Tuesday, August 14, 2007

Worthiness and Second-Order Desires

Richard wishes to argue that my account of morality in terms of desires – which he identifies as “mere drives and cravings” is inadequate. He wishes to reserve a role in ethics for ‘value-desires’ – a type of desire that adds the quality that its object is somehow ‘worthwhile’.

I see you are using 'desire' in a different sense from me. I am talking about value-desires, i.e. those ends we are drawn to by a perception of their worthiness. You are talking about mere drives, or craving-desires.

My response is that these value-desires do not exist. Instead, the phenomena that Richard is talking about and wishes to capture under ‘value-desires’, I capture under the heading of second-order desires or desires that certain desires exist.

These second-order desires can be desires-as-ends that other desires exist, but most often they are desires-as-means that other desires exist. They come from the recognition that some desires tend to fulfill other desires, and some desires do not. The desires that tend to fulfill other desires are desires that people generally have reason to endorse and have the real-world properties that Richard classifies as ‘value-desires’. The desires that do not tend to fulfill other desires, and perhaps even thwart other desires, are Richard’s ‘mere drives or cravings’.

ADHR provides an example of this in saying:

The error, I think, is that proper desires are, contra Hume, amenable to reason. If I realize that I can’t achieve both of two proper desires I have, in my experience, one of those desires has been extinguished or, at least, de-emphasized. Consistency is one feature of rationally; so, in these sorts of cases, inconsistency alters proper desires.

Now, without addressing the question of whether it is ‘contra Hume’ or not, when there are two incompatible desires this tells us something of the value of each desire as a means. Each desire contributes to the thwarting of the other, and thus each desire provides a reason for action to get rid of the other.

Which desire should a person get rid of?

Well, I have a desire to have a healthy body, and I have a desire for chocolate. These desires are in conflict. In deciding between them, I note that the desire for a healthy body will tend to fulfill other desires. However, my desire for chocolate does not. As such, I recognize that the desire for a healthy body has more value as a means than my desire for chocolate. I have more and stronger reasons to give up my desire for chocolate than my desire for a healthy body. In fact, the desire for chocolate is one which I would choose to get rid of.

As for the possibility that a desire will simply disappear when the agent recognizes a conflict, there are far too many examples when this is not the case. Poor eating habits, gambling, drugs, alcohol, tobacco, sexually transmitted disease, all point to cases where a person may wish that desire that conflicts with others would simply disappear, but does not.

Now, ADHR did not say that all desires are like his ‘proper desires’ when they come into conflict. He has not denied the existence of what he has called drive-desires. However, the only testimony he offers is a personal testimony that he has witnessed such a phenomenon.

I would like to politely suggest that one look at this with the same attitude that one would use in looking at the claim, “I know that ghosts exist; I saw one.” I will not deny that ADHR saw something. I would question the accuracy of his interpretation.

We can get something like the suppression of one desire when we realize that it is in conflict with another when, instead of having two desires, we have one desire that can e fulfilled in two different ways. A person can have a desire for sex. He may not care whether it be with Sam or Jesse. Either option fulfills his desire and, thereby, causes the other desire to ‘disappear’. However, this was not a conflict in ends, it was a conflict in means.

The main problem with Richard’s alternative is the impossibility of coming up with any sense of ‘worthiness’ other than ‘being such as to fulfill the desires in question.’ Allegedly, objects of desire can have a property of ‘worthiness’. The simple recognition that something has this property can motivate us to bring it about. Indeed, the recognition of the ‘worthiness’ of an object of evaluation is supposed to be enough to generate a motivation to pursue it – a ‘value-desire’, or a ‘motivational belief’.

No, sorry, I don’t see any reason to go down that road. There is no real-world evidence that compels this type of metaphysics. Second-order desires – the endorsement of desires because they tend to fulfill other desires, either directly or indirectly, has all of the real-world explanatory power we need without postulating any strange entities.

Part of his justification for this belief is that humans are

. Perhaps animals' motivational system is like that, but humans are rather more complicated. We reason about our ends, and not just our means. Some ends are not wholly arbitrary, but rather pursued because we judge them to be worth pursuing.

Yes, we do ‘reason about our ends’. However, there is no such thing as a pure end. Any end we adopt has effects – it is a means to the fulfillment of other ends – or a barrier to their fulfillment. So, we are wise to reason about the value of an end as a means, and to adopt those ends that tend to promote other ends, while inhibiting those ends that tend to thwart other ends.

I will agree that animals do not have the capacity to reason about the value of their ends as means. Yet, animals still (in an unreflective – unphilosophical way) exploit the power of second-order desires. One creature in a colony acts in ways that aggravate others. The others respond with condemnation and punishment. The first animal then learns his place – learns how to behave as a member of the colony.

Because these animals are unreflective about such things, they do not use these tools as efficiently as we are capable of using them. They cannot sit down and debate whether it really is a good idea to be inhibiting this desire or promoting that one. “Maybe it has some long-range or secondary benefits that outweigh its prima-facie badness?” However, the raw materials are there, and they do not require introducing some type of ‘worthiness’ into our world view.

Richard’s objections come with a recommendation

If you like, you could accommodate this within your Humean framework by adding the bridging claim:

(B) Necessarily, a rational agent desires to do what's worth doing.

Then our reasoning is merely changing our beliefs about how to fulfill this antecedent desire, rather than creating any new desire-as-ends.

In order to respond to this objection precisely, I need to know what Richard has in mind for a ‘rational agent’ and for a quality of ‘worth doing’. Yet, I suspect that whatever definitions he offers, I am going to reject this proposition.

Others have reason to make it rational for me to do what is worth doing. Here, by ‘rational’ I mean that an action is such as to fulfill my own desires, while ‘worth doing’ means that those same desires are those that tend to fulfill the desires of others. In fact, I would agree that this is the purpose for moral practices – to promote in people those desires that they have reason to promote, those desires that tend to fulfill other desires. To the degree that they are successful, then what is rational for me will be that which is worth doing.

However, there is no guarantee of success. To the degree that moral institutions fall short, to that degree we have people for whom it is rational to do things that do not fulfill the desires of others; that even thwart the desires of others.

I do not know if Richard would accept this view of rationality. I suspect that he would not accept this account of ‘worth doing’. He has something else in mind. I will answer that this ‘something else’ he has in mind isn’t real. It is as much of an invention as God, and it is just another way to try to give special significance to ends that really obtain their value in the way I described above.

Richard suggests that I assign his quality of worthiness to my own project of making the world a better place than it would have otherwise been.

Much as you will refuse to acknowledge it, I suspect that your own desire to "make the world a better place" is an instance of this. It is a worthy project, much more so than puffing on a cigarette would be.

He knows me well enough to know that I will ‘refuse to acknowledge’ this option. My desire to make the world a better place has no quality of ‘worthiness’ has he would define it. It has the quality of being a desire that people generally have reason to promote (based on their desires). Any attempt to write anything more into that involves stepping into the realm of fiction.

In fact, the differences between our two accounts invite me to ask the following question:

Is there a case in which something fulfills one of our accounts, but not the other? Is there a case in which a desire that tends to fulfill other desires is also a desire for that which is not worthwhile? Or a desire that tends to thwart other desires is worthwhile?

If not – if the two sets are co-existent, I can get rid of Richard’s concept of ‘worthwhileness’ because I simply do not need it. It does no work.

However, if it is true, then we have ‘worthwhileness’ promoting people to do harm by promoting them to pursue that which thwarts other desires, or inhibiting them from doing good by inhibiting them from doing that which fulfills other desires.

This is not a strong argument. It is not an argument at all, since it begs every question that can be begged. However, it does provide an important illustration at what is at stake. Can something have ‘worthwhileness’ that gets in the way of desires that tend to fulfill other desires or promote desires that tend to thwart other desires?

The real argument against this concept of ‘worthwhileness’ is that, if it exists, what is it? What does it take for a proposition identifying something as ‘worthwhile’ to be true or false? How do we recognize it when we see it? How is this alleged capacity to recognize ‘worthwhileness’ made compatible with the fact of human evolution? What is happening when one person identifies a state of affairs as being ‘worthwhile’ while another identifies it as ‘a waste of time’? Who is right? Can they both be right?

These types of questions argue against adding ‘worthwhileness’ to our ontology unless we absolutely have to. I hold that we do not have to. We can do everything we need to do by evaluating the capacity of desires to fulfill or thwart other desires – and we don’t have all of these pesky questions.

Value desires are not a new type of desire susceptible to reason. They are regular desires being evaluated according to their usefulness in fulfilling other desires.


Richard Y Chappell said...

Hi Alonzo, thanks for your thorough response. I'm inclined to agree that fulfilling desires generally is what turns out to be worthwhile. But you seem to want to establish this through mere semantic stipulation; I think it is a substantive fact, and one about which we could (in principle) be mistaken. That is, I use terms like 'ought' and 'value' in a more general sense that does not presuppose any particular moral theory (such as DU). I take 'rationality' to be a similar primitive term, that cannot be reductively defined in terms of anything else. (Though we can at least point to related notions such as 'coherence'.)

Now, I'm not sure that second-order desires suffice to explain the phenomena I've pointed to. I take you to be proposing that what I call a value-desire is really just a craving that you also happen to have a higher-order craving for. That doesn't really seem to do justice to the phenomenology at all. But there is a more fundamental problem: it would seem possible to have the sorts of values I'm talking about without any second-order desires at all (perhaps because one lacks mental concepts altogether, and so is only capable of propositional attitudes towards the external world).

But let me get right on to the most important point: regardless of your meta-ethical views, i.e. even if you are an error theorist about normativity (or what you elsewhere call 'intrinsic value'), there's no denying that we have normative concepts. You simply believe that those concepts don't correspond to anything in reality, i.e. you deny that anything is "really worthwhile" (in my sense). Still, there's no denying that some people disagree with you, and that they at least believe (perhaps falsely) that some things are really worthwhile.

So my point is this: anyone who understands this normative concept of mine necessarily has a desire to match it. It would be incoherent for someone to say, "I think that X is intrinsically valuable but I just don't care." (If they don't care, they must disagree that it inherently warrants their care, i.e. that it has intrinsic value in this sense.) This is so even if I'm mistaken and in actual fact there is no intrinsic value in the world. Still, you can motivate me to do something simply by getting me to BELIEVE that something is intrinsically valuable.

Alonzo Fyfe said...


One major change in the way I describe desire utilitarianism since the IIDB days is this:

I used to argue that values are relationships between states of affairs and desires as a matter of semantic stipulation, as you say here. However, that simply does not work. People at least have concepts of other types of value and use them in their arguments. Those statements would not make sense under this initial assumption.

So, now, I argue that 'should' and 'ought' make reference to reasons for action. It turns out that desires are the only reasons for action that exist. However, if there were any other reasons for action (intrinsic value, divine properties) they would be legitimate premises in an argument about what we should or should not do.

That desires are the only reasons for action that exist is an empirical claim - again, as you propose. However, I have found no reason to add to it.

Some of the phenomenology you pointed to can be explained by the fact that people have beliefs that things have inherent worth, just as they have beliefs that events in the world are influenced by God. So, they speak as if these entities are real. Yet, this does not make these entities real. Instead, these ways of speaking are mistaken. I hold to an 'error theory' about these statements, as you point out.

It is a mistake, in other words, to take a self-proclaimed 'appreciation of the inherent worth' of something at face value, just as it is a mistake to think that a person who claims to have seen a ghost has actually seen a ghost. Ghosts are such unlikely entities that it is far more reasonable to believe that what the person thought was a ghost was, in fact, something else. All of his claims about the ghost appearing to speak to him or motioning him to follow must be understood in this light as well.

I agree that people can have a desire to do that which has intrinsic merit. And they can believe that a state of affairs has intrinsic merit. If we convince such a person to change his beliefs about the merit of a particular state, then we will change his motivation to bring about that particular state.

I also agree that people can have a desire to please God. And they can believe that a state of affairs is pleasing to God. If we convince such a person to change his beliefs about the merit of a particular state, then we will change his motivation to bring about that particular state.

Yet, in neither case does this argue that intrinsic values are real.

Nor does it give us any reason to postulate a 'different type of desire'. The desire to do that which has intrinsic merit, or to do that which pleases God, is no different than the desire to make one's child happy, or the desire to have steak for supper. They just happen to take as their object a type of value that does not exist.

ADHR said...


The fact that there are cases where neither of a pair of conflicting desires is extinguished does nothing to undercut the claim that there are cases where one (or both) is extinguished. It's dogmatic to deny personal testimony on the basis of theory unless the theory is clearly stronger than the testimony. But whether the theory is stronger is exactly what's at issue. So, looking at the theory....

Your interpretation of desire-conflict employs second-order desires. Since you're not altering the nature of a desire, it follows that these are just plain ol' proper desires which take other proper desires as their objects. This gets you into trouble because the order problem just repeats at the higher order. (If the second-order desires are different in kind, you may have a way to block this move.) If you invoke second-order desires to account for the extinguishing of a first-order desire by another first-order desire, then you need third-order desires to account for cases (which are, at least, logically possible) where second-order desires are extinguished by competing second-order desires. And then, since third-order desires can come into conflict just as easily as any other kind of desire -- they are, after all, still just proper desires with a certain kind of object -- you need fourth-order desires to account for the resolution of conflicts between them. And then fifth-order, and sixth, and seventh, and so on. Thus, second-order desires ultimately fail to explain anything -- because they require third-, fourth-, fifth-, etc., order desires, without end. Our search for explanation never stops, and no explanation is ever even close to complete.

Furthermore, this picture, which which follows logically, is psychologically bizarre. It's questionable whether human psychology can accommodate the necessary fifth-, sixth-, seventh-, etc. order desires. From this, it must follow either that human fourth-order desires never conflict (felicitous, but unlikely) or humans can't resolve these conflicts at all (which flies in the face of experience; we don't see inanimate people on a regular basis).

Richard doesn't need to appeal to an inherently motivating property of worthiness or value more generally in order to come up with a better account. (Although it seems that he wants to. Bloody internalists!) All he needs to point to is the existence of pro-attitudes (value-desires) which recognize the property of worthiness. Worthiness doesn't have to produce these pro-attitudes inerrantly; there may be external factors which are also required before pro-attitudes exist. My position is actually even stronger than this, because I've claimed that intentional action is the justification for attributing a pro-attitude. That is, I've inverted the story: we infer something about the psychology of an agent based on what he has already done. Value need not be inherently motivating for this story to work because the motivational impact of value is a fait accompli once someone has acted.

Furthermore, one need not be committed to a "strange entity" view of worthiness. (To be pedantic, worthiness is a property, not an entity.) More substantively, it's still possible that worthiness can be analyzed in a manner consistent with reductive (or non-reductive) materialism. All that's ruled out, really, is elimination. Your ontological arguments don't touch views that provide materialistic analyses of value.

Anonymous said...

“These second-order desires can be desires-as-ends that other desires exist…” –Could you give an example of this, since it reads like a self-contradiction?

Uber Miguel said...

Fyfe explained what he refers to as second-order desires:

"second-order desires or desires that certain desires exist."

He also said that these second-order desires are commonly desires-as-means:

"most often [second-order desires] are desires-as-means that other desires exist."

But, they can also be desires-as-ends whenever the second-order desire is the result desired irregardless of whether or not it indeed also actually directly causes the fulfillment of the first-order desire. Remember, a good desire is one that tends to fulfill other desires.. not just the ones that, indeed, directly fulfills other desires.

Desires-as-ends can probably be thought of as when we focus our desires on a method instead of the intended product. Exercising for the sake of it (second-order) irregardless of why you started exercising in the first place (first order) such as the prospect of living a longer life.. that sort of thing.

Alonzo Fyfe said...


The problem you raised here is similar to a problem once raised in epistemology - the study of beliefs.

The idea is that we justify a proposition by connecting it to other propositions - justifying a conclusion by listing its premises.

Each premise is itself a conclusion drawn from prior premises.

The argument went that there must either be a peculiar kind of premise that needs no other premises to defend it, or we suffer an infinite regress.

Epistemologists came up with a third option.

Coherence theory.

There are no foundational proppositions. Instead, we have a network or a web of propositions, each connected others. Those propositions with the most and the strongest connections are the best justified. Those with few, fragile connections are not justified at all.

I hold that we can avoid the problem with an infinite regress of desires in the same way. Each desire is to be evaluated by its tendency to fulfill or thwart other desires.

I do not call this coherence because it has an important difference from beliefs. With beliefs, if two people believe different things, there is incoherence. In a universe with perfect coherence of beliefs no beliefs would contradict each other.

On the other hand, desires can sometimes best reach harmony by diversity. When it comes to eating chicken, I like dark meat. My wife likes white meat. Our desires are in harmony precisely because we have likes and dislikes that avoid conflict, at least when it comes to eating chicken.

So it is that if we all desired the same thing, this would lead to competition and conflict. However, in some cases, if we desire different things, our desires can be in harmony.

As for your positive view, I'm afraid that I am not understanding it well enough to comment on it. That is, other than the phrase, 'Bloody internalists!'. I understand that well enough and I agree with that part. But I am having trouble with the rest of it.

Alonzo Fyfe said...

Atheist Observer

A desire is a propositional attitude - a mental state that takes a proposition as an object. So, "I desire that I am eating chocolate" expresses a mental attitude that motivates me to make the proposition "I am eating chocolate" true.

A second-order desire is a desire that takes another desire as an object. That is to say, a desire is mentioned in the proposition that is the object of a second order desire.

So, "I desire that I do not desire that I am eating chocolate" represents a second-order desire. The desire takes another desire (in this case, the desire for chocolate) as its object.

This second-order desire can be a desire-as-end (I simply do not like being somebody who desires chocolate), or it could be a desire-as-means (getting rid of my desire for chocoalate would be useful, in that it will tend to fulfill other desires).

Examples of a second-order desire-as-end would be somebody who has an aversion to homosexuality who happens to be homosexual. He may simply dislike the fact that he is gay. Or he may simply dislike the fact that he likes alcohol. Or he may wish that he had a stronger affection for his wife. He need not desire these things because of some further ends. He may desire them for their own sake just as he may desire a chocolate cake for its own sake.

Anonymous said...


Please look at your examples again. Without other desires from which these were derived, none of them make sense. Why would anyone have an aversion to homosexuality if they have no desires associated with it? Why would someone dislike the fact that they like alcohol if they have no desires associated with it? Why would you want to have a stronger love for your wife if you have no desires associated with loving her? When you look closely, none of these make sense as ends, they only make sense as means to some other ends.
While desire utilitarianism may address a number of problems with other ethical philosophies, until it encompasses an accurate view of the creation, development, and relationships among desires, its practical uses will be limited.

ADHR said...
This comment has been removed by the author.
ADHR said...

[Sorry, the formatting went sideways when I posted this originally. This is the same comment as the above one, just formatted legibly!]


I think the idea is supposed to be that someone can "just desire" to have or not have particular desires, just as one can "just desire" to have particular objects. So, you can desire not to sexually desire members of the same sex because this fulfills other desires, or you can desire not to sexually desire members of the same sex period. I'm sympathetic to the claim that Alonzo needs a fuller account of the psychology underlying his ethical views.


I don't think I'd characterize epistemology as the study of beliefs. It's the study of knowledge and, insofar as knowledge requires beliefs, then the study of beliefs. Philosophical psychology has things to say about beliefs that epistemologists don't tend to talk about. (Of course, this may just be a wholly artificial division between subdisciplines, such as commonly emerges in the academy.)

The problem I'm pressing is distinct from the infinite justificatory regress of beliefs. In that case, we start with a belief and look for the reason(s) for it. We then look for the reason(s) for the reason(s), and so on. It's quite legitimate to argue this, at some remove, goes round in a circle and our original belief, perhaps in concert with some others, ultimately justifies itself.

It doesn't work for the higher-order desires case, though. Second-order desires aren't desires that provide justification (or some analogue) for first-order desires, they're desires that take as their object first-order desires, just as first-order desires can take as their object states of affairs, physical objects, properties, and so on and so forth. In the epistemological case, we're not dealing with beliefs about beliefs; we're dealing with beliefs that entail or support beliefs.

Now, there's clearly s nothing logically wrong with a third-order desire, i.e., a desire that has as its object a second-order desire. And there's a clear analogy between the use of a second-order desire to solve conflicts between first-order desires and the use of a third-order desire (and so on) to solve conflicts between second-order desires (and so on). That is, the reasoning is the same: in order to resolve the conflict, we go to a higher-order desire. Once you've invoked the move to a higher-order desire as a conflict-arbitrator, I'm not sure how you can ever justify stopping it at any particular level. Why not at fourth-order instead of second? Why not eighth?

Indeed, why not first? If coherence (or a desire analogue of epistemological coherence) is going to allow you to avoid invoking third-order (and so on) desires to solve ever more complex conflicts, why can't it stop the hierarchy completely? Why can't this coherence-esque feature do all the work on its own? (FWIW, Michael Smith ends up with a solution rather like this. I'd really recommend his book as a good, thorough treatment of the possibility of a pseudo-Humean account of moral reasons.)

The positive view I'm sketching really involves just three claims. First, that belief-desire explanations of intentional action are backwards. We should start with action, not with psychology. Second, that non-reductive materialist accounts of value are the most plausible accounts available. Third, reasons internalism (of just about any stripe) is false, and externalism (of at least one kind) is true.

The first I raised to highlight the assumption you (and many others) make, namely that we have to start in the mind and work our way outward. It almost looks (to me, anyway) like a last lingering vestige of Cartesianism. My view is that, really, the only evidence we ever have to go on is action. We see what people do (including ourselves, such as the operations of our minds) and we then attribute psychological states to them in order to make sense of what they have done. The reason we don't need to consider this any more than a mere attribution is that taking it more seriously requires adding in another, fairly substantive, claim: namely that psychological states, in addition to being reasons for action, are also causes of action. (And, indeed, that causation always goes forward in time, but that's a complex problem in its own right.) Once you do away with the idea that reasons cause action, it's entirely possible that reasons are psychological states that we attribute in order to account for what people do. Dennett has a view like this; he holds that by attributing psychological states to people, we come to see their actions as part of some larger, and epistemically useful, patterns.

The second I was raising as a possible fall-back for Richard. You were arguing that Richard doesn't have anywhere to go in his account of "worthiness" except to some crazy dualism or sui generis moral property, such as the kind of thing GE Moore believed in. He certainly could go there, but he doesn't have to. Most people who believe values are real (and many of those who believe they are relative) believe in some sort of non-reductive materialism. That is, they hold that certain underlying physical conditions constitute moral properties. It's a flat rejection of Moore's open-question argument: while you certainly could still ask, intelligibly, "but is moral property x really just physical properties a, b and c?", doing so reveals that you don't really understand what makes moral property x come into existence. Peter Railton has a view like this on the realist side; Gilbert Harman on the relativist.

The third I take it you have some sympathy with. There's something psychologically and metaphysically mysterious about properties that just produce action, or even that provide some motivation for action (unless we want to analyze motivation completely causally, I suppose). This is just JL Mackie's so-called "argument from queerness". I say so-called because it's not much of an argument; it really just draws attention to the problem of accounting for this mysterious power of action-production. What I'd like to say is that, in addition to the relevant properties, you also have to have an agent who has the appropriate capacities to respond to the properties in the appropriate way. That is, motivation doesn't come with reasons; motivation comes with reasons held by a certain kind of reasoner. An analogy might help here. (It'd certainly help me!) Radio waves, on their own, don't constitute music. Radio waves received by a particular kind of device, listened to by a particular kind of creature, do constitute music. Analogously, reasons, on their own, don't motivate action. Reasons perceived by a particular kind of creature, capable of particular kinds of responses, do motivate action.

Richard Y Chappell said...

Alonzo - put in terms of 'reasons for action', then, my claim is that beliefs about these reasons are intrinsically motivating. (Equivalently: anyone with the concept of normative reasons necessarily has a desire to act as they have most reason to.) You compare this to "a desire to please God", but that desire is utterly contingent. It would be perfectly coherent for someone to believe in God, and yet still lack the desire to please him. So beliefs about God are not necessarily motivating, the way that beliefs about reasons are.

As I tried to emphasize in my first comment, I am not here arguing that intrinsic value is real. I'm merely arguing that the - undenyably real - beliefs about intrinsic value are necessarily motivating. It follows that motivation can be shaped by reason, by changing one's beliefs about what they have reason to do, regardless of their contingent desires.

Alonzo Fyfe said...


There is one sense in which I can make sense of your claims above.

"Intrinsic value" typically means "A property intrinsic to an object or state of affairs whereby the mere awareness of it would motivate an agent to pursue it."

Because of this, it would certainly be odd for a person to say, "X has intrinsic value, and I am aware of it, but I am not motivated to pursue it."

However, this still does not justify a third type of mental state - a motivational belief. It simply means that real-world use of the 'intrinsic value' concept will only make sense when applied to things that the agent is desire-motivated to pursue anyway.

In fact, 'intrinsic value' in practice is a rhetorical trick whereby an agent identifies an object that he likes and wants to see promoted and seeks to imply that those who do not desire what he desires is somehow defective. So, homosexual acts are identified as 'intrinsically bad' is used as a tool for calling homosexuals defective.

This account does not require that there actually be any self-motivating beliefs. Rather, self-motivating beliefs are simply another part of the myth of intrinsic values. Instead, there is a concept of 'intrinsic value' that people only apply to things that they are already desire-motivated to pursue anyway.

Anonymous said...


You say “I think the idea is supposed to be that someone can "just desire" to have or not have particular desires, just as one can "just desire" to have particular objects. So, you can desire not to sexually desire members of the same sex because this fulfills other desires, or you can desire not to sexually desire members of the same sex period.”

My point is while that is philosophically or logically possible, in the real world it doesn’t happen. It doesn’t happen because desires don’t just pop into existence from nothing. We’ve evolved internal punishment and reward systems that keep us alive and procreating, but once you get beyond those, our desires are derived.

To say, “you can desire not to sexually desire members of the same sex period” is to say you have no reason to have this desire. If you have no reason to have that desire, you will not have it. We don’t go around randomly creating desires for no reason.

ADHR said...


I'm not sure why you think we can't have desires for no reason at all. (Indeed, on Alonzo's picture, we have to have desires for no reason at all.) If desires are psychological states then, plausibly, they're connected somehow to states of the brain. States of the brain can be caused by all sorts of things, not all of which will count as reasons. (Injury and illness, for example, can cause states of the brain.) So, it's not just logically but empirically possible that a desire can just appear, without any reason at all.

Furthermore, it's always possible that, even if we derive a desire for what seems to be a reason, we could be wrong. That is, we may take something as a reason to have some desire, and yet be mistaken: it was no reason for that desire after all.

Anonymous said...


I don’t know why you say Alonzo thinks we have to have desires for no reason at all, but then I haven’t read all his posts.
Alonzo does describe desires as propositions rather than brain states, but I tend to agree more with you that brain states are what we are talking about.
Perhaps our disagreement is semantics. You say brain states may have causes but not reasons. I consider them the same thing. I don’t insist that our desires are all logically reasoned out, but that they all have causes. As you point out, what we believe to be the reasons may well be wrong, but that is not proof that no reasons or causes exist.
When things directly trigger positive or negative brain states then we can say we have found the ends that require no further causes or reasons. It is my view that in normal life we don’t encounter desires about desires that directly trigger brain states without some mediating mechanisms that are what actually brings about these brain states. So if I have no other desires or feelings about homosexuality, I am not, without reason or cause (conscious or unconscious, mental or physical), going to suddenly have a desire not to have desires relating to it.

ADHR said...


Alonzo thinks that desires are arational, which would be nonsense if all desires are had for reasons. If it is the case that all desires are had for reasons, then all desires are at least derivatively rational (or irrational, if the reasons are bad ones).

I was taking Alonzo to mean that desires were propositional attitudes which, unless we want to be Descartes or Berkeley or someone, have to be brain states (possibly at some remove). If he thinks they're propositions, then the ontology would have to be really quite strange. There'd be desires, which are propositions, and then attitudes towards those desires, which would be... what, exactly? It's hard to say.

If you equate reason with cause, then I'd agree that desires can't be had without reasons. (Well, to the extent that I'd allow desires exist. I don't actually think they do, but that's a whole other issue.) I'm convinced, though, that reasons and causes are distinct. That is, something can be caused and yet have no reasons; and, similarly, something can have reasons and yet be uncaused. (Although, in our world, the latter is probably just a logical possibility, as we have yet to discover anything that is uncaused.)

Alonzo Fyfe said...

I suppose I should say something here.

Desires are propositional attitudes, which are brain states. Specifically, a desire that P is an attitude that the proposition P is to be made or kept true.

Desires certainly have cause-reasons, but they do not have end-reasons. Our aversion to pain has a cause-reason that includes a few million years of evolution creating a system where we respond negatively to certain transmissions to the brain and are motivated to change our behavior to keep them from happening again.

However, desires do not have end-reasons. Desires themselves are the only end-reasons that exist. (The proposition P that is the object of a desire that P being an 'end' of human action.)

There is no end-reason to the aversion to pain or the desire for sex or the fondness for high calorie food. There are only cause-reasons.

And, I am sympathetic to the idea that desires do not exist, and that we need a better theory of human action. However, until a better theory comes along, we are stuck with the theories we have.

Newton's theories turned out to be wrong. However, until Einstein's theories came out, Newton's theories were good enough and better than nothing.

ADHR said...


I'm not sure what you mean when you say desires "have" cause-reasons, but not end-reasons. Since, as you've said already, I can desire that I desire that p, it seems that I can have an end-reason for a desire -- namely, that second-order desire. Did you mean to say that desires are cause-reasons? That I can make sense of, although I think you're on dangerous ground if you collapse (as this would) causation into teleology.

FWIW, as far I understand these things, Newton's theories aren't wrong, just incomplete. That is, they describe a set of situations that are only a subset of the situations that Einstein's theories describe. That's one way a theory can be supplanted; another way is the move from Ptolemaic to Copernican astronomy. The crystal-spheres model still works, if you're willing to allow for some strange physical behaviour and difficult mathematics, but the current astronomical model is preferable because (amongst other things) it's a lot simpler.

My point is that shifts between theoretical constructs are complex, and it's rare to find out a theory was never any good. It's more common to find that the theory was incomplete (Newton) or was in some important way inferior to a competitor (Ptolemy).