On My View About Moral Requirements

Responding to Talis again

Talis has a new article out critiquing my view on moral requirements and reasons. While I am not explicitly mentioned in the article itself, he says in the comments “It was in my recent writing and conversation with Travis Talks that this view came to mind. He wants to reduce requirement talk to desire talk.”

Before I get into his critiques, I’ll just briefly lay out my view. Firstly, I would like to clarify my view is not meant to be a thesis about what ordinary people are doing when they engage in talk about moral reasons or requirements. It is solely an account of how I personally use this language.

I consider myself a subjectivist in terms of how I personally use moral language. If I say that someone has a moral obligation to do X, what I’m saying roughly is that I have a preference for them to do X. If I say that someone has a moral reason to do X, I’m saying that doing X would be conducive to achieving some goal of mine.¹

With that said, let’s get into Talis’ article:

Sapping Moral Requirements Of Their Normativity

Talis accuses my view of sapping moral requirements and reasons of normativity - their essential feature. But on my view, normativity itself is just reducible to statements about evaluative attitudes. So I don’t accept that my account of moral requirements and reasons saps them of normativity. If Talis means that I’ve sapped them of irreducible normativity, then I don’t think that was ever any irreducible normativity there in the first place.

This objection is a bit like saying gastronomic subjectivism is problematic because it saps gastronomic statements of irreducible normativity. I don’t find it to be a problem at all.

Talis says:

One implausible entailment of this reductionist view is that our own desires–on their own–can determine that someone else has a moral requirement to do something. Brian having a desire for Jane to save drowning children determines that Jane has a moral requirement to save drowning children.

To be clear, it’s not that Brian having a desire for Jane to save drowning children makes that it such that Jane has a moral requirement to save drowning children simpliciter. It’s that Brian having a desire for Jane to save drowning children makes it such that Jane has a moral requirement to save drowning children relative to Brian’s moral standards. I see nothing implausible about this whatsoever.

For instance, suppose someone has a desire that I choose a certain flavour of (vegan) ice cream. If we suppose that I don’t care to do what they want me to do (and there are no other considerations on the table) it doesn’t seem that I’ve got any reason whatsoever, no requirement whatsoever, to do what they want me to do.

I have no problem with saying that someone has a reason or obligation to choose a particular flavor of vegan ice cream relative to some set of evaluative standards. Doesn’t seem unintuitive to me.

I think it ultimately comes down to the view that I endorse; things like reasons and requirements always have some normative weight to them. That is, they bear down on what I should be doing in a way that is legitimately authoritative.

I would also say that moral reasons and requirements have normative weight and are legitimately authoritative. It’s just that that I think whether something holds normative weight is itself a relative matter. :P

Desires About Non-Moral Agents and Impossible Actions

Talis claims that framing moral requirements and reasons in terms of desires runs into problems when we consider desires regarding creatures that aren’t moral agents and impossible actions.

We presumably don’t think a dolphin has a moral obligation to save a drowning child even if we think it would be preferable for the dolphin to save the child from drowning. Likewise, we would find it preferable if someone flew to the top of a burning building and saved the victims. However we wouldn’t say that such a person has a moral obligation to do this if they don’t possess the ability to fly.

Our future AI overlords didn’t give me exactly what I wanted with this prompt but I’m including it anyway because I found it funny.

I am also disinclined to use the language of moral reasons and requirements in these situations. I don’t see this as a super big problem. I can just say that while statements attributing moral reasons or requirements are indeed statements about my desires, they refer to a subset of my desires which exclude desires about non-moral agents and impossible actions.

An Implication of Accurate Requirement Talk?

I think it’s common sensical to believe that if someone says “you have a moral requirement to X”, and what they say is true, then it also implies that your own statement “I have a moral requirement to X” is also true.

For instance, suppose someone tells me that I have a moral requirement to save children, and what they say is true. This seems to imply that my own statement “I have a moral requirement to save children” is also true. However, this would be mistaken under the reductionist view.

Under the reductionist view, if someone says that I have a moral requirement to save children, all they mean is that they have a desire that I save children. However, this need not imply that I have a desire to save children. In fact, under the reductionist view, when I say that “I have a moral requirement to save children”, I’m not at all talking about what other people desire–I’m talking about what I desire.

I don’t have the intuition that Talis is trying to get at here. Whether most people have this intuition is an empirical question - for which no empirical evidence is provided. Even if we suppose that ordinary people do in fact have this intuition, there are a few things to note about this.

First of all, just because this is their initial intuition doesn’t mean it would be their intuition post-reflection. There were a few entailments of subjectivism that I initially found unintuitive but stopped seeing as unintuitive once I thought about it more. Second of all, even if this remains intuitive to most people upon reflection, I don’t see how this is a problem for my view.

Recall that at the beginning of my article I said that this is solely an account of how I personally use moral requirements and reasons talk. It may well be the case that most people use moral language in the way that Talis describes. My view is not meant to be an account of what most people do when they engage in moral language. As I’ve said before, I’m largely agnostic when it comes to a semantic thesis about moral language in general.

If some subjectivist wants to claim that people in general use moral requirements and reasons talk in the same way I do, this might pose a problem for that view. But I don’t make that claim, so it’s not a problem for my view.

Moral Disagreement

Talis says:

On a related note, the reductionist view doesn’t seem to be able to account for disagreements about what moral requirements we have.

Jane: “You have a moral requirement to save children”.
Me: “I don’t have a moral requirement to save children”.

Under the reductionist view, there would be no disagreement here, as all Jane is saying is just that she has a desire that I save children, whilst I’m just saying that I don’t. But clearly there seems to be a disagreement here.

This is a common objection to subjectivism. I’ve made a video responding to this objection and I’ve also briefly discussed this objection with Talis on stream.

The implicit assumption behind these objections is that in order for something to qualify as a disagreement, there must be a proposition in contention. I see no reason to accept such an assumption.

I don’t think that in rejecting that disagreement must require a proposition in contention, I’m operating under some proprietary notion of disagreement or engaging in conceptual re-engineering of some kind - I think my view here is perfectly in alignment with ordinary usage of the term. Indeed, available empirical evidence seems to support this claim:

In recent work, however, John Khoo and Joshua Knobe (2018) have cast doubt on this assumption; their experiments indicate that subjects do not see disagreement as requiring exclusionary content—a single proposition that one party accepts and the other rejects.

Not All Desire Talk Is Equivalent To Moral Requirement/Reason Talk

Talis names some potential objections one might have to his arguments - saying the following:

It might be objected that the reductionist view only posits that talk of moral requirements is reducible to talk about our desires, and not that all talk of our desires refers to talk of moral requirements. Thus, in saying that I want someone to do something, I needn’t be saying that they’re morally required to do it. In some cases I am saying this, but not in every case.

As stated earlier in the article, I endorse this objection.

The issue I have with this objection is that it raises a seemingly insuperable demarcation problem. How could we demarcate desire talk of a moral requirement kind and a non-moral requirement kind, when they’re just essentially the same thing? To see this point, consider the following:

Talk of moral requirements is just equivalent to talk about one’s own desires, and as a result, there is no difference in meaning from moral requirement talk and desire talk. Hence the demarcation problem for those who hold the reductionist view, and wish to maintain that some desire talk is not equivalent to moral requirement talk–namely talk about us having desires that non-agents perform certain actions and inactions, and that others complete tasks they cannot accomplish.

Talk of moral requirements is just equivalent to talk about a subset of one’s own desires.

If a gastronomic subjectivist claims that gastronomic statements are reducible to statements about preferences, surely they do not mean that the statement “I prefer that I win a million dollars” is a gastronomic statement. That they only refer to a subset of statements about preferences is implicit.

So this objection seems to just be based on a misunderstand of what I meant when I said that talk of moral requirements is reducible to talk of preferences.

In conclusion, I do not find Talis’ objections to be particularly compelling. Some I don’t consider to be problems at all, others I think can easily be dealt with by just specifying what kind of desires one is talking about when they engage in moral requirements and reasons talk.

1. In saying this, I don’t mean to imply that my account of reasons is standard for subjectivists. I’m just explaining what talk of reasons amounts to on the specific construal of subjectivism that I adhere to.