Is Practical Irrationality Possible in the Neo-Humean Account of Practical Reason Zhong Shiwen

Department of Philosophy, Zhejiang University

Introduction

Some philosophers have argued that principle of instrumental reason cannot direct our

adoption of ends. It may be argued that this would make practical irrationality impossible.1

When there is no principle to direct our choice on ends, we cannot say that someone violate

instrumental principle if he doesn’t take necessary means to the end. Because he may just

adopts something new as his end. Then this principle of practical reason is not normative and

cannot direct our action. The reason is that whenever someone hadn’t taking means to the end,

what we can say is that he has adopted a new end. But some philosophers, especially

Humeans respond that principle of instrumental reason is actually able to leave room for

practical irrationality when we realize that there is requirement of coherence and understand

desire appropriately.2

In this paper I would argue that Humean replies are not sufficient to answer these

objections, offered by Korsgaard, that instrumental principle cannot stand alone in order to be

normative. Some philosophers claim that even though there is no requirement of adopting

ends, there is requirement of coherence between means and ends. Some people would choose

worse means to achieve the end and will something impossible instead of what is easier to get.

They would be practically irrational in this sense. I’ll argue that the former situation is not

possible because willing means to the end is the same thing as willing most efficient means to

1 ChristineKorsgaard,TheNormativityofInstrumentalReason,TheConstitutionofAgency2 e.g.MichaelSmith,’InstrumentalDesire,InstrumentalRationality’,ProceedingofAristotelianSociety;AlanGoldman,Reasonfromwithin,OxfordUniversityPress;MarkSchroeder,SlaveofthePassions

the end; and the latter situation is too not irrational or cannot be evaluated by practical

rationality. Some philosophers claim that instrumental principle also could direct our

adopting ends because we would pursue the maximum of satisfaction of desire. I will argue

that maximum of satisfaction is based on false assumption about practical rationality. And it

couldn’t provide us with a standard to evaluate sets of desires. All these are what I will argue

at section 2. Meanwhile, in the section, I will argue that if there is really such requirement of

coherence, it must be based on our commitment to a desire in the first place. According to

Humean replies, there is an account of desire which could handle this question and try to

answer this objection to commitment to desire. This account is dispositional account of desire.

Therefore, in section 3, I will argue the dispositional account of desire is capable of making

sense of practical irrationality. But according to its logic, it will follows that the normativity

of instrumental principle has to be grounded on categorical imperative. I will try to argue that

the Kantian understanding of practical reason--namely, that both hypothetical imperative and

categorical imperative both are principle of practical reason--is on the right track.

1. Is it possible to be practical irrational?

In ‘The Normativity of Instrumental Reason’, Korsgaard has argued that the Humean

account of practical rationality would actually make practical irrational impossible.

According to the Humean view of motivation, what could be counted as reason for action

must be internal to existing motivation, and what could provide motivation to us is desire

instead of pure reason or belief.3 Actually, this idea has an important implication about

3 Oneoffamousargumentsforthispoint,seeBernardWilliams,Internalandexternalreasons,inMoralLuck.AndImeanultimatesourceofmotivationaredesires.Becausesurlyreasonorbeliefcouldprovidesusinstrumentaldesires,forexample,whenIwanttodrinkwaterandIhaveabeliefaboutmeansofrealizingitistobuyoneatstore

description of psychological states that cause action: what an agent does is to satisfy

whatever she desires mostly. This description tells us a non-normative fact about human

agent, because this is a description of fact that how human beings can be motivated. It tells us

what feature human being have rather than give them a standard that they can succeed or fail

to meet. Just like when we say that human beings are mortal, it tells us a fact about human

beings rather than a standard to follow. It is impossible to violate this truth, because someone

would not be a human being were she immortal. It also means that reasons are not able to

determine intrinsic desires (desires for its own sake) or ends, because reasons only play a role

in how to satisfy desire. If this desire-belief model of rational action is right, and motivation

of an action is provided by the combination of a belief and a desire, then there is only one

principle of practical reason, namely, instrumental principle: If you will an end, then you

ought to will the means to the end.

However, according to Korsgaard, this account of practical reason would result in a

problem. The instrumental principle itself doesn’t give us reason to determine our ends. So

this principle has to operate with some idea about how our ends are to be determined4.

However, there is only one principle of practical reason, namely, the instrumental principle,

which is unable to determine your ends. Then your ends would be determined by your

motivation and the motivation would move you to act. Your ends could only be what you are

going to pursue, as Korsgaard indicates:

ontheothersideofstreet,thenIwilldesiretogoacrossthestreet.4 IthinkKorsgaardassumedthatwillinganendalwaysprovidereasonstotakenecessarymeans.Butthisquestioniscontroversial.Seee.g.KieranSetiya,CognitivismaboutInstrumentalReason;R.JayWallace,Normativity,Commitment,andInstrumentalReason;JosephRaz,TheMythofInstrumentalReason;JohnBroome,NormativeRequirement.ButitisaproblemforbothKorsgaardandHumeantheory.I’mnotgoingtodiscussthisproblemhere.

The problem is coming from the fact that Hume identifies a person’s end as what he wants most, and the criterion of what the person wants most appears to be what he actually does. The person’s ends are taken to be revealed in his conduct. If we don’t make a distinction between what a person’s end is and what he actually pursues, it will be impossible to find a case in which he violates the instrumental principle.5

Her objection is that there should be distance between what a person actually did and what he

really want to do.6 If there is no such distance, we cannot judge whether someone violates

instrumental principle or not. For example, if I want to get a high score in final exam, I need

to go to read books. But a friend asks me to go to see my favorite movie. I finally go to watch

movie with my friends. In this situation, if there is no principle to direct my choice of ends,

getting high score or watching movie, I cannot be regarded as violated the instrumental

principle. Because I just give up my end to get higher score and take watching movie as my

end. The Humean account of practical reason think that whether a consideration can be

counted as reason depends on whether it will actually move an agent to act in fact, it will only

gives us an explanatory reason to action. By giving this kind of reason, we simply tell others

what is the motivation of an agent. If whatever an agent did is what she/he took be a

necessary mean to some desire, then how could an agent violate the instrumental principle. If

I cannot violate the instrumental principle which is to be seen by the Humean as the only

principle of practical rationality, it seems to follow that practical irrationality would become

impossible.

5 ChristineKorsgaard,TheNormativityofInstrumentalReason,TheConstitutionofAgency,p42.6 SomephilosopherssaysthatwhenKorsgaardtalksaboutthattheinstrumentalprinciplerequiredustotakemeanstoends,sheignoredthedistinctionbetweenhavinganendandpursinganend.e.g.JonathanDancy,PracticalReality,p46.Idon’tthinkthatKorsgaardhavemadesuchmistake.Forherownideaabouthypotheticimperative,shehasaddresseddifferencebetweenwillinganendandsimplydesiresomething,seee.g.TheSourceofNormativity,p36;forhercriticismtoHumeanism,shemeanswhataperson’sendisactuallywhathedidandwantedmostratherthaneverythinghewanted.Mayhewantsalotofthings,buthisendiswhathefinallydecidestopursueorhe’sgoingtopursue,Ithinkshehasmadeclearaboutthispoint.Ithink,inKorsgaard’sidea,havinganendispursuinganend.Ifyoujustwantsomethingordesiresomethinganddon’tcommittobringitabout,itisnotendatall.

Indeed, we can say that someone is irrational in taking some action because this action is

not a means to achieve the ends that she wants. But this is irrationality in theoretical

reasoning.7 True practical irrationality should look like this, as Korsgaard’s describes:

Howard, who is in his thirties, needs medical treatment: specifically, he must have a course of injections, now, if he is going to live past fifty. But Howard declines to have this treatment, because he has a horror of injections. Let me just stipulate that, were it not for his horror of injections, Howard would have the treatment. It’s not that he really secretly wants to die young anyway, or anything fancy like that. Howard’s horror of injections is really what is motivating him. ... [L]et’s again stipulate that he has not miscalculated or made a mistake. He sees that, if he were governed by considerations of prudence, he would have the injections: he agrees that a long and happy life is a greater good than avoiding the injections. But he still declines to have them: he chooses “his own acknowledg’d lesser good” 8

If we think that whatever actually motivated Howard is his reason, then he is perfectly

rational in this regard. He pursues what he wants mostly even when he acknowledges that

this is lesser good. Anyone who always gives in to temptation and abandons desire he

promised to bring into being would be rational, because his choice of the new end is

motivated by what actually moves him. As I have mentioned, we could always give a

motivation to an action. If the person were not moved by the consideration which caused him

to take the action, what we could say is that this consideration is not reason for him.

Therefore, we could never judge that someone is practically irrational. Actually, it means that

once an agent acted, she would be practically rational. And it is impossible for human agents

to avoid action and choice.9 So, in Korsgaard’s own words, the problem with Humean

account is not just that people don’t in fact ever violate the instrumental principle. Hume is

7 Forthispoint,seee.g.ChristineKorsgaard,Skepticismaboutpracticalreason,CreatingtheKingdomofends,p318-320;DerekParfit,OnWhatMatters,Vol.1,Chapter58 ChristineKorsgaard,TheNormativityofInstrumentalReason,TheConstitutionofAgency,p399 ChristineKorsgaard,Self-Constitution,p1.

actually committed to the view that people cannot violate it.10 So the account of practical

reason would make practical irrationality impossible. But if it is indeed a principle of

practical reason, it is supposed to be normative, to guide our actions. If the instrumental

principle is a what we are necessary to follow and unable to violate, how can it guide our

actions or deliberations? So it is not a normative principle to us.11

Then the Humean view meets two dilemmas. The first dilemma is that either instrumental

principle is trivially normative to us or it draws a distinction between what someone actually

pursues and what he really wants. The first way to distinguish them is to introduce end we

have reasons to do. The second way is to ascribe normative force to what we really wants. It

is that you should do what you really want even when you are tempted not to. So the second

dilemma is that the two ways in which the distinction could be drawn has to go beyond

instrumental principle. Korsgaard concludes that the instrumental principle at least could not

stand alone, unless something attaches normativity to our ends. And in her eyes, this thing is

unconditional principle.12

2. Requirement of Coherence

However, some philosophers disagree with this view. They think that even if there is only

one principle of practical reason, i.e.—the instrumental principle—practical irrationality is 10 ChristineKorsgaard,TheNormativityofInstrumentalReason,TheConstitutionofAgency,p40.11 Ibid,p31.Andsomepeopledoubtthatwhyanormhastobecouldbeviolated.Maybeitislikeprincipleofmath.Agreatmathematicianabsolutelywouldfollowbasicmathlawsandwouldn’tviolateit.Butsuchlawsarestillnormstothem.AndKantseemsalsoholdthat“ought”cannotapplytoperfectrationalwillforitwouldnotbesubjecttotemptation.HereIagreewithKorsgaardthatwhatmakesperfectrationalwillsubjecttonormativityisfreedominsteadofimperfection.Agreatmathematicianofcoursewouldnotviolatedbasicmathlaws,buthecandisregardthemifhewants.SeeIbid,p52,n39.12 Ibid,p64-66.Andinher‘Afterword,2008’,Korsgaardaddedthatinstrumentalprincipleisnotonlycouldn’tstandalone,butalsonotseparablefromthecategoricalimperative.Itjustpicksoutanaspectofthecategoricalimperative.Itdoesn’thaveindependentstatus.Thereisonlyoneprincipleofpracticalreasonthatiscategoricalimperative.Ithinkthispointisveryimportant.Korsgaardregardsinstrumentalprincipleasconstitutiveprincipleofaction.Ithinkitispossibletoviolateinstrumentalprincipleinherinterpretationonlybyaddedthispoint.Thisisanotherissue,Iwillnotdiscussitonthispaper.

still possible. They hold that we are able to evaluate different ends without any principle

beyond the instrumental principle.

The first strategy they employ is to argue that someone willing some impossible end are

practically irrational. And there is no any other requirement except the instrumental principle.

When an agent acts intentionally or desires something, she is in a distinctive psychological

condition--goal-directed state. To be in a goal-directed state, one is not going to match or

represent the content of such a states as the way things are in the world. On the contrary, one

is going to make the world match or fit one’s content of such a state.13 This means that an

agent would take steps to realize her goal if it has not been realized in the world. And Smith

claims:

Someone who desires a certain states of world obtains as a result of the world’s being one of the way is it can be which will lead to the occurrence of that state of affairs, and who believes that the way the world can be which will lead to the occurrence of the desired state of affairs is the p way. For this person to fail to desire that the world be the P way is, it seems to me, quite literally for them to be in an incoherent state of mind.14

In his view, when we intrinsically desire something, requirement of coherence would make

us adopt and desire necessary means to achieve such desire.15 Smith goes on to argue that

requirement of coherence is not limited to desire to means. It could be more extensive.

Necessity between means and ends is different from logic necessity. A means may be

necessary for achieving an end. But it is not guaranteed that once I take it I will satisfy the

end. When an agent wills an end, if she realizes that there are at least two ways to achieve the

13 Discussionaboutapplicationof“directionoffit”topropositionalattitudesorworldseee.g.MichaelSmith,TheHumeanTheoryofMotivation,sect6;JohnSearle,Intentionality,p7-914 MichaelSmith,InstrumentalDesires,InstrumentalRationality,p9815 Somephilosophersdisagreewithidea.Theydon’tthinkthereissuchcoherencebetweendesiresandmeans/endbeliefs.ForthisdiscussionseeJosephRaz,TheMythofInstrumentalRationality;NikoKolondy,TheMythofPracticalConsistency.Iwillnotdiscussthisproblem.WhatIwilldiscusshereisifhiscoherenceinterpretationofinstrumentalrationalitycouldbeanormativeadoptionofends.

end, and she is more confident about one of ways, coherence would requires her to desire the

way in which she’s more confident. Or she may be practically irrational.16 Jonathan Dancy

makes a similar position, arguing, that even if there is no requirement of adoption of ends, an

agent still can be practically irrational when she doesn’t adopt a way which she thinks to be

better.17 In their view, the agent is still possible to be practically irrational even there is no

principle to direct our choice on ends.

However, if we insist that our motivation to an action could be provided only by

combination of desire and belief, which implies that our motivation to an action is whatever

could satisfy our desire then it seems a little weird how such situation can be possible. I think

the mistake they have made is that they seems to regard willing necessary means to the ends

and willing most efficient means to the end as two separated parts. But I think these two parts

are the same thing, or at least inseparable. In what follows I will try to argue for this point.

Supposed that Smith and Dancy have assumed that Korsgaard is right at this point: if there

is no requirement of adopting ends, then whatever agent does is to taking necessary means to

her ends.18 Then the situation they proposed is impossible. The problem we meet here is how

to understand “willing means” appropriately. If we simply regard means as methods which

we take and which would achieve our ends, then I agree willing means to the ends is different

from willing most efficient to the ends. But the relation between means and ends are not

necessary in strict sense. We cannot say that once we take the certain means, I will achieve

16 MichaelSmith,InstrumentalDesire,InstrumentalRationality,p98-9917 JonathanDancy,PracticalReality,p46-4818 ActuallySmithdeniesthatthereisnowaytodirectourchoiceonendseventheonlyprincipleofpracticalreasonisinstrumentalprinciple.Iwilldiscusshisideaaboutthispointlater.Buthere,Ionlyattempttoarguethesituationthatsomeonewouldtaketheworsemeansthanmeansshe’smoreconfidentisimpossible.SoIwillputhisideaabouthowtodirectourchoiceofendstemporarily.

the end one hundred percent. It is still possible that we fail to achieve the end even if we have

taken the ‘necessary’ means to the end. It follows that we take means as something with

different degree of possibilities to achieve the ends. It means that when I regard something as

a method to achieve the end, I evaluate the chance to achieve the end by taking such way at

the same time. So when we regard something as means to an end, they are not simply ways to

achieve the end, but ways with different degree of possibilities. For example, if someone

wants to eat fish today, she has two methods: she can go to market buy one or go to catch one

by herself in the river. Because she is aware of means/end relation is not necessary but only

possible, she has to evaluate chance to satisfy her desire for fish. She’s aware that her skills

on fishing are terrible and she has no intrinsic desire to fishing. But it is also possible that

market is closed or fish is sold out. Right now, it is morning on an ordinary day, it seems that

there is little chance that market is closed or fish is sold out. So she will judge that going to

market is much more possible to secure her achievement of end. It is the possibility of

achieving ends that determines whether something can be counted as means or not. If all an

agent wants to do is to satisfy her desire, the reason she wills the means is that it is possible to

achieve the ends by this way. If the agent has no intrinsic desire to the certain means, what a

mean matters to the agent is the chance or possibility to achieve the end by taking such mean.

When we take something as means to pursue our ends, we are evaluating the possibility of

achieving ends by this way at same time.

Therefore, if whatever an agent, all she wants to do is simply to promote her desires, then

what she deliberates to do is how to promote desires. The answer to this question is such

means could promote the end in such degree of possibility. Whenever we judge what is

means to the ends, we are evaluating the possibility to achieve the ends by this means. What

can motivate us to adopt some mean consists in the fact that we will achieve some given end

with maximal possibility once we adopt this mean. If what motivates us to adopt some means

is possibility, other-things-being-equal, the higher degree of possibility is much more

attractive to agent. So willing means to ends is inseparable from willing most efficient means

to ends. The instrumental principle means willing the means to the ends we will, where the

means we will here is most possible means to pursue our ends. It is because we hope to

satisfy some desire that we will the means to satisfying the desire, and what we will consider

firstly in instrumental reasoning is what chance we could have. I think this is a description

about how human beings conduct instrumental reasoning. That’s the way in which we are and

how we are constituted by nature. An agent can make mistake in evaluating the possibility or

be unable to tell which is more possible. This is a problem in theoretical reasoning. However,

if she already believes that there is one method which is more possible than others, and she

accepts the instrumental principle then she would not will other ones.19 This would, as a

matter in fact, never happen in human agents. It means that human agents cannot be

practically irrational in this sense as Dancy and Smith think.

And Smith proposed another situation in which the instrumental reason is able to judge the 19 Tojudgeifsomethingisameanstotheend,wehavetojudgeifitispossibletoachieveendbythisway.Andwealsohavetoevaluatemoredetaileddegreeofpossibility.Butitdoesn’tmeanwecannotfailatthisstandard.Sometimespeoplemayberushonmakingdecision.Theyonlydistinguishwhatispossibleandimpossiblethenignoretoevaluatedegreeofpossibility.Actually,Idon’tthinkpeoplecouldintendtoignoretoevaluatethedegreeofpossibilityexceptthissortofsituation.Because,firstly,ifwedon’tevaluatedegreeofpossibility,weareunabletodistinguishimpossibleandpossiblemeanscompletely.Maywecouldtellobviouslyimpossibleandpossiblemeans,wecannotdosowhentheirdifferenceisnotapparent.Likegrading,wecouldtellwhodoesn’twriteanythingwouldfailtest,butifsomeonewritessomethingonpapers,wehavetoexamineitcarefully.Second,ifwetakeallpossiblemeansequal,wearenotbeabletodecidewhichonetotakeifotherconditionisequal,whichforcesustoevaluateitsdegreeofpossibility.

ends. Suppose we have two intrinsic desires, and one of them is stronger than the other one.

But I’m more confident about connection between weaker desires and means, and if this

confidence is stronger enough, it is more practically rational to will the means to the weaker

desires, and to dilute the stronger intrinsic desires.20

Firstly, I want to distinguish ends hard to achieve and ends impossible to achieve. Some of

ends are very hard to achieve, for example, bringing peace to the entire world, to be greatest

philosopher and so on. This means that it may be too complicated to achieve these ends. For

example, to bring peace to entire world, you may have to understand different nations’

culture and background, understand their psychology to avoid conflict and take too many

things to consider. It would perhaps require your enormous effort and persistence. To be a

greatest philosopher, you may have to always read and think, and have to resist temptation.

Even so there’s still no way to guarantee the accomplishment of such an end. Anyway, some

of ends are very difficult to achieve, but it is not completely impossible. There is still chance

to achieve it even though it may be very small. On the other hand, some of ends are just

impossible. There is no way to achieve them. These ends may violate physical laws or logical

principle. If someone wills to lift himself by pulling his hairs or be a happily married single,

then these ends are deemed to be impossible to achieve. There is no way or chance to satisfy

them at all.

For the ends hard to achieve, I don’t think that there could be any rational requirement on

willing the end easier to pursue. When someone wills something too great or difficult to

20 MichaelSmith,InstrumentalDesire,InstrumentalRationality,p100

achieve, but she keeps on trying to figure out methods to fulfill these ends and take every

possible chance to realize it, she does nothing irrational no matter whether she realized the

end or not. Some philosophers think that when we judge someone is irrational we refer to the

possibility of whether she deserves criticism like stupidity.21 In this sense, obviously she

doesn’t deserve such criticism. Even if we don’t appeal to irrationality in this sense, but only

refer to doing something wrong, she’s still not irrational. Actually, every great progress made

in history--no matter scientific, technological or political and so on—all go through

innumerable failures and generations’ effort. If there is requirement on willing the end which

we are more confident to achieve, all people who contributed to this achievement would be

irrational. It is in virtue of their courage to challenge and their perseverance that they are able

to make such great achievement. It’s true when someone believe he cannot satisfy his desire

A. Meanwhile, his desire B is also very strong, may be just next to A. And he’s very

confident to achieve B. It is rational for him to pursue B rather than A. However, if he still

chooses to pursue A, and he dose his best to contribute to achieve A. It also seems hard to say

he is irrational. No matter whether he abandon A or insist on A, he can be counted as rational.

So I don’t think that there is requirement on willing ends that are easier to pursue.

In the latter situation, ends impossible, I don’t think that willing this sort of ends could be

evaluated or judged by practical rationality if the only principle of practical reason is the

instrumental principle. When someone wills something impossible, there is simply no any

way to promote these ends. If there is no means to the end, what action can be counted as

21 Forexample,DerekParfit,OnWhatMatters,Vol.1,p33-35

practically rational or irrational action? If the principle of practical reason is nothing but the

instrumental principle (willing necessary means to the end you will), it follows that what is

practically rational to do is to will means to promote your end. But when there is no way to

achieve the end, it seems that the only thing we can do is sitting there, do nothing and simply

wish the end would be achieved (of course achievement is not caused by us). But siting here

and doing nothing is not the way to achieve the end. It doesn’t fulfill the requirement of

instrumental principle. Sitting there and doing nothing is neither practically irrational.

Because doing so obviously cannot be counted as violating instrumental principle. Some may

say it is practically irrational if someone takes actions in such circumstance. However, if

someone actually takes actions while being aware of impossibility of the ends, it also seems

weird to say that he violated the instrumental principle. Though all actions he takes are not

means to achieve the end, he doesn’t choose not to will the means to the end. If there is any

means to the end, he will take it. Therefore, if there is no requirement on ends, then this kind

of desire—willing impossible ends—is not irrational at all. Actually, I think that they cannot

be judged by practical rationality at all. Besides these situations, some of philosophers

disagree that there is no way to evaluate choice of intrinsic desire in Humean notion of

practical rationality.

However, according to Smith, requirement of coherence are not limited to coherence

between means and ends. He also thinks that there is coherence between desires. Such

requirement would make us be able to evaluate our desires. Some theorists support this view

and deny the claim that we cannot distinguish rational ends from irrational ones without

appealing to principle beyond instrumental principle.22

First of all, let me use an example to illustrate how one of strategies they employ may

work. Bruce has a strong intrinsic desire to gamble tomorrow for a day. But he also has an

important meeting to attend and he cares about his career. Moreover he loves his wife and

doesn’t want to disappoint her since she hates gambling. And he also cares about his health

and room since gambling always full of smoking and so on. To fulfill his desire to gamble,

several other desires of his own would be interfered. Then we would be able to judge that this

desire is irrational and he should restrict this desire in a certain way. The requirement of

coherence among desires makes a normative constraint on our adoption of ends. Donald

Hubin claims that ultimate ends can be instrumentally irrational in the sense that one has

reason to extinguish them.23 Alan Goldman also comments that deeper concerns may be

better satisfied by an option that mixes or coordinates satisfaction of the more specific

desires.24 It’s obviously rational for someone who likes money to perfer 100 dollars more

than to 1 dollar. Similarly, we may have various ends. But not all ends are compatible with

one another. Satisfaction of some ends would frustrate satisfaction of larger compatible set of

desires. It’s obvious rational to promote satisfaction of larger set of desires if we want to

satisfy our desires. What we ought to pursue is a maximum satisfaction of desires.

Korsgaard thinks that this view would meet a problem: what the maximum satisfaction

22 e.g.AlanGoldman,Reasonfromwithin;DonaldHubin,TheGroundlessNormativityofInstrumentalRationality23 DonaldHubin,TheGroundlessNormativityofInstrumentalRationality,JournalofPhilosophy,vol.98,No.9,2001,p,457.24 AlanGoldman,ReasonfromWithin,p.64,Inmyexample,itseemsIdon’tassumedthere’sadeepconcernbehindthedesires.Butwecanassumewhatbehindhisvariousdesiresisdesiretobehappy.Allofthesedesirescouldcontributetohishappiness.However,Idothinkthatdeepconcernneedtobediscussedparticularly,Iwilldosolater.

is.25 According to such view, the rational decision is to pursue maximum of satisfaction of

our desires. But how can we measure satisfaction of which sets of desires is larger? The items

of these sets have prime facie normative weight—given how strongly you desire them. But

this prime facie weight doesn’t settle question of normative ranking, for example a very

strong desire may be suppressed for the sake of maximizing the total.26 The coherent way for

this idea to rank total and particular satisfaction is to balance the individual weight of

satisfaction and weight of total satisfaction. However, the satisfaction of desire is ambiguous.

Because “satisfaction” could refer to objective or subjective state. When your desire is in fact

realized, objective satisfaction is achieved. When you are aware that objective satisfaction is

achieved, you would feel sort of pleasure, which is subjective satisfaction.27 So subjective

satisfaction is conceptually dependent on objective satisfaction. If we are going to maximize

subjective satisfaction, its importance depends on importance of objective satisfaction. If we

take objective satisfaction as what we claim to maximize, the maximum of satisfaction means

maximizing the number of satisfied desires. But some of desires are much more important to

our lives, maximizing satisfaction must have something to do with giving priority to the

things that matter more to us. Therefore, we are going to assign weight to desires, the weight

should be based on something psychological, something about our own attitudes toward them.

Then the measure must be provided by some subjectively identifiable or psychological

quantum instead of degree of satisfaction. Such subjectively identifiable measure is intensity

25 SeeChristineKorsgaard,TheMythofEgoism,TheConstitutionofAgency,Especiallyp75-76,p95-9826 Ibid,p7527 Ibid,p95.AnditisimportanttonoticethatKorsgaardmakeitclearthatthisideaisdifferentfromhedonism.Sothepleasureisbasedonknowledgeofobjectivesatisfactionobtainedinsteadofgeneralpleasure.

of desire.28 But if we take the intensity of desires as the standard to direct our choice on ends,

we at least meet two problems.

Firstly, we should remember that the intensity of desire is not simply what we pursue, but

it is also what motivate or force us to act. If we face two desires, and the intensity of one of

them is stronger than the other one, I will inevitably be motivated by the one with stronger

intensity. It is a description about how we act. What we are going to pursue rationally is the

maximum satisfaction of desires which turns out to be the highest degree of intensity of

desires. In other words, what you are going to pursue is what you want most. Then we go

back to the problem we start from again, namely, how it is possible for the instrumental

principle to be violated. It is possible that one desire is stronger than others, but the added

intensities of other desires are stronger than this one. When someone was motivated to act

according to desire with higher degree of intensity, she didn’t realized that if she adds up all

intensities of other desires, the degree of intensities of these added desires is much higher

than intensity of that single one. Or sometimes she has abstract desire which generates

various specific desires, and perhaps she hasn’t realized that what she really wants is that

abstract or deeper desires, mistaking what she tries to pursue as these specific desires

themselves. As Alan Goldman says, options are to be preferred when they can be predicated

to maximize the satisfaction of one’s deepest and broadest concerns over times.29 The agent

would violate this principle in this sense. However, as Korsgaard has pointed out, this idea

28 Ibid,p95-98.HereKorsgaardtriestoarguethatsatisfactionisnotwhategoistclaimtobeneutral.ButIwillusethisconclusiontoargueproblemthatdesire-basedtheorywillmeet.29 AlanGoldman,ReasonfromWithin,p64

was based on false assumption about practical rationality.30 In such cases, the agent is not

such that she will not be motivated by ends with highest degree of intensity of desires or her

deepest concern she really wants. She is someone who doesn’t know what she really wants or

what her deepest concern is. Once she knows what she really wants or what her deepest

concern is, she will inevitably act in conformity to it. If we regard this situation as practically

irrational, then practical rationality turns to discover or reveal knowledge of our

psychological states. However, practical reason is aimed to determine what we ought to do.

Uncovering our psychological states or content seems to be the aim of theoretical reasoning.

The other problem is that this standard cannot explain a kind of irrationality in the example

about Howard in section one. Howard acknowledges that to accept injection in order to have

a longer happy life is greater good than his fear of injection. But the desire to avoid injection

is stronger and he finally chooses to reject injection. If the requirement of coherence finally

requires us to pick up the desire with highest intensity, it seems that Howard is rational. But

it’s quite ridiculous, because this choice is not reasonable even from Howard’s own

perspective. It’s common that we feel guilty when we abandon something we sincerely hope

and choose something in conflict with what we want. In the other words, weakness of will is

impossible if we try to direct our choices of desires by their intensity. It’s true that when we

think someone who gives in to temptation and pursues a desire which is conflict with some

desires that he had been committed to before, we think he breaks the requirement of

coherence. But this means that we have already give some desires a status above others. So

30 ChristineKorsgaard,TheMythofEgoism,TheConstitutionofAgency,p74.

even if a desire is stronger than the desire I decided to pursue before, I still should pursue

what I promise to satisfy. Only by giving a desire status above others, the requirement of

coherence can require us to hold that it is irrational to pursue a desire, no matter how strong

its intensity is, if it conflicts with desire we promised to bring.

Anyway, I don’t think that the requirement of coherence could help us to evaluate our

desires, because it cannot provide us with a clear standard to measure various desires. Such

requirement asks us pick up a maximal, coherent and unified desire set. But we just cannot do

so. For if we take largest number of satisfied desires as such set we certainly ignore priority

we attribute to different desires. If we take this priority attributed to desires as its intensity,

we would rule out the difference between practical reason and theoretical reason. Actually I

don’t deny that the requirement of coherence could evaluate desires to some degree. But it

has a condition, which is that we have to be committed to a desire in the first place, and then

such requirement of coherence could judge a desire should be abandoned or had by

examining whether it is in conflict or compatible with the desires we have been committed to.

There is another view, advanced by some followers of Hume, which seems is able to

overcome these two problems. I will discuss this view in the next section, trying to argue that

if it works, it will have to ground the normativity of the instrumental reason in the categorical

imperative.

3. Dispositional Desire and Motivation of Action

Some theorists think that if we have understood desire appropriately, then these objections

are based on misunderstanding about desires. Many philosophers interpret desire as brute

force or impulse. For example, Harry Frankfurt indicates:

However imposing or intense the motivational power that the passions mobilize may be, the passions have no inherent motivational authority. In fact, the passions do not really make any claims upon us at all. Considered strictly in themselves, apart from whatever additional impetus or facilitation we ourselves may provide by acceding to them, their effectiveness in moving us is entirely a matter of sheer brute force. There is nothing in them other than the magnitude of this force that requires us, or that even encourages us, to act as they command.31

But many people disagree with this interpretation of desire. They prefer to support the thesis

that desire in the Humean account of practical reason is dispositional desire rather than brute

or immediate motivational force. Motivation produced by desire can be affected by vivid

sensory or imaginative representation of its object. The vivid sensory or imaginative

representation can increase the intensity of desire. Motivational forces at the moment of

action actually are combination of baseline provided by the strength of dispositional desire

and the effect of vivid image of the objects. If an agent has stronger dispositional desire for A

than dispositional desire for B, then she will, generally speaking, choose A over B. But it is

not always so. If the agent is in some circumstance which brings her some vivid sensory

representation of B, the representation may amplify motivational force of B. Then she may be

motivated to pursue B rather than A. For example, Natasha is on a diet. Her dispositional

desire to care about her health and body shape is stronger than her desire to eat chocolate.

She could keep a healthy diet until chocolate is present before her. The presence of chocolate

would amplify her desire for it. Finally she chooses to eat chocolate. The Humean account of

practical rationality refers to the baseline of dispositional desire as motivation we should act

on rather than motivation at the moment of action. Proponents of this account think that it

31 HarryFrankfurt,Autonomy,Necessity,andLoveinNecessity,VolitionandLove,p137.

could escape criticism to the effect that practical irrationality is impossible. Because rational

action is an action caused by dispositional desire which is not motivational force of action at

the moment. This account of relation between desire and motivation successfully distinguish

what someone actually wants and really wants without making appeal to any other principle

than the instrumental principle.32

The first question which the account of Humean practical rationality meets is that it has to

abandon what Mark Schroeder called Classical Argument about the Humean theory of

practical reason.33 This argument implies that having reason requires having motivation and

desires. It would incur a commitment, that is, rational action is to do what is most motivated

to do, which makes irrationality impossible. Those who advocate the view of dispositional

desire think that if we interpret Humean rational action as action according to dispositional

motivation then we can escape from this criticism. But can this theory of practical reason

really escape from this objection as they think? I don’t think so.

The problem that who advocates view of dispositional desire would meet is how we can

commit ourselves to a desire, turn it to be dispositional desire and make it to have normative

force to us. When we are in calm passion, and feel one of desires is stronger than other

desires, we would take this desire as an end we want to satisfy. But when we take it as end,

we are committed to achieving this desire. According to the proponents of this theory, we can

see that dispositional desire cannot be something that we try to satisfy at the moment. It has

to exist over time and we have to continue to keep such desire, which means that we have to

32 FortheentireargumentseeNeilSinhababu,TheHumeanTheoryofPracticalIrrationality,JournalofEthicsandSocialPhilosophy,Vol.6,No.1,201133 seeMarkSchroeder,SlaveofthePassions,Chapter7-8

try to follow it and satisfy it in the future. When there are other conflicting desires arising in

us, we have to resist them and insist on dispositional desires. This gives us a standard that we

can fail or success to meet. In Kant’s words, willing an end is committing yourself to be the

cause of that end. Willing an end implies a commitment to taking available causal

connections in order to achieve the end. Therefore, commitment to taking means to the end is

constitutive of willing. If you don’t will the means to the end, you are not willing the end at

all. I think that proponents of view of dispositional desire take continuing to desire something

is constitutive of dispositional desire to it in the same way. Dispositional desire is not an

occurent impulse or immediate motivational force. We have to continue to desire something

for certain time. If we don’t continue to desire it, we don’t have dispositional desire for it at

all.

We can see that proponents of view of dispositional desire take it as what we really want

rather than occurent impulse. Such dispositional desire represents who we are, at least part of

who we are. If we can keep desiring something that we take to be the object of our

dispositional desire, we maintain the unity of the self. If we are tempted by something else

and abandon dispositional desire, we would lose our identity, at least part of it. However,

where is authority of this dispositional desire from? It’s true that when we have a

dispositional desire for something, we have to continue to keep the desire. And we take this

desire is what we really want. But the ‘me’ in the future would be different from the ‘present’

me. Because who I am is determined by content of my consciousness. The content of

consciousness may be different at different moments. Today, I may take desire for x as my

dispositional desire, and identify with it. But tomorrow, I may find that I have another desire.

I try to take this desire as what I really want and representing who I am. As I have argued,

dispositional desire requires that I continue to desire the same thing in the future. Then the

‘me’ in the tomorrow will have to be constrained by the present ‘me’.34 How can I make the

‘present’ me a status above me in the future? What is the source of this authority? Actually, if

we regard dispositional desire as representing our identity and regard it is what we really

want, then dispositional desire is constitutive of unity of the self. The dispositional desire

doesn’t represent our identity at a moment. It represents our identity over time which means

‘me’ at different times could be identified by the same thing. Then it is possible for us to say

that dispositional desire is what we really want rather than occurent impulse. By having such

dispositional desire, I could integrate ‘me’ at different times into a unified identity or self

over time. If we don’t own dispositional desire, we would be wanton that follows whatever

impulse arose in us35. To own dispositional desire is to have a unified self. But I think the

unity of the self depends on a formal version of categorical imperative which is principle that

our maxim must be willed as universal laws.

Firstly, I want to make clear what maxim is. We human beings are self-conscious agents,

which means that when we try to satisfy a desire or pursue an end, we would realize that we

are ‘doing action A in order to achieve purpose P’. This is what Kant calls maxim. When we

try to pursue our dispositional desire, we act on a maxim that can be universally willed. For

34 AsimilarcriticismseeJ.D.Velleman,DecidingHowtoDecideinThePossibilityofPracticalReason35 Hedonismisdifferentfromsomeonewhoawantonwhocompletelyfollowsdesires.Becausetheyactuallyactonaprinciple:alwaysactondesiresthatprovidelargestpleasure.Forentireargument,seeAndrewsReath,Hedonism,Heteronomy,andKant’sPrincipleofHappiness,inhisAgencyandAutonomyinKant’sMoralTheory.

example, if Natasha tries to satisfy her dispositional desire to keep fit, she should refuse to eat

desert. She needs to follow this maxim—refuse to eat desert in order to keep fit—in all

similar situations unless at some extraordinary condition. Whenever Natasha has desire to

keep fit, now matter how slight it is, she has to refuse to eat desert except that there some

extraordinary situation like in which, for example, someone kidnaps her and threatens to kill

her if she doesn’t eat desert. If she don’t universally wills this maxim, whenever there is

anything that could seduce her eat desert while she has desire to keep fit, she is not

committed to the desire to keep fit at all. If my commitment to maxim is not universal, I will

always be seduced by temptation, I cannot say I commit to my dispositional desire at all.36

Dispositional account of desire requires that there be an integrated self over time, which

means that there must be something normative that can be shared by me at different times. If

such a maxim or reason embodying it is not of universality in some sense, then we cannot

integrate ourselves at different times. So maxim or reason embodying such maxims must be

universal, and then my present and future self would be integrated as a unified self. Thus, it

seems that maxims on which we act must be universalizable if we are to make sense of the

unity of the self or personal identity. Moreover, if I want a reason to be normative to me at all

times, it can be also said that reason is normative to all agents. If my reason is not

universalizable, it cannot be normative to other agents. Meanwhile, if my reason is not

universalizable, it is not normative to ‘me’ at different times. If the reason of the present me

is not normative to the future me, the future me of course can reject this reason. Then I am

36 DetaileddiscussionaboutparticularwillinganduniversalwillingseeChristineKorsgaard,Self-Constitution,section4.4.

unable to integrate myself over time again. So I think that if I could regard dispositional

desire as what I really want, the maxim which I am to act when I try to pursue what I really

want, it must be laws that can be universally willed. Or there cannot be dispositional desire at

all. It’s apparent that the instrumental principle still cannot alone. It can only works with the

help of the categorical imperative.

4. Conclusion

In this paper, I have tried to examine various proposed situation about practical

irrationality if there is only one principle of practical reason—the instrumental principle. I

have argued that it is impossible for us to be practically irrational in choosing means.

Because willing means to the end is just willing most efficient means to the end. And there is

no practical rational requirement on choice of ends that we are more confident to achieve. I

also have argued that the claim, if there is only the instrumental principle, we still can be

irrational in choosing ends, meets a lot of problems. Maximum of satisfaction of desire is

based on a false assumption of practical rationality. To be irrational in this sense is theoretical

irrational. Dispositional account of desire take dispositional desire to represent what we really

wants. But the only way that dispositional desire can represent who we are and what we

really want depends on some formal version of the categorical imperative.

References Anscombe,G.E.M,Intention,Ithaca:CornellUniversityPress,1957Blackburn,Simon,RulingPassions,Oxford:ClarendonPress,1998Dancy,Jonathan,PracticalReality,Oxford:OxfordUniversityPress,2000Frankfurt, Harry G. ‘Autonomy, necessity, and love.’ Necessity, volition, and love, Cambridge:CambridgeUniversityPress,1999 Goldman,Allan,ReasonfromWithin:DesireandValues,NewYork:OxfordUniversity,2009Hampton,Jean, ‘OnInstrumentalRationality’, inJ.B.Schneewind(ed.),Reasons,Ethics,andSociety,

Chicago:OpenCourt,1996Hubin,Donald,TheGroundlessNormativityof InstrumentalRationality, Journal of Philosophy, vol.98,No.9,2001Kant, Immanuel, Groundwork of the Metaphysics of Morals (Cambridge Texts in the History ofPhilosophy).TranslatedandeditedbyMaryGregorwithanIntroductionbyChristineM.Korsgaard.Cambridge:CambridgeUniversityPress,1998Korsgaard, Chrisitine M, ‘The Normativity of Instrumental Reason’ in The Constitution of Agency,NewYork:OxfordUniversityPress,2008-----‘TheMythofEgoism’inTheConstitutionofAgency,NewYork:OxfordUniversityPress,2008-----‘ActingforReason’inTheConstitutionofAgency,NewYork:OxfordUniversityPress,2008-----‘Skepticism about Practical Reason’ in Creating the Kingdom of Ends, Cambridge: CambridgeUniversityPress,1996-----SelfConstitution:Agency,IdentityandIntegrity,NewYork:OxfordUniversityPress,2009-----TheSourcesofNormativity,Cambridge:CambridgeUniversityPress,1996Nagel,Thomas,ThePossibilityofAltruism,Princeton:PrincetonUniversityPress,1970Parfit,Derek,OnWhatMatters,Oxford:OxfordUniversityPress,2011Quinn,Warren.‘Puttingrationalityinitsplace.’inMoralityandaction,Cambridge,CambridgePressUniversity,1993Raz, Joseph,TheMythof InstrumentalReason, JournalofEthicsandSocialPhilosophy,Vol.1,No.1,April2005Reath, Andrews, ‘Hedonism, Heteronomy, and Kant’s Principle of Happiness.’ In Agency andAutonomyinKant’sMoralTheory,Oxford:ClarendonPress,2006Schapiro,Tamar,WhatareTheoriesofDesireTheoriesof,AnalyticPhilosophy,Vol,55,No.2,2014Schroeder,Mark,SlaveofthePassions,Oxford:OxfordUniversity,2007Schueler,GeorgeF.Desire:Itsroleinpracticalreasonandtheexplanationofaction.Cambridge:TheMITPress,1995. Setiya,Kieran,CognitivismaboutInstrumentalReason,Ethics,Vol.117,No.4,July,2007Sinhababu,Neil,TheHumeanTheoryofPracticalIrrationality,JournalofEthicsandSocialPhilosophy,Vol.6,No.1,2011Smith, Michael ’ Instrumental Desire, Instrumental Rationality’, Proceeding of Aristotelian Society,SupplementaryVol.78,2004Velleman,J.David,‘DecidingHowtoDecide’inThePossibilityofPracticalReasons,Oxford:ClarendonPress,2000Wallace,RJay,‘Normativity,Commitment,andInstrumentalReason’inNormativityandWill,Oxford:ClarendonPress,2006Williams, Bernard, ‘Internal and External Reason’ inMoral Luck: Philosophical Papers 1956-1972,Cambridge:CambridgeUniversityPress,1981