Freud’s psychoanalysis,contemporary cognitive/social psychology,and the case against introspection

Contemporary cognitive/social psychology

What psychological literature calls “contemporary cognitive/social psychology” cannot be considered a paradigm in the Kuhnian sense, that is, a defined and fixed pattern of the theoretical commitments endorsed and shared by the scientific community (i.e., the theoretical claims and the theoretical terms that the scientific community adopts, as well as the experimental procedures and standards; Kosso, 1992, p. 132; Kuhn, 1962/2012, p. xiii; Reiss & Sprenger, 2014). This is because, in spite of similarities between the proposals of various authors, contemporary cognitive/social psychology does not seem to have been structured in a coherent unified theoretical framework. Nonetheless, at the core of many of today’s cognitivist-oriented approaches, there is the rather vague idea that people can achieve self-knowledge not by introspecting their mental states but mainly by taking the position of an external observer and observing their own behavior. In other words, it is postulated that the processes through which people acquire knowledge about their minds are the same processes through which they acquire knowledge about the minds of others (Schwitzgebel, 2014).

One of the most popular versions of this idea is Bem’s Self-Perception Theory. Bem argues that, in cases of cognitive dissonance, people tend to resolve the conflict between two contrasting attitudes or between a belief and an overt behavior by taking the position of an external observer and not by being introspective about their own mental state. That is, “an individual’s attitude statements may be viewed as inferences from observations of his own overt behaviour and its accompanying stimulus variables. As such, his statements are functionally similar to those that any outside observer could make about him” (Bem, 1967, p. 186; see also Bem, 1972, p. 2). Bem’s position is based on two postulates: first, that people come to “understand” the minds of others (their emotions, attitudes, thoughts, etc.) by making inferences from observing the behavior of others “and/or the circumstances in which this behavior occurs” and second, people who find that their “internal cues are weak, ambiguous, or uninterpretable” (1972, p. 5) are in the same position as an external observer, that is, one who must necessarily refer to external cues in order to infer the mental states of others. In other words, someone with a first-person perspective and an external observer with a third-person perspective are in the same functional position. As Wilson (2002, p. 205) clearly exemplifies, according to the Self-Perception Theory it is as if a professional musician at a wedding infers that she is performing because she will be paid and not because she has a special feeling for the bride or enjoys spending time in churches. Bem’s view appears to be mainly anti-introspectionistic, although it does not rule out completely that people can appeal to their introspections or subjective states. Nonetheless, its insistence that there is no substantial difference between an external observer observing an individual’s behavior and the individual herself observing her own behavior can be interpreted at least as a claim about the unreliability of introspection. For Bem it is better to appeal to the observation of behavior when we must assess ourselves than to trust our introspection.

Bem’s theory was very influential on the cognitivistic-oriented social psychology of the 1970s. This is clear in one of the most quoted articles in psychological literature (Nisbett & Wilson, 1977). Nisbett and Wilson (1977) propose an extensive review of many studies from different areas of psychology and various anecdotal examples mainly on the issue of causal judgment (Greenwood, 1990, p. 41). Their aim is to show that people do not actually rely upon any true introspection when they must infer about their own mind or that of other people. Rather, when people are not able to adequately verbally report on the mental processes intervening between a stimulus and a response (Nisbett & Wilson, 1977, pp. 246–248), they appeal to some a priori or implicit causal theories about how a certain stimulus is a plausible cause of a certain response (1977, pp. 248–249). These theories may have different origins: they could come from a culture having explicit rules for determining the causal relationship between a stimulus and a response or providing implicit theories about such a relationship or from the individual’s empirical observations of co-variation between stimuli of a general type and responses of a general type (1977, p. 248; Nisbett & Bellows, 1977, p. 616). It is worth noting that, because of their appeal to these theories when they must make a verbal report about the relationship between a stimulus and a response, people do not make any true introspection even when such a report is correct and reliable. ² Further, because these theories can be socially or culturally shared by both an individual and other people from the same community, they serve as a basis for both the individual’s inferences about her own mind and other people’s inferences about her behavior (Nisbett & Bellows, 1977, p. 616). Thus, similarly to Bem’s Self-Perception Theory, because both individuals and other people who act as external observers appeal to these implicit or a priori causal theories, “it is unnecessary to assume that subjects’ reports about the influences on their attitudes and behavior are based on information any more privileged than that available to observers” (Nisbett & Bellows, 1977, p. 623). In other words, although Nisbett and Wilson do not deny that people can introspect in some situations (1977, p. 256), they defend a view according to which at best there is no substantial difference between the knowledge coming from introspection and that coming from the observation of overt behavior.

Other skeptical considerations on introspection can be found in Gopnik’s work. She points out that, although a long tradition in philosophy and psychology has suggested that we have a direct access to our internal mental states and to the offspring of our thoughts and behavior, this view cannot be correct. Rather, we can have information about ourselves and others by perceiving and inferring “but in both cases there is a kind of symmetry in the information about ourselves and others” (Gopnik & Meltzoff, 1994, p. 167). Gopnik’s position is based on the interpretation of many experimental and observational studies with infants and children. According to her, these studies converge in reporting that, throughout development, there is a strong parallel between an understanding of our own minds and an understanding of the minds of others. Children show a similar pattern of knowing themselves and others: when they are asked to report the same things about themselves and others, they are able to make attributions about the same mental states to themselves and others at the same time and, when they make errors in these attributions, they make the same errors about both themselves and others (Gopnik, 1993b, p. 274). This is evident in the case of a “false belief task,” in which children are asked to make predictions about how a deceptive object will appear to other children. For example, an experimenter shows children a candy box that turns out to be full of pencils. She asks the children what another child would think when they first saw the box. Three-year-olds appear to be unable to understand that another person’s beliefs can be false because they constantly respond that the other child would think the box is full of pencils (Gopnik & Meltzoff, 1994, p. 175). In other words, “children make errors about their own immediately past states that are similar to the errors they make about the states of others” (p. 177). More or less around the age of three and half, a change seems to occur: children start to develop a representational model of the mind that allows a new understanding of their minds and the minds of others (Gopnik, 1993a, pp. 5–6). This model is composed of our beliefs, desires, pretenses, and so on, that we experience from a first-person perspective. Very briefly, this model is nothing but our commonsensical account of psychological states (Gopnik, 1993b, p. 269). We are able to understand our mind and that of others directly, without any inference. However, Gopnik argues that this is an illusion (1993b, pp. 271–273): the acquisition of the representational model of the mind does not provide us with a direct way to understand our mental state and that of others because our verbal reports about the causal determinants of our behavior are generally incomplete, misleading, or wrong (Gopnik, 1993a, p. 9). This means that our understanding of our minds and those of others cannot be direct but only inferential, that is, based upon our observation of overt behavior (Gopnik, 1993b, p. 277). Although Gopnik explicitly argues that “there may be cases in which psychological states do lead directly to psychological experiences” (1993a, p. 12), she argues that a direct introspective access is very limited at best and that we are able to understand our minds and other people’s minds only through the representational model of the mind.

Similar positions can be found in the work on the so-called “introspection illusion” by Pronin and colleagues. This work can be considered a sort of revival of the cognitivistic-oriented social psychology of the 1970s. In fact, in a well-documented review that looks like a sort of updated version of Nisbett and Wilson’s (1977), Pronin (2009) reports a large body of empirical evidence from various areas of psychology in order to show that people cannot have an introspective access to their minds and, when they try to appeal to their introspections, they tend to make mistakes and attribution errors (p. 3). This is due to the four components of the illusion (pp. 4-6): (a) introspective weighting—when people have to assess themselves and their minds, they generally tend to be over-confident about their introspections; (b) self/other asymmetry—when people have to assess others, they generally do not rely upon introspection but upon the observation of overt behavior; (c) behavioral disregard—people generally tend to disregard overt behavior when they have to make assessments about themselves and to consider it when they have to assess others; and (d) differential evaluation—people generally tend to overestimate their own introspection and to underestimate that of others.

The effects of the introspection illusion occur in the context of the attribution of mental states to ourselves and to other people. When we assess ourselves, we tend to focus on (and thus to overvalue) our thoughts, feelings, etc., rather than our overt behavior. On the contrary, when we assess others, we tend to focus on their overt behavior rather than their thoughts, feelings, and so on (Pronin & Kugler, 2007, p. 566). Thus, it can be argued that, because of its limitations, our introspective capacity must be considered as a hindrance in viewing the world outside ourselves (and, of course, the people who live in it) objectively. In fact, this capacity leads us to “have the impression that we see issues and events objectively, as they are in reality” (Pronin, Lin, & Ross, 2002, p. 369). According to Pronin, this can be a source of problems such as “conflict, discrimination, and barriers to self-knowledge and social intimacy” (2009, p. 2.; see also Pronin & Kugler, 2007, pp. 576–577; Pronin et al., 2002, pp. 379–380 on this point). For this reason, Pronin proposes an education on the limits of introspection in order to overtake the biases this capacity can lead to and thus to focus much more on the observation of overt behavior than on introspection in the assessment of our mental states (Pronin & Kugler, 2007, pp. 574–575). Pronin and the authors dealing with the introspection illusion explicitly deny that introspection can be a privileged source of knowledge and self-knowledge. Instead, they stress the dangers and risks its appeal can lead to. They see introspective access as highly biased and unreliable. Thus, the observation of overt behavior is conceived as a more reliable source for dealing with our minds and those of other people.

Each of these four approaches deals with introspection as a folk capacity we use in our daily interactions. None of these approaches denies that people are able to introspect on some contents of their minds and thus that an introspective capacity exists in human beings: ³ all four approaches assume that this capacity is at best very limited and highly error-prone and cannot account for the extent of our minds. Nonetheless, it is not clear whether this position is based on theoretical and philosophical arguments or on empirical evidence. In fact, Bem appears to base it mainly on theoretical grounds, whereas Gopnik, Nisbett and Wilson, and Pronin base it on empirical grounds. In both cases, the result appears quite controversial and difficult to interpret. For example, in Bem, it is unclear whether a cognitivistic framework is compatible with the principles of Skinner’s radical behaviorism. Gopnik, Nisbett and Wilson, and Pronin question whether it is plausible to apply the same criteria for assessing objective/intersubjective evidence to subjective evidence. Further, it is important to stress that the anti-introspectionistic position of each of these four approaches appears to be connected to the fact that they postulate that most of our mental work takes place at an unconscious level, which is in principle inaccessible to introspection and can be inferred only from its behavioral effects. This is evident when Bem writes of “internal programs or rules” (1972, p. 28), Nisbett and Wilson of the distinction between “content and process” (1977, p. 255) or “adaptive unconscious” (Wilson, 2002, pp. 17–41), Gopnik of “psychological states” (1993a, p. 1), and Pronin of “non-conscious processes” (2009, p. 28, 52). This is a common claim of all these cognitivistic-oriented approaches (Lo Dico, 2016, pp. 51–55, 65–67) that permits the exclusion of subjectivity from the realm of scientific psychology and thus the depiction of the mental only in third-person terms according to the standards of natural sciences.

In summary, it can be argued that, according to all these approaches, most human judgments, emotions, thoughts, and behaviors are caused by the unconscious mind, and, because people do not have a reliable introspective access to it, they must infer, hypothesize, or guess in order to grasp how it works—and this leads to mistakes and attribution errors (Gopnik, 1993a, p. 12; Nisbett & Wilson, 1977, pp. 251–252; Pronin & Kugler, 2007, pp. 571, 575–577; Wilson, 2002, p. 106). All these approaches postulate a distinction between two mental systems: the former is open to introspection and describable in terms of the first-person perspective having no (or at best very limited) causal role in the determination of our mental states and behaviors; the latter is inaccessible to introspection and only inferable from the observation of overt behavior, describable in the terms of the third-person perspective, and having the only (or at best the main) causal role in the determination of our mental states and behavior.

Freud’s psychoanalysis

If someone were asked to briefly summarize the aims of Freud’s psychoanalysis as a therapeutic method, a possible answer could be:

Its intention, indeed, is to strengthen the Ego, to make it more independent of the Super-Ego, to widen its field of perception and enlarge its organization, so that it can appropriate fresh portions of the Id. Where Id was, there Ego shall be. (Freud, 1933, p. 80)

Supporters and critics of Freudian psychoanalysis generally agree that the conscious mind (no matter if conceived as “Cs./Pcs.” or “Ego”) can access the unconscious mind (no matter if conceived as “Ucs.” or “Es”) through introspection. For example, one of the pillars of Gruenbaum’s critique of Freudian psychoanalysis is that Freud was convinced that in a successful analysis “the judgments of introspective self-observation are deemed a veritable cornucopia of insightful self-knowledge, precisely because these judgments are now, ⁴ even though not entirely, undistorted” (Gruenbaum, 1980, p. 355). Or, in his re-interpretation of Freudian theory, Bettleheim (1983) argues that “Freud based his work mainly on introspection – his own and that of his patients. Introspection is what psychoanalysis is about” (p. 19). Or, in the eighth chapter of his history of psychology (dedicated to psychoanalysis), Leahey (1992, pp. 207–244) notes that, more or less during the same period as Freud, many researchers introduced the experimental method in order to avoid subjectivity, to replace armchair introspection with experimental rigor, and thus to found psychology on an objective/intersubjective base. On the contrary, Freud appeared to have replaced armchair introspection with couch introspection in which “the introspective observer is … a sick individual wishing to be cured of neurosis, rather than a trained observer committed to the advancement of science” (Leahey, 1992, p. 210). Finally, in a recent paper, Almond (2015) points out that “Freud’s methods [were] introspection and speculation from clinical data” (p. 44).

If the ability to be introspective involves having conscious and direct access to our mind, then, according to Freud’s 1933 statement, ⁵ it seems that the aim of psychoanalysis is to foster such an access and help patients to “take consciousness” of their unconscious thoughts and processes. It follows that the psychoanalytic method must deal with subjectivity and the first-person perspective (of both the patient and the analyst, of course). However, in spite of the apparent agreement among commentators who consider Freud’s method as a form of introspection, the question here is whether or not this way of knowing the mental is coherent with Freud’s theoretical (namely, metapsychological) framework. This is a tricky question because Freud does not directly deal with this topic and does not appear consistent on it throughout his work. Thus, the answer this paper will present will be tentative. It will start with three noncontroversial facts about Freud’s work. The first is that it is certain that Freud was well aware of the issue of introspection: he knew the experimental work of Wundt and the philosophical work of Brentano. ⁶ In particular, it is worth noting that during his training in medicine at the University of Vienna, Freud attended Brentano’s lectures in philosophy when Brentano was working on his book, Psychology from an Empirical Standpoint (1874/1995; Brennan, 1994, p. 229; Cofer, 1985, p. 159; Glymour, 1991, p. 47). In this work, Brentano proposes taking an introspectionistic approach when describing consciousness from a first-person perspective (Huemer, 2014; Mulligan, 2004, pp. 69–74) because, unlike physical phenomena, mental phenomena must be conceived as always conscious, unitary, and intentional (Brentano, 1874/1995, pp. 59–77; Mulligan, 2004, p. 70). According to Brentano, this approach was the main source of knowledge for psychologists. His idea of the mind is compatible with all three conditions for introspection proposed by Schwitzgebel (2014) and described above. In fact, the mind is presumed to deal only with mental phenomena and not physical ones (mentality condition), belong only to a single person and no one else (first-person condition), and be accessible only within a limited temporal window (temporal condition), that is, only in the moment in which a person can have or act with a certain intention and in which a mental phenomenon occurs in her mind.

The second fact is that Freud does not often deal with terms such as introspection, inner observation, self-knowledge, etc., and, when he does, he appears quite skeptical (and sometimes ambiguous) in doing so. For example, in “The Future of an Illusion,” he explicitly states:

But scientific work is the only road which can lead us to a knowledge of reality outside ourselves. It is once again merely an illusion to expect anything from intuition and introspection; they can give us nothing but particulars about our own mental life, which are hard to interpret. (Freud, 1927, pp. 31–32)

This suggests that for Freud, introspective data cannot provide reliable empirical evidence to confirm his theories. So, which sort of evidence must be used for testing Freud’s hypotheses? This question is connected with the third fact about Freud’s work: he often refers to detailed descriptions of overt behavior and the circumstances in which it occurs. Perhaps various clinical cases and some reports from “The Psychopathology of Everyday Life” can be considered as good examples of this:

For many years a reflex hammer and a tuning fork have been lying side by side on my writing table. One day I left in a hurry at the end of my consulting hour as I wanted to catch a particular suburban train; and in broad daylight I put the tuning fork in my coat pocket instead of the hammer. The weight of the object pulling down my pocket drew my attention to my mistake. (Freud, 1901, p. 165)

The fact that Freud does not make many references to introspection as a method of inquiry and pays a lot of attention to overt behavior indicates that the most reliable source of evidence for psychoanalysis must be overt behavior. As Hartmann points out:

The data gathered in the psychoanalytic situation with the help of the psychoanalytic method are primarily behavioral data; and the aim is clearly the exploration of human behavior. The data are mostly the patient’s verbal behavior, but include other kinds of action. They include his silences, his postures … and his movements in general, more specifically his expressive movements. (1959, p. 21)

However, one can object that at last what occurs in the analytic setting, that is, the complex dynamics between the patient and the analyst, is based upon their introspections expressed via verbal behavior. In this light we can understand one of the main techniques of psychoanalysis, free association. Very briefly, this technique consists of inviting the patient to give voice to all the thoughts that enter her mind without any exception. The patient can produce them spontaneously or on the basis of an invitation from the analyst (Laplanche & Pontalis, 1967/1973, p. 169). According to some authors, Freud interprets this method as an introspective modality through which the patient can observe herself and verbally communicate her thoughts (see, e.g., Spacal, 1990). The main argument in favor of this interpretation relies on the fact that the patient is asked not to reflect on her mental states but to simply observe them. She must avoid every criticism or judgment toward them and allow many ideas to enter her consciousness spontaneously (Freud, 1900, p. 102).

However, this description of free association appears to be in contrast with two of the main features of introspection, the first-person condition and the temporal proximity condition. About the first feature, the point is that the patient is not actively focused on her mental states or processes but rather must passively “emerge” the contents of the unconscious into consciousness (Loewenstein, 1963, p. 452). This suggests that, in order to freely associate, the patient must assume a passive stance, that is, the position of an external observer of the unconscious contents coming into consciousness. In this sense, free association consists in simply “bringing into consciousness the repressed material” (Freud, 1925, p. 40) and in reading “off all the time the surface of … consciousness” (Freud, 1923, p. 238). The method aims to have the patient eliminate conscious control over the flow of her thoughts (Grossman, 1967, p. 19) and voluntarily choose her thoughts (Laplanche & Pontalis, 1967/1973, p. 170). Consciousness is not postulated to provide direct and immediate access to the unconscious (Laplanche & Pontalis, 1967/1973, p. 178). Rather, it is seen as a sort of recipient of the manifestations of the unconscious (or the repressed material). In this sense, in Freud’s view, the influence of the first-person perspective on the patient’s verbal reports should be eliminated or, at least, highly reduced. ⁷ It is worth noting that this is the same for the analyst, who can gather the material coming from the patient’s free association and use it to construct his interpretation (Grossman, 1967, p. 26; Hoffer, 2006, p. 4). In summary, when the patient is asked to freely associate, “he is not to tell us selected things. In fact, he is not to be active at all; he is to do nothing except make an effort to prevent giving expression to impulses which rise within him” (Fenichel, 1953/1987, p. 321). The second feature, the temporal proximity condition, can be more controversial because it is true that “the patient … is required to say out loud everything which goes through his mind” (p. 321). The information coming from the patient can appear truly introspective because the patient is asked to say what immediately comes to her mind, in that specific moment when the analyst suggests that she freely associate. Nonetheless, it is important to note that Freud stresses the fact that free association deals with so-called repressed material, with what the patient has forgotten because of repression. In fact, those ideas that appear in free association “are of particular value in discovering the forgotten material” (Freud, 1923, p. 238). In other words, Freud considers free association to be a method not only to deal with the patient’s current and ongoing mental states but also with “his memories, thoughts and feelings, wishes and fears, the people in his past and present environment, etc.” (Loewenstein, 1963, p. 455; see also Freud, 1901, p. 249). As we can see, the range of mental states that are understood through free association is much wider than what is understood through introspection within a restricted temporal window. It can even be argued that what Freud actually aims at with free association is mainly (some traces of) the contents of the repressed unconscious and (some fragments of) the patient’s past memories:

In carrying out the technique of psychoanalysis, we continually require the patient to produce such derivatives of the repressed as, in consequence of either their remoteness or of their distortion, can pass the censorship of the conscious. Indeed, the associations which we require him to give without being influenced by any conscious purposive idea and without any criticism … are nothing else than remote and distorted derivatives of this kind. (Freud, 1915a, pp. 149–150)

In summary, it can be said that free association does not rely upon introspection and appears to be devised in order to avoid the possible distortions we can make when we verbally report the contents of the unconscious mind. This idea of the distortions of consciousness in verbal reports is a constant in Freud’s work, clearly in his distinction between manifest and latent content of dreams (Freud, 1915–1916, pp. 83–233) and his analysis of parapraxes (pp. 15–82). In these cases, what manifests in our consciousness (what is immediately accessible to introspection) is not the true meaning of a certain dream or the real causes of a certain act. Freud seems to be skeptical about introspection as a folk device for making attributions about our mental states and behaviors and those of other people. Perhaps this skepticism comes from his having understood consciousness as something that has a peripheral place in his model of mind. In fact, consciousness is “a sense-organ for the apprehension of psychical qualities” (Freud, 1900, p. 236; see also Laplanche & Pontalis, 1967/1973, pp. 84–88) that appears to be too narrow to support and allow such a demanding capacity as introspection. Consciousness and introspection appear able only to grasp some fragments of the unconscious that can be understood through the work of interpretation:

A small piece of the large and complicated psychical structure of unconscious dream-thoughts has made its way into the manifest dream as well … It is the business of the work of interpretation to complete these fragments … into a whole. (Freud, 1915–1916, p. 120)

This seems to be the reason why, for Freud, we can only deal with the unconscious mind indirectly: on the basis of the material coming from the analytic sessions (but also from the observation of our everyday actions), it is possible to make inferences regarding its functioning. In this sense, all the interpretations and constructions the analyst suggests to the patient throughout the sessions can be seen as conjectures about the workings of the unconscious:

The analyst finishes a piece of construction and communicates it to the subject of the analysis so that it may work upon him; he then constructs a further piece out of the fresh material pouring in upon him, deals with it in the same way and proceeds in this alternating fashion until the end. (Freud, 1937, p. 260)

It is worth noting that Freud stresses the provisional and hypothetical character of the constructions: “We do not pretend that an individual construction is anything more than a conjecture which awaits examination, confirmation or rejection” (1937, p. 265).

Just as contemporary cognitive/social psychology does, Freud’s psychoanalysis postulates a neat distinction between two mental systems: the former is very limited but open to introspection, describable in terms of the first-person perspective, and having a restricted role in the determination of our mental states and behaviors; the latter is inaccessible to introspection but can be known through inferences from the observation of behavior or from the verbal data of the analytic sessions, describable in terms of the third-person or objective/intersubjective perspective, and having the main (perhaps the only) causal role in the determination of our mental states.

Similarities between contemporary cognitive/social psychology and Freud’s psychoanalysis

Because contemporary cognitive/social psychology and Freud’s psychoanalysis have different theoretical and methodological backgrounds and aims, it is not surprising that the authors supporting contemporary cognitive/social psychology do not tend to mention Freud and his work. However, two exceptions can be found in the tenth chapter of Nisbett and Ross’s 1980 book (pp. 228–248) and in the first chapter of Wilson’s 2002 book (pp. 1–16). Nisbett and Ross argue that they share with Freud the idea that most of our mind is inaccessible to consciousness and introspection (pp. 242, 245). However, they point out that the kinds of unconscious depicted by the two approaches are different. The one endorsed by Nisbett and Ross (and by contemporary cognitive/social psychology in general) is “neither verbal in nature nor directly accessible to introspection” (1980, p. 245). Rather, the one postulated by Freud is inaccessible because of the defense mechanism of repression and thus a “direct access to unconscious facts [is] a possibility. There is a ‘memory-trace’ to which the therapist’s suggestions can be connected” (Nisbett & Ross, 1980, p. 246). In other words, Nisbett and Ross argue that the main difference between their idea of the unconscious and that of Freud is that, in their perspective, there is “no memory-trace to be primed, excited or retrieved. The unconscious facts … can only be inferred, not discovered” (1980, p. 246) Quite similarly, in his book Wilson proposes a brief comparison between these two unconscious systems: on the one hand, Freud’s unconscious is directly accessible to us because

through careful introspection … we can penetrate the haze that obscures our true feelings and motives. No one claims that such introspection is easy. … But when such insight is accomplished, often with the aid of a therapist, people have direct access to their unconscious desires. (Wilson, 2002, p. 15)

On the other hand, the kind of unconscious Wilson focuses on is completely closed to introspection and can be known only “by looking at our behavior and how others react to us, and coming up with a good narrative” (2002, p. 16). In other words, Wilson argues that we can know our unconscious minds only through inferring and theorizing from the observation of our behavior and the behavior of other people. A similar position is expressed in Gruenbaum’s 1980 paper (pp. 357–367): on the basis of the results reported in Nisbett and Wilson’s 1977 paper, he argues that Freud’s theory of introspection as a method for gathering material for confirming theories is unwarranted because “even after a successful psychoanalysis … the patient has no privileged epistemic access to the actual mechanisms involved in his changes” (p. 367). Gruenbaum argues that, because the experimentally informed review of Nisbett and Wilson (1977) has shown that our unconscious mind cannot be introspectively accessed and that our knowledge of it comes from a priori and culturally shared theories about its functioning, Freud’s theory cannot be validated on the basis of the patient’s and the analyst’s introspective reports about the causal dynamics taking place in the patient’s mind (1980, p. 364). However, the criticisms against psychoanalysis that are presented here appear to be based on a misinterpretation of Freud’s position, that is, the fact that the analytic method is a form of introspection. Despite Freud’s ambiguity on the definition of the unconscious, throughout his work he appears to be skeptical about the possibility of directly accessing the unconscious. In order to deal with this, he proposes a way of knowing the unconscious based on inferences from material coming from the patient. This material should be as free as possible from the influence of the analyst’s and the patient’s consciousness and subjectivity. Thus, in Freud’s perspective, there is little room for introspective reports on this material.

Another similarity between Freud’s psychoanalysis and contemporary cognitive/social psychology is that, because of the denial of introspection as a means of self-knowledge, people appear to be able to know themselves by learning to observe their own behavior. In other words, both approaches invite people to assume the position of an outside observer able to report on their behavior. The main difference between the approaches is that, for psychoanalysis, it is the analyst who assumes this position and helps the patient to do the same, whereas, for contemporary cognitive/social psychology, it is the person herself who must assume the position of an external observer on her own behavior starting from the observation of other people’s behavior (and thus through a continuous comparison between herself and others). However, it is clear that such a difference seems to be due to practical rather than theoretical reasons: Freud’s psychoanalysis aims to be a therapeutic technique, whereas contemporary cognitive/social psychology does not. In fact, the idea of the unconscious mind as the main determinant of our behavior, the rejection of introspection as a method of inquiry, and the appeal to the observation of behavior as a way of gaining knowledge about ourselves all seem to be common assumptions in both Freud’s psychoanalysis and contemporary cognitive/social psychology.

Concluding considerations

The aim of this article is not to provide a historical reconstruction of the connections between Freud’s psychoanalysis and contemporary social/cognitive psychology. Rather, it is to assess the epistemic status of introspection in psychology in light of the comparison between two apparently distinct psychologies such as Freud’s psychoanalysis and contemporary social/cognitive psychology. This aims to resolve with the issue of the scientific status of these psychologies. In order to be recognized as scientific, both these psychologies aim to exclude or at least strongly reduce the influence of conscious subjectivity on empirical evidence and to explain human behavior by using unconscious mentality. This way, supporters of both approaches can use both intersubjectively shared data and intersubjectively shared constructs and thus satisfy one of the most important requirements of scientificity, that of publicity. Perhaps it is for this reason that both Freud and supporters of contemporary cognitive/social psychology tend to look at introspection with a high degree of suspicion: it does not appear to be a scientific method because it is strictly connected with consciousness and subjectivity (Costall, 2006, p. 646). This means that both approaches share a “scientific ideal” according to which science must be based upon public and objective (or intersubjective) methods, concepts, and data. Both can be scientific only on the basis of the standards of natural sciences such as physics, chemistry, and biology. ⁸ In this sense, in both approaches the unconscious is a theoretical notion equivalent to those used in physics and other sciences for describing the unobservable world at the basis of observable phenomena (Kosso, 1992, Chap. 4, pp. 69–86). Such a world can be reached only through theoretical inferences on observable data and described in terms of law-like propositions. The question here is whether or not the appeal to an unconscious mind and to the sort of inferences it implies is warranted. Is it plausible to assume that most of our mind is unconscious? Although it is undeniable that people are often unaware of the workings of many mental states and processes, it is difficult to assess to which degree that is actually the case. In other words, more or less explicitly, both current-day cognitivistic-oriented social psychology (see, e.g., Newell & Shanks, 2014) and psychoanalysis (see, e.g., Atwood & Stolorow, 2014) seem to question the fundamental assumption of the unconscious mind, especially at the methodological level. The big question that recent literature asks is whether the inquiry methods about the mental that the two approaches have used until the present excluded or underestimated consciousness and all its implications such as subjectivity and the first-person perspective in the name of the ideal of natural science (Marcel, 2003, pp. 168, 178). A positive answer to this question would imply a radical reconsideration of both psychoanalysis and contemporary cognitive/social psychology.