Ethics posed the question of whether trust was rational, right and morally necessary but it was incapable of resolution. Some preference for trust can be perceived in the statements cited in Chapter 1,1 but such a preference cannot be expressed without reservations. There are obviously some cases which call for trust and other cases which call for distrust. This is rationally indisputable. It therefore follows that trust cannot be a maxim for conduct which is valid without exception. Ethics must, therefore, presuppose that whether in particular instances one should trust or not follows from the objective features of the situation, from common human understanding. Therefore, even if a general rule is formulated about trust as a principle, the decision as to whether it should be followed or not must be delegated and left to the situation. Furthermore, it must be presumed that the situation, and in particular the object of trust, displays sufficient objective features to ground a judgement that can serve the development of trust, and that these features can have the same meaning for all people and the same relevance for trust. Only thanks to these basic suppositions, which conceive the world as consisting of multiple forms and pervaded by malice and deceit but nevertheless socially objectified and stable, could the ethical mode of treating the problem be meaningful.
One could keep to the old view that this way of treating the problem is determined by its subject of reference and, to that extent, is correct. To demand more precision than the subject offers would be irrational and could lead to erroneous, excessive abstractions. This long-established argument is not to be refuted by direct rebuttal. But that does not settle the questions of its cognitive value, of whether it is even answering the right question of what its basic premises are, and, above all, of how it stands in relation to the structure of the social order which it presupposes in its premises.
A moral principle which also acknowledges its opposite must be reliant on practical situations that can be unanimously and unequivocally interpreted, and where it can be firmly settled whether the principle or its opposite is to be applied. In so far as situations for such interpretation are lacking, the principle loses both its value as a means of orientation and its normative function. For it could provide justification for each and every decision. Consequently, the normative character of such maxims becomes important only if, and so far as, one can expect social situations where the factual and social complexity has already been reduced to a considerable extent. This can be the case in a relatively simple social world, or, again, in fairly strongly regulated parts of a more complex reality – for example in the field of law, or in organizations.2 Trust is, however, typically demanded precisely where other means for simplifying the orientation of action and guaranteeing it fail to work. If one adds that the social world in general nowadays is, with some reservations, much too complex to permit an ethics of principles to function as a theory of action, then it becomes questionable whether we should continue to apply the ethical mode of thinking to the problem of trust. In particular, lowering the scope of the social dimension through social psychology and phenomenology to provide insight into the impact of social relationships upon all experience, makes it doubtful whether an ethic which seeks to offer instruction in correct personal decision-making is adequate for our problem of the rationality of trust.3
If it is not, then the rationality of trust and distrust must be seen anew and in a different way as a problem. To that end one must first discover where the cardinal point of the problematic, the heart of the difficulty, lies. Obviously social reality is much too highly differentiated for it to be abstracted into a simple but instructive ethical maxim for decisions regarding trust. The problem of the orientation and directing of action must be solved by very many differentiated means. The sciences of human action can no longer allow themselves the illusion of laying down to the actor, if only in terms of vague principles, how to act correctly, in some direct, immediate, fashion, of telling him what he ought to do. Scientific-analytic and practically oriented perspectives must be more rigorously separated and consciously worked out separately from each other. Such a difference in style between theoretical analyses and practical information-processing is no drawback to the collaboration of theory and practice, but rather precisely the basis for it, and a meaningful division of labour in the joint mastering of an exceedingly complex environment.
If one focuses the relationship of theory and practice, so understood, upon the problem of complexity, then it seems obvious to locate the problem in the difference between modes of selection, and thus to summarize it in a nutshell.4 A usable division is then offered by the distinction between system-perspective and action-orientation which seems to be gaining ground in recent developments in the social sciences. Psychology and sociology have strong theoretical tendencies to becoming sciences of personal and social action systems respectively – to becoming sciences which in their systems theory include unconscious, latent and incongruous action-oriented perspectives, which make comparative research possible, recognize structural contradictions in systems, and in all this achieve a potential for complexity which overtaxes the actor-in-a-situation. Under the pressure of this expansion of the field of vision, other sciences which pursue prescriptive aims must transform themselves, by means of stricter specialization and a greater awareness of their problems, into theories of the rational reduction of complexity – into decision theories. In their final form they endeavour to work out and establish a calculus which the actor can apply in predetermined, constructed ‘model-like’ situations, without having to think through and decide anew on the functional context of his behaviour in each case. He must be able to presuppose that system structures and decision programmes have already taken over this part of the work from him, that, in other words, the social world is organized.
What, then, does the pronouncement ‘rational’ refer to in such a social order when the problems are set out in this way, and to what does trust refer? If one were to take as a yardstick the concept of rationality in decision-making theories – be it that of rational choice in the employment of means, or that of optimality – one would from the outset fall into a too narrow conceptual frame of reference which cannot do justice to the facts of trust. Trust is not a means that can be chosen for particular ends, much less an ends/means structure capable of being optimized. Nor is trust a prognosis; its correctness cannot be measured by the occurrence of the predicted event and after some experiences have been reduced to probability values. These types of techniques, which are significant within the framework of decision-making models, have, as does trust, the function of reducing complexity. They are functional equivalents for trust, but not acts of trust in the true sense. As far as they extend, trust is unnecessary. They can replace trust, just as, conversely, the need for trust as a complementary way of absorbing uncertainty is a result of the limited effectiveness of those decision-making techniques. Trust is, however, something other than a reasonable assumption on which to decide correctly, and for this reason models for calculating correct decisions miss the point of the question of trust.
In a more widely conceived sociological theory of rationalization, the preparatory work for which is missing in the prevailing empiricaldescriptive orientation of sociological research, the evaluation ‘rational’ could follow from functional analysis. All activity which helps to orient human action more meaningfully in an exceptionally complex world, thus increasing the human capacity for comprehending and reducing complexity, would then have to be judged rational. This can only be done by means of system formation. So, from this point of view, the label ‘rational’ would not refer to decisions about particular actions but rather to systems, and functions for maintaining systems.5 We will take this idea as a basis, especially as it was already implicit in our functional analysis of trust, and see what it leads to.
Trust is rational in regard to the function of increasing the potential of a system for complexity. Without trust only very simple forms of human co-operation that can be transacted on the spot are possible, and even individual action is much too sensitive to disruption to be capable of being planned, without trust, beyond the immediately assured moment. Trust is indispensable in order to increase a social system’s potential for action beyond these elementary forms. Completely new types of actions, above all such as are not immediately satisfying and hence have to be artificially motivated, become possible in a system which can activate trust. Through trust a system gains time, and time is the critical variable in the construction of complex system structures. The satisfying of needs can be delayed, and nevertheless guaranteed. Instrumental action, oriented towards distant effects, can become institutionalized if the temporal horizon of a system is suitably extended by means of trust. The availability of liquid financial resources, power and truth, all mechanisms dependent on trust, makes possible an indifference on the part of the system towards numerous events in the environment and thus a gain in reaction time.
This judgement of trust as system-rational can nevertheless not remain unqualified and be interpreted simply as an assertion about trust. Of course, trust is never the only mechanism for the reduction of complexity; the need for trust depends on the availability, or non-availability, of functional equivalents. Trust requires numerous auxiliary mechanisms of learning, symbolizing, controlling, sanctioning; it structures the processing of experience in a way which demands energy and attention. Above all, however, not just in individual cases, but much more at the system level, trust depends on the inclination towards risk being kept under control and on the quota of disappointments not becoming too large. If this is correct, then one could suppose that a system of higher complexity, which needs more trust, also needs at the same time more distrust, and therefore must institutionalize distrust, for example in the form of control.
Accordingly, system rationality cannot be attributed to trust alone. It lies rather on a level that encompasses both trust and distrust, namely in the binary schematization of a more elemental relation to the world into the structured alternatives of trust and distrust.6 The advantages of such a schematization should be compared with more strongly formalized and specialized binary codes – for example that of right/wrong or that of true/untrue. In all these cases the definitions of the situation which are set against each other are at first logically inconvertible entities.7 By means of binary schematization they are treated, however, as if one could be turned into the other by mere negation. Thus the transition from the one form to the other is made easier, both come closer together precisely in that they are thought of as opposites, and in this lies the gain in rationality. For the simplicity and directness of the transition to the opposite makes tolerable a greater risk in determining the system. The relative inferiority, the relatively low ‘level of technicality’, of the trust mechanism in comparison with the truth-code or the legal-code can be seen, among other things, in the greater difficulty of reversing negation – trust is easier to transform into distrust than is distrust into trust.
Such thoughts are distantly analogous to the position which used to be adopted in ethics, namely, that trust should be the rule and distrust the exception, that trust is therefore to be preferred in cases of doubt but must leave room for distrust. However, there is a difference in that these considerations cannot be converted into the currency of directives which action may follow. In deciding about the particular case, trust and distrust exclude one another. Their relationship must therefore be constructed by a science of action such as ethics in the sense of an either/or, according to the rule/exception schema. When related to systems and seen as general mechanisms, trust and distrust can be increased side by side, in so far as aspects and situations can be sufficiently differentiated. Admittedly, no further indications can be gained on this abstract level of consideration as to whether one should trust or distrust in any particular case.8 Only very much more exact analyses of particular systems could help to form a basis for a decision on that issue. Systems theory can, however, found a judgement on the rationality of the trust mechanism and on the particular system conditions under which it can perform its function.
A system-rational increase in the effectiveness of trust will depend on all the aspects of trust formation which we have treated in the course of our inquiry and cannot catalogue once again here. Nevertheless, perhaps the decisive question remains open as to whether and how it is possible for trust and distrust to be co-ordinated through the creation of systems, and so increased in tandem. For reasons of general systems theory which cannot be adequately developed here, two closely connected processes ought to be decisive. These are the differentiation of the system from its environment, i.e. the drawing of boundaries, and the inner differentiation of the system, i.e. the functional specification of its subsystems and mechanisms.
It is a basic thesis of systems theory that systems constitute themselves by means of the distinction between inner and outer and maintain themselves by stabilizing this boundary.9 If we conceive rationality in the sociological sense as system rationality, it is plausible to seek in this inner/outer distinction a rational criterion for the distinctive location and the joint increase of trust or distrust. In working out this idea, care must be taken to distinguish different system references. The reference system whose inner/outer distinction we are now considering is not the system which trusts or distrusts. It is not therefore a matter of distinguishing external and internal conditions for trust, which was dealt with earlier. We are referring now rather to the case where a system, be it a person or a social system, takes part in another social system, so that it becomes enmeshed into a social system through membership and is called upon to decide, in questions of trust, whether from the point of view of this membership system trust is required for system-internal or for system-external processes. Although system nestings of many sorts are involved, we can, to simplify matters, consider the case of the membership of people in organizations.
As a first, rough, approximation, one can suppose that internal processes earn and maintain more trust than external ones, that one trusts one’s colleague in his role more than an outsider, a fellow club or party member more than a stranger. That such a differentiation is widespread and can be rational, that the dividing line between the familiar and the non-familiar person supports it, is not something to be dismissed out of hand. One finds system boundaries characterized as trust boundaries particularly in all social systems engaged in operations which are not supposed to be revealed to the outside or are even illegal and therefore have to be kept secret.10 A functional analysis of the problems of complex social systems certainly does not give this simple picture. For many systems it can be vitally necessary that they are in a position to invest trust in their environment so that they can take part in relationships which are only to be achieved through mutual trust. The system members must then be able to display trust externally also. In the lower positions in modern large organizations, for example, internal relationships can be programmed in such detail that trust between members becomes virtually unnecessary because uncertainty in behaviour is overcome by other mechanisms. For some systems also, it is precisely in their internal relationships that substantial injections of distrust are needed for them to remain alert and capable of innovation, so as not to fall back into their customary pedestrian ways of relying on one another.
The inner/outer differentiation should therefore not be equated unproblematically with the boundary between required trust and justified distrust. With greater social differentiation and system specialization especially, there exists a tendency for internal distinctions to increase and for those connected with the environment to decrease.11 Even then, however, system boundaries make possible a different strategy for the apportioning of trust and distrust in the case of internal processes than in the case of external ones, in the sense that members trust or are distrustful in a different way and for different reasons. The internal world of a system is quite different from the external world and, therefore, no one is obliged to be ‘consistent’ in trusting beyond the system boundaries. One can, for example, accept without question the opinion of one’s colleague about a technical matter but nevertheless not risk lending him money ‘personally’. System boundaries act as critical thresholds, in the sense discussed above,12 at which familiarity and trust can switch to distrust, system trust to personal trust or distrust, or distrust to indifference.
A greater measure of trust can be established internally by means of selective processes for the choice of members within the framework of the selection criteria. On the other hand, on this basis a sharp distrust of a quite specific kind can be elevated to the level of formal duty, for the purposes of supervision, for example. In the case of external relationships such system-structural reasons for trust and distrust are irrelevant and things come to depend more on the specific learning and confirming of trust relationships in the contact between system and environment, on the freedom of traffic allowed by the system boundaries, or on the strength of the system and the cover it guarantees its members in the case of distrustful behaviour leading to disruptions and conflicts in relationships with the environment.
It is precisely this differentiation of the approaches to trust and distrust which is, from the point of view of the system, rational. For it assists it in preserving the higher level of inner order in comparison to its environment, or, in other words, in stabilizing, in an extremely complex environment, a simpler, less complex system-order which is suited to human capacities for action, and in reducing the complexity gap between system and environment. In the repertoire of system strategies and latent functions which serve such system maintenance, trust naturally plays only a limited role, both internally and externally. The more effectively the environment of a system is protected by more encompassing systems from too intense, unpredictable fluctuations, the more effectively the system can move over to directing its actions by internally rationalized decision-making techniques, and replace trust with calculations of probability. Whether, and under what conditions, such a process of substitution is itself rational can likewise only be judged within the framework of our comprehensive idea of rationality, whereby trust, which has apparently come about ‘irrationally’, may appear rational if and in so far as it performs the functions which serve the maintenance of the system.
A further aspect which can contribute to the rationalization of trust and distrust in social systems has already been mentioned on numerous occasions in the preceding pages and must now be worked out in its particular problematic. This is the specifiability of the respects in which one trusts or distrusts. As soon as, and in so far as, the boundaries of a system are defined, it can allocate specific functions internally to subsystems and mechanisms and thus define more precisely actions, situations and roles with regard to which trust or distrust is expected. What would be absurd as a general rule of conduct, or could not be motivated, can be achieved within the boundaries of specific systems which themselves enjoy the trust of their members.
Admittedly, trust and distrust are, in the way in which complexity is reduced, confused in principle and oriented towards concrete persons or groups or the objects and events symbolizing their trustworthiness. That does not mean, however, that trust relationships cannot be restricted to specific aspects. One can trust someone in matters of love but not in money matters; one can trust his knowledge but not his skill, his moral intention but not his ability to report objectively, his taste but not his discretion. The reason for this specificity can be simply that trust was learned in this restricted sphere and has foundered in other respects. It may, however, also go back to a selective interest of the system to which the parties concerned belong and which even structures the learning process itself.
The possibilities of system rationalization seem to lie in such a pre-structuring and legitimation of specific opportunities for trust and distrust. In this way systems are able to provide for trust and distrust alongside one another, indeed, interlocking them in a variety of ways so that they intensify one another. Where the mobility of a two-year-old son is concerned, different members of a family have a legitimate distrust and, at the same time, mutually trust their distrust. In organizations, checks can be set up which operate under a specified order to distrust, and here too place others, often even the people who are themselves controlled, in a position of trusting the functioning of this distrust.13 There are even roles, for example, those of researchers or judges, where, within the specific horizon of their task, trust is allowed to be treated as distrust, and where reports about trust are regarded with distrust.14 Trust in systems as a whole can, as we saw, depend decisively on trust being interrupted at critical points and distrust being switched on. Conversely, only in systems which are trusted can distrust be so institutionalized and restricted that it is not regarded as personal and reciprocated, so remains protected from escalating into conflicts.
If the effect of these various mechanisms is to be increased, their combination must be secured independently of personal motivational structures and proclivity for risk. This can only be done through organization, which brings into play new, impersonal, motives for action. To this extent, moreover, a differentiation of the system to be trusted is a prerequisite for extensive inner specification. Organization in no way makes trust and distrust superfluous but it depersonalizes these mechanisms. The person who trusts no longer does so at his own risk but at the risk of the system. All he has to do is to take care still that no detectable mistakes creep into his bestowing of trust. The person who distrusts no longer does so by going back to personal modes of reduction, such as personal animosity, hostility or safety precautions, but does so on the basis of the system, which has already programmed in advance the mode of behaviour for cases of disappointment, and guards the distruster against any excess.
Characteristically, the organization in which trust and distrust are the object of expectations with regard to aspects of a work task does not completely remove the choice between these two possibilities from the actor. Even the truster must retain a modicum of distrust – he must, for instance, intervene if his colleague gives an opinion which is obviously false. And neither can the distruster, as in the case of supervision, take his justified distrust to extremes without becoming a liability to the organization. What in ethics is expected from the nature of things is performed by means of the organization and specification of the system, that is, clear directions in a particular case as to whether trust or distrust is appropriate and rational.
All in all, system theory accomplishes more than ethics in that it makes comprehensible the system’s specification mechanisms. It too, in the final analysis, cannot tell the actor how he should act and whether he should trust or not. It does, however, possess the possibility of showing clearly how systems can be set up in which, despite high complexity, it can be left to the actor to decide whether to trust or not. Systems are rational to the extent that they can encompass and reduce complexity, and this they can only do if they possess understanding of how to make use of trust and distrust without placing too heavy demands on the person who finally shows trust or distrust, that is, on the individual.
These considerations lead us back to our starting point, to the problem of social complexity. Historically as well as factually, trust takes on many various shapes. It has a different character in archaic social systems from what it has in civilized social orders; it can be trust which arises spontaneously, or which is personal and built up in a tactical-insightful manner, or it can be trust in general system mechanisms. It avoids a clear-cut ethical instruction. Only from the point of view of its function can it be understood as a whole and compared with other functionally equivalent mechanisms. Trust reduces social complexity by exceeding available information and generalizing expectations of behaviour in that it replaces missing information with an internally guaranteed security. It thus remains dependent on other reduction mechanisms developed in parallel with it, for example those of law, of organization and, of course, those of language, but it cannot, however, be reduced to them. Trust is not the sole foundation of the world. But a highly complex yet nevertheless structured representation of the world could not be established without a fairly complex society, which in turn could not be established without trust.