Deep Dive 1.2: Considering “Determinism is Self-refuting”

Among the many reasons why digging deep into this topic may be fruitless, the most prescient one is the possibility that determinism is self-defeating. If it is self-defeating, there would be few reasons to venture further. At a recent late-night chat, I got into a conversation that became immediately stymied because, with each point that I tried to make in favor of determinism, my assertion of any argument in favor of determinism would defeat its very thesis. (While it might be difficult to imagine a seemingly more unproductive conversation, I can easily recall the endless numbers of static conversations I’ve had with Christians.) Regardless of the frustrating nature of the conversation, it seems necessary to take the argument seriously since it damns the thesis. The first two articles considering the self-defeating thesis that I found, written by Churchland and Popper respectively, were a response to the refutation (by “the refutation” I mean the argument refuting determinism by calling it self-defeating) and a further response by an original purveyor of the argument. Since the fruitfulness of their back and forth amounted only to an amusing repartee, I searched for a stronger formulation of the argument. I found a presentation of the refutation by Joseph M. Boyle who attempts to lay out the clear reasoning of “the refutation” without simply asserting the refutation as prima facie obvious.

Boyle notes the determinism thesis is not obviously self-refuting, but he provides fairly decent justification which requires a response from advocates of determinism. He correctly notes that the general argument for determinism’s self-refutation begs the question by assuming that “rational assent is incompatible with causal determination.”[1] He notes further that some rational assertions do not “involve any choice.”[2] One clear example he provides for this case considers judgments about the manner with which we accept what we see as part of our immediate experience. While, epistemologically, I may not be able to provide a justification as for whether or not what I see is true (an account of an objective reality outside of my perception), an assertion containing the facts of my immediate experience is not something that I am in a position to deny. In short, I cannot deny that my current immediate experience includes a laptop on which these words are being typed. More specifically, I cannot choose to deny it. Certainly, it might be possible that I say that there is no laptop, but I would be lying to myself. Another potent example might be considerations regarding one’s taste buds. When I drink this expertly brewed cup of coffee from my favorite shop, I feel comfortably warm and the brew tingles on my tongue in a pleasurable way. I can not deny this sensory experience without lying to myself, or convince myself that I do not like the warmth or the taste. It is this type of evidence, he asserts, that the determinist could rely on to prove his thesis because, as he writes, “[determinism] is proposed as based upon–and required by–the evidence. In general, philosophical positions are not things we choose to accept or reject, but are propositions which the evidence and argumentation lead us—in the ideal case, inevitably—to accept.”[3] This understanding leads him to assert that for the refutation of determinism to succeed, the argument must be developed upon the reasoning that:

“[A]sserting propositions is an inherently normative thing; it is not simply believing a proposition, but being convinced that one is justified in believing it, [sic] and that others ought to accept it as true. Since there is a kind of norm which seems to imply the capacity for free choice, if the norm involved in the assertion of determinism is a norm of this kind, then determinism could not be asserted without invoking a norm that implied the capacity for free choice.”[4]

To understand the above proposition better, it will be necessary to explain what Boyle means by “free choice.”

According to Boyle, free choice is “radical”[5], separate from all relevant causal factors[6], and is akin to the type of free choice moral responsibilities asserted by biblical traditions[7]. He also defines “the notion of self-refuting” as “[arising] when these propositions go wrong in virtue of the incompatibility with the aspects of their own performance to which they refer.”[8] With these assertions and definitions in mind, he asserts how one might prove determinism to be self-defeating. Boyle concludes that the affirmation of “No Free Choice” (he calls it Nfc) requires an unconditional character, and this unconditional character damns the Nfc thesis because “it is proposed as reasonable and as something any reasonable person should accept.”[9] He further argues that “the affirmation of Nfc is rational and justifiable only by virtue of arguments which support it.”[10] Ultimately, he concludes that the norms required to assert determinism would be null if the thesis of determinism were true and that any attempt to urge others to accept it should be given up.[11] So, it seems I ought to give up my erstwhile attempt of convincing others to accept the determinist thesis. However, I have two possible responses to his argument: first, I contend that his definition of “free choice” is incoherent and second, I argue that the norms he states may not be as necessary as he asserts in order to argue affirmatively for determinism.

As noted above, Boyle argues for a “radical” perception of “free choice” for which others have argued is incoherent. If such choice is divorced from all relevant causal factors—nature and nurture—then choice is at best random. Certainly, any choice one makes include, necessarily, the person’s experiences, interpretations of the real, and genetic dispositions. Further, to avoid this type of random chance, his type of choice requires a type of dualism. One must necessarily assert that something independent of the physical causes in the brain is the origin of these free choices; otherwise, these choices are nothing other than a Lucretian swerve. So either choice derives from individuals in a causal manner through genetic dispositions and experiences OR they are separate from such causes and therefore random. Boyle cannot have it both ways. Consider a variation of Nagel’s example of possible dessert choices at the end of a cafeteria line consisting of a slice of chocolate cake or an apple. Whichever one’s choice, it seems dependent upon at least two factors: the individual’s dispositions towards the cake or fruit which she did not choose and her previous experiences with those foods. If the individual were to separate out all genetic dispositions and experiences, the choice would then be random, and whichever one she chooses would only be a matter of luck. Ultimately, his definition of choice does not seem sufficient enough to be deemed “free.”

As per my second response, it may be sufficient to say, that Boyles assumption rests in the empirical fact as to whether people possess the ability to determine what propositions to believe and which propositions seem plausible to them. In my discussion with the proponent of the self-defeating argument that I mentioned earlier, I attempted to assert that I remain unsure as to why I find some arguments convincing and others lacking. At a fundamental level, certain assumptions about the world must be accepted as bedrock truths a priori. If not, I’d be locked in an ever regressive need of justification. (An argument that evidence is useful cannot be proven with evidence. If someone denies that evidence is useful, no evidence exists that would prove its utility.) Sam Harris comments on this as well in one blog post arguing that “comments of this kind…suggest that people think they have control over what they believe…as if the experience of being convinced by evidence and argument were voluntary.” Boyle’s argument fails because it relies on an assumption that people arrive at their beliefs through some kind of out of body, objective experience, where arguments are weighed and conclusions are reached solely on their respective objective merits and that they have control over their mechanisms of belief. However, this is not the manner in which people accept propositions as beliefs. And even though it could be the case that some individuals would be incapable of accepting the thesis of determinism, it would not negate the facts of the determinism thesis. He seems to require determinism to fight in an arena where no other argument must battle. While claiming to escape the problems faced by initial counters to the refutation, he only really generates limitations based on norms that have little bearing on the thesis of determinism and that require unproven metaphysical presuppositions.

Ultimately, I think that, at best, his approach might provide an epistemological problem for determinism rather than a metaphysical one. To call it self-refuting is to really only say that it is impossible for one to assert positively the thesis of determinism rather than to say that determinism is impossible as a matter of fact. It could be the case that humans are determined creatures AND that humans cannot assert that they are determined creatures. Further, one wonders what Boyle would make of studies that reveal that humans often lack the ability to reason effectively. An interesting article by Elizabeth Kolbert published in the New Yorker discusses how rarely humans change their beliefs to align with the evidence. She reports that people often reject disconfirming evidence and welcome evidence that aligns with their views.  She discusses how “[t]housands of…experiments have confirmed…[that people can’t think straight]” by discussing the texts The Enigma of Reason, The Knowledge Illusion, and Denying to the Grave. While this fact is disconcerting, one possible method we have of combatting this, as Kolbert notes, is through the methodology of science, a system that attempts to get around cognitive biases and see facts clearly. The problem that ultimately confronts Boyle in light of this, is that science repeatedly points toward a determined universe and significant evidence would be needed to show how humans would be excluded from such causes.

Boyle, Joseph M. “Is Determinism Self-Refuting?” Edited by Steven J. Bartlett and Peter Suber. Self-Reference (Springer Netherlands) 21 (1987): 193-208.

[1] (Boyle 1987, 194)
[2] (Boyle 1987, 194)
[3] (Boyle 1987, 194)
[4] (Boyle 1987, 195)
[5] (Boyle 1987, 197)
[6] (Boyle 1987, 196)
[7] (Boyle 1987, 197)
[8] (Boyle 1987, 198)
[9] (Boyle 1987, 202)
[10] (Boyle 1987, 204)
[11] (Boyle 1987, 204)

4 thoughts on “Deep Dive 1.2: Considering “Determinism is Self-refuting”

  1. “Incoherent” suggests two ideas that contradict each other. The paradox arises from imaginary contradictions. The only thing that can contradict determinism is indeterminism. The only thing that can contradict free will is a will that is subject to (subjugated by) the will of another or some other undue influence (hypnosis, brain tumor, etc.).

    “Freedom from causation” is an oxymoron. Without reliable cause and effect, we could not reliably cause any effect. That is, we’d have no freedom to do anything at all! Therefore, any suggestion that to be “truly” free we must be free of reliable cause and effect would be irrational. Because “free” can never possibly mean that, it never does.

    Deterministic causation exists at three levels: physical (passive), biological (purposeful), and rational (deliberate). And it is probable that the causes of every event can be explained either by one of those levels or by interactions between them.

    Free will is normally associated with deliberate decisions, those that result from a mental process of deliberation and choosing. The most meaningful and relevant cause of these choices is the process of choosing, where multiple options are evaluated by some criteria and a single choice (or singular will) results.

    Choosing is a deterministic process. The result is causally determined by the chooser’s own purpose and the chooser’s own reasons. When this happens, we say the chooser freely determined his or her own will. But when forced at gunpoint to act against one’s own will, we say the will is not free.

    This operational definition of free will is commonly understood and correctly applied by just about everyone. Everyone recognizes coercion to be a meaningful and relevant constraint upon one’s will. Everyone recognizes that extraordinary influences, like hypnotic suggestion, or mental illness, may compromise one’s ability to make one’s own choices.

    But causal necessity is not a meaningful constraint. What we will inevitably do is “what we would have done anyway”. It is exactly identical to us just being us, doing what we do, and choosing what we choose. This is not something that we can or should be “free of”.

    So, once we get determinism properly defined (the belief that objects and forces behave reliably) and once we get free will properly defined (the ability to decide for ourselves what we will do), there is no conflict. We have two separate concepts, each expressing its own truth about the world.

    When I decide for myself what I will do, according to my own purpose and my own reasons, then it is authentically me that is controlling the choice. There is no conflict between (a) the fact that it was indeed me that was the final cause and (b) that my choice was theoretically predictable by anyone with sufficient knowledge of how I think and feel.

    The only reason you are getting resistance to determinism is because you are insisting, incoherently, that determinism conflicts with free will. It does not.


    1. Thank you for your reply, Mr. Edwards. I appreciate your position, and I think we agree on much more than you seem to let on in your post. I think we differ on our definition of free will because I define it as the ability to have done otherwise. You seem to be more willing to accept a compatibilist definition. You may be right when you suggest that most people define free will in the way that you do. I think Neil Levy has argued that his studies mete that out. Anecdotally, I have experienced that people take free will to mean that they could have done otherwise in a given situation. My experience is not scientifically relevant, and all I can say is that my post is in response to that definition of free will.

      I’m curious how far the rabbit hole you go down when you say: “Everyone recognizes that extraordinary influences, like hypnotic suggestion, or mental illness, may compromise one’s ability to make one’s own choices.” You seem to be accepting that undue influence on the brain can “take away” someone’s ability to make free choices, but I’m not sure what a reasonable limit to the undue influence is. If, for instance, I am inundated with commercials for Wendy’s cheeseburgers, and I decide, as I drive down the street to eat at Wendy’s while not consciously considering the commercials, did I make that choice freely? Had Hardees been more heavily advertised, might I not have “decided” to choose that burger joint instead. Ultimately, I’m not sure what the bright line is between “normally construed mentality” and “abnormally construed mentality.”

      You say towards the end: “When I decide for myself what I will do, according to my own purpose and my own reasons, then it is authentically me that is controlling the choice. There is no conflict between (a) the fact that it was indeed me that was the final cause and (b) that my choice was theoretically predictable by anyone with sufficient knowledge of how I think and feel.” I am not sure how to square this with those individuals you mentioned above; specifically, with the case of the person at gunpoint. Does not the man at gunpoint meet your criteria? After all, if he chooses to acquiesce to the gunman’s request, he does so for his own purpose (he wants to live) and for his own reasons (he thinks that doing what the gunman says will advance his purpose). Similarly, if he chooses to not follow the gunman’s orders, does he not do so for his own purpose (he wants to die) and for his own reasons (he thinks disobeying the gunman will advance his purpose). Even the case of mental illness may satisfy your criteria. A mental illness is necessarily part of a person’s identity (there is no them outside of their mental illness if I apply Leibniz’s Law), and therefore, it is their authentic self that makes their choices. It may be the case that asserting a “choice” is “free” as long as it derives from a person proves too much.

      As an aside, next Wednesday (May 24), I will be publishing a post evaluating an extended essay called Hard Luck by Neil Levy. He argues that it is not determinism that takes away our choice, but rather luck. I’d be interested to hear your thoughts on that post.


      1. Advertising would not be an “undue” influence. Everyone is exposed to it, and yet no one buys everything that is advertised (otherwise you’d be having a Wendy’s, a Burger King, AND a Big Mac for lunch). There is a category of advertising called “subliminal”, where images or auditory cues are embedded in an advertisement or program to prime the subconscious, but below conscious awareness. These are outlawed in the UK according to Wikipedia.

        As to mental illness, there are many varieties which would not compromise your ability to correctly decide between right and wrong. These are matters argued by competing experts during a trial. A person may experience a hallucination where they believe someone is attempting to kill them right now such that they kill the person in self-defense. Someone else may feel drawn to commit an act that they know to be illegal, and the issue would be whether it was an inclination they could resist rather than a true compulsion they could not.

        In any case, to the degree that the choice was a matter of deliberate decision, the correction would be via programs that address what went into that decision. To the degree that the choice was beyond the person’s control, they may be committed for mental therapy. In both cases, the issue is what corrective measures (if any) would be most effective.

        After the Boston Marathon bombing in 2013, the Tsarnaev brothers hijacked a car and forced the driver at gunpoint to assist in their escape. To correct the driver’s behavior, all that is required is to remove the gun from his head. To correct the surviving Tsarnaev, who had spent months convincing himself that blowing up the innocents watching the Marathon was a “good” thing to do, will likely take years, and that is assuming that he wanted to change.

        Most cases are clear cut, or already settled by precedents. A few cases may still present uncertain conditions that we must take our best guesses to work out.

        Causation itself is never an “undue” influence. Everyone understands things happen for reasons that we can often discover and do something about. Nothing can ever happen without causation, so while a specific cause may be extraordinary, causation itself is quite ordinary.

        And, of course, the person who deliberately chose to commit the crime could always have done otherwise if he had so chosen. 🙂 But that deserves its own comment.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s