Navigazione – Mappa del sito

HomeNumeri68Part II: Trusted institutions. Ap...Why (and how to) trust Institutio...

Part II: Trusted institutions. Appropriatness and rationality of trust in institutions

Why (and how to) trust Institutions? Hospitals, Schools, and liberal Trust

Pierre Lauret
Traduzione di Charlotte Jestin
p. 41-68


This is a paper about how we relate to institutions. Its aim is two-fold: accounting for what it is to ‘trust an institution’, and cashing out the right attitude to have towards public institutions. The descriptive account shows that ‘trusting institutions’ is a complex and ambivalent phenomenon, which oscillates between proper trust (as a two-place relation) and mere reliance, depending on the social function of the institution at hand. The normative proposal highlights the merit of a liberal form of trust in public institutions, as opposed to totalitarian and libertarian attitudes. To do this, the paper, reviewing a large set of public and private institutions, focuses on two cases, healthcare and educational institutions.

Torna su

Testo integrale

1Since Hobbes, the assumption that public institutions are necessary for ‘trust’ has been a commonplace of political theory and political philosophy. This trust is generally understood to be a decisive quality of social relations in a situation where others are not reduced to one-dimensional figures of potential aggressors and encounters with others are not anticipated unambiguously and self-fulfillingly as violent. Public institutions are expected to interrupt the catastrophic logic of the state of nature, as explained by Hobbes. All public institutions, regardless of their function, share a common and originary purpose: to facilitate relationships of trust between social agents, thus promoting social life, cooperation, and adjusted competition.

2There is sound reasoning in assuming that public institutions can only build trust if they themselves benefit from it. When the institutions of the State have a reputation for corruption, investors tend run away. When a segment of the population stops trusting their law enforcement, the police are often still able to maintain order, albeit at the cost of profound disorder, which would include arbitrary interference by police forces, resentment against the State and a feeling of injustice on the part of the concerned population. Admittedly, the investment of trust that institutions necessitate need not be blind. The functioning of public institutions is quite compatible with the liberal tradition of mistrust towards them. This mistrust works as a principle of control. The model of absolute power embodied by a Hobbesian ruler might make one believe that the state need only employ fear to guarantee the individuals’ trust in their reciprocal contractual relations. However, individuals in this unfriendly state model rely on the trustworthiness and rationality of the actions of their sovereign. A State built on fear is not a Hobbesian State. It is the State that is both absolute and unpredictable. This State can produce little trust.

3However, these common assumptions about the relationship between institutions and trust are not exactly clear. What does it mean to ‘trust’ in an institution? Trust is often thought of as a relationship between people. This relationship is linked to ‘personal’ qualities: loyalty, sincerity, benevolence, justice… Moreover, the contemporary debate on trust mainly seeks to define trust in terms of its role in human relations. A notion of trust, however, seems to be relevant not only to mankind, but equally to machines (‘I trust my car’), equipment (‘trust the rope, it won’t give in’), and institutions (trusting in, or perhaps not trusting in, law enforcement and the justice system). Is this concept used in the same way when applied to such different realities? Is inter-personal trust the baseline model of trust more broadly?

4A first way of clarifying the issue is to avoid making the status of ‘human person’ a condition of the relationship of trust. We can doubt that one relates to institutions as human beings. They are ‘legal persons,’ which means above all that they have a legal status, responsibilities, and rights. People work within them, but they are not ‘people’ themselves. Even by restricting the extent and intensity of trust, is it plausible that Paul would trust the police the same way he trusts his father? At first glance, it is difficult to admit that one has a personal, or close to personal, relationship with the police. And yet, one often talks about having or losing trust in the police.

5We could distinguish two things:

  • Trust in a type of being that one considers as an agent, and not necessarily a human person. An agent is an entity able to affect us, and to which one attributes a certain degree of autonomy and responsibility. It can be a person, but also a sophisticated robot, or, precisely, an institution.

  • Reliance in beings that one considers as functional objects and not as agents: a car, a rope, a compass, a metronome, a calculating machine. This last case is interesting: it prompts us to distinguish, among holders of artificial intelligence, between agents and objets. Agents, are the ones we attribute a high degree of autonomy (AlphaGo, DeeperBlue) to, and objects the ones from which we don’t expect initiatives that go further than, say, spell checking. (And we do not want the latter to take other initiatives, such as modifying our texts, or sending e-mails without their control. One of the reasons for our trust in functional objects is that they do not have autonomy: individuals want them to fulfil a function determined according to rigid operating rules. Even if they do not know these rules, they assume that it is through them that the object performs its function in a reliable manner. And if one knows them, they can use them to control its functionality and reliability. This point will prove important for our relationship to institutions.)

6Trust and reliance are standardly both thought of as three-place relations. This seems right in the case of objects: I rely on that object to do something. I count on it to meet specific expectations. It seems less appropriate when it comes to agents. Even when a relationship with a person – or an agent – is above all functional, for example in the case of a plumber or mechanic, one can always think of trust as a two-place relation: I trust this person and this trust places them in a horizon of expectations that is open, and not situated (see on this point the persuasive arguments of Domenicucci and Holton 2017). I trust my calculator to perform the right calculations, and that is all. I trust my plumber, not only to repair my sink but also for many other things: to be honest, to stifle a call of fire that would coincide with the moment of his intervention, or to warn me that the window pane on the first floor is broken. The horizon of expectations is open to a number that is not infinite, but undetermined, of situations I do not think of, or not necessarily, when I think I trust my plumber. So, I cannot describe this relationship as a three-place relation: the set of arguments likely to fill the third place is too indeterminate, and my relationship of trust consists in not listing them one by one to justify my trust. This is what the autonomy attributed to the agent implies. It is more plausible, and more convenient, to consider trust in an agent as a two-place relation.

7The case of the public institutions, then, seems unclear. On the one hand, one sees them as agents and not as objects: they are able to affect us, and they are attributed a certain degree of autonomy and discretion. On the other hand, it is tempting to consider that the way we relate to them is three-place: the police are trusted, or not, to maintain order by respecting the principles of the rule of law. Individuals have a functional relationship with institutions, and they want to both rely on them and have control over them. They grant them autonomy, but it remains limited or controlled, in the context of situated and restricted expectations. One refers to them as agents expected to fulfil specific mandates. Trust in them is inseparable from the mandate they are entrusted with. To trust them is to expect that they will not do what they are not supposed to do, that they will not go beyond, or that they will not fall short of their mandate.

8However, it is not obvious that one’s trust in an institution can be so narrowly defined. I can continue to trust my system of parliamentary democracy, or in the rule of law in my country (actual institutions and not mere ideas), even if I find that they fulfil badly, or not at all, some of their missions. The case of trust in institutions, therefore, seems complex. We seem to consider them as a class of agents towards whom our trust is more determined than a binary relationship. I will say more about that in the first Section.

  • 1 It can only be laid out roughly, and would call numerous developments.

9As a first step, I unpack more precisely what is meant by ‘trust in institutions’ – for which I introduce the term of art Institutional Confidence. To do so, I will rely on the difference between ‘trusting’ and ‘relying upon’. I will then wonder whether trust in institutions should be considered a norm, a civic ideal. As I will bring a positive, but particularized, answer to this question, I will finally explain how, in my opinion, it is necessary to live according to – and to perpetuatw – this norm. In this last part,1 my remarks will focus less on the nature of the specific relationship of trust than on a style of relations with public institutions, which, to my eye, seems to correspond to a civic ideal granted to the political regime of modern liberal democracy. I will therefore pass from a descriptive analysis to a normative reflection, which relates to political theory, and not to the sociology of institutions. This is also why I will analyse examples from common experience, and not methodically collected empirical data. I will start with the case of the hospital, then move on to the more complex case of the school.

I. The hospital

10Individuals entrust to public institutions things and beings that are crucial them. Their bodies, when an intervention in a hospital is required; their children when they go to school. This assumes a high degree of trust, and one can first look at the reasons for that trust, without reducing those reasons to a calculation of odds, and trust to a particular form of risk-taking, as economists rightly do. Although ‘trust’ is understood to mean a relation that is irreducible to a calculation of risks, trust has its reasons: it is not necessarily arbitrary or blind. It can be built up over time and justified, without implying a form of risk-calculation.

11I will begin by looking at the case of the hospital because its social function is simpler and clearer than that of the school. I will see that when it comes to institutions, the appreciation of their social function is a very important element of the trust placed in them.

12In the case of the hospital, one can distinguish four reasons for trust, which correspond to the different dimensions it relates to:

  1. the social function of the hospital: it must provide care, not kill, engage in dangerous experiments on people, or imprison them. This trust is based on common sense: to believe that hospitals have a secret function, hostile to the ill and the patients, would be considered a form of paranoia. Common sense receives the help of the professional rules of medical ethics (the Hippocratic Oath), and legal control. Quite regularly, the public is informed of cases involving abuse of power by health care workers, such as the clandestine administration of lethal substances for the purpose of euthanasia. This information is such as to show that trust should not be blind, but also that it has reasons, since these cases are exceptional, and detected by control procedures.

    • 2 Rawls clearly identifies the relationship between the institutional system that organizes the basic (...)

    the competence generally attributed to the staff of public hospitals. It depends on the efficiency of the recruitment system, and before that, of the training. Trust in the public hospital envelops trust in other institutions: the law, which has just been mentioned, and now public education, along with the degrees it delivers. An institution works in coordination with others, within a system, that owes its cohesion to certain general principles. Part of the trust one places in this institution is explained by these principles, as will be seen in this essay.2

  2. Trust in competence, or in authority, is a personal relationship: one trusts a particular doctor or nurse. But this personal relationship can be fostered, and preceded, by an impersonal one: it makes sense to say, before any personal relationship, that one trusts hospital doctors because they have received a very high level of training, that only the best can access these positions, that their function gives them a great experience, etc.

  3. the well-functioning state of technical devices and instruments. They depend on the funding of public hospitals, i.e. from the public resources allocated to them.

13Trust refers here to a political and social choice concerning the existence of public care institutions. In a schematic, or ideal way, one could say that we trust the public hospital because we want to have that trust, and we want the State to provide the corresponding financing.

  • 3 See the works of two specialists in child psychology (Bowlby 1980) and (Laznik 2013). Convergence b (...)

14being considered as a person and as an individual. One could speak here of a condition of dignity: to be treated as a person who lives individually a certain situation, and who deserves a quality of attention. Psychological evidence support a common-sense assumption: children’s trust, self-confidence, and appropriate trust in others, depend very much on their feeling of being an object of attention from their parents.3 Entering the hospital is in itself an infantilizing process: we need care; we put ourselves in the hands of others, we accept to be manipulated, transported, cleaned. Childhood means being in need of others, and certain situations, such as illness, remind us that from this need we cannot be completely free.

15Autonomy may be for us an ideal; not independence, in the sense of autarchy. Aristotle sees in friendship the compensation of dependence: the autarchic god has no friends. In the same vein, the importance of trust lies in this dependence on the other, which characterizes childhood, and that illness reminds us of. In the case of children, to be taken into account depends on the parental love, and how the parents enjoy the presence of the child. In the case of a hospital institution, we rely on the rules of ‘institutional culture’, according to which the staff has to pay attention to people, listen to them, offer them a smiling face and speak to them with kindness. The comparison with childhood indicates that this is not necessarily a superficial or secondary element of trust. Presumably, the dignity condition plays an important role in choosing a residential care institution for a dependent elderly person.

16Basing one’s trust in the hospital on these four orders of reasons does not seem to proceed from a particularly suspicious or insightful mind. It is good sense, or common sense. It would be inappropriate to have trust in a hospital that is thought to be incompetent or defective. Under the Nazi regime, parents of people suffering from mental illness or mental deficiency had good reasons to be wary of asylums and specialized institutions, whose social function was suspect. Under certain conditions, you cannot rely on the hospital.

17By putting forward this expression of ‘relying on’, I introduce the hypothesis that the logic of the reasons for trust differs depending on whether the trust is addressed to a person or an institution, even if we consider the institution as an agent. In the case of a person, the reasons justify the attribution of qualities (honesty, sincerity, benevolence, competence, etc.), or dispositions, which are individuated or incorporated in that person. Trust is then a personal binary relation (x T y), on the basis of which the person is expected to act in a certain way, in a more or less wide range of situations that it is not necessary to determine beforehand to justify our trust. Because x trusts y, when a certain situation arises, they expect y to act in a certain way. But this expectation is open, general and undetermined: x thinks that it is not necessary to contractually define the acts expected from y. There is no need to stipulate or control these acts. To a varying extent, x allows y to act at their own discretion, by granting them a degree of initiative. One must therefore distinguish the reasons for our trust in a person, and the expectations that this trust induces. They do not equate: in the case of trust as a binary relationship, the reasons I have to trust a person are not that I rely on them to do such and such, to fulfil such mandate or perform such mission. It is rather trust as a personal relationship that is likely to engender expectations in such and such situation that presents itself. This is the logic of the very particular binary relationship that is trust. This is not the case for an institution like the hospital: our trust is equivalent to our well-weighed expectations. We rely on the hospital to perform its function with competent staff, using the adequate equipment, and paying sufficient attention to the individuals being cared for. It is this system of expectations that we call trust. This trust is then developed, not because we would trust the hospital as a person, but because we rely on the means implemented, the internal rules of operation and external control procedures, to regulate its effective operation and adjust it to its function and our expectations. So, my hypothesis is that while we trust individuals, we rely upon the hospital to meet certain expectations. We entrust to it our body on the basis of a mandate or a commission, and we trust it, as long as we expect this mandate or this commission to be fulfilled satisfactorily.

18If the question is to understand and especially to justify our trust in an institution, then the most important question is the nature and the determination of the mandate we entrust to it: what are we really counting on, and can this reliance be analysed in a list of specific expectations? This question seems essential, and complicated, when it concerns a school, because it relates to the social function of this institution, which I made the first reason for the trust in a public institution. The social function of a school is more complex and more opaque than that of a hospital. What is expected is similarly more complex, more opaque, and also more diverse according to people and social positions. Therefore, before deepening the nature of trust in institutions, and the type of mandate or expectation to which it corresponds, it is necessary to dwell on the relationship between trust and the social function of institutions.

19What is at stake in the trust individuals can give them?

20In the case of the hospital, the patient relies on the response to a personal and particular need: a need for care. Since this need stems from the vulnerability of our nature, it is a social need, not an individual one, especially since the State cannot be indifferent to the health of the population in many ways. With the progress of medicine, the social need for hospitals or body care centres has replaced that of churches in charge of the salvation of the souls. On this point, there is a convergence between expectations of the civil society and State’s social control. It’s a convergence between a problem of social justice and a question of State administration of population welfare.

  • 4 The variation of the principles of distributive justice according to the types of good to be distri (...)

21As a result, trust in public hospitals raise a traditional issue in social justice: how far should public institutions guarantee individuals’ rights (and specifically claim rights), e.g. a right to healthcare? This raises a problem of solidarity, and of the choice of the distributive justice principle corresponding to the type of goods constitutive of healthcare: to each according to their needs, to each according to their ability to pay for services (as is often the case for dental care), to each according to their merit (their past sanitary conduct, their life expectancy).4 Several answers are possible and serve as effective rules in different countries for their healthcare system.

22Trust in the public hospital is therefore based on the way in which one thinks that this institution fulfills a social function, and in a fair way. For a supporter of democratic equality according to Rawls, or the complex equality according to Walzer, this trust comes down to ‘relying on’, in two senses: on the value of care, but also on the access to care. The special relationship of trust in one institution entrusted with a specific function encompasses a more global relation to the principles of justice that institutions must achieve and respect.

23Trust thus means being assured that if anyone needs an expensive surgery, they will be entitled to it, in satisfying conditions of safety, efficiency and dignity.

24By using the expression ‘being assured that’, I emphasize the two forms of relationships associated with the idea of ‘trust in institutions’: on the one hand, the dimension of mandate or commission, which I will call ‘relying on’ and which concerns the function and the functioning of the institution; on the other hand, the insurance dimension, which is the conviction of a reduction of the risks related to one’s vulnerability, insofar as the institution fulfils a protective function. It is in these two senses that one can say that the institution produces trust.

25I do not think that either of these two senses accounts for trust in one person: trusting someone is neither a particular form of risk-taking linked to a calculation of odds (on the Hobbesian model of the conditions of possibility of cooperation) nor a system of specific and determined expectations (Domenicucci, Holton 2017). But when it comes to an institution, an impersonal agent, these two modalities of trust are crucial, and if they are absent, trust becomes difficult to understand, or even blind in some cases.

26This last remark encourages, before examining this difference more fully, to address a consequence of this conception of trust in institutions: the appropriateness, or lack thereof, of this trust, according to risks and expectations. It seems normal to experience apprehension before surgery, which is often both more dangerous and riskier than repairing an automobile, for example. More generally, many diseases arouse apprehension, as long as their fate is partly unknown and their symptoms contain some ambiguity or uncertainty.

27It is up to the institution to respond to this apprehension. But the trust that one attaches to this answer cannot be blind:

  • it depends on the assessment we make of the state of the institution, its staff, and its equipment. This assessment may vary from one institution to another. People who break their leg in medium-sized towns sometimes prefer to be transported by ambulance to a big city, or to a hospital they know, rather than being operated on the spot, even though this only involves a routine intervention;

    • 5 See (Goffman 1961). A total institution is a place of residence and work where a large number of in (...)
    • 6 The case of the prison may be debatable. But would you deny the difference between a prison, and a (...)

    it is dependent on the variable sensitivity to the effects of authority, which have been studied by psychology: the authority of the title and the white coat is significant in certain situations, including being a patient in a hospital. In case of complex pathologies, or serious or risky interventions, it seems prudent to consider a second opinion. But this initiative already assumes a certain ease or social authority, since it amounts to allowing some mistrust of the practitioner who gave the first opinion. In addition, there is not always a choice, and this choice tends to disappear in the context of the hospital, which is similar to what Erving Goffman calls a total institution,5 except in terms of the duration of the stay. Entering the hospital is giving entrusting yourself to an apparatus of people, rules and things prescribing to us clothes (often a ridiculous gown open in the back), the menu and time of our meals and more generally all the rhythm and the use of our time, which deprives us of our intimacy (all kinds of people come in and out of our room at will, and often without knocking), and which submits us to a hierarchy where we are at the very low end (we cannot tell the housekeeping staff to come back later). Of course, a total institution is not as such a binding institution. By this I mean an institution that neither expects nor produces, or at least not primarily, trust, but rather obedience – for example, prison or slavery. The nature of the binding institution is that it does not work in the interests of those it constrains.6 As such, it cannot build their trust, nor does it expect it. It cannot be given a purpose similar to that of benevolence in relations between people. The hospital is thus not a binding institution. But in its operation as a total institution, it risks getting closer to it. It is precisely the function of what I have called the condition of dignity to give the relationship a sense of trust, not mere obedience. For several years now, it has incidentally been a demand of patients’ associations to be taken into account as a person, and not only as a body or mind affected by such pathology. This claim is about the relationship between trust and authority, which should not be mechanical. Because if it is, whether it is negative or positive, it is not appropriate;

  • finally, in the case of the hospital, trust should not be foreign to the awareness of the risks, especially those which, though potentially annex, can nevertheless be very important. As long as the hospitals have not solved the problem of nosocomial diseases, it is inevitable that people who are aware of this problem consider a surgery, even benign, as a serious risk-taking, in a context of not insignificant uncertainty.

28The purpose of these considerations is to recall that trust requires not only a descriptive analysis but also a normative reflection. It is in this respect that it usually raises questions because it is, first of all, a relation or a practical disposition. From a practical point of view, what interests us is not whether trust is a particular type of risk-taking, a three-place relation, or a two-place one, but rather to place our trust appropriately. Whatever the essence or nature of the trust, it has not only a being, but a value, and different modalities: it can be blind, reasonable, tested, and so on.

II. The nature of trust in institutions

29The first consequence of the previous considerations is that, in my opinion, what we generally term ‘trust in public institutions’, i.e. Institutional Confidence, is actually a three-place relation of reliance. I actually ‘rely on this institution to Phi’. The way we relate to institutions does not fit the two-place relationship distinctive of trust (Domenicucci, Holton 2017). The authors show, convincingly in my opinion, that trust is a specific and binary relationship, whose two places are occupied by autonomous agents. This sui generis relationship of trust should not be conflated with the three-place relation of ‘reliance’. ‘Counting on’ is a three-place relation: we rely on y to Phi (to be honest, to manage well our domain, to be knowledgeable in forestry, in short, to be a good bursar or steward, if you are a landowner). The unorthodox thesis of the two authors is that it is not because we trust y to do Phi1, Phi2… Phin that finally, we can say that we trust y. From the three-place relation, one cannot understand the two-place; it is rather the other way.

30What is the difference between relying on and trusting?

  • Relying on is a three-place relation, which cannot be reduced to a two-place relation. We can count on y for a lot of things. But this does not have the significance or the consequence of relying on them simpliciter. We always rely on someone to do something, and the notion of ‘relying on’ is linked to the action. The reliance is based on a mandate, not on a personal relationship. The logic of the mandate is not: ‘I rely on y; so I rely on y to Phi’. But: ‘I entrust to y mission Phi because I count on y to carry it out’.

  • Trusting is a two-place relation, which can give rise to all sorts of three-place trust relations in very different contexts (I trust you, so I lend you my car, I lend you my house, or money, I entrust you my son for the holidays, etc.). The relationship precedes these contexts, and it justifies our actions in these contexts. Finally, three-place trust is not the addition to the trust of the reliance relationship. Trust and reliance are two different relationships: it is different to trust a friend with your child because you trust them, and to entrust them to a host family you do not know, because you rely on the organization putting you two in touch to carefully select families. ‘Three-place’ trust does not rest on the same reasons or the same logic as ‘relying on’. It is a specific relationship of trust, a consequence of the two-place trust, one of its manifestations in a particular situation.

31To summarize, trust is a disposition towards an agent to whom we recognize a certain degree of autonomy. It is primarily a binary relation:

  • it does not obey the logic of mandate or mission: it does not relate to the execution of actions or specified functions – unlike ‘reliance’;

  • it grants the other agent a use or discretionary exercise of their aptitudes, judgments, feelings, in a more or less extensive field according to our relation to them. In other words, it does not rely on control, regulation, or contract procedures; and it dispenses with it. Thus, a marriage contract is neither the condition nor the manifestation of trust between the spouses, but the mark of its limit.

32Can one transfer this provision of trust from people to institutions such as banks, corporations (e.g. Monsanto), governments, public institutions? Is it appropriate? It must be remembered here that when such institutions ask to be trusted, they ask to be allowed to act or function without control, regulation, or inspection; or without fixating their action by imperative mandates. With regard to public institutions, I will present some reasons to answer negatively to the first descriptive question; and consequently to the second one, which is normative.

33First of all, it is necessary to well distinguish the case of public institutions.

34There is a big difference between a government and a public institution. Trust in a government is either political support for MPs, which is a very different thing from trust; or trust in people – in which case we are dealing with a two-place relation if we accept the hypothesis of Domenicucci and Holton. It is not about trust in an institution (the government), but in people who exert executive power. Can one reduce relations to institutions to relationships with people within these institutions? I will show that one cannot, and I have already suggested that one must distinguish trust for personal reasons (I trust this surgeon), and for institutional reasons (the French School of Surgery arouses my trust).

35But first let us deal with the case of commercial companies and banks, which is complex. If one takes companies such as Monsanto or the pharmaceutical company Servier, their main purpose (the rational search for maximum profit) and their sector of activity (chemistry, pharmacy) is likely to create distrust. Moreover, the caveat emptor maxim, well received in Anglo-Saxon law, encourages us to do so. In addition, depending on the sector of activity of companies, this maxim requires being tempered by legal rules, and public control procedures. It can be satisfied when it comes to ties, not pesticides, drugs, or automobiles. When a company sells potentially dangerous products for profit, it can only build trust if it meets specific trust criteria. These can be of two types:

  • a risk reduction, through legislation and control. The legal regulation aims to ensure that the relationship between a pharmaceutical company and the public is not a kind of state of nature where all means are good for profit;

  • a set of good practices constituting a corporate culture, which is built over time. An example can be the trust in German car brands: their solidity, reliability, longevity. Is this a form of trust similar to the one I have in my plumber? No, because in the case of Mercedes or Bmw, the reasons for the trust lie in our expectations of the products of these companies: we expect their cars to have certain qualities because this is part of their reputation built in time. These expectations are the fabric of our trust. We do not, or we should not attribute to them provisions or qualities such that we expected that, in a range of delimited but not determined situations, they would act in a manner that we would consider satisfactory, or well-intentioned. I have seen that the distinction between reasons and expectations is characteristic of trust as a binary relation: binary trust is based on reasons, but these open a more indefinite horizon of expectations. So, I do not trust Mercedes as a person or agent deemed to be trustworthy.

  • On the other hand, and disturbingly so, mistrust towards a company seems analogous to the mistrust towards a person. If a company is known for its lack of transparency, for knowingly disseminating false information, and for engaging in intense lobbying, it cannot engender trust, it even breeds mistrust. Yet, ‘I do not trust Monsanto, or Servier’ has the same meaning as ‘I do not trust the person x’. Here one can observe the difference between the reasons for mistrust, which are determined, and the expectations, which are not. Hence I can say that because of its lack of transparency and its information policy, I do not trust Monsanto in general, simpliciter. In the case of lucrative enterprises, I therefore seem to notice an asymmetry between trust, which is a reliance, and mistrust, which is general. Trust-reliance is tied to certain expectations; I rely on Mercedes or BMW for specific practices. In Monsanto or Servier’s case, mistrust seems to be involved in the attribution of a moral personality to the company. The brand no longer functions as the abbreviation of a list of well-defined guarantees, but as the proper name of a moral entity from which nothing good is expected. That is why we are inclined to support a strengthening of the regulation and the control towards this kind of companies, which seek to evade it a little too much.

36Finding an asymmetry in the analysis of trust and mistrust in for-profit firms may be a sign that something is wrong with this analysis. But for the moment, I do not see how one could relate to a firm otherwise than on the mode of the reliance, i.e. a three-place relation; and it seems sensible to me to say that we do not trust such or such company in the mode of a two-place relation. I offer the following explanation: from a company, because of its social function, one should positively expect products with certain qualities, possibly with an aura of prestige, and not a general style of appropriate attitudes; on the other hand, in the negative mode, one can have reasons to think that with such an enterprise one should expect the worse. By nature, a company can only build trust as an agent of production; but certain acts or ways of behaving provoke a reprobation that targets the enterprise as a moral agent, in much the same way as it would a person. This asymmetry therefore means that the social function of companies does not allow them to display the moral qualities that base trust as a two-place relation.

  • 7 Merton 1973.

37This is why the situation is different when it comes to non-governmental organizations, public utility associations, or interstate organizations. It is not in the same sense that we can say that we trust Mercedes (if we are a taxi driver who always buys vehicles of this brand), and Amnesty International, or the Ipcc. In the latter two cases, the trust is partial, because it is delimited by the function claimed by these institutions; but it establishes itself in a quasi-personal mode since it aims at moral qualities identical or analogous to those attributed to individuals. The function of the institution corresponds to a mandate, it induces a three-place relation: we rely on the Ipcc to produce reliable information and forecasts on climate change, its causes, its consequences; we rely on Amnesty International to conduct justified opinion campaigns on violations of individual freedoms by States. But fulfilling these mandates requires dispositions, or qualities, which have a moral value and go beyond mere technical skills, or even the love of a job well done. These dispositions give rise to trust, not reliance. If we rely on an institution to fulfil a positive function, it is possible to develop a different relationship with the institution, a trust relationship justified by the constant dispositions that we attribute to it. These dispositions are not necessarily virtues in the Aristotelian sense of dispositions of character. They can depend on regulations, or standards of practice. It is more reasonable to trust the Ipcc reports because of the actual ethical rules of scientific work, as outlined by Merton than because one believes in the perfect intellectual honesty of all its members.7 To trust in an institution, we do not need to give it a ‘moral character’ (an èthos, according to Aristotle). It would be an undue personification. It suffices that the function it claims presupposes a set of practices invested with a moral meaning to be fulfilled: veracity, knowledge, verification, courage, relevance, and skill. It is these practices, their systematic nature, and their regularity that one qualifies as ‘dispositions’. Such provisions lead to more general expectations than those relating to a particular skill. It is this type of expectation that characterizes trust as a two-place relation: it is not linked to specific expectations or mandates. Rather, it constitutes a horizon of open expectation for everything relating to the fulfilment of a claimed function.

38The link between the positivity of the social function (three-place relation) and the moral dispositions constituting a two-place relation (which can be negative, as for Monsanto, – or positive, as for Amnesty International) is even clearer if one compares the Ipcc and banks. To entrust one’s money to the bank is not the same thing as entrusting one’s savings to Bernard Madoff, or one’s notary, as do Balzac’s characters:

  • by virtue of the law, and their social surface, the banks cannot disappear with our money, or play it at the casino, or build Ponzi pyramids. The rules of the social world do not provide them with the same possibilities as individuals;

  • we have trust in their economic strength. But that does not mean that we give them a character and trustworthy dispositions – while we can do that with our banking advisor, a person to whom we can attribute qualities of honesty, benevolence, and competence. This means even less that we think it is in their character or nature to act in the sense of the public good, and that it is not necessary to subject them to public regulation and control. Trust in banks may be appropriate, but it is not directed at personal dispositions.

39As a result, one can lay down the following rules. For an institution to instil personal trust, it must meet three conditions:

  1. to have a notoriously definite positive social function;

  2. to resort to practices invested with a positive moral meaning, exceeding (but not excluding) technical competence or instrumental rationality;

  3. to have accredited in time its constant dispositions to function according to these practices.

40These three conditions articulate a positive mandate (likely to induce a reliance), moral dispositions, and an identity in time that allows considering the institution as a quasi-person. Conversely, the quasi-personal mistrust of an institution presupposes the condition of identity, an equivocal social function, and immoral dispositions (a set of regular and morally reprehensible practices or inducing mistrust – such as deception, secrecy, lobbying).

41Do public institutions like the hospital and the public education system meet these conditions? In my opinion, they do not.

42To understand this, let us start from the importance of reliance in the construction of trust in an institution. Domenicucci and Holton argue that trust is not reliance reinforced by a particular qualification – as it is standardly treated. One cannot think of trust starting from reliance. It is not even a form reliance extended to many expectations – even though one can imagine that an agent who has repeatedly satisfied various types of reliance may empirically become the object of our trust. But this is neither necessary nor sufficient, and we are still dealing with two styles of heterogeneous relationships. But in the case of an institution, trust cannot be dissociated from reliance, for a substantive reason. Without falling into naive finalism or functionalism, the institution does exist for a function: its nature is determined by what it is for. So, the trust in an institution is linked to its function, whereas the trust in a person is trust in who they are, and this ‘who’ cannot be defined by ‘what it is for’ (its purpose). In the case of the institution, the primordial relation is the reliance or lack thereof. This is why, without a mandate, there can be no Institutional Trust. This does not contradict the thesis that ‘reliance’ and ‘trust’ are two heterogeneous and irreducible relations, which do not rest on the same reasons and do not induce the same system of expectations.

43Can we rely on ‘reliance’ on institutions to build social trust? The first condition is that the social function of the institution at hand is positive. This is the case for the hospital. In addition, I explained at the beginning of this article that trust in the hospital is not only about technical skills and material conditions of functioning, but also specific moral dispositions. But it has also been said that this Institutional Trust is usually based on regulation (the rules of ethics) and control: it is not the same thing to trust the French School of Surgery, and M. X, a surgeon that we know personally. However, Institutional Confidence based on regulation and control combines the calculation of risks and the reliance but does not have a personal character. Finally, the hospital may have built an identity and a reputation in time, but in the same way as Mercedes or Bmw: the reasons for the trust are superimposed on the expectations of what it consists of, which is the characteristic of the reliance.

44Generally speaking, if, like Domenicucci and Holton (2017), one wants to define trust precisely as a two-place relation, different from a three-place relation (relying upon somebody to do something, or entrusting somebody with something), it becomes problematic to think that this relation can be developed with an institution, i.e. an agent that is not a person, or a quasi-person. I have tried to identify the conditions where it is possible, i.e. where an institution may be given quasi-personal moral dispositions. Out of these very particular conditions, trusting an institution seems absurd, or dangerous, because an institution has neither the constant moral dispositions (set of good practices attested in time) nor the identity of a person. I have explained that the main problem of trust, however we unpack it, is that it must be appropriate. One should not bind oneself as a person to institutions that not only are not people, but that, moreover, cannot be subject to moral customization, because they do not function as persons, and have neither their dispositions nor their identity. The fact that an institution may have plans and objectives, and appear fair or unfair, means that it intervenes in the mode of action and that it has certain characteristics of the agents. But these are not the ones that make a person worthy of ‘trust’. Because it takes actions, the institution poses a problem of trust; but not personal trust.

  • 8 See Winnicott 1990.

45This is, of course, different from the personal trust that we can have towards people who work in an institution. This trust is partial but very personal. It is personal, as it is not addressed to an institution but to an individual person in it. It is also partial, because it does not target the entirety of this person. As Winnicott placidly notes, psychoanalysts, who have to be reliable in their work, are not, in their private life, more reliable than anyone else, and the same holds for a nurse or a social worker.8 This does not mean that we trust x, as an analyst, or in their role as an analyst, and not as a person. Because we do not really know what that role is, or the qualities and skills it requires. We would struggle to specify our trust in a series of fixed expectations. And again, the reasons for trust do not match the horizon of trust they produce. The correct wording, therefore, seems to be: we give the person x trust in the context of their function, or it is delimited by their function. Thus a notable characteristic of institutions appears: they can make the people who work there worthy of trust, without them being necessarily regarded as saints. Trust is addressed to the person, but it is well defined by a set of practices and rules and does not necessarily extend beyond. The institution thus correlates an appropriate trust to a bounded trustability. By its operation, its rules, its practices, the institution can make people trustable, in a way that remains fairly light for them. In the context of my work as a teacher, it is up to me to be trustworthy, and not just to be competent, or able to fulfill a number of missions. It is also me that students trust, as a person to whom they attribute certain constant dispositions to knowledge, pedagogy, benevolence, freedom, justice, etc. But this relationship is delineated by a framework of rules and practices that depend on the operation and function of the institution. Trust can take place without the students seeing me as a saint or a hero, and without me needing to be.

46This last remark leads me to address my last question. I hope I have shown that reliance in a public institution is not the same thing as trust in one person, and that specific conditions are needed for an institution to build a relationship of trust, or mistrust, that would be quasi-personal. But what does it mean to rely on a public institution like the education system? And what is the relationship between this impersonal trust and the relationships of trust that can be established between people within the institution? The reasons for trust in an institution must be examined more closely. As this instance of Institutional Confidence is mere reliance, it comes down to understanding the expectations towards the institution. But the individual’s expectations do not depend only on the function and functioning of the institution. They also vary according to the use they have of it, the interests they pursue as users of the institutions. Our questions then take a normative turn: what is a good institution? And what are the good ways to trust institutions?

III. The ideal of trust and the nature of the social world

47The ability to trust some people who run an institution is not a criterion or a sign of the quality of this institution. Since trust is personal, it can very well be given to individuals while the institution in which they work is unsatisfactory. However, there is certainly a relationship, or a proportion, between the degree of reliance or trust being given to an institution, and the possibility of establishing trust relationships with the people who run it. It seems obvious that in an extermination camp, prisoners have very limited possibilities of trusting performers, and this is confirmed by all the memorial literature on this subject. One could therefore consider that the ideal of an institution is that it is reliable and that it favours the relationships of trust with its staff. This is undoubtedly what is expected of an institution for elderly people that are dependent, thus vulnerable: that the institution fulfils its missions, and that we can, as much as possible, personally trust its agents.

48Here I introduce the hypothesis that personal trust is an ideal of relationships, and that institutions can contribute to it, without producing or guaranteeing it. Can one say that trust in people, as a two-place relation, is a provision that must be the norm of relationships? This does not seem plausible. It cannot be a duty to trust in everyone we come into contact with; and I will show that being a ‘trustworthy’ person is more an ideal than a duty because the qualities that make you trustworthy are not easy to define, and much less to acquire. Yet, a social world where ‘trust reigns’, a kind of anti-state of nature where people’s reliability is presupposed, seems eminently desirable. But what kind of trust is it?

49One must distinguish between what is a norm for people, and what they presuppose in their relations to others. For example, according to Kant, the duty to be truthful obviously does not imply the duty to believe that everyone is truthful – which would be a duty of credulity. However, his main argument for basing the duty of truthfulness is that without veracity , the human world would not only be undesirable, but also impossible. Without veracity, linguistic communication loses all meaning: since no one believes anybody, nobody listens to anyone. Now, we have seen that human beings can undoubtedly attain autonomy, in different senses of the term; but not independence, autarchy. This interdependence passes through the use of speech. So the human world presupposes veracity as the ordinary regime of speech. It is because of this presupposition, and to support it, that it is a duty: we must be truthful because others expect us to be, and for it to continue to be expected, i.e. so that the human world is possible. Since in fact, we speak to each other, it is because this presupposition is active, effective. It is not a duty (of credulity), a moral attitude, but a pragmatic presupposition, and a disposition. How this provision is acquired, is a question. But it is plausible to think that it is, and it furthermore varies according to the context. The French moralists of the 17th century, La Bruyère, and La Rochefoucauld, describe the Court as a world where one should not believe anyone and trust no one. Their analysis is not anthropological, but social: what they describe is not the nature of man, but a bad form of sociality, a state of nature policed by manners.

50Can the reasoning on truthfulness be applied to trust? Not exactly.

51First, being trustworthy is more complex than being truthful. To trust someone’s advice (a wine, furniture or painting merchant, a lawyer, a financial adviser) is a more complex relationship than believing it to be true: one can esteem that X is truthful and benevolent, and still not rely on their advice, because one thinks they know nothing about it, or that they speak without thinking. Veracity and absence of malevolence are but conditions of trust, though not all of its conditions (which do not form a closed set of necessary conditions, as we have seen). Trust is directed at an agent who is assigned an open set of dispositions and is given discretion to judge and act without control: they are given the choice of wine, investment, or strategy. However, these provisions that make a person trustworthy are not always within our reach, and we can also in good faith delude ourselves on our account. Being trustworthy thus cannot be a duty. It is rather a personal ideal. We can try to make ourselves relatively trustworthy.

  • 9 See Hirschman 1977.

52Second, is this relative trustworthiness a presupposition of the relationship and cooperation that characterize the human world? Hobbes does not think so. He thinks we can only make human beings predictable, and that is enough. But on the one hand, the world of Hobbes supposes a universal regulation, which in fact is possible only by an absolute power (an effective control). On the other hand, it rests on a mutilated anthropology, which does not give the right to pity or to limited benevolence. To Hobbes’ world, one can oppose Hume’s and that of the theorists of public utility and interests.9 Trust is assured not by the fear of the policeman, but by the combination of three factors: the rules of interest, which make trustability the best policy; the internal discipline by which individuals accustom themselves to subjugate their violent and erratic passions under the thoughtful consideration of their interests, which confers calm, continuity and coherence to their psychic life and conduct; extended benevolence and the sense of public utility, on the basis of a natural disposition to sympathy, the culture of which returns in part to the institutions.

53In Hume’s world, the ideal of relations is not the reduction of risks by the State, legal and essentially repressive control of conduct (Hobbes). Nor is it the contractualization of all relationships, where trust in x is equivalent to the belief in x’s dispositions to satisfy all the clauses of a contract, i.e a mission. Hume’s world is a liberal one, characterized by the free pursuit of self-interest, individual discipline, and the external context of the rule of law (as opposed to the authoritarian state and absolute ruler of Hobbes). In the liberal world, the ‘relative trustability’ of each is both the ideal and the presupposition of relationships. It would be ideal for the rules of the social world to result in each place and function corresponding to a relatively trustable individual. In fact, we presuppose that this is usually the case. The liberal world tends to favor the relations of cooperation and competition, without making them rest exclusively on the fear of the policeman or contractualization. It is a world of free personal relationships, and it is invested with a political, moral and even existential value. To justify this last idea, I will schematically oppose this liberal world to three types of possible social worlds:

  • the world of Hobbes, which we have already seen. Viewed from the liberal world, the Hobbesian subject appears as a fearful and paranoid being, convinced that no human being is trustable and that therefore only an absolutist state can make cooperation possible;

  • the Soviet world as described by Vassili Grossman in Life and Fate, characterized by the assumption that no one is trustable, because everyone is likely to be a state spy, an informer, or a traitor in power. In this world, as in La Bruyère’s court society, the individual must constantly practice self-control or self-censorship, and can never speak freely. This is not a world of free social relations;

    • 10 An individual who is both a boxing promoter and boxer manager is emblematic of the scam world becau (...)

    the world of the scam, personified by the famous promoter and boxing manager Don King, well known for signing his boxers endless contracts, with all kinds of conditions in small print.10 It’s a world where, because no one trusts anyone, everything is under contract.

54By comparing these worlds, one can see that the presupposition of trust is associated with a political choice, and even a moral one, about the type of social world one wishes to live in, and the modality of our ordinary social relations. This is why a liberal education insists on trust – not credulity – towards individuals. From the parents’ point of view, a child who is always defiant and restive is a disappointment or a failure. Trust is therefore a valued attitude, associated with an idea of desirable personal relationships.

55One can therefore wonder about the institutions required by a liberal world. Indeed, trust depends on the context, as we have just seen. It does not depend only on the institutional context. I have just underlined the importance of education. Trust as a presupposition of the relationship also varies according to natural differences, such as day and night: we will be more likely to be wary of a person who approaches us in the street at night than in daylight. But it is impossible to think that institutions do not play a role in establishing a presupposition of trust at work in our relationships with others; and as seen from the beginning, they can only produce trust by arousing it for themselves.

56However, one must not forget here that trust in institutions does not depend only on their function and functioning. It also varies according to individual expectations, i.e. what individuals expect from institutions, and how they relate to them. The question of trust in institutions surrounds not only the reasons for which an institution can inspire trust but also the modalities of appropriate trust. I will show this regarding the school.

IV. The appropriate trust in educational institutions

57I hope to have shown that school is not an institution that can be the object of a personal type of trust. Unlike the hospital, it does not fulfill any of the conditions for this:

  1. the social function of the public education system is much more ambivalent and opaque than that of the public hospital. On the one hand, it is not captured in a transparent manner by the consciousness of users and citizens: it is not ‘constituted’ by this consciousness, it requires to be established by a methodical empirical inquiry, as demonstrated by Durkheim. We are dealing with an opaque social fact. On the other hand, the social function of the school is controversial. Bourdieu and his disciples have posited, arguably, that in France the school also fulfils a function of selection and reproduction, participating in a more general system of social domination. Many empirical surveys corroborate, if not this interpretation, at least the table of social inequalities at work in the institution. Finally, it would certainly be desirable for citizens and users to think that the education system is indeed satisfactorily fulfilling a training function, which will comply the desire to be armed for the social game while paying attention to the provisions and particular qualities of each. However, this is clearly not the case. One does not need to be a sociologist to know that the education system often undermines the self-confidence of the students. It does not promote self-confidence, nor trust in others, things that would make it possible to take part positively in the school world, then in the social world, both being cooperative and competitive.

  2. it does not use a regular and systematic set of good practices to fulfil its ambivalent social function. Certainly, we all have the emotional memory of a few teachers who are at the same time wise, pedagogical, and respectful of the dignity of students. However, did it not concern only exceptional teachers? Do we not think that our children are much more likely to come across bad or very bad teachers in school than bad or very bad surgeons in the hospital? Are we not aware of controversies over methods for learning to read, which place in the public space the hypothesis that perhaps generations of schoolchildren have learned to read in a wrong way? The education system therefore depends, for its optimal functioning, on the quality of the individuals who make it work, and all kinds of school policies, which concern the teaching methods, the orientation of the pupils, or the sectorisation. As these policies vary over time, depending on governments and pedagogical trends, it is impossible to identify the institution as a body of good practices that are stable over time.

    • 11 It differs in this respect from certain educational institutions, private and often religious, whic (...)

    therefore, the education system cannot be the subject of moral customization, unlike the Ipcc or Amnesty International. It is an anonymous institution, whose function and functioning vary according to too many factors to give it a quasi-personal identity over time.11

58As a result, trust, if it exists, will instead combine the appreciation of risk reduction, and reliance on the school to fulfill a number of missions. This possible trust will relate to various realities. Simplifying these realities, I will bring them back to four:

  • the function of the school, i.e. the training of students. It covers the competence of the staff and the good condition of the premises and equipment;

  • security: people entrust their children to school: they expect them to be safe there, and that the institution is not a place where physical and moral violence reigns;

  • equity, equal treatment of students, equal application of rules, lack of favoritism;

  • the dignity, the consideration of individuals in their individuality, and this in two aspects: on the one hand, the absence of arbitrary interference and persecution on the part of the authorities; on the other hand, the possibility to receive training adapted to the particular qualities of the individuals whom the school identifies and emerges, without this training being a place of relegation that considerably restricts the social perspectives of individuals.

59In other words, trusting the school is relying on it to be a safe place for training and socialization. By this, I mean a socialization according to what Durkheim calls the twofold requirement of contemporary societies: on the one hand organic solidarity, i.e. the way in which the individual integrates into society as a whole and respects its regulation; on the other hand, differentiation, individualism, and human rights. We know that the modern school has always had the greatest difficulty in fulfilling this complex vocation and that Durkheim has been preoccupied with it in his sociology of education texts. It is clear that the realization of these four conditions is far from being achieved, and that it represents rather an ideal, that the institution can only merely approach. One can also note that, depending on the case, trust is more a matter of calculating risks (in the case of security) or reliance (in the case of equity: we rely on the school to operate according to fair rules for assigning marks and places). But above all, apart from security, which is a fairly unambiguous state of relations between people, a great equivocity strikes the things we should entrust to the school. Is equity in the distribution of grades and places not compatible with the reproduction or even amplification of social inequalities? What are the conditions of ‘solidarity’, which for Durkheim means that the individual accepts the social place that falls to them (integration), and the rules of the social game (regulation)? Finally, what is a good training?

60This equivocity corresponds, on the users’ side, to a plurality of ways of placing their trust in the institution. Let us schematically distinguish three modalities of trust, three ideal-types:

  • totalitarian trust completely and absolutely subordinates the demands of the individual to those of the State and its institutions. The institution is always right, it is not subject to any challenge or criticism, and its vocation is to shape individuals according to a normative model defined by the State. One should not only think here of so-called ‘totalitarian’ modern States, but also of what Rousseau calls, in Emile, the education of the citizen (as opposed to the education of man), and which refers to education in ancient republics, such as Rome and Lacedaemon;

  • the libertarian trust is the opposite of the previous one: the individual trusts the institution insofar as it serves their private interests. They conceive of trust in the form of contracts concluded between them and the institution, which they undertake to respect the rules, provided that they guarantee the achievement of objectives judged by them to be advantageous, in accordance with their precise expectations;

  • the trust, which we will call liberal for lack of a better word, in a sense which will require to be specified, keeps at a distance the two preceding modalities of trust. It refuses credulity and blind conformity, it preserves, as a precious thing, a critical dimension; but it does not subordinate its adherence to the achievement of missions and objectives placed solely in the perspective of individual advantage.

61The distinction of these three modalities of trust has a normative meaning. Liberal trust is built on criticism of the other two modes of trust in the institution.

62Totalitarian trust may seem either caricatured or out-dated. It may be, however, an ever-attractive pole of our relation to institutions. Aren’t we regularly attracted by such totalitarian trust, in much the same way as by the effects of authority mentioned with regard to the hospital? Schoolteachers are well aware that some parents develop a quasi-totalitarian mode of trust in the school, especially since they see it as a means of social climbing without knowing the codes. If the school is regarded as the instrument of a social salvation whose paths are unknown, then it is always right, and its rules and verdicts must be respected unconditionally so that children must fully comply with them.

63Yet totalitarian trust is no more suited to modern sociality than to the nature and function of the institution:

  • from an internal point of view, the training function implies a form of trust without which teaching would be very difficult. The teacher and the psychoanalyst alike: if you do not trust them, their work will be neither fruitful nor even possible. This trust always has a personal character. The idea that one would trust a teacher, regardless of their person, to accomplish a certain number of assignments, seems very abstract, and I will also show that it is based on a wrong conception of training. I do not think we should deal with an abstract figure, the teacher, invested of certain qualities on the basis of their status. We should rather be conscious that we are dealing with a person, who is always individual, in whom one has a trust delimited by the frame of their function (and potentially more). Be that as it may, this trust has nothing to do with a blind follow-up or a sheepish credulity. The teacher is not a guru. They do not ask for unconditional trust in who they are, but a reasoned trust in what they say, especially since, as Durkheim recalls, the values of individualism envelop reason and the spirit of examination. In the teaching relationship, within a liberal education, trust is a flexible presupposition, which must regularly watch over its appropriateness;

    • 12 See Bourdieu 2012; 2013; 1994.
    • 13 See Boltanski 2009.

    From an external point of view, the education system does not only fulfill its central and immanent training function. As an institution of the State, it participates in and to its authority, which according to Bourdieu is the ultimate certification body.12 Bourdieu is perhaps exaggerating the State’s certifying power in a liberal society in which the market plays an important role of evaluation (what Foucault terms ‘veridiction’). But in fact, it is the State that issues the diplomas, empowers the experts, buys works of art to put them in its museums, etc. In this sense, according to Luc Boltanski, State institutions fulfill a ‘semantic’ function; they define what is the reality, what matters, and the right ways of describing it.13 Because of this ‘semantic function’, the institutions of a liberal society call for a constant critical debate about how they operate and certify. They tend to form ‘State spirits’, but a liberal State cannot claim totalitarian trust. On the contrary, its institutions will appear all the more trustworthy if they admit criticism. Precisely because reliance in the institution is largely related to its function, if it escapes the critical questioning on this function, and on its way of fulfilling it, then one cannot rely on it.

64Libertarian trust seems safe from credulity and blind conformity since it consists of an essentially instrumental relationship with the institution. The libertarian user does not differentiate between public institutions, car firms, insurance companies. They count on a certain number of services, and the more explicit and detailed they are, the better it is: quality courses, the best teachers (who must thus be recruited), profitable school trips, a socially chosen public, etc. It is a reliance relationship, a contractual relationship, and the concern of the user is to not be ripped off, to get value for money, or for the efforts they have made to obtain a place in the institution.

65However, this contractual mode of trust has its limits. Teachers may be happy to be considered as ‘the best’, but as such, they probably feel that they should be trusted in a less contractual way, and they should have some degree of initiative in their training work. From the teachers’ point of view, it is problematic to see their work being reduced, by the users or by the institution itself, to a series of ‘missions’ or ‘objectives’, as if one could objectify in this way what is ‘good training’, or the right way to teach. It is then apprehended as a sequential process entirely driven from the outside by objectives that precede it. But in the reality of a teacher’s job, things do not happen, or should not happen, this way.

66Training is a process that is reflected in its course itself, depending on the students, the difficulties of the topics covered, and also for more undetermined purposes than a series of objectives. There are different ways to be aware, understand, and know how to do something. A teacher does not ask for the trust of a Mission: Impossible character, they are not heroes with superpowers that will allow them to carry out the missions they are entrusted with. They need to be allowed a certain degree of discretion and initiative, although framed by their function, in defining the ends they pursue and the means they engage. From a certain degree of complexity in the missions, these require the ‘discretion’ in the use of the means that characterize trust of the personal type. In other words, trust as a three-place relation, with a series of mandates or missions, meets its limit when it comes to too complex tasks, or too vague, to be reduced to an ordered series of instructions.

67I would not trust an education system that would represent its training function in a table of objectives, and I would entrust my children to it with apprehension. I would prefer to think that the institution, in addition to the knowledge and skills that it has the function of transmitting, also allows students to ‘become’, intransitively, and that teachers are, in their work, penetrated with this concern.

68Of course, this becoming cannot be left to itself, since by nature it involves hardships, difficulties, and that it is the function and the necessity of education to make it possible to overcome them. I would expect the educational institution to have an understanding, an empirical knowledge of these hardships, and to accompany them. It is up to it to enter them in a collective or social time, to mark the stages and the ends (to learn to read, to count, etc.), without the process that it leads being entirely determined by external objectives that would be fixed in advance. It is by assuming this responsibility to accompany the trials, while at the same time thoughtfully arranging the time when we confront them and where we gradually overcome them, that the institution really ‘accompanies’ the individual, be it the pupil or the teacher. In doing so, while endowing them with initiative, it eases their responsibility by casting this initiative in an empirically structured framework. It is a very important thing to know how to read, but we can learn it quietly, and at one’s own pace. If I thought that the education system conveyed this awareness of training to its staff and students, that would be a reason to trust it, in the sense of relying on it.

69All in all, I went from the observation of the need for trust in institutions to a reflection, admittedly embryonic, on the conditions of a liberal trust. The latter is not only the business of institutions. It also depends on individuals, their expectations, their way of relating to public institutions such as hospitals and schools. However, it is up to the institutions to care about this relation, to cultivate it, to develop practices that induce what I call liberal trust, rather than claiming totalitarian trust in their authority, or to try to accredit themselves by offering a range of goals and services to an informed consumer audience.

70I argued that the form of Institutional Confidence we develop with the school and healthcare systems is not a two-place relation, analogous to the trust in a person. In some respects, such as personal safety, the so-called trust in institutions is a calculation of risks, and thus risk-taking made reasonable by the awareness of risk reduction. In other respects, like their social function, Institutional Confidence is a three-place relation: we rely on the institution to fulfill its function, and our reliance equals the sum of our expectations – which is a characteristic of such a relationship. But when this function is opaque, complex, and ambivalent, it is impossible to specify the reliance in a system of well-defined expectations, or in a completely explicit contractual mode. This is why liberal, and not totalitarian or libertarian trust, seems appropriate. What is central in liberal Institutional Confidence is reliance on the institution to define and regulate the practices that allow it to fulfil its own function, while leaving a latitude of initiative to its agents. This latitude is proportional to the complexity of the function. In the case of the school, what we expect from the institution displays the indeterminacy of our reliance, without being transformed into a binary trust. Actually, we do not clearly know what a good education looks like, even if we know that it involves the confrontation with certain ordeals that occur according to an order and a temporality. Here we are dealing with a particular case of reliance as a three-place relation, since this reliance cannot, or should not, be thought and formulated in terms of mandate, mission, or contract. It is because of this indeterminacy of reliance that trust in the school needs the development of personal relationships of trust with teachers and students.

  • 14 It goes without saying that trust in the education system depends on many factors independent of th (...)

71Finally, liberal trust envelops the consciousness of the ambivalence of many State institutions, such as schools. This is why it is strengthened, and not diminished, by the institution’s ability to accept criticism and to draw inspiration from it to improve itself. The prospect of improvement is inescapable: the conditions in which an institution is worthy of a high degree of trust are difficult to bring together in the case of school. It can only be ‘good enough’ to build trust, and it is always difficult to gauge this ‘enough’.14

Torna su


Boltanski, L., 2009, De la critique, Précis de sociologie de l’émancipation, Paris, Gallimard.

Bourdieu, P., 2012, Sur l’état, Cours au Collège de France 1089-1992, Paris, Points essais.

Bourdieu, P., 2013, Manet. Une révolution symbolique, Paris, Seuil.

Bourdieu, P., 1994, ésprits d’Etat, in Raisons pratiques, Paris, Points essais: 99-145.

Bowlby, J., 1963, 1973, 1980. Attachment and Loss, New York, Basic Books.

Canto-Sperber, M., 2017, L’oligarchie de l’excellence, Paris, Puf.

Domenicucci, J., Holton, R., 2017, Trust as a Two-Place Relation, in P. Faulkner, T. Simpson, The Philosophy of Trust, Oxford, Oxford University Press.

Goffman, E., 1961, Asylums: Essays on the Social Situation of Mental Patients and Other Inmates, New York, Doubleday.

Hirschman, A., 1977, The Passions and the Interests: Political Arguments For Capitalism Before Its Triumph, Princeton, Princeton University Press.

Laznik, M.-C., 2013, Pulsion invocante avec les bébés à risque d’autisme, in G.C. Crespin (ed.), La voix. Des hypothèses psychanalytiques à la recherche scientifique, “Cahiers de Préaut”, 1, 10: 23-78.

Merton, R.K., 1973, The Normative Structure of Science, in N.W. Storer (ed.) The Sociology of Science, Chicago, Chicago University Press.

Walzer, M., 1983, Spheres of Justice. A Defense of Pluralism and Equality, New York, Basic Books.

Winnicott, D.W., 1990, Deprivation and Delinquency, London, Routledge.

Torna su


1 It can only be laid out roughly, and would call numerous developments.

2 Rawls clearly identifies the relationship between the institutional system that organizes the basic structure of society and general principles of justice.

3 See the works of two specialists in child psychology (Bowlby 1980) and (Laznik 2013). Convergence between these two authors is all the more interesting that they come from different theoretical resources (ethology for Bowlby, Lacanian psychoanalysis for Laznik).

4 The variation of the principles of distributive justice according to the types of good to be distributed is one of the aspects of the ‘complex equality’ advocated by (Walzer 1983).

5 See (Goffman 1961). A total institution is a place of residence and work where a large number of individuals, placed in the same situation, cut off from the outside world for a relatively long period, lead together a secluded life whose modalities are explicitly and meticulously regulated. Some examples are prisons, concentration camps, asylums and monasteries, but also boarding schools, orphanages, etc.

6 The case of the prison may be debatable. But would you deny the difference between a prison, and a hospital or school? The army is an equivocal case.

7 Merton 1973.

8 See Winnicott 1990.

9 See Hirschman 1977.

10 An individual who is both a boxing promoter and boxer manager is emblematic of the scam world because it pursues two conflicting interests, which should make its two functions incompatible. A promoter seeks to propose the most interesting fights, therefore the most risky for each boxer. A manager seeks to protect his boxer as much as possible.

11 It differs in this respect from certain educational institutions, private and often religious, which claim a tradition and a culture maintained through time. I can tell myself that I have a high idea of ​​the institution or what it should be, but for me, it makes no sense to say that I have a quasi-personal trust in National Education, or even in its institutions such as the institution where I teach and those where I studied. I have no personal trust in the Sorbonne or the École Normale Supérieure. But I think that it makes sense, for some people at least, to trust in institutions that claim moral values ​​giving them an identity in time. In this case, the institution is credited with an identity that encompasses not only the quality of teaching but also the transmission of moral values, concern for justice, attention to people, charitable benevolence, etc. Such institutions fulfil the conditions of moral customization.

12 See Bourdieu 2012; 2013; 1994.

13 See Boltanski 2009.

14 It goes without saying that trust in the education system depends on many factors independent of the users’ expectations: the unemployment rate, the reasons for believing in a relative equality of opportunity, the more or less oligarchic character of access to the most prestigious sectors. Monique Canto-Sperber (2017: 240) asserts that in France, selective sectors, such as the classes préparatoires and the grandes écoles, ‘manage to instill in their students trust in the value of the training they have acquired and the diploma sanctioning it’. I wondered what we could hear by this ‘trust’. I tried to show that it was not a personal confidence, and that it should not be reduced to a blind adherence, a social follow-up, or a calculation of interest – which it can be.

Torna su

Per citare questo articolo

Notizia bibliografica

Pierre Lauret, «Why (and how to) trust Institutions? Hospitals, Schools, and liberal Trust»Rivista di estetica, 68 | 2018, 41-68.

Notizia bibliografica digitale

Pierre Lauret, «Why (and how to) trust Institutions? Hospitals, Schools, and liberal Trust»Rivista di estetica [Online], 68 | 2018, online dal 01 mars 2019, consultato il 23 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