Some Of Our Books

Categories

« A Searchable Database of Philosophers? | Main | Featured Philosopher: T.M. Scanlon »

May 13, 2013

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Hi Jussi,
nice post. A question to see what you are after: Are you thinking that all forms of subjectivism - there is a great variety of such views, is there not? - have the same responses available to them, and that the same objections would apply to all of them? I am asking especially since it seems to me more common to offer subjectivist views about what is good (for people) or about what people have reasons to do than of what is wrong; and so I am wondering if you have any particular authors' views in mind, or if you are thinking in terms of a general type of view (rather than the views of any particular set of authors).

Hi Sven

Hope you are well. Yes - there is a great variety of subjectivists views. I think one motivation for why people have tended to been subjectivists about good for or reasons rather then about right and wrong is that they have thought that mind-dependence is more acceptable about those notions than of morality. But, if subjectivists can give this kind of responses to the objection, then it seems to me that you could be just as well a subjectivist across the board.

Of the forms of subjectivism, I reckon this will work more for agent subjectivist views rather than patient subjectivist views. The latter probably makes the truth of the subjunctive moral claims to be a function of the different hypothetical attitudes which the individuals have in the hypothetical circumstances.

The other distinction between kinds of subjectivism is to do with what the moral utterances are about: whether they are about the speaker's attitudes or the natural properties towards which the speaker has those attitudes. I reckon these both kind of views can use the response just formulated in slightly different ways.

I didn't have a particular authors in mind, but I reckon maybe Harman's view could for example give this kind of a response to the objection. I'm sure there are many many others too.

Hi Jussi,

While I'm sure there are many more moves available to the non-subjectivist, I am under the impression that the problem of disagreement is often invoked in response to views like this. That is, I think X is wrong, you think it is right, and in spite of what both of us think, we are somehow both correct. However, one interesting question in the spirit of your post is this: could the subjectivist draw on the same sort of resources that Blackburn and Gibbard do in order to make sense of disagreement? On the face of it, it looks like they can't, but I'm not completely sure about that.

A few thoughts:
1) Wouldn't actualized subjectivism still imply a certain kind of (arguably) objectionable dependence of the *truth* of *others' claims* of rightness/wrongness on THEIR attitudes?
2) What about the epistemic possibility of your having different attitudes? Wouldn't you have to deny that kicking dogs would be wrong even if *it turned out* that I (actually, now) approved of it?
3) Related to (3) -- there's a paper by Peacocke in which he argues that any sort of mind-dependence (including the kind posited by the actualized subjectivist) is inconsistent with moral facts being knowable a priori. He thinks this is an unwelcome implication of mind-independence. FWIW.
4) I don't know if this is *morally* objectionable exactly, but isn't there something weird about the moral status of others' actions depending on the attitudes of some particular, non-idealized person. I mean, what's so special about me, now?
5) Just out of curiosity, your statement of subjectivism seems to assign the same cognitive significance to Kicking Dogs is Wrong and I Disapprove of Kicking Dogs. Is that what you want?

Andrew: Well, one thing that is special about you is that you ask quite good questions.

Isn't the question not whether or not we CAN understand B as an "internal, moralising claim.", but whether that is the ONLY way to understand it? Blackburn (and, of course, Dworkin) need to claim that we can't understand B as an external meta-theoretic claim about the nature of wrongness for their response to work. Maybe that's right. Maybe it's not. I have no settled opinion about it. The question for @Now-subjectivists is whether they can plausibly say the same thing.

Also, what Andrew said in 5.

I just wrote a reply to your reply, Jussi, that was somehow swallowed by cyberspace. (Or that seems to have been, so it if it ends up showing up, sorry for also posting this comment.) What I remarked on was the part of your response to my first comment where you said "I think one motivation for why people have tended to been subjectivists about good for or reasons rather then about right and wrong is that they have thought that mind-dependence is more acceptable about those notions than of morality. But, if subjectivists can give this kind of responses to the objection, then it seems to me that you could be just as well a subjectivist across the board." What I said was roughly the following.

I would have suspected that what might drive people to accept subjectivism about good for and reasons, but not about rightness and wrongness, is that the former matter is more personal, whereas the latter is of a more interpersonal nature - especially if you regard, as many people do, morality as regulating interaction with the help of sanctions (or different modes of holding each other responsible for things we do or not do). Morality - the people in question are likely to think - has to do with approving and disapproving things alright, but there had better be some interpersonally sharable reason outside of our state of approving or disapproving things that we can point to unless we expect people to take our subjective attitude of approving or disapproving of the things in question as a sufficient reason for regulating their actions accordingly. But those reasons we would refer to, in defending these approvals and disapprovals, would not have any need of being wholly mind-independent, but could instead (and should, it might be added, instead) be minds-depedent. (The reasons relevant in the intersubjective case could have to do with things such as what takes into account the pleasures and pains of all affected parties, what satisfies the maximum of desires of those involved, or what lets all remain in possession of agential mental properties; all of which would be minds-dependent properties, but not simply states of approval or disapproval on the part of particular moralizers.)

Dave, you say that *now*... ;-)

Hi Jussi,

Great post! This is in fact what I argue (in my book) follows pragmatically from my own end-relational theory, which is neither strictly subjectivist nor expressivist (but is, I think, a close relative to Harman's view, which may confirm one of your speculations).

I had a great conversation with Jon Tresan in Edinburgh last year, where he argued that while the counterfactual form of B is unproblematic for this kind of view, an indicative conditional does pose a serious problem; i.e. (B') 'If I don't disapprove of kicking dogs, then it isn't wrong'. (I hope I'm not misremembering his argument). I don't think this conditional is any more problematic, however. If the antecedent is false, then the meaning of 'wrong' in the consequent is fixed by my actual disapproval of kicking dogs, and so the conditional can't be true unless it is trivially true by virtue of a false antecedent. If the antecedent is true, then the meaning of 'wrong' in the consequent is fixed by my actual attitudes which don't include disapproval of kicking dogs. This implies that the conditional would be true--but I would be perfectly happy to accept this in that case, since (by hypothesis) I don't disapprove of kicking dogs. (Furthermore, I would deny that the truth of the consequent depends on the truth of the antecedent, since I would say that kicking dogs would not be wrong even if did disapprove).

However, this highlights what is probably the major objection to this kind of account, which Nick and Andrew both point out: how we report and engage with others' normative claims. In the imagined scenario I might be perfectly happy to accept the conditional as true, but you wouldn't.

Whether you're an expressivist, subjectivist, or "quasi-expressivist", you need a different account of disagreement and truth assessments. But expressivism gives the basic template: what you disagree with/report/assess is not (simply) a proposition, but (at least in part) an attitude. (That's to answer Nick by saying "yes, the subjectivist can appeal to the same resources as the expressivist"; Gunnar Bjornsson and I try to show this in our 2010 paper on disagreement).

Some responses to Andrew:
Re (3): I haven't read the Peacocke paper, but I claim that my own version of this view does recognize the existence of apriori moral facts. These are facts about what can be known apriori to increase the probability of my actual preferred end E (e.g. like E itself--hence it's knowable apriori (from my moral point of view) that E is good, for its own sake.)

Re (4): I'm sure it seems weird to a committed (robust) moral realist, but for anybody who thinks that moral status is always perspective-dependent, what's so special about you now is just that it's your present attitudes that determine your moral perspective!

Anyway, I'll be interested to see how this conversation develops. Thanks for starting it!

Nick

indeed - the disagreement problem is the biggie. I was actually writing a paper on that topic when I started to think about this dependency problem. For what it's worth, there are already interesting responses to that problem. Jackson (2008) and Dreier (2009) give one type of solution and Brogaard (2008) and Finlay & Bjornsson (2010) give another. I'm hoping that I can improve on those but this is still very much work in progress. But, yeah, you are right, nothing in this post per se addresses that problem.

Andrew,

I agree with David - you are asking great questions. Here's a first shot at them.

1 actually gets us to the disagreement problem. Consider the following exchange:
Rick: Kicking dogs is right.
Jussi: No, that's false.

I take it that the challenge for subjectivism is to explain why my response is appropriate even if Rick does have pro-attitudes toward kicking dogs. I see it as the same challenge as it is for expressivists to explain why my response is an appropriate way of denying what Rick says even if on their view moral utterances are not true or false in any robust correspondence sense.

There's different ways of doing this, but here's just one. On Brogaard's and Finlay/Bjornsson view the reference of 'that' in this kind of discussions is context sensitive and it can thus pick out different propositions in the context. In this case, my utterance does not pick out the proposition Rick literally asserted but rather the proposition that I am not against kicking dogs which I then deny by using the word 'false'. This proposition is salient in the context according to them because the purpose of moral discourse is promoting and projecting one's values.

In effect, what this response relies on is that we should understand talk about truth and falsehood too in an internal moralising way rather than in external metaethicising way. This again is the very same move as expressivists. So, again, it seems like there is parity here.

I have to think about 2 and 3 more but here's what I am inclined to say now. I would want to deny that there is such an epistemic possibility in which I don't disapprove of kicking dogs actually now. I'd like to say that I know that I am against kicking dogs. If anything I believe has a claim to be knowledge, then this seems to be one of them. Something is then epistemically possible if it's not ruled out by what I know - and thus that I am not against kicking dogs seems like an epistemic impossibility.

This makes me think that the conditional in 2 has an impossible antecedent and thus both:
(i) If it turned out that I actually now didn't disapprove of kicking dogs, it wouldn't be wrong to kick dogs.
(ii) If it turned out that I actually now didn't disapprove of kicking dogs, it would still be wrong.

come out as trivially true. If that's right, subjectivists can accept that mind-dependent views have problems with a priori moral truths but deny that their view is such.

With 4, I've tried to deny that subjectivism entails any kind of dependency so this worry might have to be formulated in some other way. Here's few responses though from the expressivist playbook:

Firstly, you can deny that there is anything 'mere' about our current actual attitudes. After all, these are the most fundamental cares and concerns we have.

Secondly, there is room in the subjectivist position for idealisation. One type of subjectivism would say that wrongness judgments are beliefs about what one's fully coherent, unified and informed version would disapprove of. This is close to Smith's view but he of course avoids subjectivism by thinking that these are beliefs about what all rational advisors would converge on to disapprove.

With 5, that's the way Schroeder describes very basic subjectivism in Being For and Noncognitivism in Metaethics. There are other forms of subjectivism which deny the sameness of cognitive significance (you migth say something here about different modes of presentation and the like). But, the simpler form of subjectivism was sufficient for making the current point.

Jack

that seems right to me, but I do not see what the difference between expressivism and subjectivism could be on that matter. If Blackburn is right that we cannot leave behind our evaluating perspective, then I don't see why subjectivists couldn't say the same thing.

Hi Sven

well, there are probably many reasons to be a subjectivist. Judgment-internalism might be one for example.

In any case, if you think of the interpersonal role of ethics, this might motivate you to be a relativist of some sort where the content of moral beliefs isn't one's own pro-attitudes but rather those of the majority of a certain group. The thing is that these views too face a similar mind-dependence challenge. You might think that they would entail the truth of a claim like if we all had different attitudes, then some other things would be wrong. But, if I am right, the relativist can respond to this challenge in the same way as the subjectivist.

Hi Steve

thanks! When I was thinking about this, I had a nagging doubt that someone had probably already thought of this and the likelihood was that it was you. I did read your brilliant book some time ago so maybe that's where you got this in some subconscious way. Should have gone back to check. Sorry in any case.

In any case, I agree with what you say about the indicative conditionals (I don't see them as a worry for the reasons you point out) and the disagreement problem. I've enjoyed recently thinking about your and Gunnar's paper even if I think I can perhaps slightly improve on that account of moral disagreements in the subjectivist framework but that's another story.

Steve, when Jon said there was still a problem with indicative conditionals, did he mean material conditionals? Or have I misunderstood your reply?

Many thanks for this second reply, Jussi. I wasn't thinking so much in terms of the semantics of "wrongness" but rather about the practical functions served by moral discourse, or the practical use of the moral concepts, which was why I didn't myself mean to bring up relativism (thought of as a sort of semantic theory) when I spoke of morality as something of a more interpersonal nature in contrast to more intrapersonal thoughts about good for and (personal) reasons. But that, of course, took my comment off topic since you are here more interested in the semantic side of things. Anyway, many interesting points are popping up, and I am following the thread with great interest.

Jussi, no worries. I doubt you read my chapter on disagreement anyway, since I haven't circulated it. And I tend to agree that it's possible to improve on what Gunnar and I argued. I'm now more inclined to put some weight on the possibility of semantic blindness, in explaining responses like "that's false"--which Gunnar doesn't particularly like.
BTW, another (earlier) source for this kind of view is a couple of papers (2004, 2009) by Denis Robinson, which I only learned about recently. And then there's von Fintel & Gillies' idea of a cloud of propositions.

Jamie, as I recall, Jon emphasized that they were indicative rather than subjunctive conditionals. I'm not sure/don't recall how he preferred to interpret indicative conditionals (other than to distinguish them from subjunctives), which is why I considered the material interpretation.

Steve - yes, I'm working on trying to apply the von Fintel & Gillies line to subjectivism and relativism in ethical discourse but we'll see where this gets me.

Steve, okay. But indicative conditionals do still look like a problem, to me.

I am fairly sure that I do indeed disapprove of kicking dogs. But suppose you somehow make me a little worried about whether I do or not. (Maybe the subjectivism in question adverts to a somewhat idealized version of my pro-attitudes, or you get me to read confusing stuff about introspection.) I think I would pretty confidently assert,

Even if I don't disapprove of kicking dogs, it's still wrong to kick them.

Don't you think?

Jamie

quick question (I should revise indicative conditionals). I thought one alternative way of thinking about them was the suppositional theory. So, on this view (I hope I get this right), you can assert 'if A, then B' if when you suppose that A you think that B is likelier than not.

So, applied to your case (and perhaps this will work in Andrew's case too), I assume the epistemic possibility that I do not disapprove of kicking dogs. I then ask, given this assumption, is it more likely than not that kicking dogs is wrong?

I take it that in your case you still believe with high degree of credence that you are against kicking dogs. Wouldn't this make assertible that assuming that I do not disapprove of kicking dogs, it is still wrong to kick them? This would make it assertible that even if I don't disapprove of kicking dogs, it's still wrong to kick them.

Maybe I am missing something. I worry that this response collapses indicative conditionals to subjunctive ones. I guess here is the question:

When we are assuming that I don't disapprove of kicking dogs I am asking is it still wrong to kick them. As an subjectivist, do I have to ask this question from the first-personal perspective of the assumed situation in which I don't disapprove of kicking dogs or can I ask it from the perspective in which I believe (perhaps with not fully certain and even perhaps falsely) that I am against kicking dogs? If the latter is ok, then I don't see the problem.

Jussi,
I am finding this confusing. As I understand what you are suggesting, it seems like it's what an expressivist would suggest. I don't really see how a subjectivist can deliver the 'right' answer, applying the suppositional test.

Sorry Jamie - for not being clear. Here's just one question. To evaluate the probability of the conditional, we are asked to suppose that A. In this case, this is 'I don't disapprove of kicking dogs'. I am wondering what is that we are presupposing when we suppose this.

Is this first-personal? Do I proceed from my deliberative perspective *as if* I didn't disapprove of kicking dogs and then proceed to ask whether kicking dogs is wrong?

Or, is this third-personal? Am I looking over a version of myself who doesn't disapprove of kicking dogs when I consider the epistemic possibility in which I don't disapprove of kicking dogs and then proceed to ask whether kicking dogs is wrong in that scenario I am looking at in which I don't disapprove of kicking dogs?

If the presuppositional test works in the first way, then I agree that this is trouble for subjectivists. If it works in the second way, then it seems like if you have even a partial belief that you disapprove of kicking dogs the right indicative mind-independence conditional becomes warrantedly assertible.

It's true that the latter test sounds like what expressivists are thinking when they do subjunctive conditionals. But, then it seems to me that the subjectivists could understand the suppositional test for indicative conditionals in the latter way too. But, I'm sorry, things are getting murky here.

One last point. I think whoever is pushing the indicative conditionals objection would be required to say something about how they understand indicative conditionals. It would then be for the subjectivists to investigate whether the mind-independence conditional would come out as true or false on their view (and also how damaging this would be).

Jamie: okay, good point, I concede this. Perhaps then the indicative conditional objection is telling against the simple subjectivist/expressivist version of this strategy. It doesn’t seem possible that a speaker could coherently harbor serious doubts whether p while sincerely asserting something whose felicity assumed that p. I think my own version of the view –combining a relational semantics with a subjectivist pragmatics – can explain your datum (and avoid Tresan’s objection), but when I tried to articulate this, I realized that my theory seems to succeed only insofar as it diverges from the kind of simple view that Jussi was asking about, which perhaps would be too tangential for this post.

So in short, we can accommodate your datum if we allow either /both that the attitude tied to ‘wrong’ is distinct from mere disapproval of the action (e.g. concern for some more basic end—like that you don’t inflict pain on innocent creatures), and that in some contexts ‘wrong’ might not be tied to a speaker’s actual current attitudes. (Things are rather hectic right now, with our workshop coming up next week, so I might be a bit slow defending/ explaining myself).

Jussi,

I might be having a concentration problem, but I can't quite follow what you're saying. I think I understand the first-person version, but as you describe the third-person version it sounds more like a counterfactual conditional. And although I do think something like that first-person version of indicative conditionals is sometimes in play, I don't think I can get that reading on this one. Not sure, but off hand I can't get it.

I think what I do when I assess the indicative conditional is this. I suppose, form the supposition, that I do not disapprove of dog-kicking. Okay, this is pretty odd, since I figure I know what I disapprove of, but perhaps I am mistaken, so, fine, I form the supposition. Now, having formed this supposition -- "adding it hypothetically to my stock of knowledge" -- I ask myself, "Is dog-kicking wrong?" And I answer, of course it is.

I disagree with you on the last point. I think we can all understand indicative conditionals, except possibly some very exotic ones that we need help with. The critic pushing the objection can just use the indicative conditionals that we already know how to use, and offer no particular theory of them. The critic can say, "this indicative conditional is pretty plainly false, but your theory appears to render it true." That's the objection.


Hello Jussi

Great post and comments all around. I am still stuck on subjunctives. You claim that actually rigidified speaker subjectivism is not committed to subjunctive conditionals like:

B) If I didn’t morally disapprove of kicking dogs, then it wouldn’t be wrong.

In your original post, you say that “there’s no reason for the subjectivists to claim the consequent reports my attitudes toward kicking dogs as they are in the hypothetical situation. Rather, it can report my current actual attitudes toward dogs being kicked in those circumstances.”

I do think there is a reason the consequent should report your hypothetical/counterfactual attitudes, given the counterfactual non-disapproval of kicking dogs described in the antecedent. On what seems to me an ordinary understanding of counterfactual conditionals, there should be a relevance relation between the antecedent and the consequent. The consequent should describe what would be the case, supposing that the counterfactual state of affairs described in the antecedent is the case. This relevance relation explains why a conditional like this sounds false: “If the Eiffel Tower were in New York, then it would be in France.”

My sense is that it changes the subject to interpret the consequent of B) as reporting your current actual attitudes. The consequent of B) is a claim about what the moral status of kicking dogs WOULD BE, given your counterfactual non-disapproval of kicking dogs. The subject matter of the consequent, as I hear it, is not a claim about what the moral status of kicking dogs actually is, given your actual moral disapproval of kicking dogs. But if your moral attitudes were different, doesn’t your view force you to concede that the moral status of kicking dogs WOULD BE different too?

Hi Andres,

well, I would want to answer no! As far as I can see, your comments brings out two points nice: one is that the antecedent has to be relevant for evaluating the consequent in the subjunctive conditional and the second is that we must be able to draw the distinction between claims about what would be wrong and what is wrong. It seems like the proposal I gave can do both of these things and yet say no to your last question.

There are two evaluated situations. In one, I disapprove of kicking dogs (this is the actual world) and in the second I do not disapprove of kicking dogs (this is the world which the antecedent of the subjunctive conditional asks us to consider).

If I ask, 'Is kicking dogs wrong?', I want to know is it wrong in the actual world. On the subjectivist view, this is a question of do I disapprove kicking dogs when it happens in this world in which I disapprove kicking dogs? The answer to this question is clearly yes. Note that I cannot use this question to ask about the other world. I cannot ask 'If I didn't disapprove of kicking dogs, is kicking dogs wrong?'.

So, the question 'I didn't disapprove of kicking dogs, would kicking dogs still be wrong?' directs me to evaluate the second situation in which I don't disapprove of kicking dogs. This must be the case for the relevance condition to be satisfied. Because I am asking about the wrongness of kicking dogs in this different situation, I am not asking whether kicking dogs is wrong. So, there is a difference between would and is here. But, still, I do now disapprove of kicking dogs in those circumstances which I can report by saying it would be wrong to kick dogs even if I didn't disapprove of kicking dogs. I could not express this thought by saying it is wrong to kick dogs even if I didn't disapprove of kicking dogs.

So, it seems like this kind of subjectivist view doesn't have to concede that the moral status of kicking dogs would be different if I had different attitudes even if (i) the relevance condition is satisfied and we are evaluating the hypothetical situation of the antecedent and (ii) we are asking would kicking dogs be wrong in that scenario rather than asking is kicking dogs wrong. Thus, it seems to me like there is logical space for this kind of combination of subjectivism and mind-independence.

Hi Jamie

thanks. This is extremely helpful and clarifies where the goal-posts are. I have to think more about how indicative conditionals work in first-personal cases more broadly and how the subjectivists could give a subjectivist interpretation of the 'of course it is' statement which comes out as a result of your clear test for the indicative conditionals.

The comments to this entry are closed.

Ethics at PEA Soup

PPE at PEA Soup

Like PEA Soup

Search PEA Soup


Disclaimer

  • Unless otherwise indicated, the views expressed in any given post reflect the opinion of only that individual who posted the particular entry or comment.