Psychoanalytic Facts as Unintended Institutional Facts

Applications Are Always Confirmations

“[A]pplications of analysis are always confirmations of it as well,” Sigmund Freud wrote in the Neue Vorlesungen zur Einführung in die Psychoanalyse, expressing his confidence that psychoanalysis could be successfully applied as a new autonomous research method in anthropology, literary studies and other disciplines in the humanities (Freud 1933, SE XXII, 146), and de facto confirmed by countless applications of his theories to cultural and social phenomena. ¹ In her insightful study Freud and the Institution of Psychoanalytic Knowledge, Sarah Winter points out that Freud’s official explanation of the remarkable expansion of the psychoanalytic method was based on two considerations:

According to Freud, the psychoanalytic work in interpreting dreams has shown, in the light of analogies with ‘linguistic usage, mythology and folklore,’ that ‘symbols seem to be a fragment of extremely ancient inherited mental equipment’ and that ‘the use of a common symbolism extends far beyond the use of a common language’ (Freud, SE VIII, 242). (Winter 1999, 216)

Moreover, Freud himself stressed the scientific and disciplinary importance of dream research to the field of normal psychology:

If dreams turned out to be constructed like symptoms, if their explanation required the same assumptions—the repression of impulses, substitutive formations, compromise formation, the dividing of the conscious and the unconscious into various psychical systems—then psychoanalysis was no longer an auxiliary science in the mind of psychopathology, it was rather the starting point of a new and deeper science of the mind that would be equally indispensable for the understanding of the normal. Its postulates and findings could be carried over to other regions of mental happening; a path lay open to it that led far afield, into spheres of universal interest (Freud, SE XX, 47; also quoted in Winter [1999, 217]).

Many theorists still find Freud’s “analytic method”—often referred to by Freud as a “technique”—useful for interpreting human phenomena. Indeed, almost every anthropological phenomenon has been given numerous and conflicting psychoanalytic interpretations. At the same time, many have sensed that the psychoanalytic method exceeds a threshold, an upper limit beyond which its interpretations merely create meanings. The psychoanalytic constructions no longer offer revealing insights, they are not supported by independent evidence gathered in other fields of inquiry or by other hermeneutic or scientific methods, and the findings cannot be used to support hypotheses or theories outside the psychoanalytic field proper. When a technique can be used to successfully apply its key theoretical concepts to every human phenomenon, when every version of the theory finds “confirmations” of its central claims, and when none of these confirmations can be productively integrated with findings obtained by other methods of inquiry, the foundations of the method and its immense success demand a critical explanation. While all hermeneutic theories agree that whatever falls under the label of “meaning” is a human-created phenomenon, one can still provide a strong argument for the claim that the object of a hermeneutic approach or methodology should not be created by the technique that purports to uncover it. The latter claim cannot simply be dismissed by the trivial fact that every interpretation reflects creative insights of the interpreter. ² The immense success of psychoanalysis qua hermeneutic method requires an independent explanation that goes beyond Freud’s own justifications and that connects various independently discovered facts about the history of psychoanalysis. In Le dossier Freud, a brilliant survey of the turbulent history of psychoanalysis, Mikkel Borch-Jacobsen and Sonu Shamdasani (2006) show that the self-confirmatory and meaning-productive character of the psychoanalytic technique was a central and persistent objection, which had been levelled against psychoanalysis from its advent. Some of Freud’s earliest critics raised this objection, including Alfred Adler, Richard von Krafft-Ebing, Carl Gustav Jung, Albert Moll, and even Freud’s one-time friend and intellectual sparring partner, Wilhelm Fliess. Sooner or later, these critics realized Freud’s failure to provide a convincing reply to the objection that psychoanalysis had not legitimated its purported capacity to understand everything. Indeed, the best Freud could offer as a counterargument was the distinctively psychoanalytic gambit that these objections were uniformly motivated by psychological resistance to the theory (see Borch-Jacobsen and Shamdasani [2006, chap. 2] for an extensive overview). Freud’s reply illustrates how apt the objection was: the Freudian concepts and the hermeneutic technique made it possible to even understand its critics and preemptively discount their criticisms. Freud commits himself here to a particularly seductive combination of the circularity and ad hominem fallacies. Despite this criticism, levelled by psychiatrists, philosophers of science, and historians alike, some intriguing questions still remain unanswered: why Freud and other psychoanalysts were right when they held that their own psychoanalytic theory could be used to interpret any cultural phenomenon. Why is it that nothing can escape the psychoanalyst’s attention? And why does the combination of theory and technique leave, in Frederick Crews’ memorable words, “an academic interpreter without even a mathematical chance of having nothing to say” (Crews 2006, 61)?

A meta-analysis of the cultural success of outcomes of the psychoanalytic hermeneutic method should also be able to explain why the interpretations cannot be fruitfully integrated in nonpsychoanalytic theories, like neurology, evidence-based medicine, sociology, linguistics, and cognitive science. Critics who reject psychoanalysis as a pseudo-hermeneutic (Cioffi 1998, Macmillan 1997), or redescribe its therapeutic effects as an outcomes of insight placebos (Jopling 2008), do notice but insufficiently explain its remarkably closed character: psychoanalytic theories and interpretations are impenetrable by other disciplines (recall Freud’s earlier point about psychoanalytic claims needing “no independent confirmation”) and its interpretations are irrelevant in other disciplines. Sociology, cognitive psychology or cultural studies can perfectly neglect psychoanalytic interpretations of phenomena without loss of evidence for their theories.

A meta-analysis should also explain why the psychoanalytic method creates the impression of “having understood” any phenomenon submitted to psychoanalytic treatment. A particularly strong claim in this respect was made by literary theorist Norman Holland, who held that “the phantasy psychoanalysis discovers at the core of a literary work has a special status in our mental life that moral, medieval, or Marxist ideas do not . . . the crucial point, then, . . . is: the psychoanalytic meaning underlies all the others” (Holland, cited in Olsen [1986, 204). “Now you understand!” seems to be the meta-hermeneutic message of the psychoanalytic interpreter who presents himself, as Jacques Lacan famously put it, as the “subject supposed to know” (le sujet supposé savoir). Since Freud himself held to the idea that “understanding and cure almost coincide” (see Freud [1933, SE XXII, 145]),it would pay to look closely not just at how the analyst was supposed to proceed in therapy, but also what the underlying theoretical structure and presuppositions of the hermeneutic method are (see Roazen [1975, 131]). Although we cannot dispute that psychoanalytic interpretations can be perfectly justified in the light of Freudian theories, and that there are interesting social, cultural, and perhaps even political constraints on what will count as an “acceptable” psychoanalytic interpretation within a community, it does not follow that they adequately explain the interpreted phenomena. What accounts for this intriguing phenomenon and what disguises the illicit move from justification to explanation?

A final set of questions concerns crucial liaisons between the hermeneutic structure of the psychoanalytic method and striking social features of the (Freudian) psychoanalytic edifice. Critical historians of psychoanalysis have tried to link the enormous success of the discipline not only with the particular therapeutic lacuna Freud discerned, but also with the rapid emergence of societies that held Freud as their undisputed master (Roazen 1975; Breger 2000; Borch-Jacobsen and Shamdasani 2006). In 1902 Freud founded the Wednesday Society,which rapidly grew and was renamed in 1908 as the Vienna Psychoanalyic Society. Other psychoanalytic societies quickly followed. But what was striking, as many pointed out, were two distinctive traits that would set apart the psychoanalytic movement from other scientific societies: first, there was—as many of Freud’s contemporaries testified—the almost religious atmosphere, and, secondly, the intolerance of dissent and opposition in public or print. Recalling the Wednesday evening meetings, Max Graf (the father of Freud’s famous child patient, “Little Hans”) wrote that “there was an atmosphere of the foundation of a religion in that room. . . . Freud’s Pupils were his apostles. . . . Good-hearted and considerate though he was in private life, Freud was hard and relentless in the presentation of his ideas” (Max Graf, quoted in Roazen [1975, 193]). Freud insisted on absolute loyalty (Roazen 1975, 308). The hermeneutic framework was embedded in a movement with distinctively cult-like traits, closely controlled by a charismatic leader who was surrounded by disciples, “like the Paladins of Charlemagne, to guard the kingdom and policy of their master” as Breger (2000, 209) colorfully put it. While many have considered this an idiosyncratic (and perhaps deplorable) trait of Freud as Founding Father, we suggest there are deeper explanatory links between the quasireligious character of the movement, Freud’s opposition to criticism, and the very existence of psychoanalytic facts. Rather than being an accidental feature of the theory and the method, the social features of psychoanalysis are (we argue) crucial for the introduction and maintenance of the psychoanalytic facts Freud “discovered.” Every psychoanalytic school has in some way or another developed distinctive sect-like traits, and the fractioned history of psychoanalysis is mainly one of competing societies and contested claims of orthodoxy (Breger 2000; Borch-Jacobsen and Shamdasani 2006).

Unintended Institutional Facts

Our explanatory hypothesis involves an application of John Searle’s theory of institutional facts (Searle 1995, 2010; Lagerspetz 2006). The explanatory strategy rests on an inference to the best (because unifying) explanation of the well-known and independently confirmed phenomena described in section 1: how can the following, prima facie unconnected phenomena be given a unified explanation: the seeming capacity to understand everything, the epistemically closed character of the theory, the difficulty of integrating psychoanalytic findings in other disciplines, and the remarkable social structure of the psychoanalytic movement? All these features suggest that the key hermeneutic claims of Freudian psychoanalysis help create institutional facts, in the precise sense developed by John Searle in The Construction of Social Reality (1995) and, more recently, Making the Social World (2010).

Searle’s theory of institutional facts builds on insightful suggestions provided by Elisabeth Anscombe (1958), and was further explored and modified by philosophers like Tuomela (2002) and Lagerspetz (2006), among others. While Freud himself consistently presented his findings as natural facts and psychoanalysis as a science that discovers those facts, we hold that a Freudian interpretation operates successfully in virtue of the unintended creation and maintenance of institutional facts, in a precise technical sense to be explained below. The continued existence of those institutional facts requires the creation of a complex pattern of shared beliefs among the “believers,” who contribute further “confirmations” of the truth of the claims instigated by Freud. This explains why psychoanalysis functions as a tightly controlled thought system, with its various schools uniformly characterized by strong hierarchical relations, absolute loyalty to the master/founder, and the expulsion of dissidents. The institutional facts have the function of making sense of you (your dreams, actions, mishaps, etc.), they create obligations and permissions, determine whether you are psychologically healthy or not, etcetera.

This assessment of the outcomes of the psychoanalytic method does not rest on a global constructivist approach to science or hermeneutics. Social constructivism provides an implausible account of science and must be rejected on independent grounds (Kukla [2000]; Boghossian [2006]; and see Boudry and Buekens [2011] for a critique of social constructivism in science and psychoanalysis). ³ In the conclusion we will briefly come back to this issue and explain why a psychoanalytic pronouncement, but not, for example, the postulation of a particle in physics, can result in the creation of an institutional fact.

Let us begin with an observation about the structure of a psychoanalytic interpretation. Since psychoanalytic interpretations are almost instantly recognizable, it suffices to introduce simple examples to illustrate their remarkably uniform surface structure: a salient and to some extent remarkable anthropological phenomenon X is assigned a distinctive psychoanalytic property Y: ⁴

Given any able interpreter’s competence to construct narrative justifications that link these concrete psychoanalytic identifications with theoretical claims of Freud (or Lacan, or . . .), the theoretical claims come to justify the redescriptions of phenomenon X as psychoanalytic phenomenon Y. In 2005, cultural theorist Jerry Flieger put the critic of Freud in the X-position and obtained the following unsurprising result:

Indeed, the vitriolic attacks (on psychoanalysis) (= X, our addition) may in fact be considered in terms that Freud himself contributed to cultural discourse—as examples of classic denial (= Y₁) (Freud is not my intellectual father), paranoid generalization (= Y₂) (Freud is to blame for everything), or intellectual hysteria (= Y₃) (Freud reduces everything to sex). (Flieger [2005, 9]; her italics, our additions)

The natural question to ask here is whether a perfectly justified interpretation of X as psychoanalytic phenomenon Y (and supported by a suitable narrative that connects central claims with the phenomena interpreted) explains phenomenon X. We will return to this intriguing question in the final section of this article. For now, we wish to remark on an intriguing feature of these identifications, viz. the fact that they are framed in terms of what Clifford Geertz (1983) calls experience-distant concepts, which contrast with experience-near concepts:

An experience-near concept is, roughly, one that someone—a patient, a subject—might himself naturally and effortlessly use to define what he or his fellows see, feel, think, imagine, and so on, and which he could readily understand when similarly applied by others. An experience-distant concept is one that specialists of one sort or another—an analyst, an experimenter, an ethnographer, even a priest or an ideologist—employ to forward their scientific, philosophical, or practical aims. “Love”is an experience-near concept, “object cathexis” is an experience-distant one. “Social stratification” and for most people in the world even “religion” (and certainly “religious system” are experience-distant; “caste” and “nirvana” are experience-near, at least for Hindus and Buddhists. (Geertz 1983, 57)

David Jopling, who draws our attention to Geertz’s distinction, adds that experience-distant concepts in psychodynamic therapy

include concepts such as unconscious forces, resistance, repression, denial, regression, transference, reaction formation, reversal, sublimation and splitting. These concepts come to play a central role in interpretation and insights. Clients learn to think of themselves in terms of these new concepts, so much so that what they first encounter as an experience-distant concept upon first entering treatment may evolve into an experience-near concept. (Jopling 2008, xxiv)

For our purposes, it is important to draw attention to another feature of Geertz’s distinction: the experience-distant concepts that figure in interpretations do not come with independent criteria for what will count as evidence for the truth of claims or interpretations in which they occur. In other words, outside the theory that defines what counts as evidence in terms of other experience-distant concepts, no independent application criteria exist. The theory introduces the concepts, and the interpretations rest on accepting the theory to the extent that—as Jopling rightly points out—the original experience-distant concept becomes an experience-near concept. Trained psychoanalysts and psychoanalytic interpreters characteristically “see” the Freudian meanings in what they interpret. ⁵

Although the experience-distant character of the central concepts in psychoanalysis is important, this feature by itself does not have much explanatory power. The hidden power of experience-distant concepts comes to light when we combine the observation with the characteristic speech act in which these concepts figure. The characteristic outcome of a psychoanalytic interpretation is a proposition of the form “X is, turns out to be, or should be, identified with psychoanalytic phenomenon Y,” where the Y-position is occupied by an experience-distant psychoanalytic term (or cluster of terms) derived from the psychoanalytic theory that forms the background theory. The identifications presented are based on carefully selected contextual evidence unearthed by the psychoanalytic technique, but which, on closer inspection, turn out to be just more psychoanalytic identifications of the same type. Thus, dream symbols are presented as evidence for the identification of dreams (X) as wish-fulfilments (Y), but the symbols themselves are X-components within propositions of the form “X is a symbol for/stands for Y,” where Y designates a psychoanalytic concept. More importantly, however, the psychoanalytic identifications are presented as contents of descriptive speech acts. These speech acts often take the form of assertions, but can also sometimes be hypotheses or conjectures. This suggests that they are based on empirical inquiry, and are therefore put forward as refutable statements and inductive generalizations. If the psychoanalytic assertion or conjecture is true and justified, we gain psychoanalytic understanding of a designated phenomenon. This brings us to the second element in our account of psychoanalytic facts: their declarative origin.

Propositions—here uncontroversially thought of as the contents of assertions—can be objects of different types of speech acts (Searle 1969). When used in assertions, they are contents of speech acts with a mind-world direction of fit—they are true if, and only if, the world is as the assertor claims it to be (Searle 1969; Humberstone 1992). But propositions can also function as the content of declarative speech acts which state that X from now on counts as Y, as in the declarative statement that this (the referent of the X-term) is (i.e., will from now on counts as) the referent of the Y term, (e.g., money, property, a border, or a valid contract) (Searle 1995, 2010; Smith 2003). ⁶ Used in descriptive speech acts of which assertions are the prime example, the proposition that X is Y describes a fact or phenomenon; however, used in declarative speech acts, the proposition helps create a new institutional fact—the declarative speech act has both a mind-world and world-mind direction of fit. ⁷ For example, uttering “Bedtime now!” (said to children at around 8 p.m.) creates a miniature institutional fact (the time to go to bed for the children) when accepted by them. The real or fictional declarative “This is (i.e.,from now on counts as) one euro” created a new monetary unit in the Eurozone (which is, by the way, also an institutional entity). ⁸ The creation of institutional facts via declaratives requires that the speaker—the person who issues the declarative—have the relevant authority to issue the declarative. Moreover, possessing this authority sometimes requires the occupation of a specific institutional role, in order to successfully declare that X counts as Y (only the US President can appoint a member of the Supreme Court).

Declarative language use with the purpose of intentionally creating institutional facts is part and parcel of our speech act repertoire and to a large extent responsible for the wealth of bona fide institutional facts that surround us: money, property, borders, contracts, world records, tenured professors, and enrolled students. Of course, as Searle (1995, 47) points out, not all institutional facts are explicitly and consciously introduced by declaratives. For our purposes, it is sufficient that some institutional facts are introduced in this manner, and that all institutional facts could be so introduced. Institutional facts should be contrasted with brute or natural facts that exist independently from our attitudes directed at them (Anscombe 1958). While natural facts (objects, properties) are discoverable by us and can become bearers of institutional properties, institutional facts are (explicitly or implicitly) created by us and are of an abstract nature. Institutional properties are not observable, although our observations of many objects, events, and properties are inevitably laden with concepts derived from descriptions under which they are institutional facts (“I see that this coin is worth one euro,” or “I am witnessing Tom and Jerry’s marriage.”). Lagerspetz (1989, 9) points out that “terms which are used to refer to institutions are in some sense like theoretical terms,” and Barry Smith (2003, 17) makes the additional point that the mass of social facts that surround us form “a huge, invisible ontology.”

Institutional facts introduced by (explicit or implicit) declaratives continue to exist or are maintained (Searle 2010, 102) only if very specific attitudes in the introducing agent and his intended audience are present or can be induced (see also Pettit [1993], Searle [1995], Tuomela [2002], and Lagerspetz [2006], whose account we will follow in this section). ⁹ Their Performativity (as a feature of institutional facts) is based on a shared attitude toward the institutional fact, and may contribute to the truth of a sentence describing the fact. The stock example is money: if a group accepts that certain pieces of metal count as money, then, under the appropriate circumstances, these objects are specimens of money for that group. Accepting that certain objects count as money by members of a group is not based on independent evidence that the coins serve as money: rather, the shared attitude creates the institutional facts, which certain sentences then go on to describe correctly. The second feature of institutional facts is Reflexivity: if a sentence describing the institutional fact is true, the relevant attitude must be present in the speaker (Searle 1995, 32-34; Tuomela 2002). The third feature is QualifiedRealism: institutional terms refer to real (abstract) objects and properties; their referents are not fictional entities and they are not intended as fictions. Statements about money, interest rates, property, borders, or—as we shall argue—distinctively psychoanalytic entities like penis envy, a dream’s latent content, or sublimation of the sex drive—can thus be literally true or false. But the objects whose truth and falsity they represent do not exist independently of all representation. A global error theory about institutional facts does not give an adequate account of our beliefs and statements about institutional facts (Searle 1995, 2010, 90ff.).

This brief sketch of the cognitive make-up of a group accepting pieces of paper as money or a river or virtual line as a border does not imply that its members need to explicitly know, nor need to be explicit about the fact that they are maintaining the existence of (explicitly or implicitly introduced), institutional facts. In other words, the threefold structure that guarantees the continued existence of a (system of) institutional fact(s) need not be seen or recognized by the participating group as necessary conditions for what they take to exist. Moreover, they need not be aware of the role of their shared, coordinated beliefs as contributing to the truth of sentences describing those facts (Searle 1995, 45ff). It is, however, possible for them to come to realize that what they take to be money or a border is an institutional fact, and they can also come to realize that an implicit or explicit declarative of the form “X counts as Y” lies at the institutional fact’s origin, and that what they took to be a “brute” or natural fact turns out to be an institutional fact. (In this sense, the conceptual distinction proposed by Searle and others is a kind of Wittgensteinian reminder of what we knew all along, albeit implicitly).

What Searle (1995) calls the function assigned to X in the formula “X counts as Y” cannot be performed solely in virtue of X’s physical properties; it requires “our agreement or acceptance that it (the function, FB & MB) can be performed” (Searle 1995, 42). When the archaeologist speculates that a piece of metal found onsite functioned as money, he is making delicate assumptions about quite specific shared beliefs (and attitudes) in a former society. Of course, the particular artifactual shape of the piece of metal or the place where it was found is excellent evidence for the hypothesis that the pieces were artifactual, but knowledge of its physical properties will never be sufficient to qualify it as a piece of money. Thus, although institutional facts and the embedding institutional frameworks supervene on the physical properties of their realizers, the concepts required to account for them qua institutional facts (i.e., concepts involving shared beliefs) are not reducible to physical concepts. Since institutional facts are not reducible to collections of brute facts, an institutional framework is “closed” in two directions: there are no nonmental properties that are necessary and sufficient for something to count as this or that institutional fact; and, conversely, an entity’s institutional status is never evidence for the obtaining of non-institutional properties of its concrete, material bearers (indeed, it might even be argued that some institutional facts have no material bearers at all—see Smith (2003) for further discussion). This does not exclude the fact that institutional facts have intended and nonintended noninstitutional causal consequences (excessive inflation can cause poverty and famine), and the creation of institutional facts sometimes requires substantial modifications of natural facts in order for the latter to be manageable realizers of institutional properties. ¹⁰

The final step in our explanatory model is that a group can come to realize that they maintained a system of institutional facts, which were mistakenly qualified by them as natural facts. Searle makes the following observations:

The process of the creation of institutional facts may proceed without the participants being conscious that it is happening according to this form. (Searle 1995, 47)

Most of these things (the creation of institutional facts, FB & MB) develop quite unconsciously, and indeed people typically are not even aware of the structure of institutional reality. It often works best when they have false beliefs about it. So there are a lot of people in the United States who still believe that a dollar is only really money because it is backed by all that gold in Fort Knox. This is total fantasy, of course. The gold has nothing to do with it. And people hold other false beliefs. They believe someone is king only because he is divinely inspired, or even believe that marriages have been made by God in heaven, and so on. I am not trying to discourage them because often the institution functions best when people hold false beliefs about it. (Searle [2001, 37-38]; our italics)

Note, first, that Searle holds that false beliefs about the nature of a practice can be useful for the continued existence of the practice, which, as Lagerspetz (2006) correctly points out, falls short of admitting their necessity for conceptual and/or practical reasons (A community can be fully aware of being involved in the creation of specific institutional facts). The situation is even more complex than Searle suggests: an unintentionally created institutional fact can explicitly and consciously be taken to be a natural fact, which may later be revealed (“unmasked,” one is inclined to say here) as a (“mere”) institutional fact. The latter point allows for the possibility that a fragment of a practice thought of as a natural fact can become an intended and explicitlyrecognized institutional fact. When an unintended institutional fact “survives” after being explicitly revealed as institutional fact, its function can often be improved and further useful extensions of the practice can be introduced. On the other hand, when it becomes common knowledge among participants that a practice, explicitly presented as involving natural facts, was based on the creation and maintenance of institutional facts, continuation of the practice will be problematic: it will either die out, be abandoned, or even be explicitly rejected (we shall briefly return to this point in the conclusion). Many institutional facts do not disappear after it has been explicitly realized by the relevant community that they are indeed institutional facts (a point to which we will briefly return in the conclusion).

We will now show that the cluster of phenomena that Searle describes in the last quote—namely, that people typically are not aware of the structure of the institutional reality that they are involved in, and that an institution sometimes functions best when people hold false beliefs about it (although this should not be generalized!)—help to characterize the psychoanalytic practice as a distinctively hermeneutic practice. Freud’s central theoretical identifications were declaratives, presented as empirically testable assertions. The Freudian declaratives served to introduce a complex system of institutional facts which, when accepted by others, came to be known as the central psychoanalytic truths. So, while Freud himself was responsible for their introduction, he alone could not maintain the continued existence of these facts. Their continued existence required that others contribute to the truth of sentences describing the fact by accepting what Freud declared to be the case. In other words, it was necessary for others to come to believe that which Freud thought he merely described but in fact unintentionally created. That is, to put it bluntly, why there had to be a psychoanalytic movement.

Let us pause here to explain the potentially “unintended” character of institutional facts. The first thing to notice is that, as Sally Haslanger (2006, 91) succinctly puts it in a related analysis of the social construction of concepts like race and gender, “our meanings are not always transparent to us.” What was not immediately transparent for Freud, and his intended audience, was the specific nature of the speech act he performed when he said that human phenomenon X is—or counts as—psychoanalytic phenomenon Y. While he explicitly maintained that he was merely describing facts, he was, in fact, creating new entities within an unfolding institutional framework. This particular form of nontransparency regarding the force of institutional statements is made possible because the proposition that X is Y can also be used in speech acts with descriptive force (mind-world direction of fit). We contend that it is the declarative character of the original Freudian psychoanalytic speech acts—the ones that initiated the institutional facts—that was neither transparent to Freud, nor to the psychoanalytic interpreter who applies Freudian theories in his hermeneutic practice. Freud’s immensely important and effective “factive rhetoric”—typified by his insistence on having discovered the meaning of dreams, his description of the technique as unearthing the psychoanalytic meaning of symptoms and dreams, and the analogies with archaeology and puzzle-solving (cf. infra)—played a key role in hiding from view the phenomenon we just described. ¹¹

The institutional character of psychoanalytic facts becomes even more plausible when one takes a closer look at the social acceptance-conditions responsible for their continued existence or maintenance, to use Searle’s concept (Searle 2010, 102). As mentioned before, psychoanalysis’ perspicuous institutional character includes the striking cult-like character of the enterprise, its strongly hierarchical internal organization, and the felt urge to disseminate psychoanalysis as if it were a religion. All these features, well documented by the historians of psychoanalysis, eminently enhanced the carefully controlled distribution of the beliefs necessary to contribute to the existence of what made the sentences describing psychoanalytic facts true (cf. supra: the Performativity feature). The “cult-like” structure of psychoanalysis turns out to be a predictable consequence of the fact that a system of institutional facts cannot be maintained on the basis of empirically founded beliefs, or be supported by independent evidence. Any viable system of institutional facts requires the careful steering and coordination of the supporting beliefs, held by a collective responsible for their continued existence. Since no independent evidence supports these beliefs—as was already noted early on in the history of psychoanalysis and later explicitly argued by critics of psychoanalysis (cf. supra)—various nonepistemic strategies had to be introduced to stabilize the contents of beliefs that contribute to the truth of sentences describing psychoanalytic facts. Faithful followers apply the central declaratives and “find new confirmations”; committees ensure that the central declaratives are not modified; dissident figures likely to undermine the continued existence of the institutional facts—by modifying or rejecting the central declaratives—must be renounced. We leave it to the reader to reread the intriguing and well-documented features of the sociology of the psychoanalytic culture as further extensions and applications of our institutional account of psychoanalytic facts. In the next paragraph we connect the abstract character of institutional facts with the closed character of psychoanalytic theories. Our favorite example is Freud’s brilliant introduction of his best-known institutional fact: that dreams are wish-fulfilments.

The Invisible Ontology of the Wish-Fulfilling Dream

In The Interpretation of Dreams (published in 1899), Freud made it immediately clear that he was breaking with the past by distinguishing the “manifest” dream and its “latent” dream-content. What is often described as his “basic hypothesis” can be roughly summarized as follows: a dream is produced when an unacceptable, repressed, infantile wish becomes active while asleep. When the unacceptable wish threatens to break through into consciousness, a process that safeguards sleep intervenes. The transforming process, the “dream work,” distorts the wish so that it appears in the dream in a disguised form. The analyst is capable of interpreting the dream by following the associations of the dreamer, and by applying his knowledge of dream symbols.

The relevant question that arises in the context of our inference to the best explanation is whether these are empirical hypotheses, or constitute a carefully composed set of declaratives (the central declarative being “dreams are wish fulfilments”), which—when duly accepted and systemically maintained by participants—create Freud’s necessarily invisible (abstract) ontology of dream contents. Interestingly, Freud himself made ambivalent claims about their status. He often presented his central claims as hypotheses, supported by “clinical observations” (or so he described his own interpretations of his dreams and those of his patients). If readers accepted the evidence, they themselves were expected to adopt the same epistemic attitude of believing the interpretation. On the other hand, Freud inadvertently stressed a key feature we have assigned to institutional facts: no evidence is required to create the central institutional fact, here disguised as a “postulate.” Consider the subtle ways in which Freud draws his readers into acceptance of his basic premises in the first lecture of the New Introductory Lectures (published in 1933):

We have—quite arbitrarily, it must be admitted—made the assumption, adopted as a postulate, that even this unintelligible dream must be a fully valid psychical act, with sense and worth, which we can use in analysis like any other communication. Only the outcome of our experiment can show whether we are right. If we succeed in turning the dream into an utterance of value of that kind, we shall evidently have a prospect of learning something new and of receiving communication of a sort which would otherwise be inaccessible to us. (Freud, SE XXII, 8-9)

That dreams are fully valid psychical acts—that is, wish-fulfilments—is “quite arbitrary” and “adopted as a postulate.” But given this postulate, the interpretation of a particular dream as a wish fulfilment cannot represent an attempt to find evidence for a hypothesis (as what happens in a genuine experiment). Rather, the set of postulates presents a hermeneutic challenge to the potential analyst. The challenge amounts to creating an interpretation that is consistent with the central institutional fact that dreams are wish fulfilments. Neither proof nor refutation of the declarative statement is possible, because there is no such thing as independent evidence for the institutional facts on the basis of which interpretation proceeds.

Having declared that dreams are wish-fulfilments, Freud could create more specific institutional facts consistent with the initial postulate, but present them as assertions whose content would constitute evidence for the postulate itself. The core structure of the Traumdeutung thus turns out to be a brilliantly woven network of institutional facts: dreams (X) have (count as having) manifest content and latent content (Y). The unconscious wish is its latent content, therefore X’s manifest content is (counts as being) censured by the unconscious (Y). Objects, situations, or persons occurring in the manifest content (X) are (count as) symbols for often sexually laden acts or objects (Y). The analyst can freely select from the patient’s associations to create a narrative in which the internally consistent system of declaratives appear as a coherent interpretative description of the patient’s vicissitudes. Wittgenstein’s suggestion that dreams might as well be interpreted as expressing unconscious fears, rather than desires, illustrates the arbitrary nature of Freud’s initial declarative (Wittgenstein 1967), and reveals another key feature of the institutional character of Freudian psychoanalysis: if dreams merely count as manifestations of unconscious wishes, alternative frameworks could be developed in which dreams would count as manifestations of unconscious fears, and reveal the initial “postulate” as a merely arbitrary fact. Wittgenstein’s comment suggests that he too must have sensed the institutional character of the Freudian edifice. Freud’s conclusion, near the end of chapter 4 of the Traumdeutung, that “a dream is the (disguised) fulfilment of a (suppressed or repressed) wish,” is not only presented as the key to interpreting all dreams (a description that is akin to his belief that the Oedipus complex is “universal”), but as the central institutional fact, with which interpretations had to cohere. The truth about the unconscious as revealed in dreams was created by Freud and maintained by his followers.

Hiding from view the institutional character of the wish-fulfilling dream was further enhanced by Freudian comparisons and analogies that strongly suggest that latent content is out there, to be found (and not constructed) by the psychoanalyst. Freud’s famous puzzle-metaphor eminently served this purpose:

I have a picture-puzzle, a rebus in front of me. It depicts a house with a boat on its roof, a single letter of the alphabet, the figure of a running man whose head has been conjured away, and so on. Now I might be misled into raising objections and declaring that the picture as a whole and its components parts are nonsensical. A boat has no business to be on the roof of a house, and a headless man cannot run . . . but obviously we can only form a proper judgement of the rebus if we put aside criticisms such as these of the whole composition and its parts and if, instead, we try to replace each separate element in some way or other. The words which are put together in this way are no longer nonsensical but may form a poetic phrase of the greatest beauty and significance. A dream is a picture puzzle of this sort. (Freud, SE IV, 277-78)

Rebus-metaphors and archaeological analogies (“saxa loquuntur!” [“stones speak”]—Freud, SE III, 192) effectively concealed from Freud and his followers how the technique inevitably resulted in the unintended creation of new institutional facts—the interpretations of dreams—rather than the discovery of natural phenomena. The “free associations” generated on Freud’s couch provided a rich source of personal anecdotes and factoids to construct narratives in which these institutional facts could appear as natural facts. Unsurprisingly, there are no dream-symbols in Freud’s theory that are inconsistent with the dream as wish-fulfilment (the core institutional fact). Interestingly enough, the puzzle-metaphor had already figured in the notorious Seduction Theory (Freud, ca. 1896). Regarding the “scenes” that had to be remembered by his patients, Freud wrote:

It is exactly like putting together a child’s picture-puzzle: after many attempts we become absolutely certain in the end which piece belongs in the empty gap; for only that one piece fills out the picture and at the same time allows its regular edges to be fitted into the edges of the other pieces in such a manner as to leave no free space and to entail no overlapping. In the same way, the contents of the infantile scenes turn out to be indispensable supplements to the associative and logical framework of the neurosis, whose insertion makes its course of development for the first time evident, or even, as we might often say, self-evident. (Freud, SE III, 205)

By postulating the infantile scenes as “indispensable supplements to the associative and logical framework of the neurosis,” it seems as though Freud was putting forward a hypothesis based on an inference to the best explanation. However, one might as well claim that Freud declares what the neurotic symptoms of his patients had to count as. ¹² Insofar as Freud’s patients came to believe his interpretations were empirically adequate accounts of neurotic phenomena, they could benefit from the therapy. This phenomenon eminently illustrates Searle’s earlier observation that many people may hold false beliefs about institutional facts or practices in which they engage (“the institution often functions best when people hold false beliefs about it”—cf. supra). In Freud’s case, the false belief was that only empirical claims and testable theories about natural psychic phenomena were put forward. ¹³ This does not exclude that institutional facts may have causal consequences in those whose beliefs help maintain those facts—hence the real (and sometimes beneficial) effect of the Freudian interpretation on his patients.

In a famous letter to his one-time intellectual ally Wilhelm Fliess, dated September 21, 1897, Freud introduced what arguably turned out to be the central institutional fact that later came to characterize the core classical Freudian psychoanalysis: neuroses find their origin in infantile sexual fantasies. Once this declarative began to be accepted by others (first his patients, then a small circle of adherents, and later a whole network of psychoanalytically minded doctors, psychiatrists, and lay analysts), neuroses (X) became manifestations of infantile sexual phantasies (Y). From 1899 onward, dreams became wish-fulfilments. Similarly, in 1902, slips of the tongue obtained the status of symptoms of unconscious drives, desires, or beliefs and consequently, the patient’s free associations became revelatory of his or her unconscious. In Drei Abhandlungen zur Sexualtheorie (1905) sensual acts of babies and toddlers (X) became manifestations of the oral, anal and genital phases (Y), a controversial institutional fact presented as evidence for—and consistent with—another institutional fact: the Freudian libidinal drive. Carl Gustav Jung later introduced the collective unconscious, and Alfred Adler the inferiority complex. In 1924, Otto Rank “discovered” the trauma of birth, an institutional fact which proved to be inconsistent with Freud’s own framework. When Ernest Jones, Freud’s first biographer, pointed out that Freud could not find any evidence at all for the trauma of birth, Freud understandably interpreted Rank’s claim as an empirical hypothesis. Had he himself accepted that dreams and symptoms might simply count as manifestations of the Rankian trauma of birth, he would surely have accepted, as evidence, whatever further institutional facts Rank presented in support of his central institutional fact. Note again the experience-distant character of the concepts that occupy the Y-position in these psychoanalytic pronouncements.

Once Freud’s declaratives were duly accepted by others (e.g., by the Wednesday Society in Vienna, the first followers in Zürich (Karl-Gustav Jung, Eugen Bleuler), the audiences attending his lectures, and his patients), the institutional fact-generating declaratives concerning wish-fulfilling dreams became literally true (recall our Qualified Realism about institutional facts—cf. supra). Stressing the evidence-based nature of dream interpretations engendered the belief that what he had “uncovered” were empirical facts; but acceptance unavoidably helped to maintain the (unintended) institutional fact that equated dreams with wish-fulfilments. Contrary to Freud’s own suggestion that neurology would eventually confirm the existence of postulated mechanisms like repression and sublimation, his ontology had to remain abstract and invisible (we shall never find evidence for the existence of abstract, institutional facts in soggy grey matter.). It thus comes as no surprise that empirical evidence for the Freudian unconscious has never been discovered. (No one should be tempted to confuse the ingredients of the Freudian unconscious with perfectly empirical facts about subpersonal processes in the brain.)

Earlier, we pointed to the closed character of psychoanalysis: psychoanalytic “findings” cannot be used as evidence in other disciplines or fields of study, and the psychoanalytic technique never became part of the repertoire of bona fide scientific or hermeneutic methods in other disciplines. But there is another phenomenon: the distinctive feeling (experienced by those who apply the theory) that a psychoanalytic interpretation of a phenomenon X marks the end of inquiry, and further suggests (at least for those who accept it) that no additional work needs to be done to understand the phenomena occupying the X-position. “Now I understand!” is the feeling when a phenomenon X is placed under a psychoanalytic label Y. In fact, no additional work can be done, for the resulting identification, an institutional fact, will never count as evidence outside the field of psychoanalytic inquiry. Furthermore, no Freudian insight will be epistemically supported by facts outside the realm of psychoanalytic institutional facts. The latter aspect of the closed character of the theory follows from the general observation that extra-institutional facts never confirm the existence of, or provide evidence for, institutional facts. Notice, however, that the closed character of psychoanalysis qua institutional system still allows that the psychoanalytic practice may have extra-psychoanalytic causal consequences (e.g., when used in healing practices). While it is certainly true that the central psychoanalytic concepts became experience-near concepts for those who accepted the theory (cf. Clifford Geertz’s distinction, explicated supra), it does not follow from this phenomenon that the central claims became more empirically verifiable. A better explanation would be to see the shift from experience-distant to experience-near concepts as an unavoidable cognitive side effect of accepting and maintaining Freudian institutional facts. Those who accepted Freud’s pronouncements not only helped in creating and maintaining psychoanalytic facts; they came to think of countless phenomena in terms of the psychoanalytic concepts that functioned as central ingredients of the theory.

Discussion: Illusions of Understanding, and a Comparison with Social Constructivism

Our assessment of psychoanalysis as a system of carefully maintained institutional facts connects some well-known independent objections voiced by, among others, Karl Popper and Adolf Grünbaum. Popper’s critique of psychoanalysis as a pseudoscience inspired philosophers like Paul Ricoeur and Jürgen Habermas to requalify psychoanalysis as a hermeneutic method, and they dismissed Freud’s own positivistic interpretation of his theories as a Selbstmissverständnis. An intriguing argument, in defence of the proposed “hermeneutic” reading of psychoanalysis, has been put forward by—among others—philosophers like Jim Hopkins, Marcia Cavell, Thomas Nagel, and Richard Wollheim. According to them, psychoanalysis’ key concepts and principles should best be seen as nonconservative extensions of our folk psychology—the ineliminable hermeneutic practice par excellence that helps us to better understand ourselves and others (Carruthers and Smith 1996; Hutto and Ratcliffe 2007). ¹⁴ As Winter points out, “(t)he cultural prominence of psychoanalysis has arisen from the persuasiveness of that thought style beyond the confines of professional psychotherapeutic practice, so that psychoanalytic knowledge also appears in the guise of a popular-cultural approach to self-understanding” (Winter 1999, 17). Folk psychology we define here as a complex cluster of meta-representational concepts and constitutive principles, which develop at a very young age. These elements purportedly make it possible for us to understand each other, by ascribing mental states such as beliefs, desires, intentions, emotions, and intentional actions via rationalizing actions and empathy. Our key objection to the hermeneutic countermove is that folk psychology makes sense of distinctively natural phenomena (beliefs, desires, actions, and emotions, cognitive capacities to empathize with others). The Freudian hermeneutic technique, on the other hand, introduces artificial concepts and creates, via declaratives accepted by the psychoanalytic community, the very phenomena it claims to explain. One cannot therefore argue that it is because a psychoanalytic concept can be assigned to a certain phenomenon (and that this designation can be perfectly justified in terms of a psychoanalytic background theory) that psychoanalytic interpretations yield genuine understanding as the application of natural folk psychological concepts do. This is consistent with the observation that applications of psychoanalytic concepts in a therapeutic setting can produce beneficial effects for patients: the Freudian therapy, like many other psychodynamic therapies, has a significant placebogenic effect (Jopling 2008). Our analysis thus suggests that a more fundamental Selbstmissverständnis lies at the heart of psychoanalysis: what the method explicitly presents as interpretations of natural meaningful phenomena turns out to rest on the unintended creation of institutional facts. There is an important sense in which Freud never quite understood what he did. The construction of the institutional context that made this Selbstmissverständnis possible was a genuine tour de force: on the one hand, Freud had to present his claims as hypotheses backed up by epistemic reasons; on the other hand, given the obvious weaknesses of his arguments and evidence, and the quickly acknowledged arbitrary nature of his interpretations by his critics, he had to create a community of adherents who accepted his pronouncements to maintain the existence of the facts he unintentionally created.

Why is it so tempting to think that we understand human or cultural phenomenon X “better” when presented with its psychoanalytic interpretation? A psychoanalytic interpretation links theoretical claims of Freud with redescriptions of concrete human phenomena. The illusion of thereby having understood, that is, having explained a particular phenomenon, derives from a subtle confusion between justifications (a relation between beliefs) and explanations (a relation between facts). Carl Hempel (1965) has pointed out that it does not follow from the fact that if X’s belief that p justifies his belief that q, that (the fact that) p explains why q is the case. My belief that the thermometer is sinking today surely justifies my further belief that it will be cold tomorrow, but the movements on the scale do not explain why it is going to be cold tomorrow (that fact is ultimately explained by meteorology and physics). Justifications provide reasons for believing that p while explanations yield understanding of the fact that p (Lipton 2004). Applied to psychoanalysis: knowledge of Freud’s theories perfectly justifies psychoanalytic redescriptions of phenomena that figure in the X-position, but it does not follow that the “facts” Freud uncovered also explain the X-phenomenon. However justified those redescriptions may be in the light of a psychoanalytic background theory, they lack explanatory force. Confusing one’s justification for the belief that q with an explanation of why q is the case, explains why a well-conducted psychoanalytic interpretation leaves one with the impression that nothing else remains to be said—that one now fully understands the phenomena.

“But does not your analysis amount to a form of social constructivism?” We strongly reject this interpretation for basically Searlean reasons. For starters, a Searlean analysis of the ontology and epistemology of institutional facts is part of an analysis of the ontology of social reality, and differs from the social constructivist’s implausible antirealist credo that all (scientific) facts are socially constructed. Searle assumes a firm and plausible distinction between brute facts, which exist independently of human intentionality, and institutional facts, which come into existence when human beings collectively award what Searle calls status functions (the referents of the terms that occupy the Y-position in declaratives) to parts of reality. A nuanced theory of social facts and social reality starts “with the fact that we are biological beasts” (Searle) and then (and only then) asks “how is it possible in a world consisting entirely of brute facts, of physical particles and fields of force, to have consciousness, intentionality, money, property, marriage, and so on” (Smith 2001, 22). Social constructivism, on the other hand, is a controversial theory on the production of scientific knowledge in general, and entails expansive ontological and epistemological claims that are widely disputed (see Kukla [2000]; Boghossian [2006]; Boudry and Buekens [2011]). We do, of course, acknowledge that many psychoanalysts have turned to social constructivism to redescribe and support both the practice and the status of the supporting theory (see Moore [1999] and Eagle [2003] for an internal critique of such proposals). Moreover, we also reject that the natural psychic phenomena that acquire a specific institutional status within the psychoanalytic practice are themselves social constructions (see also Murphy [2006]), and neither does our account entail that psychic phenomena themselves—the phenomena that figure in the X-position—are social constructions. First, notice that the original psychic phenomena Freud studied—from innocent dreams and slips of the tongue, to depressions and recurrent fears and delusions—are observable outside psychiatric and psychoanalytic contexts. The psychological phenomena that form the objects of study in psychology and psychiatry exist independently of theories about these phenomena. They may be recognized as deviant phenomena by friends and family of sufferers, though perhaps not in any manner that can be precisely described by them, and certainly not fully explicable by them (Murphy 2006). They do not come into existence because a designated group accepts their existence. Our folk psychology, which is so effective when it comes to make others intelligible (Turner 2010), is clearly not suited or equipped to interpret, let alone explain, the nature of and causal mechanisms underlying mental disorders and deviant behaviour. ¹⁵ Neither can social constructivism about psychiatric phenomena explain why disorderly thoughts or behaviour are noticed in every culture (Roth and Kroll 1986) and are, in this sense, perfectly detectable by us (again: “detected” does not amount to “being adequately explained”). ¹⁶ It is precisely because the proposed nonconservative extension of folk psychology proposed by psychoanalysis introduces a complex system of institutional facts that psychoanalysis is not an extension of our natural mental economy. And there is, of course, the fascinating question as to which version of psychoanalysis would offer the best extension of our folk psychology—a question never satisfyingly answered by proponents of the “folk-psychological extension”-thesis.

Concluding Remarks: On Self-Validating Thought Systems

Why is a critique of psychoanalysis as a pseudo-scientific endeavour (Popper), or as a potentially falsifiable and largely falsified theory (Grünbaum 1984, 1993, 2006) not enough? Our starting point was the fascinating but largely neglected hermeneutic power of psychoanalysis, its seeming capacity to “understand” or “make sense” of almost any cultural or anthropological phenomenon, a curious and problematic feature neglected by critics who confine themselves to assessing Freud’s claims against his own empirical and scientific ambitions. Belief systems can be self-validating in many ways. A belief system may encourage its supporters to discount contrary evidence (Boudry and Braeckman 2011), may have self-fulfilling effects (like placebo effects), or may partly constitute its own truths via the Searlean acceptance structure (a possibility explored and defended here). Each of these phenomena can be found in psychoanalysis: immunization strategies (“you repress the Freudian insights”) illustrate the first way, while the placebogenic effects of the therapy—predicted effects occur because of psychological mechanisms that involve belief in the existence of healing powers—illustrate the second self-validating ingredient (Jopling 2008). We showed how the third self-validating mechanism is at work in psychoanalysis: we take a number of facts and phenomena to be indicative of how the Freudian enterprise can be seen as creating a system of unintended institutional facts, an option explicitly left open in Searle’s theory of institutional facts.

Three key observations must be added to the classic “pseudoscience” or “falsified theory” objections, in order to illuminate the plausibility of yet a “third way” in which psychoanalysis has self-validating aspects. First, there is a remarkable social structure that surrounds key psychoanalytic claims, how they were defended, and shielded off against rival psychoanalytic claims. This feature links the unscientific and/or mostly false character of the theory (under the “system of natural facts”description, that is, presented as empirically testable facts about an independent reality) with the distinctive recognition and acceptance conditions proper to Searlean institutional facts. So, while there is an important sense in which a critique of psychoanalysis could stop at the point where the core beliefs are seen as either blatantly false, too vague, or unfalsifiable, an important self-validating dimension of the theory would be missed if we did not look further: the relevance of social acceptance mechanisms needed to stabilize what is true in psychoanalysis.

The second element is the empirical observation—illustrated in this article—that the key Freudian concepts can be applied everywhere, that every anthropological phenomenon X can be given at least one psychoanalytic interpretation Y. This suggests that the X- and Y-components in the relevant identification are more or less arbitrarily related, which is yet another feature of Searlean institutional facts. ¹⁷ The Y-concepts that figure in psychoanalytic identifications prima facie function like theoretical concepts (because they do not refer to observable features), but they are better seen as experience-distant concepts (in the sense defined by Geertz [1983]) with application conditions based upon what is accepted as true within a community, rather than on the basis of empirical data. This process eventually leads to what Borch-Jacobsen has called an empty theory, which denotes a theory the core concepts of which have become arbitrary signifiers that can mean almost anything in anyone’s hands (Borch-Jacobsen 2005).

There is a key reason why an institutional approach to theoretical concepts and entities in (say) high energy physics would be completely off the mark: no physicist would claim that the Higgs particle came into existence because every informed physicist believes that it exists, and the reason is that it clearly designates an entity, the existence of which can be demonstrated via experiments that operate with empirical criteria. Moreover, such concepts and hypotheses are introduced in the context of empirical theories that have been well confirmed on the basis of empirical data. Thirdly, if the Higgs particle were indeed an intended Searlean institutional fact, it would be incoherent for scientists to seek to confirm its existence on the basis of physical evidence, that is, evidence described in physical terms, for an institutional property Y cannot be derived from or be reduced to physical properties (let alone physical properties of its bearer X). ¹⁸

A key advantage of an institutional approach to psychoanalytic facts (not to be confused with a social constructivist one!) is that it explains how and why we—qua outsiders, not qua believers—can have true or false beliefs about a reality unintentionally created by Freud, and why our beliefs about Freudian unintended institutional facts can be true or false. Although Freud alone could have developed a falsifiable theory, a tightly controlled community that accepted his claims was required to turn his pronouncements into a hermeneutic system capable of creating unintended institutional facts. A comparison with certain religious artefacts and their institutional properties may be helpful here: the fact that Mount X is sacred is, on a plausible account of sacredness, an institutional fact (the connection between Mount X and its sacredness is arbitrary, and X would not have been holy had there not been a community which recognizes and/or accepts that X is sacred). But those responsible for the sacredness of X will, when pressed on the issue, deny that its sacred character is an institutional property. In fact, the very suggestion that X’s sacredness is merely institutional might be perceived as a disrespectful or even defamatory remark.

This might explain a potential misunderstanding about our use of “truth” in the context of a theory of institutional facts. The proposition that dreams are wish fulfilments can be considered as the content of an empirical claim about a natural fact and is, under that interpretation, very probably false (although one never knows what a future neuroscience will teach us!). On the other hand, when the proposition that dreams count as wish fulfilments figures in a declarative speech act (issued by Freud and duly accepted by others [Searle 1995, 2010]), the proposition is true, because under those conditions dreams become wish fulfilments, just as pieces of paper become money when certain declaratives are accepted. Put in a nutshell: without joint acceptance, the classic empiricist critique of Freud remains valid. Adding joint acceptance of Freudian declaratives, a new and very powerful self-validating belief system comes into view, one that links the arbitrary connection between human phenomenon X and psychoanalytic concept Y, with the distinctive social culture of psychoanalysis and its distinctive hermeneutic capacity to “understand” everything. Additionally, it can be noted that just as money has a function, psychoanalytic interpretations have functions: they help you “understand” phenomenon X, allow you to see X in a different light, and/or teach you “how to live with X.” They also create permissions and obligations proper to becoming “a psychoanalytic patient.” Psychoanalytic facts, like other institutional facts, have deontic powers attached to them (Searle 2010).

Our approach suggests an empirical hypothesis: if it is revealed that a practice is based on unintended institutional facts—and part of exposing it may consist in debunking the explicit scientific aspirations of the theory—the practice might gradually disappear or collapse. We think this is de facto the case with psychoanalysis, but there are more options. One reaction to the exposure of the “constructive” character of psychoanalytic interpretations was to appeal, by its defenders, to global social constructivism, which holds that every scientific theory constructs its own reality. Hence, the “institutional facts” created by Freud and his followers do not differ from other scientific facts (but see Moore [1999] and Eagle [2003] for criticism of this move from within psychoanalysis). But this is nothing less than an immunization strategy. The account presented in this article explains why so many psychoanalysts were happy to redescribe their project in broadly social constructivist terms, after Popper and Grünbaum exposed the underlying theory as a pseudo-science or its central claims as being largely falsified. Their cognitive strategies are comparable with those of religious believers who, having been convinced that their core beliefs were false or unfounded, saw a welcome opening in “symbolic” readings of its central statements. We are inclined to see this as further empirical evidence for the claim that, when a system of beliefs does not correspond with an independent reality, it need not necessarily collapse: the truth-makers of some of its central claims (abstract institutional facts) can come to be seen as created and maintained by collective acceptance of the relevant declaratives. Our diagnosis leaves open the possibility that psychoanalytic claims can be reevaluated as being about a system of intended institutional facts, just as religious believers can continue to accept that a mountain is sacred, despite the fact that they realize that its holiness is based on some form of collective acceptance. These (and other) issues will be explored in a follow-up on this article that extends the theory to religious institutional facts.