In response to:
The Scar of Sigmund Freud from the October 9, 1980 issue
To the Editors:
Edward Rothstein [NYR, October 9] claims that Freud emerges unscathed from my “Epistemological Liabilities of the Clinical Appraisal of Psychoanalytic Theory” (Psychoanalysis and Contemporary Thought, December 1979). Let me consider the merits of his reasoning.
I wrote (p. 508) that “the purported insight achieved by the patient is not the product of a process of veridical self-discovery, but rather reflects the patient’s conversion to the therapist’s interpretation.” As is patent from my context, I thereby impugned the psychoanalytic reconstruction of the patient’s actual personal history, but not the existence of his self or his personal identity, as Rothstein would have it. Nor do I gainsay the reality of the self as a theoretically inferred entity by asserting with Nisbett and Wilson that introspection does not afford a person privileged epistemic access to the discernment of the causes of his (her) own mental states, as compared to the avenues available to others who observe him (her) externally. Hence I am hardly committed to David Hume’s ontological dissolution of personal identity into a bundle of perceptions, let alone to the literary “deconstructionist” perdition with which Rothstein saddles me.
As William James tells us (Principles of Psychology, I), “The word introspection need hardly be defined—it means, of course, the looking into our minds and reporting what we there discover. Every one agrees that we there discover states of consciousness” (italics in original). And those who assign crucial probative significance to self-observation do so generally, because they maintain that it affords the subject privileged epistemic access to the workings of his (her) mind, while only indirect, inferential avenues to these processes are open to outside observers. Initially, Rothstein does seem to allow that Freud placed just such essential reliance on introspection, since Rothstein speaks of “the introspection which Freud saw as so crucial to his own self-analysis or to his patients’ confirmation of his interpretations.” How then does Rothstein manage to reason that Freud would nonetheless have been quite unruffled by Nisbett and Wilson’s epistemic derogation of privileged access to the causal dynamics of our minds?
He contrives a spurious compatibility by idiolexical legerdemain, first denying the label “introspection” to the patient’s production of free associations and then misconstruing Freud’s technical term “reflection” so as to exempt the deliverances of the analysand’s Freudian reflections from the purview of Nisbett and Wilson’s epistemic disparagement of introspection. But, contrary to Rothstein, Freud’s construal of the scope of introspection hardly excluded the ideas that emerge when the patient associates “freely” by suspending the exercise of conscious censorial selection among his associations, an activity of selection which Freud technically dubbed “reflection” (S.E. 4, pp. 101-102; 12, p. 115; 16, p. 287; 23, p. 174). Nor did he banish this selective activity from the purview of introspection. Indeed, he characterized his advocacy of the psychoanalytic quest overall as a “call to introspection” (S.E. 16, p. 285)! And the analysand’s quest, whose objective Freud epitomized by “Where id was, there ego shall be” (S.E. 22, p. 80), avowedly requires free association no less than the stated kind of reflection. Moreover, since Freud paid high epistemic tribute (S.E. 18, p. 238; 20, p. 41) to the stream of free associations from patients who follow his “fundamental rule of analysis,” he claimed probative reliability for these introspective data (S.E. 23, p. 174). Yet Rothstein categorizes the process in which “we abandon reflection and allow involuntary ideas to emerge” as a “suspension” of introspection! And he insists on confining introspection to “a reflective analysis of the external expressions of mental life.”
By this terminological stratagem, he then speciously denies that Freud ever claimed probative reliability for introspection in the received Jamesian sense of the term employed by Nisbett and Wilson. Concurrently, he blithely ignores that just this Freudian reliance on self-observation as one of the pillars of clinical validation has been devastatingly undermined by empirical findings adduced in my paper: The analyst Judd Marmor has commendably marshaled telling evidence showing that the patient’s “free” associations furnish only spurious confirmation of psychoanalytic interpretations, because they are so highly contaminated by the therapist’s influence after all (Archives of General Psychiatry 22, 1970, p. 161).
It is an utter commonplace that while the patient’s motivations are still being obfuscated by repressed conflicts, his analyst will place little credence in his purported introspective account of the promptings of his own actions. Yet Freudians hold that improved introspection goes hand-in-hand with the patient’s progress toward their therapeutic goal of insight. And it is precisely because analytically conducted self-observation is deemed to afford privileged access to insightful self-knowledge that such refined introspection is judged probatively essential to the clinical validation of the theory. In this vein, the analyst Charles Rycroft explicitly applies the label “self-observation” honorifically to convey “objective self-scrutiny,” when defining “Introspection” in his Critical Dictionary of Psychoanalysis. No wonder that Freud rated the “trustworthiness” of a self-analysis so highly: “In my judgment the situation is in fact more favorable in the case of self-observation than in that of other people” (S.E. 4, p. 105). And furthermore: “In that way one acquires the desired sense of conviction of the reality of the processes described by analysis and of the correctness of its views” (S.E. 15, p. 19). The super-ego is the special agency which Freud credits with the capability for veridical self-observation of the ego (S.E. 17, p. 235; 22, pp. 58, 60).
How can Rothstein deny that Nisbett and Wilson would emphatically discount the confirmation which Freud claimed for his theory of parapraxes on the basis of introspective self-observations made by Storfer, himself, and Lou Andréas Salomé (S.E. 6, pp. 118, 162-163, 168)? For reasons best known to himself, Rothstein seems to think that any such prima facie clash of view is obviated by the role which Freud accorded to the subject’s linguistic rendition of the findings of his (her) conscious introspection. But why should such codification detract from the privileged epistemic access championed by Freud? Evidently Rothstein’s emphasis on the mediation of language is a red herring which serves to sidetrack that Freud did attribute genuine probative value to introspective data. Besides, Rothstein is clutching at straws when trying to document from a 1912 letter to Ernest Jones that Freud would even demote introspection at least as much as Nisbett and Wilson. He cites a cryptic passing remark from that letter without its context (cf. Jones’s Freud, vol. 2, pp. 95-96). In context, however, the remark appears as Freud’s flattery of Jones’s acumen for having discerned a possible dynamically unconscious motive which had eluded Freud himself, a tribute avowedly inspired by Freud’s gratitude for an unspecified major favor from Jones. But on Rothstein’s construal, Freud’s aside would belie well-articulated tenets that Freud laboriously enunciated in diverse publications: For further details see the new enlarged version of my paper in Noûs, vol. 14, September, 1980, pp. 307-385, esp. pp. 354ff.
At the outset, Rothstein altogether misportrayed me as having confined the clinical data which Freudians deem probatively cogent to those furnished by the analysand’s introspections. His ensuing rehearsal of other kinds of clinical data thus carries coals to Newcastle. And this rehearsal as well is unavailing to sustain his thesis that “Nisbett and Ross’s [Wilson’s] experiments, and Grünbaum’s argument, have very little effect then on Freudian theory,” a conclusion hardly shared by a number of recognized analysts who have read my paper.
University of Pittsburgh
Edward Rothstein replies:
To anyone who is plunged into the midst of these issues through Professor Grünbaum’s letter, they must seem arcane, entwined, and impenetrable. The truth is, I fear, far more difficult.
Grünbaum’s arguments and my intentions are so different that we are almost speaking different languages. In my essay, I was offering an interpretation of Freud while raising questions about the nature of biographical interpretation, autobiographical activity and the process of psychoanalysis. One theme of my essay was the difficulty of giving such an order to a life, and the even greater difficulties of confirming such interpretations. I began by quoting Freud: “Biographical truth is not to be had,” and ended by saying: “The most important questions still remain. How is an interpretation verified? When, for example, in looking at the words and acts of our lives do we recognize ‘truth’?… Can psychoanalysis be verified?” Where I end is where Grünbaum begins.
I was also attempting to clarify a limited portion of Grünbaum’s critique by refining the idea of “introspection.” My “idiolexical legerdemain” was an attempt to prevent ideological confusion. My “red herring” was intended to be a meaty trout. I was attempting to deal with implications of Grünbaum’s argument, without addressing the “probative” question or engaging in a critique of Nisbett and Wilson’s paper or Grünbaum’s interpretation of it. Such questions, Grünbaum’s letter makes clear, have to be raised.
Nisbett and Wilson themselves do not mention psychoanalysis. They argue, more generally, that while an individual has privileged access to all sorts of details about himself through introspection—memories, for example, or physical sensations—when it comes to understanding more complex mental processes—like causes of actions or feelings—there may be limited introspective access to such information. In fact, they write, “one has no more certain knowledge of the workings of one’s own mind than would an outsider with intimate knowledge of one’s history and of the stimuli present at the time the cognitive process occurred.”
A report may claim to be due to introspection but it is more probably an account of activity deemed plausible according to accepted theories of behavior. I may say, for example, “The reason why I like this person is that he has a pleasant appearance; and there is no use in your arguing with me because I know myself.” Nisbett and Wilson argue that such a report may not only be incorrect about the reason for liking that person, but incorrect about the presumption that it comes from self-examination. They set up experiments in which they can reply: “Ah, you think you are introspecting here, but you are not. We have controlled the circumstances so we know you like this person not because of his appearance but for other reasons.”
Their critique of introspection is not limited to simple feelings. They appeal to accounts of creativity, quoting, for example, Jacques Hadamard’s discussion of mathematical discovery, to show that such processes are also not open to introspection. The artist and scientist do not know how they arrive at their ideas.
Classical psychoanalytic theory also holds (with some variations) that a patient’s introspective reports are often unreliable about the causes for his actions. But after psychoanalysis is completed, “untutored introspection” is replaced by a kind of “tutored introspection” (which Grünbaum calls “analytically rectified introspection”). The patient learns the truth about himself; he is able to scrutinize his mental processes objectively. After an analysis, a patient, with such rectified introspection, would be able to confirm psychoanalytic interpretations of his behavior. He would be able to say, “Why of course that interpretation is correct. I know through examining myself what I did not know before. I wanted to kill my father and marry my mother.” That confirmation of the interpretation would have a significance in clinical proofs of psychoanalysis, because such introspection is “privileged”—it cannot be argued with; its private claims, certified by the theory, are open to no question. The patient himself is able to see the aetiology of the neurosis.
It is this appeal to privileged introspection for proof of psychoanalytic theory that is, according to Grünbaum, put in doubt by Nisbett and Wilson’s arguments. The analyst cannot prove his interpretation to be correct by appealing to the patient’s corroboration, because, as Grünbaum writes, the “accuracy of the introspective reports of self-observation is so poor as to suggest that if there is any introspective access at all, it is insufficient to yield generally reliable reports.”
When a patient does offer such an introspective report, Grünbaum continues, following Nisbett and Wilson’s argument, it is likely to be a construction resulting from a theory of behavior found plausible. The patient’s report on his Oedipal conflict is simply the result of a conversion to the analyst’s theory. Grünbaum writes: “The verdicts of purported introspective self-observation are reached a priori via the causal schemata supplied by the pertinent intellectual subculture…. Each patient is a docile receptacle—almost a Lockean tabula rasa—for the tenet that his therapeutic gains were wrought by the characteristic treatment factors singled out as being remedial by the therapeutic theory in question.” A patient is convinced of the reasons for his cure. His introspective confirmation of the theory has no probative value.
There are a number of problems with these arguments which I did not address at all in my essay, and which I can only touch on here. First of all, Nisbett and Wilson’s experiments may not unambiguously support their conclusions. In one of their “most remarkable” demonstrations, for example, they attempted to show “that the manipulated warmth or coldness of an individual’s personality had a large effect on ratings of the attractiveness of his appearance…yet many subjects actually insisted that cause and effect ran in the opposite direction.” Subjects were shown a filmed interview with a college teacher speaking English with a European accent. Half the subjects saw him answering questions warmly and agreeably. The other half “saw an autocratic martinet, rigid, intolerant, and distrustful of his students.” Of course, those who saw the “warm” attitude liked the teacher more. But most of them also said that the “warm” teacher had attractive mannerisms, accent, and appearance, while others said the identical traits were irritating. The liking or disliking of the teacher, Nisbett and Wilson write, affected the judgments of “his appearance, his mannerisms, and accent, but subjects denied such an influence” and asserted instead that these factors actually affected their judgment. They add an exclamation mark to their conclusion.
This experiment is adduced as evidence that we don’t know our reasons for affections and such, and that our introspective claims about such complex matters may be formulated without any access to mental processes. What it shows instead is how such experiments can have very little to say about larger issues. I doubt very much if any person could possibly act “warm” or “cold” without some corresponding changes taking place in his appearance, mannerisms, and accent. A rigid autocratic martinet would hardly have the same appearance, mannerisms, or speech as a warm, understanding, and agreeable professor of European history. The subjects’ claims were probably more correct than the experimenters’.
The dubious conclusion of this experiment does not, of course, dismiss the entire series of experiments interpreted by Nisbett and Wilson, but they all have to be carefully scrutinized, particularly if they are to carry such heavy epistemological freight. Grünbaum himself is too easily inclined to accept the conclusions. Their application to complex human interaction may, in fact, be severely limited.
But assume it is not, and assume Grünbaum is correct in his own application of their conclusions. Then, of course, introspective reports would not be decisive in clinical validation of psychoanalysis. Grünbaum’s letter insists that I attacked his critique of the probative value of introspection. In fact I left the question of validation completely open. As my context made clear, I certainly did not “misportray” Grünbaum, by implying that he would believe there is no other path to clinical validation. I don’t even deny that Freud himself may have “claimed probative reliability for introspection in the received Jamesian sense of the term employed by Nisbett and Wilson.” I never discussed the question of “probative reliability.”
In fact, I actually agree with Grünbaum that introspection—both tutored and untutored—may not provide infallible information about ourselves. My essay argued that truth is very difficult to obtain and more difficult to confirm. Moreover, I even agree with Grünbaum (though for different reasons) that the theoretical appeal to tutored introspection as a privileged proof of an analyst’s interpretation is of minimal scientific importance. The theory itself circularly determines when an introspective report is tutored or untutored on the basis of its agreement with the theory, and indeed determines whether a tutored introspective report is accurate or not. All scientific theories, of course, determine to some extent how “probative” evidence is to be selected and interpreted. But interpreting tracks in a cloud chamber is a different procedure from that of an analyst when he interprets every introspective report after a certain moment as true, while every one before that theoretically determined moment is held open to question.
But as I argued, the interpretation of Freudian theory is not disrupted by Grünbaum’s general denigration of the powers of introspection to reveal mental processes. The controversy hinges upon a misunderstanding. The word “introspection” needs more elucidation. Freud himself was far from consistent. In his idiomatic uses of the word he was not always as aware of its problematic nature as he was in his more metapsychological moments.
It is not enough to say, for example, that the word “introspection” means, “of course,” the “looking into our minds and reporting what we there discover,” as William James had it. The nature of the “looking into” is the problem. And, despite James’s assurance, not all of us would agree that “we there discover states of consciousness.” What is a state of consciousness? Do I really find such a state when I look into my mind? How do we, in fact, look into our minds? I am speaking here not of the simple recognitions of sensation, but of the complexities of mental life which Grünbaum and Freud are intent upon. How do we know anything about our selves? When I say, for example, “I am thinking about” some disturbing problem in my life, how do I know what process is taking place, and what am I doing when I am “thinking”?
The great irony of all this is that I think (and I think Freud would have thought) that Nisbett and Wilson are right in the colloquial sense. We don’t have simple introspective access to our mental processes or to mental causation. Freud’s entire understanding of the mind is based upon that premise; the unconscious is never known “in-itself.” For Freud, mental processes are always inferred, never observed. In his essay “The Unconscious” he writes: “Some special hindrance evidently deflects our investigations from ourselves and interferes with our obtaining true knowledge of ourself.” If psychoanalysis is a “call to introspection,” it is not a call in the colloquial sense attacked by Nisbett and Wilson and Grünbaum, but in a much more technical sense, founded upon Freud’s belief in the mechanical nature of the mind and the power of interpretation to alter the mechanism.
Psychoanalytic “introspection” is a process that is founded first on the suspension of ordinary introspection and then on detached observation. Interpretation of dreams, for example, proceeds by “abandoning all these purposive ideas which normally govern our reflection” and then “taking note of whatever involuntary thoughts may occur to us in connection” with a part of the dream. The chain of associations, mechanically produced, is then reflected upon. But the reflection takes place not in some amorphous introspective realm, but with a concrete text of associations. The procedure for interpretation of dreams is a model for all psychoanalytic interpretation. And psychoanalytic “introspection,” in the key sense of the term, is nothing other than this process—one which is only partly dependent upon “looking into oneself” in the colloquial sense. Again in “The Unconscious,” Freud writes: “All the acts and manifestations which I notice in myself and do not know how to link up with the rest of my mental life must be judged as if they belonged to someone else.”
This view is based upon a theory of mind which is far more sophisticated than is usually thought. In infancy, according to Freud, consciousness could receive excitations from perception of the outside world or from excitations of pleasure or unpleasure. But the “bitter experience of life” and the needs of infantile desire ultimately led to the development of another internal system in which mental processes are detached from simple experience of pleasure or unpleasure. These “secondary processes” are processes of thinking, in which ideas are connected without being distracted by intensities of excitation; new relations are established to free the individual from the simple rule of desire and hallucination. In the seventh chapter of The Interpretation of Dreams Freud argues that the “belated” appearance of such secondary processes is responsible for the fact that the “core of our being” remains inaccessible to us. He also argues that even thinking processes only reach consciousness (i.e., become perceptions) when linked to language. Language provides the surface through which we experience ourselves. As he wrote two decades later in The Ego and the Id, “The part played by word-presentations now becomes perfectly clear; by their interposition internal thought-processes are made into perceptions.”
These views would take more space to examine. But in Freud’s system, introspection brings our mental lives to consciousness through language. When we think about something, we either think in language, or we are unaware of our thoughts. Bringing mental activity into the realm of language, making articulate the inarticulate, is part of the activity of introspection.
The expansion of the interior dominion of language is related to the expansion of the domain of the ego; the non-verbal “wish” is “socialized,” so to speak, brought into the realm of public articulation. We may indeed be converted to a theory, as Grünbaum would have it, because our understanding of ourselves is shaped by our language and the theories embedded in it. Psychoanalytic interpretation takes place in that public realm; it occurs in dialogue not dream; it is confirmed in the scrutiny of speech. As the inarticulate is brought into speech (through suspension of conscious interference) it is interpreted—connections are observed, repetitions outlined, causality inferred. The realm of “introspection” is curiously objective. Psychoanalysis does not take place simply through “looking into our minds and reporting what we there discover.”
The psychoanalytic “self” is, then, no more accessible to direct experience than interior mental life is for Grünbaum and Nisbett and Wilson. But the differences are significant. For Grünbaum the “self” would seem to be a “theoretically inferred entity” that is simply inaccessible; its shape is imposed from without and is then “sensed” within—supposedly through introspection. All the complexities of our mental life—our beliefs, our understandings about our consciousness or our actions—have little epistemological value. A “self” would be a construct as based upon the beliefs of the “pertinent intellectual subculture” as Grünbaum believes patient reports are, and Nisbett and Wilson believe our reports about our interpretations of our more common processes are. And the entire procedure of psychoanalysis (the “call to introspection”) would be a sham.
The psychoanalytic “self,” however, grants a privileged epistemological status to “introspection.” A self is still a theoretically inferred structure, but it speaks. The rhetorical structures of the mind give voice to an idiomatic language, shaped from within, rather than from without. The individual’s language gives a translated vision of the unarticulated realm lying at the base of experience.
The question of whether a given “interpretation” of that speech has any sort of epistemological superiority to another is, of course, the most difficult one. Is the psychoanalytic interpretation “true” or verifiable? Is it, to a certain extent, a “conversion” in the way it shapes the articulation of the pre-linguistic activities of the mind? The result of tutored introspection may not have the sort of singular claim to truth that Freud and some of his followers make for it. But the origins and nature of interior mental forces may well be revealed, as psychoanalytic theory claims, through interpretation of their expressions in language and behavior. Whether adequate standards for assessing such interpretations can be developed by the scientific and psychoanalytic community at large remains, however, an open question; we cannot infer that no such standards can be derived from reflection on clinical practice.
The nature of psychoanalytic interpretation, then, which I tried to touch or in my essay, requires more examination. How is an account of one’s past interpreted? How indeed is any text interpreted and what status do such interpretations have? Somehow, psychoanalysis claims, such interpretations have concrete effects on illness. A cure is a transformation worked upon the mind through speech, exercising its power like some ancient magical incantation.
But Freud clearly believed that the meanings of language and the forces pressing on the mind from within are intricately linked, as Paul Ricoeur has pointed out. The eruption of laughter after a good joke, which so intrigued Freud, may be similar to the catharsis of psychoanalytic cure. I still end where Grünbaum begins. But one way to approach the question of the confirmation of interpretations is through Freud’s philosophy of language. When does language speak truth? Wittgenstein and Freud had more in common than their presence in fin-de-siècle Vienna.