Psychoanalysis

An in-depth examination of psychoanalysis proper — its foundations, practice, evidence base, and limitations

What Is Psychoanalysis?

Psychoanalysis is the original depth psychotherapy and the most intensive treatment in this series: a multi-year undertaking conducted at a frequency of three to five sessions per week, classically with the patient reclining on a couch, the analyst out of view, and a single instruction governing the work—to say whatever comes to mind, without censorship. Everything else in modern psychotherapy descends from it, including most of what was invented in opposition to it; the companion document on psychodynamic psychotherapy covers the broad, once-or-twice-weekly adaptation that constitutes most "analytic" practice today. This document concerns the intensive original, which survives as a living—if diminished—clinical practice, a training system, and a body of theory whose vocabulary (denial, projection, repression, transference, defense mechanisms, the unconscious) long ago escaped the clinic into the culture at large.

What distinguishes psychoanalysis proper from its psychodynamic descendant is not the theory, which they share, but the conditions of treatment and what those conditions are designed to produce:

Frequency and duration. Three to five weekly sessions across, typically, three to seven years (sometimes longer). The rationale is not bureaucratic maximalism: high frequency is held to keep the work continuous rather than episodic—the patient does not spend each session catching the analyst up—and to deepen the process to a level weekly therapy cannot reach.

The couch and the fundamental rule. Reclining, without the analyst's face to monitor, the patient is freed from the regulating effects of ordinary conversation; free association—the "fundamental rule" of saying everything—becomes possible in a way face-to-face dialogue resists. The setup is an instrument for making the patient's inner life, including its evasions, audible.

The transference neurosis. The defining clinical claim: under analytic conditions—frequency, the analyst's relative anonymity and restraint, the open-ended horizon—the patient's core conflicts do not merely color the relationship (as in any therapy) but progressively organize themselves around the analyst, reproducing the original pathology in miniature, live, where it can be analyzed rather than reconstructed. Psychoanalysis is, on its own account, the treatment that deliberately cultivates the illness's full appearance in the room in order to resolve it there.

Character, not symptom, as the target. Psychoanalysis aims at structural change: the reorganization of personality—defensive style, internal object world, capacity for love, work, and play—of which symptom relief is expected to be a consequence. Its proper indication has therefore always been less the discrete disorder than the person: chronic, pervasive, characterological suffering.

The analyzed analyst. Uniquely among treatments, the practitioner must undergo the treatment: a personal training analysis of the same intensity is the center of analytic education, on the argument that an instrument calibrated by self-knowledge is the only protection against the analyst's own unconscious running the treatment.

Historical Development

Freud: from hysteria to the structural model

Psychoanalysis emerged from Freud's 1890s collaboration with Josef Breuer on hysteria—the discovery, via the "talking cure," that symptoms carried meanings and dissolved when their occasions were remembered and their affects expressed. Across the next forty years Freud built and repeatedly rebuilt the system: the abandonment of hypnosis for free association; The Interpretation of Dreams (1900) and the topographic model (unconscious/preconscious/conscious); the theory of infantile sexuality and the Oedipus complex; the papers on technique (1911–1915) codifying transference, resistance, and the analyst's stance; and the 1920s revisions—the death drive, and the structural model of id, ego, and superego, with anxiety recast as the ego's signal and defense as its repertoire. The early movement's politics were as formative as its science: the breaks with Adler (1911) and Jung (1913) established a pattern—doctrinal dispute resolved by schism—that has marked the field's institutional life ever since.

The diaspora and the schools

Nazism scattered the largely Jewish European movement, transplanting it to Britain and the Americas, where it diverged into the great schools (sketched at greater length in the psychodynamic companion): ego psychology, dominant in postwar America, with its systematic account of defense and adaptation; the Kleinian school in London, tracking primitive phantasy, envy, and the paranoid-schizoid and depressive positions, extended by Bion into the theory of containment and thinking; the British Independent/middle group (Winnicott, Fairbairn, Balint) and its developmental humanism; Kohut's self psychology in Chicago; the relational turn from the 1980s onward; and, standing apart, Lacan in France—"the return to Freud" through structural linguistics, the variable-length session, and an idiom of legendary difficulty—whose tradition dominates analytic practice in France and Latin America while remaining marginal in the Anglophone clinical mainstream. Contemporary practice in most institutes is frankly pluralist: working analysts borrow across schools, and "common ground" debates (what, technically, do all analysts actually do?) have occupied the field since the 1980s.

Institutions, hegemony, and decline

The International Psychoanalytical Association (founded 1910) and its constituent institutes built a guild structure—the tripartite training of personal analysis, supervised cases, and seminars—that has produced both rigor and notorious orthodoxy. In American psychiatry, psychoanalysis enjoyed an extraordinary mid-century hegemony: department chairs, board examinations, and the first two DSMs were saturated with it, and until 1988 the American Psychoanalytic Association effectively restricted training to physicians (a lawsuit by psychologists ended the restriction—ironically enforcing Freud's own position in The Question of Lay Analysis). The hegemony collapsed between 1970 and 1990 under converging pressures: the pharmacological revolution and DSM-III's atheoretical turn; the rise of trial-tested brief therapies; managed care; and the field's own evidential complacency. Psychoanalysis today is a small specialty—analytic caseloads have contracted, candidates are fewer and older, and most institute graduates practice mainly psychodynamic therapy—while its intellectual stock has partially recovered in unexpected quarters: attachment research, mentalization, infant studies, and neuropsychoanalysis (Solms and colleagues' program correlating analytic constructs with affective neuroscience, including Solms's revisionist work on dreaming and the drives). One further institutional fact belongs in any honest history: organized psychoanalysis pathologized homosexuality for decades after the rest of psychiatry began retreating from that position—APsA did not fully reverse course until the early 1990s and formally apologized in 2019—an episode that remains instructive about what guild certainty can do when insulated from evidence.

The Theoretical Model

The full theoretical apparatus is shared with—and summarized in—the psychodynamic companion document (unconscious process, conflict and compromise, defense, transference, internalized object relations, development). What follows is the additional load that psychoanalysis proper places on that shared base.

The dynamic unconscious, stated carefully

Modern cognitive science concedes a descriptive unconscious—most mental processing is non-conscious—but the analytic claim is stronger: a dynamic unconscious, in which specific contents are actively kept from awareness because of what they mean, and return in disguise (symptom, dream, slip, repetition). Contemporary analysts defend the claim with converging lines—experimental work on motivated cognition and repressive coping styles, the clinical ubiquity of self-deception, neuroscientific models of affect suppression—while critics regard it as the theory's perennially unproven center. The working clinical version is more modest and more defensible: people systematically do not know important things about their own motives and patterns, the not-knowing is functional, and a method exists for finding out.

Regression, repetition, and the transference neurosis

Analytic conditions are designed to permit therapeutic regression: the loosening of adult conversational defense so that earlier modes of feeling and relating become accessible. Within that loosening, the repetition compulsion—the observed tendency to reproduce formative relational configurations rather than merely remember them—delivers the pathology into the relationship as transference neurosis. Treatment then proceeds by the slow alternation Freud named in a single title: Remembering, Repeating and Working-Through. The analyst's classical stance—neutrality (not taking sides among the patient's internal factions), abstinence (not gratifying transference wishes in place of analyzing them), and relative anonymity—exists to keep the transference legible: the less the analyst imposes, the more clearly the patient's templates display themselves. All three principles have been softened in contemporary practice, where the analyst's participation is acknowledged as ineliminable and the stance is reconceived as discipline rather than blankness.

Dreams, symbol, and psychic reality

Psychoanalysis retains the dream as "royal road"—not via fixed symbol dictionaries (a caricature the field itself rejects) but as the patient's own condensed imagery, opened by association. More broadly the tradition insists on psychic reality: what shapes a life is not only what happened but what it meant and what was made of it in fantasy—a doctrine with a dark history (Freud's contested turn from the seduction theory) and a sound clinical core (two siblings, one event, two fates).

Structural change

The promised outcome: not symptom subtraction but durable reorganization—defense maturation, integration of disowned aspects of self, a transformed internal object world, expanded capacity to love and work. Operationalizing this has been the field's modern preoccupation, yielding measures (reflective functioning; the Shedler-Westen Assessment Procedure; OPD structural axes) that make the claim at least partially testable—Section 7's concern.

How Psychoanalysis Is Used

Indications and analyzability

Classical psychoanalysis was built for the neuroses—chronic anxiety, obsessionality, hysteria, inhibition—and its modern indication has migrated toward character: entrenched personality pathology of moderate severity, chronic relational and work inhibition, recurrent depression with characterological roots, perfectionism and self-defeat that have survived briefer treatments, and—candidly—the formation of clinicians, for whom training analysis remains a principal use of the full method. The traditional gatekeeping concept, analyzability, asked whether a patient could use the method: sufficient ego strength to regress without fragmenting, capacity to observe themselves, tolerance of frustration and of the couch's deprivations, psychological-mindedness, and the practical capacities (time, money, stability) the frame demands. Widened technique (Kernberg's structural approach, Kleinian work with primitive states, analytic therapy for borderline and even psychotic organization at modified frequency and posture) has eroded the old exclusions, but the core matching logic stands—and runs both directions: patients needing acute stabilization, behavioral protocols, or crisis structure need those first, and the companion documents' treatments are first-line for the disorders they own.

The frame and the arc

The analytic frame—fixed times, the fee, the couch, the analyst's reliability and restraint—is doctrine, not furniture: its constancy is what makes deviations (the patient's and the analyst's) interpretable. A full analysis moves through a recognizable arc: an opening phase of settling into the method and the deepening of transference; a long middle of repetition, interpretation, and working-through, in which the same constellations return in rising and falling intensity; and a termination phase, set by mutual agreement when criteria approach (symptom resolution, structural shift, the transference analyzed toward ordinary proportions), in which ending itself—loss, mourning, the revival of every earlier separation—becomes the final and often most productive material. Analyses end; the capacity for self-analysis is the intended residue.

Child analysis and applied forms

Child psychoanalysis (the Anna Freud and Klein lineages, fused in modern training) works at high frequency through play; its descendants include mentalization-based work with children and parent-infant psychotherapy. Beyond the clinic, "applied analysis" persists in organizational consultation (Tavistock group relations), and analytic ideas continue to irrigate the humanities—an influence irrelevant to outcome claims but part of an accurate description of what psychoanalysis is.

Who undertakes it today

Practically: candidates in training; mental-health professionals seeking depth treatment; patients of means with chronic characterological suffering and prior therapy behind them; and—via institute clinics, which offer substantially reduced-fee analysis with advanced candidates under supervision—a wider population than the stereotype allows. Several European systems (Germany most notably) still reimburse substantial analytic treatment courses, which is also why much of the modern outcome evidence is German and Nordic.

Common Practices and Methods

Free association and evenly suspended attention. The patient's task is the fundamental rule; the analyst's reciprocal discipline is Freud's gleichschwebende Aufmerksamkeit—listening without selecting, letting significance emerge rather than hunting it. The pair of disciplines is the method's engine; everything else is auxiliary.

Resistance analysis before content. The canonical sequencing: what blocks the telling is analyzed before what is untold. Lateness, silence, compliance, entertaining the analyst, flight into health—each is taken up as the work itself, not friction around it.

Interpretation, and its modern proportions. The classical hierarchy—clarification, confrontation, interpretation, with transference interpretation as the "pure gold"—still organizes technique, but process research and contemporary writing have rebalanced it: interpretations are offered sparingly, tentatively, and near the patient's awareness; the once-disparaged "supportive" and relational elements (the analyst's constancy, containment in Bion's sense, rupture and repair) are acknowledged as mutative in their own right; and the dose of transference interpretation is calibrated to the patient—research suggesting, counterintuitively, that more is not better, particularly for healthier patients.

Work in regression. Tolerating, containing, and using the deepened states the frame invites—including erotic, hostile, and despairing transferences at full intensity—without retaliation, seduction, or collapse. This is the technical justification for the training analysis: the analyst's own analyzed conflicts are the containment's load-bearing wall.

Dream work and the analysis of enactment as described in Section 3 and the companion document, conducted here with the time and frequency to follow material across days rather than weeks.

Termination technique. Setting a date and analyzing what the date mobilizes—the method's last demonstration that, in this treatment, even ending is material.

Psychoanalysis Among Its Neighbors

Versus psychodynamic psychotherapy. Same theory, different dose and ambition: weekly therapy works focally and in the displacement (patterns discussed mostly as they occur outside); analysis works globally and in the transference neurosis. The honest contemporary question is not which is "deeper" but for whom the marginal benefit of the full method justifies its cost—Section 7's dose-response data are the relevant evidence.

Versus the rest of this series. Every prior document's treatment is, historically, a reaction to this one: CBT's protocols against analytic open-endedness; DBT's explicit structure against analytic ambiguity; the existential school against drive theory's reductions; even ACT's defusion echoing, at speed, what analysis calls observing ego. The reactions corrected real faults; the corrected faults should not obscure what the original got right and the successors quietly retain—the relationship as instrument, the meaningfulness of symptoms, the patient's participation in their own suffering.

The Research Evidence

Psychoanalysis is the hardest treatment in this series to evaluate by RCT standards, for structural reasons stated plainly: multi-year treatments resist randomization and blinding; the patients who seek analysis resist random allocation; the outcome claims (structural change) outrun symptom scales; and the field came to outcome research late and reluctantly. What exists:

Randomized and quasi-randomized comparisons. The Helsinki Psychotherapy Study (Knekt and colleagues)—the most informative single project—randomized patients among short-term therapies and long-term psychodynamic psychotherapy, with a self-selected psychoanalysis arm followed in parallel: short-term treatments worked faster; long-term therapy overtook them by year three; psychoanalysis showed the greatest gains only at five-to-ten-year horizons—suggestive of a real dose effect, permanently qualified by the non-randomized analytic arm. The Munich Psychotherapy Study (Huber) randomized depressed patients among psychoanalysis, psychodynamic therapy, and CBT: broadly comparable acute outcomes with advantages for the analytic treatments on some long-term and personality-level measures. The German LAC depression study (Leuzinger-Bohleber) compared long-term analytic and cognitive-behavioral treatment of chronic depression under preference and randomized allocation, finding substantial, roughly equivalent improvement in both—with high remission persisting at long follow-up. The Stockholm (STOPPP) and German DPV follow-up studies add large naturalistic cohorts showing sustained post-termination improvement after analysis, with the sleeper-effect pattern familiar from the psychodynamic literature—and the same vulnerability to selection and regression artifacts.

Meta-analytic synthesis. De Maat and colleagues' reviews of psychoanalysis proper conclude that it is associated with large pre-post symptomatic and moderate personality-level improvement, maintained at follow-up—on evidence overwhelmingly observational. The defensible summary sentence for a website: long-term analytic treatment produces durable improvement in suitable patients; that it produces more than well-delivered briefer or weekly treatment is suggested by dose-response and follow-up data but is not established by randomized comparison.

Process and dose findings. High frequency shows associations with deeper process and, in some cohorts, better outcome—confounded everywhere by selection. Transference-interpretation research (Høglend's trials, in weekly therapy) complicates doctrine usefully: benefit concentrated in patients with poorer relational functioning, with healthier patients doing as well or better without. Mechanism measures (reflective functioning, defense maturation, SWAP-assessed personality change) move with analytic treatment in the expected directions—the field's most genuine empirical progress.

Cost-effectiveness. The Nordic and German data suggest reduced healthcare utilization, medication use, and disability after long-term analytic treatment, partially offsetting cost; against the counterfactual of equivalent symptomatic outcomes from far briefer treatment for many patients, the economic case for the full method remains the literature's weakest front and should be presented as such.

Criticisms and Controversies

The scientific status of the theory

The classical indictments—Popper's unfalsifiability charge and Grünbaum's more penetrating critique (the theory is testable but the couch cannot test it: clinical data are contaminated by suggestion, and the "tally argument" linking cure to true insight fails)—remain the serious philosophical core, joined by Crews' and other historians' documentation of constructed evidence in the founding case studies. The modern field's answer is partial reformation: extra-clinical testing (attachment, mentalization, defense research), manualized derivatives in trials, and abandonment of the indefensible content. The unreformed remainder—school loyalties adjudicated by authority rather than data, explanation that retrofits any outcome—still exists and still merits the critique.

Cost, length, and opportunity cost

The decisive practical criticism: four sessions weekly for five years is an enormous expenditure of money, time, and life, available to almost no one without subsidy, and the evidence that this dose outperforms one-tenth of it is suggestive at best. The burden of proof sits with the longer treatment, and for most patients with most disorders it has not been met; analysis's defensible modern niche—chronic characterological suffering, prior treatment failure, the training of clinicians, and patients who choose depth with informed eyes—is far narrower than its historical self-conception.

Guild culture and its casualties

The institute system has produced documented pathologies: idealization of training analysts, doctrinal policing, schism as the default mode of dispute, and—most seriously—the decades-long pathologization of homosexuality and exclusion of gay candidates, reversed only in the 1990s and formally apologized for in 2019. Classical theory's treatment of women (penis envy and the rest) is conceded ground even internally. A tradition whose method is the analysis of self-deception proved strikingly capable of institutional self-deception; contemporary institutes are smaller, humbler, and more pluralist, but the history belongs in the account.

Suggestion, dependency, and harm

The intensity that gives analysis its power concentrates its risks: transference dependency that outlives its analytic use; interminable treatments without external review; the suggestion problem Grünbaum named—patients of every school produce material confirming their analyst's theory; boundary violations, whose most damaging documented cases in psychotherapy's history cluster in high-intensity, low-oversight treatment; and the recovered-memory catastrophe discussed in the companion document, whose doctrinal roots lay here. Modern safeguards—ethics codes, consultation norms, the field's own literature on boundary theory (Gabbard)—are real and were largely built in response.

Marginality and the two futures

Within psychiatry, psychoanalysis's residual influence is mostly indirect—formulation, the psychotherapy competencies in residency training, the personality-disorder treatments. Its candidates are few; its patients fewer. The field's plausible futures divide: a continued contraction into a self-credentialing humanities discipline, or the research-engaged path (mentalization, neuropsychoanalysis, the German trial culture, single-case process science) that treats the method as a hypothesis still worth testing. Both futures are visibly underway in different institutes.

What Patients Can Expect, and Practical Considerations

What treatment feels like. Unlike anything else in this series: the couch, the analyst mostly silent and out of sight, no agenda but the rule to speak freely—initially strange, often frustrating by design, and over months capable of a depth of self-encounter weekly formats rarely reach. The relationship will become intense and will itself become the work; patients should know this in advance, as a feature and a demand.

The practical commitment. Three to five sessions weekly, typically three to seven years; full fees are substantial, but institute clinics in most major cities offer analysis at deeply reduced rates with supervised advanced candidates—the field's best-kept access secret and a legitimate option to name on a website. Analysis coexists with medication management without difficulty in modern practice.

Who should consider it—and who should not. Consider: chronic characterological difficulties that have survived adequate briefer treatment; clinicians in formation; patients seeking maximal depth with informed consent about evidence and cost. Route elsewhere first: any presentation owning a first-line protocol (the rest of this series), acute instability, and anyone for whom the frame's demands would constitute hardship rather than commitment.

Finding qualified care. Graduate analysts of APsA- or IPA-component institutes (or candidates in supervised training via institute clinics); directories through APsA and the IPA.

Conclusion

Psychoanalysis is simultaneously the most criticized treatment in psychiatry's history and the source of more of modern psychotherapy than any other single tradition. Judged as its founders advertised it—a general science of mind and the uniquely curative treatment—it failed, and the failures (evidential complacency, guild orthodoxy, real casualties) are part of its record. Judged as what it demonstrably is—the maximal-dose form of a treatment family whose weekly form is empirically co-equal with CBT, with observational and dose-response evidence of deep and durable change in suitable patients, and a training system that still produces the field's most disciplined clinical listeners—it retains a defensible, narrow, honest place: for chronic characterological suffering, for treatment-refractory depth work, for the formation of therapists, and for patients who, knowing exactly what the evidence does and does not show, choose the long road. A psychiatry website serves its readers by presenting both judgments—and by knowing which patients each one is for.

Selected References and Further Reading

1. Freud, S. (1900/1953). The Interpretation of Dreams. Standard Edition, Vols. 4–5. Hogarth Press.
2. Freud, S. (1911–1915/1958). Papers on technique. Standard Edition, Vol. 12. Hogarth Press.
3. Greenson, R.R. (1967). The Technique and Practice of Psychoanalysis, Vol. 1. International Universities Press.
4. Mitchell, S.A., & Black, M.J. (1995). Freud and Beyond: A History of Modern Psychoanalytic Thought. Basic Books.
5. Gabbard, G.O., Litowitz, B.E., & Williams, P. (Eds.) (2012). Textbook of Psychoanalysis (2nd ed.). American Psychiatric Publishing.
6. Knekt, P., et al. (2011). Quasi-experimental study on the effectiveness of psychoanalysis, long-term and short-term psychotherapy on psychiatric symptoms, work ability and functional capacity during a 5-year follow-up. Journal of Affective Disorders, 132(1–2), 37–47.
7. Huber, D., Henrich, G., Clarkin, J., & Klug, G. (2013). Psychoanalytic versus psychodynamic therapy for depression: A three-year follow-up study. Psychiatry, 76(2), 132–149.
8. Leuzinger-Bohleber, M., et al. (2019). Outcome of psychoanalytic and cognitive-behavioural long-term therapy with chronically depressed patients: A controlled trial with preferential and randomized allocation. Canadian Journal of Psychiatry, 64(1), 47–58.
9. Sandell, R., et al. (2000). Varieties of long-term outcome among patients in psychoanalysis and long-term psychotherapy (STOPPP). International Journal of Psychoanalysis, 81(5), 921–942.
10. de Maat, S., et al. (2013). The current state of the empirical evidence for psychoanalysis: A meta-analytic approach. Harvard Review of Psychiatry, 21(3), 107–137.
11. Høglend, P., et al. (2008). Transference interpretations in dynamic psychotherapy: Do they really yield sustained effects? American Journal of Psychiatry, 165(6), 763–771.
12. Fonagy, P. (2015). The effectiveness of psychodynamic psychotherapies: An update. World Psychiatry, 14(2), 137–150.
13. Grünbaum, A. (1984). The Foundations of Psychoanalysis: A Philosophical Critique. University of California Press.
14. Popper, K. (1963). Conjectures and Refutations. Routledge.
15. Crews, F. (2017). Freud: The Making of an Illusion. Metropolitan Books.
16. Solms, M. (2021). The Hidden Spring: A Journey to the Source of Consciousness. W.W. Norton. [And Solms, M. (2018). The scientific standing of psychoanalysis. BJPsych International, 15(1), 5–8.]
17. Kernberg, O.F. (2016). Psychoanalytic education at the crossroads. International Journal of Psychoanalysis (and related papers on institute reform).
18. Drescher, J. (2008). A history of homosexuality and organized psychoanalysis. Journal of the American Academy of Psychoanalysis and Dynamic Psychiatry, 36(3), 443–460.
19. Gabbard, G.O., & Lester, E.P. (1995). Boundaries and Boundary Violations in Psychoanalysis. American Psychiatric Press.
20. Eagle, M.N. (2011). From Classical to Contemporary Psychoanalysis: A Critique and Integration. Routledge.