Friday, January 20, 2006

Morality and Reasons for Action

In an email, Richard Chappell, the owner and operator of the philosophy blog Philosophy et cetera (which I highly recommend), asked that I clarify my position on the relationship between morality and reasons for action.

I have broken the relevant parts of his email out into what I think are the five central claims.

C1: I am guessing that you accept what we might call "Rational Egoism", or the view that one only has reason to fulfill their own desires.

C2: But then you must agree that the amoralist (who has no desire to be moral) has no reason to be moral if he can get away with being wicked.

Then he asks,

C3: Why not say that other people's interests (which you would explicate in terms of their desires) can provide us with (normative) reasons for action, just as our own can, whether we like it or not?

Mr. Chappell also adds the following:

C4: Of course, only our own desires can be (explanatory/motivating) reasons causing our action. . .

C5: . . .but that doesn't entail anything about what normative reasons for action there might be, at least not without some extra assumptions.

Answering this question will require two fine-grained distinctions.

Having a Reason vs. There Exists a Reason

First, I want to distinguish between “having a reason” and “there exists a reason.”

There is To answer this question, I would like to propose a fine distinction that Mr. Chappell blurs in his question. It is a distinction between “having a reason” and “there exists a reason”.

The only reasons for action that are real – that exist – are desires.

However, there is a distinction between the desires that a person has, and the desires that exist. I have all sorts of desires – thus, I have all sorts of reasons for action. However, it is clearly not the case that my reasons for action – my desires – are the only desires that exist. The planet is filled with all sorts of beings, each with their own desires (reasons for action). Those desires also exist.

Therefore, the set “has a reason” for any particular agent is an extremely small subset of the “reasons that exist”.

Desires and Motivating Reason – Rephrasing C2

Second, I want to distinguish between having a desire to be moral, and having a motivating reason to do the right thing.

A desire to be moral is one specific desire out of countless possible desires. I will interpret this as a “desire that I do the right act”. A person with a desire to be moral only needs to know that a particular act is a right act, and he will have a motivating reason to do that.

Yet, an individual can lack a desire to be moral, and still have a motivating reason to do the right act. A parent’s desire “that my child is healthy” is not identical to a desire “that I do the right act.” A parent who has the first desire but not the second will be motivated to care for his child’s health, without having a desire to be moral.

From this, I infer that C2, as written is false.

C2(a): The amoralist (who has no desire to be moral) can still have a reason to be moral, even if he can get away with being wicked, but only if being moral will fulfill some other desire that the amoralist has.

“Rational Egoism” – C1 and C4

In C1, Mr. Chappell used a definition of “Rational Egoism” that is ambiguous.

One possible definition – the definition under which I would say that C1 is true, is: R1: The reasons for action that a person has are his or her own desires.

This, however, would make C1 (accepting rational egoism) and C4 (accepting that a person’s motivating reasons are his or her own desires) identical.

Mr. Chappell cannot consistently assert:

  1. C4 is “of course” true.
  2. C1 is logically equivalent to C4
  3. C1 implies C2
  4. C2 is false

If we have to break this contradiction, where do we break it at?

I would suggest that we break it at (2).

“Rational egoism” contains two parts; only one of which is contained in C4. The other is:

R2: The only desires that a (rational?) being has are those that concern his or her own welfare.

When we add R2, we can see how C1 implies C2.

The “rational egoist” has no motivating reason not to be wicked when he can get away with it – because the rational egoist has no other-regarding desires.

However, I can reject C2 because I can (and do) reject the second component of “rational egoism” – the absence of other-regarding desires.

Back to Reasons: C1 vs. C5

Now, please note the subtle shift in Mr. Chappell’s language between C1 and C5.

In C1, Mr. Chappell uses the phrase “has [a] reason”. In C5, he uses the phrase “what normative reasons for action there might be.”

This is the distinction between the reasons that a person has, and the reasons that exist.

I have already written that the reasons that a person has is a small subset of the reasons that exist. From this, I agree that the fact that only the reasons that a person has can be the immediate cause/explanation of his actions does not imply that these are the only reasons that exist. Other reasons can (and do) exist. They simply cannot be the immediate cause/explanation of any agent’s actions.

However, these other reasons that exist can be a distant cause of a person’s actions in two ways; one blunt, and one subtle.

(1) The blunt way in which other reasons can affect an action is through the fact that “being a danger to fulfilling the desires of others makes me a threat, and gives them reason to do me harm.”

That is to say, if I thwart their desires, I am at risk of suffering all sorts of harm that they may visit upon me.

(2) The subtle way in with other reasons can affect an action is that those other reasons give other people an incentive to mold my desires.

These are motivating reasons for them to cause me to have desires that will tend to fulfill their desires, and to inhibit in me desires that will thwart their desires. Thus, even though my own desires are the immediate cause of my action, what those desires are molded, in part, by what others have reason to cause my desires to be.

On these grounds, I note that when a parent scolds a child, the parent uses the statement, “You should be ashamed in yourself. The parent is talking about the desires and aversions that society has reason to cause each individual to have, not the reasons that the person being scolded actually does have. From the point of view of the person being scolded, moral claims concern the other reasons that exist, not the reasons that the agent has.

Furthermore, I argue that this roughly – very roughly – defines the difference between law and morality. There is a large degree of overlap but, roughly, law defines the threats we make against those who do wrongful actions, whereas morality concerns the desires and aversions we seek to cultivate in people that will prevent them from being a threat to others.

One significant area of overlap relates to the concept of just and unjust laws. The difference here rests on whether the laws established in (1) are those laws that a “good person” – as defined in (2) – could support. Laws can, in fact, be just or unjust.

Summary

So, here is my view.

A0: The only reasons for action that exist are embedded in desires.

A1(a): The only reasons that a person has for any action are those embedded in his or her own desires.

A1(b): The desires (or reasons) that a person has may be other-regarding; an agent can have a desire that another person be well off.

A1(c): If “rational egoism” means “A1(a) is true”, I accept “rational egoism”. If “rational egoism” means “A1(a) is true and A1(b) is false”, I reject “rational egoism.”

A2(a): The amoralist (who has no desire to be moral) can still have a reason to be moral, if the other reasons he does have are other regarding or reasons that tend to fulfill the desires of others.

A2(b): One has to assert that A1(b) is false to get to the conclusion that a person with no desire to be moral has no reason to be moral.

A3(a): The set of reasons that exist is much larger than the set of reasons that an agent has.

A3(b): Other people’s interests (desires) are (normative) reasons for action, just as our own can, but they cannot be the immediate cause of that action.

A4: Of course, only our own desires can be (explanatory/motivating) reasons causing our action. This actually simply restates A1(a).

A5: However, this does not entail that no other normative reasons for action exist, at least not without some extra assumptions. Indeed, other reasons (desires), in addition to those that are the immediate cause of an agent’s actions, certainly do exist.

A6: Moral statements are not statements about the reasons for action the agent or the person being spoken of have. They concern the reasons for action (desires) the subject should have given the reasons that exist; what others have a reason to cause the person being subject to moral praise or condemnation to have.

8 comments:

Richard Y Chappell said...

Hi Alonzo, thanks for your thorough explanation. I should clarify that by "rational egoism" I merely meant the claim in A1(a), allowing for the possibility of other-regarding desires that would lead one to act morally. This still seems problematic because it allows for the possibility of someone having no reason to be moral (if they didn't want to, and if it wouldn't advance their other ends). Moreover, on the standard view of normative reasons, one ought to do whatever they have most reason to do. It would then follow that there might be someone who ought not to be moral. (In the "all things considered" normative, rather than "adverbial" moral, sense of 'ought'. I explain the distinction here.)

Now, I'm not sure that you've actually discussed normative reasons in this post. I think you've instead been talking about motivating reasons throughout. This would explain why you thought A1(a) and A4 equivalent. I meant for C1 to be talking about normative reasons, whereas C4 is about motivating reasons.

On the standard view, a normative reason is a fact that "counts in favour" of something. (Then, what we "ought" to do is that action which is best supported by reasons.) For example, suppose I could benefit many people by clicking my fingers. This fact would clearly "count in favour" of my clicking my fingers; thus I have (normative) reason to so act. This is so even if I have no desire to help others, and so lack any motivating reason to click my fingers. My actions are caused by my own desires, and so only they can be motivating reasons for me. But other people's interests can nonetheless be normative reasons that count in favour of my acting one way rather than another.

(I think you might have been assuming that we are necessarily motivated by the reasons for action that we have. But I reject that view. We might have reasons for action that we're unaware of -- as when you didn't know the stove was hot -- or that we simply don't care about, say if we're selfish or callous. Those reasons contribute to what we should do, even if they fail to motivate or contribute to what we in fact do.)

(I also suspect that you might deny that any "normative reasons" exist, in my sense of the term. After all, they're much on a par with intrinsic values. You might instead try to redefine the term, as you have with "ought", to mean something purely descriptive. But you would then be talking about something difference from what other philosophers mean by the terms, at least according to what I've called the "standard view".)

Finally, I'm not sure your distinction between “having a reason” and “there exists a reason” makes sense, once you consider how they relate to some particular agent's action. Suppose I am wondering whether to X. What difference is there between my "having" a reason to X, and "there existing" a reason for me to X? These are surely saying the same thing.

Perhaps, if X would fulfill your desires, you have a reason to X (or hope that someone else does). But that does not mean that there exists any reason for me to X. Rather the reason that exists is a reason for you (to X, or hope that I X). Any "existing" reasons for me to X are (by definition) reasons I "have" to X. They mean the exact same thing; there is no distinction here. Any other reasons in existence are reasons for some other action (say, that distinct act wherein you X, or hope that I X); they can't be reasons for this (my) action.

Put another way, I'd ask you to clarify your "A3(b): Other people’s interests (desires) are (normative) reasons for action, just as our own can, but they cannot be the immediate cause of that action."

Which "action" do you mean here? Can other people's interests be normative reasons for my actions? If so, this contradicts A1(a). Alternatively, if you are merely pointing out that other people's interests can be normative reasons for other actions (say, their actions), then this rather trivial suggestion doesn't say much, and indeed makes A5 quite misleading. (When we ask whether there are other normative reasons which count in favour of this action, it's simply changing the subject to point out that there are other normative reasons for other actions!)

But never mind that, the main point I wanted to make is that A4 is not just a restatement of A1(a), because normative and motivating reasons are not the same thing.

Alonzo Fyfe said...

I should clarify that by "rational egoism" I merely meant the claim in A1(a). This still seems problematic because it allows for the possibility of someone having no reason to be moral.

I am assuming that you mean logical possibility. All we have to do is imagine a creature with only one desire -- a desire to do harm to others. This is a creature that has no reason to be moral. This does not seem the least bit problematic.

If you are saying that some sort of reason must exist that every creature must have that gives it a reason to be moral, this is an ontological claim, and cannot be proved or disproved by looking at logical possibilities. We cannot simply define things into existence.

(I think you might have been assuming that we are necessarily motivated by the reasons for action that we have. But I reject that view. We might have reasons for action that we're unaware of -- as when you didn't know the stove was hot -- or that we simply don't care about, say if we're selfish or callous.

My view is that we act so as to fulfill our desires given our beliefs; however we aim to fulfill our desires. Thus, false beliefs can prevent us from actually fulfilling our desires.

If the agent has an aversion to the state that would result from putting his hand on the hot stove, then he has a reason not to do so. If he does not know that the stove is hot, he still has a reason not to put his hand on the hot stove. That is to say, the result of putting his hand on the hot stove will create a state to which he is adverse (even if he is not aware of the fact that it would create such a state).

It would appear that you would classify the aversion to the state that would result as a 'normative reason' rather than a 'motivating reason'.

If this is the case, then I am talking entirely about normative reasons, and am not talking about motivating reasons at all. "Good" relates states of affairs to desires that exist, whether one is aware of that relationship or not.


On the standard view, a normative reason is a fact that "counts in favour" of something. (Then, what we "ought" to do is that action which is best supported by reasons.)

I hold that "ought" is an ambiguous term. There are several different meanings of the word "ought" -- each relating to a different subset of the set of all 'normative reasons' (desires). Such as the distinction between practical 'ought' (considering only the 'normative reasons/desires that I have), and moral 'ought' (considering all of the normative reasons/desires that exist).

For example, suppose I could benefit many people by clicking my fingers. This fact would clearly "count in favour" of my clicking my fingers; thus I have (normative) reason to so act.

In the case of moral 'ought' -- yes, it would. In the case of practical 'ought' -- no.

However, it does give others a practical-ought reason to make it the case that you do have a motivating reason to snap your fingers, to the degree that they have the power to alter and shape your motivating reasons, which they do.

The mere fact that other people exist with motivating reasons to cause you to snap your fingers, means that (in a sense), you almost certainly do have a motivating reason to snap your fingers -- as a part of a bargain in which they do something for you, or to avoid the harm that they would visit upon you if you do not. However, these are all contingent, which is why I say almost certainly.


I also suspect that you might deny that any "normative reasons"…

It turns out that I do not, though we may disagree with the other properties that a normative reason has.

The question of whether I have redefined the term remains open to question.

I'm not sure your distinction between “having a reason” and “there exists a reason” makes sense….

I see no basis for this claim. "Having a reason" refers to the desires that one actually has. "There exists a reason" refers to the desires that exist. Clearly, there is a distinction between the desires that I have, and the desires that exist.

When actually making a decision about what to do, a person will seek to do that action that best fulfills his desires, and will in fact do that action that best fulfills his desires, given his beliefs.

The case of the hot stove illustrates this distinction, where the person who believes that the stove is not hot, and who believes that putting his hand on the stove will fulfill some other desire, will put his hand on the stove, even though he has a reason not to.

A person with complete and true beliefs will act so as to fulfill his own desires -- the desires that he has -- the reasons for action that he has. To the degree that he cannot fulfill all of them, he will fulfill the more and the stronger of those that he has.

However, a person with complete and true beliefs will not necessarily act so as to fulfill all of the reasons that exist. Indeed, if he has a desire to harm others, he may act so as to thwart the desires of others, to fulfill his own desire to do harm.

The way to get a person to act in ways that fulfill the desires of others is to make sure that he has desires that tend to fulfill the desires of others. That way, in his quest to fulfill his own desires, he benefits others as well. Those 'others' clearly have reasons to see to it that each agent has such desires.


Any "existing" reasons for me to X are (by definition) reasons I "have" to X. They mean the exact same thing; there is no distinction here.

I think that the issue here is an ambiguity in the phrase "reasons for me". I think that this phrase is ambiguous, and its ambiguity lies along the same lines as the "reasons that exist" and "reasons that I have".

There is a "reasons-for-me" sense of this phrase that is, in fact, the same as "reasons of mine" or "reasons that I have" (or "desires that I have").

There is another sense of this phrase that means "reasons that count in favor of my" which is not the same as "reasons of mine". This sense relates to the broader sense of "reasons that exist" (or "desires that exist").

In this case, the "'existing' reasons-for-me to X" is, in fact, the reasons that I have for doing X. Yet, it does not deny that there are also "'existing' reasons that count in favor my doing X'" that are not reasons that I have.

The fact is, there is a difference between the desires that I have, and the desires that exist. If our vocabulary blurs this distinction, or refuses to let us recognize the fact of its existence, the problem rests in our vocabulary. We cannot use conceptual analysis as a "proof" that such a distinction does not exist -- that the set of reasons that I have is identical to the set of reasons that exists.

Can other people's interests be normative reasons for my actions? If so, this contradicts A1(a).

If you are talking about normative reasons that you have -- of course not. If you are talking about normative reasons that exist, then yes.

But never mind that, the main point I wanted to make is that A4 is not just a restatement of A1(a), because normative and motivating reasons are not the same thing.

Under this clarification, nothing that I wrote concerns "motivating reasons" as you have defined the term here.

In fact, I would argue that "motivating reasons" as you define it, are not that important. Where I have talked about "motivating reason" I was talking about a reasons that would have an effect on a person's actions if his beliefs were true and complete. You seem to be defining the term as those reasons that actually affect person's actions. If a false belief causes a person to put his hand on a hot stove, you would not count his aversion to pain as a "motivating reason"; whereas, I would.

I think that this is a semantic dispute that does not affect the argument. I can switch to your terms and still make the same claims -- though with different words.

Richard Y Chappell said...

I do think we're talking about thoroughly different things. I'm happy to grant your broader use of 'motivating reason', it's still only talking about a descriptive matter. Normative reasons go beyond this. They are conceptually distinct from desires (though there may be a metaphysical connection, such as a supervenience relation, there).

For example, you write: "he still has a reason not to put his hand on the hot stove. That is to say, the result of putting his hand on the hot stove will create a state to which he is adverse"

I am concerned about your locution "that is to say". You first and second sentences are not synonymous. Whether one entails the other will depend on our theory of reasons. But I worry that you are merely defining them to be equivalent, and thus aren't really dealing with our independent conception of reasons at all!

This also relates to your use of the word 'ought'. I agree that it has many different uses. As explained in my linked post (see prev. comment), most of these are 'adverbial' uses, e.g. "morally ought", "prudentially ought", etc. None of those reflect the fundamental normative concept. Rather, to say that you "R-ally ought to X" is merely to make the descriptive claim that X is a requirement of framework R (e.g. morality, prudence, whatever). It leaves open the question of whether one really ought to follow that framework of requirements. (For example, sometimes we ought not to do what we "legally ought", as MLK recognized.)

The fundamental normative concept concerns what we really ought, all things considered, to do. Normative reasons are those facts which help determine one of these 'ought' facts. Note that these are genuinely normative facts. Normative claims are not synonymous with any descriptive claims whatsoever. "Reason" and "desire" do not mean the same thing. (Though, depending on what value theory is true, it might be that some descriptive facts determine the value facts. This is where supervenience comes in.)

Because normative reasons are connected to this normative ought, if you reject the latter (as I believe you do) then you reject the former too. Your theory is not really a normative one at all. You just adopt normative terms and redefine them to mean the same thing as various descriptive terms (e.g. desire). That seems a bit of a waste of time (why not just use the descriptive terms all along, and avoid the confusion?).

When I say that someone has reason to be moral, I am saying something distinct from the claim that morality would help fulfill their desires. (I am even saying something distinct from the claim that "acting morally would help fulfill desires generally", though I think there is a metaphysical connection between the two facts.)

Finally, you say: "The fact is, there is a difference between the desires that I have, and the desires that exist."

That is certainly true, and I wouldn't dream of denying it. What I do deny is that "reason" and "desire" are synonymous, or co-substitutable in the above sentence. My previous comment showed that there is no difference between reasons that I have to X, and reasons that exist for me to X. Indeed, to save time I will simply stipulate that these two sentences mean the same thing. If a fact counts in favour of my action, then, by definition, this provides me with a reason for so acting. I might not like it, and I might refuse to act on it. But that just goes to show that "normative reason" and "motivation" are two distinct things. It's possibly that my reasons could conflict with my desires, because my reasons might take into account a broader set of facts, perhaps including other people's desires. This will depend on our theory of reasons. It's not something you can infer just from the meaning of the terms.

Lippard said...

I don't think intentions reduce to desires... see Michael Bratman's _Intentions, Plans, and Practical Reason_ (1987, Harvard Univ. Press).

Alonzo Fyfe said...

Richard

I agree that "reasons" and "desires" are not synonamous.

When I say that desires are the only reasons for action that exist, I do not mean to claim that this is true by definition. Yet, it is still true. (Or, at least I have no evidence for, and no reason to suppose the existence of, any other type of reason).

You write, "When I say that someone has reason to be moral, I am saying something distinct from the claim that morality would help fulfill their desires."

Okay. I agree with this.

However, if you were to claim that, "Agent has a reason to be moral" is sometimes true in the real world in situations where "Being moral will fulfill Agent's desires," your statement would be false.

In other words, a statement like, "Agent has a reason to be moral independent of what will fulfill his desires" is false in the sense "Agent has a soul that is independent of his physical body" is false.

If a type of reason for action did exist that was not a desire then it would, indeed, be a "reason for action" but would not be a "desire". The coherence of this statement shows that you are correct in saying that they are not synonamous.

Yet, these other types of reasons for action still do not exist, and no person has a reason to do anything but that which will fullfill his desires.

Furthermore, all sorts of conceptual analysis on the meaning of "soul" can do nothing to establish the proposition "souls exist". Similarly, all sorts of conceptual analysis of "normative reasons" will not establish the existence of any reason for action other than desire fulfillment.

Alonzo Fyfe said...

Richard

I thought I would add some clarification on whether normative reasons exist.

I believe that desires are normative reasons and, since desires exist, normative reasons exist.

However, no other normative reasons exist.

The claim that 'normative reasons' is not synonamous with 'desires' still allows for the possibility of desires being normative reasons, just as 'cat' is not synonamous with 'leopard' allows for the possibility that leopards are cats.

If, however, you want to make the claim that desires cannot be normative reasons -- that the concept of normative reasons exclude desires -- then this will imply that normative reasons do not exist.

However, I would still say that I can do anything with desires that morality tends to want to do with 'normative reasons'.

This is similar to -- if you want to define 'mind' in such a way that it cannot possibly be the brain, then I will assert that 'minds' do not exist. Yet, still, every physical event that we attribute to 'minds' can be explained in terms of brain functions instead.

Richard Y Chappell said...

Alonzo, I'm still unsure why you don't allow for the possibility that someone else's desires might nevertheless provide me with normative reasons for action (independently of my own desires). You have asserted your position, but I don't see any arguments for it. Their desires exist just as mine do. Why are mine so special? (It isn't even clear why things other than desires couldn't constitute normative reasons. What is it about desires that gives them, and nothing else, this special normative status?)

Also recall that normative reasons are defined such that we really ought ("all things considered") to do what we have most normative reason to do. It then follows from your view that everyone really ought to be prudent, and not moral (if this conflicts with their own desires). That seems implausible. Surely if anything has normative signifiance, then morality does. If that seems too metaphysically "spooky" somehow, then - it seems to me - you would do better to embrace a fully-fledged normative skepticism. Conversely, if we recognize our own desires as having genuine normative significance, then why not extend this to other things too? Once normative reasons are in our ontology, there's no point being stingy in their application.

Alonzo Fyfe said...

Richard

At this point, I do not see your arguments as being anything other than semantic differences.

I have desires. Other people have desires.

"Prudential ought" considers all of the desires that I have (and will have).

"Moral ought" evaluates desires relative to all desires that exist.

It is consistent with this to say that other people's desires are relevant to answering moral ought questions.

I find it odd and confusing to say that I have a normative reason to do something based on somebody else's desires. Another person's desire is as disconnected from me as his left finger. A statement to the effect that I have a reason based on his desire is as odd as saying that I have 65 billion fingers because that is the total population of fingers.

You say 'normative reasons are defined such that we really ought to do what we have most normative reason to do.'

What is this really ought? How was it established that we really ought to do that which we have most normative reason to do?

Anyway, your solution seems to fix this definition of 'normative reason,' then to expand the set of relevant desires by saying that the desires of other people count as a normative reason that a person has, yielding what I think is a statement that is absurd as claiming that I have 65 billion fingers.

My solution is to stretch the concept of normative reason, and say that 'ought' does not concern reasons that the individual has, but reasons that exist. I am not committed to saying that an individual has the desires/reasons of 6.5 billion people. I am committed to saying only that the desires exist.

How do we resolve these types of disputes?

We don't need to. These options are fully consistent. Both individuals in this case are saying the same thing; they are just saying it in two different languages. The dispute is over which language seems the more intuitively obvious and less confusing. The dispute is not over any matter of substance.

On the other hand, neither of these two views are consistent with the idea that a person really ought to do that which fulfills his own desires and disregards the desires of others. Both respondents above reject this view. Niether should be interpreted as defending or implying it.