In my experience, maybe 90% of the time when someone says they can prove something true with a logical syllogism, the syllogism they present is hopelessly fallacious. There seems to be a ubiquitous failure mode caused by a popular belief that syllogisms can prove things. Sure, in an extremely limited and narrow sense, they can; hence the illusion that seduces so many into thinking they always can, that this is how one proves things, which becomes the belief that they must do, which then inspires them to conjure something resembling a syllogism reinforcing their belief. They don’t stress-test the syllogism for fallacies. Often they don’t even test it for validity; but I’m here speaking even of syllogisms that at least have a valid structure.

The problem is that almost everything in the world is not a logically necessary truth but a question of empirical fact, and empirical conclusions can only be reached inductively, not deductively; yet the usual syllogisms people think to use are not inductive arguments. Deductive syllogisms can really only extract information from the premises; they cannot circularly prove the truth of those premises. Worse, as those premises often have a nonzero probability of being false, and standard deductive syllogisms aren’t designed to coherently commute these probabilities into the conclusion, you have to construct syllogisms as “if, then” statements (“if” the premises are true, “then” the conclusion is true), which only tells you that the conclusion is already inherent in the conjunction of the premises. In other words, you are just extracting the consequences of your own assumptions. You aren’t actually proving anything factual. That would require proving the premises true. And that’s a lot harder to do “with a syllogism.”

Most people don’t get how probabilistic or inductive reasoning even works. In a sense, the relative simplicity of deductive syllogisms seduces people into using them when they should be avoiding them as useless to the purpose. And so you get a lot of really terrible syllogisms, that people (even professional philosophers) believe “prove” what they purport. Some of the most common failures when doing this are equivocation fallacies (covertly switching the meaning of a key word somewhere between premise and conclusion), question-begging (assuming the conclusion true, just differently worded, in a premise), and false lemmas (claiming a list of options exhausts all possibilities, when in fact an important possibility—indeed often the one most likely to be true—has been omitted).

So anytime you see anyone claiming they can prove something with a syllogism, you are warranted in doubting it. Approach with extreme skepticism, and vet the hell out of the syllogism they present. Do the stress-test on it they probably didn’t. Verify the premises are all very likely to be true; that the meaning of words hasn’t changed from beginning to end; that a single premise isn’t simply assuming the conclusion; and that when options have been enumerated, none have been left out; and so on. Maybe you’ll find they’ve landed on a sound and valid syllogism, warranting belief in its conclusion. But usually you won’t.

Today I will use “deductive” arguments against the existence of free will as my example base. And for the following I will reference and rely on yesterday’s article where I laid out much of the groundwork, Free Will in the Real World … and Why It Matters, the gist of which is that in real world applications (whether a court of law or interpersonal relations or self-assessment), when people say “free will” they only mean “with informed consent,” not any kind of power to act uncaused.

Those who want to deny free will exists thus operate on a fundamental unspoken equivocation fallacy: that if they can prove “uncaused actions” don’t exist, they have proved “autonomy” and thus “consent” don’t exist. But autonomy and consent have nothing to do with acting uncaused. And “free will” in real world application only ever refers to the latter, not the former. This means free will deniers are like people insisting ferries don’t exist by proving faeries don’t exist. But in the real world, ferries are all we really care about, because they exist, we observe them, and we use them. Faeries have no effect on the world, and thus the concept of them has no use. Free will in the real-world sense exists—we can empirically observe when it exists or doesn’t, and we use it in countless practical applications. Free will in the ivory tower sense is even more useless and inapplicable to reality than faeries. So why is anyone still talking about the faeries? They should only be talking about the ferries.

Example 1: Jonathan Pearce

Atheist philosopher Jonathan MS Pearce has attempted his own “syllogism” to disprove the existence of free will (in A Syllogism for Determinism), and ironically, he says “the core to” his “disbelief” is “the philosophical incoherence of the idea of free will.” So he’s even an atheist because he falsely believes in the nonexistence of free will. To be fair, he has conceded that if we define free will as compatibilists do, then he agrees free will exists. So his error appears to be living in the ivory tower, and thus not noticing that in every real world application “free will” only means what the compatibilists note it means. There is therefore no use for the “ivory tower” definition. So why are we still talking about it?

Pearce might answer, “Because I need to debunk obscure, incoherent notions of God that depend on a non-existent idea of free will” (see God’s Own Free Will). But that’s not a good reason. If the problem is that someone has assigned to God an incoherent idea of free will, the solution is to point out that that idea of free will is incoherent and does not align with any practical application of the term—what is commonly called “libertarian free will,” or LFW, simply isn’t the sense of “free will” anyone is using in ordinary life, in courts of law, or any material context whatever. So the solution is not to say “free will doesn’t exist,” as that is as fallacious as saying “mammals don’t exist because faeries are mammals and faeries don’t exist.” The apologist has the wrong idea of free will, an idea completely divorced from reality, even from logic. That’s what we should be calling them out for. The dangers of doing it the other way around I already explained yesterday.

Of course, one could also take Pearce to task for an even more basic error here, which is that theism, even Christianity, in no way depends on that illogical notion of free will (as Christian apologists have already pointed out). So you actually can’t disprove it that way; all you can do with such an argument is reform its theology, leaving its purported truth otherwise intact. That’s not a sound way to end up an atheist. It’s also perplexing to see a philosopher argue “If in any meaningful sense I could not have done otherwise, then God punishing me eternally is rendered utterly incoherent.” Because “God punishing me eternally” is just as incoherent even if libertarian free will existed. So why does it even matter whether “LFW” existed? The problem is not LFW; it’s irrationally incommensurate punishments. Pearce thus seems obsessed with a complete irrelevancy.

But my concern today is with Pearce’s resort to “Argument by Syllogism,” in which Pearce says “what I really want to do is encapsulate these ideas into a robust syllogism” disproving free will, and then gives us this:

First:

(1) Every human choice or action is an event.

(2) Every event has its explanatory cause.

(3) Therefore, every human choice or action has its explanatory cause.

Building upon (3), we have our second syllogism:

(3) Every human choice or action has its explanatory cause.

(4) To have explanatory cause is not to be free.

(5) Therefore, human choice or action is not free.

Okay. So, he just declares “To have an explanatory cause is not to be free” but one of the very points at issue is whether “to have an explanatory cause is not to be free.” So he hasn’t actually proved anything; all he’s done is assert an unproven. But it gets worse: “Every event has its explanatory cause” renders his syllogism no better than a circular argument—because it simply presumes the very thing his opponent, a Christian advocate of LFW, denies. What good is that? His opponent will just say Premise (2) is false, that “some” events are uncaused, in particular human choices—and Pearce’s argument is cooked. Pearce cannot presume LFW nowhere exists in order to prove LFW nowhere exists. Pearce can’t recover from this failure with this or any syllogism. He is simply stuck having to prove there are no events in the entirely of existence that lack causes.

And that’s a tall order. It can’t be deductively done. There are serious physicists who even deny it—for example, many live options for interpreting quantum mechanics allow that some events, at some level of analysis, are uncaused. And that doesn’t even require speculating that human decisions are quantum events (though that’s logically possible, and thus not refutable by syllogism); the relevant point is that Premise (2) has never been proved true to enough certainty to maintain that there are no possible uncaused events. If the outcome of a quantum wave form collapse has no cause, what else in the universe might have no cause? Indeed, in a universe with gods and miracles in it, the odds actually go up that uncaused events might exist within it, because the possibilities for such a thing then substantially increase. If a God existed, we really could not say he is subject to our laws of neurophysics. He could be analogous to a quantum mechanism many of whose decisions are indeed uncaused.

Now, of course, I agree these things are all very unlikely. But I base that conclusion on inductive, empirical arguments—on evidence—not deductive syllogisms. What the brain does when making decisions simply does not track quantum wave form collapse, at all (except in ways unrelated to the point here). It manifestly involves classical macro-causation on fairly straightforward information processing models. People don’t just “randomly” do things. They always do things for reasons. And those reasons are always caused by other materially observable things: knowledge (information), a material personality (physical neurological thresholds and structures), skills (one’s information processing mechanisms, which all obey classical laws of computation, and are demonstrably physically manifest in their brain at far above quantum scales). So quantum free will is as fictional as faeries. Even more broadly, evidence well establishes supernatural alternatives are wildly implausible (see Naturalism Is Not an Axiom of the Sciences but a Conclusion of Them and The Argument from Specified Complexity against Supernaturalism).

But you have to argue to these conclusions with evidence. You can’t “prove” them with syllogisms; and you certainly can’t simply assert them as premises. That gets you nowhere. Pearce’s syllogism would never persuade any rational Christian to agree with his conclusion; and it should never do. It simply gainsays the Christian; it does not prove them wrong. The one thing in dispute remains completely unresolved: are human decisions uncaused events? Are they an exception to the rule Pearce presumes about events generally? His syllogism cannot answer that question. It therefore accomplishes nothing. So why did he think it would? I can only guess he must have been seduced by that argument form, then failed to stress-test his own application of it.

Missing the Ferries for the Faeries

A more productive route would be to grant the Christian’s premise of LFW, and then demonstrate that that logically entails a contradiction in the Christian’s worldview: LFW entails decisions are uncaused; if decisions are uncaused, they can’t have been caused by reasons; ergo God has never had any reasons to do anything he supposedly did. That’s a much bigger problem, because the Christian can’t escape that conclusion—not by any logically sound means anyway. Their only way out is to drop this nonsense about LFW (as many Christians do), or concede God’s decisions are all just random whims—they cannot even have been caused by his “good” or “just” nature, or his “intelligence” or “wisdom” or “omniscience” either, as any of that would then be a cause, thus negating LFW. Even if God only “sometimes” acts without reasons, those decisions are still reasonless; and the others are still caused and thus not products of LFW. Arguments like this demonstrate that LFW is simply internally incoherent. It does not track logically any practical notion of free will.

Consider for example: LFW entails decisions are uncaused; if decisions are uncaused, they can’t have been caused by anyone’s knowledge or intentions; ergo no one’s decision has ever been made in consequence of their knowledge and intent; but knowledge and intent are essential elements of guilt; ergo no one can ever be guilty of anything. So much for hell making sense.

This illustrates how incoherent and ill-thought-out Christian ideas of free will are. In the real world, it is only when knowledge and intent cause a decision that we deem a person to be acting with free will; and it is the absence of such causation that negates free will. Exactly the opposite of LFW. If I have an epileptic fit and push someone into the street, “I” did not cause that. It has causes (which bypassed “me,” my conscious assent—my knowledge and intent), but needn’t, as that’s irrelevant to the conclusion of my innocence. Even if epilepsy were the supernatural condition of “uncaused decisions” to move my body, this would negate my free will, not establish it. I only have free will when my will is caused, in particular by my knowledge and intentions—information and motivation—and not someone else’s knowledge and intentions, nor by some causal process that bypasses mine.

I cover the absolute necessity of decisions being caused in order to be free in detail in my section on free will in Sense and Goodness without God. Quite simply, LFW is fatally incoherent. It destroys any basis for responsibility in a foolish attempt to establish responsibility on some other grounds than what we all know to be true. It’s just irrational handwaving. A curtain hiding Oz.

Christians don’t want to face the fact that the only logically possible alternative to “is caused” is “is random,” and it can never make any sense to punish people for random events they didn’t cause—that no part of their apparatus of informed consent allowed to happen. Christians try to weasel out of this conundrum by claiming decisions are “partly” caused by reasons, by information and desires, but not “wholly,” but that is simply describing different degrees of the same defective apparatus: if I am going to act contrary to my knowledge and intent “at random,” for no reason, I cannot be responsible for that action—even if that bizarre random event happens only 80% of the time or 20% of the time, 100% of the times it occurs, “I” (the sum of my memories, desires, character) won’t have caused it, so I won’t be responsible for it. LFW thus reduces all human decisions to just another kind of epilepsy. Which renders LFW nonsensical. It’s useless theologically. It’s useless legally. It’s useless morally.

It doesn’t even work to try and say (as some do) that we have a desire for every possible outcome, and LFW selects at random (because that is what “uncaused” means) which desire causes our decision. That’s just a different kind of “spiritual epilepsy,” where even my desires are wholly uncaused by my character, my reason, or any kind of information or proclivity you could identify as distinctly “me” so as to hold “me” responsible for it. If actions are simply the product of a “desire roulette” in our brains the outcome of which we, as persons, have no causal control over—that our tendering of consent makes no causal difference to—then we could not be held responsible for any of our actions. And this is factually true: courts of law would deem such actions as “not in our control” and thus not find us guilty. Not because the “insanity defense” holds that if a mental illness caused our action we are innocent (the insanity defense never operates that way), but because the “insanity defense” holds that if we did not tender informed consent to what we did because we physically couldn’t, we were never in control of it in the very sense everyone else is.

Tourette syndrome is an example: where it is physically impossible to control our utterances, we are not held responsible for them; which is observably different from everyone else who can control their utterances. The difference is that those who are in control consent to what they say (the ability of our conscious consent to causally impact what we do is what we mean by controlling our behavior), so what they say and what they want to say are in alignment, hence we can conclude what they said was caused by what they (as a person) wanted; whereas a victim of Tourette syndrome has no such physical capability: no matter how adamantly they don’t want to say something, they say it anyway. Their desires and their behaviors are not in alignment; so we cannot claim their desires caused their behavior, and therefore we cannot hold them responsible for it.

This failure to go look closely at what “free will” means in real world practice has thus caused the perverse reversal of reality—wherein causation is required for free will to be deemed present—into absurdity—wherein free will requires the absence of that very causation. And that renders free will nonsensical. How can we infer you knew you were committing the crime and wanted to commit it, if we must first declare no knowledge or desire can ever have caused you to have done it?

This is how the LFW advocate is destroying the practical real world concept of free will. They are in essence doing just what creationists do when they claim a “theory” is just “a guess” and therefore the theory of evolution is not a scientific fact. Christians contrive a wholly incoherent definition of free will that’s completely divorced from every practical real world application of free will, and then use that deviant monstrosity to push some other absurdity, like the desert of hell or the innocence of God. The last thing we should be doing to combat that nonsense is to agree with them that “theory” just means “a guess” or that “free will” means LFW. Both are divorced from the reality of what those words mean in any pertinent context. In reality free will means with informed consent, not “free of causation.” No court of law summons physicists to determine whether a criminal’s actions were free of causation; to the contrary, they require prosecutors to prove the suspect’s actions were caused by their knowledge and intent—because that is what free will really means.

The Stanford Encyclopedia of Philosophy

My closing examples come from the Stanford Encyclopedia of Philosophy online, which has an entry for Arguments for Incompatibilism by Kadri Vihvelin. The principal error pervading the “syllogisms” presented there are two equivocation fallacies over what “control” and “could have done otherwise” mean—by ignoring what they mean in the real world. In the real world, to control something does not mean what the incompatibilists Vihvelin surveys present it as meaning. That point I already thoroughly covered in Free Will in the Real World … and Why It Matters, so I won’t repeat that here. But in just the same fashion, in the real world, “could have done otherwise” also means something other than what the incompatibilists Vihvelin surveys present it as meaning.

In courts of law, when it is asked whether a criminal “could have done otherwise” this always means if they had wanted to. That is in fact absolutely essential to establishing guilt in a court of law. Because the very thing one must prove to prove guilt is the presence of an informed desire to commit the crime, what is called in legal jargon mens rea, “a guilty mind” or simply “criminal intent.” It is the physical fact that the criminal could have done otherwise had they wanted to that establishes in fact they did want to commit the crime and thus are guilty of it. If it is proved in court that the accused could not have done otherwise even if they had wanted to, that would produce an acquittal—because it means the criminal act is no longer evidence of criminal intent. In other words, if the accused could not have stopped what happened—if it would have happened no matter what the accused wanted or did—then it cannot be shown they intended it to happen, much less that their intending it caused it. Other evidence could be adduced, e.g. their saying they intended the outcome, but that only goes back to proving the same point: courts only care whether a criminal could have done otherwise had they wanted to; they aren’t looking for magical wizards who can choose to do what they have no desire at all to do.

If it were the other way around, if someone “could have” not committed a crime even if they had no other desire than to commit the crime, the whole concept of requisite intent would become meaningless. Because that would entail the converse possibility, that someone who didn’t commit a crime could still have committed the crime even if they had no desire to. Which means we would be incapable of knowing if anyone was a criminal. On this model of agent physics, whether you do or don’t commit a crime (whether you “could have done otherwise” in either direction) is no longer causally connected to any intent to commit a crime. It’s just random. We could not infer from the commission of a crime that someone wanted to do that or knew they were doing it, because they “could have done otherwise” regardless of anything they wanted or knew. If such a power existed, we could rarely establish anyone’s guilt. Every accused person could just appeal to this strange physics of human behavior, that people just “do things” wholly uncaused, and therefore no one can prove anyone intends to do anything or knew they were doing it. Mens rea could never be proved. Everyone would go free. That’s the consequence of LFW.

Ironically, it is claimed this is the consequence of determinism, too—but it isn’t. Any accused person who appealed to determinism to argue they “could not have done otherwise” would have no legal argument—the law cares not whether you “could not have done otherwise” in that sense, it only cares whether you knew what you were doing and wanted to do it anyway. It is the causal role of your intent that alone makes you guilty of anything you have done. That’s precisely the very thing we are aiming to hold people accountable for. As a famous Supreme Court Justice once said to a suspect claiming he was “fated” to knowingly and intentionally commit a crime: well, he was therefore fated to go to prison for it. The justice’s logic is unassailable.

Determinism argues nothing against the logic of criminal justice or the assignment of responsibility anywhere, at any level in our society. The accused criminal claiming he “could not have done otherwise” as a defense is simply committing an equivocation fallacy: in law, the “could not have done otherwise” defense only holds when it’s proved you “could not have done otherwise even if you wanted to.” You have to prove intent had no causal role in the events, that they could not have been stopped no matter how much the accused wanted to stop them. Philosophers commit this same confusion, thus wholly failing to understand the entire reality of moral and criminal responsibility in practical fact, which is always about knowledge and intent, not “defying fate.”

This confusion pervades the incompatibilist arguments in the Stanford article. Although Vihvelin gets the point I made earlier: that syllogistic “proofs” like Pearce’s suffer from the particular flaw of presuming the conclusion in the premises. Vihvelin presents an example:

  1. [A manipulated victim] doesn’t act freely and, for that reason, is not morally responsible for what he does.
  2. If determinism is true, there is no relevant difference between [that victim] and any normal case of apparently free and morally responsible action.
  3. Therefore, if determinism is true no one ever acts freely or is morally responsible for what he does.

As Vihvelin concludes, “this argument works only if the second premise is true,” but that is precisely the point in contention: whether the second premise is true. So you can never “prove” the conclusion that way. This is just argument by assertion; not a proof of anything. In actual fact, there is an obvious material difference between someone who has been manipulated into committing a crime unwittingly, and someone who committed the crime knowingly and willingly. And in fact that difference is precisely the difference made by free will in every real world milieu. Free will means acting with knowledge and intent—acting with informed consent. That is precisely what the “victim” in the syllogism doesn’t get to do, so they don’t have free will in precisely the way someone who knowingly and willingly committed a crime does. Premise 2 is therefore false. And anyone who paid attention to reality would know that.

Vihvelin points out the same flaw befalls every other syllogism they find attempted in the literature. For example, the next one discussed starts with the first premise, “We act freely (in the way necessary for moral responsibility) only if we are the ultimate sources (originators, first causes) of at least some of our choices,” but that is precisely what is to be proved; you can’t therefore prove it by merely asserting it. There is actually no such principle, in law or anywhere. Responsibility is not a function of “ultimate causes,” as if only the Big Bang can be responsible for anything.

Courts of law don’t care why or how you became a criminal, for example; they only care whether you are a criminal. All they want to know is whether you knew it was wrong and chose to do it anyway. If circumstances constrained your freedom in some way—for instance you were starving and only stole food for that reason, or you were born into and raised by a family so criminal and abusive your path toward criminality was all but assured—you cannot get an acquittal (you’re still found guilty), but your punishment might be reduced to compensate you for what you suffered that drove you to your crime. Because we always deem wanton sociopaths and gratuitous violators as worse than victims of circumstance; but we still classify all criminal minds as what they are: criminal minds. You cannot ontologically erase the fact that you were proved malevolent in precisely the way those not being convicted of crimes haven’t. The social causes underlying your criminality matter, but are an issue for governments and communities, not trial juries, who are only tasked with assessing the outcome, not its deeper social causes.

This does still mean we should all be taking steps to help arrange society so as to manufacture fewer dangerous people; but at the same time it does not matter why someone is dangerous, you must protect yourself from them all the same—because they simply are dangerous. On any coherent understanding of justice, merely retributive justice is pointless, even evidentially counter-productive, and only evidence-based, outcome-measure-driven goals of segregation, deterrence, and reform have moral merit in any system of punishment. Which is also why hell is an absurdly irrational, and indeed fundamentally unjust and immoral, response to criminality—particularly for an agent with infinite wisdom and unlimited capabilities and resources. Even fallible, limited, resource-starved humans have come up with far wiser and consequentially justifiable punishment systems. Hell just isn’t compatible with any widely believed concept of God. But that would be the case even if LFW were a thing. Because there is actually no logical connection between LFW and retributive justice. Which is why no one has ever presented one. All we get is handwaving. Hence the problem there isn’t LFW. (Nor does this disprove the existence of God, BTW; it only disproves the existence of Hell.)

Vihvelin then surveys the highly convoluted syllogistic Consequence Argument of van Inwagen, which had to go through several revisions even to achieve formal validity (a common problem when you attempt convoluted and confusing syllogisms), and in the end was still caught out resting on a fallacious false premise, that “I had the ability to do something (raise my hand) such that if I had exercised my ability, [the past] would have been different” has to be identical to “I had the ability to do something (raise my hand) such that if I had exercised my ability, I would thereby have caused the [the past] to be different,” and since it isn’t (those two statements are not the same things), van Inwagen’s syllogism collapsed into a nothingburger. He was simply trying to deny “I had the ability to do something (raise my hand) such that if I had exercised my ability, [the past] would have been different” as an available ontological option; but since it plainly is an ontologically available option, you can’t prove incompatibilism—because that being an ontologically available option is compatibilism.

Indeed, that’s exactly what courts of law, for example, are saying: “you are guilty if you could have done otherwise had you wanted to,” i.e. if past facts had been different than they were, and you had lacked criminal intent, instead of possessed it, then you wouldn’t have committed the crime. Hence your possessing that criminal intent, and it causing your action, is precisely what establishes that you acted freely to commit a crime. Only if you could not have done otherwise even had you wanted to would that not be the case, and you would then be acquitted—for want of proving your intent. After all, if the crime happened no matter what you intended to do, you can hardly be proved responsible for it happening. You might be; secretly, maybe you liked that outcome and thought you were bringing it about, but that can’t be proved in court if the same thing would have happened anyway (unless, as I noted earlier, other evidence of your intent can be adduced, as can be done in Ronald Opus style scenarios, but that’s irrelevant to my point).

Conclusion

Deductive syllogisms are almost always useless, and usually only hide errors of reasoning under attractive technicalia, rather than preventing them as one would otherwise expect “logical rigor” to do. You should always be extremely suspicious of them, and extensively challenge their validity and soundness, before using them or even taking them seriously. Arguments “against free will” are just one area where this is evident. The same can be found in almost any domain where syllogisms are deployed, from theology to metaethics and beyond. Indeed, almost (and I do emphasize almost—because there are exceptions) the only use syllogisms have is as a means of exposing how your reasoning is going astray, by formalizing it into a syllogism and then finding the resulting flaw—as you can always trust it’s very likely there will be one. Which means what is most useful of all about syllogisms is that you can use them to diagram and thus evaluate your own reasoning or someone else’s—precisely to identify what one’s assumptions are, and their dependence on them.

In the present case, arguments against free will simply never use “free will” in any real world sense, but in some bizarre, incoherent sense that has never had any practical application. In the real world, having free will is about whether you consented to what you did—whether you did it knowingly and intentionally. Which actually requires causal determinism to exist. Which is the exact opposite of “LFW”, which perversely claims free will requires the complete absence of causation—that to be free we must make decisions wholly at random, bypassing every relevant cause of responsible action, including knowledge, desires, character, everything, even reasons for acting. That’s irrational. And useless. Toss it. And get back to using the only definition of free will actually employed in practice. Knowing how syllogisms work can help you discover this fact. But they won’t by themselves prove anything.

-:-

You can peruse Pearce’s replies through links provided in comments, and follow that comment’s breadcrumb for my thoughts, beyond what my article already explains or links to.

§

To comment use the Add Comment field at bottom, or click the Reply box next to (or the nearest one above) any comment. See Comments & Moderation Policy for standards and expectations.

Discover more from Richard Carrier Blogs

Subscribe now to keep reading and get access to the full archive.

Continue reading