Psychoanalytic Therapy

Abstract

Psychoanalytic therapy is a form of psychotherapy that treats distress by making unconscious conflict conscious, working from the theory of mind Sigmund Freud began building in the 1890s. This article treats it as a clinical and scientific object: the method Freud derived from free association and the analysis of transference, the structural model of id, ego, and superego, and the modern psychodynamic treatments that descend from it. It follows the tradition from the cathartic treatment of hysteria, through the codification of the analytic setting, to the evidence base assembled by meta-analysis and equivalence testing. It also distinguishes classical psychoanalysis from the shorter psychodynamic psychotherapies now used in routine practice. Three interactive demonstrations model structural conflict, transference, and the size of a treatment effect.

Keywords: transference, psychodynamic therapy, unconscious conflict

Psychoanalytic therapy is the clinical application of psychoanalysis: a family of treatments that trace psychological symptoms to conflicts outside awareness and work to bring those conflicts into the open, where they can be understood and revised. Its founding wager, set out by Josef Breuer and Freud in the case studies of the 1890s, is that symptoms have meaning, that a neurotic complaint is the disguised expression of a wish or memory the patient cannot consciously tolerate, and that recovering that meaning in the presence of a receptive listener is itself curative (Breuer & Freud, 1895). Everything that distinguishes the tradition, the couch, the injunction to say whatever comes to mind, the close attention to the patient's feelings about the analyst, follows from that single premise and the method built to act on it.

What Psychoanalytic Therapy Is

Psychoanalytic therapy is a treatment in which a trained clinician helps a patient recognise and rework patterns of thought, feeling, and relating that operate outside awareness and that generate symptoms or self-defeating conduct. It is defined by a set of assumptions rather than a technique menu: that much of mental life is unconscious, that early experience shapes adult patterns, that people resist the very self-knowledge that would help them, and that the relationship with the therapist becomes a live arena in which those patterns replay and can be changed. Shedler's synthesis identifies the features that distinguish it in practice, a focus on affect and its expression, attention to avoidance and resistance, the search for recurring themes, an emphasis on past experience and on the therapy relationship itself, and the exploration of wishes and fantasy (Shedler, 2010).

The word psychoanalysis is reserved for the intensive, open-ended treatment Freud codified, typically several sessions a week over years, with the patient on a couch and the analyst largely out of view. Psychoanalytic or psychodynamic psychotherapy denotes the wider family that applies the same theory and technique less intensively, usually once or twice weekly, face to face, and often with a defined focus and a time limit. Gabbard's standard text frames the difference as one of degree rather than kind: the same understanding of unconscious conflict and transference informs both, and the choice between them turns on the patient's goals, resources, and capacities rather than on a different theory of cure (Gabbard, 2017). This article treats the family as a whole while marking where the intensive and the brief forms diverge.

Types of Psychoanalytic Therapy

Beyond being a subject in its own right, Psychoanalytic Therapy is a formal category in the National Library of Medicine's Medical Subject Headings, which files it at tree position F04.754.709, beneath Psychotherapy, and hangs its recognised narrower kinds beneath it. These subtypes are a scheme built to index the literature, not a claim about the mind's natural joints; they cut across one another, so a single treatment may draw on several at once, and the category mixes a bedrock technique with a whole derived school. Table 1 lists the direct children of the descriptor; neither yet has its own article on this site, so both are shown as plain text.

The pairing is heterogeneous by design. It places a foundational technique, free association, which is a component of nearly every analytic treatment rather than a treatment in its own right, beside transactional analysis, a distinct mid-century school with its own vocabulary and clinical procedures. Inclusion in the index records that a literature exists to be catalogued, not that the two are parallel therapies or that either is separately validated; the sections that follow concern the mainstream psychoanalytic and psychodynamic treatments for which the outcome evidence is strongest.

Origins: Catharsis, Free Association, and the Unconscious

The tradition begins with a single case. Between 1880 and 1882 Breuer treated a young woman he called Anna O., whose hysterical symptoms eased when she traced each one, under hypnosis, back to the forgotten circumstance in which it had arisen and discharged the accompanying emotion, a process she named the “talking cure.” Breuer and Freud generalised the observation into a theory: hysterics “suffer mainly from reminiscences,” from memories too painful to be kept in consciousness whose strangulated affect is converted into bodily symptoms, and the cure lies in recovering the memory and releasing the affect (Breuer & Freud, 1895). This cathartic method was the seed of everything that followed, but Freud soon abandoned its central tool. Hypnosis worked unreliably and only for some patients, so he replaced it with free association, instructing patients to report every thought that occurred to them, however trivial, shameful, or irrelevant it seemed. The associations were never truly free: where they faltered, circled, or dried up, Freud inferred the pressure of a resistance, and behind the resistance a repressed idea the patient was unconsciously working to keep out of awareness.

Freud's first map of the mind, the topographic model, divided it by accessibility to consciousness. Figure 1 shows its three regions: a narrow conscious zone of present awareness, a preconscious of material that can be recalled at will, and a large unconscious of wishes and memories held down by repression and reachable only indirectly, through slips, dreams, and the drift of association. The therapeutic task, on this picture, is to enlarge the first region at the expense of the third, to make the unconscious conscious, and free association is the instrument that lets the analyst work back from the surface of speech to what lies beneath it.

The Structural Model of the Mind

The topographic model could not easily explain why the ego's own defences, the forces doing the repressing, are themselves unconscious. In 1923 Freud recast the theory in structural terms, dividing the mind not by depth but by function into three agencies whose conflict generates symptoms (Freud, 1923). The id is the reservoir of instinctual drives, seeking immediate gratification with no regard for reality or morality. The superego is the internalised voice of parental and social prohibition, the conscience and the ideal, which condemns and forbids. Between them the ego mediates: it answers to the id's demands, the superego's strictures, and the constraints of external reality at once, and when it cannot reconcile them it deploys defence mechanisms, repression foremost among them, to keep the conflict out of awareness at the cost of a symptom.

On this account a neurotic symptom is a compromise formation: a partial, disguised discharge of a forbidden id impulse that simultaneously satisfies the superego's need to punish and the ego's need for safety. The clinical aim follows directly, “where id was, there ego shall be,” to expand the reach of the reflective, reality-testing ego over territory previously governed by blind impulse and rigid prohibition. The first demonstration makes this dynamic manipulable: raising the id's drive or the superego's severity increases the pressure on the ego, which responds by binding more of the conflict in defence, and the readout tracks the residual anxiety and the degree of symptomatic compromise that results.

Transference and the Analytic Relationship

Freud's most consequential clinical discovery was that patients do not merely recall their formative relationships in analysis; they re-enact them, casting the analyst in roles drawn from their past and reacting to the analyst as though he were a parent, a rival, or a lover. He named this transference and, after initially regarding it as an obstacle, came to see it as the treatment's central instrument: the neurosis is not described from memory but revived in the consulting room, where it becomes available for observation and change (Freud, 1912). When a patient's response to the analyst is out of proportion to anything the analyst has actually done, the surplus is transference, a template from an earlier relationship imposed on the present one, and interpreting that surplus lets the patient see a pattern in action rather than merely hear about it.

Two features make transference workable. First, the analyst's relative neutrality and restraint keep the screen comparatively blank, so that what the patient projects onto it can be recognised as coming from the patient. Second, the same mechanism runs in reverse: the analyst's own unconscious reactions to the patient, or countertransference, once treated as a contaminant, are now used as data about what the patient evokes in others. The second demonstration models the projection directly. A neutral therapist response is held fixed while the reader varies how strongly a past-relationship template is activated; as activation rises, the patient's perceived meaning of the identical response swings from benign toward hostile or rejecting, and the gap between the response and its perception is the quantity an interpretation would target.

From Psychoanalysis to Modern Psychodynamic Therapy

The treatment Freud practised was long, intensive, and open-ended, and for much of the twentieth century its length and its resistance to measurement drew the charge that it was untestable dogma. The response, from the 1960s onward, was the development of briefer, focused psychodynamic psychotherapies that retain the core theory, unconscious conflict, transference, defence, while imposing a time limit and an agreed focus. These short-term dynamic treatments meet once or twice a week over months rather than years, sit the patient and therapist face to face, and target a circumscribed problem, yet they work with the same material: the recurring relational pattern, the avoided feeling, the resistance that appears the moment the work approaches something important. Barber and colleagues document how this manualisation made the tradition researchable, turning an artisanal practice into a set of defined procedures that a trial could compare against a control (Barber et al., 2013).

What survived the compression is the mechanism the tradition has always claimed as its own: change through insight, the patient's growing understanding of how a past conflict shapes present life, consolidated by working through, the repeated recognition of the same pattern across the many situations in which it recurs. This distinguishes psychodynamic therapy sharply from the behavioural and cognitive treatments that target symptoms and beliefs directly. The distinction is real but not absolute, and the two families increasingly borrow from each other; what remains characteristic of the psychodynamic approach is its conviction that lasting change requires understanding the hidden purpose a symptom serves, not merely removing it.

The Evidence Base

For decades the efficacy of psychoanalytic treatment was argued from case reports and clinical conviction, exactly the ground on which the wider field had learned to distrust its own claims. Eysenck's notorious 1952 review made the charge explicit: surveying the outcome literature, he claimed that patients given psychotherapy, with psychoanalysis prominent among the treatments he indicted, recovered no more often than those left untreated, a provocation that set the empirical agenda for the next half century (Eysenck, 1952). The picture changed as short-term psychodynamic therapy was manualised and subjected to controlled trials, and the accumulated results now support a firm conclusion. An early meta-analysis by Leichsenring and Leibing found psychodynamic therapy effective for personality disorders, with large pre-post effects sustained at follow-up (Leichsenring & Leibing, 2003). For the intensive, long-term form specifically, Leichsenring and Rabung pooled the controlled studies and reported a large between-group effect favouring long-term psychodynamic psychotherapy over shorter or less intensive comparisons in complex and chronic disorders (Leichsenring & Rabung, 2008). A Cochrane review of short-term psychodynamic therapy for common mental disorders found significant benefit over controls across somatic, anxiety, and depressive presentations, with gains that tended to grow at long-term follow-up (Abbass et al., 2014).

The depression literature has been examined most closely. Driessen and colleagues, updating the evidence for short-term psychodynamic therapy in depression, found it more effective than control conditions and not significantly different from other active treatments at termination (Driessen et al., 2015), and Fonagy's overview reached the same verdict across the disorders for which trials exist (Fonagy, 2015). The strongest test of that equivalence came from Steinert and colleagues, who set out to test not whether psychodynamic therapy beats a control but whether its outcomes are statistically equivalent to those of other empirically supported treatments, and found that they fell within a pre-specified equivalence margin (Steinert et al., 2017). The third demonstration lets a reader work with the effect sizes these syntheses report, converting a standardized mean difference into the percentile standing and the number-needed-to-treat that give it clinical meaning.

Worked Example

Consider a controlled trial of short-term psychodynamic therapy for depression that reports a between-group effect size of d = 0.60 at termination, a value in the mid-range of the syntheses above. What does that number mean for a patient? An effect size in standard-deviation units can be read in three complementary ways, each derived from the same normal-overlap assumption.

The first is Cohen's U₃, the percentile of the control distribution reached by the average treated patient. It is the normal cumulative probability of the effect size, Φ(0.60) = 0.726, so the average treated patient ends up better off than about 73% of untreated controls, against the 50% that would hold with no treatment. The second converts the effect size to a correlation, r = d / √(d² + 4) = 0.60 / √4.36 = 0.29, the point-biserial correlation between receiving treatment and improving. The third makes that correlation concrete through the binomial effect size display, which reads r as a difference in success rates around a 50% base: success rises from (1 − 0.29)/2 = 36% in the control condition to (1 + 0.29)/2 = 64% with treatment. That 28-point difference implies a number-needed-to-treat of 1 / 0.287 ≈ 3.5: roughly one additional patient recovers for every three or four treated who would not have recovered otherwise. A “modest” effect size of 0.60, in other words, corresponds to a clinically substantial shift, and the same arithmetic lets any reported effect be translated into these terms.

Discussion

Psychoanalytic therapy occupies an unusual position in clinical science: the oldest of the modern psychotherapies, long treated as the least empirical, it now carries a respectable and growing evidence base for its briefer forms. The outcome question has largely been answered, short-term psychodynamic therapy helps, and helps about as much as its established rivals, so the live disputes have moved inward, to mechanism. The tradition's distinctive claim is that it works through insight into unconscious conflict, yet insight is precisely what is hardest to measure, and demonstrating that symptom change is caused by the growth of self-understanding, rather than by the alliance, the hope, and the ritual it shares with every therapy, remains an open problem. Yakeley's appraisal of psychoanalysis in modern mental-health practice frames the challenge squarely: the treatment must justify its cost and intensity in a system that increasingly demands mechanism as well as outcome (Yakeley, 2018).

Two cautions temper the optimism. First, the evidence is strongest for the manualised short-term treatments and thinnest for the intensive, long-term psychoanalysis from which the theory descends, so the empirical case does not straightforwardly vindicate classical practice. Second, like all psychotherapy research, the field is exposed to allegiance effects and publication bias, and its trials are fewer and often smaller than those of cognitive behavioural therapy. The honest summary is that psychoanalytic therapy has earned its place among the empirically supported treatments for common disorders while leaving its signature theoretical claim, cure through the recovery of unconscious meaning, still to be demonstrated as a mechanism rather than assumed as a rationale.

Current Directions

Three lines of work define the field's active front. The first is the move from efficacy to equivalence and beyond: having shown that psychodynamic therapy matches other treatments on average, researchers are now asking for whom it does better or worse, and the umbrella review by Leichsenring and colleagues assembles the accumulated meta-analyses to grade the treatment's status disorder by disorder against updated evidentiary criteria (Leichsenring et al., 2023). The second concerns the harder cases the mainstream trials often exclude. For borderline personality disorder, where psychodynamically derived treatments such as transference-focused and mentalization-based therapy were designed, the comparative evidence is still consolidating: Cristea and colleagues' meta-analysis found that psychotherapies for the disorder produce real but moderate effects and that no single approach clearly dominates (Cristea et al., 2017).

The third direction is mechanism itself. Process researchers are pursuing the tradition's central claim empirically, testing whether measured gains in insight, reflective functioning, or the resolution of a focal conflict mediate symptom change, and whether transference interpretation specifically adds to outcome. This is where the modern literature is thinnest and where the tradition's future credibility mostly rests: converting a century-old theory of how change happens into mediational evidence that meets the standards now applied to every psychotherapy. The equivalence findings bought psychoanalytic therapy a seat at the table; demonstrating its mechanism is the work that would justify keeping it there (Steinert et al., 2017).

Key Researchers

Allan Abbass. Professor of psychiatry at Dalhousie University and founding director of its Centre for Emotions and Health; a leading investigator of short-term psychodynamic therapy and its Cochrane evidence base. ORCID - Google Scholar - Wikipedia

Jacques P. Barber. Professor emeritus at the Derner School of Psychology, Adelphi University; a psychotherapy-process researcher and co-author of the psychodynamic therapies chapter in Bergin and Garfield's handbook. ORCID - Google Scholar - Wikipedia

Peter Fonagy. Head of the Division of Psychology and Language Sciences at University College London; developer of mentalization-based treatment and author of a major update on the effectiveness of psychodynamic psychotherapies. ORCID - Faculty Page - Google Scholar - Wikipedia

Sigmund Freud (1856-1939). Viennese neurologist and founder of psychoanalysis; originator of free association, the analysis of transference, and the structural model of the mind. Wikipedia

Glen O. Gabbard. Clinical professor of psychiatry at Baylor College of Medicine; author of the standard textbook on long-term psychodynamic psychotherapy. Faculty Page - Wikipedia

Falk Leichsenring. Professor of psychotherapy research at the University of Giessen; author of the meta-analyses and umbrella reviews that anchor the modern evidence for psychodynamic therapy. ORCID

Jonathan Shedler. Clinical professor of psychiatry at the University of California, San Francisco; author of the widely cited review of the efficacy of psychodynamic psychotherapy. Google Scholar

Glossary

Cathartic method.

Breuer and Freud's original treatment, in which recovering the forgotten memory behind a symptom and discharging its associated emotion relieved the symptom; the direct precursor of free association.

Countertransference.

The analyst's own unconscious emotional reactions to the patient, once regarded as an obstacle and now used as evidence about the feelings the patient evokes in others.

Defence mechanism.

An unconscious operation of the ego that keeps a threatening impulse or idea out of awareness, such as repression, denial, or projection, typically at the cost of a symptom or a restriction of experience.

Ego.

In the structural model, the agency that mediates among the demands of the id, the superego, and external reality, and that deploys the defences; the seat of reality-testing and reflection.

Free association.

The fundamental analytic technique in which the patient reports every thought that arises without selection or censorship, so that unconscious material can emerge in the gaps and detours of speech.

Id.

In the structural model, the wholly unconscious reservoir of instinctual drives that seeks immediate gratification without regard to reality or morality.

Interpretation.

The analyst's verbal intervention that gives the patient a previously unconscious meaning of a symptom, association, or transference reaction; the principal tool by which insight is offered.

Psychoanalysis.

The intensive, open-ended treatment Freud codified, conducted several times a week with the patient on a couch, from which the briefer psychodynamic psychotherapies derive.

Psychodynamic psychotherapy.

The briefer, less intensive application of psychoanalytic theory and technique, usually once or twice weekly and often time-limited, that constitutes most contemporary analytic practice.

Repression.

The foundational defence mechanism, the active exclusion of an intolerable wish or memory from consciousness, whose return in disguised form Freud held to underlie neurotic symptoms.

Resistance.

The patient's unconscious opposition to the progress of the treatment, visible when associations falter or the work approaches a repressed idea; its analysis is a central part of the therapy.

Structural model.

Freud's 1923 division of the mind by function into id, ego, and superego, whose conflict produces symptoms as compromise formations; it superseded the earlier topographic model.

Superego.

In the structural model, the internalised agency of parental and social prohibition and ideals, functioning as conscience; its severity contributes to guilt and to symptom formation.

Transference.

The patient's redirection onto the analyst of feelings and patterns drawn from earlier relationships, which revives the neurosis in the consulting room and makes it available for interpretation.

Unconscious.

In the topographic model, the region of the mind holding wishes and memories kept from awareness by repression and reachable only indirectly, through dreams, slips, and free association.

Working through.

The repeated recognition of the same unconscious conflict across the many situations in which it recurs, by which a single insight is consolidated into durable change.

Frequently Asked Questions

What is psychoanalytic therapy?
It is a form of psychotherapy that treats distress by uncovering and interpreting conflicts outside conscious awareness, using techniques such as free association and the analysis of transference that derive from Freud's clinical method (Shedler, 2010).

How is psychoanalysis different from psychodynamic psychotherapy?
Psychoanalysis is the intensive, open-ended form, several sessions a week on the couch, while psychodynamic psychotherapy applies the same theory less intensively, usually once or twice weekly and often with a time limit, and is the form used in most modern practice (Gabbard, 2017).

What is transference?
It is the patient's redirection onto the therapist of feelings and relational patterns drawn from earlier relationships, which Freud came to regard as the treatment's central tool because it revives the patient's difficulties where they can be observed and changed (Freud, 1912).

What are the id, ego, and superego?
They are the three agencies of Freud's structural model: the id holds instinctual drives, the superego holds internalised prohibition and ideals, and the ego mediates between them and external reality, producing symptoms when it cannot reconcile their competing demands (Freud, 1923).

Does psychoanalytic therapy actually work?
Yes; controlled trials and meta-analyses find that short-term psychodynamic therapy produces reliable benefit over control conditions for depression, anxiety, and personality disorders, with effects that often grow at follow-up (Abbass et al., 2014).

Is it as effective as other therapies?
For the disorders in which it has been tested, its outcomes are broadly equivalent to those of other established treatments; one study designed specifically to test equivalence found psychodynamic therapy's results within a pre-specified margin of empirically supported comparators (Steinert et al., 2017).

How does psychoanalytic therapy claim to produce change?
Its distinctive mechanism is insight, the patient's growing understanding of how an unconscious conflict shapes present life, consolidated through working through, rather than the direct symptom targeting of behavioural and cognitive treatments (Fonagy, 2015).

What is the main open question about it?
Whether its signature mechanism can be demonstrated empirically, since showing that symptom change is caused by insight into unconscious conflict, rather than by the factors it shares with all therapies, remains difficult and is the focus of current process research (Yakeley, 2018).

References