It is often thought that one central advantage of expressivism over subjectivism is that expressivism can make sense of moral disagreements. Whereas according to subjectivism, people end up talking past one another, expressivism enables speakers to express disagreements in attitude as Stevenson famously put it. This orthodoxy has been recently challenged in two ways. Subjectivists have tried to create new ways of making sense of disagreements, and it has turned out that the traditional expressivist accounts of disagreement are more problematic than previously thought. The latter issue has become even more pressing because of the negation problem. The questions of when two people disagree and when one person holds inconsistent attitudes seem to be very much the same question, and so many expressivists have thought that by giving an account of disagreement they can also give an account of inconsistency. In a recent paper entitled “Disagreement” (PPR) and in a corresponding chapter on disagreement in his new Impassionate Belief book, Mike Ridge has tried to develop a new account of disagreement (which he calls "disagreement in prescription") to solve these worries. I want to argue below that this account fails because it commits the conditional fallacy.
Here’s Ridge’s approximate version of the view (the later complications will not affect what I’ll say below):
Two people, A and B disagree in prescription about D’s phying in C just in case in circumstances of honesty, full candor and non-hypocricy, A would advise phying in C and B would advise psying in X, where phying and psying are incompatible.
Intuitively, this is a pretty attractive view. Think of a case in which Jane thinks that Mary ought to tell a lie in the circumstances she is in, whereas Jill thinks that Mary ought not to lie in her situation. Presumably, Jane would advise Mary to lie whereas Jill would advise her not to do so. This suggests that we could use these dispositions to give incompatible advice to make sense of what the disagreement between Jane and Jill consists of. For what it’s worth, I’d want to run the order of explanation the other way. I would want to say that Jill and Jane would give different advice just because they disagree about what Mary ought to do.
In any case, the reason I am sceptical about Ridge’s view is that it is formulated in terms of a counterfactual conditional and philosophical theories that are formulated in this way usually fail because they commit my favourite objection – the conditional fallacy. The right hand side of Ridge’s account first places A and B into idealised hypothetical conditions. The theory then says that A and B disagree if they do certain things in those new circumstances. The problem is that placing A and B in the idealised circumstances changes them and so what A and B do in the new circumstances will no longer be relevant for whether they disagree in the actual circumstances.
To see the problem consider the following two debates:
Ann: Harry ought to be honest to Larry.
Ben: That’s not true. He ought not to be honest to Larry.
Mark: Kerry should to give advice to Pam.
Val: that’s not true. Kerry shouldn’t advice anyone.
Intuitively, Ann and Ben disagree and so do Mark and Val. The problem is that Ridge’s account can’t make sense of their disagreements. Ann would advise Harry to be honest to Larry. In the real world Ben is against honesty. However, when we place Ben in the idealised world we have to change him so that he too is an honest person. Presumably honest people are for honesty. So in the idealised circumstances Ben too would advise Harry to be honest to Larry. This means that the right hand side of Ridge’s view is not satisfied and so Ann and Ben don’t disagree on his view in the actual world.
The same goes for Mark and Val. Mark would advise Kerry to give advice to Pam. In the real world, Val is against giving advice but in the hypothetical idealised circumstances we have to make her willing to give advice so we have to make her to be for advising others. This means that in the idealised circumstances Val too advices Kerry to give advice. So, against the disagreement between Mark and Val disappears on Ridge’s account.
Here’s what Ridge says about worries of this sort:
“These definitions are couched in terms of conditionals. One might worry that these should be read as counterfactual conditionals, and then object that in the nearest world in which a given person would offer advice of any kind, his state of mind would be quite different. This, though, is not the intended reading. The idea is rather that we keep the states of mind of A and B fixed and ask, given those states of mind, what each of them would advise D to do, if they had to advice one way or another, and moreover had to do so honestly, candidly, and without hypocrisy of any kind (Ridge 2014, 187).”
I don’t think this response works. Firstly, I am little worried about being told that I should not read a counterfactual conditional as a counterfactual conditional. If the theory is not based on a counterfactual conditional, then why formulate it in terms of one? The second thing to note is that the fix in the end of passage just replaces one counterfactual conditional with another. It says that when we place the agents in the idealised conditions, we make no psychological changes (we keep attitudes fixed) but rather we change the external circumstances so that they have to give advice honestly. (By the way, the only way I can make sense of this requirement to give advice is that now in the idealised circumstances there is a threat: think of a demon insisting that you give honest advice or they shoot you.)
The problem is that the conditional fallacy still doesn’t go away. Consider the following case:
Imagine that Freda’s states of mind is such that she is disposed to advice people not to tell lies except in situations in which she has to give advice one way or another and she has to do so honestly, candidly and without hypocrisy of any kind. In these situations, she gets so nervous that she advises people to tell lies. Erin in contrast is always disposed to advise people to tell lies. Then, assume that we are in an ordinary situation in which Freda doesn’t have to advise anyone even if she has an opportunity to do so. In this situation, Freda and Erin have the following discussion:
Freda: Olly should not tell a lie to Polly.
Erin: You’re mistaken. Olly should tell a lie to Polly.
Intuitively, Freda and Erin can sincerely say these things given their attitudes towards lying (Freda is against Olly lying to Polly whereas Erin is for this). And, intuitively they disagree. However, the revised version of Ridge’s view doesn’t support this intuition. If Freda had to give advice honestly, she would advise everyone to lie because of her nerves (note that Ridge explicitly says that we keep their states of minds fixed – I assume this includes the disposition to get nervous when under pressure and to advice people to lie in that case). And so, on Ridge’s view, there’s again no disagreement. So the conditional fallacy is still a problem. This makes me doubt that there’s a way to makes sense of disagreements in terms of dispositions to advise in idealised circumstances – the conditional fallacy is too much of a problem.