Freud Was Making It Up (But He Wasn't Wrong About Everything)

Sigmund Freud was a cocaine enthusiast who built an unfalsifiable theoretical system from a sample size of upper-class Viennese women, declared it a universal science of the mind, and spent the next four decades defending it against all comers with the ferocity of a man who knew, on some level, that the whole thing was held together with narrative charisma rather than evidence. He was also the person who introduced the idea that mental suffering has meaning, that the stories people tell about their pain are data, and that the relationship between patient and clinician is itself a therapeutic instrument. The field that replaced him kept the prestige and threw out everything that mattered.

The cocaine years were not incidental

Freud’s early enthusiasm for cocaine was not a youthful indiscretion. It was a window into his epistemological method. In 1884, he published Uber Coca, a paper praising cocaine as a treatment for depression, morphine addiction, digestive disorders, and fatigue. He recommended it to friends, colleagues, and patients. His friend Ernst von Fleischl-Marxow became addicted to cocaine on Freud’s recommendation and deteriorated badly. Freud never fully reckoned with this.

The cocaine episode reveals something important about Freud’s relationship with evidence. He was a clinician who generalized from personal experience and small samples with a confidence that the data did not support. He tried cocaine, felt better, observed that some patients felt better, and declared it a wonder drug. When the evidence went against him, he moved on to the next theory rather than revising his epistemological method. The pattern would define his career.

What makes the cocaine period revealing is not just the bad science. It is the response to the bad science. When colleagues criticized his cocaine advocacy and pointed to Fleischl-Marxow’s deterioration, Freud did not revise his position based on evidence. He defended, deflected, and eventually moved on to psychoanalysis, carrying the same methodological habits into a new domain. The tool changed. The relationship to evidence did not.

The unfalsifiable framework was a real problem

Karl Popper identified psychoanalysis as a paradigm case of pseudoscience precisely because of its unfalsifiability. A scientific theory makes predictions that could, in principle, be wrong. If the predictions are wrong, the theory is revised or discarded. Psychoanalytic theory makes no predictions that could be wrong. Every outcome confirms the theory. The patient gets better; the analysis worked. The patient gets worse; they are resisting. The patient shows no change; the unconscious conflict is deeper than initially assessed.

This is not a trivial criticism. Unfalsifiability means a theory can never be tested, which means it can never be improved through contact with evidence, which means it will reflect the biases and assumptions of its creator for as long as it persists. Freud’s biases were substantial. His theories about female psychology were projections of Victorian gender ideology dressed up as science. Penis envy, the castration complex, the idea that women’s moral development is inferior to men’s because they lack the anxiety of castration; these were not observations. They were the assumptions of a nineteenth-century Viennese man who treated his cultural prejudices as universal truths.

The psychoanalytic establishment defended these ideas for decades. Not because the evidence supported them. Because the framework made it impossible for evidence to not support them. The defense mechanisms of the theory mirrored the defense mechanisms the theory described. Denial, projection, rationalization. The theory was its own best patient.

Freud’s followers made the problem worse. Jung took the mythology in a more mystical direction, adding archetypes and a collective unconscious that made the framework even less testable. Lacan made it incomprehensible, wrapping psychoanalytic concepts in linguistic theory so dense that criticism could be deflected as failure to understand rather than failure of the theory. Klein developed object relations theory, which introduced productive clinical concepts but remained embedded in the unfalsifiable structure. The tree kept branching. None of the branches grew toward testability.

The mythology machine served a function

Freud built a mythology of the mind. The id, the ego, the superego. The Oedipus complex. The death drive. The unconscious as a seething repository of repressed desires. These were not discoveries. They were narrative constructs; metaphors that organized clinical experience into a compelling story about what it means to be human.

The story was powerful because it gave suffering meaning. Before Freud, mental illness was either a moral failing or a brain disease. You were sinful or you were broken. Freud offered a third option: you were conflicted. Your symptoms were not random malfunctions. They were communications from a part of yourself that you had disowned. The phobia, the compulsion, the paralyzing anxiety; these were messages. They had content. They pointed toward unresolved conflicts that, if brought to consciousness and understood, could be resolved.

This was a genuinely important idea. The notion that symptoms have psychological meaning, that they represent something rather than simply being something, transformed how clinicians engaged with suffering. It made the patient’s subjective experience relevant to treatment. It made listening a clinical skill. It made the therapeutic relationship an instrument of change rather than a delivery mechanism for pills.

The specifics were often garbage. The Oedipus complex as a universal developmental stage is not supported by cross-cultural evidence. The seduction theory, which Freud first proposed and then abandoned, created a legacy of confusion about childhood sexual abuse that damaged clinical practice for generations. Dream interpretation as the “royal road to the unconscious” has no empirical basis beyond the circular logic of the psychoanalytic framework itself. But the meta-level insight; that the mind has depth, that what is conscious is not all there is, that symptoms can be communications rather than malfunctions; this was directionally right in ways that matter.

What the DSM replaced was not nothing

The DSM-III was necessary. Psychoanalytic diagnosis was a mess. Two analysts could evaluate the same patient and produce radically different formulations, each internally consistent, each unfalsifiable, each reflecting the analyst’s theoretical commitments more than the patient’s actual condition. Reliability was essentially zero. The DSM-III fixed this by making diagnosis a mechanical process: check the symptoms, count the boxes, assign the code.

But the fix came with a cost that the field has never acknowledged. The DSM-III didn’t just remove psychoanalytic jargon. It removed the idea that symptoms have meaning. A patient presenting with panic attacks is evaluated against the checklist for Panic Disorder. Do the attacks come on suddenly? Is there a fear of dying or losing control? Are there physical symptoms; heart racing, sweating, shortness of breath? Check enough boxes and the diagnosis is assigned. The question of what the panic attacks mean; what they represent in the context of this person’s life, relationships, history, and conflicts; is not part of the diagnostic process.

Freud would have asked different questions. What were you doing when the first attack occurred? What were you thinking about? Who were you with? What did you lose recently? What are you afraid of that you have not admitted to yourself? These questions are not on the checklist. They are not reimbursable. They do not generate diagnostic codes. They are also, in many cases, clinically essential.

The fifteen-minute medication check that dominates contemporary psychiatric practice is the logical endpoint of the DSM’s design philosophy. If the diagnosis is a checklist and the treatment is a pill, the clinical encounter can be compressed to the minimum time required to review symptoms and adjust dosage. The patient’s story is not relevant to the diagnostic algorithm. The patient’s meaning-making is not billable. The economic incentive structure of insurance-driven mental health care selected for the DSM’s mechanical approach and against Freud’s narrative approach, not because one was more effective but because one was cheaper to administer.

The transference insight was real

Freud’s single most important clinical observation was transference: the phenomenon in which a patient unconsciously redirects feelings, expectations, and relational patterns from early relationships onto the therapist. The patient who becomes angry at the therapist for being late is not just responding to the therapist’s lateness. They are replaying a pattern of expectations about whether important people show up, whether their needs matter, whether authority figures can be trusted.

Transference was not a theoretical construct imposed on clinical data. It was an observation that therapists across orientations and decades have consistently reported. The patient begins treating the therapist as though the therapist is someone else. The emotional intensity exceeds what the current relationship warrants. The patterns repeat. The recognition of these patterns, when it occurs, is therapeutic. The patient sees, in the microcosm of the therapeutic relationship, the relational template they have been running unconsciously.

Modern psychotherapy research has validated the therapeutic relationship as the single strongest predictor of treatment outcome across all modalities. Not the specific technique. Not the theoretical orientation. The relationship. The alliance between therapist and patient predicts outcome more strongly than whether the therapist practices CBT, psychodynamic therapy, EMDR, or any other branded approach. This finding is a vindication of Freud’s core clinical insight, though the field has been reluctant to frame it that way because Freud carries too much baggage.

The stories people tell about their pain are data

Freud understood something that the DSM framework structurally cannot accommodate: the narrative a person constructs about their suffering is itself clinically significant. Two people can meet identical criteria for Major Depressive Disorder and have completely different relationships to their depression. One experiences it as something that happened to them; a chemical event, a brain malfunction. The other experiences it as something meaningful; a response to loss, an expression of grief that was never completed, a collapse following the withdrawal of a relationship that had been holding them together.

These are different clinical situations. They may respond to different interventions. The person who experiences depression as a brain malfunction may benefit most from medication. The person who experiences depression as unprocessed grief may benefit most from a therapeutic relationship that helps them complete the grieving. The DSM treats them identically because the DSM cannot see the difference. The checklist is mechanism-agnostic and meaning-blind.

This blindness has consequences. It produces a system that cannot distinguish between grief and depression, between a reasonable response to an unbearable situation and a brain malfunction, between suffering that needs to be treated and suffering that needs to be heard. It drives the overprescription of medication for conditions that might respond better to relational treatment. It produces a clinical culture in which the question “what does this symptom mean to you” sounds like a throwback rather than a diagnostic instrument. It trains a generation of clinicians to see symptoms as items on a list rather than communications from a suffering person.

Freud was making it up. The Oedipus complex is not universal. Penis envy is projection. The death drive is metaphysics. The specific content of Freudian theory is largely wrong, frequently offensive, and deservedly criticized. But the meta-insight; that mental suffering has meaning, that the patient’s story matters, that the therapeutic relationship is the treatment; these were genuine contributions that the field abandoned when it replaced Freud’s mythology with Kraepelin’s filing cabinet.

The DSM gave psychiatry scientific respectability at the cost of clinical depth. It replaced an unfalsifiable theory with an unvalidated taxonomy and called it progress. The categories are reliable. The categories are reimbursable. The categories may not correspond to anything real in the brain or the body. But they are easier to teach, easier to code, and easier to defend in court than the messy, subjective, narrative-dependent clinical process that Freud, for all his failures, understood was where the actual healing happened.

The irony is that the therapeutic modalities with the strongest evidence base; CBT, DBT, EMDR; all depend on the therapeutic relationship to function. The technique manual is useless without a clinician who can build rapport, tolerate distress, and hold space for the patient’s experience. The relationship Freud identified as central to healing remains central to healing. The field just stopped talking about it because relationships are hard to manualize, hard to bill for, and hard to study in a randomized controlled trial.

The field traded a mythology that took suffering seriously for a billing system that doesn’t.