The Guise of the Guise of the Bad

Transcription

1 Ethic Theory Moral Prac (2018) 21: The Guise of the Guise of the Bad Sergio Tenenbaum 1 Accepted: 14 January 2017 / Published online: 28 November 2017 # Springer Science+Business Media B.V., part of Springer Nature 2017 Abstract It is undeniable that human agents sometimes act badly, and it seems that they sometimes pursue bad things simply because they are bad. This latter phenomenon has often been taken to provide counterexamples to views according to which we always act under the guise of the good (GG). This paper identifies several distinct arguments in favour of the possibility that one can act under the guise of the bad. GG seems to face more serious difficulties when trying to answer three different, but related, arguments for the possibility of acting under the guise of the bad. The main strategies available to answer these objections end up either undermining the motivation for GG or failing to do full justice to the nature of perverse motivation. However, these difficulties turn out to be generated by focusing on a particular version of GG, what I call the Bcontent version^. But we have independent reasons to prefer a different version of GG; namely, the Battitude version^. The attitude version allows for a much richer understanding of the possibility of acting on what we conceive to be bad. Drawing on an analogy with theoretical akrasia and theoretical perversion, I try to show how the attitude version can provide a compelling account of perverse actions. Keywords Guise of the good; philosophy of action. Practical reason. Perversion. Desire 1 Introduction The guise of the good view (GG) says, roughly, that all that we desire and pursue is desired and pursued under the guise of the good. Philosophers often insist that those committed to GG must deny the possibility of acting under the guise of the bad, and, consequently, must reinterpret, or somehow redescribe beyond recognition, very ordinary phenomena. Velleman has an elegant characterization of these apparent counterexamples to GG: The assumption that desire aims at the good... misdescribe(s) examples of perverse desire. Consider the figure of Satan in Paradise Lost, who responds to his defeat with the * Sergio Tenenbaum sergio.tenenbaum@utoronto.ca 1 University of Toronto, Toronto, ON, Canada

2 6 S. Tenenbaum cry, "Evil be thou my Good." If Satan ever loses sight of the evil in what he now desires, if he ever comes to think of what he desires as really good, he will no longer be at all satanic; he'll be just another well-intentioned fool. (Velleman 1992, p. 18) It seems natural to think that certain common, or at least well-understood pursuits, aim at the bad, and that only under the influence of a theory could we lose sight of this possible object of our will. Velleman takes it to be part of our understanding of the extreme form of evil represented by Satan that Satan does not aim at the good. The lines from Milton might suggest that Satan pursues the bad under the guise of the good, but if they are taken too literally, Satan s pursuit of the bad seems to imply that Satan pursues the bad by embracing a manifest contradiction. Satan would thus be worse than a fool: he would seem to display a pitiful incompetence with the basic rules of logic. At any rate, GG seems committed to an interpretation of Satan as making this basic theoretical mistake. In this paper, I first identify different kinds of (apparently) perverse pursuits, and examine how GG could account for them. I argue that a certain plausible understanding of satanic motivations and of Augustine s motivation in the pear theft episode can pose a serious challenge to GG. To meet this challenge, I distinguish two versions of GG: the Bcontent version^ and Bthe attitude version^. The attitude version turns out to be independently plausible and to provide a compelling account of both the conceptual possibility and the irrationality of perverse action. 1 2 Guise of the Good and the Various Pursuits of the Bad There have been a number of recent defences of GG; I ll not try to rehash them all here. 2 For my part, I take GG to be essential to understanding the unity of reason in the theoretical and practical realm, as well as understanding intentional agency as a manifestation of a rational power. On this understanding of GG s role, the thesis is not an empirical generalization, but a claim about the a priori nature of rational agency. 3 GG can be a theory about the nature of desire or the nature of intentional agency, but I focus here on versions that accept both. In particular, I assume that GG accepts the following claims: [Want] S wants to φ only if S conceives φ-ing to be good. [Intend] S intends to φ only if S judges φ-ing to be good. [Do] S φ-s intentionally only if S judges φ-ing to be good. These claims are often accepted because the good is taken to be the formal end of practical reason or intentional agency; that is, in pursuing or wanting anything an agent is pursuing something she takes to be good. 4 Opponents of GG either propose a different formal end, 5 or 1 It is worth noting that the view presented here is committed only to the claim that the pursuit of the bad as such is irrational; it is not committed to ethical rationalism. In other words, the view presented here does not imply that evil actions are irrational (since evil actions generally do not have the bad as such as their aim) or that moral knowledge is rational knowledge. 2 For my own, see Tenenbaum (2007, 2008, 2012). For others, see Lavin and Boyle (2010); Raz (2010); Schafer (2013); Wald (2017). 3 But to accept this claim is, of course, not to say that the theory is immune to counterexamples or empirical challenges; we might suspect that our understanding of rational agency is flawed if it seems to conflict with what appears to be ordinary instances of agency that are incompatible with the theory. 4 For a clear exception, see Boswell (n.d.). 5 Velleman (1992)

3 The Guise of the Guise of the Bad 7 argue against the very idea that there is a formal end to desire or practical reason. 6 These arguments often consist in finding forms of imperfect or less than ideal human agency that are supposed to refute GG. Philosophers have argued that GG cannot make room for akrasia, depressed agents, agents with conceptual limitations such as children and non-human animals, etc. 7 But the possibility of perverse agents seems to be a particularly difficult challenge for GG. How could one see their pursuit of what they take to be bad to be good in any way? There are various ways in which an agent can be said to be pursuing what she takes to be bad, and in order to understand how GG can accommodate perverse pursuits, we need to first have a more systematic account of the different ways in which one pursues something that is, or is thought to be, bad. In the rest of this section, I provide such an account. 2.1 BDe re^(object): Sometimes we pursue X, because X is an instance of Y, and X is bad but not because it is an instance of Y. J. L. Austin (Austin 1964) gives the (fictitious?) example of his helping himself to a second bombe for dessert when the number of bombes served is the same as the number of guests. Austin presumably judged his act was wrong or bad; one should leave enough and as good for the other guests at the table. But Austin did not take the second bombe in order to deprive other guests of their fair share of food; rather, he took the extra bombe because it was delicious, or for some similar reason. Being an act of consuming delicious food is certainly not what makes the act wrong. There are many ways in which GG can make room for Austin s action. One obvious possibility is to weaken GG so that it requires that the agent pursue an action insofar as she conceives it to be in some way good, rather than good overall or good simpliciter. 8 However, such proposals significantly weaken GG, and they might not be compatible with the overall motivation for the view. 9 But stronger views can also accommodate this type of akratic action. 10 Let us make a distinction between prima facie states such as desires, and all-out or overall states such as actions or intentions. On many versions of GG, in desiring X we do not judge X to be good, but we are merely in a state in which X appears or seems good to us. 11 Intending to X or X-ing intentionally, on the other hand, is, or entails, 12 an all-out judgment that X is good. Once we understand desires and intentions or intentional doings in this manner, we can represent practical reasoning in the model of defeasible reasoning. Desires provide us with warrant for certain conclusions or Bdefaults^13 that can be defeated or undermined. On this picture, my desire to eat delicious food warrants the following inference: Bombe This is a bombe (an instance of delicious food) I eat the bombe 6 Most explicitly, Setiya (2007), but see also Baker (2015). 7 For two classical versions of such arguments, see Stocker (1979); Velleman (1992). 8 See Clark (2010) 9 See Baker (2015) on this issue. 10 See Tenenbaum (1999, 2007) for a detailed exposition and defence of the basic framework I present here. Some of the inferences below follow similar lines than the ones proposed by Davidson (1980). 11 Scanlon (1998); Stampe (1987); Tenenbaum (1999, 2007). 12 Since the view I prefer is that the action (or intention) is the judgment (see p for the more precise version), I ll drop the parenthetic remark Bor entails^ from here on. 13 For an account of the relevant notion of default, see Horty (2007, 2012).

4 8 S. Tenenbaum Since we want to be clear that this is a form of the guise of the good view, it will be better to provide a version of this inference with good as part of the content of the desire and of the conclusion, while making the desire appear explicitly as one of the premises in the inference. As we ll see later, this is not the version I ll defend: Bombe 1 Eating delicious foods is good (desire) Eating the bombe is eating delicious food Therefore Eating this bombe (as I am doing right now) is good My desire to refrain from performing unjust acts should defeat the presentation of eating the bombe as good provided by my desire for eating delicious food. 14 So given all that I know and want the correct inference should be: Bombe 2 Refraining from unfair acts is good Eating the bombe is unfair Refraining from eating the bombe (as I m doing right now) is good Given that practical reasoning is non-monotonic, the validity of Bombe 1, in the absence of defeaters and underminers, does not make the following valid: Bombe 3 Eating delicious foods is good (desire) Eating the bombe is eating delicious food Eating the bombe is unfair * eating this bombe (as I am doing right now) is good We can now represent Austin as making the inference represented in Bombe 1. He acts under the guise of the good, but the fact that he knows that eating the bombe is bad is represented by his knowing that Bombe 2 is valid and Bombe 3 is invalid. Of course, given that Austin knows that eating the bombe is unfair and that the presence of this Bdefeater^ invalidates the reasoning in Bombe 1, his accepting this reasoning is irrational, but this is exactly as we would expect: Austin is exhibiting a form of akrasia,andakrasia is a species of irrationality. In fact, we can even add that Austin knows (theoretically) that he should reason as in Bombe 2 rather than as in Bombe Of course, there is much more to be said about this framework, but I ll assume that some version of this model can account for the cases under 14 Using (an adaptation of) John Horty s model Horty (2007, 2012), we can say that the default BX is unfair Don t X^ ranks higher than the default BX is eating delicious food Do X^ 15 This preserves many aspects of Davidson s account in Davidson 1980, but dispenses with reliance on Principle of Continence and with ascribing any kind of comparative judgment to the akratic agent. This version of the account is thus immune to the objections raised against it in McDowell (2010); namely, that Davidson saccount implausibly ascribes to the akratic agent a comparative (all-out) judgment that the akratic action is better than the enkratic action. McDowell thinks that this problem is a consequence of Davidson s commitment to having an evaluative proposition as the content of an intention. As it ll become clear, I am sympathetic with McDowell s rejection of this understanding of the content of an intention (though I have doubts that Davidson is committed to it), but I think one can accept it and steer safely away from the problematic implication of Davidson saccountof akrasia.

5 The Guise of the Guise of the Bad 9 which we knowingly pursue the bad, when the bad is merely the Bde re^ object of our pursuit; that is, cases in which we pursue something for some attractive feature but we know that the object of our pursuit is all-things-considered bad. 2.2 De re (Property) Let us suppose that Inigo Montoya is pursuing revenge for his father, and let us assume that Inigo Montoya, at least in his more reflective moments, understands that the pursuit of revenge killing is in fact bad. In other words, Inigo Montoya is attracted to the very feature of an act (revenge killing) that makes his act bad. Here our original strategy would not work; after all there is no otherwise valid reasoning that has been defeated by the presence of competing consideration. The equivalent of Bombe 1 in our original argument would have a premise that the agent would not endorse: (1A) Revenge killing is good* Killing Rugen is an act of revenge killing killing Rugen (as I am doing right now) is good Still, an analysis very similar to the one for the BDe Re^ Object case applies here. In desiring to avenge his father, revenge killing does appear to Inigo Montoya to be a good act. However, he understands, upon reflection, that this perspective on the good is illusory, much like someone who realizes that the perceptions in the desert of bodies of water in the horizon are not to be trusted. 16 In such cases, one s perspective has been undermined rather than defeated: the epistemological equivalent of such cases would be cases in which rather than having counterevidence, I have higher-order evidence that what I take to be evidence should not be trusted (because I have taken a drug, or because I know I am susceptible to some kind of illusion). We can represent experiencing this Bappearance^ as a matter of accepting or holding something like the following: Prima facie, revenge killing is good But Inigo Montoya also knows that this appearance is undermined. So if he now reasons as in (1A), he must have inferred the major premise from an appearance that he knew to be a mere appearance. He would be like someone who infers that there is water on the horizon on the basis of what she knows to be a mirage. This is doubtless a form of irrationality, but there is nothing incoherent in us in describing the agent as relying on an unwarranted premise. 2.3 De Dicto When Milton s Satan says: BEvil be thou my good^, he does not seem to be interested in a specific object that happens to be evil. He seems to be interested in pursuing the bad for its own sake: what attracts Satan to an act is, it seems, exactly its being an evil act, rather than its being an act of killing, an act of theft, or anything like that. Moreover, Satan does not seem to be making any kind of mistake or inferential faux-pas. He acts in full knowledge of the evil of his acts and he performs them ingenuously and unhesitantly. 17 And yet, GG is committed to attribute to Satan endorsement of something like the following: 16 For a more detailed discussion of these issues, see Tenenbaum (1999, 2007). 17 This point is emphasized by Velleman (1992).

6 10 S. Tenenbaum Acting badly (performing a bad act) is good (is a good act) But this is a manifest contradiction. It is unclear that attributing the clear-eyed acceptance of a manifest contradiction to an agent is a coherent option for the theorist. But even if it is, it would be strange that Satan turns out to be simply logically incompetent. Satan was not supposed to be an example of an agent with a dismal grasp of elementary logic: it seems that GG implies that Satan s character is not a coherent possibility. One might have doubts whether this is an implausible conclusion, 18 but I will argue that even if it is, GG can provide a plausible account of the kind of perverse motivation that Satan exemplifies. More precisely, I ll try to show both that (a) many of the ordinary, or at least recognizable, examples of agents intentionally pursuing the bad cannot establish that these agents are pursuing the bad for its own sake, and (b) what I ll call Bthe attitude version of GG^ allows for any plausible form of perverse motivation. 3 Interlude: Pro Tanto Bad Before we move on, it is worth looking at a slightly different set of cases that might seem to present a related challenge to GG. Sometimes we act in ways that are typically or intrinsically bad, but performing these acts involves no kind of irrationality or mistake. To use Aristotle s example, throwing the cargo overboard is an action that is typically bad (as Aristotle says Bno one would willingly throw cargo overboard^19 ) and can only (at least in normal cases) be pursued as an instrumental means to do something good; we do not pursue throwing cargo overboard because of the features that make it bad (loss of income) but for something different that it brings about (the survival of the ship through the storm). However sometimes we pursue something because it is an instance of something that is typically bad exactly because of the features that normally make it bad. One obvious case is punishment. Of course, there are many theories of punishment, but I assume that some forms of punishment are supposed to be the deserved response to a certain action. As I m packing a lunch for my daughter to take to school, I notice that I haven t been shopping and the only thing I can pack for her is a canned sardine sandwich. Disgusted with my failure as a parent, I decide I should have the same lunch as my daughter, even though I could have easily bought lunch for myself at the cafeteria. As I experience the foul texture of canned sardines and their repugnant flavor, I think: Bthis is exactly what I deserve for being such a bad parent^. Of course, the punishment can only be delivered if the taste is indeed repugnant and the texture is actually foul; it would not be an instance of punishment otherwise. Still, what I pursue is something that I conceive to be good. There are many ways to understand the value of deserved punishment. One possibility is to follow a roughly Moorean path, 20 and argue that the good as a whole of the organic whole constituted by my bad action combined with my suffering overrides the badness of the pain that I suffer. Adding the value to the organic whole in this manner allows the Moorean to conclude that the punishment is good on the whole. Alternatively, one can say that a certain form of desert can ground a Bvalue inversion^: if I acted badly, then some experiences I undergo that would typically be good are, in virtue of my sins, bad, and vice-versa. The 18 Kant seems to rule out this kind of motivation for human beings, but he is not committed to the overall impossibility of a being for whom evil can be itself an incentive. See Kant (1998, p.58(6:36)) 19 Aristotle (1999, p. 1110a:11) 20 Moore (1966, pp ).

7 The Guise of the Guise of the Bad 11 important point is that it is possible that I conceive an act to be good even if some of its constituents are essentially bad, at least considered in isolation. These cases do not present a challenge to any version of GG, but it is sometimes easy to confuse them with instances of pursuing the bad for its own sake Cases of Perversity There have been a number of cases that philosophers have put forward as cases of perversity. Before we move on, it is worth noting that there is a very general problem that must be faced by any attempt to show the coherence of pursuing the bad de dicto by means of example. Examples of agents who are intelligibly pursuing bad ends are, of course, always cases in which particular agent is pursuing a specific bad ends. But why would we think that, for any such case, it is the badness itself that attracts, rather than the property that makes it bad. In other words, why should we think that any such a case is a case of de dicto pursuit of the evil, as opposed to a case of de re (property) pursuit of evil. For instance, (Sussman 2009) canvasses a number of ordinary examples of pursuit of minor pain: playing with a loose tooth, picking a scab. But why must we describe such cases as being attracted to badness, rather than to the particular feeling of unpleasantness, which, in these circumstances, we take to be good? In order to find a case in which we are convinced we have a proper BDe Dicto^ case, we need to find that the agent is responding to the badness itself, but this generates its own problems. After all, if one is responding to the badness itself, then it seems that the badness will attract however it is instantiated; there ll be something to be said for any other form of badness. But this is certainly not the case in these examples. The badness of, say, bad art or even other forms of mild pain, do not attract the scab-picker in the same way. Of course, this does not show that the attempt to understand these cases as pursuing something for the sake of its badness is incoherent or unsustainable. But the fact that the agent is motivated only by this particular instance of acting badly does show that the Bde dicto^ description of the case is not particularly compelling. Satan is often mentioned as an example of an agent who pursues the bad for badness sake. Let us look again at Velleman s description of Satan: [Satan] cannot reasonably be interpreted as adopting new estimates of what s valuable that is, as resolving to cease judging evil to be evil and to start judging it to be good. If Satan ever loses sight of the evil in what he now desires, if he ever comes to think of what he desires as really good, he will no longer be at all satanic; he'll be just another well-intentioned fool. The ruler of Hell doesn't desire what he wrongly thinks is worthy of approval; he desires what he rightly thinks isn't. (Velleman 1992 p. 18) But Satan s example also raises doubts. Although people seem to be divided on this issue, Satan is probably a fictional example at the very least, our knowledge of his motivation seems to come from storytelling. But arguably Satan is a possible agent, perhaps even just an extreme version of possible human motivations. Of course, a great deal of Satan s motivation 21 Sussman (2009) argues that camp attitudes to certain movies or enjoying kitsch would be cases of pursuing the bad that pose a problem for GG. But these cases have a structure similar to the case of punishment. A certain complex attitude to something that is bad (something like appreciating the utter failure of a movie making, or the enjoyable aspects of something that one recognizes as a failure) makes for a good response, and thus a response worthy of pursuit. In such cases, the naïve enjoyment of these forms of art would simply be an inadequate aesthetic response.

8 12 S. Tenenbaum consists in quite ordinary desires: hatred for, and envy of, a more powerful being, revolt against submitting to another (even if benevolent) being, motivate many ordinary human beings. 22 However, Satan is supposed to also have the motivation to pursue the bad or evil as such. But could he really corrupt the soul of the innocent, without any thoughts of showing off his immense power, with no thought of sticking it to the Man (God), motivated not by hatred or by an attempt to assert his independence, but simply because it is bad? If someone insists that Satan s motivationis really just to pursue the bad as such, aren t we tempted to ask, Bwhat is the good of it... what s the good of being bad^, 23 aren t we tempted to look for a deeper motive? In fact, the view seems only plausible if we attribute to Satan the pursuit of a specific form of badness, Bmoral badness^, rather than badness as such. For it seems false to say that Satan pursues the bad simpliciter; after all, Satan is not motivated to eat foul food, watch paint dry, or sing off key; these activities are also clearly bad (though not Bmorally^ bad). 24 A de dicto concern for the bad should, at least in the absence of defeaters, speak in favour of any bad activity just as a concern for healthy living speaks in favour of any healthy activity. 25 Perhaps the best example of an attempt to show the intelligibility of the pursuit of evil is the celebrated pear episode in Augustine s Confessions. Augustine describes his and his friends motivations in stealing pears as follows: I loved my own perdition and my own faults, not the things for which I committed the wrong, but the wrong itself. (Augustine 1961, p. 47) It was not the pears that my unhappy soul desired. I had plenty of my own, and I only picked those so that I might steal. For no sooner I picked them I threw them away, and tasted nothing in them but my own sin, which I relished and enjoyed. If any part of one of these pears passed my lips, it was the sin that gave it flavor. (Augustine 1961, p. 49) Augustine s description of this episode is certainly compelling, and he carefully rules out various possible aspects of the act that one might propose as his reason to steal the pears: he doesn t do it for profit (Bno sooner I picked them I threw them away^) or for gastronomic pleasure (Btasted nothing in them but my own sin^). Some philosophers have argued that Augustine himself intends the pear episode to be a challenge to, but not a refutation of, GG 26 ; indeed, very much like in Anscombe s analysis of Satan s motivation, we find Augustine s actions more intelligible as an instance of rebellion or social ritual, then as the pursuit of bad for badness sake. So Satan s and Augustine s motivation seems to be a form of rebellion, a form of expression of their freedom, their Bintact liberty in the unsubmissiveness of... [their] will.^27 But if God commands that one pursues the good, then one can show one s Bintact liberty^ only by pursuing some form of badness. 28 In this way, even if the object of the pursuit is not badness 22 Of course one might argue that these motivations present a problem for GG. Responding to such concerns would take us far astray. Moreover, if GG cannot make room for these motivations, we do not need to consider the possibility of pursuing the bad for badness sake; we already know the view is doomed. 23 Anscombe (2000, p. 75). 24 I make a similar claim in Tenenbaum (2009). 25 Sussman (2009) correctly points out that a concern for the bad does not imply that one favours engaging in pointless activities, but it does seem to imply at least some inclination, or favourable response, to the pursuit of other bad activities. I return to this point momentarily. 26 See, for instance, MacDonald (2003). 27 Anscombe (2000, p.75) 28 This requires some conception of freedom as liberty of indifference, or, as in Reinhold s view, freedom needs to be conceived as the freedom to act well or badly Reinhold (2005).

9 The Guise of the Guise of the Bad 13 itself, it is essential that it be something that is bad. After all, a pursuit of an object that is good could not assert one s freedom in relation to a supremely benevolent being; it is only by pursuing something that is bad that one could show one s independence from such a being. If this is an intelligible way of interpreting Augustine s or Satan s desires and actions, they do seem to raise a problem for GG. We could represent the content of, say, Augustine s attitudes as follows: (FREE) Rebellious acts of pursuing the bad are good Stealing the pear is a rebellious act of pursuing the bad Stealing the pear is good But even though this is no longer a case in which someone is simply pursuing the bad for badness sake, it seems no less problematic for GG. For, after all, the major premise seems to include a manifest contradiction, and if we accept Anscombe s view that an agent knows not only what she is doing intentionally, but also why she s doing it, 29 the agent would be fully aware of the manifest contradiction that her action embodies. Now, GG can and must accept that perverse actions are irrational; it is worth noting that it is essentially a consequence of any form of ethical rationalism that any type of evil action is irrational. But explaining the intelligibility of Augustine s action in terms of explicit acceptance of (and possibly belief in) a manifest contradiction seems, at least at a first glance, rather implausible. I propose we make a distinction here between a weak and a strong version of the possibility of perversion: Strong Perversion Some intentional actions A are such that the action is fully explained by the fact that the agent believed that in doing A she was doing something bad, and that the agent desired to engage in a bad action. 30 I argued above against the idea that there are any compelling examples of Strong Perversion. However, it seems significantly more plausible to accept the following claim: Weak Perversion Some intentional actions A are such that the action is partly explained by the facts (1) that the agent believed that in doing A she was doing something bad, and (2) that the agent desired to engage in an action only insofar as it was bad. Accepting the possibility of weak perversion does not commit us to the view that the bad as such attracts, but only that it is an essential component of what attracts. It is significantly more plausible that our understanding of the possibility of Satanic motivation, or the less grandiose form of perversion involved in Augustine s theft of the pears, involves seeing it as an instance of Weak Perversion. But even Weak Perversion seems to require that GG attribute beliefs in manifest contradictions to these agents. The manifest contradiction depends on good appearing in the content of the attitudes that leads the agent to the conclusion that a certain act is good. Some versions of GG interpret the thesis as saying that if an agent desires (or intentionally does) X, she has an attitude (a belief, or 29 Anscombe (2000) 30 Depending on one s favourite action theory, one might prefer a different formulation of Strong Perversion. The essential point is that Strong Perversion claims that the Bad on its own can be an agent s reasontoact.

10 14 S. Tenenbaum some other attitude) with the content BX is good^, 31 but these are not the only ones. On other versions of GG, this is not true. According to what I ll call Bthe attitude version^, 32 Bgood^ is a feature of the form of the attitude, not its content. This view takes the relation between, on the one hand, desiring, intending, or acting intentionally, and, on the other hand, good, to be analogous to the relation between believing and true. In having a belief that p one takes p to be true even though is true is not (or at least does not need to be) part of the content of one s belief that p. In the same way, in intending X, desiring X, or doing X intentionally, one holds X to be good even though good is not (necessarily) part of the content of these attitudes. 33 Of course, there are many questions to be raised about such a view. There is no uncontroversial understanding of the relation between belief and truth; we need to spell out in more detail the analogy between truth and good here. 34 But for our present purposes, the main features of the attitude version of GG are: (i) good does not appear in the content of the relevant practical attitudes, and (ii) good and true have analogous roles in, respectively, practical and theoretical reason. But this is enough to make room within the attitude view for interpreting Augustine s actions in an unproblematic way. The basic form of the Augustinian thought is given by the following schema, in which the agent s attitudes appear in square brackets and are followed by the content of the attitude: (FREE)* [DESIRE] Performing rebellious acts [BELIEF] Stealing the pear is a rebellious act [INTENTION] Stealing the pear The attitudes themselves are forms of Bholding true^ and Bholding good^ and commitment to GG is manifested by understanding intention and desire as a form of holding or taking the content to be good. 35 This can be made explicit by replacing the specific attitude by its general form: (FREE)** [HOLDS GOOD] Performing rebellious acts [HOLDS TRUE] Stealing the pear is a rebellious act [HOLDS GOOD] Stealing the pear Although neither good nor bad appear in the contents of this inference, the agent who makes this inference, as we said above, holds the contents of the first premise and the conclusion to be good. Moreover, Brebellious acts^ are essentially acts that pursue something that is not good and thus the agent is also committed to the badness of her pursuits. One could complain that even though the agent does not believe a manifest contradiction, the agent in this interpretation still displays irrationality, as she expresses 31 See for instance, Oddie (2005); Raz (2010). 32 This label is used in Wald & Tenenbaum (2018). 33 For versions of this view (or close relatives), see Schafer (2013); Tenenbaum (2006, 2007, 2008); Wald (2017). Kriegel (2018) argues that Brentano holds a version of this view as well. 34 I develop these ideas in more detail in Tenenbaum (2006, 2007, 2008, 2012). 35 For a similar proposal with respect to the relation between belief and truth (with some complications), see Shah and Velleman (2006).

11 The Guise of the Guise of the Bad 15 incompatible commitments. Any view committed at least to the rather weak claim that we ought to pursue the good (or what we judge to be good) or even the claim that we ought to pursue what we have good reason to pursue (or what we judge we have good reason to pursue) will have to say that perversion is a form of irrationality; they will be, on these views, a violation of a basic normative requirement. The problem for GG was the implausibility of attributing to Augustine or Satan a belief in a manifest contradiction; or, relatedly, the problem was that the GG made such characters impossible rather than irrational. The attitude version neither renders them impossible nor makes them theoretically irrational; instead, it explains their irrationality as conflict between the form and the content of their practical attitudes. 36 In the next section, I spell out some of the structure of the attitude view and how it treats other forms of irrationality, in particular, akrasia. I then try to clarify the attitude version account of perversion by examining whether similar forms of irrationality could be present in the theoretical realm. Other advocates of GG might argue that I am granting too much: perhaps we should deny that there are any plausible versions even of Weak Perversion. But I hope that the considerations below speak in favour of the attitude version even for those who are not impressed by Satan s evil nature or Augustine s youthful sins. 5 The Attitude Version and Perversion Let us look at the following instance of good theoretical reasoning 37 : DAISY & [BEL] Daisy is a duck & [BEL] I already knew Daisy in 1975 & [BEL] Ducks never live more than 30 years & [BEL] Daisy is dead It seems that truth plays a role in explaining why this is an instance of good reasoning. Although the details of the explanation are complex, the inference is warranted at least in part because the truth of the premises guarantee, are Bconducive to^, or evidence for, the truth of the conclusion. Of course, neither B[BEL]^ nor Btrue^ is part of the content of the attitude in question. But the fact that the attitude in question is one of belief is essential to understanding the reasoning in question as good reasoning. If I merely wished I had known Daisy in 1975, the inference to the conclusion would be a case of wishful thinking, and it would be a sign of a cruel heart, rather than of a logical mind, to move from A-C to the intention that Daisy is dead. A view that takes belief to 36 At least according to some interpretations of Kant, this would be the same form of incompatibility between form and content that Kant takes to be present in any instance of a choice of an evil maxim. See Engstrom (2009). 37 I am, of course, assuming the reasoning occurs after 2005, otherwise it would not be good reasoning.

12 16 S. Tenenbaum have a constitutive relation to truth is in a good position to explain why basing beliefs on other beliefs in this manner is rationally warranted. 38 If we took belief to be an attitude in which one holds the content true, we could even represent the reasoning in the following way 39 : DAISY* & [HOLDS TRUE] Daisy is a duck & [HOLDS TRUE] I already knew Daisy in 1975 & [HOLDS TRUE] Ducks never live more than 30 years & [HOLDS TRUE] Daisy is dead According to the attitude version of GG, the relation between good on the one hand, and desire, intention, and intentional action on the other hand, should be understood in a similar manner. In having these attitudes one conceives, or holds the content, to be good, and the validity and soundness of one s reasoning is partly determined by the relation between these attitudes and the good. 40 So if we go back to Bombe 1, we get the following inference: Bombe 1 * [DES] Eating delicious food [BEL] Eating the bombe is eating delicious food [INT] eating this bombe (as I am doing right now) Here the inference is directly from the contents of the desire and the belief to the content of the intention or intentional action. And similarly, as in the case of belief, the fact that desires and intentions are forms of conceiving or holding their objects to be good is supposed to explain the correctness of the reasoning. On the attitude version of GG, we can also present the reasoning as follows: Bombe 1 ** [HOLDS GOOD] Eating delicious food [HOLDS TRUE] Eating the bombe is eating delicious food [HOLDS GOOD] eating this bombe (as I am doing right now). 38 I am obviously ignoring a great number of complications. For one, the exact connection between truth and belief is itself the subject of much debate (I discuss this issue in more detail in Tenenbaum 2006, 2008, 2012.See also Wald 2017). Gilbert Harman, among others, has argued that the rules of reasoning are radically different from the laws of logic Harman (1986). But even in Harman s view the aim of truth is partly determinant of the rules of good reasoning. 39 Again, I am ignoring obvious complication such as whether one can make sense of the attitude of Bholding^ without thinking of it in terms of belief. For more on this issue, see the references in footnote 34, Wald (2017), and Shah and Velleman (2006). 40 This is not very different from the idea that the good practical inference is Breason-transmitting^.Infact,some philosophers argue for a Bguiseofreasons^ view instead of a guise of the good^ view. See Gregory (2013).

13 The Guise of the Guise of the Bad 17 Of course, desire and intention or acting intentionally must be different attitudes, so there must be a difference between them that is not presented in the reasoning. One possibility is to think that desire in which we hold something to be merely prima facie good, 41 but, for our purposes, which view we adopt on this matter is unimportant. The main question is whether such a view can represent perversion in a way that the content version of GG cannot. Since the attitude version is based on an analogy with theoretical reason, it might be worth asking if we can represent theoretical perversion. Let us start with the simpler case of theoretical akrasia. In analogy to how we represented practical akrasia, theoretical akrasia would look something like this: Roulette [BEL] The number 32 has not been rolled all night. [BEL] Roulette rolls are causally independent from each other [BEL] 32 is the most likely number to be rolled now Although the first premise could ground the conclusion in certain contexts, clearly the second premise defeats it as a potential reason for the conclusion. And insofar as you re aware of this fact, if you hold the conclusion on the basis of the first premise, you re being irrational. This seems to be exactly a case of akrasia: you form an attitude on the basis of what you know to be worse or insufficient grounds. Some philosophers have denied the possibility of theoretical akrasia. 42 But once we accept that we can hold beliefs that are not warranted by the evidence, it s notclearwhyitis conceptually impossible that you hold a belief on grounds that you know to be insufficient or to have been defeated. Moreover, just as ordinary experience seems to speak in favour of the possibility of practical akrasia, it seems also to give us numerous examples of theoretical akrasia. How many of us Bknock on wood^, act at times as if karma ruled the universe, prefer not to make confident assertions that one will be awarded a prize, or hesitate to bet on last week s lottery winning number? And, to give an example close to the heart of many practical philosophers, how many of us two-boxers in Newcomb s puzzle would grab both boxes without trepidation if we ever encountered such a situation in real life? 43 But just like the case of practical akrasia is not a case of practical perversion, theoretical akrasia is not the same as perversion, not even weak perversion. A case of perversion in the theoretical realm would have to be a case in which an agent holds a belief not only on what she recognizes to be weaker grounds, but partly because they re weaker grounds. Interesting enough, there are arguments against this possibility that parallel the case against perversion put forth by GG theorists. If belief aims at the truth, it would be impossible to form beliefs while being fully aware that one is forming them on grounds that are unrelated to their truth; in particular on any grounds that one knows not to be truth-conducive. Thus the possibility that one forms beliefs exactly because they lead us away from the truth has to be deemed conceptually impossible. 44 But 41 See Davidson (1980). My own view is that desiring is conceiving something to be good from a certain perspective, and that such inferences can be explained by means of some form of non-monotonic logic. See Tenenbaum (2007). 42 For instance Hurley (1992). 43 Empirical evidence suggests that we do engage in the kind of irrational inferences that two-boxers attribute to one-boxers. See the famous experiments in Quattrone and Tversky (1984) and the experiments in Fernbach et al. (2014) among others. 44 See Shah and Velleman (2006) on doxastic deliberation.

14 18 S. Tenenbaum could we have a form of weak perversion in the theoretical realm? It is not clear that anything in the nature of belief rules it out. Perhaps, we could have also a form of rebellion in which one tried to assert theoretical independence exactly by forming judgments in their own way, unconstrained by the accepted rules of inquiry. Various forms of rebellion against intellectual authority, from revolt against a supposedly enslaving conception of Bthe truth^ to antipathy towards experts who call on us to defer to them, or sheer Whitmanean exuberance, 45 might tempt us to express the Bunsubmissiveness^ of our power of judgment by rejecting accepted forms of inference. So we could have something like the following: Anti-Experts Experts say that Brexit would have awful consequences I remember being better off before Britain joined the EU It is rebellious to ignore the experts. Britain will be better off outside the EU This would fall short of Strong (theoretical) Perversion, as the ground for the belief (BI remember being better off before Britain joined the EU^) is not itself the falsity of the belief or even some ground that the agent recognizes not to be truth-conducive. However, it is a form of Weak Perversion given that blocking the defeating testimony of the experts in the above inference depends essentially on the fact that expert testimony is a well-established method of inquiry. Here too the agent does not believe a manifest contradiction, but the commitments expressed in these judgments manifest irrationality (assuming that we understand that an Bexpert^ on a subject is someone whose opinion is more likely to be correct than the opinion of a non-expert). 46 Of course, I do not want to engage here in a priori psychology of holding various kinds of Brebellious^ beliefs. Just as in the case of practical perversion, we might doubt that this is the best interpretation of any actual case of believing against evidence. But also just like in the case of practical perversion, it is not clear that we can rule out this possibility purely on conceptual grounds. 6 Conclusion Of course, many of the issues raised here deserve (and have received in the literature) more detailed discussion. As I said above, the aim was not to provide a full defence of GG, but to try to clarify the various ways in which we seem to pursue the bad and argue that none of them poses a threat to GG, or at least not to the attitude version of GG. But I hope this also helps us see some of the advantages of this version of GG. The attitude version makes fewer demands on the content of practical attitudes, and makes no demands on the theoretical attitudes accompanying them; desires or intentions need not have any explicitly evaluative content, and they need not be accompanied by any evaluative belief. For this reason, it allows for a wider array of (rational and irrational) attitudes that ground these practical attitudes, as well as 45 Following the famous lines of Walt Whitman s BSong to Myself^: BDo I contradict myself/very well then I contradict myself/(i am large, I contain multitudes).^ 46 Of course, since this particular inference involves probabilistic reasoning, the irrationality is more complex; we need at least further assumptions that there is further evidence defeating or undermining the testimony of the experts. I am certainly not arguing that BLeavers^ reasoned according to this pattern.

15 The Guise of the Guise of the Bad 19 a wider array of beliefs that can be held (rationally or irrationally) while holding these practical attitudes. This is the other side of what I take to be the main advantage of the attitude version: it allows us to maintain GG while accepting the all too plausible claim that in desiring and intending our mind is directed to the objects of our desires and intentions rather than to thoughts about the good. 47 Acknowledgements I would like to thank Jessica Moss, Jennifer Nagel and Sarah Paul, the audience and the participants of the conference BActing Under the Guise of the Bad^, and two anonymous referees for this journal for extremely helpful comments on earlier versions of this paper. References Anscombe GEM (2000) Intention, 2nd ed. Harvard University Press, Cambridge Aristotle (1999) Nicomachean ethics, 2nd ed. Hackett Pub. Co., Indianapolis Augustine (1961) Confessions, edited by Pine-Coffin RS. Penguin Books, Baltimore Austin JL (1964) A plea for excuses. In: Gustafson D (ed) Essays in philosophical psychology. Anchor Books, Garden City, pp 1 29 Baker D (2015) Akrasia and the problem of the Unity of reason. Ratio 28(1):65 80 Boswell, P (n.d.) Affective Content and the Guise of the Good, Universite de Montreal. Clark P (2010) Aspects, guises, species, and knowing something to be good. In: Tenenbaum S (ed) Desire, practical reason, and the good. Oxford University Press, Oxford, pp Davidson D (1980) How is weakness of the will possible? In: Davidson D (ed) Essays on actions and events. Oxford University Press, New York, pp Engstrom S (2009) The form of practical knowledge: a study of the categorical imperative. Harvard University Press, Harvard Fernbach PM, Hagmayer Y, Sloman SA (2014) Effort denial in self-deception. Organ Behav Hum Decis Process 123(1): Gregory A (2013) The guise of reasons. Am Philos Q 50(1):63 72 Harman G (1986) Change in view: principles of reasoning. MIT Press, Cambridge Horty JF (2007) Reasons as defaults. Philosophers Imprint 7(3):1 28 Horty JF (2012) Reasons as defaults. Oxford University Press, Oxford Hurley S (1992) Natural reasons: personality and polity. Oxford University Press, Oxford Kant I (1998) Religion within the boundaries of mere reason and other writings. Edited by wood a & di Giovanni G. Cambridge University Press, Cambridge Kriegel U (2018) Brentano s philosophical system: mind, being, value. Oxford University Press, Oxford (forthcoming) Lavin D, Boyle M (2010) Goodness and desire. In: Tenenbaum S (ed) Desire, practical reason, and the good. Oxford University Press, New York, pp MacDonald S (2003) Petit larceny, the beginning of all sin. Faith and. Philosophy 20: McDowell J (2010) What is the content of an intention in action? Ratio (new series), XXIII: Moore GE (1966) Principia Ethica. Cambridge University Press, Cambridge Oddie G (2005) Value, reality, and desire. Oxford University Press, New York Quattrone GA, Tversky A (1984) Causal versus diagnostic contingencies: on self-deception and on the voter s illusion. J Pers Soc Psychol 46(2): Raz J (2010) The guise of the good. In: Tenenbaum S (ed) Desire, practical reason, and the good. Oxford University Press, Oxford, pp Reinhold KL (2005) Letters on the Kantian philosophy. Cambridge University Press, Cambridge, Edited by Ameriks K Scanlon T (1998) What we owe to each other. Harvard University Press, Cambridge Schafer K (2013) Perception and the rational force of desire. J Philos 110(5): Setiya K (2007) Reasons without rationalism. Princeton University Press, Princeton Shah N, Velleman D (2006) Doxastic deliberation. Philos Rev 114(4): Stampe D (1987) The Authority of Desire. Philos Rev 96(3): See (Wald 2017) for further elaboration of this claim.