Forthcoming in Philosophical Issues: A Supplement to Noûs, 2022. TRUST AS PERFORMANCE J. Adam Carter University of Glasgow Abstract. It is argued that trust is a performative kind and that the evaluative normativity of trust is a special case of the evaluative normativity of performances generally. The view is shown to have advantages over competitor views, e.g., according to which good trusting is principally a matter of good believing (e.g., Hieronymi 2008; McMyler 2011), or good affect (e.g., Jones 1996; Baier 1986), or good conation (e.g., Holton 1994). Moreover, the view can be easily extended to explain good (and bad) distrust, where the latter is understood as aimed (narrow-scoped) forbearance from trusting. The overarching framework – which assimilates the evaluative norms of trusting (and distrusting) to performance-theoretic norms – supplies us with an entirely new lens to view traditional philosophical problems about what is involved in trusting and distrusting well and badly, and thus, places our capacity to make progress on problems in this area on a new footing 1. INTRODUCTION Trust is indispensable to the success of almost every kind of coordinated human activity, from politics and business to sport and scientific research. It is accordingly important that we know how to do it well – and how to avoid doing it badly. But the question of what it is to trust well is not easily separable from the question of what kind of thing trust is. And the matter of what kind of thing trust is — viz., whether it is a belief1 , or some non-doxastic attitude or stance (of one kind or another, perhaps affective 2 , perhaps conative 3 ) 1 See, e.g., Hardin (2002), Hieronymi (2008), and McMyler (2011). For a more moderate doxastic account, see Keren (2014; 2020). 2 Some notable examples here include Baier (1986) and Jones (1996). 3 See Holton (1994). 1 is divisive.4 It is also, as Jon Kvanvig (2016) puts it, somewhat ‘stupefying’ (2016, 8) given that the various distinct things trust has been identified with are also very different from each other. In response to considering the menu of disparate options about the nature of trust in the literature, Bernd Lahno (2004) writes, ‘any adequate theory of trust must include behavioral, cognitive and affective dimensions or aspects’ (2004, 30); by contrast, and more pessimistically, Thomas W. Simpson (2012) suggests that ‘There is a strong prima facie case for supposing that there is no single phenomenon that ‘trust’ refers to, nor that our folk concept has determinate rules of use’ (2012, 551). Unsurprisingly, given the disparity of opinion about the ontology of trust, what we find in the philosophy of trust are various incompatible (and starkly different) pictures of the evaluative normativity of trust – viz., of what the relevant standards are that good trusting, as such, should be expected to meet.5 Generally speaking, evaluative norms – unlike prescriptive norms, which prescribe conduct – regulate what it takes for a token of a particular type of thing to be good or bad with regard to its type, where the ‘goodness’ here is attributive in Geach’s (1956) sense – viz., the sense in which a sharp knife is a good knife, qua knife, regardless of whether it is good or bad simpliciter. (Likewise, in this sense, a known belief is a good belief, regardless of whether it would be good or bad simpliciter – viz., as it would be were the content of the knowledge instructions for igniting a terrible bomb.)6 Without a defensible picture of plausible evaluative norms for trusting, we’re poorly positioned to say when trusting is good or skilled as an instance of trusting ; we would be relegated – at least in our evaluative theorising – to seek out conditions under which trust is good or bad simpliciter – e.g., by investigating, like we might with anything else, when it paradigmatically leads to good (and bad) consequences, and if so what they happen to be.7 It should be no surprise that philosophers of trust 4 For a representative picture of this divisiveness, see, e.g., Simon (2020) and Faulkner and Simpson (2017). 5 For overviews, see Carter and Simion (2020, sec. 2), McLeod (2020, sec. 3); see also Faulkner (2010; 2014), Frost-Arnold (2014), Hieronymi (2008), and Hinchman (2020). 6 For a helpful overview of the prescriptive/evaluative norm distinction, with reference to attributive as opposed to predicative goodness, see McHugh (2012, 22) and, as this distinction applies to belief specifically, Simion, Kelp, and Ghijsen (2016, 384–386). 7 We would be in poor shape philosophically if we asked only how belief is good simpliciter and not what makes something good qua belief (then there would be very little epistemol- 2 have attempted to go further than this – i.e., further than exploring (e.g., as social psychologists have8 ) how trust might be good because, e.g., it enhances cooperation9 – typically by first taking some kind of stand on the ontology of trust – a stance on what kind of thing it is – which is then used as a kind of ‘blueprint’ for thinking about good trusting as such. According to doxastic accounts of trust, trust is essentially a kind of belief, a belief about the object of trust, e.g., that the trustee will take care of things as enstrusted.10 Good trusting, on simple doxastic accounts of trust on which belief of a particular sort is necessary and sufficient for trust, will just be a kind of good believing; that is, it will be an instance of the very thing – belief – whose attributive goodness it is always appropriate to assess by looking at its rationality, asking whether it’s true or known, whether it coheres with other justified beliefs, whether its production manifested epistemic virtues, etc.11 Rationality, reliability, truth, coherence, knowledge, etc., – paradigmatic evaluative norms of belief, as such – are (on these views) also norms that would regulate what counts as good trusting, in so far as trusting is believing.12 Moreover, if trusting is a kind of believing, then what the proponent of the doxastic account of trust tells us counts as good trusting will need to reflect any constraints on beliefs as such, including that we cannot bring them about via direct control. According to non-doxastic accounts of trust, it’s false that trust is a kind of belief, even if trusting sometimes accompanies belief. And by extension, it is false that some type of good believing is what it is that the evaluative normativity of trust should be thought of regulating. But here things complicate quickly. Some non-doxastic accounts of trust maintain that trust is essentially an affective attitude, or an emotion – some even hold that whatever affective attitude it is, it must not be accompanied with beogy!). The same goes for emotion; it is of philosophical importance what makes an emotion good or bad as such, not merely simpliciter. 8 See, for example, Berg, Dickhaut, and McCabe (1995) and Braynov and Sandholm (2002). For a philosophical incorporation of trust’s importance to decision-theoretic cooperative behaviour, see Faulkner (2011Ch. 1). 9 See Hardin (2002). Cf. Cook, Hardin, and Levi (2005Ch. 2). 10 For a recent and helpful overview of doxastic accounts of trust, see Keren (2020). 11 These are just some representative evaluative norms of belief. For related discussion, including of prescriptive norms of belief, see Simion, Kelp, and Ghijsen (2016), Whiting (2013), Benton (2014), McHugh (2012), Shah and Velleman (2005), and Gibbons (2013). For criticism, see Glüer and Wikforss (2009) and Papineau (2013). 12 As Karen Jones (1996, 2–5) caputres this idea, what we say about the nature of trust, viz., whether it is a belief, contrains what we say about the rationality of trust, given that belief already has clearly defined standards of rationality. See also Keren (2014) for related discussion. 3 lief. Other non-doxastic accounts of trust maintain that trust is a nondoxastic, non-affective conative attitude – e.g., a kind of moral stance – and so neither a belief nor a non-doxastic affective state. In light of the above, it looks very much like the endeavour of getting a clear grip on what good trusting involves threatens to fall into disarray; after all, the evaluative norms of belief bear little to no resemblance to the evaluative norms of, e.g., hoping, being optimistic, adopting moral stances, etc.13 But perhaps all is not as bad as it seems. Here is the plan for what follows. Section 2 demonstrates several insuperable problems for accounts of the evaluative normativity of trust that fall out of doxastic accounts of the nature of trust. Section 3 shows that different problems arise for attempting to extract an account of good trusting by looking at the attributive goodness of various non-doxastic attitudes (both affective and conative) that have been identified with trust. Rather than to simply try to select and then make do with the lesser of the known evils, I will – in Section 4 – suggest a way forward, one that involves the identification of trust as a performative kind. If we think of trust as a performative kind, we avoid the problems that face accounts of the evaluative normativity of trust that are restricted to theorising about good trusting as a species of good believing, good hoping, good emoting, good conation, etc. And this is the case even if trust sometimes or even usually involves combinations of these attitudes (both doxastic as well as non-doxastic) or stances. Finally the view that trust is a performative kind is shown to lend itself naturally to a plausible view of good and bad distrusting ; I conclude by showing how this application of the framework to distrust works and what the payoffs are. 2. GOOD TRUSTING AS GOOD BELIEVING: THE DOXASTIC ACCOUNT So is trust a kind of belief? Let’s sharpen this initial question in two ways; first, by bracketing two-place trust (i.e., X trusts Y) and focusing on threeplace-trust: schematically (S trusts X to 𝜙).14 Second, for simplicity, let’s 13 Consider, for example, some fundamental disanalogies that will bear on what the respective evaluative norms will look like. There is a disanalogy between belief, and emotions on the one hand, and the adoption of a moral stance, on the other, when it comes to voluntariness (e.g., Alston 1989). However, adopting a moral stance lines up with emotion and other affective attitudes when it comes to direction of fit (see, e.g., Humberstone 1992). 14 The distinction between two- and three-place trust was drawn initially by Horsburgh (1960). According to one popular way of thinking about relationship between two- and three-place trust, three-place trust is fundamental in the sense that two-place trust is explained in terms of it. For some representative examples of three-place fundamentalism, 4 consider just cases of interpersonal three-place-trust, which have been a central focal point in the philosophy of trust, and which involve one person trusting another person to – in a broad sense – take care of something, 𝜙, as entrusted15 , and which further involves (unlike in cases of mere reliance) subjecting oneself to the possibility of betrayal.16 In the specific case of testimonial trust – of special interest in social epistemology – when a hearer forms a testimony-based belief on a speaker’s say-so17 , the something she trusts the speaker for is the truth, or perhaps knowledge18 , of what she says. On a simple way of thinking of the relationship between testimonial trust and three-place interpersonal trust generally, the former is just an instance of the latter, an instance where ‘the truth’ is plugged in for 𝜙 in the schema, and which betraying the hearer’s trust involves misinforming her. With these caveats aside, let’s now consider the strong doxastic account of three-place interpersonal trust (hereafter, trust) according to which trust is essentially a belief, viz., a belief that the trustee will take care of things as entrusted. This kind of proposal – variations of which have been defended by Russell Hardin (2002), Pamela Hieronymi (2008), and Benjamin McMyler (2011) – is strong because it takes believing something about the object of trust to be type-identical with trusting. And there are some marks in its favour. For one thing, in paradigmatic cases of testimonial trust, the hearer trusts what the speaker says only if the hearer believes that the speaker has told the truth. And, more generally, as Hieronymi (2008) notes, if you entrust any kind of task to someone while believing they won’t do the thing, it seems you’re not really trusting them to do it.19 see, e.g., Baier (1986), Holton (1994), Jones (1996), Faulkner (2007), Hieronymi (2008), and Hawley (2014). Cf., Faulkner (2015). 15 The locution ‘as entrusted’ is meant to encompass views on which the trustee counts as taking care of things as entrusted only if doing so in a particular way, including, e.g., out of goodwill (Baier 1986; Jones 1996) or in conjunction with a belief that one is so committed (e.g., Hawley 2014). The present proposal – which is theoretically neutral on this point – is compatible with opting for either such kind of gloss. 16 As Annette Baier (1986) puts it, interpersonal trust involves subjecting oneself ‘necessarily to the limits’ of another’s goodwill (1986, 235), and in a way that differs from the kind of reliance we place in mere objects. For related discussion, see McMyler (2020, 129, 134). 17 For an overview of what qualifies as a testimony-based belief, see, e.g., Graham (2016, 172–173). 18 See, e.g., Kelp (2018) for a view friendly to this suggestion. 19 In support of this line of thought – viz., that one’s trust tracks one’s belief that the trustee will prove trustworthy – Hieronymi offers the following case-pair involving the betrayal of a secret. ‘SECRETS: Consider two cases. In one, I fully believe you are trustworthy; 5 Unfortunately, regardless of whatever else we might say for or against a strong doxastic account of trust, there are serious problems for the idea that trusting well is principally a matter of believing well, viz., of holding rational, reliable20 , true, coherent, etc., beliefs about the object of trust – and regardless of what features in the content of that belief (i.e., that the trustee will encapsulate the truster’s interests21 , or prove trustworthy out of goodwill22 , etc.) To appreciate this point, consider that any belief, as such, is better than it would be otherwise if it complies with the paradigmatic evaluative epistemic norms of belief, e.g., norms that hold that beliefs ought to be supported by evidence and known.23 However, (a) complying with a standard evidence norm24 (i.e., on which evidential support improves a belief’s quality) fails to improve trust for the reason that there is a constitutive tension between trusting and complying with evidential norms; and, second, (b) complying with the knowledge norm, specifically, undermines (or: moots) trust (rather than improves it) because it eliminates vulnerability. in the other, I have doubts about your trustworthiness, but, for other reasons (perhaps to build trust in our relationship, perhaps because I think friends should trust one another, or perhaps simply because I have no better alternative), I decide to tell you my secret. Suppose that, in both cases, you spill the beans, and that you do so in the same circumstances, for the same reasons (2008, 230).’ According to Hieronymi, once we hold fixed both (i) the ‘importance of the good entrusted’; and (ii) ‘the wrongness of the violation,’ then: ‘[…] it seems plausible that one’s degree of vulnerability to betrayal tracks one’s degree of trusting belief … further, this seems to be because, in the second case, there was less trust to betray (2008, 230–231).’ If the degree of one’s trust is, as Hieronymi thinks, positively correlated with the degree of one’s belief the trustee will prove trustworthy, then this counts in favour of the strong doxastic account, which would straightforwardly explain this correlation. 20 Of course, a norm of reliability will be an example of an ‘externalist’ norm on good trusting just as reliability is a paradigmatically externalist norm on good believing, in that that in virtue of which one satisfies the norm needn’t be reflectively accessible to the truster. See McLeod (2002, 91–100) for discussion. 21 (Hardin 2002). 22 (Jones 1996; Baier 1986). 23 Because we are discussing evaluative rather than prescriptive norms here, the ‘ought’ should be read as a kind of ‘ought to be’ rather than an ‘ought to do’ – viz., in the sense that a good knife ought to be sharp. For some useful discussion of the difference here, see, along with McHugh (2012, 22) and Simion, Kelp, and Ghijsen (2016, 384–386), and Schroeder (2011, 5–8) for evaluative norms featuring ‘ought’ claims specifically. 24 As Wanderer and Townsend (2013) put it, ‘No matter how the norms of Evidentialism are construed, trust invariably seems to stand in tension with them’ (2013, 7). See also Booth (2007). For a simple expression of an evaluative evidence norm on belief, take the following from Jonathan Adler (1999): ‘One’s believing that p is proper (i.e. in accord with the concept of belief) if and only if one’s evidence establishes that p is true’ (1999, 51). Alternatively, see Richard Feldman (2000): ‘When adopting (or maintaining) an attitude towards a proposition, p, a person maximises epistemic value by adopting (or maintaining) a rational attitude towards p’ (2000, 685). 6 On the first point, consider the following case due to Jeremy Wanderer and Leo Townsend (2013): PARANOID PARENT. A paranoid parent […] organises a babysitter for their child, and then proceeds to spend the evening out monitoring their babysitter’s antics remotely, via a ‘nanny-cam.’ The paranoid parent is not only a lousy date, but also a lousy trustor; in performing the seemingly rational act of broadening the evidential base relevant to her judgments of trustworthiness, she is, precisely, failing to trust the babysitter (2013, 1). The kind of belief that the proponent of a strong doxastic account identifies with trusting is such that the paranoid parent improves its quality qua belief – by strengthening the evidence basis for the belief – only by at the same time doing something that apparently undermines her trust. As Wanderer and Townsend put it, cases like PARANOID PARENT indicate that part of ‘what it is to trust’ is to refrain from complying with evidence norms on belief (2013, 2).25 So, you can’t by complying with such norms thereby trust better. Likewise, consider knowledge as a norm governing what counts as good belief – a position embraced by, e.g., ‘knowledge-firsters.’26 If the kind of belief that the proponent of a strong doxastic account identifies with trusting satisfies the knowledge norm, it arguably ceases thereby to qualify as trust. But this is not because of anything to do with trust’s relationship to evidence. Rather, it is to do with a constitutive tension between trusting and securing an outcome. If you have – put roughly – some kind of ‘guarantee’ that it is impossible for X to betray your trust, then as the thought goes, you are thereby no longer trusting them to do anything. This idea, viz., that trusting essentially involves subjecting oneself willingly to non-negligible vulnerability – at the very least, as Baier (1986, 25 This idea is sometimes captured in terms of a prima facie incompatibility between trusting and monitoring. As Baier (1986) vividly expresses the idea ‘Trust is a fragile plant […] which may not endure inspection of its roots, even when they were, before inspection, quite healthy’ (1986, 260). Belief, by contrast, not only withstands, but improves through inspection of its roots. 26 See, e.g., Williamson (2013, 5). A typical way that this view is defended by knowledgefirst philosophers involves two steps: first, there is a defence of the view that justification is the primary norm governing belief; and then, there is the further and crucial step that a involves a defence of the thesis that a belief is justified if and only if it is known. See Williamson (2016) and Sutton (2007) for representative statements of this idea. For an overview, see Benton (2014). 7 244), notes, to the limits of another’s goodwill, though also to the limits of her competence27 – is mostly uncontroversial.28 As Hardin (1992) summarises: As virtually all writers on the subject agree, trust involves giving discretion to another to affect one’s interests. This move is inherently subject to the risk that the other will abuse the power of discretion (1992, 507). But then if trust essentially involves rendering oneself vulnerable to betrayal, it is hard to see how – by coming to know that the trustee has taken care of things as entrusted – the trustor has improved any trust she might have had prior to acquiring this knowledge as opposed to having simply rendered her trust moot. Here is another problem for the idea that we can – as the proponent of the strong doxastic account must permit – profitably defend an account of what it is to trust well that is constrained by facts about what it is to believe well. The reasoning behind this second problem – call it the argument from voluntariness – goes as follows. Belief is never subject to arbitrary voluntary control. And that means that norms of believing never regulate what it takes for something subject to arbitrary control to be good or bad with regard to belief. If trust is a species of belief, then norms of good believing are always applicable to trust. Some cases of trust are subject to arbitrary voluntary control. So, norms of belief sometimes do not regulate trust. The idea that belief is not subject to arbitrary voluntary control is platitudinous, and is central to marking the difference between ‘belief’ and ‘make belief.’29 However, the idea that trust is, at least sometimes, subject 27 Perhaps also: to the trustee’s capacity to remain in conditions conducive to her cooperation, even if we hold fixed both good will and competence. For example, you might entrust someone to repay a debt. They are capable and willing, but fail to repay the debt due to an unexpected natural catastrophe. 28 For various expressions of this idea, see, along with Hardin (1992), e.g., Baier (1986, 244), McLeod (2020, sec. 1), Nickel and Vaesen (2012, 861–862), Carter (2020b, 2301, 2318–2319), Carter and Simion (2020, sec. 1a), Becker (1996, 45, 49), Dasgupta (1988, 67–68), Dormandy (2020, 241–242), Kirton (forthcoming), O’Neil (2017, 70–72), Potter (2020, 244), and Hinchman (2017). Cf., Pettit (1995, 208). 29 For a classic defence of this position, see Williams (1970). See also Scott-Kakures (1994) and, for an overview, Vitz (2008). It is worth noting, as Heil (1983, 355–356) points out, that we often use language to talk about belief – such as duty-based language – that seems to imply a kind of voluntariness. Such language is found in Descartes – who seems in the Fourth Meditation to speak of belief through the will (see Cottingham 2002, 352–355) but it is also used widely by contemporary writers (e.g., BonJour 1980, 60–61) who say (in various 8 to arbitrary voluntary control is something we could give up only on pain of failing to countenance therapeutic trust – viz., where one trusts (e.g., a teenager with no established track record of reliability) with the intended aim of bringing about (or increasing) trustworthiess.30 To the extent that therapeutic trust is voluntary in exactly the sense in which belief is not, as the worry goes, good believing does not provide us any kind of blueprint for good trusting.31 A third and final argument against the assimilation of good trusting to good believing focuses on cases of trusting through doxastic suspension. For example, suppose you see your friend holding a bloody knife, standing next to a body, after which you accuse your friend of the murder. Your friend, appreciating how overwhelming their guilt looks in light of the evidence, implores you to not rush to judgement until you hear the full story. They ask you to trust them – and wait until you hear the full explanation before drawing any conclusions whatsoever about what you’ve just seen. Suppose you do then trust your friend. In doing so you are explicitly not forming a belief about whether they will prove trustworthy in this case, nor are you forming a belief that they will or will not betray your trust in any way. You trust through doxastic suspension, such that the suspension from belief constitutes your trusting. Belief, as a kind of affirming, categorically precludes suspension.32 Thus, the trust you place in your ways) that affirmation about whether something is so without suitable ‘inspection’ of what one is affirming violates an epistemic duty, perhaps a duty to be epistemically responsible in one’s belief formation (Kornblith 1983, 34–37). It is a mistake though to think that this kind of talk implicates the idea Williams challenges of having direct arbitrary control over belief (though, cf., Vitz 2010). One helpful way to see why is to consider how judging whether something is so is both an intentional action but not subject to arbitrary control. For a detailed discussion on this point, see Sosa (2021, 32, 105 n. 59) and Sosa (2007, 88–91). 30 The first notable discussion of therapeutic trust as a species of trust is due to Horsburgh (1960, 5, 7–8, 12). See also Jones (2004, 5–7) and Frost-Arnold (2014, 1960–1963). 31 Therapeutic trust cases also raise another problem for strong doxastic accounts, which concerns the relationship between trust and expectation. As Peter Railton (2014) puts it, ‘Belief that p is a degree of confidence […] in a representation, p, that gives rise to and regulates a degree of expectation that things are or will be as p portrays them’ (2014, 145). Therapeutic trust, however, often does not involve any such expectation. What this means then, in addition to that the strong doxastic account is problematic in its own right, is that an account of good trusting can’t be an account of something that essentially involves an expectation. Granted, some proponents of doxastic accounts have sidestepped entirely issues to do with therapeutic trust by biting the bullet (Hieronymi 2008) and simply denying that therapeutic trust is genuine trust. See, however, Frost-Arnold (2014) for criticism of this strategy. 32 For discussion, see Turri (2012) and Carter (2018). For an extended treatment of forbearance, see Sosa (2021Ch. 3). 9 friend here isn’t something we could account for as good or bad trust in terms of norms governing belief, e.g., by asking if the belief counts as knowledge, or if it is rational. It’s looking like trusting well doesn’t line up very well with believing well. Is it of any use to weaken the account – such that trusting well might be understood as a function of satisfying at least (some) norm of belief along with perhaps some other norms? An initial hurdle, of course, is that we’ve already seen that there are reasons to think good belief (or any kind of belief for that matter) may not be necessary for trusting, well or otherwise. But, even if those problems could be overcome, we’d need to know exactly what those other norms are. After all, the good cracking of an egg, even if necessary for a good cake, provides us little clue for what the standard is for a good cake.33 Let’s now look at some non-doxastic norms, norms on trust whose motivation is sourced in very different, nondoxastic approaches to the ontology of trust.34 3. ALTERNATIVE NORMS ON GOOD TRUSTING: NON-DOXASTIC ACCOUNTS 3.1. Good trusting as good affect Non-doxastic accounts of the nature of trust embrace a negative and a positive thesis. The negative thesis, common to all non-doxastic accounts, is just the denial of the claim that belief (i.e., that the trustee will prove trustworthy) is central to trust of the threeplace interpersonal variety. What distinguishes non-doxastic accounts from each other is the positive theses they maintain about the nature of trust. Perhaps the most common non-doxastic proposal-type maintains that trust is, rather than a belief, an affective attitude.35 33 Granted, if belief is necessary for trust, then we will know that when a norm on good believing is not complied with when one trusts, then the quality of that trust is to that extent defective. This is information about the evaluative normativity of trust, but it is not a useful guide to what good trusting involves, unless we know in addition what other norms, beyond norms of belief, would need to be complied with to trust well. 34 Might it be worth one final push for the line that norms of good trusting are doxastic norms – by pointing out that (i) trusting testimony is a paradigmatic form of trust; and (ii) that trusting testimony is something you do if and only if you actually uptake what the speaker says? The answer here is ‘no.’ The reason is that it might just be that testimonial trust incidentally involves belief because of what testimony demands but that such belief isn’t essential to trust as such. Compare: trusting someone with a secret involves sharing the secret with them; but trust doesn’t essentially involve anything like this. We should thus be wary about generalising norms of trust, as such, from the testimonial case where believing the word of another is the standard shape that trusting takes. 35 For some representative defences of this kind of view, see de Sousa (1987), Calhoun (1984), Rorty (1980), and Lahno (2001). 10 On Karen Jones’s (1996) influential account, that affective attitude is optimism ‘that the goodwill and competence of another will extend to cover the domain of our interaction with her’ (1996, 4).36 For Lawrence Becker (1996), the relevant affective attitude is, instead, ‘a sense of security about other people’s benevolence, conscientiousness, and reciprocity’ (1996, 43). For Victoria McGeer (2008), it is a kind of ‘hope,’ for Guido Möllering (2001), a ‘leap of faith.’ Despite their differences, each of these affective attitude accounts of trust implies that any standard for good trust – a standard often captured by our talk of ‘justified trust’ – will take a different shape from the standards we expect good or justified belief to meet. Here’s McGeer (2008): The question of whether our trust can be justified, then, becomes a question of whether certain feelings towards others can be justified, which is not to say they can’t be, but rather that their justification conditions are different from, and perhaps not as stringent as, those on belief or on belief-based predictions of reliability (2008, 241). Likewise, as Jones (1996) puts it: we can be justified in trusting even when we would not be justified in predicting a favourable action on the part of the one trusted. Our evidence for trusting need not be as great as the evidence required for a corresponding justified prediction. In this respect trusting is more like hoping than like predicting (Jones 1996, 15, my italics). As we’ll see, there is a serious problem for the thought that good trusting is a species of good hope. And the reasons why this is so generalise to other positively valenced affective attitudes about others actions and intentions – including optimism and a sense of security–, such that we should be sceptical that good trusting is something we can encapsulate under the heading of good affect. So when is an instance of hope good hope? An initial reaction here might be as follows: “hope is good if it makes you better off with respect to getting the thing you’re hoping for.” This line of thought, however, faces 36 Jones clarifies that this kind of optimism she takes to be central to trust needn’t involve any tendency to – as optimism is often taken to imply one would do – ‘look on the bright side,’ given that, in the context of a very difficult joint task, you could trust someone, through your optimism directed at their goodwill, without any optimism about the success of the task (1996, 6). 11 a challenge in the form of Luc Bovens’ (1999) decision-theoretic ‘dominance argument’ against the value of hoping. Just suppose you that, for some projected state of the world, 𝜎 you have a choice between (i) hoping for it or (ii) not hoping for it. The projected state, 𝜎 will either come about or it will not. If it does not come about, then you would have been worse off having hoped than not having hoped, given that you will then be left with a greater sense of frustration after hoping than after not hoping. But suppose that 𝜎 does come about. Are you thereby better of having hoped? Perhaps, the contrary is the case. As Bovens puts it: […] is there anything to be gained from having hoped for it? In hoping for something, I tend to fill in the contours in the brightest colors. Suppose that my hopes come true, but not precisely in the bright colors that I had pictured. Had I not hoped for anything, I would have been delighted. But having hoped as I have, I experience a sense of frustration rather than satisfaction (1999, 670). But then the idea is that, whether the state of the world does or does not come about, I am always better off not having hoped for it rather than having hoped for it. Hence, by dominance, I should not hope. Let us zoom out for a moment to see why this line of reasoning looks like it poses a problem for the thought that good trusting is a kind of good hoping. An ex ante theoretical desideratum on any account of what good trusting involves is that the value of good trust isn’t ‘swamped’ by the value of trust. However, the dominance argument seems to show that the idea that good trusting is a kind of good hoping will inevitably fail to meet this desideratum. This is because, if the dominance argument goes through, it looks like the value of hoping will trivially swamp the value of good hoping, given that hoping will never be better than not hoping. Thus, as the thought goes, it is not plausible that good trusting is a kind of good hoping. But let’s not get carried away. Maybe there is a simple reply to the dominance argument, which is as follows: the dominance argument goes through only if we are assuming no causal dependency between states of the world and choices.37 But, as the reply continues, there often is such dependency. When you are tied up and locked in a room, a hopeful as opposed to a defeatist attitude might cause you to explore alternative 37 Bovens anticipates this line of reply in his (1999, sec. III.a). 12 ways for you to escape which you wouldn’t otherwise have considered, the exploring of which might then lead you to stumble upon a way that actually works. And, mutatis mutandis, for many more mundane circumstances where exploring additional alternatives is instrumentally valuable in achieving a hoped-for objective. Unfortunately, the above reply will be of little use for the proponent of ‘good trusting is a kind of good hoping.’ This is because, in the specific case of trust, there is not plausibly any such causal dependency between the relevant states of the world (whether, holding fixed that one has trusted, the trustee proves trustworthy) and whether one hopes that she has. And this is the case even if sometimes or even often, in cases of hoping, there is such a causal dependency. (Compare: your hoping might straightforwardly influence how you perform; but it won’t influence how your trustee performs). Granted, your hoping the trustee proves trustworthy might lead you to direct attention to your trustee’s performance qua trustee. But – and this is a point we’ve explored in Section 2 – to the extent you’re monitoring their performance, you’re not trusting them. In sum, then, it looks like the dominance argument remains a problem for the proponent of ‘good trusting is a kind of good hoping.’ Yet, here is a further card for such a proponent to play. “Granted, once you’ve trusted someone, then hoping they’ll prove trustworthy will never be a better strategy than refraining from hoping. However, hoping can itself lead to good trusting in the first place. Consider that without some hope, one might never trust at all, insofar as trusting involves incurring risk. In this respect, hoping isn’t ‘causally idle’ vis-à-vis the trustee’s proving trustworthy; it is an enabling condition38 for this result’s coming about.” The continuation of this line maintains that good trusting is good hoping when (and only when) one’s hoping enables successful trust moreso than it enables trust that is then betrayed. This is, at least, a prima facie coherent picture for how we might think of good trusting in terms of good hoping. But this picture quickly breaks down under scrutiny. For one thing, if the trust-relevant value of hoping is to be found outside of trusting itself – such that we locate its value prior to one’s trusting, as an enabling condition for that trust to have initially to come about, then identifying this value in hoping simply fails to qualify as a vindication of what good trusting is which adverts to good hoping. 38 For a notable articulation of enabling conditions, see Dancy (2004, 51, 64, 172). 13 The proponent of the idea that good trusting is a species of good hoping, then faces a dilemma. On the one hand, lies the dominance argument, which seems to imply that the value of mere hoping will trivially swamp the value of good hoping – a result at tension with the ex ante theoretical constraint on an account of good trusting which is that it exceeds the value of mere trusting. But, the dominance argument assumed no causal dependency between states of the world and hoping. While there is no such causal dependency in the special case of trusting, as hoping the trustee will prove trustworthy doesn’t increase the likelihood the trustee will in fact do so, there is a causal dependency in the sense that hoping can function as an enabling condition for incurring risk constitutive of trusting. But, and this is the other horn, if one sidesteps the dominance argument in this way, one is then giving a story about how hope has some trust-relevant value, but is no longer giving a vindication of what good trusting is which adverts to good hoping. The above dilemma looks difficult to overcome for a proponent of the view that good trusting is a species of good hoping. But – even setting the dilemma aside entirely – there remains a further reason why such a proposal is sure to face an uphill battle. This final reason has to do with an additional kind of ex ante constraint on good trusting – one that involves rationality and risk assessment. As we saw in Section 2, it is platitudinous that trusting essentially involves subjecting oneself to non-negligible risk of betrayal. This is platitudinous about trust in a way that, by parity of reasoning, the claim that knowledge is factive is platitudinous about knowledge. But given that trusting essentially involves incurring some risk, we should thereby expect that good trusting, as such, will be incompatible with poorly navigating these risks that, by trusting, one thereby incurs.39 But here is where, again, the prospects for assimilating good trusting to good hoping look dim, as hoping tends to make us worse at the sort of risk assessment good trusting plausibly demands of us. This is for two reasons, which are related. First, and as Bovens notes, the very act of hoping for something inclines us to a predictable error in reasoning, which is to ‘overestimate the subjective probability that the [hoped for] state of the world will come about’ (1999, 680). A well studied way in social psychology in which this kind of overestimation occurs is via the mechanisms of the availability heuristic.40 But perhaps even worse, rationally speaking, 39 For discussion on this point, see Carter (2020b) and Coleman (1990). a seminal discussion, see Tversky and Kahneman (1973). As studies by Vaughn (1999) suggest, the heuristic is strongest in cases where the outcome of an event is uncertain 40 For 14 is that hoping for an outcome has been demonstrated to encourage – as McGeer puts it – ‘superstitious ideas of our own agential powers’ such that we are led, via hoping, to overestimate the sense in which our hoping itself raises the likelihood that the hoped for event will come about – and this is two rational mistakes bundled into one. That hoping is, psychologically, a kind of invitation to misperceive the causal efficacy of our own agency (in connection with the hoped for event) is a common view in the psychology of hope (e.g., Charles R. Snyder et al. 1991), and it reveals an important way in which hoping of any sort stands to throw a spanner in our capacities for risk assessment (vis-à-vis the hoped for event) that good trusting can’t plausibly afford for us to compromise.41 This concludes the case for rejecting the idea that the evaluative normativity of trusting is going to line up with the evaluative normativity of any kind of hoping. Let’s now generalise. What goes for hope plausibly goes – mutatis mutandis – for affective states in the neighbourhood of hope, including faith and optimism directed at the object of trust. Just consider that an assimilation of good trusting to good faith or optimism will inevitably face both (i) variations on the dominance dilemma; as well as the (ii) norms of rationality of risk assessment objections. After all, regarding (i), it looks like the value of good faith and good optimism will trivially swamp the value of mere faith and mere optimism whenever the relevant states of the world about which one is optimistic or faithful are causally independent (as they necessarily are in the case of trust42 ) on one’s having that faith or being optimistic. Although faith and optimistic, like hope, can be efficacious in prompting one to then trust in the first place when one might not have, by pointing to this efficacy one is no longer characterising good trusting, as such, with reference to either of these attitudes that might prompt trust. Regarding (ii), it suffices to note – with reference to the availability heuristic – that faith and optimism, no less than hope, will tend to incline a truster to overestimate the – which will always be the case when one is trusting given that trusting essentially involves the incurring of some risk. 41 This is not to say, of course, that hoping can’t have beneficial practical consequences, apart from anything do do with the relationship between hoping and trusting. For a discussion of some of the benefits of hoping, see McGeer (2004) and, for psychological benefits specifically, C. Rick Snyder (1995). For the present purposes, the fact that hoping stands to imperil risk assessment in the case of trusting – when the probability of the trustee proving trustworthy is independent of whether we hope they prove trustworthy – does not bode well for any view on which good trusting, which will presumably preclude defective risk assessment, is meant to be a species of hope. 42 This is due to the constitutive tension between trusting and monitoring. For discussion, see Section 2. 15 likelihood the trustee will prove trustworthy, and in doing so, come into tension with the kind of good risk assessment that good trusting, as such, plausibly demands. In sum, the evaluative normativity of trusting simply does not line up with the evaluative normativity of any kind of positively valenced affect of the sort that trust has been identified in the literature. 3.2. Good trusting as good conation The standards that regulate what counts as good trusting must be other than those that regulate good belief or good affect. What about good conation – viz., goodness with respect to some motivational state or states that one has?43 Within the category of non-doxastic accounts of the nature of trust, affectbased theories like Jones’ are but one type of proposal. A different albeit prominent non-doxastic account is the ‘participant stance’ account, due to Richard Holton (1994), and which takes trust not to consist in the manifesting of any affective attitude, per se, but rather, in a kind of ‘normatively laden stance’ that implies a readiness to react in certain fitting ways to the trustee’s e.g., betrayal or cooperation44 ; as Holton puts it: […] you have a readiness to feel betrayal should it be disappointed, and gratitude should it be upheld (1994, 67). Is trusting perhaps a matter of doing this in a good way – viz., is it a matter of good readiness to feel (certain kinds of) reactive attitudes? We can envision at least two dimensions of conative quality here. One dimension concerns how ready one is to feel betrayal if trust is disappointed, gratitude should it be upheld. Along this dimension, presumably, the readier, the better. A separate dimension of quality concerns not the extent of the readiness, but what one is ready to feel. Here, the gold standard would seem to be a matter of fittingness: what one is ready to feel is betrayal (rather than something else) if and only if trust is betrayed, gratitude (rather than something else) if and only if trust is upheld. 43 For relevant discussion on conative states and (some representative views about) how they are taken to be motivating, see, e.g., Rosati (2016, sec. 3), Björklund et al. (2012), and Mele (2003). 44 A recent participant stance view that is difficult to taxonomise straightforwardly is due to Berislav Maruši฀ (2017), and which is strictly a kind of doxastic account, though one that incorporates elements of a participant-stance account, in the sense that the account maintains that trust is a belief held from the participant stance. From the perspective of assessing what it is to trust well, this kind of doxastic participant stance proposal will be committed to the position that norms of good trusting must incorporate norms of good believing. However, as I argued in Section 2, this commitment turns out to be problematic. 16 Bearing in mind these two dimensions of conative quality implied by the participant stance view – in short, ‘readiness quality’ and ‘fittingness quality’ – consider now the following case: THREE EASY MARKS: X, Y, and Z share a common flaw: deep-seated naivety. Too easily and often, each trusts unreliable websites, used car dealers, and people peddling get-rich-quick schemes. For simplicity, suppose all three are betrayed 90% of the time when they trust, and that they trust to the exact same extent – viz., none distrusts more than any other. But each trusts differently with reference to the two key quality dimensions of trusting that are implied by the participant stance view. X is consistently ready, when X trusts, to feel disappointment at perceived betrayal and gratitude at perceived trustee cooperation; however, X ’s perceptions are not well calibrated with reality. X too often misjudges when the trustee has in fact betrayed versus cooperated; consequently, X too often, though very readily, fits disappointment with cooperation, gratitude with betrayal. Put simply: X scores high in ‘readiness quality,’ poor in ‘fittingness quality.’ Y is in the opposite position. Y ’s perceptions (of betrayal and cooperation) are, unlike X ’s, well calibrated with reality; but Y is inconsistently ready, whenever Y trusts, to actually feel disappointment at betrayal when Y (accurately) perceives it and gratitude at cooperation when Y (accurately) perceives it. Put simply: Y scores high in fittingness quality, but low in readiness quality. Z shares, ex hypothesi, the common flaw of naivety with X and Y, however, Z scores as high as X in readiness quality and as high in Y in fittingness quality. Here are two observations about the THREE EASY MARKS case. First, all three are – in an obvious sense – bad trusters! All three, we are assuming, trust in ways that lead to betrayal more often than not.45 This, crucially, includes Z, whose overall conative score is impeccable – that is, Z does great, and clearly better than either X or Y, by those combined metrics that would seem to matter for good trusting on the participant stance view – viz., readiness to feel certain fitting attitudes in response to betrayal and cooperation on the part of the trustee. 45 Note that we are assuming the social-epistemic environment for trust here is a normal one and so is not unusually epistemically hostile. What best explains their poor reliability is thus not going to be any abnormal features of their environment. 17 The proponent of the participant stance view has a few moves available in reply, but none is promising. One move is to simply bite the bullet and say that Z represents good trusting in virtue of the dimensions of trusting quality that distinguish Z favourably from X and Y. This move, though, looks like a non-starter, given what we’ve already stipulated about Z ’s betrayal ratio. A more sophisticated move would be to insist that Z ’s goodness as a truster as represented by Z ’s admirable readiness to feel certain appropriate attitudes to the trustee, given different ways the trustee might behave, distinguish some important dimensions of good trusting, even if good conation doesn’t capture all good ways in which one might trust. I think we should regard this more sophisticated reply with some suspicion, however. The reason why can be put in terms of an additional ex ante desideratum we should expect any account of good trusting to satisfy: namely, that an explanation of what makes for good trusting can’t be orthogonal to the value of a trust’s being successful - i.e., the trustee’s taking care of things as entrusted. Here a brief analogy to epistemic norms will be of use. Evaluative norms of belief are obviously not orthogonal to successful belief – viz., true belief and knowledge. Consider, for example, the evaluative epistemic norms that aim to capture justified belief – e.g., these norms tell us that justified beliefs are ‘reliably produced’ beliefs; ‘beliefs that fit the evidence,’ etc. Both of these are, as Sanford Goldberg (2015) notes, ‘standards of success in connection with our pursuit of truth (and avoidance of error),’ or perhaps in connection with our pursuit of knowledge (and avoidance of ignorance).46 Put differently: it is because in believing we aim at truth and knowledge that the evaluative norms of belief capture (in different ways) standards of success in connection with these rather than some other aims. The norms are not orthogonal to, but rather importantly constrained by, what counts as successful attainment of the aim of the kind of attempt one makes by believing. But, as THREE EASY MARKS illustrates, the norms of good conation – of which the norms of good trusting will be a proper subset on the participant stance view – are entirely orthogonal to successful trusting; this is because the satisfaction of conative norms (i.e., readiness to feel certain attitudes in response to trusting outcomes) floats entirely freely of the aim, in trusting, that the trustee take care of things as entrusted. ‘Z ’ in our example case illustrates this, maximially satisfying conative norms 46 For a recent function-driven defence of this view, see Simion (2019). 18 while trusting in ways that rarely ever result in the attainment of the aim Z makes an attempt at attaining in trusting. 4. TRUST AS PERFORMANCE {- Good trusting is not something we can capture in terms of good believing, good affect, or good conation. So where do we go from here? Attempting to salvage any of these views, with some special pleading, does not look particularly attractive. Neither does opting for some kind of disjunctive proposal. But the good news is that we needn’t resort to such strategies. This is because there is a simple view that gets us everything we could want – and more – out of a view of the evaluative normativity of trust, and with none of the baggage that comes with any of the other views. Here is the key thesis I will defend and further develop in what follows: (†) Trust is a performative kind. The evaluative normativity of trust is a special case of the evaluative normativity of performances generally. Several key ideas here need some unpacking. In this section, I will: (a) briefly outline the normative structure of performance, giving special attention to the three central evaluative norms that apply to any performance type: success, competence, and aptness; (b) sketch and defend the thesis that trusting is a performance-type, and in doing so, characterise (with reference to (a)) the three central evaluative norms that apply to trusting: successful trust, competent trust, and apt trust; (c) show how the key thesis (†) satisfies key desiderata on any account of good trusting which other proposals canvased in Sections 2-3 (i.e., good trusting as good believing, good affect, good conation) failed to meet; (d) demonstrate how the idea that trust’s evaluative normativity is a species of performance normativity has the power to explain not only good trusting, but also good distrust, both deliberative and implicit. Let’s take these in turn. 19 4.1. Performance normativity A certain kind of normativity – performance normativity – is applicable to all performances that are attempts with aims. A simple example is the archer’s performance of shooting an arrow at a target. There are three central ways we can evaluative this performance, as an attempt.47 First, we can evaluate the attempt against the norm of success. An attempt is a ‘better’ attempt if it succeeds in attaining the aim internal to the kind of attempt it is, than if it fails – viz., a shot that hits the target is better than one that misses.48 Second, we can evaluate the attempt against the norm of competence. Regardless of whether an attempt actually succeeds in attaining its aim – that is, regardless of how the attempt stacks up against the success norm – the attempt is better if competent than if incompetent. A competent attempt – to a first approximation – will issue from a disposition to succeed (at attaining the aim internal to the performance-type) reliably enough when one tries in normal conditions. Third, an attempt is a better attempt if is not just successful and competent, but apt, viz., successful through competence rather than luck. These three evaluative norms – success, competence, and aptness – point to three distinct ways that any performance might be good. The ‘goodness’ here is attributive goodness; it applies to performances qua the kind of aimed attempt they are. The executioner’s skilled movements might be successful, competent as well as apt, while at the same time reprehensible. Finally: it is important to note that the formula ‘success + competence = aptness’ is incorrect. Suppose the archer’s shot is fired competently but – due to a freak gust of wind – is blown off target. But then, due to a second freak gust of wind, is blown back on course, so that the arrow lands in the bullseye. Here the shot is competent and successful, but not apt, because the success is not through competence but through luck.49 47 The key ideas of performance normativity originated from Ernest Sosa’s 2005 John Locke lectures, which were later published as A Virtue Epistemology (2007; 2009) and refined in Sosa (2015) and, in more recent work, redescribed as telic normativity in Sosa (2021). For overviews of recent work on performance normativity, see Kelp (2020) and Vargas (2016). For critiques and developments of performance normativity, see, e.g., Chrisman (2012), Kelp et al. (2017), and Carter (2020a). 48 See Sosa (2020) for the most recent presentation of the evaluative normativity of attempts as attempts. 49 Performances that are successful and competent but inapt have a ‘Gettier’ structure, where the success is disconnected from the good method used. For discussion, see Sosa (2007, Ch. 2; 2010a, 467, 474–475) and Greco (2009, 19–21; 2010, 73–76, 94–99). Cf., Pritchard (2012, 251, 264–268). 20 4.2. Trust as performance Consider a simple case of three-place interpersonal trust which is betrayed. Suppose you trust your friend with a secret, and you find out later that your friend spilled the beans. There is a clear since in which your trusting your friend with that secret did not succeed in attaining what it was at which, by trusting them with that secret, you thereby aimed. They did not – put generally – take care of things as entrusted. Their having done so would have involved, in this case, their not repeating what you had told them. Recast in the language of performance normativity: your trusting here didn’t do very well by the lights of the evaluative norm of success, and in a way that is broadly analogous to how an archer’s shot would be better if it hit the target than not, a belief better if it is accurate (true) rather than inaccurate (false). Accordingly: The Evaluative Success Norm of Trust (ESNT): S’s trusting X with 𝜙 is better if X takes care of 𝜙 as entrusted than if X does not. But just as missed shots and false beliefs can be competent despite failing to secure the relevant aim, likewise, trust can be competent even when it is not successful. When it is competent, it will derive from exercise of trusting skill 50 , which one has only if is disposed to trust successfully reliably enough when one trusts in proper shape and properly situated. Why is trusting skill indexed to ‘proper shape and situation?’ Compare: it does not count against your having the skill to drive a car if you would fail to perform reliably when attempting to drive if drugged and placed on slick roads. Likewise, it doesn’t count against your skill to trust well if normal bounds of – to a first approximation – risk, effort and skill (required of the trustee) are not present. In a bit more detail: it doesn’t count against someone’s having a skill to trust well if the truster would not trust successfully reliably enough in conditions where the (a) risk to the truster is excessively high and gains of betrayal are enormous; or where the level of (b) effort or (c) skill that would be required by the trustee to take care 50 Note that Sosa’s own terminology has shifted over the years when it comes to skill, in connection with shape and situation. For one thing, earlier discussions (e.g., 2010a, 465, 470) use the term ‘seat’ rather than ‘skill.’ For another, Sosa sometimes describes what one has, when one has the skill (i.e., what one retains even when in improper shape and while improperly situated) as an innermost competence (2015, 83), distinct from the complete competence one has when one’s skill is conjoined with proper shape and situation. By contrast, the term inner competence (2017, 191) is meant to pick out what one possesses when they possess both skill and shape, but not the situational element of the complete competence. For general discussion on these points, see Sosa (2015Ch. 3) and (2017, Ch. 12). 21 of things as entrusted is abnormally high.51 When a skilled truster is in proper shape and properly situated, the truster has the (complete) competence to trust well, and not merely the skill to do so. Trusting that issues (non-deviantly) from such a competence is good, qua trust, in an important respect – viz., the very same respect in which other kinds of competent performances are good (qua their performance type) even if it is one of those times where the performance does not succeed. This is implied by the more general performance-theoretic idea that a given attempt is better if it issues from a disposition to reliably attain its aim in normal conditions than otherwise. Thus: The Evaluative Competence Norm of Trust (ECNT): S’s trusting X with 𝜙 is better if S trusts X with 𝜙 competently than if S does not. Because (as noted in Section 4.1) any performance could be both successful and competent without being apt – which is of a higher quality qua performance than either successful or competent performance, or a conjunction of them – the same goes for trust. For example, suppose you competently trust a reliable colleague to pay back a loan on a particular date (say, 1 January). On 31 December, your colleague is in an accident which causes total amnesia. Struggling to regain memory, your colleague begins to remember who you are and then simply fabricates a specific memory (which luckily happens to be veridical) that they owe you money which must be repaid by 1 January. Because the friend is of a good and trustworthy character, they are on the basis of this fabricated but veridical memory moved to repay the loan, which they do. Your trusting them is thus successful, they have taken care of things as entrusted; the trust is also competent; but qua performance it nonetheless falls short in that it is not successful through competence, but successful just by dumb luck (i.e., that the trustee fabricated a veridical memory rather than a fictitious one).52 Thus, in addition to ESNT and ECNT, trust is also evaluable with respect to the following norm of aptness: The Evaluative Aptness Norm of Trust (EANT): S’s trusting X with 𝜙 is better if S trusts X with 𝜙 aptly than if S does not. 51 For related discussion on normal boundaries within which good trusting is valued, see Carter (2020b, sec. 6). 52 Another variation on this kind of case, with the same results, will appeal not to amnesia but to what Sven Bernecker (2010, 137–138) calls ‘trace creation’ and ‘trace implantation’ – where memory traces are created in vitro and implanted. While this is perhaps less plausible than amnesia – it makes for a cleaner case given that there is no worry that the amnesia would undermine one’s trustworthy character. 22 Apt trust is a kind of achievement, a success through competence. A common view in the axiology of achievements is that the value of an achievement does not reduce to the value of attaining the relevant success any old way (including through luck even when the attempt was a competent attempt – as in the ‘lucky success’ case above).53 And this idea is captured nicely by EANT, according to which the attributive goodness of apt trust asymmetrically entails the attributive goodness of successful, competent trust – just as we should expect it would.54 In sum, then, the key claims advanced thus far are that (i) trust is a performative kind; (ii) the evaluative normativity of trust is a special case of the evaluative normativity of performances generally; and (iii) ESNT, ECNT, and EANT capture three distinct evaluative norms against which any instance of (three-place interpersonal) trust can be evaluated as better or worse as an instance of trusting – with EANT representing a higher standard of good trusting than ESNT and ECNT. Let’s consider now (a) how this view accommodates desiderata competitor accounts could not in explaining good trusting; and (b) how it offers a further advantage, which is that it can be easily supplemented – by adding just a bit more to the picture outlined so far – in order to give us evaluative norms of distrust, which regulate what counts as a good instance of forbearing from trusting. 4.3. Taking stock Before adding anything further to the picture just developed, let’s see how it fares against the problems (from Sections 2 and 3) facing the competing views of good trusting surveyed. vs. the doxastic account There were problems with the idea – implied by doxastic accounts of trust – that the evaluative norms of trust are (a subset of) evaluative norms of belief. In short, we saw that there is a constitutive tension between trusting and complying with evidence and knowledge norms of belief. These problems are are not applicable to the performance-theoretic account, which does not assimilate good trusting to good believing in the first place. Moreover, the proposal is not committed – problematically, as the doxastic account is – to predicting that good trusting will be a function of doing something well involuntarily to 53 For discussion, see, e.g., Greco (2010Ch. 6), Sosa (2010b), Pritchard (2009a; 2009b), and Bradford (2013; 2015a; 2015b). 54 Put another way: the idea that apt trust is better than inapt trust implies that apt trust is better (qua trust) then either mere successful trust or mere competent trust, or (as the veridical memory case above suggests) mere successful and competent but inapt trust. 23 the same extent that belief is involuntary. Finally, the proposal is not challenged by cases – distinctively problematic for doxastic accounts – where good trust is achieved by suspending belief rather than by believing anything about the object of trust well. vs. the affective account One worry for the assimilation of good trusting to good affect was that such a proposal is in tension with a plausible ex ante constraint on an account of good trusting, which is that the value of good trusting should exceed, on any plausible account of what good trusting involves, the value of mere trusting. However, the dominance argument threatened to show – in tension with this constraint – that the value of mere hoping will trivially swamp the value of good hoping, and mutatis mutandis for other positive affect.55 The ‘trust as performance’ view does not succumb to the dominance argument, given that each of EANT, ESNT, and ECNT is better than mere trust that is neither apt, successful or competent – and the value of neither successful nor competent trust, nor the conjunction of the two, swamps the value of apt trust. Further, a problem for the would-be assimilation of good trusting to good affect is that we should expect that good trusting, as such, will be incompatible with poorly navigating risks that, by trusting, one thereby incurs. However, as we saw in Section 3.1, the prospects here aren’t promising, given what social psychology tells us about the kinds of rational mistakes that the very act of hoping or being optimistic vis-a-vis an outcome inclines us towards. On the performance-theoretic view, successful trust is of course compatible with poor risk assessment56 (in the sense that, analogously, e.g., hitting the bullseye and guessing correctly are compatible with using poor form and getting lucky), but – crucially – competent trust and apt trust are not. vs. the conative account As the THREE EASY MARKS case illustrated, the norms of good conation – of which the norms of good trusting will be a proper subset on the participant stance view – are entirely orthogonal to successful trusting, given that the satisfaction of conative norms (i.e., readiness to feel certain attitudes in response to trusting outcomes) floats 55 This argument assumed no causal dependency between states of the world and hoping. See Section 3 for further discussion. 56 This is because trust is successful so long as the trustee takes care of things as entrusted. Although there is an evaluative dimension of trust, as such, on which trust is better if successful than not (as captured by ESNT), it is false that successful trust implies good risk assessment at all. Bad trust can sometimes, luckily, result in successful trust, just as – by parity of reasoning – a wild guess will sometimes turn out to be true. 24 entirely freely of the aim, in trusting, that the trustee actually take care of things as entrusted. The performance-theoretic account, by contrast, takes that aim as the normative starting point – in that it is with reference to this aim that we understand not only the evaluative norm of successful trust, but also by extension competent and apt trust. An ecumenical advantage Question: “Surely apt trust, as well as very often successful and competent trust, will require good believing – including accurate and reasonable beliefs about the trustee’s good will and competence – as well as (often) combinations of affective and conative attitudes. How does the performance view countenance this observation?” Reply: Trusting will often, like other performances, require the good execution of other subsidiary activities. But the performance itself is evaluated with reference to the standards of success, competence and aptness, qua the performance type it is, and not with reference to the standards that regulate what make the subsidiary activities that are often necessary for good trust good as the kind of things they are.57 While the performance-theoretic account doesn’t make the mistake of assimilating good trusting to good 𝜙-ing for some 𝜙 (or any set of 𝜙-ings) the execution of which is among the subsidiary activities that good trusting usually involves, it does accommodate the data point that good trusting will plausibly often, as Bernd Lahno (2004) puts it, ‘include behavioral, cognitive and affective dimensions or aspects’ (2004, 30). Other proposals are comparatively more restricted in how this data point could be accommodated – as each predicts the goodness of good trusting will primarily be a matter of the goodness of one of these things but to the exclusion of others. In this respect, the performance-theoretical account can claim a kind of ecumenical edge. 5. EXTENDING THE VIEW: THE NORMATIVITY OF DISTRUST According to (EANT), S’s trusting X with 𝜙 is better if S trusts X with 𝜙 aptly than if S does not, and S’s trust is apt just in case it is successful because competent. 57 By way of analogy, suppose we sought an account of good promising (see, e.g., Heuer 2012a; 2012b). There are a range of things that good promising will typically involve. A good promiser will often have some subset of the following: (a) beliefs (i.e., about is being promised and to whom, about what means must be taken to fulfil the promise), (b) abilities (i.e., to make good on what is promised, and so to take the relevant means to doing so); (b) affect (i.e., to feel obligated to the promisee in light of having made promise); (c) conation (i.e., a readiness to feel, e.g., regret if the promise is not kept), and so on. 25 Let’s now take things a bit further. Consider that trust can be apt even when trust is implicit. We very often trust others with small things without ever consciously deliberating about whether to have trusted them in the first place; and we can do this better or worse – in ways that are successful, competent and apt. Compare: we believe many things are particular ways around us, implicitly, which guide action despite our having never attempted, through any conscious deliberation, to ‘settle the question’ for which a stance (i.e., on whether p – e.g., that the table is an arm’s length away) would constitute an answer.58 Some of these beliefs are apt, as they are when their correctness manifests competence, others aren’t. Within a performance-theoretic framework, the aim at which implicit trust constitutively attempts to secure is best understood as teleological, not intentional.59 And it is the teleological aim of implicit trust with reference to which we assess implicit trust for success, competence and aptness. No intentional aim is needed for such performance-theoretic assessment. By way of comparison, and as Sosa (2021, 25, fn. 12) notes, we can assess our implicit or ‘functional’ beliefs for success, competence and aptness – those that guide behaviour below the surface of conscious reflection – not because a thinker intentionally aims at anything, but just because teleologically our perceptual systems aim at correctly representing our surroundings. On many more substantial matters, however, our trust is not merely implicit but deliberative, and it is deliberative in a way that is broadly analogous to how some of our considered inquiries (i.e., inquiries in to ‘whether p’ questions) are deliberative. In such cases, we consciously consider whether to judge or suspend (intentionally omitting judgement). Likewise, when one faces a salient choice whether or not to trust someone X with something 𝜙 one deliberates on whether to trust X with 𝜙 or whether instead to forbear (intentionally omit trusting.) Let us continue this analogy further. Following Sosa (2015, Ch. 3; though 58 For an extended discussion on this point in connection with performance epistemology, see Sosa (2015Ch. 3, n. 5). 59 On this point, it is useful to consider Sosa’s (2015) remarks on what is needed for performative assessment as follows: ‘functional states can have teleological aims. Thus a state of alertness in a crouching cat may be aimed at detecting vulnerable prey. Whether as a state it can count as a “performance” in any ordinary sense is hence irrelevant to our focus on “performances” that have an aim and to which we may then apply our AAA aim-involving normative account. All that really matters for this latter is that the entity have a constitutive aim, whatever may be its ontological status or the label appropriately applicable to it in ordinary parlance’ (2015, Ch. 5, n. 5). For other discussions of functional and teleological assessment, see Sosa (2017, 71–72, 129–130, 152) and (2021, 24–31, 52–58, 64, 110, 118). 26 for recent updates see 2021; 2020), it is plausible that when we deliberately judge whether p, we intentionally aim not just at getting it right any way, but at getting it right aptly (for performance-theoretic virtue epistemologists: aiming to get it right aptly is tantamount to aiming at apt belief – knowledge 60 –, rather than at truth any old way).61 This is, by way of an athletic comparison, just as a basketball player62 aims by shooting not merely to make the shot any old way, but to make it competently, to make a well-selected shot.63 And, plausibly, mutatis mutandis, for deliberative – rather than mere implicit – trust: in deliberately trusting, we intentionally aim, in trusting, not merely at successful trust (like a basketball player chucking from half court, or an inquirer who aims at truth through a guess), but at aptness – viz., at apt trust. The relevant performance-theoretic analogies are thus: implicit belief aims (teleologically) at truth, deliberative belief (judgement) aims (intentionally) at apt belief (knowledge). Implicit trust aims (telelogically) at successful trust, deliberative trust aims (intentionally) at apt trust. Implicit belief, when apt, is knowledge; judgemental belief, when apt, is apt belief (knowledge) of a higher quality (i.e., fully apt belief, or knowing full well 64 ) – what results when one aptly attains the aim (aptness) of judgemental belief. Implicit trust, when apt, is apt trust; deliberative trust, when apt, is apt trust of a higher quality (i.e., fully apt trust) – when one aptly attains the aim (apt trust) of deliberative trust. Here is how all of the above connects with the question of good distrust. As a first point of note: the right way to characterise the way we aim at aptness, when we deliberate about whether to make the relevant attempt (or not) in the first place is in terms of a biconditional aim: to 𝜙 iff one’s 𝜙-ing would be apt. Just as we can actually make some attempt, X, in the endeavour to attain that biconditional aim, we can also forbear from X’ing in the endeavor to attain that very same biconditional aim – viz., to 𝜙 iff one’s 𝜙-ing would be apt. For example, the inquirer pursues the ‘positive’ Jamesian aim (attaining aptness) by (positively) affirming whether p in the endeavour to affirm if and only if doing so would be apt; but the inquirer can also con60 For a recent defence of the idea that inquiry is knowledge-aimed, within a wider knowledge-first virtue epistemology, see Kelp (Forthcoming). 61 Cf., Schechter (2019) for criticism. 62 Unless, of course, time is expiring. 63 The coach will berate a player who chucks it from half court, even if it goes in. For discussion, see Sosa (2015Ch. 3) and Carter (2016). 64 This term is coined in Sosa (2010c). 27 tribute to the biconditional aim by contributing to its subsidiary (negative Jamesian) aim, avoiding inaptness, by forbearing on whether p, and doing so also in the endeavour to affirm if and only if doing so would be apt.65 We are getting close now to seeing how distrust is itself is something we can evaluate for success, competence, and aptness – given that (put generally) forbearances, like the performances of which they are omissions, can be aimed. But first, there an ambiguity that needs addressed.66 The following locution <forbear from X’ing in the endeavor to attain aim A> is crucially ambiguous between a narrow-scope and a wide-scope reading. Narrow-scope forbearance should be read as: (Forbearing from X ’ing) in the endeavour to attain a given aim A. By contrast, wide-scope forbearance should be read as: forbearing from (X ’ing in the endeavor to attain a given aim A). Widescope forbearance from trusting is something akin to the Pyrhhonian analogue in the case of human cooperation as opposed to inquiry. It is a wide-scope abstaining from trusting simpliciter, thus, including from trusting in the endeavour to attain any aim. Call this Pyrrhonian mistrust, a omission from trust – though not an omission, in the endeavour to trust if and only if that trust would be apt, of trusting. By contrast, narrow-scope forbearance from trusting, but not wide-scope forbearance, is constitutively aimed forbearance from trusting.67 Take a simple case of three-place interpersonal narrow-scope forbearance from trusting: when I consciously deliberate whether to trust the stranger with my keys, and intentionally forbear, my forbearance is aimed; I forbear in the endeavour to, by forbearing, avoid trusting inaptly. Call this narrow65 See Sosa (2021Ch. 3) for a detailed discussion of this idea within telic virtue epistemol- 66 See Sosa (2020, 49) for discussion. ogy. 67 It is worth noting a connection between, on the one hand, the distinction between wide- scope/narrow-scope forbearance from trusting, and, on the other, Jane Friedman’s (2013) remarks on how to characterise a certain kind of agnosticism of interest in epistemology. As Friedman puts it: ‘[…] the sort of neutrality or indecision that is at the heart of agnosticism is not mere non-belief and can only be captured with an attitude. This means that the attitude will have to be one that represents (or expresses or just is) a subject’s neutrality or indecision with respect to the truth of some proposition. This will have to be either a sui generis attitude of indecision, or some other more familiar attitude’ (2013, 167). On the present proposal it is worth noting that whilst wide-scope forbearance from trusting, mere omission, needn’t involve any positive characteristics, narrow-scope forbearance – much like the kind of positive attitude of indecision that Friedman associates with agnosticism – will often have (some combinations of) attitudes, affect, and/or conation. In this respect, narrow-scope distrust is like both the agnosticism that Friedman describes, as well as trust itself (see Section 4.3 for discussion). 28 scope kind of aimed forbearance from trust deliberative distrust; it is (forbearing from trusting) in the endeavour to avoid inapt trust, and not forbearing from (trusting in the endeavour to avoid inapt trust). Deliberative distrust is subject to the evaluative norms of success, competence, and aptness. The success norm on deliberative distrust says: S’s (deliberative) distrusting X with 𝜙 is better if S’s forbearing from trusting X with 𝜙 avoids inaptness than if S doesn’t. (Deliberative distrust fails if, (i) one deliberatively distrusts X with 𝜙; and (ii) were one to have trusted (X with 𝜙), one’s trusting (X with 𝜙) would have been apt.) That would have been bad distrust along the success dimension. This is so even if that distrust was competent. The competence norm on deliberative distrust says: S’s (deliberative) distrusting X with 𝜙 is better if S forbears from trusting X with 𝜙 competently than if S doesn’t; this will require that S’s deliberative distrusting X with 𝜙 manifest S’s competence to (narrow-scope) forbear in ways that reliably lead to avoiding inaptness, when the truster is in proper shape and properly situated. The overly cynical person might forbear from trusting a reliable trustee successfully (and avoid inaptness given that the reliable would-be trustee would have failed to prove trustworthy on this occasion) but would still fail to do so competently. Finally, the aptness norm on deliberative distrust says: S’s deliberative distrusting X with 𝜙 is better if S deliberatively distrusts X with 𝜙 aptly than if S does not. Question: judgemental belief is to deliberative trust as implicit belief is to implicit trust. But what about this analogy: Deliberative trust is to implicit trust as deliberative forbearance is to ? The answer, of course, is implicit distrust, which rounds out our picture. Implicit distrust, like implicit trust, is teleological rather than intentionally aimed forbearance from trusting. Widescope forbearance from implicit trusting differs from Pyrrhonian mistrust (widescope forbearance from deliberative trust) not necessarily in respect of involving some different policy or stance, per se, but simply with respect to what kind of trust is, in fact, omitted.68 Narrow-scope implicit distrust is constitutively aimed, teleologically, not (as is narrow-scope deliberative distrust) at avoiding inaptness, but at avoiding unsuccessful trust – viz., trust where the trustee fails to take care of things as entrusted. As such, implicit distrust is evaluable as successful, competent and apt in connection with the aim of avoiding 68 Possibly, an individual could wide-scope forbear from one kind of trusting without widescope forbearing from the other. 29 unsuccessful trust: implicit distrust is better if successful (i.e., if not forbearing would have resulted in unsuccessful trust) than if not, competent than if not (i.e., if one’s forberance manifested a competence to not too easily forbear when, had one refrained from forbearing, one would have trusted successfully) than not, apt than if not. In sum, by modelling the evaluative normativity of distust performancetheoretically as norms of forbearance, we get the following picture: (i) first, widescope forbearance from trusting (Pyrrhonian mistrust and non-Pyrrhonian mistrust) is not performance-theoretically evaluable; (ii) narrow-scope forbearance from trusting (just like trusting itself) comes in two varieties, deliberative distrust and implicit distrust, each of which is performance-theoretically evaluable, for success, competence and aptness in connection with (a) the intentional aim of avoiding inaptness (in the case of deliberative mistrust); and (b) the functional/teleological aim of avoiding unsuccessful trust (in the case of implicit distrust). In order to make these distinctions in distrust quality more concrete, I’ll conclude with an example case illustrating each – beginning with the wide-scoped varieties of forbearance (i.e., from deliberative and implicit trust) that lie outside performance-theoretic evaluation. Widescope forbearance from trust (Pyrrhonian mistrust and Non-pyrrhonian mistrust) Pyrrhonian mistrust, i.e., forbearance from (deliberative trusting in the endeavour to trust iff one would do so aptly), is simply an omission from deliberative trusting. In the three-place case of interest, it is an omission from deliberatively trusting X with 𝜙. I’m using the term ‘Pyrrhonian’ only because what it omits is the taking of an intentional stance – (broadly) analogous to the kind of omission from intentional judgement that characterises a Pyrrhonian withdrawal from intentional judgement (viz., of whether something is so). What one omits is not necessarily implicit trust (though omissions from deliberative trust are compatible with omissions from implicit trust). For example featuring an omission from mere deliberative trust but not implicit trust, suppose a misanthrope’s plan is to withdraw from all cooperation; the plan is, however, flawed, as the misanthrope implicitly trusts a neighbour N with a small task, T, though she would not have done so had she deliberated. Even though the misanthrope omits deliberative trust in the case of N with T, she (despite herself) fails to omit implicit trust of N with T. While her implicit in this case is performance-theoretically assessable for success, competence and aptness, her non-Pyrrhonian mistrust, viz., her forbearing from (deliberative trust in the endeavour to attain apt trust) – is not. 30 Contrary to the situation of the misanthrope, who exhibits Pyrrhonian mistrust69 , we can imagine a recovering misanthrope in the inverse position: she wide-scope forbears from implicit trust but not from deliberative trust. Her misanthropic tendencies are so deeply ingrained, that she – assume ex hypothesi – is simply incapable of implicit trust. She trusts, if ever, only with a strong will to do so, to overcome these tendencies. Suppose, then that with such will overcoming her tendencies, she intentionally, deliberately, trusts her neighbour N with T. Omitted here is implicit trust, not deliberative trust. While her deliberative trust of N with T in this case is performance-theoretically assessable for success, competence and aptness, her non-Pyrrhonian mistrust is not. Narrow-scope intentionally aimed forbearance from trusting: Deliberative distrust (successful, competent and apt) Suppose A is deciding whether to trust to trust a friend, B, to watch A’s very young children (W ) for the weekend. The choice is complicated, but after deliberation on such things as risk of betrayed trust, difficulty of task, trustee goodwill and competence, etc., A decides not to trust B with W, and does so in the endeavour to trust if and only if that trust would be (not merely successful) but apt. This intentional forbearance from trusting succeeds even if B would have succeeded in taking care of the children for the weekend, and so long as B would very easily have failed. This is because, if B very easily would have failed, then had A trusted, she would not have avoided inapt trust, even though she would have attained successful trust. However, A’s deliberative distrust of B with W fails if were A to have trusted B with W, her trusting would have been apt. Likewise, A’s deliberative distrust of B with W is competent if that distrust reliably enough would have avoided inaptness, and apt iff A’s successful deliberative distrust of B with W manifests this competence. Narrow-scope functionally aimed forbearance from trusting: Implicit distrust (successful, competent and apt) One simple example case that can be used to illustrate narrow-scope functionally aimed forbearance from trusting involves bias-driven testimonial injustice.70 Suppose a member 69 I am using ‘mistrust’ in the case of widescope forbearance (compared to mere distrust) to indicate the wholesale character of the relevant omission from trusting – in that widescope cases omit not only trusting but also omit (omission in the endeavour to trust iff doing so would be apt). 70 See Fricker (2007) for the canonical presentation. It is worth noting that an alternative kind of example that might be used to illustrate this species of implicit distrust involves implicit distrust of experts. For discussion of this phenomenon, and some of the biases that 31 of a marginalised group, M, testifies that p in an open trial, in which A is a racist juror.71 Due to racism, A implicitly distrusts M ’s testimony, including M ’s assertion that p, despite never intentionally determining whether to trust p (any more than, say, A is reflecting on whether to believe that the testifier is speaking, something A believes implicitly). In addition, were A to deliberate about whether to trust p specifically, A would have been triggered to reflect on racism in a way that would have offset the racist and implicit tendency to forbear. In such a case, A’s implicit distrust is successful if, had A trusted M ’s testimony’ that p, A would have ended up believing falsely whether p. Likewise, the implicit distrust is not successful if, had A trusted M ’s testimony that p, A would have ended up believing truly whether p. Likewise, A’s implicit distrust of B with W is competent if that distrust reliably enough would have avoided unsuccessful trust, and apt iff A’s implicit distrust of M with p manifests this competence. If we assume, ex hypothesi, that M did testify truthfully that p in trial, A’s implicit distrust will be inapt because unsuccessful. Because racially driven implicit bias is unreliable72 , A’s implicit distrust of M whether p, will, in addition, fail to be competent. 6. CONCLUDING REMARKS Two main objectives have been pursued here, one negative, the other positive. The negative objective has been to demonstrate problems with ways of theorising about good trusting (viz., as good belief, good affect, and good conation) that would be predicted by various traditional approaches to the nature of trust. I’ve attempted against this background to reorient our theorising about good trusting, by defending – favourably against the competitor views criticised – the view that trust is a performative kind, and that the evaluative normativity of trust is a special case of the evaluative normativity of performances generally. A key aspect of the positive view is that there are three central evaluative norms that apply to trusting: successful trust, competent trust, and apt trust, which track three distinct ways that trust, as such, is better than it would be otherwise - and thus three distinct ‘zones’ of quality. A further ecumenical advantage of the proposal is that it can accommodate some of the more attractive features of competitor accounts, including that good trusting and distrusting very lead to it, see Baghramian and Croce (forthcoming, sec. 4). 71 Note that the details of this case are importantly different from a more familiar case of a racist juror in social epistemology, due to Lackey (2007, 598), the latter of which is meant to illustrate the phenomenon of ‘selfless assertion.’ 72 See, for instance, Saul (2017), Munroe (2016), and Díaz and Almagro (2019). 32 often require (combinations of) belief, affect, and conation. In the final section, I demonstrated how the ‘trust as performance’ view, with some additional elaborations, has the power to explain not only good (and bad) trusting, but also good (and bad) distrust, where the latter is understood as aimed (narrow-scoped) forbearance from trusting. The resulting picture – a framework that assimilates the evaluative norms of trusting (and distrusting) to performance-theoretic norms – supplies us with an entirely new lens to view traditional philosophical problems about what is involved in trusting and distrusting well and badly, and thus, places our capacity to make progress on problems in this area on a new footing.73 73 This paper was written as part of the Leverhulme-funded ‘A Virtue Epistemology of Trust’ (#RPG-2019-302) project, which is hosted by the University of Glasgow’s COGITO Epistemology Research Centre, and I’m grateful to the Leverhulme Trust for supporting this research. 33 REFERENCES Adler, Jonathan E. 1999. “The Ethics of Belief: Off the Wrong Track.” Midwest Studies in Philosophy 23: 267–285. Alston, William P. 1989. Epistemic Justification: Essays in the Theory of Knowledge. Cornell University Press. Baghramian, Maria, and Michel Croce. forthcoming. “Experts, Public Policy and the Question of Trust.” In Routledge Handbook of Political Epistemology, edited by Michael Hannon and Jeroen De Ridder. London, UK: Routledge. Baier, Annette. 1986. “Trust and Antitrust.” Ethics 96 (2): 231–260. Becker, Lawrence C. 1996. “Trust as Noncognitive Security about Motives.” Ethics 107 (1). University of Chicago Press: 43–61. Benton, Matthew A. 2014. “Knowledge Norms.” Internet Encyclopedia of Philosophy. Berg, Joyce, John Dickhaut, and Kevin McCabe. 1995. “Trust, Reciprocity, and Social History.” Games and Economic Behavior 10 (1): 122–142. doi:10.1006/game.1995.1027. Bernecker, Sven. 2010. Memory: A Philosophical Study. Oxford University Press. Björklund, Fredrik, Gunnar Björnsson, John Eriksson, Ragnar Francén Olinder, and Caj Strandberg. 2012. “Recent Work on Motivational Internalism.” Analysis 72 (1): 124–137. doi:10.1093/analys/anr118. BonJour, Laurence. 1980. “Externalist Theories of Empirical Knowledge.” Midwest Studies in Philosophy 5: 53–73. Booth, Anthony Robert. 2007. “The Two Faces of Evidentialism.” Erkenntnis 67 (3). Springer: 401–417. Bovens, Luc. 1999. “The Value of Hope.” Philosophy and Phenomenological Research 59 (3): 667–681. Bradford, Gwen. 2013. “The Value of Achievements.” Pacific Philosophical Quarterly 94 (2). Wiley Online Library: 204–224. Bradford, Gwen. 2015a. Achievement. Oxford University Press, USA. Bradford, Gwen. 2015b. “Knowledge, Achievement, and Manifestation.” Erkenntnis 80 (1). Springer Netherlands: 97–116. doi:10.1007/s10670014-9614-0. 34 Braynov, Sviatoslav, and Tuomas Sandholm. 2002. “Contracting With Uncertain Level Of Trust.” Computational Intelligence 18 (4): 501–514. doi:10.1111/1467-8640.00200. Calhoun, Cheshire. 1984. “Cognitive Emotions?” What Is an Emotion, 327– 342. Carter, J. Adam. 2016. “Sosa on Knowledge, Judgment and Guessing.” Synthese, August. Carter, J. Adam. 2018. “On Behalf of Controversial View Agnosticism.” European Journal of Philosophy 26 (4): 1358–1370. doi:10.1111/ejop.12333. Carter, J. Adam. 2020a. “De Minimis Normativism: A New Theory of Full Aptness.” The Philosophical Quarterly. Carter, J. Adam. 2020b. “On Behalf of a Bi-Level Account of Trust.” Philosophical Studies 177 (8): 2299–2322. doi:10.1007/s11098-019-01311-2. Carter, J. Adam, and Mona Simion. 2020. “The Ethics and Epistemology of Trust.” Internet Encyclopedia of Philosophy. Internet Encyclopedia of Philosophy. Chrisman, Matthew. 2012. “The Normative Evaluation of Belief and the Aspectual Classification of Belief and Knowledge Attributions.” The Journal of Philosophy 109 (10): 588–612. Coleman, James S. 1990. “Relations of Trust.” Foundations of Social Theory, Cambridge, London, 91–116. Cook, Karen S., Russell Hardin, and Margaret Levi. 2005. Cooperation Without Trust? Russell Sage Foundation. Cottingham, John. 2002. “Descartes and the Voluntariness of Belief.” The Monist 85 (3): 343–360. Dancy, Jonathan. 2004. Ethics Without Principles. Oxford University Press on Demand. Dasgupta, Partha. 1988. “Trust as a Commodity.” In Trust: Making and Breaking Cooperative Relations, edited by Diego Gambetta, 49–72. Blackwell. de Sousa, Ronald. 1987. The Rationality of Emotion. MIT Press. Díaz, Rodrigo, and Manuel Almagro. 2019. “You Are Just Being Emotional! Testimonial Injustice and Folk-Psychological Attributions.” Synthese. Springer, 1–22. Dormandy, Katherine. 2020. “Exploitative Epistemic Trust.” In Trust in Epistemology, edited by Katherine Dormandy, 241–264. 35 Faulkner, Paul. 2007. “A Genealogy of Trust.” Episteme: A Journal of Social Epistemology 4 (3). Edinburgh University Press: 305–321. Faulkner, Paul. 2010. “Norms of Trust.” In Social Epistemology, edited by Adrian Haddock, Alan Millar, and Duncan Pritchard. Oxford University Press. Faulkner, Paul. 2011. Knowledge on Trust. Oxford University Press. Faulkner, Paul. 2014. “The Moral Obligations of Trust.” Philosophical Explorations 17 (3). Taylor & Francis: 332–345. Faulkner, Paul. 2015. “The Attitude of Trust Is Basic.” Analysis 75 (3): 424– 429. doi:10.1093/analys/anv037. Faulkner, Paul, and Thomas Simpson. 2017. The Philosophy of Trust. Oxford University Press. Feldman, Richard. 2000. “The Ethics of Belief.” Philosophy and Phenomenological Research 60 (3). JSTOR: 667–695. Fricker, Miranda. 2007. Epistemic Injustice: Power and the Ethics of Knowing. Oxford University Press. Friedman, Jane. 2013. “Suspended Judgment.” Philosophical Studies 162 (2). Springer: 165–181. Frost-Arnold, Karen. 2014. “The Cognitive Attitude of Rational Trust.” Synthese 191 (9). Springer Verlag. doi:10.1007/s11229-012-0151-6. Geach, Peter T. 1956. “Good and Evil.” Analysis 17 (2). JSTOR: 33–42. Gibbons, John. 2013. The Norm of Belief. Oxford University Press. Glüer, Kathrin, and \AAsa Wikforss. 2009. “Against Content Normativity.” Mind 118 (469). Oxford University Press: 31–70. doi:10.1093/mind/fzn154. Goldberg, Sanford. 2015. “What Is the Subject-Matter of the Theory of Epistemic Justification?” Epistemic Evaluation: Purposeful Epistemology. Oxford University Press, 205–223. Graham, Peter J. 2016. “Testimonial Knowledge: A Unified Account.” Philosophical Issues 26 (1). Wiley Online Library: 172–186. Greco, John. 2009. “Knowledge and Success from Ability.” Philosophical Studies 142 (1). Springer: 17–26. Greco, John. 2010. Achieving Knowledge: A Virtue-Theoretic Account of Epistemic Normativity. Cambridge University Press. 36 Hardin, Russell. 1992. “The Street-Level Epistemology of Trust.” Analyse & Kritik 14 (2). De Gruyter Oldenbourg: 152–176. Hardin, Russell. 2002. Trust and Trustworthiness. Russell Sage Foundation. Hawley, Katherine. 2014. “Trust, Distrust and Commitment.” Noûs 48 (1). JSTOR: 1–20. Heil, John. 1983. “Doxastic Agency.” Philosophical Studies 43 (3). Springer: 355–364. Heuer, Ulrike. 2012a. “Promising-Part 1.” Philosophy Compass 7 (12). WileyBlackwell: 832–841. doi:10.1111/j.1747-9991.2012.00524.x. Heuer, Ulrike. 2012b. “Promising - Part 2.” Philosophy Compass 7 (12). WileyBlackwell: 842–851. doi:10.1111/j.1747-9991.2012.00523.x. Hieronymi, Pamela. 2008. “The Reasons of Trust.” Australasian Journal of Philosophy 86 (2). Taylor & Francis: 213–236. Hinchman, Edward. 2017. “On the Risks of Resting Assured: An Assurance Theory of Trust.” In The Philosophy of Trust, edited by Paul Faulkner and Thomas W. Simpson. Oxford: Oxford University Press. Hinchman, Edward. 2020. “Trust and Will.” In Routledge Handbook on Trust and Philosophy, edited by Judith Simon. New York: Routledge. Holton, Richard. 1994. “Deciding to Trust, Coming to Believe.” Australasian Journal of Philosophy 72 (1). Taylor & Francis: 63–76. Horsburgh, H. J. N. 1960. “The Ethics of Trust.” The Philosophical Quarterly 10 (41). Blackwell Publishers Ltd Oxford, UK and Boston, USA: 343–354. Humberstone, I. Lloyd. 1992. “Direction of Fit.” Mind 101 (401). JSTOR: 59–83. Jones, Karen. 1996. “Trust as an Affective Attitude.” Ethics 107 (1). University of Chicago Press: 4–25. Jones, Karen. 2004. “Trust and Terror.” In Moral Psychology: Feminist Ethics and Social Theory, edited by Peggy DesAutels and Margaret Urban Walker, 3–18. Rowman & Littlefield. Kelp, Christoph. Forthcoming. Inquiry, Knowledge, and Understanding. Oxford: Oxford University Press. Kelp, Christoph. 2018. “Assertion: A Function First Account.” Noûs 52 (2). Wiley Online Library: 411–442. Kelp, Christoph. 2020. “The Epistemology of Ernest Sosa: An Introduction.” Synthese. Springer, 1–8. 37 Kelp, Christoph, Cameron Boult, Fernando Broncano-Berrocal, Paul Dimmock, Harmen Ghijsen, and Mona Simion. 2017. “Hoops and Barns: A New Dilemma for Sosa.” Synthese. Springer, 1–16. Keren, Arnon. 2014. “Trust and Belief: A Preemptive Reasons Account.” Synthese 191 (12). Springer: 2593–2615. Keren, Arnon. 2020. “Trust and Belief.” In The Routledge Handbook of Trust and Philosophy, edited by Judith Simon, 109–120. Kirton, Andrew. forthcoming. “Matters of Trust as Matters of Attachment Security.” International Journal of Philosophical Studies, forthcoming, 1–20. Kornblith, Hilary. 1983. “Justified Belief and Epistemically Responsible Action.” The Philosophical Review 92 (1). JSTOR: 33–48. Kvanvig, Jonathan L. 2016. “The Idea of Faith as Trust.” Reason and Faith: Themes from Richard Swinburne. Oxford University Press Oxford, 4–26. Lackey, Jennifer. 2007. “Norms of Assertion.” Noûs 41 (4): 594–626. Lahno, Bernd. 2001. “On the Emotional Character of Trust.” Ethical Theory and Moral Practice 4 (2). Springer Netherlands: 171–189. doi:10.1023/A:1011425102875. Lahno, Bernd. 2004. “Three Aspects of Interpersonal Trust.” Analyse & Kritik 26 (1). De Gruyter Oldenbourg: 30–47. Maruši฀ , Berislav. 2017. “Trust, Reliance and the Participant Stance.” Philosopher’s Imprint 17 (17). Ann Arbor, MI: University of Michigan. http://hdl.handle.net/2027/spo.3521354.0017.017. McGeer, Victoria. 2004. “The Art of Good Hope.” The Annals of the American Academy of Political and Social Science 592 (1). Sage Publications: 100–127. McGeer, Victoria. 2008. “Trust, Hope and Empowerment.” Australasian Journal of Philosophy 86 (2). Routledge: 237–254. doi:10.1080/00048400801886413. McHugh, Conor. 2012. “The Truth Norm of Belief.” Pacific Philosophical Quarterly 93 (1). Wiley Online Library: 8–30. McLeod, Carolyn. 2002. Self-Trust and Reproductive Autonomy. MIT Press. McLeod, Carolyn. 2020. “Trust.” In The Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta, Fall 2020. Metaphysics Research Lab, Stanford University. McMyler, Benjamin. 2011. Testimony, Trust, and Authority. OUP USA. McMyler, Benjamin. 2020. “Trust and Authority.” The Routledge Handbook of Trust and Philosophy. Routledge. 38 Mele, Alfred R. 2003. Motivation and Agency. Oxford University Press. Möllering, Guido. 2001. “The Nature of Trust: From Georg Simmel to a Theory of Expectation, Interpretation and Suspension.” Sociology 35 (2). Cambridge University Press: 403–420. Munroe, Wade. 2016. “Testimonial Injustice and Prescriptive Credibility Deficits.” Canadian Journal of Philosophy 46 (6). Cambridge University Press: 924–947. Nickel, Philip J., and Krist Vaesen. 2012. “Risk and Trust.” In Handbook of Risk Theory, edited by Sabine Roeser, Rafaela Hillerbrand, Martin Peterson, and Per Sandin. Springer. O’Neil, Collin. 2017. “Betraying Trust.” In The Philosophy of Trust, edited by Paul Faulkner and Thomas W. Simpson, 70–89. Oxford, UK: Oxford University Press. Papineau, David. 2013. “There Are No Norms of Belief.” In The Aim of Belief, edited by Timothy Chan. Oxford University Press. Pettit, Philip. 1995. “The Cunning of Trust.” Philosophy and Public Affairs 24 (3). Wiley-Blackwell: 202–225. Potter, Nancy Nyquist. 2020. “Interpersonal Trust.” In The Routledge Handbook of Trust and Philosophy, edited by Judith Simon, 243–255. Routledge. Pritchard, Duncan. 2009b. “Knowledge, Understanding and Epistemic Value.” Royal Institute of Philosophy Supplement 64. Cambridge: Cambridge University Press: 19–43. doi:10.1017/s1358246109000046. Pritchard, Duncan. 2009a. “The Value of Knowledge.” The Harvard Review of Philosophy 16 (1). Equinox Publishing Ltd.: 86–103. doi:10.5840/harvardreview20091616. Pritchard, Duncan. 2012. “Anti-Luck Virtue Epistemology.” The Journal of Philosophy 109 (3): 247–279. Railton, Peter. 2014. “Reliance, Trust, and Belief.” Inquiry 57 (1). Routledge: 122–150. Rorty, Amélie Oksenberg. 1980. Explaining Emotions. Vol. 232. Univ of California Press. Rosati, Connie S. 2016. “Moral Motivation.” In The Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta, Winter 2016. Metaphysics Research Lab, Stanford University. https://plato.stanford.edu/archives/win2016/ entries/moral-motivation/. Saul, Jennifer. 2017. “Implicit Bias, Stereotype Threat, and Epistemic Injustice.” The Routledge Handbook of Epistemic Injustice. Routledge, 235–242. 39 Schechter, Joshua. 2019. “Aiming at Aptness.” Episteme 16 (4). Cambridge University Press: 438–452. Schroeder, Mark. 2011. “Ought, Agents, and Actions.” The Philosophical Review 120 (1). Duke University Press: 1–41. doi:10.1215/00318108-2010017. Scott-Kakures, Dion. 1994. “On Belief and the Captivity of the Will.” Philosophy and Phenomenological Research 54 (1): 77–103. Shah, Nishi, and J. David Velleman. 2005. “Doxastic Deliberation.” Philosophical Review 114 (4). Duke University Press: 497–534. doi:10.1215/00318108-114-4-497. Simion, Mona. 2019. “Knowledge-First Functionalism.” Philosophical Issues 29 (1). Wiley Online Library: 254–267. Simion, Mona, Christoph Kelp, and Harmen Ghijsen. 2016. “Norms of Belief.” Philosophical Issues 26 (1). Wiley Online Library: 374–392. Simon, Judith. 2020. The Routledge Handbook of Trust and Philosophy. Routledge. doi:10.4324/9781315542294. Simpson, Thomas W. 2012. “What Is Trust?” Pacific Philosophical Quarterly 93 (4). Wiley Online Library: 550–569. Snyder, C. Rick. 1995. “Conceptualizing, Measuring, and Nurturing Hope.” Journal of Counseling & Development 73 (3). Wiley Online Library: 355–360. Snyder, Charles R., Cheri Harris, John R. Anderson, Sharon A. Holleran, Lori M. Irving, Sandra T. Sigmon, Lauren Yoshinobu, June Gibb, Charyle Langelle, and Pat Harney. 1991. “The Will and the Ways: Development and Validation of an Individual-Differences Measure of Hope.” Journal of Personality and Social Psychology 60 (4). American Psychological Association: 570. Sosa, Ernest. 2007. A Virtue Epistemology: Apt Belief and Reflective Knowledge, Volume I. Oxford University Press. Sosa, Ernest. 2009. Reflective Knowledge: Apt Belief and Reflective Knowledge, Volume II. Oxford University Press. Sosa, Ernest. 2010a. “How Competence Matters in Epistemology.” Philosophical Perspectives 24 (1): 465–475. Sosa, Ernest. 2010c. Knowing Full Well. Princeton University Press. Sosa, Ernest. 2010b. “Value Matters in Epistemology.” The Journal of Philosophy 107 (4). JSTOR: 167–190. Sosa, Ernest. 2015. Judgment & Agency. Oxford University Press UK. 40 Sosa, Ernest. 2017. Epistemology. Princeton: Princeton University Press. Sosa, Ernest. 2020. “Default Assumptions and Pure Thought.” Manuscript. Sosa, Ernest. 2021. Epistemic Explanations: A Theory of Telic Normativity, and What It Explains. Oxford University Press. Sutton, Jonathan. 2007. Without Justification. MIT press. Turri, John. 2012. “A Puzzle about Withholding.” The Philosophical Quarterly 62 (247): 355–364. doi:10.1111/j.1467-9213.2011.00043.x. Tversky, Amos, and Daniel Kahneman. 1973. “Availability: A Heuristic for Judging Frequency and Probability.” Cognitive Psychology 5 (2). Elsevier: 207–232. Vargas, Miguel Ángel Fernández. 2016. Performance Epistemology: Foundations and Applications. Oxford University Press. Vaughn, Leigh Ann. 1999. “Effects of Uncertainty on Use of the Availability of Heuristic for Self-Efficacy Judgments.” European Journal of Social Psychology 29 (2-3). Wiley Online Library: 407–410. Vitz, Rico. 2008. “Doxastic Voluntarism.” In Internet Encyclopedia of Philosophy, edited by Unknown Unknown. Vitz, Rico. 2010. “Descartes and the Question of Direct Doxastic Voluntarism.” https://philpapers.org/rec/VITDAT. Wanderer, Jeremy, and Leo Townsend. 2013. “Is It Rational to Trust?” Philosophy Compass 8 (1). Wiley Online Library: 1–14. Whiting, Daniel. 2013. “Nothing but the Truth: On the Norms and Aims of Belief.” In The Aim of Belief, edited by Timothy Chan. Oxford University Press. Williams, Bernard. 1970. “Deciding to Believe.” In Problems of the Self, edited by Bernard Williams, 136–151. Cambridge University Press. Williamson, Timothy. 2013. “Knowledge First.” In Contemporary Debates in Epistemology, edited by Matthias Steup John Turri, 1–10. Blackwell. Williamson, Timothy. 2016. “Justifications, Excuses, and Sceptical Scenarios.” The New Evil Demon. Oxford University Press, Oxford. Citeseer. 41