Morality in the Real World 12: Action-Based Theories of Desire

by Luke Muehlhauser on February 15, 2011 in Ethics,Podcast

In episode 12 of Morality in the Real World, Alonzo Fyfe and I explain the theory of desire at the heart of a desirism, a variety of action-based theory of desire.

Download Episode 12

You can also listen to this podcast at, or subscribe in iTunes or out of iTunes. See the full list of episodes here.

Every five episodes we answer audience questions, so please do post your questions and objections below. Make sure your questions address the topics of this episode only. If we plan to address the subject of your question in later episode, we will not answer it in the next Q & A episode. You can also leave your question in audio and we will play it back during the Q & A episode and respond to it: call 413-723-0175 and press 1 to leave a voicemail.

Transcript of episode 12:

ALONZO: Hey Luke.

LUKE: Hey Alonzo.

ALONZO: So… this is the second of ten episodes where we look at that entity on which desirism is built: desire.

LUKE: Right. In our last episode, we made a number of claims that we said are true about desires – that desires are propositional attitudes, that they motivate action, that agents act so as to fulfill the most and strongest of their desires given their active beliefs, that reason is the slave of desires, and so on.

ALONZO: That we did.

LUKE: Okay, I have a question about that. Are these claims simply something that you made up and threw out there because you thought they would make a neat moral theory? Is this Alonzo Fyfe’s theory of desires?

ALONZO: No, it’s not. These claims make up what is called the standard view of desire in the philosophy of mind.

LUKE: Okay. But, the standard view is not the only theory. There are a lot of competitors – including the theory that beliefs and desires don’t exist at all. The standard theory about beliefs and desires isn’t even a clear winner. It’s not two laps ahead of the competition or anything like that. It’s ahead by, maybe, a couple of lengths and the race has just started.

ALONZO: Okay, that’s true. But it’s still the standard view, and even a leading opponent of that theory, Timothy Schroeder, calls it “the standard view.” It’s not an intellectual crime to go with the opinion that is the most widely accepted among experts in the field.

LUKE: The way Schroeder puts the standard view, it says:

For an organism to desire P is for the organism to be disposed to take what actions it believes are likely to bring about P.

ALONZO: Okay, let me explain this a bit: When philosophers talk about P that way, they mean P to stand in for a proposition like “Israel and Palestine are at peace.” So the standard view would say that for an organism to desire “Israel and Palestine are at peace” is for the organism to be disposed to take what actions it believes are likely to bring about “Israel and Palestine are at peace.” Right?

LUKE: Right. Another example of an action-based theory of desire comes from Robert Stalnaker, who says,

To desire that P is to be disposed to act in ways that would tend to bring it about that P in a world in which one’s beliefs, whatever they are, were true.

ALONZO: That’s the standard view. What we said is:

Claim Number 3: A desire-as-end is an attitude towards a proposition that causes an agent to seek to create or preserve states of affairs in which the proposition is true.

This making or keeping a proposition true fills the same role as “bringing about P”.

LUKE: Right.

ALONZO: We also said:

Claim Number 5: An agent intends to act so as to fulfill the most and strongest of the agent’s desires-as-ends, given the agent’s active beliefs.

LUKE: Right. And that fits the claim that desires motivate agents to act in ways that they believe will bring about P. Agents would like to bring about P, but false beliefs sometimes get in the way and cause them to fail to bring about P.

ALONZO: I still have some concern about what Schroeder means by the phrase, “to take actions”. I want to make sure that we are not including such things as reflex actions and involuntary muscle spasms as “actions” in this sense. We’re talking about intentional actions here.

LUKE: Of course.

You know, a lot of this actually traces back to the 18th century Scottish philosopher David Hume.

ALONZO: Well, I think that Hume had most of this stuff right. This is Hume’s theory of intentional action. What Hume missed that provided the key to my way of thinking is the idea that desires are propositional attitudes. Desires motivate agents to make or keep true particular propositions.

LUKE: Yeah, Schroeder writes about these propositional attitudes in his article. He writes:

According to most theories, desires are always desires for conceivable states of affairs. A desire for tea is a desire for a certain state of affairs one has in mind: that one drink some tea. A desire for a new pair of skates is likewise a desire for another state of affairs: that one own a new pair of skates. And so on. This idea is also expressed with phrases such as ‘desires are attitudes toward propositions’ or ‘desires have propositional content’.

ALONZO: Right. That’s what is meant by the claim that desires can be expressed in the form of an attitude towards a proposition. An agent has a desire that “Israel and Palestine are at peace” or, in Shroeder’s example, an agent has a desire that he own a new pair of skates.

We also make the claim that beliefs can be expressed as propositional attitudes. That was contained in our first two claims:

Claim Number 1: Beliefs and desires-as-ends exist as propositional attitudes.

Claim Number 2: A belief is an attitude towards a proposition that takes the proposition to be true.

LUKE: Schroeder doesn’t specifically talk about beliefs in his entry on desires. But a quick visit to the Stanford Encyclopedia of Philosophy entry on Belief shows that it says:

Most contemporary philosophers characterize belief as a “propositional attitude”.


Contemporary analytic philosophers of mind generally use the term “belief” to refer to the attitude we have, roughly, whenever we take something to be the case or regard it as true.

So again, we’re just using the standard view. It’s not the Alonzo Fyfe Theory of beliefs and desires.

ALONZO: Nope. I don’t claim to have made any progress at all in the theory of desires. I am simply using the ideas that other people have presented . . . ideas that represent the most widely accepted view of what desires are, but, as we said, not the only view.

LUKE: Right. And one of the objections that might be raised against the standard theory is that I don’t always act on my desires. We mentioned in the last episode that I may want some chocolate cake, but I may go to the fitness center instead.

ALONZO: Well, yes. But that’s not actually a problem with the theory. Remember Claim Number 5:

An agent intends to act so as to fulfill the most and strongest of their desires-as-ends, given their active beliefs.

Your desire for chocolate cake might not qualify for membership in the set of “most and strongest.” As such, it gets sidelined – overruled by the most and the strongest desires. You will still feel a tug towards the chocolate cake, showing that the desire is there and working, but you will not act on it unless there is a way of doing so as a part of the set of most and strongest desires.

LUKE: So, that’s not a problem for action-based theories of desire. What Schroeder says is:

The strength of a desire is typically said to be constituted by the desire’s causal power regarding the control of action: for one desire to be stronger than another is for the agent to be disposed to act upon it, rather than the second desire, in a situation in which (a) all else is equal, and (b) the agent believes that each desire is satisfiable by a distinct action, and (c) the agent believes that the desires are not jointly satisfiable.

ALONZO: Right, there, so: are we done?

LUKE: Not even close.

Another claim we made was:

Claim Number 7: In addition to desires-as-ends, there are desires-as-means: desires for states that, given one’s active beliefs, will lead to the fulfillment of an end. Desires-as-means are constituted by desires-as-ends and beliefs.

ALONZO: Yep. You’ll find Schroeder expressing that claim in his entry on desire as well:

Some desires are for states of affairs that are wanted for themselves: these are intrinsic desires. It is generally agreed that pleasure is desired for its own sake, and it is plausible that many people also desire the welfare of their children, the success of their favorite sports teams, and the end of injustice, and desire them all intrinsically.

LUKE: And he continues:

Normally, however, one calls a desire ‘instrumental’ when one means. . . that the end is desired merely as a means to some other end, and not at all for its own sake. . .

ALONZO: Yep. Now, I think we need to explain that what Schroeder calls intrinsic desires, I call desires-as-ends. What he calls instrumental desires, I call desires-as-means. It appears that Schroeder’s terms are becoming the jargon of those who study this area, and we may have to adopt them for that reason. However, I find them confusing.

LUKE: Yeah. We talked about that in our first season. “Instrumental desire” sounds like it should mean “useful desire”, but it doesn’t. It means a desire for something insofar as the thing that is desired is useful – not insofar as the desire itself is useful.

And “intrinsic desire” sounds like the desire itself is intrinsic to the agent – it is a part of the agent’s makeup. But instrumental desires are also a part of the agent’s makeup. There’s no difference about that. So I think talking about desires-as-ends and desires-as-means is less confusing.

ALONZO: Well, I could tell you a story that might make some sense of Schroeder’s terms, if you are interested.

LUKE: Is it a good story?

ALONZO: You’ll have to tell me.

A long time ago in ancient Greece, philosophers recognized a difference between things valued as ends and things valued as means. They assumed that ends had value in virtue of their intrinsic properties, and they called these “intrinsic values”. Things valued as means they called “instrumental values.”

As the idea that value is grounded on desires gained traction, these concepts of intrinsic and instrumental values turned into talk of intrinsic and instrumental desires.

LUKE: That is a good story, but I’m not convinced that you can prove that theory about history.

ALONZO: Probably not. But it helps our understanding of the link between intrinsic and instrumental desires as Schroeder and others talk about them on the one hand, and what has been traditionally known as intrinsic and instrumental value on the other.

Either way, in the Stanford article, you will find the distinction we make between desires-as-ends and desires-as-means expressed as a distinction between what Schroeder and others call intrinsic and instrumental desires.

LUKE: Now, there’s an important point about these claims that I think needs to be addressed. In presenting the standard view for the Stanford Encyclopedia of Philosophy, Schroeder raises an objection:

[A] number of philosophers have suggested that desires are only one psychological state able to initiate action, so that it is a mistake to identify desires with psychological states disposing us to actions. Some of these philosophers have focused on the negative point, that what might be called ‘true desires’ do not exhaust the possible motivational states.

ALONZO: Okay . . . first, this notion of “true desires” sets off a few alarm bells. This sounds like we are heading towards an application of the No True Scotsman Fallacy.

LUKE: Okay. That’s a fair warning. For the sake of our listeners, the No True Scotsman Fallacy is a way of trying to salvage an argument by restricting the definition of some term. For example, I might say that no Scotsman would ever be a Nazi. You show me a picture of a Scotsman who is a Nazi, so I answer, “Well, no true Scotsman would be a Nazi.” What I have simply done is redefined the term Scotsman – I have made the definition a lot narrower – and made it true by definition that a so-called ‘true Scotsman’ cannot be a Nazi.

ALONZO: I am worried that if somebody starts to claim that no ‘true desire’ would motivate an agent in a particular way that he might be pulling the same move.

For example, Wayne Davis distinguishes between two types of desires: volitive desires and appetitive desires. He then takes one of them to be ‘true desires’ and casts the other out of the realm of ‘desires’. It no longer belongs. After he casts it out, he says that there are things other than “true desires” that are capable of motivating action. However, this is just semantics.

LUKE: Is that really what he says?

ALONZO: Yes. Note that, what Davis was arguing for was a definition of desire that best accounts for our intuitions as to its use. His conclusion was that our use of the term ‘desire’ is one that makes this distinction, and is a use that allows that some entities that are not ‘true desires’ to motivate action. This is how we talk, according to Davis.

LUKE: That brings up the point I wanted to talk about. A lot of philosophy is concerned with something called “conceptual analysis,” where the goal is to create a kind of Super Dictionary made up of definitions for terms that not only capture every nuance of people’s intuitions about the meaning of each term and our actual practice of using that term, but definitions that are also logically consistent and, if they’re meant to refer to things in the real world, do actually successfully match up with the real world.

ALONZO: Right. Whether or not a theory of a term passes muster is determined by whether or not the theory provides this best account of use, consistency, and reference.

LUKE: Yeah: use, consistency, and reference. And that’s why you’ll read all kinds of arguments about how, for example, we can intuitively imagine a distant planet where water isn’t made of H2O but XYZ and yet we would still call it ‘water’, and so that has implications for what the best SuperDictionary definition of “water” is, and so on.

But I’ve gotta say, that’s not my project. I’m not doing philosophy to create a Super Dictionary that captures every nuance of what people’s intuitions say about the meaning and use of English terms.

ALONZO: Me either.

LUKE: I am more than happy to scrap a common usage of a term in favor of a definition that lets us talk about things in the real world more clearly and precisely.

This is what happened, for example, when we stopped talking about dolphins as being a kind of fish, because we realized that reality isn’t actually carved up that way. Dolphins have more in common with cows than they do with carp. And so even though we had always talked about dolphins being fish, and that’s what our intuitions said, we changed the way we use that term “fish” because we can communicate better when the borders around the meaning of our terms roughly match up with borders that exist in reality.

So I’m not actually looking for a theory of desire that helps us create a Super Dictionary’s entry for the term ‘desire’ that, you know, fits the world and captures all our intuitions about its meaning. I’m after a theory of desire that helps us predict and explain the world around us, even if we have to throw out some of our intuitions about desire because those intuitions turn out not to match reality very cleanly.

So back to Davis, if Davis’s two types of desires – volitive and appetitive – are both propositional attitudes that cause an agent to act so as to make or keep true the proposition that is the object of that desire, then, strictly speaking, desirism doesn’t have much of a reason to care whether Davis is right or wrong about these two types of desires – and we have no reason at all to care whether he calls one of them ‘true desires’ or not.

If he wants to talk about desires in that way, that’s fine, and I can just translate his language about desires into my language about desires, and we won’t necessarily be disagreeing about any matters of fact. We’ll just be talking slightly different languages about the same phenomena.

ALONZO: So, a substantive objection to desirism would have to take the form of a claim that there are states that cause an agent to intentionally act a particular way, but that state is not a propositional attitude, or it is a propositional attitude that somehow motivates an agent but it doesn’t move the agent to make or keep that proposition true.

We haven’t established – and we are not even going to try to establish – that an action-based theory of desire represents the best Super Dictionary definition of ‘desire’ – one that captures all our intuitions about the meaning of ‘desire’ and so on.

What we’ve presented it as a theory that gives us the best hope of explaining and predicting the actions of intentional agents in the world around us. The two questions to answer are: (1) Is there a different way of carving up conceptual space that makes our communication with each other more effective? And (2) How well does it do at explaining and predicting events in the world around us?

LUKE: Right. And, I would like to tackle that second question first. Tim Schroeder’s book, Three Faces of Desire, has a section in which he identifies eight objections to action-based theories of desire. I think it would be useful to look at those objections and categorize them in terms of “inadequate Super Dictionary definition” types of objections versus “failure to explain and predict intentional action” types of objections, and see what we can do to answer that second type of objection.

ALONZO: I can do that.

LUKE: Alright, talk to you next time!

Audio clips

(in order of appearance)

  • “Hour Five” from Somnium by Robert Rich
  • “Whipping the Horse’s Eyes” from A Feast of Wire by Calexico
  • “Butterflies & Hurricanes” from Within by William Joseph

* marks royalty-free music. With copyrighted music, we use only short clips and hope this qualifies as Fair Use. Fair Use is defined in the courts, but please note that we make no profit from this podcast, and we hope to bring profit to the copyright owners by linking listeners to somewhere they can purchase the music. If you are a copyright owner and have a complaint, please contact us and we will respond immediately. The text and the recordings of Luke and Alonzo for this podcast are licensed with Creative Commons license Attribution-Noncommercial-Share Alike 3.0, which means you are welcome to republish or remix this work as long as you (1) cite the original source, and (2) share your remix using the same license, and (3) do not use it for commercial purposes.

Previous post:

Next post:

{ 8 comments… read them below or add one }

Yair February 15, 2011 at 11:31 am

Since you will consider objections to this standard view in the next episode, it seems inappropriate to raise ones at this stage. Instead I will only note that you aren’t applying what your’e defining.

According to the letter of your definition, a person can only have one or at most a few desires at any one time – the desires that actually caused him to act.

Claim Number 3: A desire-as-end is an attitude towards a proposition that causes an agent to seek to create or preserve states of affairs in which the proposition is true.

A paralyzed man, by this definition, will have no desires. There are no ineffective desires; at any given time, only the “succesful” desires exist. There is no strength to desires; a desire simply exists or not. And so on.

Now, this is clearly not the desires you’re using in practice. But that means that you aren’t really applying the definition you put forward. You’re using something else, which you don’t explicate. Whatever it is, I think it’s fine. Definitely seems to be enough to proceed to analyze intentional actions. Now, can we move to metaethics already?


Luke Muehlhauser February 15, 2011 at 11:33 am


We did not say that desires cause an agent to act. We said that desires cause an agent to seek to act. This seeking to act can be defeated – for example by stronger competing desires, or by physical paralysis.


Peter Hurford February 15, 2011 at 2:20 pm

What do you make of the desires-as-means that I would still prefer not to do? For example, I have a desire to get an “A” in my Calculus class. This means that I have to study, so I acquire a desires-as-means (studying) to get at my desires-as-ends (“A”)*.

But, all else being equal, I would much rather get an “A” without having to study, because I have no actual desire to study Calculus. So I’m desiring something that I do not really desire.

My personal intuition is that “desire” is really two different words: Desire{1} means that “this is my reason / motivation for doing a given action” and desire{2} means that “all else being equal, I would still rather do this”. In this sense, I desire{1} to study because it lets me get the “A” that I desire{2}. But I do not desire{2} to study.

One way to solve this distinction is that we call everything I desire{2} a “desires-as-ends” and we call everything that I desire{1} but do not desire{2} a “desires-as-means”.

However, this would seem to redefine the relationship between desires as it is possible to desire{2} studying for its own merit of learning (I feel this way in politics, economics, and philosophy, but not math), making it not a “desires-as-means” to getting an “A”.


*: Whether or not getting an “A” is truly a desires-as-ends under your definition isn’t really the point.


Graham Robinson February 16, 2011 at 2:31 am

(I posted this in the wrong place)
I’ve found the last two podcasts very interesting and helpful. Are you planning to produce a list of the claims that identifies which are original claims and provides citations for established claims.
Many Thanks


Luke Muehlhauser February 16, 2011 at 6:29 am


Yeah, probably.


NFQ February 16, 2011 at 11:55 am

Nicely done, guys.

I realize this isn’t about atheism per se, but it reminded me of a point I often try to make, usually to Christians but to theists in general, about “what God wants.” I ask, if God wants everybody to go to heaven, why do you also believe that he made hell? Why did he give us the choice to sin? Why did he make Satan to tempt us? It sure seems like the God they believe in doesn’t actually want everyone to go to heaven, because if he did, he would have done something differently. Frustratingly, they just talk in circles and don’t really “get” my point. But your explanation of what it means to desire something is so clear, I will probably start just referring people here when it gets to that point in the conversation in the future. :)

And Peter, I think this goes to your question a bit, as well. If you want to get an A in calculus but don’t want to study, then “I want to get an A in calculus” might not be the best way to express that more complicated set of desires. Maybe you could say, “I want the best grade possible with the least amount of work,” suggesting that you were doing some kind of moral optimization problem (calc again, haha) to determine your preferred balance of work and good grades. Similarly, it seems like saying “God wants us to go to heaven” is an oversimplification (besides a probable falsehood…) — what they really should say is something more like, “God wants us to go to heaven, but not as much as he wants us to suffer and sin.” … This is the “stronger competing desires” bit that Luke mentioned in his comment right above yours.


Little James February 17, 2011 at 8:14 pm

An agent intends to act so as to fulfill the most and strongest of their desires-as-ends, given their active beliefs.

Is there some definition of “desire strength” that avoids making this claim into a tautology? “An agent acts to fulfill the strongest desires, where desire A is defined to be stronger than desire B if an agent who has desires A & B acts to fulfill A.” If it is tautological as such, it doesn’t really help explain anything about the way incompatible desires are reconciled. Or maybe I’m missing the point of this one (is it meant to be a definition for “desire strength”?)..


JS Allen February 18, 2011 at 2:28 pm

It would be useful for you to define “intentional” and defend the emphasis on “intentional” actions here. Your quotes of the standard view only talk about agents being “disposed to act” and “tends to bring about”. Rather than “tends”, however, you’re talking about what an agent intends. What does that mean, and why is it important that desirism focus only on intentional action?

He then takes one of them to be ‘true desires’ and casts the other out of the realm of ‘desires’. It no longer belongs. After he casts it out, he says that there are things other than “true desires” that are capable of motivating action. However, this is just semantics.

You seem to be making a similar distinction by excluding “reflex actions and involuntary muscle spasms”, and presumably a number of other things that don’t fit your definition of “intentional”.

If it is tautological as such, it doesn’t really help explain anything about the way incompatible desires are reconciled.

Agreed. I’m still not convinced that the whole thing isn’t one big circular definition with no empirical usefulness.


Leave a Comment