1. The Resilient Reliance theory of faith

When you put your faith in someone for something, you rely on them for it, or you are disposed to do so. This explains why faith can be risky, because you might be mistaken about whether to rely on them. Moreover, it sheds light on why faith can make you vulnerable, because they might not come through with respect to what you rely on them for. In addition, it clarifies why a charge of faithlessness can stick, because if you have no tendency whatsoever to rely on someone for something when you should, you deserve to be upbraided for a lack of faith: “have a little faith in me!” Further, it makes sense of what faithfulness is because when you are faithful to someone who has faith in you for something, you’re disposed to come through reliably with respect to what they rely on you for.

Some theorists might want to use ‘rely’ as a stative verb in this context, merely denoting a state of dependence. We prefer to use ‘rely’ as an active verb here, denoting a nonbasic action, one you perform by doing other things, e.g., relying on a personal trainer by following their advice and attending their sessions. That’s because faith implies action, and the action characteristic of having faith in someone for something is relying on them for it. We must exercise caution here, however. After all, you can have faith in someone for something even while you are not performing the act of relying on them for it, as when you retain your faith in someone as your spouse even while you are asleep or fully absorbed in something else. So, performing the act of relying is not essential to faith, but rather a disposition to perform it. In what follows, we will use ‘rely’ to denote the act and ‘disposition to rely’ to denote the disposition to perform that act.

A mere disposition to rely is not enough for faith. That’s because faith involves certain cognitive and conative states in addition to a disposition to rely. If Jim has faith in the therapist to help him with his issues, he is not only disposed to rely on her to help him, he is also in some cognitive state or other toward her helping him and he is in some conative state or other toward her helping him. But he’s not in any old cognitive and conative states. Jim cannot have faith in the therapist to help him if he wants her to help him but believes she will not do so. Likewise, Jim cannot have faith in the therapist to help him if believes she will help him but he wants her not to do so. Disbelief and disdesire are too “negative” for faith; more “positive” states are required.

At a first approximation, for Jim to be in a positive cognitive state toward his therapist’s helping him is for him to be in some cognitive state or other that represents her helping him, with four features: (i) it has the propositional content that his therapist will help him, (ii) it disposes him to take a stand on behalf of the truth of that proposition, (iii) it is responsive to his evidence for its truth and/or it is the output of a cognitive capacity the exercise of which aims at forming and retaining true positive cognitive states, and (iv) it is how it ought to be only if it is true. Belief that his therapist will come through counts as a positive cognitive state but—crucially—there are other candidates, e.g., a high enough credence or confidence that she will help him or, depending on the details, propositional acceptance, propositional trust, propositional hope, propositional reliance, imaginative assent, and beliefless assuming.Footnote ⁴ We hasten to add an important qualification. When propositional belief is the cognitive component of faith, faith does not preclude believing “thinner” propositions. For example, if Jim has faith in his therapist to help him, instead of believing the “thick” proposition that she will help him, he might believe the “thinner” proposition that she will likely help, or that she will more likely than not help, etc. Belief in the “thick” proposition that she will help is optional.

As for a positive conative state, for Jim to be in a positive conative state toward his therapist helping him is for him to be in some state or other that motivates him to rely on her to help him. Wanting her to help him counts but—crucially—there are other candidates, e.g., depending on the details, looking with favor on her helping him, being for it, a felt attraction to it, caring about it, it mattering to him, being emotionally invested in it, and affection for, or commitment to, her helping him, among other possibilities, all of which have been mentioned in the literature; even a second-order desire, i.e., wanting to want her to help him, might fit the bill.Footnote ⁵

We collect both kinds of state under the label ‘positive stance.’ For Jim to be in a positive stance toward the therapist helping him is for him to be in a positive cognitive state and a positive conative state toward her helping him.

However, a disposition-to-rely-plus-a-positive-stance is not enough for faith. You can be disposed to rely on someone for something, be in a positive stance toward them coming through, and yet lack faith in them because you are disinclined to rely on them when faced with the least bit of difficulty in doing so. But, whatever else faith is, it isn’t fickle, flighty, or weak kneed. It need not be ripped, six-pack-ab faith, but it has to have at least a little firmness to it, some measure of grit, tenacity, perseverance, pluck, moxie, pushback, steadfastness, stick-to-itiveness, or the like in the face of challenges, whether they are emotional, evidential, social, or something else besides.Footnote ⁶

Faith’s ‘resilience’—as we call it—accommodates positive correlations between faith and resilience in a variety of contexts, both secular and religious. For example, historian Teresa Morgan argues that in the ancient Greco-Roman world, pístis and fides played an essential role in relationships of mutual faith and faithfulness “at every socio-economic level,” “relationships of wives and husbands, parents and children, masters and slaves, patrons and clients, subjects and rulers, armies and commanders, friends, allies, fellow-human beings, gods and worshippers,” most evident in times of relational crisis involving “fear, doubt, and skepticism” (Reference Morgan2015, 45, 63, 64, 75, and passim). And in contemporary secular contexts—e.g., in politics, music, sports, biography, the movie industry, agriculture, and journalism, to name a few—faith frequently features resilience, as when, in his farewell speech in February 2017, President Barack Obama commended faith to his fellow citizens, “faith in America and in Americans,” and “faith … in the power of ordinary Americans to bring about change,” a commendation not only to rely on them but, in light of the preceding 2016 election, to continue relying on them despite what he regarded as the challenges to doing so represented by the incoming administration (Obama Reference Obama2017). Faith’s resilience also accommodates positive correlations between faith and resilience in religious contexts, e.g., in expressions of pístis in the Christian New Testament and ʾĕmûnāh in the Hebrew Bible, and in the biblical themes of covenant and salvation, the lives of exemplars of faith in God, and the lives of ordinary people of faith today. In addition, faith’s resilience can help shed light on the value of relationships of mutual faith and faithfulness, about which we’ll say more later.Footnote ⁷

Reflections such as these lead us to posit a theory of faith we put like this:

So, on Resilient Reliance, for Jim to have faith in his therapist to help him is for him to be disposed to rely on her to help him, with resilience in the face of challenges to doing so, because of his positive stance toward her helping him. And, for Iman to have faith in Allah to reward her devotion is for her to be disposed to rely on Allah to reward her, with resilience in the face of challenges to doing so, because of her positive stance toward Allah rewarding her.

Some clarifications are in order. First, faith’s resilience does not need to dispose us to overcome all possible challenges, only some. Moreover, our faith in someone can be bad and/or vicious because, among other things, we can be overdisposed to overcoming challenges to relying on them, not giving up when we should, and we can be underdisposed as well, giving up too easily when we shouldn’t; and we can be poor judges of when, how, and in whom to have faith for what. Further, faith can be viewed as a role-functional psychological state on our theory, one that takes as input any of a wide variety of combinations of positive stances toward someone coming through and gives as output a disposition to rely on them to do so, with resilience in the face of challenges. Finally, while faith is often epistemically irrational or otherwise intellectually not-up-to-snuff, that’s not necessary on our theory. That’s a plus, by our lights.Footnote ⁸

Since our focus is on how trust relates to faith given our preferred theory of faith, we will only briefly indicate some reasons to prefer Resilient Reliance over two instructive alternatives, namely Belief-Only and Belief-Plus. According to

Notice that, on Belief-Only, relational faith is just a propositional belief.Footnote ⁹ This is its Achilles’ heel. For it seems that you can believe that someone will come through with respect to something and yet lack faith in them for it because you are not in favor of them coming through. Even if we had believed that Timothy McVeigh would come through as a terrorist, we would never have placed our faith in him to bomb the Federal Building in Oklahoma City since we are not in favor of his doing so. Moreover, it seems that you can believe that someone will come through with respect to something and you can be in favor of them coming through, and yet lack faith in them for it because you are not the least bit disposed to rely on them for it. Even if Jim believes that his therapist will help him and even though he wants her to do so, due to any number of factors, e.g., a paralyzing fear of revealing his innermost self to anyone, he might not be the least bit disposed to rely on her to help him. Furthermore, it seems that you can believe that someone will come through with respect to something and you can want them to come through and you can be disposed to rely on them to come through, and yet not have faith in them for it because you are not the least bit disposed to be resilient in the face of challenges. Even if Iman is disposed to rely on Allah to reward her devotion, she may be so disposed only when things are going well for her and not at all in the face of the slightest evidential, emotional, social, or other challenge to relying on Allah; in that case, it seems she lacks faith.

The upshot is that Belief-Only omits three essential components of faith: a positive conative state, a disposition to rely, and a disposition to be resilient in the face of challenges. Suppose we agree. In that case, we might simply add the missing elements to get:

Notice that while Belief-Only differs in many respects from Resilient Reliance, Belief-Plus differs from Resilient Reliance in just one respect. While Belief-Plus specifies the positive cognitive state required for faith as belief that they will come through, Resilient Reliance allows others.Footnote ¹⁰

We find the specificity of the belief-condition problematic. That’s because, as we indicated earlier, other kinds of positive cognitive states can play the role Belief-Plus assigns only to propositional belief, and even when propositional belief is the cognitive component of faith, it seems that faith does not preclude believing “thinner” propositions in addition to “thick” ones. Further, Belief-Plus cannot accommodate and explain the fact that exemplars of religious faith often struggle with intellectual doubt over the course of their journey, sometimes even belief-cancelling doubt. We develop this last point elsewhere by arguing that, in the stories of the Hebrew Bible and the Christian New Testament, several characters presented as exemplars of faith in God and/or Jesus struggle with doubt, e.g., Abraham, Jesus, the woman with a hemorrhage, the Syrophoenician woman, Jairus, and blind Bartimaeus.Footnote ¹¹ Here we call attention to a modern-day exemplar, Mother Teresa of Calcutta.

In 1942, after what she took to be a calling from the Lord, she made a private vow to give herself completely to him, no matter what, and to serve him in the poorest of the poor. What she didn’t expect at the time was that the “no matter what” clause of her vow would include five decades of relational emptiness and severe doubt. It appears from her private writings that she not only experienced the felt absence of God during that time, but she also experienced doubt of a sort and degree that is incompatible with belief. “[T]here is no One to answer my prayers,” she wrote: “So many unanswered questions live within me―I am afraid to uncover them―because of the blasphemy.―If there be God, please forgive me” (Mother Teresa Reference Teresa and Kolodiejchuk2007, 187). Later she wrote:

This was not a once-off occurrence. It was her “traveling companion” for nearly her entire adult life; nevertheless, “she carried on” (157, 326, 336).

How are we to understand this? In contrast with her earlier assessment of herself as having lost her faith, she later came to a different understanding which she described with nine short words: “to live by faith and yet not to believe” (248; contrast with 187 and 193). It is not difficult to see here someone experiencing belief-cancelling doubt―quite understandable given her intimacy with pain and disease, death and suffering―and yet we plausibly also see someone resolved to rely on Jesus to come through, as her Lord, someone perhaps acting on the beliefless assumption, or perhaps even a very low credence, that the basic Christian story is true, and so keeping her vow to serve him in the poorest of the poor despite her fifty-year-long struggle with doubt.Footnote ¹²

Of course, not only exemplars of faith struggle with doubt, so do ordinary people of faith today, for which there is ample anecdotal evidence, e.g., our own experience, the experience of others we know personally, and the experience of those we don’t know personally but whose experience we can access through sources such as a variety of clergy reports and scores of easily googled self-help books that address doubting Christians, none of which would be written unless their intended audience experienced severe doubt. Beyond anecdotal evidence such as this, psychological research on religious struggle and spiritual formation reveals that being a person of faith bears a significant positive correlation to experiencing doubt at some time—whether mild, moderate, or severe—about God’s love, justice, and existence, in addition to many other religious struggles.Footnote ¹³ Kenneth Pargament and Julie Exline (Reference Pargament and Exline2022, 215) write:

Pargament and Exline (Reference Pargament and Exline2022, 217) continue:

Moreover, the most influential theory in developmental psychology of religion, James Fowler’s theory of the stages of faith—which was initially based on interviews of 600 people of faith, and many more since then—recognizes continuing to rely on God in the face of challenges, including severe doubt, as a rite of passage for post-adolescent “mature” faith (Fowler Reference Fowler1981; Seel Reference Seel2012). Further, these bodies of research continue to grow, displaying similar results.Footnote ¹⁴

Clearly enough, Resilient Reliance can accommodate the doubt experienced by exemplars of faith as well as more ordinary people of faith and, just as clearly, Belief-Plus cannot. Nor can Belief-Plus accommodate severe doubt when it shows up in everyday human-human relationships of mutual faith and faithfulness. These are serious strikes against Belief-Plus, in our view.Footnote ¹⁵

Of course, not all readers will find our theory as plausible as we do. We invite them to consider a hypothetical: If the Resilient Reliance theory of faith is correct, how does trust relate to faith?

2. Trust without relying

According to Martin Hollis, “trust is simply a matter of predictability,” as when we trust our apple trees to bear apples, not oranges, or we trust our neighbor to go to the office on weekdays (Reference Hollis1998, 10). Such “predictive” trust “is a simple matter of warranted prediction.” On a second theory, also from Hollis, this time a theory of “normative” trust, for us to trust you for something is for us not just to predict that you will come through; in addition, “[we] believe that you should—or at least that you believe that you should” (11). According to Russell Hardin, for us to trust you to do something is for us to “believe you have the right intentions toward us and that you are competent to do what we trust you to do,” where “the right intentions on your part as a person we might trust are to want to take our interests (and possibly our welfare), as our interests, into account in your actions” (Reference Hardin2006, 17). On Berislav Marušić’s theory, for you to trust someone for something is for you to believe that they will come through simply because they committed to it (Reference Marušić2017, 5).

Suppose one of these theories is correct. Then trust is just a cognitive state of one sort or another. As such, it need not involve relying on the trusted in any way, shape, or form. Thus, on these theories, trust is insufficient for faith since faith necessarily involves a disposition to rely on the faithed.

Even so, trust might be necessary for faith, on these theories. However, while faith necessarily involves some positive cognitive state or other toward the proposition that the faithed come through, each of these theories requires a more specific positive cognitive state for trust, a prediction or a belief. Moreover, Hollis’ second theory, and Hardin’s and Marušić’s theories, require propositional contents none of which faith requires. So, on these theories of trust, trust is not necessary for faith.

Upshot: on these theories, trust is insufficient and unnecessary for faith.

The substantive issue here is whether these theories rightly exclude relying from trust in any way, shape, or form. If they do, then trust is fundamentally different from faith since a disposition to rely is fundamental to faith. What can be said for and against trust involving relying?

We tend to think that just as faith involves a disposition to rely on the faithed, so trust involves a disposition to rely on the trusted, and for reasons similar to those mentioned above with respect to faith. Trust’s disposition to rely explains why trust can be risky, why it can make one vulnerable, why a charge of unjustified lack-of-trust can stick, and what trustworthiness is.

Not everyone agrees.

According to Christopher Thompson, we can see that trust does not involve relying because there are clear cases of trusting someone to do something, but their doing it does not influence your plans, and so you do not rely on them to do it (Reference Thompson2017, 649). For example, imagine that you host a dinner party. You have a well-stocked wine cellar from which you plan to serve your guests. One of them, Dionysius, promises to bring a bottle of Mouton-Rothschild Bordeaux (1970) from his cellar. Naturally enough, you trust him to do as much. However, because you plan to serve wine from your own cellar, what he does forms no part of your plans. Thus, you do not perform the act of relying on him to bring the Bordeaux. That’s trust without relying, says Thompson.

It’s hard to disagree with Thompson’s verdict that you can trust Dionysius to bring the Bordeaux without performing the act of relying on him to do so. However, it seems you cannot trust him to bring it without a disposition to perform that act. To see why, imagine that before Dionysius arrives, another guest asks for Bordeaux and you have none in your cellar. In that case, you will tell them a bottle is on its way, and you will do so because you trust Dionysius to bring it, and so you are disposed to rely on him to do so. This is crucial. For if trust entails a disposition to perform the act of relying and not the act of relying itself, then Thompson’s case is not a counterexample to theories that require the disposition but not the act. Moreover, it is more plausible to posit that trust entails a disposition to perform the act of relying rather than performing the act itself. After all, we can retain our trust in others even when we are asleep or absorbed in something else.

Marušić also argues that trust does not involve relying. Drawing on Pamela Hieronymi’s work, he writes that “trust cannot consist in or involve reliance because reliance, unlike trust, is responsive to practical reasons” (Marušić Reference Marušić2017, 2; cf. Hieronymi Reference Heironymi2008).

As indicated earlier, we agree with the disjunction that relying “is, or involves, action.” However, while it follows from Marušić’s premises―i.e., it follows from (1) relying is responsive to practical reasons and (2) trust is not responsive to practical reasons―that trust is distinct from relying, it does not follow that trust does not involve relying. For suppose that a disposition to rely is responsive to practical reasons but trust is not. Nothing in that supposition precludes the possibility that trust involves both a disposition to rely and a positive cognitive state, e.g., belief, judgment, or credence, among other candidates. So then, consider the following proposition: it is jointly possible that trust involves a disposition to rely and a positive cognitive state, the disposition to rely is responsive to practical reasons, and trust itself is not responsive to practical reasons. That possibility is left open by (1) and (2), and it has trust involving relying. So, it does not follow from (1) and (2) that trust does not involve relying.

Reflecting on our assessment of Marušić’s argument, one might well wonder: If the disposition to rely is responsive to practical reasons, and it is a component of trust, then won’t trust itself be responsive to practical reasons?

At first glance, this seems like a good point to make. However, on closer inspection, it appears to commit the fallacy of composition. It’s like arguing that since the disposition to assert p is responsive to practical reasons, and it is a component of belief that p, it follows that belief that p is itself responsive to practical reasons.

Nothing in our assessment of Marušić’s argument (so far) bears on whether his premises are true. We only supposed they were true, just to see what followed. But what about their truth? In particular, is it true that trust is unresponsive to practical reasons?

Of course, if we bring to that question a prior identification of trust with belief or positive cognitive states more generally, then we may well infer that “[t]he reasons in light of which we trust someone for something are not reasons that show trusting that person worthwhile or a good thing to do” (Marušić Reference Marušić2017, 2–3), although we might wish to resist that inference. (See Rinard Reference Rinard2019 for dissent.) But if we bring to that question a prior identification of trust with a disposition to rely due to a positive stance toward the relied upon coming through, then we may well infer that the reasons in light of which we trust someone for something can sometimes be reasons that show trusting them worthwhile or a good thing to do (because we can have practical reasons to be disposed to rely on them), in addition to theoretical reasons to be in some positive cognitive state toward their coming through, e.g., belief that they are trustworthy.

Hieronymi may well disagree. That’s because, she might say, if trust is responsive to practical reasons, “then one could adopt the attitude insincerely, and we could question the sincerity of the attitude itself (not just its expression). But that does not seem possible” (Heironymi Reference Heironymi2008, 223n16).

This too is a point worth dwelling on. However, on reflection, it seems to be a non sequitur. After all, if trust were responsive to practical reasons and complex in the way indicated above, then when we trust someone for something, we will be in a positive cognitive state toward their coming through―e.g., believe that they will―and so it does not follow that “we could question the sincerity of the attitude itself.”

Marušić might give a different argument for concluding that trust is unresponsive to practical reasons. Although there are practical reasons to get yourself to trust someone for something, we must distinguish reasons to get yourself to trust from reasons to trust. And, when we do that, we will see that the reasons in light of which we trust someone for something are not practical reasons, i.e., “reasons that show trusting that person worthwhile or a good thing to do” (Marušić Reference Marušić2017, 3).

Once again, a long pause is called for. It is difficult to deny the distinction between reasons to get yourself to believe something and reasons to believe it. Moreover, it is difficult to deny that if we regard trust as a belief―e.g., a belief of the sort envisioned by Heironymi or Marušić―then it makes sense to apply the distinction to trust, and the conclusion arguably follows. (But again, see Rinard Reference Rinard2019.) However, if we regard trust as involving both a disposition to rely and a positive cognitive state toward the relied upon coming through, will we be led to conclude that the reasons in light of which we trust someone for something are not practical reasons? Perhaps not. That’s because it will remain open to us to understand the reasons to trust in terms of both practical reasons related to trust’s disposition to rely and theoretical reasons related to trust’s positive cognitive state.

As we indicated above, we tend to think there’s good enough reason to posit that trust somehow involves relying, specifically a disposition to perform the act of relying. Further, we find uncompelling the only reasons we know of to suppose otherwise. So, in what follows we will assume that trust somehow involves a disposition to rely. But we may well be wrong. Maybe Thompson, Marušić, and Heironymi are right. If so, trust is very different from faith since faith fundamentally involves a disposition to rely while trust does not.

3. Trust just is a disposition to rely

You trust your friend to keep your secrets. In doing so, you are disposed to rely on them to keep your secrets. And that’s what trust is. Trust is neither more nor less than a disposition to rely. End of story. Period. Or so Bernard Williams says. Trust, he writes, is “the willingness of one party to rely on another to act in certain ways” (Reference Williams2002, 88). On Williams’ theory, we have a tidy answer to our title question: although trust does not entail faith since faith involves more than trust, faith entails trust since faith necessarily involves a disposition to rely and that’s what trust is.

Not so fast, most trust theorists say. Even though trust involves a disposition to rely, a disposition to rely is insufficient for trust. This is the substantive issue that stands between Williams and most trust theorists. Let’s look into the matter briefly.

Why should we suppose that a disposition to rely is insufficient for trust?

The most commonly cited reason is that there are clear counterexamples. For example, the conman is disposed to rely on his mark to send the check, but he doesn’t trust them; the villagers are disposed to rely on the food poisoner to be deterred by the new security system at the grocery, but they don’t trust him; Jack is disposed to rely on the car to start, but he doesn’t trust it; Kant’s neighbors were disposed to rely on him to set their clocks, but they didn’t trust him; etc.Footnote ¹⁶

We get different accounts of what is missing in each case, but Richard Holton’s influential response proves illustrative for our purposes: “When you trust someone to do something, you rely on them to do it, and you regard that reliance in a certain way: you have a readiness to feel betrayal should it be disappointed, and gratitude should it be upheld,” where this readiness to reactive attitudes results from adopting “the participant stance” toward the trusted (Reference Holton1994, 67). The conman doesn’t adopt the participant stance with his mark. Consequently, although he’ll feel disappointed if they don’t send a check, he won’t feel betrayed; and, although he’ll feel satisfied if they send a check, he won’t feel gratitude. He lacks a readiness to reactive attitudes. So, he doesn’t trust his mark to send the check even though he’s disposed to rely on them to do so. Mutatis mutandis, the same goes for the villagers, Jack, and Kant’s neighbors.

Do such cases reveal that readiness to reactive attitudes is essential to trust? Perhaps not, Williams might argue.

He might begin: while it is doubtless true that there are occasions when you trust someone to do something and you also have a readiness to reactive attitudes, it does not follow that your readiness partly constitutes your trust. After all, you have that readiness to reactive attitudes because you suppose they acquiesce, or would acquiesce, or should acquiesce, to your relying on them to do it. However, it is that supposition—without which you could not adopt the participant stance toward them—which generates your readiness to reactive attitudes, not your trust, which is nothing but your disposition to rely on them. Thus, the conman, the villagers, Jack, and Kant’s neighbors all trust. They just lack a readiness to reactive attitudes because they have not adopted the participant stance toward the trusted, and they have not adopted that stance because they have not supposed that the trusted acquiesces, or would acquiesce, or should acquiesce to their relying on them in the ways they do.

Williams might continue: the source of that supposition is as varied as the environments in which we live, and it might be only tacit, e.g., ensconced in the social roles occupied by those we rely on. By way of illustration, you might suppose that a particular individual is a police officer, and so suppose that they have acquiesced to citizens such as yourself relying on them in that role and that they will, therefore, protect and serve you per the oath they took when they voluntarily took on that role. If you trust that individual as a police officer, and so rely on them in that capacity, but they don’t come through, you rightly feel betrayed. But it doesn’t follow that the trust you place in them includes in itself your disposition to feel betrayed. Your trust in them as a police officer is neither more nor less than your disposition to rely on them in that role, and your readiness to feel betrayed is something else entirely external to it: namely, an appropriate response to your supposition that they have acquiesced to the norms of the role they occupy.

Williams might continue further: it might even be that, in some cases, the source of the supposition is not the norms of a social role such as being a police officer. For example, you might simply trust a particular individual as a person. Naturally, given that they are a person, you will suppose they would or should acquiesce to your relying on them as a person (even if they have never done so explicitly) and that they will, therefore, refrain from knifing or robbing you, among other things, per the norms that apply to them as a person. Consequently, if you rely on them in that capacity and they violate those norms, you rightly feel betrayed. But again, it doesn’t follow that the trust you place in them as a person involves in itself your readiness to feel betrayed. The trust you place in them as a person is nothing more than a disposition to rely on them as a person, and the readiness to feel betrayal is external to it: a fitting response to your supposition that they would or should acquiesce to the norms of being a person.

So it is, Williams might argue, that he can explain the normative phenomena Baier, Holton, and others point to without positing that the readiness to reactive attitudes resulting from adopting the participant stance partly constitutes trust itself.

Another objection to Williams’s identity theory appeals to trust’s relation to other things we care about: trustworthiness, distrust, and trust’s role in our lives. Trust is related to these things in such a way that if trust were just a disposition to rely, trust would not be so related, and so we should not identify trust with that disposition.

However, Williams has a reply ready to hand, for, on his identity theory, trust is related to trustworthiness in exactly the way that if trust were a disposition to rely, it would be so related; specifically, the way in which a disposition to rely is related to reliability. For if trust is a disposition to rely, then trustworthiness is nothing but worthiness-to-be-relied-upon, and that’s just reliability. Similarly, on his theory, trust is related to distrust in exactly the way that if trust were a disposition to rely, it would be so related; specifically, the way in which a disposition to rely on someone for something is related to the disposition to not rely on them for it. Further, on his theory, trust plays a role in our lives in exactly the way that if trust were nothing but a disposition to rely, it would play that role; specifically, the role of enabling us to rely on others for things we care about, things that, absent our disposition to rely, we are less likely to enjoy.

Suppose we grant the plausibility of these replies on behalf of Williams. We may then wonder why we should not identify trust with a disposition to rely. Here’s our answer in brief. We argued in s ection 1 that, a mere disposition to rely is not enough for faith because faith involves a positive stance toward the faithed coming through in addition to a disposition to rely. That argument, mutatis mutandis, applies to trust. Upshot: contra Williams, we should not identify trust with a disposition to rely.

In what follows, we take it that Williams’s view is incorrect. But we may well be wrong. Maybe Williams is right. If so, trust does not entail faith, but faith entails trust since faith involves a disposition to rely and that’s what trust is.

4. Trust as relying plus normative expectations and readiness to reactive attitudes

We now turn to theories of trust according to which trust involves not only relying but normative expectations and readiness to reactive attitudes.Footnote ¹⁷ For example, according to Paul Faulkner, for you to (“affectively”) trust someone to do something is for you to rely on them to do it and for you to think both that they should see you relying on them to do it as a reason to do it and that they should do it for this reason, where thinking these things brings with it a “susceptibility to those reactive attitudes that characterize trust’s betrayal” (Faulkner Reference Faulkner2007, 882–83; Reference Faulkner2014, 1978). Annette Baier’s early theory says that for you to trust someone for something is for you to value their coming through and to rely on their goodwill to do so, and for you to be disposed to feel betrayal, or let down, if they don’t come through (Baier Reference Baier1986, 234-36). As for Holton, for you to trust someone for something is for you to rely on them to come through and for you to “have a readiness to feel betrayal should [your trust] be disappointed, and gratitude should it be upheld,” a readiness grounded in “the participant stance” (Holton Reference Holton1994, 67). Later-Baier has it that for you to trust someone for something is for you “to give discretionary powers” to them, “to let them decide” whether your welfare is best advanced by relying on them to come through (Baier Reference Baier1991, 117). Katherine Hawley says that for you to trust someone to do something is for you “to believe that she has a commitment to doing it, and to rely upon her to meet that commitment,” from the participant stance (Hawley Reference Hawley2014, 10, 7). While Faulkner’s theory makes normative expectations explicit, the dispositions to reactive attitudes in the theories of early-Baier, Holton, and Hawley, and the giving-and-letting of later-Baier, arguably make sense only against a background of normative expectations.Footnote ¹⁸

Note that on these five theories trust involves normative expectations and dispositions to reactive attitudes. However, faith need not involve either. So, if trust is what these theories say it is, then trust is unnecessary for faith. Moreover, on these theories, trust need not involve resilience in the face of challenges to relying on the trusted to come through. However, faith necessarily involves resilience, so, if trust is what these theories say it is, then trust is insufficient for faith.

Upshot: given these theories, trust is neither necessary nor sufficient for faith.

However, while one may grant that on these theories trust is not sufficient for faith, one might be puzzled by our reason to think that on these theories trust is not necessary for faith. After all, like trust, doesn’t faith involve normative expectations and dispositions to reactive attitudes?

This is a good point, and we acknowledge that some theorists might take this route. But we prefer a different path. That’s because we see no good reason to posit that faith or trust necessarily involve normative expectations or dispositions to reactive attitudes. Of course, such expectations and dispositions frequently accompany faith and trust, but it seems that, as indicated in our discussion of Williams, we can explain their presence by positing implicit or explicit contextually sensitive suppositions about those in whom we place our faith and trust―specifically, that they have acquiesced, or would or should acquiesce, to our placing our faith and trust in them―and we can do that without positing that those expectations and dispositions partly constitute faith and trust themselves.

5. Trust, faith, and resilience

It is worth emphasizing that, if any of the theories of trust canvassed so far is correct, then, since none of them entails resilience, only faith and not trust involves resilience. However, some trust theorists disagree with their peers on this score.

Judith Baker distinguishes three “kinds of trust,” only one of which she takes to involve resilience, which she calls “special trust, or friendship trust.” Such trust involves a disposition to “hold beliefs in the face of counter-evidence,” as when we trust that a friend is innocent of a criminal charge because she told us they are innocent, despite “an impressive amount of evidence brought against her”; “I am biased in favor of my friend, in favor of her innocence… . I am committed to her being innocent” (Baker Reference Baker1987, 3, 10). We can put the theory succinctly:

Since it is arguably impossible to fully believe our friend is innocent because they said so without being disposed to rely on them to tell us the truth about it, that disposition is implicit in the theory.

Three observations. First, on Baker’s view, special trust necessarily involves a disposition to resist counterevidence, but generic trust does not, and so generic trust need not involve such a disposition. Second, if special trust is distinct from generic trust in the way indicated, then, even though generic trust need not involve a disposition to resist counterevidence, special trust shares something important in common with faith: resilience in the face of evidential challenges. Third, having gone this far, we might set our sights further: we might identify faith as a species of generic trust as follows. For you to generically trust someone for something is for you to be disposed to rely on them to come through with respect to it because of your positive stance toward their coming through, while faith adds resilience in the face of challenges to relying on them to the dispositional output, whether those challenges are evidential, which is Baker’s concern, or emotional, social, or something else besides. This would provide a different answer to our title question: faith entails trust, but trust doesn’t entail faith since faith is a species of trust, the species distinguished from others by necessarily involving resilience.

Some trust theorists, however, want to go further. They affirm that trust—not just one of its species, but generic trust—necessarily involves resilience in the face of challenges, notably evidential ones. On Karen Jones’s early theory, for you to trust someone for something is for you to be optimistic about their goodwill and competence as it extends to your counting on them to come through, and for you to expect that they will be directly and favorably moved by the thought that you are counting on them where that optimism involves an affective attitude, not merely a cognitive state (Reference Jones1996, 4). On Jones’s later theory, for you to trust someone for something is for you to have an affective attitude of optimism that their competence and responsiveness to your dependency will cover it (Reference Jones2019, 958; cf. Reference Jones2017, 99).Footnote ¹⁹ Crucial to both theories is her understanding of the optimism involved in trust. According to Jones, trust “essentially” involves “an affective attitude of optimism about the goodwill [and competence] of another” (early theory) or about their competence and responsiveness (later theory). She identifies that optimism with a certain kind of emotion, one that involves “a distinctive way of seeing” the trusted, which results from a “cognitive set” (Reference Jones1996, 11). Cognitive sets are mental states that collectively play “at least the following cognitive roles: they (1) focus attention, (2) direct inquiry, (3) shape interpretation, (4) structure inference and (5) shuffle action options in a hierarchy of perceived salience and desirability” (Jones Reference Jones2019, 958). As a result, trust “restricts the interpretations of another’s behavior and motives that we consider. It also restricts the interpretations we will consider as possibly applying to situations and the kinds of inferences we will make about the likely actions of another” so that trust “functions analogously to blinkered vision: it shields from view a whole range of interpretations about the motives of another and restricts the inferences we will make about the likely actions of another” (Reference Jones1996, 11–12).Footnote ²⁰ So it is that on Jones’s theory, trust in general—and not just some species of trust—necessarily involves resilience in the face of evidential challenges to counting on the trusted.

If Jones is right, then we cannot distinguish faith from trust by faith’s resilience. We might go further, for if we allow that trust involves a disposition to resist challenges to relying or counting on the trusted, that sounds a lot like faith’s disposition to resilient reliance. Further, if we allow that trust involves “two elements,” “one cognitive and one affective or emotional,” that sounds a lot like faith’s positive cognitive and conative states. We may well conclude that the differences between Jones’s theory of trust and our theory of faith amount to different ways to dot an i and cross a t. Of course, that’s not to say that the differences are unimportant. Rather, it’s just to say that the posited components of trust and faith resemble each other very closely, so much so that our taste for desert landscapes might lead us to refrain from regarding them as distinct psychological states. In this frame of mind, we may find the appetite for another, and more parsimonious, answer to our title question: trust is related to faith by numerical identity.

What should we make of this suggestion? We’ve been led to it via Jones’s claim that trust—generic trust, trust in general, trust per se—“essentially” involves a disposition to resist counterevidence, a disposition grounded in trust’s optimistic cognitive set. Is Jones right?

An initial indication that trust does not necessarily involve a disposition to resist counterevidence comes from what appear to be cases of trust in which we trust someone for something even though we are not the least bit disposed to resist counterevidence to them coming through, e.g., our trust in strangers for accurate directions, our trust in drivers to stay on the opposite side of the road, our trust in our pharmacist to provide our medication, etc. In such cases, it seems we need not be the least bit disposed to resist counterevidence. If we get evidence that they gave us the wrong directions, we typically don’t resist it; we simply ask someone else. If we get evidence that they aren’t going to stay on their side of the road, we don’t resist it; we simply swerve out of the way. If we get evidence that they don’t have our preferred cold medication, we don’t resist it; we just go to the pharmacy down the street. Of course, that’s not to say that there are no cases of trust in which it disposes us to view counterevidence through rose-colored glasses. It’s only to say that the cases to which we have drawn attention suggest that such a disposition is unnecessary for trust.

Here Jones may well want to put her foot down. After all, if trust “essentially” involves an affective attitude of optimism with a cognitive set that plays the five roles she describes, it follows that trust necessarily involves a disposition to resist counterevidence, contrary to what we have suggested. This is a good point. It occasions a deeper inquiry into Jones’s influential theory.

The first thing we’d like to suggest is that even if a cognitive set tends to cause such a disposition, it seems to be only contingently related to it. After all, what if we knew of the biasing effects that our optimistic cognitive sets can produce, and what if we aimed to avoid those effects—e.g., by training ourselves to be alert to them and to correct for them—and what if we achieved our aim? In that case, our bias-alert-correction system would accompany our trust-constituting cognitive sets, and it would eliminate any disposition to resist counterevidence we would have otherwise had.

But a more important question about Jones’s theory comes to the fore: Does trust “essentially” involve an affective attitude of optimism grounded in a cognitive set?

Perhaps not. For suppose we give our friend money to purchase the books for our reading group. Naturally enough, we want her to purchase them, and suppose we expect that she will do so given our high regard for her competence and her positive response to our counting on her to buy them. Why shouldn’t that be enough for us to trust her to buy the books? Why must we have, in addition, an affective attitude of optimism grounded in a cognitive set? After all, we’re counting on her to buy the books because we want her to buy them and we expect her to do so given our high regard for her competence and her positive response to our counting on her. Our counting on her because of our positive stance seems enough; an affective attitude of optimism grounded in a cognitive set seems unnecessary. To be sure, we might have counted on her because of an affective attitude of optimism grounded in a cognitive set. That seems enough too. We only suggest it isn’t necessary, it isn’t “essential”. We might also trust absent that particular way in which we can be in a positive stance.

Upshot: we should not identify trust with faith. For, although faith essentially involves a disposition to overcome challenges to relying on the faithed, trust does not essentially involve a disposition to overcome challenges to relying on the trusted.

6. The value of faith

Oftentimes, when we place our faith in someone—e.g., when we place our faith in someone as a spouse—we bind ourselves to them in a particular way and, when all goes well, our faith in them is met by their faithfulness to us as a spouse; moreover, their faith in us as a spouse is met by our faithfulness to them in that capacity, which enhances the bond between us. We find such relationships of mutual faith and faithfulness not only between spouses, but also between lovers and between friends; between parents and their children; between siblings and between other relatives, e.g., spouses and their in-laws; between caretakers and dependents, teachers and students, employers and employees; between associates, teammates, and soldiers; and between people and their gods or God (if such there be).

If the bonds created by mutual faith and faithfulness were easily dissolved when troubles arise, the fulfillment we enjoy by virtue of being in these relationships would be even more fragile and fleeting than it is. Imagine coming to rely on someone as a new friend only to discover that they are disposed to stop coming through in that capacity given the least bit of difficulty in doing so. Unfaithful is he that says farewell when the road darkens. But unfaithfulness is not the only way in which a relationship can suffer. Imagine someone coming to rely on you as a new friend only to discover that they are disposed to stop relying on you in that capacity given the least bit of difficulty in doing so. Faithless is he that says farewell when the road darkens. Faithlessness every bit as much as unfaithfulness can harm a relationship.Footnote ²¹

It is precisely because both faith and faithfulness dispose us to be resilient in the face of challenges to relying on and coming through for each other that relationships of mutual faith and faithfulness can help to provide both the goods that naturally arise from those relationships, e.g., the goods of marriage, friendship, collegiality, etc., as well as the stability from which we can find the security and support we cherish in such relationships, e.g., the security we need to raise a family or run a farm together, or the support we need to explore our own personal aspirations and long-term projects. Like faithfulness to another, faith in them helps to keep us in such relationships. That is its role: to enable us to continue to rely on others in the face of evidential, emotional, social, and other challenges that might otherwise undermine the relationship. And therein lies its potential value too.

Trust does not play this role. For, as we observed, we sometimes trust people without being at all disposed to continue relying on them in the face of evidential or other challenges to doing so. The answer to our title question that we favor, then, is that only faith, and not trust, disposes us to overcome challenges to relying on someone for something that matters to us. As such, only faith, and not trust, has what it takes to keep us relying on our spouses, partners, friends, colleagues, and others in the face of challenges to doing so. As such, only faith, and not trust, has what it takes to facilitate the stability out of which the security and support we need can flower.