1.0 The Halfway Fallacy
Some time ago, Bentham’s Bulldog (BB) presented the following argument in this blog post:
Suppose we’re about to discover whether moral realism is true. Someone offers you a bet. If it’s true they get a dollar, if it’s false you get 10,000 dollars. Should you take the bet?
No. Here’s why. If moral realism is true, then winning the bet actually makes you better off. You have a genuine reason to not [sic] the bet — you’d just lose a dollar. But if anti-realism is true, you are not made better off, nothing matters, and taking the bet doesn’t make you better off. You only have genuine reason in one case.
This argument is an excellent demonstration of what I call the halfway fallacy. I use the term “fallacy” here loosely, since it’s more of a presumption others are free to contest, a kind of philosophical oversight.
The halfway fallacy occurs when one argues that a particular position contrary to their own has one or more flaws or undesirable characteristics, but those flaws or undesirable characteristics are only applicable to the position if some set of claims you believe are true are in fact true, but you haven’t argued for and those the objections are directed at are free to (and in many cases probably would) reject. In other words, the problem occurs when one holds certain presuppositions that those who hold the view are free to reject (it might also be the case that they aren’t merely free to reject these presuppositions but do, and it may even be that rejecting the presuppositions is a natural and synergistic feature of the contrary view).
One might also call this the fallacy of unshared presumption. The central problem with this form of reasoning is that an argument for or against some position is based on considering the implications of someone holding one or more views contrary to one’s own, but, critically, not considering that they reject certain other presuppositions you hold. Of course, any objection you raise against a particular position is likely to be predicated on various presuppositions that someone else might reject. So one might insist every argument has this feature.
To a certain extent, that is true. Yet the presuppositions behind one’s position will often be more or less a matter of dispute between yourself and one’s interlocutors. It is one thing to presuppose the earth is round when arguing with someone about the geographic location of a particular city. It is quite another to presuppose that the Christian God exists when arguing with an atheist.
The halfway fallacy isn’t so much an inferential mistake, then, but a way of framing disputes as though one has a strong argument when one really doesn’t, often because the argument criticizes some view in virtue of that views implications given some set of presuppositions that proponents of that view are likely to reject, or that they could readily reject in virtue of approaching the philosophical dispute in question from a fundamentally different perspective, often one closely tied to the views downstream of that perspective.
Perhaps the best way to illustrate what I have in mind is an example.
2.0 Scifi Lovers and Scifi Haters
Suppose you think the following:
Science fiction movies are the only good movies.
Now imagine you encounter someone who says:
“I don’t like science fiction movies.”
The shock! The horror! What mangled mind could fail to appreciate 2001: A Space Odyssey?!
Given your views, you could then conclude that:
This person does not like any good movies.
After all, the following propositions are true:
This person does not like science fiction movies.
Science fiction movies are the only good movies.
It follows, then, that:
This person does not like any good movies.
Now, it would be one thing for you to judge that this person does not like any movies that you consider good movies. But imagine that you insisted, to this person, that given their views, that they are committed to the view that they don’t like any good movies:
Scifi Lover: “Given that the only good movies are science fiction movies, if you don’t like any science fiction movies, then you don’t like any good movies.”
How do you expect this person to react? Here’s one way they could respond:
Scifi Hater: “Yes, I concede that is an implication of my view. Given that the only good movies are science fiction movies, then, since I don’t like any science fiction movies, then, clearly, I don’t like any good movies.”
Do you think that’s how most people would respond? I suspect not. This would be a weird and implausible response. Instead, they’d be much more likely to respond like this:
Scifi Hater: “Sure, I don’t like any science fiction movies. But I think there are good movies that aren’t science fiction movies. I like movies in a variety of different genres: documentaries, dramas, comedies, and so on. I just don’t like science fiction movies.”
This is a bit more formal than how a person might actually respond, but it conveys the typical reaction someone would actually have. And what does this convey?
They don’t like science fiction movies.
They like various non-science fiction movies.
They don’t think science fiction movies are the only good movies.
They think there are good movies other than science fiction movies.
In other words, they reject (2) above, that “Science fiction movies are the only good movies.”
As such, they do not have to grant that they don’t like any good movies, for the simple reason that they also reject the scifi lover’s presumption that science fiction movies are the only good movies.
The halfway fallacy can be illustrated in this example by considering what the person who likes science fiction is presuming. They grant that the other person does not like science fiction movies. However, they continue to presume that only science fiction movies are good movies. Yet the person who doesn’t like science fiction movies also disagrees with this latter claim. That is, they do not think that science fiction movies are the only good movies. So this person is not committed to the view that they do not like any good movies because they don’t share the scifi lover’s belief that only science fiction movies are good movies.
It would be absurd for the science fiction lover to insist that this person really is committed, on their own view to agree that they don’t like any “good movies.” The science fiction lover seems to have only gone halfway in considering the perspective of a person who doesn’t like science fiction movies: they recognize that this person doesn’t like science fiction movies, but haven’t seemed to appreciate that this person disagrees with them about a further fact: that the only good movies are science fiction movies. If they were to have a dispute, the dispute could very well center on whether science fiction movies are the only good movies. It would be inappropriate to begin an inquiry under the presumption that it is simply a fact that only science fiction movies are good movies; that is one of the things the science fiction hater denies.
3.0 Supernatural magic and natural illusions
Let’s apply this same reasoning to BB’s argument. According to BB, you shouldn’t take the bet because if realism is true, winning the bet “actually makes you better off,” while if antirealism is true, you would not be better off even if you won $10,000. Why? Because if antirealism is true, then “nothing matters.” According to BB, you only have a “genuine reason” if realism is true.
There are multiple problems here. One is the use of deceptive modifiers, which I have discussed at length in an earlier post.
Note the use of the terms “actually” and “genuine.” Both of these terms are deceptive modifiers. In both cases, the presumption is that you can only be better off, that things can only matter, and that you can only have reasons to do things if realism is true. The presumption behind these terms is that the only sense in which any of these notions obtains is in some respect that just constitutes the realist’s conception of “better off,” “mattering,” and “reasons.”
In other words, if BB just means “in a realist sense,” then it is misleading and presumptuous to use this terminology. Those of us who deny that moralism is true are not obliged to grant that only the realist’s conception of being better off, mattering, and reasons are “actual” or “genuine.” Indeed, this is one of the very things I dispute.
For comparison, suppose two people are watching a magician perform magic on a stage (a comparison I’ve made to these types of remarks before). They appear to be able to make objects disappear, saw people in half, read people’s minds, and so on. We can propose at least two possible explanations for these demonstrations:
(1) They can perform supernatural magic
(2) They can perform natural illusions
Dennett (2016) relates an account from Lee Siegel (1991)
‘I’m writing a book on magic,’ I explain, and I’m asked, ‘Real magic?’ By real magic people mean miracles, thaumaturgical acts, and supernatural powers. ‘No,’ I answer: ‘Conjuring tricks, not real magic.’ Real magic, in other words, refers to the magic that is not real, while the magic that is real, that can actually be done, is not real magic. (p. 425, as quoted in Dennett, 2016, p. 66)
Siegel’s remark highlights an interesting ambiguity in the use of the term “real.” Magic could be said to be real in at least ways:
Real in the sense of being supernatural
Real in the sense of it occurring in the real world
Supernatural magic is real in the sense of (a), but not in the sense of (b). That is, insofar as one’s notion of “real” magic is magic that isn’t reducible to some prestidigitation, some sleight of hand, or mentalism that could be described in mundane, every day terms, then if a street performer is merely doing the latter, they’re not doing “real magic.”
If, on the other hand, “real magic” refers to the actual sort of thing people can do and that does occur in the real world, then a stage performer pulling off tricks and illusions is doing real magic in that respect.
When BB uses terms like “actually,” and “genuine,” these terms are used in a way that is ambiguous between two analogous readings, only with respect to realism. We have two readings of “actual” mattering or “genuine” reasons:
Actual or genuine in the sense of being stance-independent
Actual or genuine in the sense of accurately describing what the world is actually like
I believe BB’s remarks derive much of their persuasive force from exploiting this unresolved ambiguity. While BB’s position only makes sense if BB has (a) in mind, I think BB’s remarks take advantage of the ambiguity generated by the use of terms like “genuine” to give the misleading impression that to reject BB’s position is to accept that one could not be better off, that nothing could matter, and that one could have no reasons in the sense of (b) as well.
Yet the antirealist would deny (b). Antirealists do not accept that the realist’s characterization of mattering, reasons, etc. describe what the world is actually like: that’s the very thing they deny.
To sharpen the analogy, imagine you have a friend who believes stage performers do supernatural magic. If you deny this, and maintain, instead, that they perform natural illusions, what follows from this? Does it follow that you don’t believe they’re doing anything at all? After all, if “real” means occurring in the real world, then if you deny that the magician is performing “real magic,” then aren’t you denying that they’re doing anything that’s occurring in the real world?
That would be absurd. You and the person who believes magicians are doing supernatural magic can both see the magician performing the tricks. You can see them pull the rabbit from a hat. You can see them provide detailed biographical information about an apparent stranger. You can see them walk off the stage on the left, only to appear seconds later on the right side of the stage. If you deny they’re doing “real magic,” then you must be a complete idiot: how could you possibly deny what you can see with your own eyes?
Of course, to deny they’re doing “real magic” from the perspective of someone who thinks they are doing natural illusions isn’t to deny what they see with their own eyes, it is to deny the supernaturalist’s interpretation of it.
Just the same, an antirealist isn’t limited to only rejecting that things matter (or that one can have reasons or can be better off) in the sense supposed by BB or other moral realists, i.e., stance-independently, they can also deny that this is the only sense in which things can or do matter (or the only sense in which one can be better off or have reasons).
4.0 Trivial arguments and normative entanglement
Moral antirealists are, in other words, entitled to present alternative accounts of what the actual world is like, and this includes offering alternative accounts of ways in which one can be better off, or ways in which things can matter, or ways in which can be meaningfully said to “have reasons.” We don’t have to accept the realist’s account of these concepts, and when the realist uses these terms, the antirealist does not have to accept that the realist’s construal of these terms is the only accurate or possible way to construe those terms.
What BB and philosophers more generally ought to do is, when employing such terms, make their implicit presuppositions explicit, and then see how their arguments look.
In this case, “actual” and “genuine,” once disambiguated in the only way that would allow BB’s argument to actually work (since if these terms are consistent with antirealism then BB’s claims wouldn’t be true), one would have to swap them out for something like “in a realist sense” or “stance-independently.” In other words, BB could have said this, when responding to whether you should take the bet:
No. Here’s why. If moral realism is true, then winning the bet makes you stance-independently better off. You have a stance-independent reason to not [take] the bet — you’d just lose a dollar. But if anti-realism is true, you are not made stance-independently better off, nothing stance-independently matters, and taking the bet doesn’t make you stance-independently better off. You only have stance-independent reason in one case.
Once the argument is disambiguated, this argument becomes trivial.
Moral realism (or, more accurately, normative realism) just is the view that there are stance-independent normative facts, such that one can be stance-independently better off, one can have stance-independent reasons, things can stance-independently matter, and so on. Normative antirealism, conversely, is the view that there are no stance-independent normative facts. So we could rewrite this again to say:
No. Here’s why. If the correct account of normativity is that normative considerations are stance-independent, then winning the bet makes you stance-independently better off. You have a stance-independent reason to not [take] the bet — you’d just lose a dollar. But if there are no stance-independent normative facts, you are not made stance-independently better off, nothing stance-independently matters, and taking the bet doesn’t make you stance-independently better off. You only have stance-independent reason in one case.
This clarification further reinforces how trivial this claim is. It is nothing more than a statement of the implications of normative realism and antirealism, respectively, with some terminological facades to give it a bit of rhetorical punch.
A great deal of arguments in analytic philosophy are trivial in this respect. The more I have examined these arguments, the more they appear rely on a kind of terminological legerdemain: one presupposes or stipulates that some familiar term, like “genuine,” carries their philosophical presuppositions, then one appends “genuine” to some claim they expect others to be hard-pressed to deny, not because they’d look stupid or evil if they deny your philosophical position, but because the term in question has ordinary, colloquial uses that, were one to deny them, would make you look stupid or evil.
If someone said they don’t think it’s “genuinely wrong” to torture babies for fun, they’d look evil. Even if someone said they don’t think it’s “stance-independently wrong,” to torture babies for fun, they might still look evil. Yet this is likely due to normative entanglement and to Bayesian considerations (someone who says it is stance-independently wrong affirms it is wrong full stop, whereas someone who denies it is stance-independently wrong doesn’t, and thus one would have to update from such a response that this person has given some evidence that they don’t think it’s wrong in any respect).
Now suppose someone said they don’t think it’s “magically wrong” to torture babies for fun. Now, perhaps, some of the stink is starting to rub off the normative element of the remark (i.e., “wrong”). An audience might think:
Okay, they don’t think it’s magically wrong, but do they think it’s wrong (in some non-magical way?)
Finally, a person could confirm their normative stance towards the issue alongside rejecting realism, and most of the stink is likely to go away:
I don’t think it’s stance-independently wrong, but of course I think it’s morally wrong. I find it repugnant and unacceptable and would do everything in my power to oppose it.
It’s much less clear this person is evil compared to someone who simply said they don’t think baby torture is “genuinely wrong.”
5.0 Conclusion
BB’s wager is not a very convincing one. BB seems to be imagining that if someone were considering the wager, that their decisions should be based on whether something would make them better off in a realist sense, and that, if antirealism is true, they have no reason in a realist sense to favor having $10,000 because nothing matters in a realist sense.
If someone trying to decide whether to take a bet only thought things could matter, or that they could have reasons to do things in a realist sense, then perhaps they’d conclude that they shouldn’t take the bet. But this observation strikes me as trivial.
It also seems quite strange. Suppose normative realism really were false. BB’s account of this wager suggests that one’s decisions about what to do would become irrelevant. It wouldn’t matter if you took the bet or not. Nothing would matter. If this generalizes, then it wouldn’t matter if you suffered. It wouldn’t matter if you tortured people. It wouldn’t matter what you ate. You’d have no more reason to eat food you like than to drink poison.
I suspect this appraisal of the situation creates a potentially motivated asymmetry. The stakes, for a realist who thinks this way, are very high. If they’re wrong about realism, they lose everything. Literally everything. Nothing matters. It wouldn’t matter what they did. Love is nothing. Happiness is nothing. Joy is nothing. Kindness is nothing. Reality is just a nihilistic facade.
Antirealists don’t face nearly as great an incentive to resist moral realism, however. Perhaps this asymmetry matters. Perhaps it prompts a greater degree of motivated reasoning on the part of moral realists than antirealists experience, since the costs for them are effectively infinite, while the costs to me are comparatively much lower (the main cost for me would, I suppose, be embarrassment and guilt over pushing antirealism for so long). I’d like to propose this as one minor point of evidence against moral realism. Call it the “asymmetric motivation hypothesis.” According to this hypothesis, moral realists are subject to a greater potential biasing influence when evaluating arguments for and against realist and antirealist positions, compared to antirealists. All else being equal, this provides some minor evidence against the plausibility of moral realism, in that its proponents are more vulnerable (again, all else equal) to motivated reasoning than antirealists.
References
Dennett, D. C. (2016). Illusionism as the obvious default theory of consciousness. Journal of Consciousness Studies, 23(11-12), 65-72.
Siegel, L. (1991) Net of magic: Wonders and deceptions in India. Chicago, IL: University of Chicago Press.
I suspect something like this fallacy may be lurking behind a lot of casual accusations of incoherence or self-defeat—e.g., illusionism about phenomenal consciousness is incoherent because (according to opponents of illusionism) an illusion is a phenomenally conscious experience. Or determinism is self-defeating because (according to opponents of determinism) a belief with a sufficient cause can’t be justified.
As far as I can tell, "whether moral realism is true" is not an empirical matter. It is hard to imagine how we’re about to discover whether moral realism is true. I would not take this bet.