(1)

Julie choseb overa, even though she knewb was more expensive thana.

There is nothing puzzling about Julie’s choice. Perhaps Juliewas choosing among vacation options, andb was a week’svacation in Paris, whilea was a week’s vacation inPeoria. In any event, Julie evidently took the overall merits ofb to outweigh those ofa, even ifb wasinferior from a financial standpoint.

(2)

Jimmy opted ford overc, despite his judgingc to be a healthier choice thand.

Again, we find Jimmy’s decision unremarkable. Perhapscandd were competing dessert options:c, let ussuppose, was a dish of dry Wheaties, rich in fiber and whole grain,whereasd was a gossamer-light yet oh-so-rich Valrhonachocolate mousse. Jimmy obviously (and reasonably!) assessedd as the better dessert option all things considered, eventhough he knewd was less good for his health thancwould be. Nothing puzzling about that.

(3)

Joseph didf rather thane, even though he wasconvinced thate was the better thing to do all thingsconsidered.

Here, by contrast, we have a genuinely puzzling case, one we cannotmake sense of in the same way. Why would Joseph dof when heassessede as the superior course of actionall thingsconsidered? Joseph’s choice sounds so inexplicable that wemight even query whether the case has been accurately described. IfJoseph is really freely choosingf overe, we mightthink it questionable that he genuinely assessese as betterall things considered. Perhaps he actually takesf to besuperior (for him, under the circumstances), although he thinks mostpeople would opt fore or would saye was a betterchoice.

Our divergent reactions to these three examples point to somethingdistinctive about the judgment that one course of action is betterthan another. (Betteroverall, or better all thingsconsidered, that is—not simply better in some respect.) Suchjudgments appear to enjoy a special connection to the agent’sactions which other judgments do not possess. We are puzzled byJoseph’s choice precisely because we expect people’sactions—at least when freely undertaken—to reflect theiroverall assessment of the merits of the alternative courses of actionbefore them. And Joseph’s—at least as reportedabove—doesn’t. When assessment and action are said to havediverged in this way, we are often sceptical: we question whether theagent really held the course of action he considered but rejected tobe better. Moreover, even when we accept the description of the case,we find such action somehow puzzling, defective, or dubiouslyintelligible, in a way that action contrary to one’s judgmentsof financial wisdom (for example) is not. It seems that the judgmentcontrary to which Joseph acts—the judgment that one course ofaction is better than another—has what we can vaguely term aspecial character, in comparison with other judgments such as that onecourse of action would be healthier or more costly than another.

Let us give a name to the assessment of his options contrary to whichJoseph acts. Let us call his judgment thate is a betterthing to do all things considered Joseph’sbetterjudgment. (“Better judgment” does not mean“superior judgment”; it simply means a judgment as towhich option is overall better.) Joseph, then, appears to have acted,freely and intentionally, contrary to his better judgment. This isprecisely the phenomenon the philosophical tradition calls“weakness of will.”^[1] Philosophers have been perplexed by or dubious about such action fora very long time.^[2] Indeed, Plato’s Socrates famously denied its possibility in theProtagoras. “No one,” he declared, “whoeither knows or believes that there is another possible course ofaction, better than the one he is following, will ever continue on hispresent course” (Protagoras 358b–c).^[3] Philosophers have been wrestling with the issue ever since. It is notsurprising that weakness of will has such a long and distinguishedpedigree as a topic of philosophical discussion: it is both anintrinsically interesting phenomenon and a topic rich in implicationsfor our broader theories of action, practical reasoning, rationality,evaluative judgment, and the interrelations among these.

• 1. Hare on the Impossibility of Weakness of Will
• 2. Davidson on the Possibility of Weakness of Will
• 3. The Debate After Davidson
  • 3.1 Internalist and Externalist Strands
  • 3.2 Weakness of Will as Potentially Rational
  • 3.3 Changing the Subject
  • 3.4 Broadening the Discussion
• Bibliography
• Academic Tools
• Other Internet Resources
• Related Entries

1. Hare on the Impossibility of Weakness of Will

Let us commence our examination of contemporary discussions of thisissue in appropriately Socratic vein, with an account that givesexpression to and builds on many of the intuitions that lead us to besceptical about reports like (3) above. For the moral philosopher R.M. Hare—as for Socrates—it is impossible for a person todo one thing if he genuinely and in the fullest sense holds that heought instead to do something else. (If, that is—to echo theearlier quote from Socrates—he “believes that there isanother possible course of action, better than the one he isfollowing.”) This certainly seems to constitute a denial of thepossibility of akratic or weak-willed action. In Hare’s case itis a consequence of the general account of the nature of evaluativejudgments which he defends (Hare 1952; see also Hare 1963).

Hare is much impressed by what we vaguely referred to above as the“special character” of evaluative judgments, such as thejudgment that one course of action isbetter than another, orthat oneought to do a certain thing.^[4] Such “value-judgments” (Hare’s term) seem differentfrom merely “descriptive” judgments like that one thing ismore expensive than another, or rounder than another (Hare 1952, ch.7; Hare 1963, ch. 2). In particular, evaluative judgments seem to beara special connection toaction which no purely descriptivejudgment possesses. We can see that Hare’s analysis begins fromsomething like the data we rehearsed above. Here is how he developsthose data into a theory. He identifies as the fundamental distinctivefeature of evaluative judgments—the feature responsible fortheir special character—that such judgments are intended toguide conduct. (See, e.g., Hare 1952, p. 1; p. 29; p. 46; p.125; p. 127; p. 142, pp. 171; Hare 1963, p. 67; p. 70.) The specialfunction of evaluative judgments is to be action-guiding: that is, ifyou will, what evaluative judgments arefor. Hare thenproposes a more precise gloss on what it is for a judgment to“guide conduct”: an action-guiding judgment is one whichentails an answer to the practical question “What shallI do?” (Hare 1952, p. 29; see Hare 1963, p. 54 for theterminology “practical question”). What is it, though,that an action-guiding judgment must entail? In other words, whatwould count as an answer to the question “What shall Ido?” Hare holds that nodescriptive statement canconstitute an answer to such a question; only a first-personcommand orimperative can answer it (Hare 1952, p.29; p. 46). The relevant first-person imperative could be expressed inwords as “Let me doa” (Hare 1963, p. 55).

To recap the argument thus far: it is the function of evaluativejudgments like “I ought to doa” to guideconduct. A judgment can guide conduct only if it entails an answer tothe question “What shall I do?” An answer to that questionmust take the form of the first-person command or imperative“Let me doa.” Therefore, first-person evaluativejudgments must entail first-person imperatives (Hare 1952, p. 192).Now in general, if judgmentJ₁ entails judgmentJ₂, then assenting toJ₁ mustinvolve assenting toJ₂: someone who professed toassent toJ₁ but who disclaimedJ₂ would be held not to have spoken correctly whenhe claimed to assent toJ₁ (Hare 1952, pp.171–2). Assenting to the evaluative judgment “I ought todoa,” then, involves assenting to the first-personcommand “Let me doa” (Hare 1952, pp.168–9). But what is involved in sincerely assenting to afirst-person command or imperative of that type? Just as sincereassent to a statement involvesbelieving that statement,sincere assent to an imperative addressed to ourselves meansdoing the thing in question if we can:

It is a tautology to say that we cannot sincerely assent to a …command addressed to ourselves, andat the same time notperform it, if now is the occasion for performing it and it is in our(physical and psychological) power to do so. (Hare 1952, p. 20)

So: provided it is within my power to doa now, if I do notdoa now it follows that I do not genuinely judge that Iought to doa now. Thus, as Hare states at the very openingof his book, a person’s evaluative judgments are revealed by hisactions and choices:

If we were to ask of a person ‘What are his moralprinciples?’ the way in which we could be most sure of a trueanswer would be by studying what hedid…. It would bewhen … he was faced with choices or decisions betweenalternative courses of action, between alternative answers to thequestion ‘What shall I do?’, that he would reveal in whatprinciples of conduct he really believed. (Hare 1952, p. 1)

Note that Hare is not simply saying that a person’s actions arethe most reliable source of evidence as to his evaluative judgments,or that if a person didb the most likely hypothesis is thathe judgedb to be the best thing to do. Hare is saying,rather, that itfollows from a person’s having doneb that he judgedb best from among the options opento him at the time. On this view, akratic or weak-willed actions as wehave understood them are impossible. There could not be a case inwhich someone genuinely and in the fullest sense held that he ought todoa now (wherea was within his power) and yet didb. On Hare’s view, “it becomes analytic to saythat everyone always does what he thinks he ought to [if physicallyand psychologically able]” (Hare 1952, p. 169).

Butdoes everyone always do what he thinks he ought to, whenhe is physically and psychologically able? It may seem that this issimply not always the case (even if it isusually the case).Have you, dear reader,never failed to get up off the couchand turn off the TV when you judged it was really time to startgrading those papers? Have younever had one or two moredrinks than you thought best on balance? Have youneverdeliberately pursued a sexual liaison which you viewed as an overallbad idea? In short, have younever acted in a way whichdeparted from your overall evaluation of your options? If so, let mebe the first to congratulate you on your fortitude. While weak-willedaction does seem somehow puzzling, or defective in some important way,it does nonetheless seem to happen.

For Hare, however, any apparent case ofakrasia must in factbe one in which the agent is actuallyunable to doa, or one in which the agent does not genuinely evaluatea as better—even if he says he does.^[5] As an example of the first kind of case Hare cites Medea (Hare 1963,pp. 78–9), who (he contends) is powerless, literally helpless,in the face of the strong emotions and desires roiling her: she istrulyunable (psychologically) to resist the temptationsbesieging her. A typical example of the second kind of case, on theother hand, would be one in which the agent is actually using theevaluative term “good” or “ought” only in whatHare calls an “inverted-commas” sense (Hare 1952, p. 120;pp. 124–6; pp. 164–5; pp. 167–171). In such cases,when the agent says (while doingb) “I know I reallyought to doa,” he means only that mostpeople—or, at any rate, the people whose opinions on suchmatters are generally regarded as authoritative—would say heought to doa. As Hare notes (Hare 1952, p. 124), to believethis is not to make an evaluative judgment oneself; rather, it is toallude to the value-judgments of other people. Such an agent does nothimself assess the course of action he fails to follow as better thanthe one he selects, even if other people would.

No doubt there are cases of the two types Hare describes; but they donot seem to exhaust the field. We can grant that there is the oddmurderer, overcome by irresistible homicidal urges but horrified atwhat she is doing. But surely not every case that we might be temptedto describe as one of acting contrary to one’s better judgmentinvolves irresistible psychic forces. Consider, for example, thefollowing case memorably put forward by J. L. Austin:

I am very partial to ice cream, and a bombe is served divided intosegments corresponding one to one with the persons at High Table: I amtempted to help myself to two segments and do so, thus succumbing totemptation and even conceivably (but why necessarily?) going againstmy principles. But do I lose control of myself? Do I raven, do Isnatch the morsels from the dish and wolf them down, impervious to theconsternation of my colleagues? Not a bit of it. We often succumb totemptation with calm and even with finesse. (Austin 1956/7, p. 198)

(I might add that it also seems doubtful that irresistible psychicforces kept you on the couch watching TV while those papers werewaiting.) As for the “inverted-commas” case, this toosurely happens: people do sometimes pay lip service to conventionalstandards which they themselves do not really accept. But again, itseems highly doubtful that this is true of all seeming cases ofweak-willed action. It seems depressingly possible to select andimplement one course of action whilegenuinely believing thatit is an overall worse choice than some other option open to you.

Has something gone wrong? We started with the unexceptionable-soundingthought that moral and evaluative judgments are intended to guideconduct; we arrived at a blanket denial of the possibility of akraticaction which fits ill with observed facts. But if we are disinclinedto follow Hare this far we should ask what the alternative is, for itmay be even worse. For Hare, the answer is clear: our only otheroption is to repudiate the idea that moral and other evaluativejudgments have a special character or nature, namely that of beingaction-guiding. For we should recall that Hare presents all hissubsequent conclusions as simply following, through a series of steps,from that initial thought. “The reason why actions are in apeculiar way revelatory of moral principles is that the function ofmoral principles is to guide conduct,” Hare continues in thepassage quoted earlier (Hare 1952, p. 1). For Hare, then, the only wayto escape his “Socratic” conclusion about weakness of willwould be to give up the idea that evaluative judgments are intended toguide conduct, or to “have [a] bearing upon our actions”(Hare 1963, p. 169; see also Hare 1952, p. 46; p. 143; p. 163; pp.171–2; and Hare 1963, p. 70; p. 99).

The choices before us, then, as presented by Hare, are Hare’sown view, or one which assigns no distinctive role in action orpractical thought to evaluative judgments, treating them as just likeany other judgment. We might call the first of these an extremeversion of (judgment)internalism. (This label is used tomean many different things; here, I mean by it that certain judgmentshave an internal or necessary connection to motivation and to action.)By extension, we might usefully follow Michael Bratman in calling thesecond type of view “extreme externalism” (Bratman 1979,pp. 158–9).

Extreme externalism also seems unsatisfactory, however. First, itseems unable to explain why there should be anything perplexing orproblematical about action contrary to one’s better judgment,why there should be any philosophical problem about its possibility orits analysis. On this kind of view, it seems, Joseph’s choice((3) above) should strike us as no more puzzling than Julie’s orJimmy’s ((1) or (2)). As Hare puts it:

On the view that we are considering, there is nothing odder aboutthinking something the best thing to do in the circumstances, but notdoing it, than there is about thinking a stone the roundest stone inthe vicinity and not picking it up, but picking up some other stoneinstead…. There will be nothing that requires explanation if Ichoose to do what I think to be, say, the worst possible thing to doand leave undone what I think the best thing to do. (Hare 1963, pp.68–9)

But our reactions to (1), (2), and (3) show that wedo thinkthere is something peculiar about action contrary to one’sbetter judgment which renders such action hard to understand, orperhaps even impossible. An extreme externalist view thus seems tomischaracterize the status of akratic actions.

Perhaps even more importantly, however, extreme externalism hasdramatic implications for our understanding of intentional action ingeneral—not just weak-willed action. For such a view impliesthat

deliberation about what it would be best to do has no closer relationto practical reasoning than, say, deliberation about what it would bechic to do. If one happens to care about what it would be chic to do,then a consideration of this matter may play an important role inone’s practical reasoning. If one does not care, it will beirrelevant. The case is the same with reasoning about what it would bebest to do. (Bratman 1979, p. 158)

To adopt a general doctrine of this sort seems an awfully precipitousresponse to the possibility ofakrasia (that is, weak-willedaction). For it seems extremely plausible to assign to our overallevaluations of our options an important role in our choices. Man is arational animal, the saying goes; that is—to offer one gloss onthis idea—we act on reasons, and in light of our assessment ofthe overall balance of reasons. When we engage in deliberation orreasoning about what to do, we often proceed by thinking about thereasons which favor our various options, and then bringing thesetogether into an overall assessment which is, precisely, intended toguide our choice.

Or, as Bratman puts it, we very often reason about what it isbest to do as a way of settling the question of whatto do. (He calls this “evaluative practicalreasoning”: Bratman 1979, p. 156.) “One’sevaluations [thus] play a crucial role in the reasoning underlyingfull-blown action,” Bratman holds (p. 170), and to be forced todeny this would be in his view “too high a price to pay”(p. 159). As Alfred Mele similarly puts it, “there is a realdanger that in attempting to make causal and conceptual space forfull-fledged akratic action one might commit oneself to the rejectionof genuine ties between evaluative judgment and action” (1991,p. 34). But that would be to throw the baby out with the bathwater. Ifwe want to resist Hare’s conclusions, we must do so in a waywhich steers clear of the danger to which Mele alerts us. We mustnavigate between the Scylla of extreme internalism and the Charybdisof extreme externalism.

2. Davidson on the Possibility of Weakness of Will

This is just what Donald Davidson set out to do in a rich, elegant,and incisive paper published in 1970 which has had a toweringinfluence on the subsequent literature. Davidson’s treatmentaims to vindicate the possibility of weakness of will; to offer anovel analysis of its nature; to clarify its status as a marginal,somehow defective instance of agency which we rightly find dubiouslyintelligible; and to do all this within the contours of a general viewof practical reasoning and intentional action which assigns a centraland special role to our evaluative judgments. Let us see how heproposes to do these things.

First, Davidson offers the following general characterization ofweak-willed or incontinent action:^[6]

In doingb an agent acts incontinently if and only if: (a)the agent doesb intentionally; (b) the agent believes thereis an alternative actiona open to him; and (c) the agentjudges that, all things considered, it would be better to doa than to dob.

We initially described weak-willed action as free, intentional actioncontrary to the agent’s better judgment; it may be useful to seehow Davidson’s more precise definition matches up with thatinitial characterization. Davidson’s condition (a) requires thatthe action in question be intentional.^[7] Condition (b) seems intended to ensure that the action in question isafree action.^[8] Condition (c) of Davidson’s definition represents what we havecalled the agent’s “better judgment,” that is, theoverall evaluation of his options contrary to which the incontinentagent acts.

Davidson notes that “there is no proving such actions exist; butit seems to me absolutely certain that they do” (p. 29). Why,then, is there a persistent tendency, both in philosophy and inordinary thought, to deny that such actions are possible?Davidson’s diagnosis is that two plausible principles which“derive their force from a very persuasive view of the nature ofintentional action and practical reasoning” (p. 31) appear toentail that incontinence is impossible. He articulates those twoprinciples as follows (p. 23):

P1. If an agent wants to doa more than hewants to dob and he believes himself free to do eithera orb, then he will intentionally doa ifhe does eithera orb intentionally.

P2. If an agent judges that it would be better to doa than to dob, then he wants to doa morethan he wants to dob.

P2, Davidson observes, “connects judgements of what it is betterto do with motivation or wanting” (p. 23); he adds later that it“states a mild form of internalism” (p. 26). Davidson isproposing,contra the extreme externalist position, that ourevaluative judgments about the merits of the options we deem open tous are not motivationally inert. While he admits that one couldquibble or tinker with the formulation of P1 and P2 (pp. 23–4;p. 27; p. 31), he is confident that they or something like them giveexpression to a powerfully attractive picture of practical reasoningand intentional action, one which assigns an important motivationalrole to the agent’s evaluative judgments.^[9]

The difficulty is, though, that P1 and P2—howeverattractive—together imply that an agent never intentionally doesb when he judges that it would be better to doa (ifhe takes himself to be free to do either). And this certainly lookslike a denial of the possibility of incontinent action. No wonder,then, that so many have been tempted to say that akratic action isimpossible! Looking carefully, however, we can see that P1 and P2 donot imply the impossibility of incontinent actionsasDavidson has defined them. For Davidson characterizes the agentwho incontinently doesb as holding, not that it would bebetter to doa than to dob, but that it would bebetter,all things considered, to doa than to dob. Is the “all things considered” just arhetorical flourish? Or does it mark a genuine difference between thetwo judgments? If these are two different judgments, and one can holdthe latter without holding the former, then incontinent action ispossibleeven if P1 and P2 are true.

In the rest of his paper Davidson sets out to vindicate that verypossibility. The phrase “all things considered” is not, asit might seem, merely a minor difference in wording that allowsweakness of will to get off on a technicality. Rather, that phrasemarks an important contrast in logical form to which we would need toattend in any case in order properly to understand the structure ofpractical reasoning. For that phrase indicates a judgment that isconditional orrelational rather thanall-out orunconditional in form; and thatdifference is crucial.^[10] We can better see the relational character of anall-things-considered judgment if we first look at evaluativejudgments that play an important role in an earlier phase of practicalreasoning, the phase where we consider what reasons or considerationsfavor doinga and what reasons or considerations favor doingb. (For simplicity, imagine a case in which an agent ischoosing only between two options known to be incompatible,aandb.) Theseprima facie judgments, as Davidsonterms them, take the form:

PF: In light ofr,a isprimafacie better thanb.

In this schemar refers to a consideration taken to berelevant, such as the following: thata would be relaxing,whileb would be stressful. A PF judgment of this kindidentifies one respect in whicha is superior tob,one perspective from whicha comes out on top.

We should pause to note three things about PF judgments. (i) A PFjudgment is not a conclusion in favor of the overall superiority ofa (orb). Such “all-out” evaluativejudgments have a simpler logical form, namely:

AO:a is better thanb.

(ii) Indeed, no AO conclusionfollows logically from any PFjudgment. (iii) More strongly still: PF judgments do not in generalprovide sufficientgrounds for the corresponding AOconclusion. Even if an agent makes one PF judgment which favorsa overb, as in the case we imagined, she mayalso makeother PF judgments which favorbovera (say, one in which the following consideration playsthe role ofr: thatb would be lucrative, whilea would be expensive). We do not want to say in that casethat she has sufficient grounds to draw each of two incompatibleconclusions (thata is better thanb, and thatb is better thana; these are incompatible providedthe better-than relation is asymmetric, assumed here).

We have contrasted PF judgments with AO or “all-out”evaluative judgments. PF judgments are relational in character: theypoint out arelation which holds between the considerationr and doinga (orb). (We could call thatrelation the “favoring” relation.) That relation is notsuch as to permit us to “detach” (as Davidson puts it, p.37) an unconditional evaluative conclusion in favor of doinga (orb) from PF and the supposition thatrobtains. That is, we are not to understand PF judgments as having theform of a material conditional.

Davidson’s innovative suggestion is that judgments with this PFlogical form are an appropriate way to model what happens in the earlystages of practical reasoning, where we rehearse reasons for andagainst the options we are considering. And his stressing that no suchPF judgment commits the agent to an overall evaluative conclusion infavor ofa orb is useful in thinking about a caselike Julie’s ((1) above). We described Julie as knowing (andtherefore believing) thatb was more expensive thana, but opting forb nonetheless. We can imagine,then, that among the ingredients of Julie’s practical reasoningwas a PF judgment like this:

In light of the fact thatb is more expensive thana,a isprima facie better thanb.

But this PF judgment alone, as we have seen, does not commit her tothe overall judgment thata is better thanb. Forshe may also have made other PF judgments, such as

In light of the fact thatb would be much moregastronomically exciting thana,b isprimafacie better thana.

We would not want to say that Julie now has sufficient grounds both toconclude thata is better thanb and to concludethatb is better thana. Her premises all seemconsistent; they do not provide her with sufficient grounds to embracea contradiction. Thus her various PF judgments, when consideredseparately, mustnot each commit her to a correspondingoverall conclusion in favor ofa orb.

Practical reasoning, Davidson suggests, starts from judgments likethese, each identifying one respect in which one of the options issuperior. But in order to make progress in our practical reasoning weshall eventually need to consider howa compares tob not just with respect toone consideration, but inlight of several considerations taken together. That is, Julie willeventually need to consider how to fill in the blanks in a PF judgmentlike this:

In light of the fact thatb is more expensive thanaand the fact thatb would be much moregastronomically exciting thana, … isprimafacie better than ….

This PF judgment is morecomprehensive than the ones weattributed to Julie a moment ago, as it takes into account a broaderrange of considerations. (I take the term “comprehensive”from Lazar 1999.) Now in Julie’s case we can surmise how shefilled in those blanks: with “b isprima faciebetter thana.” Julie’s filling in the blanks inthat way can naturally be taken as expressing the view that the muchgreater gastronomic excitement promised byboutweighs oroverridesb’sinferiority toa from a strictly financial standpoint.

We thus need to generalize our schema for PF judgments so thatcomparative assessments ofa andb can berelativized, not just to a single consideration, but to multipleconsiderations taken together or as a body:

PF_n: In light of⟨r₁, …,r_n⟩,a isprima facie better thanb.

Notice that PF_n judgments are still relational inform: they assert that a relation (the “favoring”relation) holds between theset of considerations⟨r₁, …,r_n⟩anda (orb). Indeed, the relational character of aPF_n judgment remains even if we make it ascomprehensive as we can: if we expand the set⟨r₁, …,r_n⟩to incorporateall the considerations the agent deemsrelevant to her decision. Following Davidson (p. 38), let us give thelabele to that set. So even the following judgment:

ATC: In light ofe,a isprimafacie better thanb.

is a relational or conditional judgment and not an all-out conclusionin favor of doinga. To make a judgment of the above formdoesnot commit one to drawing an overall conclusion in favorof doinga.

We may be better able to see this by considering an analogy fromtheoretical reason. Suppose Hercule Poirot has been called in toinvestigate a murder. We can imagine him assessing bits of evidence ashe encounters them:

In light of the fact that the murder weapon belongs to ColonelMustard, Mustard looks guilty;

In light of his having an alibi for the time of the murder, Mustardlooks not guilty;

and so on. These are theoretical analogues of the PF judgmentsrelativized to single considerations which we looked at earlier.However, Poirot will eventually need to consider how these variousbits of evidence add up; that is, he will eventually need to fill inthe blanks in a more comprehensive PF_n judgmentlike this:

In light of ⟨e₁, …,e_n⟩, … looks to be the guilty party,

where ⟨e₁, …,e_n⟩ is a set of bits of pertinent evidence.Notice, though, that no such PF_n judgment actuallyconstitutes settling on a particular person as the culprit. Suppose wetake a maximally large ⟨e₁, …,e_n⟩ consisting ofall the evidencePoirot has seen, and imagine him thinking

The body of evidence I have seen points on balance toward ColonelMustard as the guilty party.

Still, to make this observation is manifestlynot to concludethat Mustard is guilty. Poirot may feel he needs more evidence beforehe can dothat.

In the same way, an ATC or all-things-considered judgment, althoughcomprehensive, is still relational in nature, and therefore distinctfrom an AO judgment in favor ofa. It is at leastpossible to make an ATC judgment in favor ofawithout making the corresponding AO judgment in favor ofa.(That is roughly the position Poirot is in.) This possibility is thekey to Davidson’s solution to the problem of how weakness ofwill is possible. For ATC is, precisely,the agent’s betterjudgment as Davidson construes it in his definition ofincontinent action. P1 and P2 together imply that an agent who reachesanAO conclusion in favor ofa will notintentionally dob. But the incontinent agent never reachessuch an AO conclusion. With respect toa, he remains stuck atthe Hercule Poirot stage: he sees that the considerations he hasrehearsed, taken as a body, favora, but he is not willing tomake a commitment toa as the thing to do.^[11] He makes only a relational ATC judgment in favor ofa,contrary to which he then acts.

What should we say about an agent who does this? Let’s return tothe three features ofprima facie or PF judgments which wenoted earlier. (i) was that a PF judgment is not a conclusion in favorof the overall superiority ofa (orb). (ii) wasthat no AO conclusionfollows logically from any PF judgment.These two features hold even of the special subclass of PF judgmentswhich are ATC judgments. Such judgments neither are equivalent to, norlogically imply, any AO judgment. This establishes that theincontinent agent who fails to draw the AO conclusion whichcorresponds to his ATC conclusion, and to perform the correspondingaction, is not committing “a simple logical blunder” (p.40). Notably, he does not contradict himself. He does, however,exhibit a defect in rationality, on Davidson’s account. Recallfeature (iii) of PF judgments: that they do not in general providesufficient grounds for the corresponding AO conclusion. But this doesnot hold of the special subclass of such judgments which areATC judgments. Drawing anATC conclusion in favor ofadoes give one sufficient grounds to conclude thata is bettersans phrase and, indeed, to doa. For Davidson proposes that the transition from an ATCjudgment in favor ofa to the corresponding AO judgment, andto doinga, is enjoined by a substantive principle ofrationality which he dubs “the principle of continence.”That principle tells us to “perform the action judged best onthe basis of all available relevant reasons” (p. 41); and theincontinent agent violates this injunction. [NEW NOTE:Davidson’s principle is stronger than it needs to be. To securethe result that the incontinent agent violates the principle ofcontinence, it is not necessary that the principle instruct ustoperform the action which satisfies that description. It issufficient if it merely enjoins usnot to perform somealternative action.] The principle of continence thussubstantiates the idea that “what is wrong is that theincontinent man acts, and judges, irrationally, for this is surelywhat we must say of a man who goes against his own bestjudgement” (p. 41). He acts irrationally in virtue of violatingthis substantive principle, obedience to which is a necessarycondition for rationality.

We must put this point about the irrationality of incontinence withsome care, however. For recall that an incontinent action must itselfbe intentional, that is, done for a reason. The weak-willed agent,then, must have a reason for doingb; indeed he doesb for that reason. What he lacks—and lacks by his ownlights—is asufficient reason to dob, givenall the considerations that he takes to favora. As Davidsonputs it, if we ask “what is the agent’s reason for doing[b] when he believes it would be better, all thingsconsidered, to do another thing, then the answer must be: for this,the agent has no reason” (p. 42). What Davidson seems to have inmind is that while the agent does have a reason for doingb(p. 42, n. 25), he lacks a reasonto do brather thana. By his own lights, he has noadequate reason fordoingb; thus he cannot make sense of his own action.“He recognizes, in his own intentional behaviour, somethingessentially surd” (p. 42). So akratic action, whilepossible on Davidson’s account, is nonethelessnecessarilyirrational. That is the sense in which it is adefective and not fully intelligible exercise of agency, despite beinga very real phenomenon.

3. The Debate After Davidson

3.1 Internalist and Externalist Strands

Davidson has certainly presented an arresting theory of practicalreasoning. But has he shown how weakness of the will is possible? Mostphilosophers writing after him, while acknowledging his pathbreakingwork on the issue, think he has not. One principal difficulty whichsubsequent theorists have seized on is that Davidson’s view canaccount for the possibility of action contrary to one’s betterjudgmentonly if one’s better judgment is construedmerely as a conditional orprima facie judgment.Davidson’s P1 and P2 do genuinely rule out the possibility offree intentional action contrary to an all-out or unconditionalevaluative judgment. But it seems that such cases exist. MichaelBratman, for instance, introduces us to Sam, who, in a depressedstate, is deep into a bottle of wine, despite his acknowledged needfor an early wake-up and a clear head tomorrow (1979, p. 156).Sam’s friend, stopping by, says:

Look here. Your reasons for abstaining seem clearly stronger than yourreasons for drinking. So how can you have thought that it would bebest to drink?

To which Sam replies:

I don’t think it would be best to drink. Do you think I’mstupid enough to think that, given how strong my reasons forabstaining are? I think it would be best to abstain. Still, I’mdrinking.

Sam’s case certainly seems possible as described.Davidson’s view, though, must reject it as impossible.^[12] Given his conduct, Sam can’t think it best to abstain; at most,he thinks it all-things-considered best to abstain, a very differentkettle of fish. But this seems false of Sam: there is no evidence thathe has remained stuck at the Hercule Poirot stage with respect to thesuperiority of abstaining. He seems to have gone all the way to ajudgmentsans phrase that abstaining would be better; and yethe drinks.

Ironically, this complaint makes Davidson out to be a bit like Hare.Like Hare, Davidson subscribes to an internalist principle (P2) whichconnects evaluative judgments with motivation and hence with action.(Indeed, in light of the difficulty raised here, one might wonder ifDavidson is entitled to consider P2 a “mild” form ofinternalism, as he claims (p. 26).) As with Hare, this internalistcommitment rules out as impossible certain kinds of action contrary toone’s evaluative judgment. Now Davidson, like Hare, does acceptthe possibility of certain phenomena in this neighborhood;but—as with Hare—critics think the cases permitted by hisanalysis fail to exhaust the range of actual cases of weakness ofwill. The phenomenon seems to run one step ahead of our attempts tomake room for it.

Those writing after Davidson have tended to focus, then, on thequestion of the possibility and rational status of action contrary toone’sunconditional better judgment.^[13] Naturally, different theorists have plotted different courses throughthese shoals. Some tack more to the internalist side, wishing topreserve a strong internal connection between evaluation and actioneven at the risk of denying or seeming to deny the possibility ofakratic action (or at least some understandings of it). Examples ofsome post-Davidson treatments which share a broadly internalistemphasis, even if they feature different flavors of internalism, arethose of Bratman (1979), Buss (1997), Tenenbaum (1999; see also 2007,ch. 7), and Stroud (2003). The main danger for such approaches is thatin seeking to preserve and defend a certain picture of the primordialrole of evaluative thought in rational action—a picture criticsare likely to dismiss as too rationalistic—such theorists may beled to reject common phenomena which ought properly to haveconstrained their more abstract theories. (See the opening of Wiggins1979 for a forceful articulation of this criticism.)

Other theorists, by contrast, are more drawn toward the externalistshoreline. They emphasize the motivational importance of factors otherthan the agent’s evaluative judgment and the divergences thatcan result between an agent’s evaluation of her options and hermotivation to act. They are thus disinclined to posit any strong,necessary link between evaluative judgment and action. MichaelStocker, for instance, argues that the philosophical tradition hasbeen led astray in assuming that evaluation dictates motivation.“Motivation and evaluation do not stand in a simple and directrelation to each other, as so often supposed,” he writes.Rather, “their interrelations are mediated by large arrays ofcomplex psychic structures, such as mood, energy, and interest”(1979, pp. 738–9). Similarly, Alfred Mele proposes as afundamental and general truth—and one that underlies thepossibility ofakrasia—that “the motivationalforce of a want may be out of line with the agent’s evaluationof the object of that want” (1987, p. 37). Mele goes on to offerseveral different reasons why the two can come apart: for example,rewards perceived asproximate can exert a motivationalinfluence disproportionate to the value the agent reflectivelyattaches to them (1987, ch. 6). Such wants may function as strongcauses even if the agent takes them to constitute weakreasons.

With respect to these questions, the challenge sketched at the end ofSection 1 above remains in full force. What is required is a viewwhich successfully navigates between the Scylla of an extremeinternalism about evaluative judgment which would preclude thepossibility of weakness of will, and the Charybdis of an extremeexternalism which would deny any privileged role to evaluativejudgment in practical reasoning or rational action. For one’sverdict aboutakrasia will in general be closely connected toone’s more general views of action, practical reasoning,rationality, and evaluative judgment—as was certainly true ofDavidson.

Views that downplay the role of evaluative judgment in action andhence tack more toward the externalist side of the channel may moreeasily be able to accept the possibility and indeed the actuality ofweakness of will. But they are subject to their own challenges. Forexample, suppose we follow Mele’s image ofakrasia andposit that a certain agent is caused to dox by motivation todox which is dramatically out of kilter with her assessmentof the merits of doingx. This agent is horrified by the ideaof doingx, say, but finds herself doing it anyway. We mightwonder whether in that case her doingx is genuinely free,intentional, and uncompelled. Such an agent might seem instead to beat the mercy of a motivational force which is, from her point of view,utterly alien. Thus, worries about distinguishingakrasiafrom compulsion come back in full force in connection with proposalslike these. (Seefn. 7 above for relevant references; Buss and Tenenbaum press these worriesagainst accounts like Mele’s in particular.) Moreover, there isthe danger, for accounts of this more externalist stripe, of takingtoo much of the mystery out of weakness of will. Even if akraticaction is possible and indeed actual, it remains a puzzling, marginal,somehow defective instance of agency, one that we rightly find notfully intelligible. Views that do not assign a privileged place inrational deliberation and action to the agent’s overallassessment of her options risk making akratic action seem no moreproblematic than Julie’s or Jimmy’s decisions, orHare’s agent who fails to pick up the roundest stone in thevicinity.

3.2 Weakness of Will as Potentially Rational

The “externalist turn” toward downplaying the role of anagent’s better judgment and emphasizing other psychic factorsinstead is connected to a second way in which some theorists writingafter Davidson have dissented from his analysis. Davidson, as we saw,viewed akratic action as possible, but irrational. The weak-willedagent acts contrary to what she herself takes to be the balance ofreasons; her choice is thus unreasonable by her own lights. On thispicture, incontinent action is a paradigm case of practicalirrationality. Many other theorists have agreed with Davidson on thisscore and have takenakrasia to be perhaps the clearestexample of practical irrationality. But some writers (notably Audi1990, McIntyre 1990, and Arpaly 2000) have questioned whether akraticactionis necessarily irrational. Perhaps we ought to leaveroom, not just for thepossibility of akratic action, but forthe potentialrationality of akratic action.

The irrationality which is held necessarily to attach to akraticaction derives from the discrepancy between what the agent judges tobe the best (or better) thing to do, and what she does. That is, heraction is faulted as irrational in virtue of not conforming to herbetter judgment. But—ask these critics—what if her betterjudgment is itself faulty? There is nothing magical about anagent’s better judgment that ensures that it is correct, or evenwarranted; like any other judgment, it can be in error, or evenunjustified. (Recall that by “better judgment” we meant,all along, only “a judgment as to which course of action isbetter,” not “asuperior judgment.”)Suppose the agent’s better judgment is itself defective. Then“even though the akratic agent does not believe that she isdoing what she has most reason to do, it may nevertheless be the casethat the course of action that she is pursuing is the one that she has… most reason to pursue” (McIntyre 1990, p. 385). In thatsense the akratic agent may be wiser than her own better judgment.

How, concretely, could an agent’s better judgment go astray inthis way? Perhaps her survey of what she took to be the relevantconsiderations did not include, or did not attach sufficient weightto, what were in fact significant reasons in favor of one of thepossible courses of action. She may have overlooked these, or(wrongly) deemed them not to be reasons, or failed to appreciate theirfull force; and in that case her judgment of what it is best to dowill be incorrect. Consider, for example, Jonathan Bennett’sHuckleberry Finn (Bennett 1974, discussed in McIntyre 1990), whoakratically fails to turn in his slave friend Jim to the authorities.Huck’s judgment that he ought to do so, however, was basedprimarily on what he took to be the force of Miss Watson’sproperty rights; it ignored his powerful feelings of friendship andaffection for Jim, as well as other highly relevant factors. His“better judgment” was thusnot in fact a verycomprehensive judgment; it did not actually take into account the fullrange of relevant considerations.

Or consider Emily, who has always thought it best that she pursue aPh.D. in chemistry (Arpaly 2000, p. 504). When she revisits the issue,as she does periodically, she discounts her increasing feelings ofrestlessness, sadness, and lack of motivation as she proceeds in theprogram, and concludes that she ought to persevere. But in fact shehas very good reasons to quit the program—her talents are notwell suited to a career in chemistry, and the people who are thrivingin the program are very different from her. If she impulsively,akratically quits the program, purely on the basis of her feelings,Emily is in fact doing just what she ought to do.^[14] That her action conflicts with her better judgment does notsignificantly impugn its rationality, given all the considerationsthatdo support her quitting the program. “A theory ofrationality should not assume that there is something special about anagent’s best judgment. An agent’s best judgment is justanother belief” (Arpaly 2000, p. 512). Emily’s actionconflicts, then, with one belief she has; but it coheres with manymore of her beliefs and desires overall. Although she may find her ownaction inexplicable or “surd,” she is in fact actingrationally, even though she does not know it.ContraDavidson, “we can … act rationally just when we cannotmake any sense of our actions” (Arpaly 2000, p. 513).

It is unclear, however, whether these arguments and examples arelikely to sway those who takeakrasia to be a paradigm ofpractical irrationality. These dissenters stress thesubstantivemerits of the course of action the akratic agent follows. Buttraditionalists may say that is beside the point: however well thingsturn out, the practical thinking of the akratic agent still exhibits aprocedural defect. Someone who flouts her own conclusionabout where the balance of reasons lies hasipso facto goneastray in her practical reasoning. Even if the action she performs isin fact supported by the balance of reasons, she does not think it is,and that is enough to show her practical reasoning to be faulty. Thedefenders of the traditional conception ofakrasia asirrational thus wish to grant special rational authority (in thisprocedural sense) to the agent’s better judgment, even if theyadmit that such a judgment can be substantively incorrect. Bycontrast, the dissenters “[do] not believe best judgments haveany privileged role” (Arpaly 2000, p. 513). We see again thecontrast between “internalist” and“externalist” tendencies in the debates over weakness ofwill.

3.3 Changing the Subject

Further revisionist strands take the agent’s better judgmenteven farther out of the picture. In an outstandingly lucid andstimulating essay published in 1999 (see also his 2009, ch. 4),Richard Holton argued that weakness of will is not action contrary toone’s better judgment at all. The literature has gone astray inunderstanding weakness of will in this way; weakness of will isactually quite a different phenomenon, in which the agent’sbetter judgment plays no role.^[15] For Holton, when ordinary people speak of weakness of will they havein mind a certain kind of failure to act on one’sintentions. What matters for weakness of will, then, is notwhether you deem another course of action superior at the time ofaction. It is whether you are abandoning an intention you previouslyformed. Weakness of will as the untutored understand it is notakrasia (if we reserve that term for action contrary toone’s better judgment), but rather a certain kind of failure tostick to one’s plans.^[16] This understanding of weakness of will changes the subject in twoways. First, the state of the agent with which the weak-willed actionis in conflict is not an evaluative judgment (as inakrasia)but a different kind of state, namely an intention. Second, it is notessential that there besynchronic conflict, asakrasia demands. You must act contrary to yourpresent better judgment in order to exhibitakrasia;conflict with aprevious better judgment does not indicateakrasia, but merely a change of mind. However, you canexhibit weakness of will as Holton understands it simply by abandoninga previously formed intention.

Of course not all cases of abandoning or failing to act on apreviously formed intention count as weakness of will. I intend to runfive miles tomorrow evening. If I break my leg tomorrow morning andfail to run five miles tomorrow evening, I will not have exhibitedweakness of will. How can we characterizewhich failures toact on a previously formed intention count as weakness of will?Holton’s answer has two parts. First, he says, there is anirreducible normative dimension to the question whethersomeone’s abandoning of an intention constituted weakness ofwill (Holton 1999, p. 259).^[17] That is, there is no purely descriptive criterion (such as whetherher action was contrary to her better judgment) which is sufficientfor weakness of will; in order to decide whether a given case was aninstance of weakness of will we must consider normative questions,such as whether it wasreasonable for the agent to haveabandoned or revised that intention, or whether sheshouldhave done so. In the case of my broken leg, for instance, it wasclearly reasonable for me to abandon my intention; that is why I couldnot be charged with weakness of will in that case.

Second, says Holton, we need to attend to an important subclass of ourintentions to do something at a future time, namelycontrary-inclination-defeatingintentions, or, as helater terms them (Holton 2003),resolutions. Resolutions areintentions that are formed precisely in order to insulate one againstcontrary inclinations which one expects to feel when the time comes.Thus one reason I might form an intention on Monday to run five mileson Tuesday—as opposed to leaving the issue open until Tuesday,for decision then—is to reduce the effect of feelings oflassitude to which I fear I may be subject when Tuesday rolls around.Then suppose Tuesday rolls around; I am indeed prey to feelings oflassitude; and I decide as a result not to run.Now I can becharged with weakness of will. Weakness of will involves,specifically, a failure to act on aresolution; this issufficient to differentiate weakness of will from mere change of mindand even from caprice (which is adifferent species ofunreasonable intention revision, according to Holton).

As a later paper by Alison McIntyre shows (McIntyre 2006),understanding weakness of will in this way casts a fresh light on theissue of its rational status. The weak-willed agent abandons aresolution because of a contrary inclination of exactly the type whichthe resolution was expressly designed to defeat. Thus, as McIntyreunderlines, weak-willed action always involves a procedural rational defect:^[18] a technique of self-management has been deployed but has failed(McIntyre 2006, p. 296). This fact alone gives us grounds to criticizethe action in virtue of thesecond feature which (for Holton)distinguishes weakness of will from mere change of mind. In cases likethese we needn’t resolve the potentially murky issue of whetherthe agent wasreasonable in abandoning her intention.

McIntyre maintains that it would be overstating the case to say thatbecause weakness of will involves this procedural defect, it is alwaysirrational (McIntyre 2006, p. 290; pp. 298–9; p. 302). Sheproposes rather that practical rationality has multiple facets andaims, and that failure in one respect or along one dimension does notautomatically justify the especially severe form of rational criticismwhich we intend by the term “irrational.” For example,consider an agent who succumbs to contrary inclination of exactly thetype expected when the time comes to act on a trulystupidresolution. (Holton gives the example of resolving to go without waterfor two days just to see what it feels like: Holton 2003, p. 42.)There will indeed be a blemish on this agent’s rationalscorecard if he eventually gives in and drinks: he will have failed inhis attempt at self-management. But wouldn’t it be rationallyfarworse for him to stick to his silly resolution no matterwhat the cost?

We can also re-examine the issue of the rationality ofakrasia in light of this analysis of weakness of will; for wecan distinguish between akratic and non-akratic cases of the latter.As McIntyre points out, resolutions typically rest on judgments aboutwhat it is best that one do at a (future) timet. If an agentfails to act on a previously formed resolution to doa att, thus exhibiting weakness of will, we can distinguish thecase in which he still endorses att the judgment that it isbest that he doa att (even though he does not doit) from the case in which he abandons that judgment along with hisresolution. In the latter, non-akratic case, the agent in effectrationalizes his failure to live up to his resolution by deciding thatit is not after all best that he doa att. McIntyrepoints out that the traditional view thatakrasia is alwaysirrational seems to give us a perverse incentive to rationalize, sincein that case we escape the grave charge of practical irrationality,being left only with the procedural practical defect present in allcases of weakness of will (McIntyre 2006, p. 291). But this seemsimplausible: are the two sub-cases really so radically different intheir rational status? Indeed, McIntyre argues, if anything, akraticweakness of will is typically rationallypreferable torationalizing weakness of will (McIntyre 2006, p. 287; pp. 309ff.).“In the presence of powerful contrary inclinations that bringabout a failure to be resolute,” she writes, “resistingrationalization and remaining clearheaded about one’s reasons toact can constitute a modest accomplishment” (McIntyre 2006, p.311). Have we witnessed the transformation ofakrasia fromimpossible, to irrational, to downright admirable?

Tamar Schapiro (Schapiro 2021, 2022) departs in a very different wayfrom the tradition of understanding weakness of will as actioncontrary to the agent’s better judgment. Schapiro thinks itmisdiagnoses and understates the defect in weak-willed action to seeit as the product of a “mistake within practicaldeliberation” (Schapiro 2021, p. 152). The problem is morefundamental: weak-willed action is better characterized as“abandon[ing] your post as deliberator” (Schapiro 2021, p.152) than as making an error in deliberation. Weak-willed action infact does not actually engage thewill at all. OnSchapiro’s dualistic Kantian moral psychology,inclination is a kind of “feeling like it” whichaffects what she calls your “inner animal,” an instinctivesubagential part of you which perceives the world as teleologicallyladen. An inclination creates motivational pressure and engages yourinstinctive agential capacities even before you, the deciding agent,have weighed in. As an agent with a free will, it falls to you totreat that inclination as raw material for the construction of endsand principles which you can take responsibility for. That is theproper work of the will. However, inclination always places your will“at a crossroads” (Schapiro quotes Kant’sGroundwork), because there is another alternative. You mightinstead abdicate your responsibility todecide what to do and“flee the burden of your freedom.” To “take the lowroad” in this way is not to act on a faulty or inferiorprinciple; it is not to act on a principle at all, indeed arguably notto act at all. When we take the low road, we flee into our animality,seeking refuge, as it were, in our inner animal and allowing ourselvesto actas if we were nothing more than our inner animal. Thisalways constitutes bad faith, however, because we can never actuallybe only our inner animal. The burden of our freedom isinescapable.

3.4 Broadening the Discussion

Recent work on weakness of will has expanded the scope of discussionin several respects, three of which we discuss below. Anotherimportant development has been the application of weakness of will togroups: see Pettit 2003, List and Pettit 2011, and Flowerman2024.

3.4.1 EpistemicAkrasia

It is natural to ask whether there could be anepistemicversion of the phenomenon ofakrasia. In an influentialdiscussion, Sophie Horowitz suggests that practicalakrasiadoes indeed have an epistemic counterpart: “Just as an akraticagent acts in a way she believes she ought not act, an epistemicallyakratic agent believes something that she believes is unsupported byher evidence” (Horowitz 2014, p. 718). That is, the latter holdsa belief of the form “p, but my evidence doesn’tsupportp.”

Is epistemicakrasia really possible, though? Somephilosophers who are happy to countenance practicalakrasiahave answered in the negative: see for instance Hurley 1989, Owens2002, and Adler 2002. David Owens (2002) argues that we lack the sortof doxastic control that would be required in order for our beliefs tobe formed freely and deliberately. But epistemicakrasiawould require control of just that kind, he suggests. Jonathan Adler(2002) makes a different argument for the impossibility of epistemicakrasia. Adler notes what he takes to be an importantdisanalogy between epistemic and practical reasoning. Conflict betweentwobeliefs that can’t both betrue mustweaken one or both beliefs; but conflict between twodesiresthat can’t both besatisfied need not weaken eitherdesire. So it is perfectly possible that a practically akratic agentwho has formed an all-things-considered judgment about what to do maystill be in the grip of a desire to do something else. However, Adlerthinks this can’t happen in the doxastic realm: one can’tremain in the grip of abelief if one views the evidence foran opposing belief as decisive. (See Levy 2004 and Ribeiro 2011 forcounterarguments to Adler.)

If there could indeed be an epistemic version ofakrasia, itis worth considering how closely it would parallel practicalakrasia, including the purported irrationality of the latter.Daniel Greco (2014) underlines that practical and epistemic varietiesofakrasia have in common that they involve a kind offragmentation or inner conflict. This supports the idea that bothepistemic and practicalakrasia are irrational. (See alsoFeldman 2005.) But just as the rational status of practicalakrasia has become contested (see section 3.2 above), sometheorists (for example Coates (2012), Christensen (2024),Lasonen-Aarnio (2014), Sinhababu (2020), Weatherson (2019), andWedgwood (2011)) have argued that epistemicakrasia couldpotentially be rational as well. This discussion has been shaped by agrowing interest in higher-order evidence — that is, evidencethat bears on what one’s first-order evidence actually supports.Defenders of the rationality of epistemicakrasia have arguedthat beliefs of the form “p, but my evidencedoesn’t supportp” can be rationally permissiblewhen one hasmisleading higher-order evidence. In such cases,they contend, a person could have good grounds both for believing thatp and for believing that her evidence doesn’t supportthe conclusion thatp. Horowitz (2014, p. 740), on the otherhand, maintains that beliefs of this form are rationally unstable.

Theorists have also appealed to epistemicakrasia to advanceor defend other positions in epistemology. Jamie Fritz (2021) uses it,for instance, in an argument for epistemic impurism: the view that“[s]ome paradigmatically epistemic properties of a belief thatp (like whetherp is epistemically rational, orwhether it is knowledge) depend on factors that are not relevant tothe truth ofp” (Fritz 2021, p. 98). Fritz thinksepistemicpurists are committed to the possibility ofrational epistemicakrasia, which he takes to be areductio. Others have drawn a variety of conclusions aboutcredences—degrees of belief—by appealing toepistemicakrasia: see Jackson and Tan 2021, Lin 2021, andFraser 2022.

3.4.2 Interdisciplinary Approaches to Understanding the Will and its Weakness

Philosophers have increasingly engaged with empirical literature aboutthe nature of willpower and how this notion is understood in ordinarylife. Part of this literature is concerned to distinguish willpowerfrom potentially related notions such as self-control,self-regulation, and resisting temptation. For example, studies byIrving et al. (2022) and Bermúdez et al. (2023) suggest thatthe folk psychological understanding of self-control centerssynchronic strategies for resisting temptation. However, empiricalevidence suggests that diachronic regulation and external scaffoldingmay be more effective at helping agents manage motivational conflictover time (see Duckworth et al. 2016, Soravia et al 2015, and Wagneret al. 2004). Indeed, it is possible that those who most effectivelyresist temptation are those who have cultivated habits so thattemptation rarely arises for them in first place. Haug (2021, pp.303–306) uses empirical evidence to suggest that individuals whoare regarded as highly self-controlled experience a kind ofmotivational harmony, the result of indirect attentional strategieswhich inhibit conflicting desires from rising to the level at whichthey must be resisted. It thus lies between what Aristotle would havedescribed as continence and temperance. (See Orlandi and Stroud 2021for a philosophical treatment.) There has also been discussion of howphilosophical understandings ofakrasia might be related todual-process models of reasoning in psychology (see, e.g., Altehenger2021a).

On the question of whether exercises of willpower or self-control mustbe effortful, the older “ego depletion” model of willpowerhas increasingly come under question in recent years. (See forinstance Dang and Hagger 2019.) Mylopoulos and Pacherie (2020),Bermúdez et al. (2024), and Bermúdez (2021) insteadcharacterize self-control as a kind of skill or a set of skills, theexercise of which need not be effortful.

George Ainslie has also addressed the question of whether willpowertakes effort. Ainslie has argued in an influential body of work (seeAinslie 1992, 2001) that a pervasive tendency toward “hyperbolicdelay discounting” leads us to prefer smaller, sooner rewardsover larger, later ones. We can counter this tendency by engaging in“intertemporal bargaining,” in which we “recursivelyself-predict” what we will do in the future in light of ourchoice today and create personal rules about which choices will countas according with our values and which will constitute lapses. Ainslieproposes models for this process that draw on behavioral economics,social psychology, and brain imaging. In a recent target article(Ainslie 2021), Ainslie argues that willpower should be identifiedwith a form ofresolve understood along the above lines; hedistinguishes this attribute from impulse suppression and from habit.The article is accompanied by numerous responses by critics.

3.4.3 Understanding Psychological Difficulties through the Lens ofAkrasia

Could certain clinical psychopathological conditions that interferewith practical reasoning be fruitfully understood in terms ofakrasia? Kampa (2020) maintains that the empirical evidenceshows obsessive-compulsive disorder (OCD) to constitute a kind ofepistemicakrasia. He suggests that we can learn aboutdoxastic attitude-formation generally from research into the natureand treatment of OCD (Kampa 2020, p. 489). August Gorman (2019) arguesthat we can gain insights into the relations betweenakrasia,compulsion, and attributability by considering agents with Tourettesyndrome, misophonia, and addiction. Gorman’s account takes offfrom a phenomenon first identified by Gary Watson. Normally, when weare acting on our desires, we are trying tosatisfy thosedesires, not justeliminate them. By contrast, when an agentacts compulsively, they act merely to eliminate the desire, not tosatisfy it (Watson 1975, p. 20, Gorman 2019, p. 140). However, actionsthat we perform merely to extinguish or eliminate a desire seem not tobe attributable to us in the usual way. (For more on how Gormandistinguishes weakness of will from compulsion, see Gorman 2023.)Though people with Tourette syndrome, misophonia, and addiction mayexperience desires that feel a lot like “neurologicalnoise,” Gorman argues that things are not black and white as towhether ignoring or acting on that “noise” can beattributed to them. This marks an important contrast betweenGorman’s view and more traditional views of attributability suchas Deep Self views, according to which we cannot be attributable foran action if we are alienated from the motivation for it.

Recognizing the moral complexity of cases ofakrasia has ledsome philosophers to consider them in discussions about the nature ofblame (see Hartford 2020). There seem to be differences inblameworthiness among akratic wrongdoers: contrast those who laterrecognize that they acted wrongly with those who are aware of butindifferent to the wrongfulness of their actions and those who areignorant that what they have done is wrong. Variation in our judgmentsabout whether agents are responsible for various akratic actions havealso been used in order to characterize addiction and other forms ofpsychopathology.

In particular, there has been growing interest in the question ofwhether addiction might be best understood as a form ofakrasia. Building on the arguments of Mele (2002), NickHeather (2016) contends that addiction shares the core features ofakratic action and should be understood as a special kind ofakrasia, one in which agents consistently act against boththeir present judgments and their prior resolutions (Heather 2016, pp.133–4). Heather thus accepts Holton’s (1999) distinctionbetweenakrasia and weakness of will, but argues thataddiction paradigmatically involves both halves of thisdistinction.

On the other hand, Edmund Henden (2016) argues on phenomenologicalgrounds that addiction should not be assimilated to weakness of will.Addiction often involves habitual behavior, and addicts may continueto take drugs even when they don’t find it pleasurable to do so.(For discussion of habitual weak-willed action from a differentperspective, see Silver 2019.) Henden contends that in these respectsaddiction is profoundly unlike giving into temptation, as paradigmexamples ofakrasia are often described. Moreover, Hendennotes that weakness of will seems rationally criticizable in a waythat addiction does not, suggesting that they can’t be the samephenomenon. Addicts do not lack strength of will: many sincerely tryvery hard to abstain from using drugs, putting in a level of effortthat would normally suffice for success at a task. Federico Burdman(2023) uses the case of addiction to remind us of an important lessonabout self-control which we flagged in the previous section,underlining that we need an inclusive conception of self-control whichrecognizes the significance of external scaffolding and othercommunity-based strategies for helping us manage motivational conflictand stick to our resolutions.

Bibliography

• Adler, J. E., 2002, “Akratic Believing?,”Philosophical Studies, 110(1): 1–27.
• Ainslie, G., 1992,Picoeconomics, Cambridge: CambridgeUniversity Press.
• –––, 2001,Breakdown of Will,Cambridge: Cambridge University Press.
• –––, 2021, “Willpower With and WithoutEffort,”Behavioral and Brain Sciences, 44(e30):1–16.
• Altehenger, H., 2021a, “Deliberative Agency,Self-Control, and the Divided Mind,”Theoria, 87:542–58.
• –––, 2021b, “The MismatchProblem: Why Mele’s Approach to the Puzzle of SynchronicSelf-Control Does Not Succeed,”Pacific PhilosophicalQuarterly, 102(2): 243–66.
• –––, 2021c, “Self-Control and theSelf,”Synthese, 199: 2183–2198.
• Antill, G., 2019, “Agency, Akrasia, and the NormativeEnvironment,”Journal of the American PhilosophicalAssociation, 5(3): 321–38.
• Aristotle,Nicomachean Ethics, Book VII, Chs.1–10.
• Arpaly, N., 2000, “On Acting Rationally Against One’sBetter Judgment,”Ethics, 110: 488–513.
• Audi, R., 1979, “Weakness of Will and PracticalJudgment,”Noûs, 13: 173–196.
• –––, 1990, “Weakness of Will and RationalAction,”Australasian Journal of Philosophy, 68:270–281.
• Austin, J. L., 1956/7, “A Plea for Excuses,” in Austin1979, pp. 175–204.
• –––, 1979,Philosophical Papers,3^rd ed., J. O. Urmson and G. J. Warnock (eds.), Oxford:Oxford University Press.
• Bales, A., and Handfield, T., 2021, “Resolutions ProvideReasons or ‘How the Cookie Monster Quit Cookies,’”Synthese, 199: 4829–4840.
• Bennett, J., 1974, “The Conscience of HuckleberryFinn,”Philosophy, 49: 123–134.
• Bermúdez, J.P., Berthelette, S., Fernández, G.,Anaya, A., and Rodríguez, D., 2024, “Temptation andApathy,” in S. Amaya, D. Shoemaker, and M. Vargas,OxfordStudies in Agency and Responsibility, 8: 10–33.
• Bermúdez, J.P., 2021, “The Skill ofSelf-Control,”Synthese, 199: 6251–6273.
• Bermúdez, J.P., Murray, S., Chartrand, L., and Barbosa, S.,2023, “What’s Inside Is All That Counts? The Contours ofEveryday Thinking about Self-Control,”Review of Philosophyand Psychology, 14: 33–55.
• Bobonich, C., and Destrée, P. (eds.), 2007, AkrasiainGreek Philosophy: From Socrates to Plotinus, Leiden, Boston:Brill.
• Bratman, M., 1979, “Practical Reasoning and Weakness of theWill,”Noûs, 13: 153–171.
• Burdman, F., 2023, “Diachronic and Externally-ScaffoldedSelf-Control in Addiction,”Manuscrito, 46(1):77–116.
• Buss, S., 1997, “Weakness of Will,”PacificPhilosophical Quarterly, 78: 13–44.
• Callard, A., 2014, “Ignorance and Akrasia-Denial in theProtagoras,”Oxford Studies in AncientPhilosophy, 47: 31–80.
• –––, 2015, “The Weaker Reason,”The Harvard Review of Philosophy, 22: 68–83.
• Charles, D., 2011, “Akrasia: The Rest of theStory?,” inMoral Psychology and Human Action inAristotle, M. Pakaluk and G. Pearson (eds.), New York: OxfordUniversity Press, pp. 187–210.
• Charlton, W., 1988,Weakness of Will, Oxford: BasilBlackwell.
• Christensen, D., 2024, “Epistemic Akrasia: No ApologyRequired,”Noûs, 58(1): 54–76.
• Coates, A., 2012, “Rational Epistemic Akrasia,”American Philosophical Quarterly, 49: 113–124.
• Dang, J., and Hagger, M., 2019, “Time to Set a New ResearchAgenda for Ego Depletion and Self-Control,”SocialPsychology, 50: 277–281.
• Daoust, M-K., 2019, “Epistemic Akrasia and EpistemicReasons,”Episteme, 16(3): 282–302.
• Davidson, D., 1970, “How Is Weakness of the WillPossible?,” in Davidson 1980, pp. 21–42.
• –––, 1978, “Intending,” in Davidson1980, pp. 83–102.
• –––, 1980,Essays on Actions andEvents, Oxford: Clarendon Press.
• –––, 1982, “Paradoxes ofIrrationality,” in Davidson 2004, pp. 169–187.
• –––, 2004,Problems of Rationality,Oxford: Clarendon Press.
• Dodd, D., 2009, “Weakness of Will asIntention-Violation,”European Journal of Philosophy,17: 45–59.
• Duckworth, A. L., Gendler, T. S., and Gross, J. J., 2016,“Situational Strategies for Self-Control,”Perspectives on Psychological Science, 11(1):35–55.
• Duckworth, A. L., White, R. E., Matteucci, A. J., Shearer, A., andGross, J. J., 2016, “A Stitch in Time: Strategic Self-Control inHigh School and College Students,”Journal of EducationalPsychology, 108(3): 329–41.
• Dunn, R., 1987,The Possibility of Weakness of Will,Indianapolis: Hackett.
• Feldman, R., 2005, “Respecting the Evidence,”Philosophical Perspectives, 19: 95–119.
• Fix, J.D., 2021, “Practical Cognition as Volition,”European Journal of Philosophy, 30(3): 1077–91.
• Fleming, P., 2010, “Hume on Weakness of Will,”British Journal for the History of Philosophy, 18:597–609.
• Flowerman, C. H., 2024, “Weakness of Political Will,”Journal of Ethics and Social Philosophy, 27(1):94–120.
• Fraser, R., 2022, “Mushy Akrasia: Why Mushy Credences AreRationally Permissible,”Philosophy and PhenomenologicalResearch, 105(1): 79–106.
• Fritz, J., 2021, “Akrasia and Epistemic Impurism,”Journal of the American Philosophical Association, 7(1):1–19.
• Gorman, A., 2019, “The Minimal Approval View ofAttributability,” in D. Shoemaker (ed.),Oxford Studies inAgency and Responsibility, 6: 140–164.
• –––, 2023, “What is the Difference betweenWeakness of Will and Compulsion?”Journal of the AmericanPhilosophical Association, 9(1): 37–52.
• Gosling, J., 1990,Weakness of the Will, London and NewYork: Routledge.
• Greco, D., 2014, “A Puzzle about Epistemic Akrasia,”Philosophical Studies, 167(2): 201–219.
• Hare, R. M., 1952,The Language of Morals, Oxford:Clarendon Press.
• –––, 1963,Freedom and Reason, Oxford:Clarendon Press.
• –––, 2001, “Weakness of Will,” inThe Encyclopedia of Ethics, 2^nd ed., L. Becker andC. Becker (eds.), New York: Routledge, pp. 1789–1792.
• Hartford, A., 2020, “Complex Akrasia andBlameworthiness,”Journal of Philosophical Research,45: 15–33.
• Haug, M. C., 2021, “Trait Self-Control, Inhibition, andExecutive Functions: Rethinking Some Traditional Assumptions,”Neuroethics, 14(2): 303–314.
• Heather, N., 2016, “Addiction as a Form of Akrasia,”in N. Heather and G. Segal (eds.),Addiction and Choice:Rethinking the Relationship, Oxford: Oxford University Press, pp.133–148.
• Heinzelmann, N., 2021, “Conflicting Judgments and Weaknessof Will,”Philosophia, 49(1): 255–269.
• Henden, E., 2016, “Addiction, Compulsion, and Weakness ofthe Will: A Dual-Process Perspective,” in N. Heather and G.Segal (eds.),Addiction and Choice: Rethinking theRelationship, Oxford: Oxford University Press, pp.116–131.
• Hill, T., 1986, “Weakness of Will and Character,” inHill 1991, pp. 118–137.
• –––, 1991,Autonomy and Self-Respect,Cambridge: Cambridge University Press.
• Hinchman, E., 2013, “Rational Requirements and‘Rational’ Akrasia,”Philosophical Studies,166: 529–552.
• Hoffmann, T. (ed.), 2008,Weakness of Will from Plato to thePresent, Washington: Catholic University of America Press.
• Holton, R., 1999, “Intention and Weakness of Will,”Journal of Philosophy, 96: 241–262.
• –––, 2003, “How is Strength of WillPossible?,” in Stroud and Tappolet 2003, pp. 39–67.
• –––, 2009,Willing, Wanting, Waiting,Oxford: Clarendon Press.
• Holton, R., and May, J., 2012, “What in the World isWeakness of Will?,”Philosophical Studies, 157:341–360.
• Horowitz, S., 2014, “Epistemic Akrasia,”Noûs, 48(4): 718–744.
• Hurley, S., 1989,Natural Reasons, Oxford: OxfordUniversity Press.
• Irving, Z. C., Bridges, J., Glasser, A., Bermúdez, J.P.,and Sripada, C., 2022, “Will-Powered: Synchronic Regulation isthe Difference Maker for Self-Control,”Cognition, 225:1–12.
• Jackson, E., and Tan, P., 2021, “Epistemic Akrasia andBelief-Credence Dualism,”Philosophy and PhenomenologicalResearch, 104(3): 717–727.
• Jackson, F., 1984, “Weakness of Will,”Mind,93: 1–18.
• Jones, K., 2003, “Emotion, Weakness of Will, and theNormative Conception of Agency,” in A. Hatzimoysis (ed.),Philosophy and the Emotions, Cambridge: Cambridge UniversityPress, pp. 181–200.
• Kalis, A., 2011,Failures of Agency: Irrational Behavior andSelf-Understanding, Lanham, MD: Lexington Books.
• Kampa, S., 2020, “Obsessive-Compulsive Akrasia,”Mind & Language, 35(1): 475–492.
• Kennett, J., 2001,Agency and Responsibility: A Common-SenseMoral Psychology, Oxford: Clarendon Press.
• Lasonen-Aarnio, M., 2020, “Enkrasia or Evidentialism?Learning to Love Mismatch,”Philosophical Studies, 177:597–632.
• –––, 2014, “Higher-Order Evidence and theLimits of Defeat,”Philosophy and PhenomenologicalResearch, 88(2): 314–345.
• Lazar, A., 1999, “Akrasia and the Principle of Continence orWhat the Tortoise Would Say to Achilles,” in L. E. Hahn (ed.),The Philosophy of Donald Davidson (Library of LivingPhilosophers, Vol. 27), Chicago: Open Court, pp. 381–401.
• Levy, N., 2004, “Epistemic Akrasia and the Subsumption ofEvidence: A Reconsideration,”Croatian Journal ofPhilosophy, 4(10): 149–156.
• Lin, H., “Bayesian Epistemology,”The StanfordEncyclopedia of Philosophy (Fall 2022 Edition), E. N. Zalta &U. Nodelman (eds.), URL =<https://plato.stanford.edu/archives/fall2022/entries/epistemology-bayesian/>.
• List, C., and Pettit, P., 2011,Group Agency, Oxford:Oxford University Press.
• Liu, W., 2022, “Ignorance in Plato’sProtagoras,”Phronesis, 67(3):309–37.
• Marshall, E., 2010, “Spinoza on the Problem ofAkrasia,”European Journal of Philosophy, 18:41–59.
• McIntyre, A., 1990, “Is Akratic Action AlwaysIrrational?,” inIdentity, Character, and Morality, O.Flanagan and A. Rorty (eds.), Cambridge, MA: MIT Press, pp.379–400.
• –––, 2006, “What Is Wrong With Weakness ofWill?,”Journal of Philosophy, 103: 284–311.
• Mele, A., 1987,Irrationality, New York: OxfordUniversity Press.
• –––, 1991, “Akratic Action and thePractical Role of Better Judgment,”Pacific PhilosophicalQuarterly, 72: 33–47.
• –––, 2002, “Akratics and Addicts,”American Philosophical Quarterly, 39: 153–167.
• –––, 2012,Backsliding: UnderstandingWeakness of Will, New York: Oxford University Press.
• Mylopoulos, M., and Pacherie, E., 2020, “Self-Control asHybrid Skill,” in A. R. Mele (ed.),SurroundingSelf-Control, New York: Oxford Academic, pp. 81–115.
• Natali, C. (ed.), 2009,Aristotle’s NicomacheanEthics,Book VII, Oxford: Oxford University Press.
• Owens, D., 2002, “Epistemic Akrasia,”TheMonist, 85(3): 381–397.
• Obdrzalek, S., 2023, “Evaluative Illusion in Plato’sProtagoras,” in V. Caston and R. Kamtekar (eds.),OxfordStudies in Ancient Philosophy, 62: 41–84.
• Orlandi, M., and Stroud, S., 2021, “Self-Control in Actionand Belief,”Philosophical Explorations, 24(2):225–242.
• Pears, D., 1984,Motivated Irrationality, Oxford:Clarendon Press.
• Pettit, P., 2003, “Akrasia, Collective andIndividual,” in Stroud ad Tappolet 2003, pp. 68–96.
• Plato,Protagoras, inThe Collected Dialogues ofPlato, E. Hamilton and H. Cairns (eds.), Princeton: PrincetonUniversity Press, 1961, pp. 308–352.
• Ribeiro, B., 2011, “Epistemic Akrasia,”International Journal for the Study of Skepticism, 1:18–25.
• Rorty, A., 1980, “Where Does the Akratic Break TakePlace?,”Australasian Journal of Philosophy, 58:333–347.
• Rosa, L., 2022, “Ambiguous Statements about Akrasia,”Journal of Philosophy, 119(11): 581–601.
• Saarinen, R., 1994,Weakness of the Will in Medieval Thought:From Augustine to Buridan, Leiden, New York: Brill.
• Schapiro, T., 2021,Feeling Like It: A Theory of Inclinationand Will, Oxford: Oxford University Press.
• –––, 2022, “What Makes Weak-Willed ActionWeak?” in T. Schapiro, K. Ebels-Duggan, and S. Street (eds.),Normativity and Agency: Themes from the Philosophy of Christine M.Korsgaard, Oxford: Oxford University Press, pp.126–146.
• Schwenkler, J., 2022, “Steering Clear of Trouble,”Philosophic Exchange.
• Silver, K., 2019, “Habitual Weakness,”Thought, 8(4): 270–77.
• Sinhababu, N., 2020, “Rational Epistemic Akrasia for theAmbivalent Pragmatist,” in D. E. Gatzia and B. Brogaard (eds.),Being of Two Minds: The Philosophy and Psychology ofAmbivalence, New York: Routledge, pp. 75–93.
• Smith, M., 2003, “Rational Capacities, or: How toDistinguish Recklessness, Weakness, and Compulsion,” in Stroudand Tappolet 2003, pp. 17–38.
• Soravia, L. M., Schläfli, K., Stutz, S., Rösner, S., andMoggi, F., 2015, “Resistance to Temptation: The Interaction ofExternal and Internal Control on Alcohol Use During ResidentialTreatment for Alcohol Use Disorder,”Alcoholism: Clinicaland ExperimentalResearch, 39(11): 2209–2214.
• Stocker, M., 1979, “Desiring the Bad: An Essay in MoralPsychology,”Journal of Philosophy, 76:738–753.
• Stroud, S., 2003, “Weakness of Will and PracticalJudgement,” in Stroud and Tappolet 2003, pp. 121–146.
• –––, 2013, “Irrationality,” inACompanion to Donald Davidson, E. Lepore and K. Ludwig (eds.),Oxford: John Wiley & Sons, pp. 489–505.
• Stroud, S., and Tappolet, C. (eds.), 2003,Weakness of Willand Practical Irrationality, Oxford: Clarendon Press.
• Tappolet, C., 2003, “Emotions and the Intelligibility ofAkratic Action,” in Stroud and Tappolet 2003, pp.97–120.
• Tenenbaum, S., 1999, “The Judgment of a Weak Will,”Philosophy and Phenomenological Research, 59:875–911.
• –––, 2007,Appearances of the Good: An Essayon the Nature of Practical Reason, Cambridge: CambridgeUniversity Press.
• Thero, D., 2006,Understanding Moral Weakness, Amsterdam,New York: Rodopi.
• Tremblay, M., 2020, “Akrasia in Epictetus: A Comparison withAristotle,”Apeiron, 53(4): 397–417.
• Wagner, J., Burg, M., and Sirois, B., 2004, “Social Supportand the Transtheoretical Model: Relationship of Social Support toSmoking Cessation Stage, Decisional Balance, Process Use, andTemptation,”Addictive Behaviors, 29(5):1039–1043.
• Walker, A., 1989, “The Problem of Weakness of Will,”Noûs, 23: 653–676.
• Wallace, R. J., 1999, “Addiction as Defect of the Will: SomePhilosophical Reflections,”Law and Philosophy, 18:621–654.
• Watson, G., 1975, “Free Agency,”The Journal ofPhilosophy, 72(8): 205–220.
• –––, 1977, “Skepticism About Weakness ofWill,”Philosophical Review, 86: 316–339.
• Weatherson, B., 2019,Normative Externalism, Oxford:Oxford University Press.
• Wedgwood, R., 2011, “Justified Inference,”Synthese, 189(2): 1–23.
• Wiggins, D., 1979, “Weakness of Will, Commensurability, andthe Objects of Deliberation and Desire,”Proceedings of theAristotelian Society, 79: 251–277.
• Wilkerson, T. E., 1997,Irrational Action: A PhilosophicalAnalysis, Aldershot: Ashgate.

Academic Tools

	How to cite this entry.
	Preview the PDF version of this entry at theFriends of the SEP Society.
	Look up topics and thinkers related to this entry at the Internet Philosophy Ontology Project (InPhO).
	Enhanced bibliography for this entryatPhilPapers, with links to its database.

Other Internet Resources

[Please contact the authors with suggestions.]

Related Entries

action |Aristotle, General Topics: ethics |Davidson, Donald |practical reason |practical reason: and the structure of actions |practical reason: medieval theories of |self-doubt, epistemic |thick ethical concepts

Acknowledgments

I am grateful to Eric Guindon and to Joseph van Weelden for veryuseful research assistance.