore falls woefully short here, devoting a paragraph to saying that we will make progress on vexing problems if we accept a lack of free will, but he doesn’t say how… And this sentence “in actuality we have no more free will than a bowl of sugar” strikes me as a little unserious seeing as organisms can produce actions, and sugar cannot. No bowl of sugar could ever plausibly be accused of voluntary manslaughter, and the whole issue of free will seems to assume the capacity to act or think. Thanks for your thoughts.
Cashmore is a brilliant understated writer. Introducing Cartesian dualism, he notes that “The mechanism by which this was achieved was, understandably, not understood, although Descartes offered some suggestions.”
Moral culpability works fine with pure determinism, since our moral systems are at root pragmatic. We evolved to dislike cheating and treachery, and thus to punish them, because (in our highly social and cooperative way of life) genes for disliking and punishing cheating and treachery outcompeted genes for not caring about them. Thence morality.
We can and do hold someone morally responsible for their conduct even if their conduct was completely entailed by the prior state of the system. The stuff about souls, and an uncaused “will”, is then a commentary about ourselves, a spiel that we have constructed, that hides the underlying truth of the matter. And philosophical attempts to reconcile that spiel with the underlying reality are misconceived and won’t work.
“We evolved to dislike cheating and treachery, and thus to punish them. . . .”
Coel, if our genes for disliking cheating and treachery were as successful at evolving as you suggest, there wouldn’t be any cheating and treachery to punish. 😊
See the concept of an “evolutionarily stable strategy”, that we’d evolve into a stable, tensioned equilibrium between cheating and attempts to clamp down on cheating.
An economical statement and compelling. Uncounted ‘spiels,’ a single story–mortality.
I agree with you that culpability works fine with determinism. I guess my confusion is that if I make a distinction between voluntary and involuntary actions, the “voluntary” is the same as free will for an organism. Yes, if you rewind the clock the events are fixed to be determined, but the world doesn’t work that way (you can’t actually rewind the clock). What evolved was an organism that can make voluntary actions because the future hasn’t happened yet. I am genuinely confused by this. It seems like the same argument to say neurons are not conscious, and if you reduce neural interactions to neurons, you don’t have consciousness, therefore consciousness is not real. But conscious and unconscious state differences are readily observable. And this is why reductionism doesn’t really work. If you wouldn’t mind enlightening me about how free will vs non-free will is different from the distinction between voluntary and involuntary neural systems, I would really appreciate it.
I don’t want to “feel better”. I want to know what the hell is going on. I’m not brave.
Just curious. Science’s very value is that it can confound our expectations, tell us what we don’t want to hear. That’s the fundamental problem in New Zealand. They want to be comforted. Reassured.
“ You can’t understand human societies, he observes, without reference to their values, or human actions without desires and intentions.”
But values, desires, and intentions are physically encoded in our brains, and are processed by our brains. If we could see precisely that coded information, the processing mechanism, the encoding of the result, and the physical ramifications (producing first-person experiences), we would be able to understand human actions.
Good that determinism is getting some airplay, even if it’s not always totally coherent. Robert Sapolsky is coming out with a book, working title last I heard: “Determined: The Science of Life without Free Will.”
“Baggini recommends [“Freely Determined”] but criticizes the author for eluding a truth that really bothers people: our inability to decide or behave other than the way we do.”
What people want in being able to do otherwise in actual situations (contra-causal free will) is to be little first causes that are not themselves determined by antecedent and current conditions, which allows placing a very strong sort of blame and taking a very strong form of credit. But of course there’s no observational or experimental evidence to suggest that we have some self-forming or behavior-initiating capacity independent of determining conditions. And were there such a contra-causal ability, its exercise would pose a mystery since the agent and their behavior wouldn’t be fully traceable to antecedents. Bottom line: indeterminism can’t add to one’s control or responsibility, so isn’t worth wanting, as Dennett would say. Good old cause-and-effect doesn’t disempower us but is rather the necessary basis for effective action.
Seeing that we *couldn’t* have done otherwise in actual situations (as opposed to counterfactual) can undercut assigning the strong sort of libertarian credit and blame that shapes so much of interpersonal and social policy, often in a very punitive direction. Which is why getting the word out about determinism (leavened by any indeterminism that might exist) and denying contra-causal free will is so important. Earlier this year Jerry posted positively about another piece Baggini wrote in Psyche defending compatibilism, but with all the necessary advisories about giving up libertarian free will and what that could entail for attitudes and social policies. Good stuff!
https://whyevolutionistrue.com/2022/05/19/julian-baggini-on-free-will/
I ‘feel’ like I have free will in the same sense that I “feel” that a table is an object. Intellectually I know it’s really just atoms no different to the atoms that surround it and atoms in turn are just excitations of quantum fields. However, for many purposes it is useful to consider a table as an object and my will as free.
What Sheldon calls “the grand hierarchy of human reality” corresponds to the worldview of ontological emergentism with its postulation of relatively autonomous levels or layers of being over and above the basic physical one, where new, physically irreducible (chemical, biological, psychological, or sociological) properties or powers occur that enable (chemical, biological, psychological, and sociological) wholes (complexes/systems) to influence the behavior of their parts (elements) by virtue of downward/top-down causation.
“…what first got me thinking about determism, and ultimately rejecting contracausal free will, was a paper by biochemist Anthony Cashmore in PNAS”
Fantastic paper. Thanks for sharing it. I loved the Einstein quote too.
Just this morning I happened to visualize our illusory free will as equivalent to our driving trains with a steering wheel, thinking that by turning the wheel clockwise, say, we make the train go right, wholly unaware of the rails that are in fact guiding the train. Nevertheless, like Cashmore, I find it hard to believe that our consciousness does not affect our behavior in any way (otherwise, why would it have been favored by natural selection? Huxley’s view seems too extreme), but, also like Cashmore, I don’t think there’s anything magical there (consciousness is just the result of the chemical activity in the brain), and certainly nothing that supports the notion of free will.
Even though his own physicalist account of action pretty much marginalizes consciousness, Cashmore feels duty bound to find something for it to do, since after all, it was naturally selected for, right? Perhaps not. What was selected for in evolution are the neural and functional correlates of consciousness, which seem to be necessary for the sort of cognition and action control that makes possible flexible and novel behavior at the human level (and at lower levels too, no doubt, like that of our primate cousins).
Consciousness itself may be a side-effect, for it isn’t at all clear that, considered as something non-identical to its neural correlates, it adds anything to cognition and action control. If consciousness is identical to its physical correlates, then it obviously has the same causal powers, so doesn’t add anything extra. But if it *isn’t* identical, then the problem of dualist mental causation has to be solved: how does something categorically non-physical feed into and influence the brain and body’s physical behavior-control network? There are no viable science-based solutions on offer. Cashmore says: “…there must be a mechanism by which consciousness does influence behavior. There must be a flow of information from consciousness to neural activity.” But he doesn’t suggest what it might be. This is best explained by the hypothesis that there is no such mechanism, and that consciousness doesn’t and can’t play a role in 3rd person explanations of action, strange as that might sound. About which see https://www.naturalism.org/philosophy/consciousness/respecting-privacy
The idea that consciousness is merely a side-effect would certainly eliminate the otherwise devilishly difficult problem of comprehending what exactly it adds. As you suggest, the neural and functional correlates of consciousness would seem enough to fully explain our behavior, from which it’d follow that our mental states are a curious epiphenomenon with no causal powers. On the other hand, it seems clear that evolution has “designed” our conscious states to operate in very specific ways, in ways that suggest that they have very specific functions and purposes. Think, for example, of driving your car in a nearly unconscious manner until a careless pedestrian suddenly “awakens” you and your full conscious attention seems to be called forth for you to handle the potential risk more deftly. We can think of numerous situations in which consciousness is activated (or not) to various degrees according to what appear to be very precise rules whose evolution would be hard to make sense of from the consciousness-has-no-purpose perspective.
As an aside, I find it tremendously difficult to think clearly about this. I think I’m hardly alone. Yet it seems like it shouldn’t be so puzzling. Natural selection may have seen to it that we be congenitally incapable of grasping this aspect of our nature (too keen an awareness of our being mere robots must be detrimental to our Darwinian fitness). The same seems to be true for most people as regards libertarian free will. As far as that is concerned, at least, I happen to see as immediately obvious that libertarian free will is an inherently incoherent notion—no science is needed to reject it, since basic reasoning should suffice. But people smarter than me see nothing incoherent in it, and posit that future advances in science might well reveal it to exist; they too, in my opinion, can’t escape the cognitive blinders that evolution seems to have firmly placed in our mind (in the minds of most people, at least) in that regard.
Everybody thinks that he is an agent who produces his own actions, everyone believes he has free will even if he knows free will is non-existent. The only practical difference we observe is that freewill-deniers are soft on crime for the right reason.
I don’t believe free will denying makes me unhappy; I think it makes me happier because I don’t have to worry much about shame and guilt. Free will deniers can live their lives without taking these negative feelings too seriously.
Well, I for one do not think that I have “free will” in the sense of an un-caused will.
I think that I arrive at a choices via the largely deterministic decision-making mechanism in my skull, akin to how a chess-playing computer arrives at choices. (Yes, I really do feel that way.)
And it makes entire sense to regard both my brain and the chess-playing computer as agents making choices, that’s what they indeed are (the fact that the choices derive from the prior state notwithstanding).
“ I don’t believe free will denying makes me unhappy; I think it makes me happier because I don’t have to worry much about shame and guilt.”
It is not exactly comforting to read that free will denial leads you to saying the above.
If you feel absolved of feeling guilt for things you have done, that logic is available for things you are about to do.
“I’m tempted to do X and I know it’s wrong but I also know I’m ultimately not responsible and know I can absolve myself of guilt once I’ve done it.”
Scientific theories are like butterfly nets, they are made to catch some aspects of reality, and they let other aspects escape through the mesh. Scientific theories strive to find predictable and deterministic behaviors, so they may easily overlook some minor aspects which don’t fit into that scheme. If we had asked scientists of late XIX century: “Does _real_ randomness exist in nature?”, they would have answered: “NO! Randomness doesn’t exist, cannot exist. All the laws of physics exclude randomness. If randomness existed, we would have discovered it in our experiments etc”. But when our investigation became more subtle, it turned out that true randomness exists. Now, possibly “infinite” layers of reality exist below our current level of understanding, so our view of the world may completely change.
It’s not wise to build our moral laws and rule our society according to our current understanding of physics. PS I’m a physicist.
I think the determinism argument boils down to whether or not the laws of physics are deterministic.
No it doesn’t, for even if some of the laws of physics are purely non-deterministic, that doesn’t give us the ability to make conscous, willed decisions. At best it adds a bit of unpredictability to what we do, but gives us no agency.
I agree with that. Free will is not determinism. If there is no determinism, you might act differently after a rewind, but that does not mean free will in the traditional sense (which is the will being an independent agent without an external cause).
Just to summarize my argument:
1) you find what you’re looking for; science looks for determinism, so it’s not strange that it tends to find determinism;
2) non-determinism seemed absolutely impossible to scientists of XIX century; now it’s accepted. In the same way, agency may be possible in the context of our future theories.
I’d like to add an observation. Quantum theory is our most fundamental theory, it describes the world at a microscopic scale. As a recent post on this site has discussed, the link between the microscopic theory and the macroscopic world is highly problematic. Wave function collapse is an unavoidable link in this chain, but it must be stressed that it is an “ad hoc” rule which CONTRADICTS the laws of quantum mechanics; according to quantum mechanics, the collapse shouldn’t happen. So we are applying a microscopic theory to our macroscopic world, going through a contradictory step. This is not very logical, in my view.
The only coherent, non-contradictory interpretation of quantum mechanics is Everett’s many-worlds interpretation (multiverse), which is not so popular among determinists (guess why?). The dichotomy between determinism and free will in a many-worlds context would certainly need a deep rethinking of the terms.
The problem is always the unknown. With the knowledge of the 19th scientists thought so-and-so. With our current knowledge, scientists think something else. Usually, you test a theory with an experiment.
Claiming that the world is fully deterministic is metaphysical speculation even when you have theories that make the assumption plausible because you do not know everything. I think you need evidence.
And I once noted jokingly, that if you want to prove that this universe is fully deterministic, you have to start the big bang again, and then wait 14 billion years to find out whether or not you are on the same place starting the big bang again.
But if the world is not deterministic, that does not mean that we have free will. I guess that is where the confusion comes from, most notably when the owner of the blog calls himself a ‘hard determinist’ but means that he does not believe in free will.
It’s not a blog but a website! Okay, let me say I’m a “hard materialist”. I’ve explained my position on this before: it’s unlikely the universe would repeat itself started at the same condition because quantum phenomena were in play, and well may be in evolution as well (e.g. in mutation).
“[N]on-determinism seemed absolutely impossible to scientists of XIX century…”
The idea of indeterminism (as a necessary condition of free agency) is very old. For example:
“In order to avoid fatalism, within his materialism and atomism, Epicurus posits an indeterministic atomic motion, the “swerve” (Fat. 22, LS 20E). As the Epicurean poet Lucretius puts it, if atomic motion were caused only by the weight of atoms and their collisions and entanglements, human beings, as well as all other animals, would be subject to necessity and unable to move themselves around as they please. But we obviously can move ourselves around as we please, and so there is another cause of atomic motion, an occasional minute swerve to the side that occurs “at no fixed region of space or fixed time” and liberates us from fate (DRN 2.251–93, LS 20F). Because determinism is false, the future is still open, and statements like “you will recover from an illness” are, at present, neither true nor false.”
Ancient Theories of Freedom and Determinism: https://plato.stanford.edu/entries/freedom-ancient/
Philosophers with false assumptions about time and causality think that scientific determinism conflicts with human abilities to select from multiple genuine options. Unfortunately, this includes some compatibilists, like Baggini. They falsely assume that in order to be able to do X, you must be able to do so holding fixed all microscopic details of the past. (Nobody thinks that you should hold all details about the future fixed, even if those are related by physical law to your action.) I explain why this is wrong, and how common sense ideas about causality fail to fit the scientific facts, on my website, which is mine. Hint, think relativity and thermodynamics, more than quantum.
Sheldon’s first sentence above is correct, but the second is not. Fortunately libertarian free will is an imaginary solution to an imaginary problem. And so is the kind of Strong Emergence (as philosophers call it) that Sheldon expresses in the second sentence.