|
|
|
Universidade Federal de Santa Maria
Voluntas, Santa Maria, v. 16, n. 2, e93794, 2025
Submissão: 23/09/2025 • Aprovação: 09/10/2025 • Publicação: 31/10/2025
3 TRUST AS A COMPLEX TWO-LEVEL EMOTION
Artigos
Trust as a Social Emotion
Confiança como uma emoção social
Leonardo de Mello Ribeiro I
I Universidade Federal de Minas
Gerais, Belo
Horizonte, MG, Brasil
ABSTRACT
The development of trusting relations is one of the central aspects of human sociality. Trust makes it possible for people to count on and cooperate with each other, creating the conditions for people to achieve goods and promote their interests and well-being. Yet, justifying trust can be a vexing task. This paper offers a proposal on the nature of trust and its possible justification that aims to combine both cognitive and emotional aspects into a unified account, seeking to reconcile two opposing tendencies in the literature, namely, rationalist and emotional views. Trust should be understood as a normative relation of a special sort and a combination of restricted rationality and emotional “amplification”. More specifically, it involves a complex two-level social emotion that plays a dual role: it partly responds to the truster’s available evidence about the trustworthiness of others, but also goes beyond this evidence by expressing a form of optimism that never eliminates vulnerability and risk. Thus, trust is a kind of practical optimism and openness cultivated within social environments. As such, the only possible justification for trust will be ultimately “subjective”.
Keywords: Trust, Emotion, Society, Cooperation, Reasons for action
RESUMO
O desenvolvimento de relações de confiança é um dos aspectos centrais da sociabilidade humana. A confiança possibilita que as pessoas contem e cooperem umas com as outras, criando as condições para que alcancem bens e promovam seus interesses e bem-estar. Porém, justificar a confiança possa ser uma tarefa complexa. Este artigo oferece uma proposta sobre a natureza da confiança e sua possível justificação que busca combinar aspectos cognitivos e emocionais em uma visão unificada, conciliando duas tendências opostas na literatura, a saber, racionalistas e emocionais. A confiança deve ser entendida como uma relação normativa de um tipo especial e uma combinação de racionalidade restrita e "amplificação" emocional. Mais especificamente, ela envolve uma complexa emoção social de dois níveis, com um papel duplo: em parte, responde às evidências disponíveis daquele que confia sobre a confiabilidade dos outros, mas também vai além dessa evidência ao expressar uma forma de otimismo que nunca elimina a vulnerabilidade e o risco. Assim, a confiança é uma espécie de otimismo prático e abertura ao outro cultivada em ambientes sociais. Como tal, a justificação possível para a confiança será, em última análise, "subjetiva".
Palavras-chave: Confiança, Emoção, Sociedade, Cooperação, Razões para agir
The development of trusting relations is one of the central aspects of human sociality. Trust serves as a kind of “social glue” that makes it possible for people to count on and cooperate with each other. It creates the conditions for people to achieve goods and promote their interests and well-being. Arguably, without trust, there would not be human relations and society as we recognize and experience them[1].
Justifying trust is one of the cruxes of social life, yet it can be a vexing task. Entire generations of philosophers, social scientists, jurists, and psychologists have attempted to do so, but have been far from reaching full agreement. This paper does not have the ambition of filling this gap, but it hopes to contribute to the debate by offering a proposal on the nature of trust and its possible justification that aims to combine both cognitive and emotional aspects into a unified account[2]. In doing so, it also seeks to reconcile two opposing tendencies in the literature on trust, namely, rationalist and emotional accounts.
Contrary to pure rationalist accounts of trust, the view defended here rejects that trust can be entirely explained and justified in rational terms. Contrary to common affective/emotional accounts of trust, it rejects that trust is entirely non-rational or non-normative and is similar to blind faith. Trust should be understood as a normative relation of a special sort and a combination of restricted rationality and emotional “amplification”. It involves a kind of practical optimism and openness cultivated within social environments. However, it lacks either full rational justification (in the sense of being justified by “non-subjective” practical reasons) or justification on its own (in the sense of bearing basic or irreducible normative entitlement). We will argue that trust involves a complex two-level social emotion that plays a dual role: it partly responds to the truster’s available evidence about the trustworthiness of others, but also goes beyond this evidence by expressing a form of optimism that never eliminates vulnerability and risk. As a result, the only possible justification for trust will be ultimately “subjective”, that is, entirely based or dependent on the truster’s own optimism.
2.1 The assumptions
Trust is widespread in human relations and human experience. As Annette Baier vividly and unmistakably once said:
Trust, the phenomenon we are so familiar with that we scarcely notice its presence and its variety, is shown by us and responded to by us not only with intimates but with strangers, and even with declared enemies. We trust our enemies not to fire at us when we lay down our arms and put out a white flag. (…) We often trust total strangers, such as those from whom we ask directions in foreign cities, to direct rather than misdirect us (…). We put our bodily safety into the hands of pilots, drivers, doctors, with scarcely any sense of recklessness[3].
Trust is a broad social phenomenon ranging from the personal level to the large-scale impersonal social level. So, a first step needed is to clarify that we will be talking about a subset of those trusting relations.
Arguably, trust occurs in its strongest form in close, affect-laden relationships. Some take it as essential to parental, friendship and other intimate relations. Some even argue[4] in favor of a developmental thesis according to which large-scale social relations of trust (including those between strangers and in impersonal settings) are the consequence of, and to be modelled on, the development of “natural dispositions” to trust found in those close relations. However, our discussion will not be developmental. Nothing we say here depends on such an assumption, although parts of our account may turn out to be compatible with that hypothesis. Moreover, trust in some of those intimate contexts may be indistinguishable from blind faith. But it is part of the thesis advocated here that trust is not blind faith. So, if there is some true developmental story that connects the “natural dispositions” of blind faith characteristic of intimate relations with large-scale social trust, there must also be some additional story about how the former transforms into the latter.
It is trust in large-scale social relations (including those between strangers, impersonal relationships and relations mediated by social structures) where the explanation and justification of trust as a social phenomenon becomes more challenging. This is the sense in which trust is fundamentally involved in social cooperation. Our focus here will be on that large-scale socially cooperative sense of trust. Nonetheless, trust as discussed here is limited to relations between persons[5].
In addition, trust should not be interpreted as a moral relation in itself. This is so not only in the trivial sense that there may be trust among peers who pursue ends contrary to what is morally prescribed, but also in the deeper sense that one’s trusting attitude does not by itself bind the trusted party. Although we cannot fully argue for this point here, it seems reasonable to suppose that a trusting attitude does not bind in the state of nature. A common opposing view treats trust as similar to a promise[6]. However, not only does it make perfect sense to ask whether one should trust a promise, but trust is also significantly different from a promise in that a promisor makes the original move of “giving their word” to the promisee. The promisor, by the very act of promising, commits themselves to act as they have promised to the promisee. This is what confers weight to the idea that promises are binding even in the state of nature—though validating that claim would require more than simply stick to that intuition. But nothing like that seems to hold of trust. In trust, the truster by the very act of trusting cannot commit the other to be trustworthy. They can only commit themselves to trust from their own perspective, irrespective of what the trusted party is disposed to do. Trust entails accepting to be dependent on and vulnerable to the other. As we will see, this is part of the peculiar nature of trust.
However, this is not to say that trusting relations cannot become moralized or normative through social norms, conventions, practices, and enforcement by law. Indeed, our typical apprehension of trust, as illustrated in the above passage by Baier, carries plenty of moral and socially normative connotations. This means that one of the central questions of this debate—namely, whether trust is a precondition for cooperation and society or the other way round—receives a twofold answer. As Gambetta[7] points out, it seems to make sense to say both that any socially cooperative engagement needs trust to occur and that social cooperation (supported by social structures) fosters and increases trust.
2.2 The problem of cooperation
The problem of cooperation, for which trust can be deemed a solution, is the following. It is a normative practical problem. One wants to promote their interests, well-being or secure goods and, to achieve that, very often (almost always) needs the cooperation of others. Others have equivalent interests and can benefit from the same cooperative scheme. But, at least in some circumstances, they can also benefit from cheating. Why should one trust another?
Hardin[8] illustrates what is at stake here by describing a case taken from Dostoevsky’s The Brothers Karamazov. We are told the story of a lieutenant colonel who manages money on behalf of the army. The colonel develops a mutually profitable relationship with the merchant Trifonov, lending him money while Trifonov does business with it. Trifonov then returns the whole amount to the colonel adding to it a gift as well as interest on the loan. When the colonel is about to leave the post, he asks Trifonov to return the last sum he lent him, but hear from Trifonov that he never received any money from him.
The relationship established between the colonel and Trifonov can be described as a mutually self-interested trusting relation, with the only aim of sheer gain for both sides. In addition, as Hardin points out, it is a relation not “backed by the law of contracts”[9]. This means that the common social structures for enforcing and protecting social cooperation are absent. While their relationship was mutually profitable and both benefited from it, they shared a mutual interest in maintaining it. But, when the relationship was about to end, and Trifonov saw an opportunity to keep all the money for himself, he betrayed the colonel. As Hardin puts it, the end of the relationship changed the equilibrium of incentives, since “the incentive to the next mover was clearly to withdraw from the cooperation. It was the lieutenant colonel’s misfortune that Trifonov was the next mover at the end”[10].
Although this was a secret relationship, occurring without the usual protection of social structures, it illustrates nicely how trust works as well as its potential risks. Trust can be fostered by social structures insofar as these structures provide protection and incentives for social cooperation. In the absence of such protection and incentives, trust may be expected on the assumption of moral and informal social norms—as the colonel might have trusted by presuming gratitude from Trifonov. The main point, however, is that, in trust, people are always to some extent vulnerable to those on whom they have decided to depend. Let us now develop this point in more detail.
2.3 The elements of trust
Trust is generally seen as having both social and psychological aspects. Socially, it refers to an assumed reliability between individuals that enables them to cooperate and sustain cooperation as their interaction continues. As Baier[11] points out, “[a]ny form of cooperative activity, including the division of labor, requires the co-operators to trust one another to do their bit”. Similarly, but taking it a step further, Lewis and Weigert[12] argue that trust is “a functional prerequisite for the possibility of society”, with its absence leading to “chaos and paralysing fear.” The idea here is that human social interaction and cooperation, given its complexity and associated risks, would lack the conditions for development and stability without trust. Trust is, thus, something socially needed.
As part of that social dimension, we must also recognize that trust does not simply occur between two isolated individuals, detached from their broader social contexts, background, and histories. Trust always occur within concrete social situations, involving networks of relationships, shared social norms, individual and group histories, as well as the social structures and institutions that organize societies. These factors influence how individuals view themselves in relation to others.
From a psychological perspective, trust is about having positive expectations concerning the behavior of others—including their competence in fulfilling what they are trusted for—in a way that involves an imbalance or asymmetry of power. To trust someone is to place oneself in a position where one’s interests and well-being depend on the choices and actions of another, with the ever-present potential for betrayal. Giddens[13] describes trust as “a particular type of confidence”, but one that inherently carries the risk of frustration due to the possibility of defection by the trusted party. Trust, thus, involves uncertainty, vulnerability and risk.
Some call this the “ambivalence” of trust. In trusting others, one acts as if the situation were unproblematic, although one cannot be certain about how the trusted will choose and act. As Möllering[14] puts it, in the face of uncertainty and risk, people trust nevertheless. In this way, trust helps manage uncertainty in cooperation, but never eliminates its risks. Luhmann[15] captures this well, stating that “no decisive grounds can be offered for trusting; trust always extrapolates from available evidence; it is (…) a blending of knowledge and ignorance”.
The sort of dependence that trust involves can be described as dependence of the truster on the choices and actions of the trusted, and the latter’s recognition of the former’s dependence. Thus, some authors[16] emphasize that trust involves the expectation that the trusted will take the interests of the truster to heart, that is, that the trusted will recognize the truster’s condition of dependence as a reason to act. This means that the relation of trust is not to be explained solely in terms of mere compatibility or convenience between the interests of truster and trusted, in the sense they their interests happen to coincide. As Hardin[17] notes, mere compatibility of interests and continuity of a relationship are insufficient for characterizing trust, because trust should involve some form of mutual commitments.
For similar reasons, authors such as Baier[18], Jones[19] and Faulkner[20] claim that trust is not mere reliance on others, in the sense of simply predicting the behavior of others on the basis of factors “external” to the trusting relation. We may rely on other people—that is, in their behavior not representing a threat to us and even promoting our interests—because we can predict their behavior as reliably motivated by individual and social conditions. All this is compatible with their having, as Baier[21] says, “ill will” toward us. This sense of reliance is involved, for example, in the prediction of shopkeepers’ behavior given their interests in not mistreating their clients or in the prediction of how someone will behave out of fear of retaliation. Actually, there are many other human motives reliable enough to justify expectations of reliance: vanity, habit, social status, etiquette, etc. However, none of these, according to some authors, are cases of trust, since they need not involve any recognition on the part of the trusted of the truster’s dependence on them as their reason to act. For the time being, let us accept these general distinctions, although we will later find reasons for revising some of them.
2.4 Theories of trust
There is a well-established tradition that analyzes trust through rational decision-making models, often drawing from frameworks like expected utility maximization[22], (Bayesian) subjective probabilities[23] and game theory[24]. Although diverging in the details, those approaches share the view that trusting someone involves a cognitive process in which an individual assesses the trustworthiness of others using available evidence to inform rational decision-making. The truster assigns probabilities, based on this evidence, to the likely actions and choices of a potential partner. For instance, Gambetta[25], a proponent of this approach, argues that trust can be understood as the truster's probabilistic prediction (even if precarious) that potential social partners (trustees) will align with the truster's interests, thereby acting in a way that meets the truster’s expectations.
Hardin[26], in a slightly different fashion, emphasizes that those who are trusted have incentives to be trustworthy because they can individually benefit from the trusting relation, and because there are social structures and institutional mechanisms that discourage betrayal (e.g., sanctions, social ostracism, reputational damage). According to Hardin, one trusts another because the trusted party has an interest in fulfilling the trust and maintaining the cooperative scheme. This is what Hardin calls “encapsulated interest”. In his words:
I trust you because I think it is in your interest to take my interests in the relevant matter seriously in the following sense: You value the continuation of our relationship, and you therefore have your own interests in taking my interests into account[27].
Let us call those rationalist models ‘pure’, given that they think the explanation and justification of trust can be given entirely in rational terms.
There have been numerous objections to those “pure” rationalist accounts, which we cannot examine in detail here[28]. However, it is worth noting that most critiques share a common point: pure rationalist accounts tend to oversimplify the problem of trust, glossing over its complexity. Those accounts frame trust merely as a matter of being rationally justified in believing someone to be trustworthy. For Hardin, trust does not add anything to trustworthiness. Trust is simply the mental state—namely, belief—that is the cognition of trustworthiness. Appropriate trust amounts to being rationally justified in believing in someone’s trustworthiness[29]. Thus, the objection goes, even when pure rationalist accounts acknowledge that there is always an element of risk and uncertainty in trusting, they treat this as a “calculated risk”, simply as probabilities assigned to the potential actions of others.
However, this rationalist perspective fails to make sense of what, arguably, are the main characteristics of trust: vulnerability and risk. These features are directly related to rational and cognitive limitations—such as incomplete information and constraints of time—that render the truster’s expectations restricted and uncertain. As Luhmann notes, “whether action on the basis of trust has been right, therefore, in the final, retrospective, reckoning, depends on whether the trust has been honoured or been broken”[30]. Thus, the “final” justification for trust can only be retrospective and beyond the truster’s control. In this respect, trust does not seem align completely with the typical “objective perspective” of rational decision theory. The uncertainty inherent in the dynamic process of social interaction characteristic of trust cannot be fully captured by probabilistic assessments or any other sort of rational calculus, even when these are not understood purely in formal terms. More specifically, trust seems to go beyond the epistemic reasons one can gather in assessing others’ trustworthiness. Because epistemic evidence of trustworthiness is always limited, trust must extrapolate beyond “calculated” or “projected” risk. Trust involves placing oneself in a risky situation where one becomes dependent on and vulnerable to the choices and actions of others.
Even Hardin’s account, which can be read as a more moderate and sophisticated version of pure rationalist accounts, falls prey to that objection. One of the advantages of Hardin’s account is that it employs a merely informal notion of rational probability assessments, which is meant to accommodate the idea that social cooperation is, on most occasions, worthwhile from the perspective of individuals. According to Hardin, the structures of society provide protection and incentives for cooperation, making such cooperation generally beneficial. He, thus, seems to be simply assuming that, if all the parties are rational in most socially cooperative engagements, they will create commitments as truster and trusted, and will benefit from that by satisfying their mutual expectations and interests. The possibility of betrayal or defection is to be interpreted in terms of irrationality of one of the parties—either because the truster is careless, negligent or insensitive to evidence about someone’s trustworthiness or because the trusted is exploitative, manipulative or insensitive to the truster’s interests. Thus, all such irrational moves are to the detriment of both truster and trusted. However, the problem is that this presumption of mutual rationality and benefit is a bold assumption made by Hardin. Not only we cannot rely on it in concrete, real-world situations, but it is also a controversial claim that would require further defense[31].
In response to those limitations of pure rationalist accounts, several authors have proposed alternative views and different ways of characterizing trust and its associated mental state. For obvious reasons, we cannot cover all those responses here. But going through aspects of some of them will be relevant for what follows.
For some authors, trust is described as involving, from the perspective of the truster, a non-cognitive attitude[32], an affective/emotional component[33] or even sheer faith[34]. Others emphasize, from the perspective of the trusted, their motivations, such as goodwill[35], appropriate consideration for the truster[36], or a practical normative response[37].
Although they differ in significant ways, all those views have in common the fact that they reject that trust can be entirely based on rational assessments of trustworthiness. Some of them hold that trust extrapolates beyond those assessments, others say that those assessments are only partly explanatory and still others claim that trust is not rational at all. This means that, if there are reasons for trusting, those reasons go beyond the reasons one can offer for believing someone trustworthy. The basic idea behind this has a double dimension: not only are those reasons always incomplete and uncertain, but they are also inappropriate to explain and justify trust.
All of them also emphasize the fact that the truster expects the trusted to recognize their dependence and vulnerability, taking this consideration as their reason to act. Although Hardin can also make sense of something akin to this idea, those accounts interpret the appropriate response to dependence and vulnerability in a different and more robust way. Hardin’s “encapsulated interest” model allows him to distinguish trust from merely coincident interests and from mere reliance on others, as we have seen. So, Hardin’s account can find room for the idea that the trusted takes the truster’s interests as a reason to act, that is, the trusted recognizes the truster’s dependence and vulnerability and, in fulfilling the trust, they act for that reason. Still, Hardin sees this relation, at the end of the day, as instrumental: the trusted fulfills the trust because they benefit from maintaining the relationship with the truster.
So, while Hardin ultimately explains that in terms of instrumental reasons—that is, in terms of mutually beneficial relationships between self-interested individuals—those alternative views reject that the dependence and vulnerability characteristic of trust can be taken as a reason to act in an instrumental way. The emphasis some of those accounts place on the concept of goodwill and appropriate consideration for the truster is revealing[38]. Accordingly, by counting on the goodwill of the trusted, the truster also expects that the trusted will take their dependence and vulnerability as intrinsic reasons to act[39]. That is to say, the trusted should have intrinsic consideration for the interests or well-being of the truster—even if momentarily, say, in a one-off interaction. The account is, thus, normative in a sense that differs from the kind of instrumental self-interested normativity of Hardin’s account.
On the other hand, one problem with those alternative views is exactly that they either completely suppress the rationality of trust—making trust entirely non-rational—or make trust dependent on goodwill and intrinsic motivations on the part of the trusted in a robust sense. Neither of these options seems plausible. Trust seems to be responsive to reasons, even if only to a certain extent. Also, not all trusting relationships depend on goodwill and intrinsic motivations in that robust sense, although some might. As we will see below, this suggests that what is essential to trust is not the trusted’s motivations in a robust sense. Trust is more fundamentally about the truster’s placing themselves in a risky situation of dependence and vulnerability, and being optimistic that the trusted will fulfill the trust, than about expecting specific motivations from the trusted.
In what follows, we will offer an account of trust that combines elements of the pure rationalist accounts and of those alternative views.
3 TRUST AS A COMPLEX TWO-LEVEL EMOTION
An account of trust that attempts to combine a role for emotion with a cognitive dimension is proposed by Lewis and Weigert[40], who argue that trust involves a combination of cognitive, emotional, and behavioral components. They contend that the emotional and cognitive aspects complement each other in producing behavioral responses: “if all cognitive content were removed from emotional trust, we would be left with blind faith or fixed hope”, while “if all emotional content were removed from cognitive trust, we would be left with nothing more than a coldblooded prediction or rationally calculated risk”[41].
Lewis and Weigert’s account is especially interesting because it retains a cognitive (and rational) element of trust. This seems to point in the right direction because, given what an emotion plausibly is, emotions contain a cognitive content as their objects. Thus, if we describe trust as involving an emotional component, we need to understand how this component is related to the cognitive one. In what follows, we will take Lewis and Weigert’s insight as the basic idea from which to develop our proposal.
To begin with, let us assume a minimal shared understanding or common ground about emotions that most theories would agree with. Emotions are about something and are typically directed toward the world or a state of affairs, that is, they have intentional objects[42]. In addition, emotions are related to evaluations that make them intelligible and allow us to individuate them, that is, they have formal objects[43]. For example, fear is directed toward something perceived (or believed or interpreted or assessed) as dangerous, anger arises in response to harm, grief is a reaction to loss, love is related to the lovable or worthy of love, indignation emerges from perceived unfairness, etc.
So, if trust has an emotional component, it should be interpreted in accordance with this minimal assumption of common ground. It should be directed toward the world and associated with evaluations as its object. At this point, it might seem a natural move to define trust simply as an emotion with trustworthiness as its object. While this idea seems basically right, it requires further qualification. Let us develop this in more detail.
Assessments of trustworthiness are clearly cognitive (and rational) in a sense. To assess someone as trustworthy involves gathering information or considerations that support such assessment. The kind of relevant information here is supposed to show, for instance, that the trusted person has proved to be truthful and competent in what they are trusted to do, has behaved as expected in past trusting relations, has taken into account the interests of those who trusted them, has a reputation for being a reliable social partner, is responsive to social incentives to be cooperative, and so on. These are all sorts of considerations that can, for example, factor into Hardin’s account of the reasons for assessing the trustworthiness of a potential social partner. Thus, it seems plausible to acknowledge that this is an essential dimension of trust—though not one that exhausts it, as we will see shortly.
On the other hand, as we have suggested, confining an account of trust to cognitive judgments of trustworthiness leaves out a central aspect of trust, namely, the vulnerability and risk into which the truster places themselves. This is so because assessments of trustworthiness will always be limited and uncertain, even if they are appropriately responsive to the available evidence and, as such, rational. So, while this cognitive dimension of trust is a necessary part of it, it cannot fully capture what trust is.
It is at this moment that a proposal like Lewis and Weigert’s becomes especially promising by adding an emotional component to trust without abandoning the cognitive one. Before developing it in more detail, we should clarify the point of bringing emotions into the picture. One common motivation to explain trust in terms of emotions stems from two related ideas: firstly, the thought that cognitive assessments of trustworthiness are not an essential component of trust; secondly, the assumption that emotions are not in any sense cognitive or rational. Both ideas are implausible. Therefore, we need to provide another motivation for bringing emotions into the picture. But explaining why those two ideas are implausible will serve as a useful intermediate step in our argument.
The point highlights a common flaw in some of those alternative views that we have mentioned earlier. A number of authors who regard trust as involving an emotional component reject a link to cognitive/rational assessments of trustworthiness. This sounds implausible because it risks turning trust into an irrational attitude or reducing it to blind faith[44]. This would leave unexplained what trust is a response to. The emotional dimension of trust would be difficult to make sense of without some connection with information about trustworthiness. Actually, it would not even be clear how people could reason and argue about trust at all. The result would be that trust could perhaps be explained, but not rationally justified at any level. There would be no room for normativity in trusting another. Thus, although assessments of trustworthiness cannot alone fully capture trust, we had better not turn trust into something entirely non-rational or irrational.
Relatedly, the view that emotions are not in any sense cognitive or rational runs counter to our best explanations of how to make sense of emotions. As we have said, emotions are related to evaluations and are responsive to them. As a result, certain emotional episodes can be said rational, whereas others can be said irrational, insofar as they are appropriately or inappropriately responsive to the evaluations that make them intelligible. This means that the evaluations provide normative criteria for judging emotions. Now, one natural way of interpreting such evaluations is to take them as cognitive and responsive to reasons. In evaluating something as dangerous, one is (at least partly) responding to reasons to believe something as dangerous. So, if we want to bring an emotional component into the picture, we must preserve a cognitive/rational component as well.
So, the next questions are: how could emotion help us explain trust in a way that complements the cognitive/rational component? How can it help us explain what the cognitive/rational component of trust leaves out, namely, to make sense of the vulnerability and risk involved in trust?
We should begin by saying that, although emotions can have a cognitive/rational component, they are not exhausted by it. Emotions cannot be reduced to rational beliefs. Although this is not the place to provide a full account, it is worth noting that emotions also include practical, motivational, phenomenological, affective, bodily, behavioral components. They can be understood as involving a complex array of dispositions.[45] Emotions typically “transform” cognitive information gathered from the world, conferring practical, phenomenological, motivational, etc. significance to it and thereby changing our way of seeing things[46]. So, there is a sense in which it is correct to say that emotions go beyond their cognitive/rational components. More to the point here, some emotions even go beyond the reasons we might have for forming them. For example, this seems to be the case of love, if we think that love can be partly based on reasons, but is not exhausted by them. Love seems always to involve some “amplification” of the evidence one can find that another is lovable or worthy of love. Something similar can be said of hope, since hope also relies on some evidence but goes beyond it. As we will see next, trust bears similarity to love and hope in this respect. So, let us now begin to unpack the proposal.
If there is an emotion as a direct response to assessments of trustworthiness, it should be confidence. So, we can say that, as a result of assessing another as trustworthy, one becomes confident that the other will choose and behave as expected. In this way, trust at a basic level should be a mixture of emotion (confidence) and cognition (assessments of trustworthiness). Confidence, as an emotion, is also a practical positive response to evidence about trustworthiness. As such, it can change the way of seeing things in the world, that is, the way of seeing information gathered from assessments of trustworthiness. Still, confidence alone is not enough to fully explain trust, given that confidence is a direct response to assessments of trustworthiness. Confidence is, thus, connected mainly with reliance. To illustrate, a high level of confidence in someone—as a response to an assessment of this person as highly trustworthy—could provide a reason to act accordingly. But this would not yet be trust. It would be a form of mere reliance or prediction of someone’s behavior in which the vulnerability and risk characteristic of trust are absent. So, cases of high level of confidence in someone might turn the interaction rational and trust dispensable. However, given that on many occasions of social interaction the levels of confidence between individual partners are not high enough, trust becomes necessary. Nevertheless, because trust is a form of reliance and prediction, confidence should be part of it—though it does not completely explain what trust is.
So, trust must also involve a higher-order emotional reaction linked to confidence and its associated assessments of trustworthiness, but one that goes beyond those assessments. This higher-order emotional reaction should take the form of optimism and openness toward forming a cooperative relationship with an individual to whom one has reacted with confidence as a response to believing them to be trustworthy[47]. In other words, trust should involve a complex, two-level psychological state: at a lower level, it is a direct emotional response (confidence) to trustworthiness; at a higher level, it is a further emotional response (optimism and openness) toward one’s own assessments of someone’s trustworthiness infused with confidence[48]. Trust is then a form of optimistic projection that one’s assessments of trustworthiness about another will prove true[49].
As should be clear, according to this proposal, the complex emotional reaction of trust, including the higher-level optimism and openness, does not ignore evidence of someone’s trustworthiness. However, the emotional response “amplifies” the evidence. The evidence of one’s trustworthiness “counts in favor of” the emotional response that constitutes trust but does not, by itself, rationally justify it. Thus, if there is rational justification for trust, it must come from elsewhere. Still, the available evidence provides a kind of “rationalizing support.” For example, if someone trusts another without any evidence of their trustworthiness or with some evidence of their untrustworthiness, they may be trusting without rationalizing support and can be rationally criticized for that.[50] This provides a minimal normative criterion for the judging the rationality of trust. Next, we will ask whether trust can be rationally justified in any further sense.
Let us recap. We have seen so far that trust is composed of a two-level emotion: it is both an emotional reaction to information gathered about someone’s trustworthiness and an emotional reaction that amplifies such information. Can trust be justified on these grounds?
Firstly, we need to remind us that the justification at stake here is practical. Accuracy of belief is relevant to determining one’s reasons to act. After all, false, incomplete or uncertain information about the world can make one misrepresent their reasons and even lead them to act against their interests and well-being. The problem with the practical justification of trust is that, although it depends on evidence for the belief in a person’s trustworthiness, the information provided by that belief is insufficient to ensure that one can avoid vulnerability and risk when trusting that person. So, if one trusts another, what can serve as a reason for the trust or justify it? Does the truster act for no reason? In terms of our proposal here, the optimism characteristic of trust is a higher-order practical response to the limited and uncertain information about someone’s trustworthiness. How could that be a rational response?
We have already identified one sense in which it could be rational. It gains “rationalizing support” from the evidence gathered about someone’s trustworthiness. This is a minimal sense of rationality and normativity. It establishes conditions under which one can be criticized for trusting someone about whom one either has no information on their trustworthiness or has some information on their untrustworthiness. But, apart from this minimal sense, is there any further sense in which trust can be practically rational or justified?
The answer is that there can be justification in a “subjective” sense, but hardly in a “non-subjective” sense. It is “subjective” in the sense that the reason to act is entirely based or dependent on the truster’s optimism, that is, on their own motivation. The optimism itself provides a “positive” reason to trust. Therefore, this justification should be based on a kind of bootstrapping or self-supporting style of reasons[51]. For our purposes here, it suffices that “non-subjective” be understood as the direct negation of “subjective”. Thus, reasons are “non-subjective” in the sense that they are based on something other than, independent of, or additional to the truster’s motivations—regardless of their nature, origin or status. In what follows, we will provide support for the claim that a reason to trust can only be “subjective” in that sense.
This suggestion can sound unsatisfactory to some who might have expected our proposal to provide a “non-subjective” rational justification for trust. While we cannot fully meet that expectation, we can offer explanations for the optimism and openness characteristic of trust. There are many possible explanations for why people are optimistic about trusting one another, found not only at the individual level but also at the social level. At the individual level, perhaps as a general empirical truth about human nature, people may be biologically and psychologically predisposed to trust. The degree of these predispositions can vary among individuals and their development may depend on environmental conditions. No matter the details, an explanation of this kind should not be downgraded or neglected. Besides, caring relations and educational patterns can be central in the development of trusting and trustworthy personalities. As an extension of that, culture and sociality expressed through social practices can be responsible for turning many trusting relations into thick social and moral relations. Finally, as sociologists such as Sztompka[52], Mitsztal[53] and Luhmann[54] hold, social structures and institutions at the macro level of society can create conditions for the development of widespread dispositions to trust.
Faulkner[55] thinks that trust can be rationally justified in a “non-subjective” sense because we live by a norm of trust and trustworthiness (which involves sincerity in conversation) as part of our social norms. This presumption, Faulkner argues, could provide “non-subjective” practical reasons to justify trust. Although such reasons are in a sense also self-supporting, they are supposed to go beyond the merely “subjective” sense. The idea would be that the attitude of trust implies the presumption that we live by that norm of trust and trustworthiness and this presumption justifies the truster’s normative expectations regarding the trusted—in particular, that the trusted will take the truster’s dependence as a reason to fulfill the trust. The attitude of trust then can be said to provide self-supporting reasons to (the act of) trust under the presumption of a “non-subjective” social norm of trust and trustworthiness. Although the attitude of trust itself gives rise to the reason to trust, it acquires its justificatory and normative role only because there is a constitutive social norm governing trust and trustworthiness (and speech and communication more broadly). The norm is not itself the reason to trust, but is what confers justificatory and normative role on the attitude of trust.
That can be true as far as it goes. It is likely that there is one such a norm of trust and trustworthiness as part of our social norms and conventions. However, this does not seem to be enough to provide a “non-subjective” justification for trust. Social norms and conventions might well justify the truster’s forming normative expectations regarding the trusted. Still, this is not to say that the truster is justified in having normative expectations that the trust will in fact be granted and fulfilled by the trusted—i.e., it is not to say that the truster is justified to trust. To think otherwise seems to depend on the assumption of an idealized version of human societies in which the compliance with such a norm is assured. But this assumption does not directly translate into concrete practice and, thus, ends up becoming otiose as a source of justification.
In other words, a truster may be justified in expecting a certain normative response from the trusted and in complaining in case the trust is betrayed. However, the truster cannot be justified, solely on the presumption that a certain norm of trustworthiness is in force, that the trusted will grant the trust and comply with the norm. Trust as a social phenomenon between persons is really about real practices and action, about concrete social situations, and deliberation under these conditions. This is ultimately what the vulnerability and risk it implies mean, as well as what the challenge raised by the problem of cooperation means. Thus, there is always an unavoidable component of prediction in trust—contrary to what Faulkner claims. Trust is primarily and essentially about the truster’s expectation that the trusted will fulfill the trust in the sense that the truster’s assessments of trustworthiness will prove true. The impossibility of eliminating uncertainty in that prediction is what makes the vulnerability and risk inherent in trust so dramatic.
Let us illustrate the point by borrowing one of Faulkner’s examples:
A reformer decides to pick a new employee from a list of recently discharged prisoners the local prison circulates, and knowingly picks someone convicted for theft. Nevertheless, the reformer trusts her new employee with the till and at the end of the first day he reassures her, ‘don’t worry, the till balances[56].
Does it sound plausible to say about this case that the reformer has “non-subjective” reasons to trust, created by her own act of trust, that the trusted employee will fulfill the trust, on the assumption of a common social norm of trust and trustworthiness, which places him under a “non-subjective” normative requirement to fulfill the trust? It seems not. It seems more plausible to say that, although the reformer may be justified in forming normative expectations about the employee, she is not justified “non-subjectively” in trusting him. Rather, she is simply being (subjectively) optimistic that the employee will grant and respond to her normative expectations, and thereby fulfill the trust.
The result achieved here may also sound unsatisfactory from the perspective of the intuitions underlying some of those alternative views discussed before, particularly with respect to the emphasis those views confer on the motivations of the trusted person. As we have seen, Baier[57] and Jones[58] claim that trust involves expectations regarding the motivations of the trusted. For them, the trusted should have goodwill toward the truster and should also take into consideration truster’s vulnerability and dependence on them. As we have said above, this latter condition means that the trusted should have intrinsic consideration for the interests or well-being of the truster. These distinctions have the main purpose of allowing those authors to differentiate trust from mere reliance. Although trust is a form of reliance, the frustration of mere reliance, according to Baier, warrants only disappointment as the appropriate response, whereas the frustration of trust warrants feeling let down or experiencing betrayal as the appropriate response. The problem, then, for the proposal defended here would be that it seems to leave too little room for the appropriate motivational responses of the trusted.
Indeed, that is true as far as it goes. But an answer to that should begin by acknowledging that, according to a common understanding of trust, we often trust strangers and purely self-interested persons. In such cases, it barely makes sense to claim that we expect goodwill or intrinsic consideration for our interests or well-being from those types of persons. It is always possible to reply by saying that those cases reflect a stretched use of the term or concept of trust, that they are not genuine examples of trust. But this answer does not sound compelling. If we wish to preserve the idea that there can be trust between strangers and purely self-interested persons, as well as between intimates, we had better start with the idea that trust is a pluralistic phenomenon. There seem to be many realizations of trust: between strangers, purely self-interested persons, intimates, acquaintances; trust as a moral phenomenon; trust grounded in normative expectations, in communitarian ties, in social norms or conventions, and so on. It seems theoretically desirable to try to preserve most of this plurality under a single general concept of trust. That would have as a consequence that there can be many different normative relations, patterns of responses, and background conditions that govern trusting relations. So, a kind of diversity of expectations with respect to the motivations of the trusted should therefore be incorporated into the whole picture of trust[59].
If that sounds plausible, then any attempt to provide an all-encompassing view on trust would need to start by focusing on the background conditions and motivations of the truster. After all, trust is an asymmetrical relation whose existence always depends on the initiative of the truster. A feature common to any variation of the truster’s role in a trusting relation is the truster’s knowingly placing themselves in a situation of vulnerability and risk in relation to others (even if this need not be consciously articulated or an object of self-awareness). In some cases, depending on the normative relations and background conditions governing the trust, the appropriate response from the trusted may well presuppose goodwill and intrinsic consideration for the truster. But this need not hold for all instances of trust. In some of them—for example, the sort of instrumental relation characteristic of Hardin’s “encapsulated interest”—motivations such as goodwill and intrinsic consideration for the truster can be dispensed with. Nonetheless, in all cases of trust, the truster knows about their vulnerability and risk and maintain optimistic expectations regarding their assessments of the trusted’s trustworthiness.
Again, that means that there is always an unavoidable component of prediction in trust. Although the truster’s optimistic expectations are embedded within a specific realization of the trusting relation, along with its presuppositions of appropriate responses from the trusted, the truster’s expectations are primarily and essentially about the trusted’s fulfilling the trust.
5.1 Credulity
Nothing said so far is supposed to be incompatible with the idea that human beings may display, as an empirical fact, a widespread disposition for credulity. Indeed, we seem to be credulous as social beings. Although this may vary according to social conditions and environments, we tend to trust our social peers blindly or instinctively, without further consideration. Social signaling and conventional behavior work as conditioning triggers for that disposition for credulity.
We have said at the beginning that our discussion would not be about developmental research on human tendencies to trust. It may turn out to be true that one such developmental story confirms human dispositions for credulity and, perhaps, even that it had some survival value in our evolutionary history. Also, this may be related with the value we actually ascribe to trust. But nothing we have argued here hinges on any answer to those questions. As we have made clear throughout, our discussion is not about credulity or any merely descriptive level of the psychological tendency to trust, but with conditions for the justification of trust.
On the other hand, we have offered reasons for not taking credulity as relevant to the main questions addressed here about the justification of trust as a response to the problem of cooperation. We have provided a series of considerations for finding credulity unacceptable as a solution to that problem. Understanding trust as credulity would come close to understanding it as blind faith. Perhaps the only significant difference would be that a blind faith account of trust seems to acknowledge the normative problem of cooperation but offers a skeptical response to it, whereas credulity accounts are merely descriptive, psychological accounts, whose argumentative strategy is to circumvent that normative problem.
The main problem with blind faith accounts, which might be extended to credulity accounts, as we have contended, is that they seem to leave no room for any level of rationality and normativity in trust. That strikes as implausible because we recognize the problem of cooperation as normative, and because the problem appears to be part of our ordinary experience of social living. Failing to recognize the problem and its normative character would render dispensable the justification of measures (both at the individual and social level) adopted to prevent frustration in trusting relations. This, however, is not tenable. We are continuously involved in assessing the trustworthiness of current and potential social partners, monitoring their behavior, and selecting those for whom we have favorable (albeit limited and uncertain) evidence to trust. That seems common ground in our social lives.
It is also important to clarify that nothing we have said here requires that people consciously rehearse or articulate their reasons before or while trusting others. Again, the thesis defended here is not merely psychological in a descriptive sense. People may be justified in trusting others even if this occurs quite automatically, displaying a kind of immediacy that might otherwise be mistakenly taken for entitlement (as a normative, justificatory concept). This last point will be the topic of the next and final section.
5.2 Entitlement
A distinct point would be to argue that trust possesses a kind of entitlement (as a normative, justificatory concept) in that it does not require reasons or considerations in favor of trusting another. Trust would bear a distinctive kind of justification in that it would be normative on its own (by virtue of being basic and irreducible). Accordingly, one would be justified in trusting another in the absence of defeaters or countervailing reasons. This would have, as a result, that we are justified in trusting others by default.[60] However, the suggestion is problematic. Let us see why.
The idea of entitlement as a justificatory concept has its origins in epistemological views about forms of justification that go beyond evidentialist requirements of giving reasons for belief. Some authors[61] argue that perception, for example, is a source of justified belief by default. Accordingly, we are entitled to believe what we perceive in the absence of defeaters or countervailing reasons. This is so because, unless otherwise shown, we have (second-order or background) reasons to believe in the reliability of our perceptual system. That is to say, unless there is evidence that one’s perceptual system is malfunctioning or operating under abnormal conditions, one is entitled to non-inferentially believe what one perceives.
By analogy, something similar could be said of trust[62]. Unless otherwise shown, we could be entitled to trust by default—so, be justified in trusting—because we have (second-order or background) reasons to believe in the reliability of our trusting relations. That is, unless there is evidence that a particular trusting relation is working under abnormal conditions, one would be entitled to non-inferentially trust others. But is that true? It does not seem so.
There are two initial “structural” problems with that suggestion. Firstly, the analogy does not stand. Trust is a relation between persons, whereas perception is not. In trust, we rely on others—more specifically, on their choices and actions, which are beyond our control and access. This is precisely what gives rise to the problem of cooperation. Nothing like that holds for perception. Relatedly, it is because trust is a practical relation of reliance between persons that it is normative in a sense in which perception could never be—even if one insists on the existence of some normative space in perception.
In addition to such “structural” problems, the view that treats trust as entitlement is problematic because in social environments, especially in a large-scale complex and fragmented society, we are always subject to betrayal and defection in trusting others. As we have seen, this is part of the basic assumptions of what trust involves. Although it is certainly true that our reasons for believing others to be trustworthy will always be limited and uncertain, a practically justified solution to this limitation can hardly be to treat trust as justified by default. Again, we seem to have no option but to remain continuously involved in assessing the trustworthiness of current and potential social partners, monitoring their behavior, and selecting those for whom we have favorable (albeit limited and uncertain) evidence to trust. That seems common ground in our social lives. Denying it would amount to understanding trust as a form of credulity. Although it might be plausible to find room for an equivalent of credulity in perception—so that something like “perceptual credulity” would bear justification by default—we have seen good reasons for not reducing trust to credulity.
In spite of all that, maybe one could offer deeper and broader considerations for taking trust as bearing justification by default. Let us consider one of the interpretations Stern[63] offers of Løgstrup’s account of trust as basic. Here is one of Løgstrup’s examples[64]:
Let us imagine that we stand facing a destroyer who is trying to win us for his cause, but we know that he will shun no means in doing so and that he is not to be trusted. Face to face with the destroyer, we discover how much effort it takes to remain on our guard. The thought that, by talking things out, we would be able to dissuade the destroyer from his destructive enterprise keeps presenting itself; there is no eradicating it once and for all[65].
The question then is why that thought, which involves an almost irresistible inclination to speak openly and trust the interlocutor, is so persistent. One answer could be that it reveals a socio-psychological tendency for credulity, as we observe in human experience of sociality. As such, it would offer nothing that would make us revise what we have already said.
However, Stern does not interpret the example as merely describing a socio-psychological tendency, but instead as an attempt to formulate a transcendental claim.[66] Accordingly, the form of life characteristic of human beings would presuppose trust and sincerity in speech and communication. As Stern says: “unless speech is open, in the sense that most people speak to each other truly and honestly, in a way that means they can be trusted in what they say, speech would be impossible as a form of life”[67]. So, if speech is constitutive of the human form of life as we recognize it—in the sense that speech makes human life possible in the first place—and if successful speech and communication are only possible through trust, then trust would be rationally justified (and would be more basic than distrust). This could now serve as a broad transcendental (“non-subjective”) argument in favor of the idea that trust carries justification by default and that we are entitled to trust in the absence of defeaters or countervailing reasons.
But does that conclusion really follow? It does not look so. We cannot explore here all the details and implications of that thesis about the nature of successful speech and communication. However, let us grant that it is true that the human form of life, as we recognized it, could neither exist nor flourish without successful speech and communication, and that trust is precondition of it. Would this show that we are practically justified in trusting anyone by default in our ordinary encounters and interactions in common human life? It would not. At most, it would show that the human form of life we all should recognize as sharing is, on the whole, dependent on trust. But this is not to say that any human being on any social occasion of interaction, facing the problem of cooperation, has a reason to trust and expect trustworthiness from any other. Similarly to what we have said in response to Faulkner, it could ground one’s forming normative expectations regarding others we trust, given our shared human form of life, but it could not justify “non-subjectively” our trusting them.
Nonetheless, nothing said here denies the possibility that certain iterated successful trusting relations, under favorable social conditions, become so ingrained that dispositions to trust get “internalized” by default in the practices of many people, perhaps even an entire society. Yet this would be a psychological and social development of trusting relations, not their “non-subjective” rational justification.[68]
Baier, A. Trust and antitrust. Ethics, Vol. 96, p. 231–260, 1986.
Barbalet, J. A characterization of trust, and its consequences. Theory and Society, , Vol. 38, N. 4, p. 367–382, 2009.
Becker, L. C. Trust as noncognitive security about motives. Ethics. Vol. 107, N. 1, p. 43-61, 1996.
Bratman, M. Planning, Time, and Self-Governance: Essays in Practical Rationality. Oxford: Oxford University Press. 2018.
Burge, T. Perceptual entitlement. Philosophy and Phenomenological Research. Vol. 67, p. 503–548, 2003.
Coady, C. A. J. Testimony: a Philosophical Study. Oxford: Clarendon Press, 1992.
Coleman, J. S. Foundations of Social Theory. Cambridge, Mass.: Harvard University Press, 1990
Darwall, S. The Heart and its Attitudes. Oxford: Oxford University Press, 2024.
Faulkner, P. Knowledge on Trust. Oxford: Oxford University Press, 2011.
Gambetta, D. Can we trust trust? In: Gambetta, D. (Ed.). Trust: Making and Breaking Cooperative Relations. Oxford: Blackwell, 1988, 213–237.
Giddens, A. The Consequences of Modernity. Cambridge: Polity Press, 1990.
Goldie, P. The Emotions: a Philosophical Exploration. Oxford: Clarendon Press, 2000.
Hardin, R. Trust and Trustworthiness. New York: Russell Sage Foundation, 2002.
Jones, K. Trust as an affective attitude. Ethics, Vol. 107, N. 1, p. 4–25, 1996.
Kenny, A. Action, Emotion and Will. London: Routledge. 1963.
Lewis, J. D.; Weigert, A. Trust as a social reality. Social Forces, Vol. 63, N. 4, p. 967–985, 1985.
Luhmann, N. Trust and Power. Chichester: John Wiley & Sons, 1979.
Misztal, B. A. Trust in Modern Societies: The Search for the Basis of Social Order. Cambridge: Polity Press, 1996.
Möllering, G. Trust: Reason, Routine, Reflexivity. Amsterdam: Elsevier, 2006.
Ribeiro, L. M. Delving into denialism: rationality, emotion, value, and trust in social context. In: Barbosa, E. (Ed.). Moral Challenges in a Pandemic Age. New York, NY: Routledge, p. 171-191, 2023.
Scherer, K. R. What are emotions? And how can they be measured? Social Science Information, Vol. 44, N. 4, p. 695-729, 2005.
Seligman, A. B. The Problem of Trust. Princeton: Princeton University Press, 1997.
Stern, R. Trust is basic: Løgstrup on the priority of trust. In: Faulkner, P.; Simpson, T. (Eds.). The Philosophy of Trust. Oxford: Oxford University Press, 2017, p. 272-293
Sztompka, P. Trust: A Sociological Theory: Cambridge: Cambridge University Press, 2000.
CONTRIBUTION OF AUTHORSHIP
1 – Leonardo de Mello Ribeiro
Leonardo de Mello Ribeiro é doutor em Filosofia pela University of Sheffield, Inglaterra
https://orcid.org/0000-0002-5912-2921 • ribeiro.lm@gmail.com
Contribuição: Escrita - Primeira Redação
HOW TO QHOTE THIS ARTICLE
Ribeiro, L. M. Trust as a Social Emotionv. Voluntas: Revista Internacional de Filosofia, Santa Maria - Florianópolis, v. 16, n. 2, e93794, 2025. Disponível em: https://doi.org/10.5902/2179378693794. Acesso em: dia, mês abreviado, ano.