Trust’s Meno problem: Can the doxastic view account for the value of trust?

ABSTRACT

The doxastic view (DV) of trust maintains that trust essentially involves belief. In a recent paper, Arnon Keren (2020) gestures toward a new objection to the view, labeled Trust’s Meno Problem (TMP), which calls into question the DV’s ability to explain the widely held intuition that trust has distinct and indispensable value. As of yet, there has been no attempt to take up TMP on behalf of DV. This paper aims to fill precisely this lacuna. I do so in three main stages. In §1 I contextualize and elucidate the problem, to which Keren gestures but does not address in detail. In §2 I disambiguate multiple possible interpretations of TMP, seeking to identify the most philosophically challenging. Finally, in §3, I argue that DV can solve even this interpretation. In order to do so, I make use ofthe highly plausible claim we find in the work of Katherine Hawley (2012, 2019): that trust pays a compliment to the trustee. The payoffs of exploring the doxastic view in the context of Trust’s Meno Problem are twofold: we better understand the nature of the problem itself, and we see that the doxastic view can give a satisfying answer.

Keywords:

• Trust
• trust’s meno problem
• value of trust
• the doxastic view
• Arnon Keren
• Katherine Hawley

Introduction

We may split the philosophical study of trust into two distinct questions: one ontological, and one axiological. The former concerns trust’s nature – what trust essentially is. Trust, it seems, is a kind of reliance, but crucially, it is not mere reliance (Goldberg, 2020). Addressing the ontological question thus involves providing an analysis of trust that adequately reflects this distinction. Proponents of doxastic views (Adler, 1994; Hieronymi, 2008; Keren, 2014, 2020; McMyler, 2011) maintain that trust essentially involves a trusting belief on the part of the trustor, i.e., a belief that the trustee will take care of things as entrusted. Proponents of non-doxastic views (Baker, 1987; Faulkner, 2007, 2011; K. Jones, 1996; McLeod, 2002), on the other hand, deny that belief is essential to trust, instead arguing for some non-doxastic attitude as the essential component (e.g., a kind of optimism (K. Jones, 1996)).

Though the two camps thus disagree over the correct answer to the ontological question, they are, for the most part, united in their silence on the axiological question.¹ It is widely agreed that trust is not merely some useful commodity. Rather, trust is, or at least seems to be, distinctly and indispensably valuable for the existence of a society in which individuals can navigate the social world – fostering the kinds of relationships they value, sharing information, engaging in meaningful collaborative projects, and so on. As John Locke put it, trust is the “bond of society” (1663; echoed in Hollis, 1998, p. 4). Sissela Bok similarly writes that, “Whatever matters to human beings, trust is the atmosphere in which it thrives.” (Bok, 1978, p. 31) Far from being merely a handy tool we might make use of in certain situations, then – like a smartwatch or a hot water bottle – trust, it seems, forms the backdrop of cooperative societies, and solidifies the relationships that shape and structure our lives. But if this is so, then the philosophical study of trust ought to distinguish it from mere reliance not only at the ontological level but at the axiological level, too.

In a recent article, Keren (2020) labels this task Trust’s Meno Problem (TMP). The Meno Problem for knowledge involves explaining the intuition that knowledge is not merely distinct from, but also more valuable than, mere true belief. What has gone overlooked in the philosophical literature on trust, however, is that a parallel intuition holds for trust as it does for knowledge. If trust is distinctly and indispensably valuable, then we must explain not merely why it is valuable, but why it is more valuable than that which falls short (Keren, 2020, p. 117). Moreover, it follows that different theories of the nature of trust will provide different accounts of its value, and thus that TMP might provide an additional axis along which we can evaluate these competing theories. (Keren, 2020, p. 117)

The TMP thus applies to all competing philosophical theories of trust. However, no sooner does Keren pose the TMP than does he suggest that it is particularly worrisome for the doxastic view (DV). This is so, the thought goes, because one may act as if one has a trusting belief without really having it, and in so doing one may secure the aforementioned value of trust. Insofar as we agree with Keren, then, that TMP represents a novel standard by which we may evaluate competing theories of trust, the search for its solution is most pressing for none other than proponents of the DV. However, this challenge has not yet been explored in any detail, neither by opponents nor proponents of the view. It is this lacuna in the literature that the current paper aims to fill.

Here is how I proceed. §1 elucidates and contextualizes the problem as it pertains to the DV. Keren gestures toward it but it is yet to be spelled out in detail. Then, having done so, §2 distinguishes between multiple possible interpretations of the problem, seeking to find the one that represents the most serious challenge to the doxastic view. To this end, I utilize Pritchard’s (2009, 2010; 2011) distinctions between multiple specifications of the original Meno problem. Finally, having clarified the problem and found the most challenging interpretation, in §3 I show that the DV has the resources to provide a satisfying answer. In order to do so, I make use of the highly plausible claim we find in the work of Hawley (2012, 2019):that well-placed trust represents a way of showing respect, or paying a compliment, to the trustee. The main payoffs of exploring the doxastic view in the context of Keren’s TMP are twofold: we better understand the nature of the problem itself, and we see that the DV, despite initial appearances, can give a good answer.

1 The doxastic view and the trust meno problem

1.1 Some qualifications

In what follows I will adopt some common assumptions regarding the nature of the attitude whose value we seek to explain. Firstly, I am thinking here of trust as a three-place, interpersonal relation, involving two people and a task.² (Hawley, 2012; Hieronymi, 2008; Holton, 1994; K. Jones, 1996; Hardin 2002; Simion & Kelp, forthcoming) Though we do often speak about cases in which X trusts Y simpliciter, my focus, as is commonplace in the literature, will be on instances in which X trusts Y to ∅. Secondly, and again simply following the dominant view, my focus will be on instances in which X and Y are people. I can rely on objects, but I do not grant them the rich trust I grant people. (Jones, 1996, p. 14; Hawley, 2014, p. 2; Simion, 2021, p. 14) Thirdly, following Baier’s seminal work (1986), I conceive of trust as constitutively involving vulnerability to betrayal. When trusting rather than merely relying, one subjects themselves to the possibility that one will be betrayed, rather than merely let down. (Baier, 1986; Becker, 1996; Dasgupta, 1988) One may well feel let down by a faulty alarm clock, but it is not appropriate to feel betrayed. (Baier, 1986, p. 235) Finally, note that the kind of trust whose value we seek to explain here will have its value not unconditionally, but rather only when it is well-placed, or justified. Moreover, notice that narrowing the focus to well-placed rather than any old trust places restrictions on both trustor and trustee. A trusting relation may fall short of the attitude we have in our sights if either (i) the trustor is unjustified in trusting, or (ii) the trustee fails to act as entrusted. Plausibly, the distinctive and indispensable value of trust is absent in instances in which I trust you while in possession of positive reasons not to trust you.³ Similarly, if I trust you and you subsequently act as entrusted, but do so out of sheer luck rather than any semblance of recognition, acknowledgment, or responsiveness to my trust, then there is a sense in which you have failed to act as entrusted.

1.2 The doxastic view

In what follows, I take these observations for granted, rather than provide any argument for them.⁴ Our focus is on three-place, interpersonal trust, where this is understood as involving vulnerability to betrayal, as well as success conditions on both the trustor and trustee’s side (hereafter, just “trust”). Our answer to the ontological question, the thought goes, ought to be able to explain why trust, so understood, has value over and above that which falls short.

We now turn to the doxastic view, of which there are multiple, competing formulations in the literature (Adler, 1994; Hieronymi, 2008; McMyler, 2011; Keren, 2014, 2020;). It will be beyond the scope of the current paper to settle the question as to which is to be preferred; instead, let’s simply establish its key tenets – some common ground occupied by all formulations.

The pivotal claim distinctive of the DV is that trust essentially involves what I will call a “trusting belief” on the part of the trustor (Adler, 1994; Hardin 2002; Fricker 2006; Hieronymi, 2008; McMyler, 2011; Keren, 2014, 2020). A trusting belief is here understood as a belief either that (i) the trustee will act as entrusted; or that (ii) the trustee is trustworthy with respect to what she is entrusted to do. The details differ depending on the view, but all insist upon the necessity of such a belief for trust.⁵ If X does not believe Y to be trustworthy in the relevant sense, then according to the DV, X does not really trust Y.

The popularity of the view is no mystery. For one thing, the DV does an excellent job of explaining trust’s systematic relationship with belief. Whatever your preferred ontology of trust, it is hard to deny that trust and belief are intimately related; if X trusts Y to ∅, why is it that X almost always believes that Y will indeed ∅, if after all trust does not essentially involve belief? Epistemic trust provides a helpful illustration: oftentimes in testimonial exchanges, we trust speakers for the truth and sincerity of their assertion. But why, if trust does not involve belief, does such epistemic trust invariably result in believing what they say? (Keren, 2020, p. 112) Moreover, the doxastic view makes sense of the way in which trust can give reason for belief. If, as the DV would have it, X’s trusting his colleague, Y, to lock up the office for the day involves a trusting belief, then it is no mystery why X’s trusting Y in this respect seems to give X reason to believe that Y will do so. This is because, as Faulkner puts it, according to the DV, “[t]rust gives a reason for belief because belief can provide a reason for belief” (2017: 113). On a non-doxastic view, which denies the necessity of a trusting belief, the intimate link between trust and belief stands in need of further explanation.

Aside from the relationship between trust and belief, the doxastic view is also on strong ground with respect to the relationship between trust and mere wishful optimism (Carter and Simion (2020). Consider a case in which Xavier, knowing full well that Yasmin is recklessly irresponsible, with a disdain for children, nevertheless trusts her to babysit his children. Moreover, we may suppose that he does so out of wishful optimism that she will one day change her ways. It seems that the most natural reading of such a case is that Xavier lost his trust in Yasmin just when he learnt that she was a terribly irresponsible individual, with a disdain for children, and that the attitude he now adopts is somewhat different to trust. The doxastic view does an excellent job of explaining this difference insofar as Xavier stops trusting when he stops believing Yasmin to be trustworthy. Maintaining and explaining this difference is more difficult on non-doxastic accounts, given that Xavier certainly still seems optimistic that Yasmin will do so, even if he no longer believes.

The DV is thus, at least prima facie, a position worth defending, and the main undertakings one inherits in attempting to do so have been well-rehearsed in the literature to date. In particular, the DV faces a challenge from considerations pertaining to trust’s relationship with evidence (Baker, 1987; Faulkner, 2007, 2011; K. Jones, 1996; McLeod, 2002, 2015), on the one hand, and trust’s relationship with the will (Faulkner, 2007; Holton, 1994; Simpson, 2012), on the other. Whether or not these objections can adequately be addressed by proponents of the DV will be beyond the scope of the current paper.⁶ Instead, my focus will be on a novel kind of objection to the view, articulated by Keren (2020), and dubbed Trust’s Meno Problem.

1.3 Trust’s meno problem

In Plato’s Meno, Socrates famously raises a quandary about the value of knowledge. If knowledge is, as it seems to be, distinctly and uniquely valuable, then what is it that makes it so? Importantly, this question is not to be divorced from the ontological question regarding the nature of knowledge. Indeed, the two are intimately connected: if one’s theory of the nature of knowledge does a poor job of answering the question regarding the value of knowledge, then so much the worse for the theory of its nature. (Kvanvig, 2003, quoted in Greco 2006, 91) The Meno Problem, as it has come to be known, is, as Greco puts it: “ … at the heart of the project of explanation. The task of explaining what knowledge is involves the task of explaining why knowledge is valuable.” (2006, 91)

According to Keren (2020), what has gone largely unnoticed is that parallel intuitions to those that generate the original Meno Problem hold for trust as they do for knowledge. The following two intuitions are widely accepted in the literature:

Trust Intuition 1: Trust has distinct and indispensable value.

Trust Intuition 2: Trust is not to be equated with mere reliance.

However, if this is right, and trust does indeed possess such value, then just as in the case of knowledge, our theory of its nature had better be able to accommodate and explain this value, and, moreover, this value had better not collapse into the value of mere reliance. This is Trust’s Meno Problem (TMP).

Keren points out that extant philosophical discussions of trust, which, when they speak of trust’s value at all, speak almost exclusively in instrumental terms (McLeod 2015), are unlikely to provide an adequate solution to this problem (Keren, 2020, p. 116). For, we can rely on others in ways characteristic of trust without really trusting them, and, presumably, in so doing, realize all of the same instrumental goods associated with trust. Accordingly, TMP consists in explaining why trust is more valuable than relying on others in ways characteristic of trust, if indeed it is. (Ibid.)

This challenge, according to Keren, looks prima facie to be particularly worrisome for proponents of the DV. The reason for this is that proponents of the view, particularly when pressed on objections pertaining to trust’s relation with the will, are especially apt to draw a sharp distinction between real trust, on the one hand, and merely acting as if one trusts, on the other.⁷ TMP, at least for such views, then consists in:

… explaining why this difference should at all matter to us: What is the additional value of relations involving trust when compared with those involving merely acting as if one trusted without believing that the person is trustworthy? Why should we be interested in the very distinction? Is a relation in which I act trustingly because I believe that you are trustworthy more valuable than a relation in which I act in exactly the same way, without having such a belief? If it is indeed more valuable, why is this so? If not, then does not mere reliance, without trust, allow us to enjoy the very same goods that trust allows us to enjoy? In that case, it would seem difficult to maintain that trust is indispensably valuable.

(Keren, 2020, p. 118)

If one can merely act as if one possesses a trusting belief, and in so doing secure the goods associated with trust, then it is not at all clear what value is added by a trustor’s possession of a bona fide trusting belief. Following Keren’s line, then, though TMP is a problem for all accounts of trust’s ontology, it looks particularly pressing for the DV.

2 Interpreting trust’s meno problem

The remainder of the paper is dedicated to arguing that, if there is indeed reason to abandon the DV, it is not on grounds of TMP. However, TMP, much like the original Meno Problem, admits of numerous possible interpretations. The purpose of the current section is to sharpen the problem by identifying its most challenging interpretation – the interpretation that poses the biggest threat to the DV. For, if the DV can solve even this interpretation, then we will have made a good case that there’s nothing to see here. Pritchard (2010) helpfully distinguishes between distinct tiers of Meno Problem for knowledge: the Primary, Secondary and Tertiary Meno Problems, as well as the closely related Swamping Problem. All of which pose different challenges and all of which, I believe, have trust analogues that are overlooked in Keren’s challenge.

2.1 Trust’s primary meno problem

The minimal reading of the original Meno Problem is that of explaining why knowledge is more valuable than mere true belief. Pritchard calls this the Primary Meno Problem. The key thought here is that if we cannot even explain why knowledge is more valuable than mere true belief, then the very project of accounting for the value of knowledge is a lost cause (Pritchard, 2010, p. 6).

Translating this into trust’s terms, we see that herein lies a simple trust-analogue. Straight out of our guiding trust intuitions falls the following, which we might call Trust’s Primary Meno Problem: that of explaining why trust is more valuable than mere reliance. If, in an attempt to explain and accommodate Trust Intuition 1, one were forced to abandon Trust Intuition 2, then one would have fallen at the first hurdle, as it were.

2.2 Trust’s secondary meno problem

Pritchard, however, points out that if one were to solve the Primary Meno Problem for knowledge by pointing out some value-conferring property X, then another problem immediately arises just in case there are Xs such that mere true belief plus X is nevertheless insufficient for knowledge. The reason for this, according to Pritchard:

… is that if the distinctive value of knowledge is due to some feature of knowledge which, with true belief, falls short of knowledge, then it seems that what we should seek is not knowledge as such, but rather that which falls short of knowledge […] But if that’s right, then why do we regard knowledge as distinctively valuable at all?

(2010, 7)

This problem is pertinent in the context of the original Meno Problem because there are states more valuable than mere true belief that are nevertheless still not knowledge – justified true belief, for example. The Secondary Meno Problem, according to Pritchard, thus consists in explaining the value of knowledge not merely over-and-above true belief, but anything that falls short of knowledge.

What is important to note, for our purposes, is that the Secondary Meno Problem, too, will arise for trust as it does for knowledge. The reason for this is that, just as with mere true belief and knowledge, there are attitudes more valuable than mere reliance which are nevertheless not trust. Take reasonable as opposed to unreasonable reliance, for example. I might either reasonably or unreasonably rely on somebody to do something. Plausibly, when I do the former, my doing so has more value than when I do the latter.⁸

If this is right, then even if we could provide an answer to Trust’s Primary Meno Problem, we would immediately face what we may now call Trust’s Secondary Meno Problem. Here it is: if one solves the Primary Meno Problem by citing some additional value-conferring property X in their analysis of trust, then whilst they may have explained the value of trust over-and-above mere reliance, they have not yet explained the value of trust over-and-above that which falls short of trust. Neither reasonable nor unreasonable reliance are instances of trust (to see this, note that, unlike trust, neither need incur any vulnerability to betrayal) but that does not mean that the former is not more valuable than the latter. If the distinctive value of trust is due to some feature of trust which, when combined with mere reliance, falls short of real trust, then it seems that what we value is not trust as such, but rather that which falls short of trust (i.e., mere reliance plus value-conferring property X, in this case reasonableness). But, of course, if that is correct, then we have failed to accommodate Trust Intuition 1, that trust has distinct and indispensable value. (Pritchard, 2010, p. 7)

2.3 Trust’s tertiary meno problem

It seems that in the case of trust as in the case of knowledge, both primary and secondary iterations of the Meno Problem require a solution if we are to adequately account for their respective value. Pritchard, however, points out a further complication, which arises when we note that even an answer to the Secondary Meno Problem does not guarantee a satisfactory account of the distinctive value of knowledge. The reason for this, according to Pritchard, is that the Secondary Meno Problem leaves open the possibility that the difference of value at issue between knowledge and that which falls short is merely one of degree rather than of kind. (Pritchard, 2010, p. 7) To characterize the difference as a mere difference in degree, according to Pritchard, is to place knowledge on a kind of a continuum with that which falls short. And to do so, the thought goes, is to fail to capture the distinctiveness of the value knowledge seems to us to have. (Pritchard, 2010, pp. 7–8) After all, if the difference is merely one of degree rather than of kind, then why the long history of epistemological focus on this particular stage of the continuum? (Ibid., 8) Accounting for our intuitions about the value of knowledge over-and-above that which falls short thus requires an explanation that spells out the difference as one of kind rather than of degree. Call this the Tertiary Meno Problem.

Now, trust’s analogues to the primary and secondary interpretations of the Meno Problem fell straightforwardly out of Intuitions 1 and 2. Whether or not trust has an analogue to the Tertiary Meno Problem, I think, is less obvious. Perhaps we are let off the hook by the fact that, as we saw at the outset, trust is considered in some sense a kind of reliance (Goldberg, 2020). Perhaps this gives us reason to focus on trust’s Primary and Secondary Meno Problems, and ignore the Tertiary interpretation.

Perhaps. But I think that, on closer inspection, there is reason to worry that trust has its own version of the Tertiary Meno Problem, too, which would be left unsolved were we to focus our attention exclusively on the Primary and Secondary versions. Note that while it is true that trust is seen as a kind of reliance, there are good grounds for thinking that this characterization is somewhat imprecise.

To see why, note that extant attempts to add that special component, X, to reliance in order that it transforms into trust are oftentimes attempts not merely to slightly tweak the attitude, but rather to transform its entire essence. So much so, one might think, that it may no longer make sense to conceptualize reliance and reliance plus X as mere tokens of the same broad type. Moreover, this is plausible on any competing account of the content of X. Look, for example, at what is supposed to happen to mere reliance once X is added such that X is considered even a contender in the literature. X must add a vulnerability to betrayal component – an evaluatively loaded concept – that is absent in mere reliance. It must also place non-negligible demands on both trustor and trustee that are absent in mere reliance. Moreover, X must transform mere reliance into the kind of attitude capable of playing the kind of role we commonly think that trust plays, i.e., the kind of attitude that can create, as Locke put it, the “bond of society”, or, on Bok’s line, the atmosphere in which what we care about can thrive. Some have thought it plausible that trust facilitates or allows for the kinds of human attachments essential to our well-being (Harding 2011; Kirton 2020, forthcoming; Wonderly 2016). If this is right, then perhaps a more accurate reading is that the difference between trust and reliance – much like the difference between knowledge and true belief – is best conceptualized as a difference in kind rather than merely in degree.

Perhaps not. What is important to note for our current purposes is that, if this reading is at all plausible, then we run the risk of failing to adequately accommodate all of our intuitions regarding the value of trust by focussing solely on the Meno Problem’s Primary and Secondary interpretations. The reason for this, of course, is that they leave open Trust’s Tertiary Meno Problem, whereas solving the Tertiary problem would also deal with the first two problems. Even if one is left unmoved by the observations in the previous paragraph, then, one will surely still agree that, all else equal, if one could provide an answer to the Tertiary Meno Problem, then this would be by far the preferable outcome. (Pritchard, 2010, p. 8)

2.4 Trust’s swamping problem

On the way to addressing Trust’s Tertiary Meno Problem, we must make one last stop to consider the much-discussed Swamping Problem for knowledge. As Pritchard (2010, p. 9) highlights, the Swamping Problem is a variant on the Meno Problem which is not necessarily dealt with in the process of dealing with the other interpretations. Given this, and so as to avoid leaving any stone unturned, best to establish whether it, too, has a trust analogue we ought to consider.

The Swamping Problem⁹ can be illustrated with a common example, involving two excellent cups of coffee. Suppose these two cups of coffee are identical in every way – in flavor, appearance, quantity, price and so on. Now, supposing we value excellent coffee, it follows that we value excellent coffee machines. Notice, however, that once we have the excellent coffee, it is no longer clear that we care about where it came from. In other words, as long as we have the excellent coffee, the fact that it was produced by an excellent machine confers no additional value upon it. (Pritchard, 2010, p. 9) Whatever value is conferred upon a coffee by virtue of its being the product of an excellent machine, this value is “swamped” by the value conferred upon the coffee by virtue of its being excellent. The broad problem for theories of epistemic value is that if whatever component you add to true beliefs in order to transform them into knowledge (the property of being reliably produced, for instance) is only instrumentally valuable relative to the further good of true belief, then in cases in which you have a true belief, no further value is conferred by the presence of your added component. (Pritchard, 2010, p. 10)

To translate once again into trust’s terms, the analogue challenge for philosophical accounts of trust is as follows: if your added component, X, which transforms less-than-trust into trust, adds value that is instrumental relative to the goods associated with trust, then it is no longer clear what value is added by the involvement of trust in instances in which these goods are already present. The worry for philosophical theories of trust can be made more pertinent by noting that trust’s value, when it is discussed, is indeed almost exclusively discussed in instrumental terms – in terms of the goods it produces, preserves, or maintains (McLeod 2015). Some spell out trust’s value in terms of social goods, of cooperation, cohesion, and so on. Dimock (2020), for example, argues that: “If trust is a cooperation-enabling resource, it is an important kind of social capital in communities.” (161) Others similarly argue that trust enables individuals “ … to work together for common purposes in groups and organizations.” (Fukuyama 1995) Moreover, and as a result of this, some argue that “high-trust” societies have stronger economies and stronger social networks in general than “low-trust” societies. (Fukuyama 1995; Inglehart, 1999) What is important for our purposes is that whichever extant account of trust’s value is correct – and there may be elements of truth in all of them – we are immediately faced with Trust’s Swamping Problem insofar as this value is instrumental relative to trust-associated goods.

Suppose that the value of trust is relative to goods pertaining to, roughly, social cohesion. Now, further supposing that we can generate and maintain social cohesion through alternate methods, it is no longer clear why trust’s value is distinct and indispensable. In other words, we would have failed to accommodate Trust Intuition 1. Perhaps it is not; perhaps the goods of trust cannot be generated and maintained in the absence of real trust. But the case for this has not been made, and would need to be, lest we leave Trust’s Swamping Problem unsolved.

Now, while it is possible that in handling Trust’s Tertiary Meno Problem we will have, in so doing, handled Trust’s Swamping Problem, it is not a given. Suppose we show that trust’s value over-and-above that which falls short is higher not merely in degree but also in kind. Still, if trust’s value is instrumental relative to trust-associated goods, then whenever these are present, the kind of value it has will be lost. Trust’s Swamping Problem is thus a distinct problem worthy of further investigation. But, perhaps more worryingly for our present purposes, it looks to sit very nicely with Keren’s original reasons for thinking that Trust’s Meno Problem was troublesome for the DV in particular. Recall that the problem for the DV consisted in explaining why the difference specifically between holding a trusting belief, on the one hand, and merely acting as if one held one, on the other. Note the following passage, from Keren, in particular:

Is a relation in which I act trustingly because I believe that you are trustworthy more valuable than a relation in which I act in the exact same way, without having such a belief? If it is, why? If not, then does not mere reliance provide us with the very same goods that trust does? If so, trust doesn’t seem indispensably valuable.

(2020, 118, my italics)

The thought, then, is that one may realize the goods relative to which trust is thought to be valuable by merely relying, just so long as one acts as though they had a belief. We now see, more precisely, the way in which Keren thinks that the worry is one which plagues the DV in particular. If you fill in the X in the equation “mere reliance plus X equals trust” in the way that the proponent of the DV is apt to, then you must explain why this attitude has more value than another which could, at least in theory, produce the same goods.

In what follows I set aside whether Keren is right that this problem is uniquely worrisome for the DV. Let us grant that it is. Suffice it to note that the disambiguation project of this section has delivered two major findings: (i) if one can solve Trust’s Tertiary Meno Problem, one will thereby have solved trust’s Primary and Secondary Meno problems; but that, (ii) in so doing, one will not necessarily have solved Trust’s Swamping Problem. In order to properly address Keren’s challenge then, the proponent of the DV must provide an adequate solution to both Trust’s Tertiary Meno Problem, and Trust’s Swamping Problem. It is this task to which I now turn.

3 The final value of trust

One of the main objectives of this paper is already secured: that of disambiguating Trust’s Meno Problem, to which Keren (2020) gestures but does not explicate. We now see that it is subject to analogues of Pritchard’s distinctions between Primary, Secondary, and Tertiary interpretations, as well as the related Swamping Problem. In this final section I argue that the DV can meet even the most challenging versions of Trust’s Meno Problem, which we have identified as both Trust’s Tertiary Meno Problem and Trust’s Swamping Problem. In order to do so, my strategy will be to explicate the oft-overlooked notion of trust’s non-instrumental value. In what follows, I will grant that trust’s associated goods could indeed be attained by merely acting as if one trusted, as opposed to really trusting. Still, I will argue that trust has non-instrumental value, value not derived from its ability to produce further goods, and that the DV does a good job of explaining this value.

3.1 Two kinds of non-instrumental value

One may be forgiven for thinking that the argument for trust’s non-instrumental value is off to a shaky start given that, for one thing, we have built success conditions into our exploration of its value. In other words, given that we noted at the outset that a plausible account of trust’s value need only explain the value it has when it goes well, does this not in itself rule out the possibility that trust per se has any non-instrumental value?

This worry can easily be addressed by noting a common distinction in the broader axiology literature between two kinds of non-instrumental value: intrinsic and final. Whilst the fact that the value of trust is hostage to certain success-related constraints might rule out its having intrinsic value, it does not rule out its having final value. Some good is said to have final value if it is valuable for its own sake – and thus non-instrumentally valuable – but is so in virtue of a relational rather than intrinsic property. Two intrinsically identical items might, therefore, differ in value in virtue of their relational properties. The common example is of Princess Diana’s Dress. (Rabinowicz & Rønnow-Rasmussen, 2000, p. 41) A dress that belonged to Princess Diana might well be considered more valuable than an identical dress that did not. This value cannot be accounted for in intrinsic terms given that they do not differ, we may suppose, in any intrinsic properties. But neither can it be accounted for in instrumental terms. The value Diana’s dress has is not derived from the goods it produces, or what it allows you to do; rather, its value is owed to its relation to Diana.

What is important to note is that the kind of trust whose value we seek to explain is, as we have noted, a three-place, interpersonal relation, which holds between two individuals (trustor and trustee) and an action (∅). There is thus scope for this kind of relation to have non-instrumental, yet non-intrinsic, final value. The attitude whose value we seek to explain might possess its distinctive value in virtue of its relational rather than intrinsic properties. The value of these relational properties, which obtain between two people and an action when all goes well, might be what distinguishes this attitude from attitudes that fall short.

3.2 Why think that trust has final value?

If it turns out that trust has final rather than merely instrumental value, then we have a promising new route for solving the most challenging versions of Trust’s Meno Problem, or so I will argue. First thing’s first, however: why think that trust has such value?

As we have seen, the little discussion there is of the value of trust has focussed almost exclusively on its instrumental rather than non-instrumental value. Is this, we may ask, because the thought that trust has non-instrumental value is prima facie implausible, or counterintuitive? Not at all. To the contrary: though underexplored in the literature, I think the notion of trust’s having non-instrumental value is, in fact, highly intuitively plausible irrespective of your preferred ontology of trust, and stands up on closer scrutiny. I now elucidate and defend this idea.

Here is a plausible thought we find in Hawley (2012; though we can see similar ideas elsewhere, e.g., O’Neil (2012, p. 311) and D’Cruz (2019, 2020)): trust constitutes a kind of compliment, or respect, to the trustor; distrust, on the other hand, constitutes a kind of insult, or disrespect. Here is Hawley:

Some of the benefits and harms generated by trust and distrust are practical – affecting our health, wealth, or enjoyment of life. But some strike more directly at who we are, at our personal integrity and autonomy. Distrust is an insult, not just an inconvenience; correspondingly, trust is a compliment. I want my friends, family, and colleagues to trust me, not just because that will make things go more smoothly for me – though it will – but also because I value their high opinion of me, and their level of trust is a key part of their opinion.

(2012, 13–14)

Here is another passage we find elsewhere in Hawley’s work:

Trust is not ethically neutral. In the right circumstances, we expect others to trust us, and it is an insult to be distrusted without good reason. Attitudes of trust and distrust carry an evaluative message, and can have far-reaching consequences of great ethical significance. Trustworthiness is an ethically admirable trait, contrasting with both dishonesty and flakiness …

(2019, 2)

There seems to me to be something deeply important in these insights from Hawley: that trust, at least when well-placed, pays a kind of a compliment to the trustee. Consider the following case:

Bar: Claire is the manager of a bar. She has equal reason to trust her two assistant managers: Lilith and Angie. Neither has ever gone over-and-above for her; but nor has either of them ever let her down. However, for whatever reason, when it comes to important matters, Claire finds herself trusting Lilith and not Angie.

Following Hawley’s insights, I think that Claire’s trust conveys a kind of a compliment to Lilith. She deems Lilith worthy of trust, at least when it comes to important matters, and that seems to be to say something important, and evaluatively positive, about Lilith as a person. Of course, Lilith may not care about, or even notice, this message. But that does not mean that Claire’s trust doesn’t carry it.

Let us further suppose that there are no positive downstream effects for Lilith of having been trusted. There is no opportunity for promotions down the line, no benefits of having gained Claire’s trust. We may even suppose that doing so will, in fact, be an inconvenience, all things considered. Nonetheless, I think that, in such a case, Claire’s trust alone pays Lilith a kind of a compliment not paid to Angie.

We should be careful not to overstate the point here. It would, of course, be an overstatement to suggest that trust must necessarily leave the trustee feeling flattered, or anything similar. However, I think it is very intuitive that at least part of what sets trust apart from that which falls short is the kind of evaluative compliment that it comes with. Given that trust comes with a vulnerability to betrayal, it follows that it comes with a kind of uncertainty. X is not vulnerable to betrayal if she has a guarantee that Y will indeed ∅. The kind of compliment which I think trust embodies, extrapolating from Hawley, is that, under this uncertainty, X deems Y to be the kind of person who will indeed ∅. This, I think, says something positive about Y qua individual. This is the sense in which Claire pays Lilith, and not Angie, a compliment.

3.3 Final value and trust’s meno problem(s)

We have seen, with the help of Hawley (2012, 2019), that it is plausible that trust’s value is not merely instrumental, but also final. If this is right, then a straightforward road to solving Trust’s Meno Problem(s) suggests itself. Before demonstrating that this road is open to proponents of the DV in particular, let me firstly explain what it looks like exactly.

In order to see why trust’s having final value solves the problem, let us return to the two most challenging interpretations of the problem. Firstly, we have Trust’s Tertiary Meno Problem, which we found to be worthy of our attention given that, if any view on trust could solve it, they would thereby have also solved Trust’s Primary and Secondary Meno Problems. Now, the reason that our findings in §3.2 help with this is that, if trust’s value consists – at least partly – in the compliment it embodies, then, at least so long as your view of trust can explain this value, it looks as though your view can differentiate trust from that which falls short in terms of a difference in kind, rather than merely degree. An attitude which embodies a compliment about another individual qua individual, I contend, looks like a different kind of attitude from one that does not. If trusting involves paying a compliment, then it is not clear that merely dialing up the degree of your reliance will ever do the same.

Now for Trust’s Swamping Problem, which looked to line up particularly well with the challenge to the DV as Keren sets it out. For this, the solution in terms of trust’s final value is more straightforward. Trust’s Swamping Problem, recall, demands an explanation of the distinctive and indispensable value of trust that is not swamped by the value of having the goods associated with trust already in one’s possession. This was particularly pertinent to the DV, recall, given that it seemed as though one may act as if one held a trusting belief – and thereby secure the goods associated with trust – without really holding one. However, if it is true that trust has final value, then even if one could secure the goods associated with trust by adopting some attitude that is less-than-trusting, one would have failed to thereby secure the full value of trust. There would, on such an explanation, be value over-and-above trust’s instrumental value.

3.4 Can the doxastic view account for trust’s final value?

What has been said so far lends support to the idea that trust has final value. Moreover, nothing so far is theory-laden; it does not lend itself to any particular view of trust. However, while our answer is not yet theory-laden, our question pertains to the DV in particular. The final piece of our puzzle thus requires exploring whether the DV can account for trust’s non-instrumental, final value.

Consider the following case, designed to show that it can.

Science lab: Simon is in charge of a lab of scientists. Due to gender bias, Simon trusts the men in his group far more than any other gender identity. However, Simon recognises that instilling an atmosphere of mutual trust is beneficial for the continued success of the lab, which, in turn, will reflect kindly on his own reputation and career. As a result, though it is never manifested in his behaviour, when he sets a task for the lab, he believes that the men will fulfil it and does not believe that anyone else will.

Here is the thought: irrespective of what Simon believes, he does not pay the characteristic trust compliment when he merely acts as if he believes; he does so only when he really believes.¹⁰ As a result, the relation between Simon and those whom he believes to be trustworthy and those whom he does not is different (and lesser) in evaluative character. This thought can be brought out by considering a scenario in which, for example, the women in his group found out that he did not really believe them to be trustworthy. Suppose they confronted him, questioning why he failed to trust them when he trusted the men on the very same evidence regarding their respective competence. We can expect that they would be decidedly unsatisfied were he to defend himself by pointing out that he never allowed this to manifest itself in his behavior. Irrespective of his behavior, the attitude he takes up toward them is different than the one he takes up toward the men.

Moreover, this difference cannot be explained by reference to the fact that the women have now found out, and are thus insulted, or offended. Though this is likely to also be true, the crucial point is that they have a legitimate complaint regarding the initial state of affairs. Their complaint regards the attitude he adopted toward them, as is brought out by comparison with that which he adopted toward the men. Furthermore, this difference is also not explained away by the downstream effects of his trust because, we may suppose, there are none. We may stipulate that no concrete difference in outcomes whatsoever results from Simon’s trusting only the men. Does this evaporate the difference in the attitude he adopts toward everyone else when compared to that which he adopts toward the men? It doesn’t seem that it does. In such cases, the disappointment of the employees is best explained by Simon’s lack of trusting belief, rather than of trusting action. He did successfully act as if he trusted, what he failed to do was believe.

Conclusion

As Hawley put it, we want our friends, family and colleagues to trust us not merely because of the downstream effects, but also because we value their high opinion of us. (2012, 14) Here I have argued that it is in this high opinion – in the form of a kind of compliment – that trust’s final value lies, and, moreover, that the DV can give a good account of such value. When we ask our friends, family, and colleagues to trust us, what we want is for them to believe that we are trustworthy with respect to the ∅ in question, not merely to act as if they so believe. It seems, then, that what we value essentially involves no less than a trusting belief; a would-be trustor who merely acts as if they have a trusting belief fails to pay the sort of compliment associated with full trust. If this much is right, then it looks as though the proponent of the DV can give a satisfying account of the distinctive and indispensable value of trust. To be clear, the purpose of this paper was not to argue that the DV is the only view that can provide such an account, or even that it necessarily provides the best one. Rather, the objective here was more modest: to show that, if there is a reason to abandon the DV, it is not on grounds of Trust’s Meno Problem.¹¹

Disclosure statement

No potential conflict of interest was reported by the author.

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

The work was supported by the Arts and Humanities Research Council (AHRC) and the Scottish Graduate School for Arts and Humanities (SGSAH).

Notes

1. A notable exception is Simpson (2012).

2. See Nguyen (forthcoming, for an argument that there is a special kind of trust that is not exclusively interpersonal. For lack of space, and so as to walk in lockstep with the vast majority of the philosophical literature on trust thus far, I set this aside in what follows.

3. For example, see Carter and Simion (2020) for discussion of “wishful optimism” cases, in which, out of sheer wishful optimism, X trusts Y to carry out some task despite Y’s having proven to be extremely unreliable with similar such tasks.

4. It is important to note that these observations should be uncontroversial insofar as they are shared by most, if not all, competing accounts of the ontology of trust.

5. Some views, Hieronymi (2008), e.g., make a sufficiency claim. I set aside such “pure” doxastic views in what follows and focus on the weaker, “impure” formulation of the view, which involves only a necessity claim.

6. For responses on behalf of the DV, see Hieronymi (2008); Keren (2014), (Keren, 2020).

7. See Hieronymi (2008), for example.

8. One may resist the analogy here by pointing out that while JTB is not sufficient, it is nevertheless necessary for knowledge, whereas on the other hand, reasonable trust is not even necessary for trust. After all, X may plausibly trust Y to ∅ even if they do so unreasonably. Now, while it is indeed possible to unreasonably trust, what’s important to point out is that in doing so, one will have taken up an attitude less valuable than full trust, i.e., the attitude whose value we seek to explain. The attitude that has distinct and indispensable value, and which needs to be distinguished from those attitudes that fall short of having such value, is not merely any old trust; it is, I contend, reasonable, or well-placed, trust. My thanks to an anonymous reviewer at Philosophical Psychology for pressing me on this point.

9. See W. E. Jones (1997), Swinburne (1999, 2000), Riggs (2002), Kvanvig (2003), Zagzebski (2003), Greco (2010), for critical discussion of the Swamping Problem.

10. In a recent paper, Bieber and Viehoff (2023) briefly address a similar question, namely, why trust-without-belief cannot serve the characteristic function(s) of trust. They argue (2023, 20) that while it makes sense to say that someone trusts someone else despite lacking a trusting belief about that individual, this not the norm, or paradigm case. The reason it makes sense to say this, according to Bieber and Viehoff, is that convincingly acting as if one has a trusting belief can serve (at least some of) trust’s functions. However, that this will fulfil trust’s function(s) is parasitic on the paradigm case involving bona fide belief. This is so given that trust-without-belief will not be able to fulfil trust’s function unless it successfully passes for trust-with-belief, and that its doing so is parasitic on trust-with-belief being the norm, or paradigm, case. If people always acted as if others were trustworthy, without believing, the practice would collapse. While this addresses an important and closely related question, my focus here is slightly different. Firstly, while I have argued here that trust characteristically involves a kind of compliment, I have not argued that this is part of the function of trust. Some item P can have value (in particular non-instrumental value) in virtue of some property X without X having anything to do with P’s function. Diana’s dress has value in virtue of its relation to Diana, but it does not function to manifest such a relation. Of course, it is an interesting and pertinent question how the function of trust relates to its value, but it is not one I have addressed here. Secondly, and relatedly, it is not clear whether, for Bieber and Viehoff, non-paradigmatic trust is nevertheless bona fide trust. Presumably, given that they aim to characterize trust along functionalist lines, then simply fulfilling trust’s characteristic functions is sufficient for bona fide trust. However, given that I do not share this functionalist ambition, my position is unambiguous on this matter: insofar as trust does not embody the kind of compliment I have explicated here, it is not bona fide trust.

11. Many thanks to Adam Carter, Rory Aird, and Katie Moody, without whose help and support I would not have managed to write this paper. Thanks also to Jesús Navarro, Tim Kearl, Dario Mortini and Aarón Álvarez-González for helpful discussion. Finally, I am extremely grateful to an anonymous reviewer at Philosophical Psychology for their thoughtful and constructive comments. Research for this paper was generously funded by the Scottish Graduate School of Arts and Humanities.