Chapter 3
Cognitive dissonance

In this chapter, we will discuss the contributions of twentieth-century sociology no longer at the level of understanding group dynamics, but at the level of the functioning of individuals. The book which is going to serve as our common thread is little known to non-sociologists. It is A Theory of Cognitive Dissonance written in 1957 by Léon Festinger.

Implications of cognitive dissonance on our vision of the human

The cognitive dissonance theory proposed by Festinger asserts that an individual experiences stress each time two pieces of information brought to his attention are contradictory, dissonant, and that he will therefore seek to eliminate this dissonance. The heart of Festinger's book is a classification of the different methods that the individual can use for this purpose.
What interests us more particularly here is that often one of the pieces of information comes from the experience of the individual, and the other from an event which suddenly contradicts it. In this case, the rational attitude of questioning the experience gained is in fact infrequent. Forgetting the element that brought the contradiction, rejecting its validity on the pretext that it is not fully proven, seeking and adopting new beliefs to reinforce those that have just been shaken, or considering that there is no contradiction on a more general picture, are much more frequent behaviours.

When synthesizing to the maximum, Festinger makes this key remark for the rest of this book: contrary to the received idea of ​​a rational individual who acts according to his reason, scientific observation shows an individual who adjusts a posteriori his beliefs to justify his actions.
To illustrate this phenomenon, let us return to one of the experiments carried out by Festinger. People are invited to participate in a perfectly boring experience, for which they are paid, some by a small amount, others by a larger amount. Then we ask all these people what they think of the experience. The result is that poorly paid people express a more positive opinion about the experience. The explanation is that when the sum is not sufficient, individuals seek other explanations to justify their action afterwards, in the present case having agreed to participate in the experience.
In the same logic of a posteriori self-conviction of his attitude, if participation in an experience requires more effort, then it will be seen as more interesting.
Here is another less intuitive example of the effect of cognitive dissonance: people who have to choose between two products will spend all the more time continuing to collect positive information about the chosen product after having made the choice, that at the time of the choice the perceived merits of the selected product were not significantly more important than those of the other, so that after a certain time, the gap between the two products would seem to them greater.
As a summary, contrary to the idealized vision that we have of the human being as a rational being at the summit of evolution, Festinger shows us an individual who seeks more to justify a posteriori his errors than to correct them. So he tends to repeat them, and the more he repeats them, the more going back becomes dissonant, therefore difficult.

This phenomenon of a posteriori consolidation of choices, regardless of their relevance, has a huge consequence which is particularly interesting for the rest of this book: what is most difficult is to have an individual modify his usual action, event if it was inappropriate and the elements to prove it relatively simple and clear.
Furthermore, what Festinger shows in the section "Maintaining incorrect convictions" in chapter X is that, in the case of complex situations, simple social support - that is, the fact that others make the same mistake as we do - is often enough to collectively maintain a wrong position. However, most of the choices for social organization are complex, so that being usual may be enough to maintain the illusion of their relevance. Basically, the "we always did like that, so it's good" allows to get rid of the arguments that question the merits of the use, as long as the situation is complex, that is to say say almost all the time.

Let's continue the list of bad news: as soon as we accept an action under duress, whether it is a strong duress like a physical threat, or a weaker one like working for a living, cognitive dissonance generally acts in the sense of gradually making us accept the merits of these actions. We will come back to this in the second part of this chapter, which is more specifically devoted to the world of work.

Finally, the problems revealed by cognitive dissonance do not stop at the difficulty of correcting errors. It should be added that the same mechanisms provide the individual with the possibility of getting rid of his responsibility cheaply, and thus of preserving his good conscience.
For example, the simple fact of not agreeing will tend to be interpreted as exonerating co-responsibility for the acts of the group, even when the person has not strongly opposed it, or even opposed it outright. Let us quote Hannah Arendt in Eichmann in Jerusalem: “'The inner emigrant' saves his conscience by not agreeing inwardly, but without changing his action ... so as not to be unmasked! " (0)
Again, we have the illusion of a rational individual, making choices consciously. This probably comes from the stories of our childhood, in which the wicked are aware of their wickedness, and assume it. Now cognitive dissonance shows us that acting against the interests of others is generally accompanied by a lie to oneself, which by using the methods listed in the theory of cognitive dissonance, generally allows the individual to act in good conscience in the end, all the more easily as the situation is complex.
In other words, the world that is going wrong is not the result of a few individuals who do evil in conscience like the bad guys of the stories of our childhood, nor even the result of the actions of which we are not proud but which we do nevertheless by opportunism, but much more the result of the enormous mass of actions that we carry out against the interest of others, and for which the lie to ourselves allows us to experience no bad conscience, quite the contrary.

Like Marx and Parkinson, Festinger starts from observations and not from speculation, then offers an explanation of the causes, but gives few solutions to change this state of affairs.
What is stranger, given the importance of cognitive dissonance theory for the understanding of human cognition, is that Festinger forgets to address two important questions related to cognitive dissonance, which will not be more so by its multiple successors who will be content to seek to invalidate or refine the initial theory.

The first question is: Festinger considers cognitive dissonance to be a fundamental component of our cognitive mechanics. He also tells us that it is not necessarily the only one, but he does not try to establish the exhaustive list of the constituents of our cognitive mechanics.
We will immediately add the generalized nepotism, or social ambition, which we saw in chapter 2, and formulate the hypothesis that these two elements suffice as foundations to build the rest of this book. More precisely, the desire for social advancement constitutes the initial motivation, and cognitive dissonance preserves our capacity for action in order to achieve it.

The second question is: in all of Festinger's book, the experiences reported are built on a sociological model, that is to say in which what is measured is the proportion of individuals choosing each offered option. At the end of the book, Festinger approaches the psychological aspect in a naive way since it only proposes to classify individuals according to their tolerance to cognitive dissonance. What we expected at this level is something much more ambitious, which would have consisted in establishing a new classification of psychological profiles, the scientific basis of which would be its effectiveness in predicting the behavior of an individual subjected to various sociological experiences.

On the other hand, we find in later developments of the theory of cognitive dissonance an interesting proposition: the resolution of cognitive dissonance would serve us to preserve our capacity to act by limiting our level of doubt. This argument perfectly matches the vision proposed in the book The Stupidity Paradox: The Power and Pitfalls of Functional Stupidity at Work that we will discuss later in this chapter.

Once these two limits have been set in Festinger's work, let us list some consequences to better measure their importance.

The starting point of a scientific psychiatry

For the moment, the fact that psychiatry is not yet a science is fairly commonly accepted, and three main approaches fight for the dominant position.
The first would be to look for scientific bases for psychiatry on the side of neuroscience. The problem is that if neuroscience meets the criteria of the scientific approach, their object of study covers only certain basic mechanisms at work in brain functioning, but not the overall complex machinery that produces our psychology. Wanting to explain psychiatry by neuroscience is a bit like trying to explain our social organization by observing only the movements of citizens.
The second consists in giving up adopting scientific bases, and considering that the therapy is reduced to a doctor-patient verbal interaction which follows conventions based on a simply plausible theory, as in Antiquity, for example psychoanalysis.
The third is the cognitive-behavioral approach, which aims to reinsert the individual by training him, and therefore conditioning him to perform actions in line with social expectations. The problem here is that the stated objective, in order to obtain the patient's agreement, is personnal fulfillment, but since we do not know how to define the behaviors that produce this personal fulfillment on a particular individual, it just promotes the behaviors considered to be the most effective in achieving social success, which have been deduced from the observation of individuals who have experienced very great success. Now on the one hand absolutely nothing says that these recipes will produce the same effects when they are applied to other individuals, and on the other hand, personal fulfillment is not limited to social success.

However, as we have just seen it, for more than 50 years, cognitive dissonance has opened the first possibility of a scientific approach in the field of psychiatry, via the establishment of a classification of personalities which would be validated by its effectiveness to predict the behavior of individuals during different sociological experiences. Now remains to do the heavy work of defining these categories or axes, and tools to characterize an individual. An axis could be for example the sensitivity of the individual to social support. In chapter 10 of A Theory of Cognitive Dissonance, Festinger explains that the more complex a situation is, the more an individual will be able to maintain an incorrect belief as long as he receives social support, i.e. other people make the same mistake as him. The example he takes is that of Japanese citizens living in the United States of America, having requested their repatriation to Japan at the end of the Second World War, and remaining persuaded, even on the return ship, that it was Japan that had won the war.
Also, the remark "we don't have anything else [other than psychoanalysis]" by the psychiatrist and psychoanalyst Pierre Sidon during the Les chemins de la connaissance radio programme on December 11, 2018 seems both revealing to us - we're not looking for - and problematic, because it aims to justify maintening psychiatry as the only branch of medicine not subject to the scientific approach, therefore "not effective at all" to use Jean-Christophe Rufin's words (1).

To illustrate the relevance of a classification based on the prediction of behavior in sociology experiments, let's take for example the case of Autism. The American DSM classification ended up more or less giving up defining this category, given the difficulties in defining the criteria. We propose to define Autism as a weaker motivation for the struggle for social status, that is to say less predisposition to the generalized nepotism seen in chapter 2, as well as a different selection of the preferred solutions in case of cognitive dissonance, which obviously remains to be clarified. What is interesting with this approach is that there are no longer normal individuals, and failing individuals, but two modes of treatment of cognitive dissonance which each have their advantages and disadvantages; which explains much better on the one hand the difficulties encountered by autistic people in a society which is not predominantly autistic, and in which the rules were therefore established by non autistic people, but also on the other hand their social interest in order to compensate for the lack of rationality of so-called normal individuals, that the sociological studies on the basis of the theory of cognitive dissonance have brought to light, but which psychiatry ignores.
In other words, the first problem of the current classification is that it is based on a normality whose only objective reality is to correspond to the behavior of the majority, therefore dominant group. And therefore, the second problem is that it thereby creates a bias which leads to characterize all the other categories in terms of unabilities. Witness the replacement of the word Autism, which literally means self-centered, with the expression Pervasive Developmental Disorders. This reveals another obstacle which for the moment prevents psychiatry from becoming a science: psychiatry does not dare to look at - and therefore present to society - the normal individual for what he is, namely very largely irrational, as this is shown by the theory of cognitive dissonance.

Spirituality as a tool for limiting cognitive dissonance

At the level of spiritual practices now, Buddhism non-duality can be reread advantageously in terms of cognitive dissonance. Indeed, it consists in getting rid of the biases identified by cognitive dissonance, that is to say, objectively welcoming information that comes to contradict our past experiences.
Likewise, meditation can be seen as an exercise aimed at lowering the level of general mental tension in order to allow dissonant information not to be immediately blocked by a wall of defensive reactions, therefore a tool to help establish harmonious functioning cognitive dissonance.
However, we will not provide more details about these subjects in this book, because we prefer to methodically lay down in chapter 22 a minimalist philosophy which takes account of the advances in science through the sociology of the twentieth century.

Cognitive dissonance in the world of work

A direct effect of cognitive dissonance in the world of work is described in the book The Stupidity Paradox: The Power and Pitfalls of Functional Stupidity at Work by Mats Alvesson and André Spicer. This book first focuses on effective behavior in very high-tech companies, or in consultancy firms with an elitist image and recruiting only people that intended the most prestigious schools: they mainly encountered bureaucracy and very ordinary operation, completely out of step with the external image of the company. Likewise, promotion officially rests on excellence, but in practice on the game of alliances. Then, the book notes that young people who arrive in these companies at the end of their studies, after a period of surprise and disappointment, generally end up adopting the culture and methods of the company and losing their critical sense. The book concludes that companies find some benefit in terms of capacity for action, at the cost of increased risk linked to the inability to take into account the signs preceeding a disaster.
In terms of promotion based on the game of alliances, it is an illustration that generalized nepotism is the norm in commercial enterprises. The rapid incorporation by young people of the discrepancy between speech and reality perfectly illustrates Festinger's point of permanent work to reduce cognitive dissonance. An interesting element to note at this level is that when interviewing employees and their superior, only the superior needs to lie about the importance of his work, which is generally done by the adoption of beliefs concerning the virtues of enlightened management. Finally, the book's conclusion announces that the benefit obtained in return for the increased risk linked to the lack of rationality observed, is the optimization of the capacity for action. This overlaps with the remarks of the addition in the new edition of A Theory of Cognitive Dissonance.

Let us now focus on the more subtle effects of cognitive dissonance in the world of work. For this, we refer to the article by John W. Meyer and Brian Rowan published in 1977 in the American Jounal of Sociology and entitled Institutionalized Organizations: Formal Structure as Myth and Ceremony. This is a summary article from which we retain several key points.
First of all, the formal organization of organizations (2) corresponds to the application of good practices whose efficiency is taken for granted without verification.
Then, alongside this formal organization displayed in an ostentatious manner, an informal organization is discreetly set up because the first does not allow the actual activity to be carried out correctly, but without putting it into question. Surprisingly enough, the organizations that agree to set up these inefficient formal organizations, based on what the article describes as myths, lose some productivity, but ultimately increase their chances of survival, because they create confidence internally and externally, and thus facilitate their access to funding. Finally, the article reports that the more an organization has incorporated mythical good practices into its formal structure, the more ritualized and inefficient controls and evaluations tend to replace  evaluations: trust has replaced rationality.
Like the book The Stupidity Paradox: The Power and Pitfalls of Functional Stupidity at Work, the article by Meyer and Rowan notes the mediocrity of effective organizations, and the discrepancy between the displayed image and real practices. Like the book, he explains it by the interest that organizations find in it. However, the article leaves several questions unanswered, which we will now try to answer.

The first question is: why and how do myths take hold?
To answer it, we will call on the tripartition notion by Georges Dumézil which we will explain in more detail in the next chapter. For the moment, let us content ourselves with the simplified representation that our societies are divided into three social groups, which correspond to the clergy, the nobility and the third estate in the Old regime. Let us repeat: if our main motivation inherited from our genetic heritage is social ascent, then, at any moment, there are people who seek to bring out a new god, because that allows to constitute a new clergy, therefore that creates socially elevated places. To understand this, it is necessary to refer to the notion of clergy in Antiquity, where to each god corresponded one or more temples and the associated clergy, as opposed to our Judeo-Christian monotheistic heritage. In the economic pantheon, we therefore regularly add new gods, such as quality, corporate citizenship, the manager who makes his teams grow, respect for the environment, e-reputation, integrated management software, data outsourcing, etc. Each time standards and diploma-based training courses will be put in place to supervise the new domain, that is to say, to sanctify the new clergy.

Then, what Meyer and Rowan show, is that organizations, whether commercial businesses or administrations, will not seek to be efficient in the various fields, but simply to comply with the new constraints, and for that, they will largely be content to employ people with the corresponding degrees, or apply the standard method of the field, without worrying about its relevance, which brings up the second question: why do commercial companies not vigorously fight against these myths which lead to the establishment of ineffective formal organizations which will ultimately undermine their results?
As seen in chapter 2, to progress socially, what is most effective is to be a strategist in the game of alliances. Since denouncing the absence of utility, or rationality, of one of the myths of the corporate culture, means declaring war on the corresponding clergy, it is generally a bad strategy; so the people who progress the best on the social ladder avoid it. In other words, they simply put their personal interest before that of the company.
If we now consider the case of a business owner who would be the sole shareholder of his business, we could say that for him, the personal interest corresponds to that of the business, so that he has less reasons to preserve these myths inside his business, since by rejecting them, he could advance his result by limiting the unproductive costs associated with it. However, a business does not operate in isolation, but in close contact with customers and suppliers, so that no longer applying the standard will expose the business manager to the risk of having to explain his choices, therefore to denounce myths, and therefore to put its interlocutors on notice to clarify the situation at their level, which they risk not appreciating at all.

Which brings us to the third question: do people who do not denounce the myths of corporate culture do so out of opportunism and hypocrisy or in good faith? It is at this level that we find the strong link with cognitive dissonance. The more individuals are aware of the mythical nature of inefficient formal organizations but presented as good practices, rational and optimal, the more they find themselves in a situation of dissonance between their knowledge and the behavior necessary to be able to make useful alliances in the professional context, therefore the more their capacity for action is reduced. The optimum for successful social advancement is therefore to believe in good faith in the relevance of inefficient formal organizations.
Another fact confirms that the majority believes in good faith in the relevance of inefficient formal organizations. If the banker did not sincerely believe in it, it would be in his interest to stop funding more widely the companies that implement these inefficient formal organizations as widely as possible, contrary to what Meyer and Rowan observed.

Other cognitive bias

The list of scientifically validated cognitive biases is long. The interest of the cognitive dissonance exposed by Festinger is that it is not a cognitive bias, but the discovery of a fundamental constituent of brain mechanics which explains a large number of these cognitive biases. We will not try here to establish an overview of these biases. To do this, just consult the corresponding Wikipedia article. On the other hand, we are going to present three biases which seem to us more particularly relevant to take into account when building a new system of social organization.

The first bias is the overconfidence of the ignorant, more cleverly called the Dunning-Kruger effect. Basically, when people are asked to assess their ability to solve a problem, the overvaluation of their own competence is all the more important as their effective competence will prove to be weak. Another side of the same effect is our tendency to think that we are more competent than others.
The mechanism of operation of this bias is quite simple to understand: in the vast majority of our analyzes and decisions, we apply heuristics, that is to say tricks to avoid having to deal with the problem in all its complexity. We save our cognitive effort, and if we did not, we would be very slow and very tired at the end of the day. If we are inexperienced in a subject, we tend to use general heuristics, and if we are experts, we tend to use much more detailed and domain specific heuristics. What changes in passing is the reliability of the involved heuristics, and the fact that if we are experts, then when the particular case shows specificities indicating that it does not correspond well to any of our available heuristics, depending on our psychological profile, we may either be alerted and switch to a more in-depth analysis mode, or apply the heuristic anyway because of the cognitive dissonance which tells us that as an expert we cannot but know. On the other hand, the inexperienced person with regard to the subject to be treated applies a broad heuristic which will apply whatever the details of the case to be treated are.
The consequence at the level of the social organization system is obvious: you cannot ask a person to assess his ability to solve a problem. In the current system, this results in the tendency to demand that the person has the appropriate diploma. However, Meyer and Rowan show that this results in purely formal controls, that is to say that one relies entirely on the expert on duty. However, we have just seen that in the event of difficulty, there is nothing to proove that he will have the social courage to say that he does not know. This is all the more true since others are very likely not to encourage it. This is the now well known drama of whistleblowers.
But the effect of overconfidence does not stop with technical experts, on the contrary. At hierarchical levels of management, the effect is even deaper, and all the more that one goes up in the hierarchy, where the cognitive dissonance invites the individual to consider that his place necessarily corresponds to exceptional skills, and that the only justification for his position and his exceptional salary becomes his "ability to make decisions".

The second bias that we have chosen to note, and whose consequences will apply following the first, is the asymmetry of risk taking expressed in the theory of perspectives. This theory says that we seek above all to maximize the probability of a gain, and to minimize the probability of a loss, but not to optimize the gain or the probable loss, i.e. the gain multiplied by its probability or loss multiplied by its probability. Put more simply, faced with a prospect of gain, we are not taking enough risk and are looking to gain something more or less for sure, even if it is not much, and conversely, facing the prospect of a loss, we take too much risk and try to avoid losing at all costs, even if it means losing a lot when we lose. The second side is well known to the general public in the form of the syndrome of the player who seeks at all costs to recover and ends up losing everything.
In the world of work, this will translate on two levels: on the one hand, a certain conservatism at the time of decision-making, which generally leads to the adoption of a mythical solution as described by Meyer and Rowan, seen as the "most likely to work" solution. On the other hand, when we find that a past choice has turned out to be bad, the preferred decision will not be to note the loss and then to start on a healthy basis, but to bias all the following decisions in order to remove or minimize the initial error.

Finally, the third cognitive bias, which we will qualify as symbolic evaluation bias, corresponds to the observation that when choosing a complex product or alternative, we tend to base our decision on a few perceived qualities of the different products or alternatives, while in use, what will determine our satisfaction is much more the absence of defect of the product or the chosen alternative. The initial reason for this methodological choice is quite obvious: anticipating possible faults requires having considered all aspects, whereas choosing more or less arbitrarily a certain number of evaluation criteria makes it possible to reduce the complexity of the decision process at will. Where there is a bias is that if by force of circumstances, we decide to evaluate in a simplistic way, we should then also modify the evaluation method so as to take this over-simplification into account; for example, by privileging the possibility, the facility, and the low cost of retrofitting the product. Or by organizing a field trial before the final decision to make the faults visible. However, we note that in practice, the over-simplification of the decision-making process is just not taken into account. This is particularly clear in the choice of computer softwares, which have become extremely complex products. It is as if what the decision maker was looking for is to give the illusion of a rational decision, and not to make a rational decision. From our point of view, this means that what Meyer and Rowan find in the production process in the company is also true in the decision process.


The process of irrational resolution of cognitive dissonance appears to us above all as the tool of blindness, and therefore of lying in good faith, that is to say a limit imposed on rationality to promote the alliance game, and therefore social ascent. However, this contradicts our intuition, which is the result of an education leading us to suppose that lying only intervenes in bad faith (3).

If for the rest we had to remember only one thing about cognitive dissonance, it is the fact that we cannot trust a person simply because he has the required skills, and even less because he has the corresponding diploma. This will become even more evident in the next chapter, which will illustrate all this by dissecting effective decision-making processes. A credible social system aiming to share progress with everyone must therefore necessarily set up an efficient mechanism to verify each reasoning leading to a decision involving the community.


Please take with care : this is just the translation to english, from the translation to french, not the initial wording of Hannah Arendt.

Jean-Christophe Rufin, during his conference at the Collège de France on January 23, 2018. He does not speak specifically of psychiatry, but of medicine as a branch of the humanities, that is to say medicine not subject to the scientific method.

Organizations means companies as well as administrations.

This education which leads us to associate lies in bad faith is itself the result of a long philosophical tradition, which begins with Plato and continues with the Stoic Marc Aurèle then Descartes. According to this tradition, the only real obstacle to clairvoyance is ignorance. In other words, it presupposes that every man has the possibility of making an informed decision as long as he takes the trouble to study the subject correctly. The only other obstacle envisaged to this natural clairvoyance is passions.
Now, cognitive dissonance is science that refutes this reassuring belief in a much more fundamental way, by revealing other mechanisms leading to lying to oneself. For example, chapters 2 and 3 of Festinger's book discuss the consequences of decision-making, and show a distortion of reality resulting from worry and not from a passion.