I agree with all this. Though I think a corollary is that whether an argument begs the question isn't something you can straightforwardly read off from the argument itself; it depends on the intended audience. One and the same argument might beg the question against some audiences, but not others. I think that's part of what makes it easy to have disputes like the original one on twitter; people may differ a lot in what audiences they're imagining for some argument.
Yes! In fact, this sort of creeps up against a worry I have about arguments in general: if we imagine a sufficiently idealized, rational agent, perhaps there are *no* non-Q-begging arguments I could give them. Like, even the money pump example: a sufficiently idealized agent might have seen that intransitive preferences obviously imply these pumps (much as we obviously see that only white swans entails no black swans). What “moves the dialectic forward” or is “independent” really is person-relative.
Haha, that reminds me of Pruss responding to Oppy on W-PSR implies S-PSR saying "well it's not question begging unless youre a super expert in modal logic like Oppy so give me a break" lol (paraphrase)
You assume that begging the question makes an argument bad, but I deny this.
> “Well, even if (2) is false, what makes this a good argument is that it makes dialectical progress.
You assume that good arguments need a possibility of making dialectical progress. I accept that question-begging arguments have no possibility of making dialectical progress. However, your view requires a substantive metaphilosophical assumption that dialectical progress is possible. I am convinced that on many (perhaps most) philosophical issues, dialectical progress is by and large impossible. In other words, there is no way to resolve these philosophical issues through purely epistemic considerations. If dialectical progress is not possible, then it's not obvious that good arguments must even have a possibility of rationally persuading someone who is not antecedently inclined to accept the position.
Second, we rationally do or may hold beliefs based on question-begging arguments. I believe my memory is reliable. Here is an argument to that effect. If my memory is reliable, then I remember a lot of what has happened in my past. My memory is reliable, so I remember a lot of what has happened in the past. I accept this argument, but it begs the question because it won’t convince someone who doesn’t already accept the reliability of memory in general. This argument won’t convince the Pyrrhonist. But if any belief is rational, my beliefs about the past are rational even though they presuppose the reliability of my memory.
> “This is what good arguments do: they move things forward, by showing that there are other, independent considerations that should move you towards my side.”
My concern is that sometimes there are no independent considerations that could move my opponent over towards my side. Think of the radical skeptic, the irrationalist, or the trivialist. If a proponent of these views is sufficiently radical, there will be literally nothing you have in common. The problem with saying that begging the question is bad is that it assumes that you must be able to convince other people based on commonalities. But I deny this. If we cannot we cannot legitimately beg the question against some people, then what do you say to the radical skeptic?
Interesting! A few thoughts, although I want to preface them by saying I don’t think non-Q-begging is a necessary property of good arguments. I think whether an argument is good is sensitive to context. In particular, it is sensitive to what the common grounds are of the person you’re speaking to. As other comments have pointed out, in some contexts an argument is question begging and in others it isn’t. Anyways:
1. You say that in some areas it is not possible to resolve issues through purely epistemic considerations. Hmmm … if I grant you this, then it sounds like you should just think that in some areas there are no good arguments! Now, to be clear, that does not mean that people can’t be *justified* in believing what they believe. They totally can — they can have a set of reasons which entail their view. But there might not be any arguments in this domain that should convince anyone else: for everyone might have no common ground, as you point out.
2. As for the radical views you point out … if we really have no common ground from which to establish independently plausible arguments, then I think the thing to say is that there *is* no argument that should convince either of us to change our minds!
I said that “there is no way to resolve these philosophical issues through purely epistemic considerations.” You think this means there are no good arguments in philosophy, but I am unsure what you mean by “good” argument. I was unclear, but what I meant is that I am unaware of any philosophical arguments for interesting conclusions that rationally compel assent. I’m reminded of Alvin Plantinga’s characterization of natural theology in Warranted Christian Belief: “I took it that these arguments are successful only if they start from propositions that compel assent from every honest and intelligent person and proceed majestically to their conclusion by way of forms of argument that can be rejected only on pain of insincerity or irrationality…(I failed to note that no philosophical arguments of any consequence meets that standard; hence the fact that theistic arguments do not is of less significance than I thought.)” I doubt philosophers can produce arguments for their interesting views that meet Plantinga's criteria. But if good arguments merely provide reasons for assent, then there are good philosophical arguments.
Cool post! FWIW I actually wrote an essay on the Moorean response to skepticism for this exam season, which involved developing an account of something similar to begging the question (though the scope is narrower), which I think echoes what you say here:
"The inference from premises P to the falsity of alternative A is justificatorily circular IFF P is not immediately known AND for all my evidence E to justify P(P)=X, I must presuppose that P(¬A)≥X.
The general idea of “presupposing” is that an inference from E to P presupposes Q just in case a belief in Q is “logically prior” to E justifying P. A way of illustrating this is to imagine yourself behind a sort of “veil of ignorance,” not knowing what evidential situation you will be in. To presuppose P(Q)=Y for E to support P(P)=X, simply means that P(P|E)=X is justified, only if you assume P(Q)=Y behind the veil."
[After writing it, I actually suspect the way I spell out "presupposing" is too strong, ruling in many bad inferences. But I'm still not sure about that.]
Hmm I must say I’m not exactly following your account of “presuppose”. With that said though, your definition of justificatory circularity seems right!
Yeah I've also had a hard time nailing it down. I've tried rephrasing it:
"The general idea of “presupposing” is that an inference from E to P presupposes Q just in case a belief in Q is “logically prior” to E justifying P. A way of illustrating this is to imagine yourself behind a sort of “veil of ignorance,” not knowing whether E obtains or not. For the inference from E to P(P)=X to presuppose P(Q)=Y then means that P(P|E)=X only if P(Q)=Y.
That is, even before finding out whether E, I have to already have P(Q)=Y to make the inference. Or applying it to justificatory circularity: Even before finding out whether E, I already hold ¬A to be at least as probable as the premises, which are supposed to establish ¬A."
Anyways that also doesn't matter, just thought it was interesting that we converged on the point that begging the question isn't inherent to an argument, but depends on the dialectical context--specifically what positions the argument is trying to rule out.
Yeah I definitely think it is sensitive to context! Although, now that I think about your account of presupposing, I want to point out that maybe the “only if P(Q) = Y” might be too weak, just because of the weirdness of the material conditional. For instance, take some necessarily true proposition (1+1=2). It might be that 1+1=2 satisfies the conditional when substituted for ‘Q’. But I think it’s kind of weird to say all our inferences presuppose that 1+1=2 (would we say they all presuppose that Goldbach’s Conjecture is true/false?). I think this is because ‘presuppose’ isn’t so “external” — maybe you need to steal some notion of ‘relevance’ from relevance logics or something, not sure. Anyways, as you say, it doesn’t matter! Main point is still the same
Interesting point! I suppose I don't think it's too strange to say that any inference "presupposes" that 1+1=2. It just seems like it's sort of irrelevant to mention, but it really is a background assumption in some sense, just one that doesn't directly bear on the inference being made. But precisely demarcating what bears on some inference will be hard, I suspect. Besides, presupposing isn't inherently a bad thing--only when the thing presupposed is supposed to be the thing you're trying to establish.
Now, all purely formal inferences are going to be justificatorily circular by this account (at least for people with consistent credences). But then they're also a special case beyond the intended scope.
Oh good point about formal inferences … yeah idk it does seem like a cost to say they’re all justificatorily circular. Although, maybe not: if someone doesn’t accept any of our (eg) mathematical axioms, I guess they’d think all our mathematical arguments and proofs suck.
Wait, the connection to our discipline’s practice of objection-minimizing is really nice! I hadnt realized this: objection minimization leads to less contestable philosophy, which in turn means we don’t get more clarity about the costs of certain views. That, ironically, is also what question begging arguments do!
I want to point out a problem with Ben Watkins’ analysis of Darwin to Jesus’ argument. DtJ offers the argument: (1) If materialism is true, the mind and brain are identical. (2) The mind and brain are not identical. (3) Materialism is not true. Watkins points out that materialism is “logically equivalent” to the proposition that “the mind and brain are identical.” I assume he means that materialism is true if and only if the relevant proposition is true. But this necessary equivalence doesn’t seem to be a sufficient condition for begging the question.
You give the example of the rationality of intransitive preferences. Suppose someone offers the following argument to me. If no rational agent has intransitive preferences, then no rational agent has intransitive preferences. No rational agent has intransitive preferences. Therefore, no rational agent has intransitive preferences. I’d say this begs the question. But then suppose my interlocutor points out that if you have transitive preferences, you can get money pumped. So my opponent argues: If you have intransitive preferences, you can get money pumped. No rational agent can get money pumped. So, no rational agent has intransitive preferences. Given Watkins’ criterion for begging the question, this argument is question-begging. There is no possibility of an agent who has intransitive preferences and who can't also be money pumped. But clearly, the latter argument isn't question-begging for me.
I think this shows that begging the question is an epistemic (and therefore person-relative) thing. It doesn’t have to do with logical relations like necessary equivalences. Otherwise, it would seem that all arguments are question-begging in virtue of their deductive validity. Suppose, then, I had never realized that materialism implies that the mind is identical to the brain. In that case, I think DtJ’s argument would not have been question-begging for me because it would have given me new information.
I agree with your conclusion — begging the question is person-relative. But just want to point something out:
(1) Watkins is just wrong that materialism is true iff the mind and brain are identical. The mind and brain could be identical, and yet that doesn’t entail that materialism is true, because something else could be non-material (abstract objects, for instance)
(2) Something similar about money pumps. Getting money pumped doesn’t entail intransitive preferences (you can have transitive preferences and still get money pumped, if your preferences don’t cohere with other principles).
But your conclusion is right. Question begging is person relative
Agreed! Well, to an extent. I agree that money pump arguments (like all arguments) can be begging the question in certain contexts / for certain agents. Just a slight nitpick though: non-representability needn’t entail that you *wouldn’t at all* pay some amount of money to go from one option to another, or be willing to take some amount of money to go from one to another. I think the thing to say is that in some cases the possibility of money pumps isnt a mark of irrationality.
I agree with all this. Though I think a corollary is that whether an argument begs the question isn't something you can straightforwardly read off from the argument itself; it depends on the intended audience. One and the same argument might beg the question against some audiences, but not others. I think that's part of what makes it easy to have disputes like the original one on twitter; people may differ a lot in what audiences they're imagining for some argument.
Yes! In fact, this sort of creeps up against a worry I have about arguments in general: if we imagine a sufficiently idealized, rational agent, perhaps there are *no* non-Q-begging arguments I could give them. Like, even the money pump example: a sufficiently idealized agent might have seen that intransitive preferences obviously imply these pumps (much as we obviously see that only white swans entails no black swans). What “moves the dialectic forward” or is “independent” really is person-relative.
Haha, that reminds me of Pruss responding to Oppy on W-PSR implies S-PSR saying "well it's not question begging unless youre a super expert in modal logic like Oppy so give me a break" lol (paraphrase)
You assume that begging the question makes an argument bad, but I deny this.
> “Well, even if (2) is false, what makes this a good argument is that it makes dialectical progress.
You assume that good arguments need a possibility of making dialectical progress. I accept that question-begging arguments have no possibility of making dialectical progress. However, your view requires a substantive metaphilosophical assumption that dialectical progress is possible. I am convinced that on many (perhaps most) philosophical issues, dialectical progress is by and large impossible. In other words, there is no way to resolve these philosophical issues through purely epistemic considerations. If dialectical progress is not possible, then it's not obvious that good arguments must even have a possibility of rationally persuading someone who is not antecedently inclined to accept the position.
Second, we rationally do or may hold beliefs based on question-begging arguments. I believe my memory is reliable. Here is an argument to that effect. If my memory is reliable, then I remember a lot of what has happened in my past. My memory is reliable, so I remember a lot of what has happened in the past. I accept this argument, but it begs the question because it won’t convince someone who doesn’t already accept the reliability of memory in general. This argument won’t convince the Pyrrhonist. But if any belief is rational, my beliefs about the past are rational even though they presuppose the reliability of my memory.
> “This is what good arguments do: they move things forward, by showing that there are other, independent considerations that should move you towards my side.”
My concern is that sometimes there are no independent considerations that could move my opponent over towards my side. Think of the radical skeptic, the irrationalist, or the trivialist. If a proponent of these views is sufficiently radical, there will be literally nothing you have in common. The problem with saying that begging the question is bad is that it assumes that you must be able to convince other people based on commonalities. But I deny this. If we cannot we cannot legitimately beg the question against some people, then what do you say to the radical skeptic?
Interesting! A few thoughts, although I want to preface them by saying I don’t think non-Q-begging is a necessary property of good arguments. I think whether an argument is good is sensitive to context. In particular, it is sensitive to what the common grounds are of the person you’re speaking to. As other comments have pointed out, in some contexts an argument is question begging and in others it isn’t. Anyways:
1. You say that in some areas it is not possible to resolve issues through purely epistemic considerations. Hmmm … if I grant you this, then it sounds like you should just think that in some areas there are no good arguments! Now, to be clear, that does not mean that people can’t be *justified* in believing what they believe. They totally can — they can have a set of reasons which entail their view. But there might not be any arguments in this domain that should convince anyone else: for everyone might have no common ground, as you point out.
2. As for the radical views you point out … if we really have no common ground from which to establish independently plausible arguments, then I think the thing to say is that there *is* no argument that should convince either of us to change our minds!
I said that “there is no way to resolve these philosophical issues through purely epistemic considerations.” You think this means there are no good arguments in philosophy, but I am unsure what you mean by “good” argument. I was unclear, but what I meant is that I am unaware of any philosophical arguments for interesting conclusions that rationally compel assent. I’m reminded of Alvin Plantinga’s characterization of natural theology in Warranted Christian Belief: “I took it that these arguments are successful only if they start from propositions that compel assent from every honest and intelligent person and proceed majestically to their conclusion by way of forms of argument that can be rejected only on pain of insincerity or irrationality…(I failed to note that no philosophical arguments of any consequence meets that standard; hence the fact that theistic arguments do not is of less significance than I thought.)” I doubt philosophers can produce arguments for their interesting views that meet Plantinga's criteria. But if good arguments merely provide reasons for assent, then there are good philosophical arguments.
Cool post! FWIW I actually wrote an essay on the Moorean response to skepticism for this exam season, which involved developing an account of something similar to begging the question (though the scope is narrower), which I think echoes what you say here:
"The inference from premises P to the falsity of alternative A is justificatorily circular IFF P is not immediately known AND for all my evidence E to justify P(P)=X, I must presuppose that P(¬A)≥X.
The general idea of “presupposing” is that an inference from E to P presupposes Q just in case a belief in Q is “logically prior” to E justifying P. A way of illustrating this is to imagine yourself behind a sort of “veil of ignorance,” not knowing what evidential situation you will be in. To presuppose P(Q)=Y for E to support P(P)=X, simply means that P(P|E)=X is justified, only if you assume P(Q)=Y behind the veil."
[After writing it, I actually suspect the way I spell out "presupposing" is too strong, ruling in many bad inferences. But I'm still not sure about that.]
Hmm I must say I’m not exactly following your account of “presuppose”. With that said though, your definition of justificatory circularity seems right!
Yeah I've also had a hard time nailing it down. I've tried rephrasing it:
"The general idea of “presupposing” is that an inference from E to P presupposes Q just in case a belief in Q is “logically prior” to E justifying P. A way of illustrating this is to imagine yourself behind a sort of “veil of ignorance,” not knowing whether E obtains or not. For the inference from E to P(P)=X to presuppose P(Q)=Y then means that P(P|E)=X only if P(Q)=Y.
That is, even before finding out whether E, I have to already have P(Q)=Y to make the inference. Or applying it to justificatory circularity: Even before finding out whether E, I already hold ¬A to be at least as probable as the premises, which are supposed to establish ¬A."
Anyways that also doesn't matter, just thought it was interesting that we converged on the point that begging the question isn't inherent to an argument, but depends on the dialectical context--specifically what positions the argument is trying to rule out.
Yeah I definitely think it is sensitive to context! Although, now that I think about your account of presupposing, I want to point out that maybe the “only if P(Q) = Y” might be too weak, just because of the weirdness of the material conditional. For instance, take some necessarily true proposition (1+1=2). It might be that 1+1=2 satisfies the conditional when substituted for ‘Q’. But I think it’s kind of weird to say all our inferences presuppose that 1+1=2 (would we say they all presuppose that Goldbach’s Conjecture is true/false?). I think this is because ‘presuppose’ isn’t so “external” — maybe you need to steal some notion of ‘relevance’ from relevance logics or something, not sure. Anyways, as you say, it doesn’t matter! Main point is still the same
Interesting point! I suppose I don't think it's too strange to say that any inference "presupposes" that 1+1=2. It just seems like it's sort of irrelevant to mention, but it really is a background assumption in some sense, just one that doesn't directly bear on the inference being made. But precisely demarcating what bears on some inference will be hard, I suspect. Besides, presupposing isn't inherently a bad thing--only when the thing presupposed is supposed to be the thing you're trying to establish.
Now, all purely formal inferences are going to be justificatorily circular by this account (at least for people with consistent credences). But then they're also a special case beyond the intended scope.
Oh good point about formal inferences … yeah idk it does seem like a cost to say they’re all justificatorily circular. Although, maybe not: if someone doesn’t accept any of our (eg) mathematical axioms, I guess they’d think all our mathematical arguments and proofs suck.
Oh yeah, also completely forgot to reply earlier, but I'd definitely be interested in reading your paper on bounded aggregation:)
Related: https://www.goodthoughts.blog/p/contestable-vs-question-begging-arguments
Wait, the connection to our discipline’s practice of objection-minimizing is really nice! I hadnt realized this: objection minimization leads to less contestable philosophy, which in turn means we don’t get more clarity about the costs of certain views. That, ironically, is also what question begging arguments do!
I want to point out a problem with Ben Watkins’ analysis of Darwin to Jesus’ argument. DtJ offers the argument: (1) If materialism is true, the mind and brain are identical. (2) The mind and brain are not identical. (3) Materialism is not true. Watkins points out that materialism is “logically equivalent” to the proposition that “the mind and brain are identical.” I assume he means that materialism is true if and only if the relevant proposition is true. But this necessary equivalence doesn’t seem to be a sufficient condition for begging the question.
You give the example of the rationality of intransitive preferences. Suppose someone offers the following argument to me. If no rational agent has intransitive preferences, then no rational agent has intransitive preferences. No rational agent has intransitive preferences. Therefore, no rational agent has intransitive preferences. I’d say this begs the question. But then suppose my interlocutor points out that if you have transitive preferences, you can get money pumped. So my opponent argues: If you have intransitive preferences, you can get money pumped. No rational agent can get money pumped. So, no rational agent has intransitive preferences. Given Watkins’ criterion for begging the question, this argument is question-begging. There is no possibility of an agent who has intransitive preferences and who can't also be money pumped. But clearly, the latter argument isn't question-begging for me.
I think this shows that begging the question is an epistemic (and therefore person-relative) thing. It doesn’t have to do with logical relations like necessary equivalences. Otherwise, it would seem that all arguments are question-begging in virtue of their deductive validity. Suppose, then, I had never realized that materialism implies that the mind is identical to the brain. In that case, I think DtJ’s argument would not have been question-begging for me because it would have given me new information.
I agree with your conclusion — begging the question is person-relative. But just want to point something out:
(1) Watkins is just wrong that materialism is true iff the mind and brain are identical. The mind and brain could be identical, and yet that doesn’t entail that materialism is true, because something else could be non-material (abstract objects, for instance)
(2) Something similar about money pumps. Getting money pumped doesn’t entail intransitive preferences (you can have transitive preferences and still get money pumped, if your preferences don’t cohere with other principles).
But your conclusion is right. Question begging is person relative
Agreed! Well, to an extent. I agree that money pump arguments (like all arguments) can be begging the question in certain contexts / for certain agents. Just a slight nitpick though: non-representability needn’t entail that you *wouldn’t at all* pay some amount of money to go from one option to another, or be willing to take some amount of money to go from one to another. I think the thing to say is that in some cases the possibility of money pumps isnt a mark of irrationality.