Navigazione – Mappa del sito

HomeNumeri68Part I: Trust and institutionsHarmonic and Disharmonic Views of...

Part I: Trust and institutions

Harmonic and Disharmonic Views of Trust

Laurent Jaffro
p. 11-26


This paper, at the crossroads of practical and epistemological questions, puts forward a non-standard approach to the study of a set of trust phenomena (trust, trustworthiness, distrust, self-trust, self-distrust…) and their interconnectedness. Two paradigmatic approaches to trust – harmonic and disharmonic – are unpacked and shown to be complementary. In contexts where the harmonic view applies, trust phenomena are mutually reinforcing. When the disharmonic view is appropriate, instead, they counterbalance one another. An analysis of Augustine’s De fide rerum quae non videntur helps introducing the former. Each view carries its own conception of the role of institutions in trust. For the harmonic approach, institutions are contexts that reinforce trust silently, whereas for the disharmonic view they are salient objects of trust. The need for an over-arching articulation of these limited views is further highlighted by appeal to another distinction, between background and decision-based forms of trust. This paves the way to what is here termed a metaharmonic theory of trust, a theory sensitive to the difference between the contexts where trust phenomena reinforce or counterbalance one another.

Torna su

Termini di indicizzazione

Torna su

Testo integrale

1Do we need proof of others’ good will in order to act on our presumption of it? This question is at the crossroads of the practical and epistemic concepts of trust. Proof is, among other things, a confirmation of that which would remain doubtful in its absence. In the epistemological debate, the sceptic believes that proof is required as long as prior trust does not guarantee validity. The anti-sceptic believes that this would be asking for proof that is superfluous or impossible. There is a certain parallel, as well as significant differences, between this discussion and the practical issues involved in trust, between the problems posed by faith and those posed by action and interaction. Epistemic trust is the credence given to information or the source of information, for example, someone’s word or our own faculties, such as memory or perception. Practical trust is a significant reduction in the uncertainty regarding the future actions of other agents.

2An ambiguity must be dispelled: good will can be conceptualized as a simple absence of ill will or as a negative magnitude of it. When I complain that a shopkeeper does not demonstrate good will in response to my request as a customer, the good will I am expecting from him is just an absence of ill will. In what follows, since Augustine first addresses the issue of trust between friends, I will take the notion of good will in the sense of the negative magnitude of ill will. In other words, I will examine the issue of the proof of a positive volitive disposition, or benevolence, that goes beyond the ‘perhaps minimal’ good will whose presumption is characteristic of practical trust, according to Annette Baier (Baier 1986: 234).

3What is the relationship between the problems of scepticism, to which some respond by mobilising trust in the domain of knowledge, and the problems of cooperation, to which some respond by mobilising trust in the domain of action? I propose that we approach this question on the basis of the opuscule De fide rerum quae non videntur (circa 400), which is attributed to Augustine. My reading will be neither scholarly nor historical, but retrospective, taking on the filter of current discussions on cooperation in the social sciences, on the one hand, and on coherentism in epistemology, on the other. A secondary undertaking is to suggest a distinction between the many roles of social institutions: they are not only a possible object of trust, but also provide a context that may or may not reinforce trust. In addition, social institutions can be subject to various forms or degrees of trust. For instance, trust with respect to a social institution can be very implicit, or much more explicit and even voluntary.

4The methodology I have adopted consists in first developing a paradigm of trust, not based on how fides is treated in Augustine in general, but specifically on how it is handled in On Faith in Things Unseen. This paradigm of trust is one that is simultaneously epistemic and practical, and is a good fit for what I call a harmonic conception of trust. In the second part, I will argue that this model must be rendered more complete and complex by introducing another understanding of trust, the disharmonic conception, in order to account for other important aspects of the phenomena of trust.

1. The Augustinian paradigm

5The title of Augustine’s opuscule is redundant, since faith by definition is a type of trust whose object is that which cannot be seen. Believing has that which is invisible or absent as its object, while seeing in the figurative sense, just as in the literal sense, has that which is visible or present as its object. Seeing in the figurative sense is a type of cognition to which all cognition cannot be reduced. This is because there is another type of cognition – belief. There are obvious truths, which are blindingly obvious to the mind, and hidden truths, which are subject to faith.

6Why this title? What is a pleonasm for Augustine is not one for the objector he is responding to. The objector, like Doubting Thomas, demands to see in order to believe. But this objection is based on a confusion, because if one needs to believe something, it is because one does not see it. In response, Augustine essentially develops epistemological considerations. He concludes by maintaining that at a higher degree of comprehension this faith, which in a preliminary analysis is not vision, is actually vision. Vision, whether it is intellectual or sensory, is then used as a metaphor for faith. There are ‘eyes of faith’ beyond bodily eyes and the eyes of the mind.

7The most remarkable thing in Augustine’s analysis is that it runs into the practical problem of trust in interpersonal interactions, even if it is from afar. But these interactions are not really of the same nature as interaction in a large, complex society. Augustine talks about the practical problem of trust in a way that seems both exotic and stimulating to us. There is a distance between trust in family and friends as Augustine talks about it, which is linked to significant social institutions, and what we understand as trust today (perhaps since Hobbes? In any case, Hobbes contributed to this understanding). This distance makes it possible to bring out certain interesting aspects of the problem.

8Augustine comes to this issue through the example of friendship. Friendship is a disposition that we do not see, and it is subject to fides, which is interchangeably faith and trust. He asks: What would happen if this trust in the other person’s disposition of friendship was absent in us? This may raise the issue, which is crucial in the social sciences, of the difference between the problems of cooperation in situations where you can adjust your actions because you are confident about the other person’s dispositions, and situations where you cannot do so because other people’s dispositions are hidden from you. In the latter situations, it is rational to adopt either, if available, the dominant strategy (that is, the strategy that yields a better payoff regardless of what the other chooses) or a Nash equilibrium (that is, a situation in which all chosen strategies are such that, taking into account the others’ strategies, each player gets the best payoff). This occurs most often because the behaviour we are expecting from others are actions that will be taken in the future, and observation of behaviour is our only means of knowing the dispositions of other agents. By having access to the other’s dispositions through fides, one gets a kind of assurance that the observation of behaviour cannot provide.

9Augustine begins by recalling the truism that the domain of knowledge goes well beyond what we see with bodily eyes. Augustine takes the example closest to him: the operations of his own mind. Augustine knows that he is thinking, whereas this operation is invisible to bodily eyes. It is not invisible, however, to the eyes of the mind. “In our mind itself, whose nature is invisible, there are innumerable things” (Augustine 1950: 81). Augustine also takes the example of faith: When I believe, I know that I believe. This transparency of faith as an operation of the mind makes it so that when we believe, we cannot doubt that we believe (Chastaing 1956). Now, divine realities are of course inaccessible to bodily eyes, but they are no less inaccessible to the eyes of the mind. If they were subject to intellectual intuition, faith would be superfluous.

10The model of self-knowledge is not sufficient to account for knowledge of divine realities. The problem of knowing divine realities is analogous to the problem of knowing other minds, and not to the problems of self-knowledge or knowledge of external, sensible realities. Augustine takes the analogy between access to the divine and access to the other literally. Hence the example of friendship, which is the interpersonal equivalent to faith in the divine.

11Here is the problem: “Tell me, please, with what eyes do you see your friend’s disposition towards you? For no disposition can be seen with bodily eyes” (Augustine 1950: 83). ‘Disposition’ is a translation of voluntas. But the other person’s disposition is not perceptible to the eyes of my mind either. It is not present in my animus; I am not conscious of it. “If you do not see it, how do you on your part requite his loving kindness, if what you cannot see you do not believe?” (Augustine 1950: 83). Doubting Thomas’ requirement applied to social interactions would produce disastrous results. It is here that the epistemic issue of trust runs into the practical issue of cooperation. Since I do not see the other person’s good will towards me neither with the eyes of my mind nor with my bodily eyes, if I do not believe in this good will, then I will not return it.

12Augustine’s pessimistic hypothesis evokes from afar the problem of how cooperation may emerge in a non-cooperative context through a tit-for-tat strategy. Robert Axelrod in The Evolution of Cooperation (1984) showed that the winning strategy in a repeated game would be for an agent to take the initiative to behave benevolently towards the partner in the first round, and then echo the other person’s response. But if the latter was not able to see the former’s benevolent conduct, then he could not respond with benevolent behaviour; or if he did respond benevolently, the former would not necessarily ‘see’ this behaviour as benevolent. In this case, benevolence could not be signalled, and thus crucial information is lacking, which results in a failure of cooperation. However, the connection between tit-for-tat and trustful conduct is superficial. For trust does not consist in observing the other’s behaviour, but in ‘seeing’ the other’s dispositions. Tit-for-tat is a case of mechanical generation of cooperation, whereas Augustinian ‘friendship’ rests upon built-in beliefs about other persons, rather than on our capacity to respond to behavioural outputs.

13The fact is that we most often reward the other person’s benevolence by reciprocating. We play a different game than the prisoner’s dilemma. We view this benevolence with the eyes of faith. Reciprocation is evidence of trust in the other person’s disposition. Observable behaviour is subject to interpretation. Now, if benevolence and friendship were not present ab initio in interpersonal relationships, people could not interpret behaviour as demonstrating benevolence. “Your friend’s faith is not appreciated by you if there is in you no reciprocating faith by which you may believe that there is in him what you do not see” (Augustine 1950: 85).

14The objector claims that he does not see the other’s friendship but recognizes it in his words and deeds. But in reality he believes; he trusts in his friend’s good will, but also in his friend’s reciprocal trust. Augustine rejects the argument that it would be enough to see benevolent words and deeds. There is a counter example on which Augustine bases his argument: I can see these words and deeds, and be mistaken. You can feign good will or hide ill will, “in expecting some advantage”. This is a strategy of deceit, which banks on the other person’s gullibility.

15Augustine thus points out that you cannot have proof of the other person’s friendship by means that are completely external to the relationship of trust that binds the friends together. You cannot put the relationship of trust on hold while you find proof of your friendship:

Indeed, since you wish to put faith in things not seen, ought not we, seeing that we do put faith in the heart of friends both when they have not been fully proven, and when we have proved them to be good by our misfortunes, even then we believe their good will toward us rather than see it? Unless it is that faith is so great that very appropriately we judge that we see (with the eyes of faith itself, as it were) what we believe, although we really ought to believe because we cannot see (Augustine 1950: 85‑7).

  • 1 This seems to agree with the empirical work of psychologists showing a mutual reinforcement over ti (...)

16Proof does not prove this friendship for me unless I believe in this friendship. Proof is therefore superfluous and impossible anyway if it is understood as a means of taking us from a situation where we doubt the friendship to a situation of trust. If we do not initially believe in the friendship, we will not interpret signs or proof through the lens of trust as signs or proof of friendship. They may instead be construed as trickery. If the other person is not already ‘seen’ as a friend through the eyes of faith, proof will not work.1

17We must bear in mind that friendship, here, is a virtue. It refers to a superlative friendship, and not to a weak benevolence, like a disposition to cooperate. This is a significant difference from the cooperation problems modelled by non-cooperative games. If we replace this thick virtue – which occurs in a strong institutional, familial, or civic context – with a disposition to interact in a non-aggressive manner with unknown partners or outside of a significant institutional framework, Augustine’s words lose their relevance. The presumption of weak, minimal benevolence in the other person does not imply belief in his virtue, but only expectations regarding his behaviour.

18The connection to the issue of practical trust is based on a thought experiment. Let us eliminate trust in the other person’s good will from interpersonal relationships. “Who will not mark what great disorder will result in them, and how dreadful a confusion will follow?” (Augustine 1950: 87). Augustine thus presents his imaginary world of widespread distrust as something that looks quite similar to the Hobbesian representation of the state of war.

For who will be cherished by anyone in mutual charity, since love itself is invisible, if what I do not see I ought not to believe? Friendship, then, will completely perish, since it is established solely on mutual love (Augustine 1950: 87).

19If friendship is absent, we cannot receive the other person’s marks of friendship since we cannot believe that he has given us marks of friendship. And if the other person’s actions are not interpreted as the actions of a friend, then the interaction will deteriorate accordingly. “If friendship perishes, the bonds of neither marriage nor of relationship nor of affinity will be held intact in the mind”. If spouses do not see one another’s love through the eyes of faith, they will not return the other’s love. Then their relationship cannot be friendship, i.e. reciprocal love. Similarly, children who do not know they are loved will not love their parents. Even beyond that, according to Augustine, all human relationships will be ruined in the absence of this belief.

20In this thought experiment, it is not friendship that is presumed to be eliminated at the outset, but the belief in the other person’s love for us. Now, there cannot be friendship without faith in the other person’s love. The consequence of this initial elimination of belief is the deterioration or disappearance of friendship. Augustine thus illustrates what I describe in Section III as a harmonic conception of trust, according to which trust and trustworthiness reinforce one another. Similarly, self-trust and trust in the other person, practical trust and epistemic trust, trust and trust in trust also reinforce one another. According to this harmonic conception, when one part of this edifice deteriorates, the whole collapses.

21Reciprocity in the relationship between A and B means that A sees B’s disposition to cooperate through the eyes of faith. If A does not see B’s cooperativeness, B will not reciprocate. What is Augustine’s position with respect to the following proposition, which I will introduce all the while accepting its anachronistic nature? Cooperation is based on a strategy of giving tit for tat, whether the behaviour is positive or negative. We imitate the other person’s response. In the beginning of the opuscule, when he talks about what is observable with bodily eyes or the eyes of the mind, Augustine considers that it is not true, in the context of friendship, that one person’s behaviour is adjusted on the basis of what he observes of the other’s behaviour. Now, if we consider Augustine’s final proposition, it would appear that friendly reciprocity is based on the vision of the other person’s love through the eyes of faith. If children do not see their parent’s love, they will not return their love. Within the institution of the family or friendship, there are only positive actions. We might be tempted to see a positive, virtuous form of tit-for-tat in the reciprocity of friendship.

22This comparison is not absurd, since Augustine alludes to the situation of uncertainty when he considers the circumstances under which “charity is uncertain and affection (voluntas) suspected” (Augustine 1950: 87) between parents and children. However, we cannot really talk about a positive, virtuous version of tit-for-tat since this is the winning strategy in contexts where one does not cognitively have access to the other person’s dispositions. We match the other person’s behavioural response when we do not know his dispositions. It is difficult to talk about ‘tit-for-tat strategy’ in a situation where we have significant information that could reassure us about the other person’s dispositions. Friends view their friendship as an ‘assurance game’, rather than the application of a tit-for-tat rule to a repeated prisoner’s dilemma.

23Let us sum up our initial findings:

  1. The profile of the interaction changes completely depending on whether we initially input belief in the other person’s love or suspicion regarding his dispositions.

  2. This analysis leads us to nuance the opposition between the practical problems of trust and the problems of evidence. These two types of issues converge in that they are both problems of seeing the invisible.

  3. Through his questioning on interpersonal fides, Augustine has thought, in his own way, about questions like ‘should we have trust in trust?’ or ‘does trust in the other presuppose self-trust?’. The operations of another mind, and a fortiori its dispositions, which cannot be an object of my vision through the eyes of my mind, can, however, be an object of my faith. I cannot trust the other person if I am not first confident in my abilities to see the other person’s love through the eyes of faith. If I doubt the other person’s love, then I will not see it. He who believes that seeing is believing cannot see anything in the sense that belief is a superior form of vision. The objector’s lack of trust in his own cognitive equipment could in practice lead him to not trust the other.

  4. This type of approach also draws attention to circularity in the edifice of trust. According to what I call a harmonic conception of trust, trust functions on a loop in line with the image of the ‘loop of trust’ used by Keith Lehrer (1997) in his defence of coherentist epistemology.

2. Examples of applications of the Augustinian paradigm

    • 2 One example: Regarding physico-theological proof, the character Cleanthes in Hume’s Dialogues Conce (...)

    In the theological domain, this discussion shows an affinity with the debate on the status of proof in the case of God’s existence. This is not at all surprising since the example of the proof of friendship serves as a model for Augustine for the need for proof in religious matters. The question is whether this proof completely establishes trust or whether it at least partially presupposes a form of trust.2

  1. In the ethical domain, several authors, ancient as well as modern, have shown that an undertaking that claims that it is possible to define the meaning of moral predicates independently of and prior to their common understanding in everyday moral life is doomed to failure. Just like I cannot ‘see’ the other person’s goodness without being ‘a friend’ in the Augustinian sense, I cannot learn the meaning of the term ‘just’ in the moral sense fully in the absence of practical dispositions for ‘justice’. This point was particularly salient in the critique of theological voluntarism in the Cambridge Platonists, for example, and also in G.E. Moore’s critique of the naturalistic fallacy (Bouveresse 2016).

  2. In the epistemological domain too, trust cannot be found if it has been lost or put on hold, even if it is only long enough to obtain proof. Augustine’s words could apply to issues of modern scepticism in their epistemological as well as moral aspects. This theme is elaborated in Thomas Reid, for whom a natural capacity to trust is the principle behind both social experience and cognitive activities. Reid simultaneously maintains that “without fidelity and trust, there can be no human society” (Reid 2010: 334), and that “we are born under a necessity of trusting” our intellectual faculties (Reid 2010: 179-180). The reliability of our faculties does not need to be established by an additional loop, no more than practical trust in the other’s ‘good faith’ requires prior proof. This would be impossible anyway, according to the First Principle of Contingent Truths in the list of Principles of Common Sense: “That the natural faculties, by which we distinguish truth from error, are not fallacious. If any man should demand a proof of this, it is impossible to satisfy him”. Indeed, proof presupposes faith in this reliability (for a discussion of the interpretations of this principle, see Rysiew 2014).

24According to Reid, since the attitudes or behaviours of trust cannot be wholly acquired, but rely on a natural capacity, it is fallacious and inappropriate to account for them in terms that are purely economic, prudential, or even purely moral:

The things essential to human society, I mean good faith on the one part, and trust on the other, are formed by nature in the minds of children, before they are capable of knowing their utility, or being influenced by considerations either of duty or interest (Reid 2010: 335).

3. Completing the harmonic view with a disharmonic view of trust

25One important question is whether trust in other people and self-trust reinforce or counterbalance one another. Augustine claims that they reinforce one another. A second question is about the relationship between trust and distrust: are they opposites or complementary phenomena? Augustine, at least in his De fide rerum, claims that they are opposites. The Augustinian paradigm should be qualified on both points.

26Indeed, there are forms of trust in other people, or institutions, which have their basis in self-distrust. People who do not trust themselves have recourse to coaches or counsellors, whom they trust. Professional therapists might be helpful, if friends or family members cannot do the job. In this Section, I try to understand the links between this kind of trust, on the one hand, and distrust, especially self-distrust, on the other.

27Thus, I distinguish between two philosophical attitudes towards trust, ‘harmonic’ and ‘disharmonic’. The distinction arises when we ask whether one kind of trust combines well with other kinds, namely whether trust in others necessarily goes hand in hand with self-trust, and also whether trust necessarily accompanies trust in trust, or whether one kind of trust (say, epistemic trust) is necessarily associated with another kind (political trust, economic trust, or practical trust in general). According to the harmonic conception, the more (or less) you trust yourself, the more (or less) you trust others, and vice versa. According to the disharmonic conception, distrust towards other agents or institutions is congruent with trusting only oneself. Conversely, those who lack self-trust, in order to overcome their difficulties, may have to trust others, not others in general, but specific categories of people and perhaps determinate persons, in whom we place our trust either because of their personal qualities or because of their institutional role, or both.

28Here is one argument in favour of the harmonic conception: Isabel trusts her own faculties and operations, including the operation of trusting other people. Thus, if Isabel trusts Margaret, Isabel’s self-trust reinforces her trust in Margaret. Conversely, were Isabel to suspect that her self-confidence was excessive, she would have reason to be more cautious in her relationship with Margaret. Also, if Isabel’s trust in Margaret was severely betrayed, then she would have reason not to trust her tendency to trust others.

29However, in support of the disharmonic conception, I would reply: Isabel’s trust in Margaret may operate independently of Isabel’s self-trust, so that, if Isabel’s trust in Margaret was severely betrayed, Isabel could still (with some luck) count on herself. Conversely, were Isabel’s self-trust to become impaired, she (with some luck) could count on Margaret.

30Let us consider another possible argument in favour of the harmonic conception, which goes like this: Trust and trustworthiness are causally and normatively dependent upon one another. The more someone is manifestly trustworthy, the more we trust that person. The more we trust someone, the more that person is encouraged to be trustworthy. Can someone who lacks self-trust be trustworthy? We might think that, were he trustworthy, he should be trustworthy for himself as well, and thus that he should trust himself. Yet he lacks self-trust. Therefore, he is not fully trustworthy. He is the rotten apple that spoils the barrel.

31This argument, which I would not deem robust, is an argument against the compatibility between lack of self-trust – let us call it self-distrust – and mutual trust. It is not a relevant argument against the compatibility between self-distrust and one-way trust from the self-distrusting person in another person, for instance a therapist. Even though we suppose that self-distrusting people are thus not fully trustworthy, I do not see why this should hinder them from trusting someone else. Self-distrust might even motivate trust in specialists.

32To sum up, according to the harmonic view, trust-tendency is a relationship that is symmetric in some sense and is also circular as far as the relationship between trust and self-trust is concerned:

    • 3 I use ‘symmetry’ in a loose sense. I thank Brian Jabarian and André Lapidus for their comments on t (...)

    Symmetry: If A tends to trust B, then B tends to be motivated to trust A.3

    • 4 One proponent of the harmonic view is Keith Lehrer, who, in the wake of his coherentist interpretat (...)

    Circularity, interrelatedness of trust and self-trust: If A tends to trust B, then, A tends to trust herself; and if A trusts herself, she tends to trends B.4

33According to a strong version of the disharmonic view, trust-tendency is neither symmetric nor circular. Of course, it frequently happens that when A tends to trust B, B is motivated to trust A, etc. However, this does not rule out the significance of cases where it does not happen. This calls to mind prisoner’s dilemma games, where it is not rational to initiate trust, and it would appear to be rational not to respond to trust with trust. Moreover, there is often some sort of causal or motivational relationship between trust and self-distrust, or between distrust towards others and self-trust.

34According to a weaker version of the disharmonic view – which I favour – trust-tendency is symmetric but not circular. If A tends to trust B, then B tends to be motivated to trust A. It would be imprudent to deny this, for this is just the way trust works. The view that trust-tendency is symmetric in this loose sense suggests that trust is a solution that is quite different from the dominant strategy in a prisoner’s dilemma, and that the parties are playing a different game, corresponding to one option in Amartya Sen’s assurance game: the hypothesis that “everyone has implicit faith in everyone else doing the ‘right thing’” (Sen 1967: 114). There are particular practical and institutional contexts in which the phenomenon of trust appears: The agents’ conduct is constrained by rules, habits, and institutional contexts, which diminish uncertainty and turn it into an acceptable risk. In these contexts, there is a causal and normative, or motivational, relationship between A’s tendency to trust B, and B’s tendency to trust A. The fact remains that trust-tendency is not circular in the above sense: As in the strong disharmonic conception, there is often some sort of causal or motivational relationship between trust and self-distrust, etc.

35This comparison between harmonic and disharmonic views of trust should take into account the distinction between background implicit trust and trust as an action. The harmonic conception draws heavily on background trust, whereas, in accordance with the disharmonic conception, trust as an intentional and voluntary behaviour remains a convenient recourse for those who lack self-trust. However, this distinction and the latter distinction between the two views of trust cannot be exactly aligned. Moreover, implicitness, voluntariness, etc. differ in degree as well as in kind.

  • 5 Annette Baier interestingly notes that “we inhabit a climate of trust as we inhabit an atmosphere a (...)

36Background trust cannot be willed. It is deeply involuntary; it is not something we decide to do; it is the state of our relationships with people, environments, or the natural world, from which doubt or suspicion is absent. Whereas mistrust forms the background to a decision to trust, insofar as it involves an awareness of uncertainty, a primitive state of trust is the background to our experience of distrust: For instance, as a result of an accident, I discover that I should not trust my legs, although until now I had trusted them, while not even being conscious of my trusting them – a case of ‘atmospheric’ trust.5

37Let us modify the example in order to tackle the question of epistemic self-trust. Let us substitute ‘legs’ with ‘mental faculties’. Thus, we may distinguish two kinds of epistemic self-trust:

  1. (Background trust) I have no reason at all to doubt that my mental faculties will allow me to correctly think, talk, argue, feel, perceive, judge, love, etc. It is not only the case that I have no reason to suspect my mental faculties: I am not even aware of it. It did not even occur to me that I have no reason, etc.

  2. (Trust as an action) I have reason to think that my mental faculties are fallible; in spite of that, I place my trust in them, for I have an even better reason to do so.

38In this shift from bodily to mental faculties, we may observe an analogy between distrust and scepticism. This debate took centre stage in 18th century Scottish philosophy. Both David Hume (the sceptic) and Thomas Reid (the anti-sceptic) claimed that there are irresistible natural beliefs, for instance the belief in the existence of external objects beyond our perceptions, which even sceptical philosophers cannot help having. Thus, we could speak of an epistemic background trust in the existence of objects or persons. One difference between David Hume and Thomas Reid is that the former considers that this background trust ought to be justified, and yet cannot be justified; whereas the latter replies that epistemic trust does not need to be justified. According to the Humean sceptic, our faculties are so fallible that we should be very careful in their use. Nevertheless, in everyday life, we act upon the premise that they are trustworthy.

39Until now, I have set out the rudiments of a philosophy of trust from a few (too) simple examples: legs, bodily faculties, and mental faculties are not agents. A fuller philosophy of trust needs to focus on the interactions between agents. An agent, roughly speaking, is at the centre of intentions, and thus a centre of will and intelligence. Were my legs or my faculties agents, then trusting them would look completely different. For then my legs or my faculties could decide to respond or not to my expectations towards them. I would not be able to know in advance what they are about to do, not only because of the intrinsic uncertainty of future events, but also because of the uncertainty that is characteristic of agency.

40When the context of interactions between agents is characterised by their acute awareness of uncertainty and risk – and thus may encourage distrust – trust as an action (Dumouchel 2005) is an option. The difference between trust and distrust, here, lies not in the informational input, but in the practical output. The informational input is a diagnosis of uncertainty. This diagnosis may lead to the option of defection: We prefer not to expose ourselves to a betrayal of trust. It is thus a case of practical distrust. The same diagnosis may lead to the option of participation: We accept the risk of exposing ourselves to betrayal in order to pave the way for cooperation. It is thus a case of practical trust. Here the voluntary nature of this form of trust is fully manifest: The only difference between trust and distrust in this situation is between the decision to escape and the decision to accept the possibility of being let down. They converge in the initial analysis and diverge in the resulting action. The difference between the two is behavioural.

  • 6 Let us note in passing that a loop between the voluntary and the involuntary kinds of trust is plau (...)

41If we turn now to atmospheric trust, the picture is quite different. This kind of trust does not differ from distrust only as a practical output of the same informational input. Background trust is at least the absence of distrust, and perhaps the opposite of distrust. The awareness of uncertainty is numbed. By what? It would be a gross mistake to answer: by trust. Trust is the explanandum, not the explanans. However, it is true that there is a phenomenon of self-reinforcement behind it, since experience, customs, practices, conventions, and institutions play a crucial role in the formation of background trust. (a) Habit is the process through which we learn both to respond to the environment and to become familiar with specific responses of the environment. Once we are accustomed to a state of affairs, we do not pay attention to difficulties as long as our expectations are not significantly disappointed; we even tolerate exceptions, failures, or mistakes. (b) Conventions may be viewed in a Humean way as having positive feedback on the mutual adjustment from which they proceed. (c) Institutions are contexts, which diminish the uncertainty of social interactions, because they may constitute sufficient conditions for what Bernard Williams (1988) calls thick trust.6

42Since Augustine’s thesis is about a world of virtue, friendship, and faith, whereas the disharmonic view fits in with a world of weakness and wariness, they do not contradict one another. The two theses may even be compatible if one world coexists with the other.

  • 7 Self-trust may be seen as a necessary condition for minimal procedural autonomy, understood as the (...)

43One reason why I put forward the distinction between harmonic and disharmonic views is that it highlights the controversy on the status and scope of practical autonomy. The harmonic conception encourages a high idea of self-control, in which autonomy and sociality are mutually integrated.7 On the contrary, the disharmonic conception fosters a more modest view of self-control, tailored to the fact of our weakness. In this latter conception – a form of self-paternalism – self-control is indirect and draws on the recourse to counsellors, pastors, etc., determinate persons, roles, or institutions.

44Let us imagine the situation of agents who could not trust themselves and who would also be aware of being disloyal to themselves as well as to others, unable to stick to their commitments, and thus so unreliable that nobody could trust them. In a word, these agents would add complete lack of self-confidence to utter untrustworthiness. Would they be also untrusting, unable to trust others? Not necessarily, according to the disharmonic conception.

45These constitutively weak agents are not so different from us. What I call a ‘hopeless case’ is not the effect of practical irrationality only, understood as an accident or as a particular pathology. It is the consequence of a bundle of common circumstances, which notably include bad luck and the temporal structure of human existence. The sense of the uncertainty of our own future actions, together with the experience of our past failures, may be sufficient to convince us that we are unreliable for ourselves as well as for others. Here distrust is not just the shadow of trust – a negative solution to the practical problem to which trust is the positive solution – but the feeling that trust as well as trustworthiness are out of reach.

46A solution comes from the resources of human psychology, with some help from the others. As Edward Hinchman puts it: “The diachronic exercise of practical reason is […] based on trust – typically on self-trust, but when self-trust fails […] it will have to be based on trust in others” (Hinchman 2003: 26). In this hopeless case, trust as an action is the correct option and is directed at a particular ‘trustee’, therapist, expert or coach. Since it is voluntary, it is something we can do even though atmospheric trust is lacking, and in the absence of evidence of trustworthiness. As Paul Dumouchel argues:

Viewing trust as an action suggests that what makes trust necessary is not so much our lack of knowledge about the world as the fact that we depend upon each other. It is true that we do not have perfect knowledge of the world, but even if I did know in advance what you will do, it does not follow that I could always avoid giving you power over me.

47Indeed, background trust, whether interpersonal or organisational or both, is also involved in people’s relationships with their therapists. Nevertheless, in hopeless cases, automatic respect for institutional authorities or for the person is not sufficient or does not work. Here trust as an action is of great use. Even though background trust plays a major role in epistemic behaviours, including those that involve recourse to the expertise of others, trust in the expertise of others, at least when it has a therapeutic or perhaps a religious dimension, is often a decision, the beginning of an action. However, my argument for the practical significance of the judicious placing of trust is anything but an encomium for sheer voluntarism, nor is it akin to highly demanding views of personal autonomy. On the contrary, my intention is to point out that when distrust, including self-distrust, is pervasive, some kind of trust is still present, and remains a technical resource that is well suited to our condition of weakness.

Concluding words

  • 8 See for instance the beginning of Book VIII of the Confessions, where a still weak Augustine finds (...)

48In the first part of this paper, I have set out the Augustinian paradigm of trust. In the second part, I have brought a different, disharmonic, conception of trust into the discussion. This seems to suggest that Augustine embodies a harmonic understanding of trust, which is highly debatable. Out of consideration for Augustine and especially for his Confessions, I must say that his insistence on the Fall and the need for God’s grace and the help of others would plead in favour of a disharmonic view. The harmonic view applies to the realm of the faithful. We need the disharmonic view to account for the rest of us, including Augustine before his full conversion.8

49Should we consider the two views to be incompatible? No. They are two parts of a more comprehensive view (which, were it not pompous, I would term metaharmonic) that accounts for the dynamic of trust in contexts where the phenomena of trust reinforce one another, as well as in contexts where they counterbalance one another. Institutions are often a source of trust: This is correctly described by the disharmonic view. The harmonic view captures the fact that institutions are also the object of implicit trust. It also accounts for the way in which institutions may warrant and reinforce implicit trust. In brief, the two views are complementary.

  • 9 See Knud Ejler Løgstrup (1997: 8) about the way in which trust in relatives and friends sets the pa (...)
  • 10 Portions of this paper have presented and discussed on four occasions: at the seminar Les Après-mid (...)

50A remaining question is which view has priority over the other. Although a more in-depth argument would require more space than can be allotted to it here, I think that the harmonic conception, because it concerns the natural core of trust, which guides further acquisition and development of trust skills,9 is conceptually central, and the disharmonic, peripheral. However, the philosophy of trust should not neglect the periphery.10

Torna su


Augustine, 1950, De Fide Rerum quae non videntur, in M.F. McDonald (ed. and trans.), Washington D.C., The Catholic University of America Press.

Axelrod, R., 1984, The Evolution of Cooperation, New York, Basic Books.

Baier, A., 1986, Trust and antitrust, “Ethics”, 96, 2: 231-260.

Bouveresse, J., 2017, Éthique et logique: Cudworth, Leibniz, Moore et la critique du ‘sophisme volontariste’, in M. Pécharman and P. de Rouilhan (eds), Le philosophe et le langage. Études offertes à Jean-Claude Pariente, Paris, Vrin: 275-302.

Chastaing, M., 1956, Consciousness and evidence, “Mind”, 65, 259: 346-358.

Dumouchel, P., 2005, Trust as an action, “European Journal of Sociology”, 46, 3: 417-428.

Govier, T., 1993, Self-trust, autonomy, and self-esteem, “Hypathia”, 8, 1: 99-120.

Hinchman, E., 2003, Trust and diachronic agency, “Noüs”, 37, 1: 25-51.

Jaffro, L., 2009, Ist die Evidenz der Existenz des Anderen ein Modell für die Gotteserkenntnis? Berkeley, Hume, Reid und das ‘Argument des Irregulären’, “Aufklärung. Interdisziplinäres Jahrbuch zur Erforschung des 18. Jahrhunderts und seiner Wirkungsgeschichte”, 21: 51-74.

Lehrer, K., 1997, Self-Trust. A Study of Reason, Knowledge, and Autonomy, New York, Oxford University Press.

Løgstrup, K., 1997, The ethical demand, H. Fink, A. MacIntyre (eds), Notre Dame, University of Notre Dame Press.

Miller, P., Rempel, J., 2004, Trust and partner-enhancing attributions in close relationships, “Personality and Social Psychology Bulletin”, 30, 6: 695-705.

Reid, T., 2010, Essays on the active powers of man, in K. Haakonssen, J.A. Harris (eds), Edinburgh, Edinburgh University Press.

Rysiew, P., 2014, Reid’s First Principle, “Canadian Journal of Philosophy”, 41, S1: 167-182.

Sen, A., 1967, Isolation, assurance, and the social rate of discount, “The Quarterly Journal of Economics”, 81, 1: 112-124.

Weckert, J., 2005, Trust in cyberspace, in R. Cavalier (ed.), The Impact of the Internet on Our Moral Lives, Albany, State University of New York Press: 95-117.

Williams, B., 1995, Formal structure and social reality, in B. Williams, Making Sense of Humanity, Cambridge, Cambridge University Press: 111-122.

Torna su


1 This seems to agree with the empirical work of psychologists showing a mutual reinforcement over time between a feeling of trust and a favourable interpretation of the partner’s motivations in married couples (Miller, Rempel 2004). This type of approach bolsters the conception that trust is a ‘seeing-as’, which is in line with Augustine’s paradigm: “If A sees B as trustworthy (with respect to A), he interprets B’s behavior as showing goodwill, or at least not ill will, toward A. If A sees B as untrustworthy, that is, A distrusts B as opposed to having no views on the matter, he interprets B’s behavior as showing ill will toward A (a typical characteristic of the paranoid A)” (Weckert 2005: 103). According to John Weckert, the disappearance of trust is a ‘gestalt switch’ (105).

2 One example: Regarding physico-theological proof, the character Cleanthes in Hume’s Dialogues Concerning Natural Religion (Part III) calls an argument that is distinct from reasoning by analogy an ‘irregular argument’. An irregular argument, which is epistemological in nature, consists in demonstrating that belief in an intelligent cause of the universe is irresistible, as is our trust in the existence of others. This irregular argument is considered to be valid by authors such as Berkeley and Reid (Jaffro 2009).

3 I use ‘symmetry’ in a loose sense. I thank Brian Jabarian and André Lapidus for their comments on the properties of trust-tendency.

4 One proponent of the harmonic view is Keith Lehrer, who, in the wake of his coherentist interpretation of Thomas Reid and the Reidian stance in his own epistemological work, claims that there is a “keystone loop of self-trust, trust of others, and the trustworthiness of ourselves and others for ourselves and others” (Lehrer 1997: 126).

5 Annette Baier interestingly notes that “we inhabit a climate of trust as we inhabit an atmosphere and notice it as we notice air, only when it becomes scarce or polluted” (Baier 1986: 234).

6 Let us note in passing that a loop between the voluntary and the involuntary kinds of trust is plausible: The initial decision to trust, like a Pascalian wager, with some luck may foster habitual trust as ‘second nature’; conversely, background trust, encouraged by social conventions and practices, may facilitate the decision to trust in a context where uncertainty resurfaces.

7 Self-trust may be seen as a necessary condition for minimal procedural autonomy, understood as the ability to think and decide for oneself (Govier 1993).

8 See for instance the beginning of Book VIII of the Confessions, where a still weak Augustine finds help and support from Simplicianus.

9 See Knud Ejler Løgstrup (1997: 8) about the way in which trust in relatives and friends sets the pattern for trust in outsiders, contrary to the common view that suggests that mistrust is the only option beyond the circle of acquaintance: “Under normal circumstances […] we accept the stranger’s word and do not mistrust him until we have some particular reason to do so. We never suspect a person of falsehood until after we have caught him in a lie”. The intimate connection between epistemic and practical trust is obvious here.

10 Portions of this paper have presented and discussed on four occasions: at the seminar Les Après-midis de Philosophie et Economie in Paris (December 2017); before the Société de Philosophy du Québec (May 2017); at the Ecole Pratique des Hautes Etudes (December 2015); at the Philosophy Department of the University of Nanjing (April 2014). I thank the organisers and participants for their comments. The paper especially benefited from conversations with Jacopo Domenicucci, André Lapidus, Aude Bandini and Brian Jabarian.

Torna su

Per citare questo articolo

Notizia bibliografica

Laurent Jaffro, «Harmonic and Disharmonic Views of Trust»Rivista di estetica, 68 | 2018, 11-26.

Notizia bibliografica digitale

Laurent Jaffro, «Harmonic and Disharmonic Views of Trust»Rivista di estetica [Online], 68 | 2018, online dal 01 mars 2019, consultato il 24 juin 2024. URL:; DOI:

Torna su

Diritti d’autore


Solamente il testo è utilizzabile con licenza CC BY-NC-ND 4.0. Salvo diversa indicazione, per tutti agli altri elementi (illustrazioni, allegati importati) la copia non è autorizzata ("Tutti i diritti riservati").

Torna su
Cerca su OpenEdition Search

Sarai reindirizzato su OpenEdition Search