Friday, May 6, 2011

What Is Desirism?

In one sense, this is an easy question: Desirism is the moral system of Alonzo Fyfe as promoted on his blog1 and in his collaborative podcast with Luke Muehlhauser.2 At the same time, it's a hard question because Fyfe's views have changed significantly over time and no definitive sketch has been kept current. My purpose here is to sketch what I think Desirism is all about, and update this post when my own understanding of his moral system changes.

The following sketch will be in my own words, since I'm interested in the ideas, not Fyfe's phrasing or style of explanation. Nor should anyone take this as anything more than a bystander's attempt at describing Desirism. I am not speaking for or against the system here.

Current version: 1.3.2 — last updated May 10, 2011

Explaining Behavior

Proponents of moral systems tend to start with some aspect of morality which they have an insight about, then they try to account for other aspects by building on this core. Desirism's starting point is an explanation of human behavior, or as philosophers would say: an action theory.3

Action theory is relevant to moral theory, since the question of how voluntary action works can bear on the question of which actions, morally speaking, we ought to perform.

I won't get into all the details of Desirism's action theory, but basically: a person's attempted actions are determined by their desires and beliefs. If I want a cookie and I believe a cookie is in the left jar instead of the right jar, I will try to open the left jar...unless I have some stronger desire which outweighs my cookie desire.

This might sound a tad obvious and silly, but it is a substantial claim. It's opposed to the notion that we can act 'from duty' against all our desires, which is how Kant defined moral action.4

Shaping Behavior

It follows from the above that if we want someone to make a different choice in a given circumstance, we must either change her desires or her beliefs.

From what I know of Fyfe, he doesn't advocate making stuff up to shape the beliefs of those who buy into it. Instead, he focuses on the process of social praise and condemnation to shape the desires of those affected by such peer pressure.

Suppose most people in a population don't like rock music. Heck, they don't even like thinking about other people listening to rock music. What's going to happen? Some of them are going to condemn rock music and, probably, some others who like rock music will start to like it a little less. (This may backfire with teenagers.)

I see a parallel here with biological evolution as a way of describing the "change in the frequency of alleles within a gene pool from one generation to the next."5 Desirism can be seen as a model for the dynamics of held-desires in a population over time.

Judging Behavior

What does it mean to say an action is 'morally wrong' or 'morally right'? This is where Desirism gets fuzzy, in part because the detailed podcast series hasn't gotten around to addressing this yet. Here's how Fyfe seems to handle normativity:

When we say a person 'ought' to do X or 'ought not' to do Y, we are implying there is a reason for taking (or not taking) an action. According to Desirism, desires are the only possible reasons for taking an action. (At least when mistaken belief isn't complicating things.) And therefore, strictly speaking, 'Sam ought to do X' is only true if X properly corresponds to Sam's desires.

Wait, what?! Doesn't this throw moral right and wrong out the window entirely if Sam only 'ought' to do whatever conforms to her personal desires?

Not so fast. Desirism locates normativity at a different level than usual. Instead of focusing on what agents ought to do in a given situation, Desirism focuses on which desires agents ought to have. The colloquial assertion 'Sam ought to do X' is taken as an elliptical form of, 'Anyone ought to have Z desire[s], which would lead Sam to do X.'

What are these desires 'anyone ought to have' and how can we justifiably claim people 'ought' to have them? We can answer both at once: the desires 'anyone ought to have' are those which — when held by individuals — fulfill other desires more so than thwart other desires. Just as the desires an individual happens to hold justifies her taking desires-fulfilling action, the full set of desires held by desiring beings justifies their holding desires-fulfilling desires.

Of course there is one big difference between these kinds of 'ought.' An individual will invariably do what she 'ought' according to her own desires. That's just how people work. But an individual might not hold the desires she 'ought' to hold according to desires considered more generally. Furthermore, she's not going to be intrinsically motivated to hold these — as Fyfe calls them — good desires. She must be extrinsically motivated, and this happens through a process of peer pressure, as explained in the Shaping Behavior section above.

When we say a person did 'wrong,' we're applying social pressure against anyone listening to not desire doing the condemned thing. Vice-versa for calling an action 'right.' According to Desirim, there is an important sense in which such judgments can serve more than just an emotive role: they can be correct or incorrect insofar as the desires being encouraged (or discouraged) facilitate other desires.

4. See about halfway through the First Section of Kant's Groundwork of the Metaphysics of Morals.
5. From


  1. "An added complexity is that the strength of the individuals' overall desires counts at the population level too. Fyfe would presumably not condemn a practice if 60% of the population has weak overall desires which are thwarted by the practice, but 40% of the population has very strong overall desires which are helped along by the practice."
    This is not desirism but desire fulfilment act utilitarianism (DFAU). There is no utility to maximise in desirism, that is why we no longer call it desire utilitarianism, which was misleading, IMHO.

    So it is not "the right act is the act which maximises desire fulfilment" - DFAU - but, my version, "the right act is the result of desires the people generally have all-things-considered and all-things-being-equal reasons to promote and not inhibit".

    Alonzo thinks that "generally" captures my two "all-" clauses but without them I think it is all to easy to misread desirism as DFAU. So I make them explicit all-things-considered (i.e. do not exclude desires of one's out-group) and all-things-being-equal (i.e discount any particular distribution of desires in the population to find the invariant desire-desire relations).

  2. Excuse this being off topic Garren, but after reading a few of your posts, I'm interested in what sort of metaethical system you personally support. Do you have a particular post(s) you can point me to that describes your views?


  3. I am eternally confused about what the desirist "good" is. The utilitarian account in your OP does not seem to fit Fyfe's often-raised rapist-world scenario, where he seems to indicate that "good" desires are desires that further other desires, in a one-one (desire-desire) analysis. What these are depends on what other desires are extant, so you have to consider the desires that exist. And the weirdest step of all is to then say that the "good" act is an act out of good desires; but then again, just as often it seems the good act is an act to spread and further good desires.

    I find this line of reasoning rather pointless. I liken it to creating a God - the desirists imagine a set of desires, a Will, and act to further His (non-existent) desires. That's... just bizarre. I like Russell Blackford's metaphor of different metaethical positions providing different tools, that are useful for different purposes; I'm not sure what purpose that definition is good for...

    However, I'm not at all certain that I "get" desirism. So I'm waiting to see Fyfe et al finally state things in a clear manner. I don't expect that to happen anytime soon. As I doubt that their definition would prove useful, I'm not too worried that it's gonna take them a few more years...


  4. @faithlessgod,

    I'm aware that Fyfe would object to the above characterization as a kind of 'act utilitarianism' (as faithlessgod pointed out). However, I have yet to understand what the current, official alternative might be. I've heard wildly different answers.

    Maybe I'll just put in multiple options for now. I certainly welcome your feedback or reading recommendations.


    I plan on doing a one post summary of my metaethical view soon. The super-short version is that I argue all proper uses of words like 'good' and 'ought' are relative to an end (or goal). Next, we can look at real world moral judgments to determine which ends are currently being used as moral ends. At this point we can either accept this status quo, or choose to reform which ends we count as moral ends. There can be no fact of the matter about which ultimate ends we ought to count, because ultimate ends are prior to 'ought.' I do advocate reforming the de facto list of moral ends, but for reasons external to metaethics.

    Recommended posts for now:

  5. Damn blogger ate my comment again I cannot use this comment system here. there is something wrong wiht it that I have not seen in other blogger comments.

  6. Apparently there are some third party comment solutions which can be put in place of the Blogger default. I'm looking into that since I've had a few complaints.

  7. Replying to the May 10th, 2011 revision...

    Of course there are a million qualifications and explanations and arguments I'd like to add, but what you've said appears to be a correct interpretation of some central parts of desirist theory.

    Note that my series on Less Wrong will probably get to some of the metaethical issues that interest you before the podcast does:

  8. @Luke,

    Heh, I'll take that. I am following your No-Nonsense series with curiosity.


    What I should have answered earlier is that I have a really hard time disagreeing with anything Stephen Finlay says about metaethics. There are things he says I'm not sold on, but I'm not sold on any conflicting position either.

    Wish I could find more criticism of his views.