Eye Movement Desensitization and Reprocessing

Abstract

Eye movement desensitization and reprocessing (EMDR) is a form of psychologic desensitization: a structured, eight-phase psychotherapy in which a patient briefly holds a distressing memory in mind while tracking a therapist's moving fingers or another bilateral stimulus. Devised by Francine Shapiro in the late 1980s, it is now a first-line treatment for post-traumatic stress disorder, recommended by the World Health Organization on the strength of meta-analytic trials. Its theory, the adaptive information processing model, holds that trauma is stored in an unprocessed form the procedure helps integrate; its most tested mechanism, the working-memory account, holds that the concurrent eye movements tax a limited visuospatial resource and so blunt the memory's vividness. This article sets out the protocol, the two theories, the efficacy evidence, and the still-live dispute over whether the eye movements matter at all.

Keywords: eye movement desensitization reprocessing, working memory, post-traumatic stress disorder, bilateral stimulation

Eye movement desensitization and reprocessing is a therapy in which a patient recalls a distressing event while simultaneously performing a distracting bilateral task, most often tracking the therapist's hand as it moves rapidly back and forth across the visual field. The recall and the eye movements are alternated in short sets, and between sets the patient reports whatever has come to mind, allowing the memory to shift in content and emotional charge until it can be brought to mind without acute distress (Shapiro, 1989a). The method is unusual among trauma therapies in asking the patient to divide attention rather than to focus it, and that feature is both the source of its clinical signature and the center of the debate about how, and whether, it works (van den Hout & Engelhard, 2012).

What EMDR Is

EMDR is a psychotherapy method, not a theory or a disorder, and MeSH classifies it as a form of psychologic desensitization, the broader family of treatments that reduce a conditioned emotional reaction by graded, controlled contact with what provokes it. What sets EMDR apart within that family is the pairing of memory recall with a concurrent bilateral task: the patient does not simply approach the feared material but holds it in mind while attention is pulled rhythmically elsewhere (Shapiro, 1989b).

The procedure has a characteristic rhythm. The therapist asks the patient to bring to mind a target memory together with the negative belief, bodily sensation, and emotion attached to it, and to rate the distress it provokes. The patient then follows a moving stimulus — classically the therapist's fingers sweeping left and right about twice a second — for a short set of twenty to thirty movements, after which they pause, notice whatever has surfaced, and report it briefly before the next set begins. Across successive sets the material typically shifts on its own, and the distress rating is expected to fall until the memory can be recalled calmly and a preferred positive belief about oneself feels true (Shapiro, 2018).

What distinguishes EMDR from ordinary exposure is this divided-attention structure. In prolonged exposure the patient concentrates fully on the trauma memory; in EMDR the memory is deliberately competed with, and the eye movements, hand taps, or alternating tones are what supply the competition. That difference is not incidental decoration but the element the method is named for and the one its mechanistic debate turns on, because a therapy that works by dividing attention demands a different explanation than one that works by sustaining it (Andrade et al., 1997).

The Eight-Phase Protocol

EMDR is delivered through a standardized sequence of eight phases, and the reprocessing that gives the method its reputation occupies only the middle of that sequence. The phases move from preparation through the paired-recall work and out to consolidation and review, and treating them as a fixed order is part of what made the therapy manualizable and testable rather than an idiosyncratic technique (Shapiro, 2018).

The first phases establish the ground. History-taking identifies the target memories and the patient's readiness; preparation builds the therapeutic relationship and teaches self-calming skills the patient can use if distress becomes overwhelming; and assessment fixes the specific image, the negative cognition, a desired positive cognition, and baseline ratings of distress and of belief. Two rating scales run through the work: the Subjective Units of Distress scale, from 0 to 10, and the Validity of Cognition scale, from 1 to 7, which tracks how true the preferred positive belief feels (Shapiro, 1989a).

The heart of the method is phase four, desensitization, where recall and eye movements are alternated in sets until the Subjective Units of Distress rating for the target falls to zero or near it. Phase five, installation, then pairs the now-neutral memory with the patient's chosen positive belief and strengthens it with further sets, and phase six scans the body for residual tension. Closure returns the patient to a stable state whether or not the target is finished, and reevaluation at the following session checks that gains have held before the next memory is chosen (de Jongh et al., 2019).

The Adaptive Information Processing Model

Shapiro's clinical theory of why EMDR works is the adaptive information processing model. It proposes an innate system that normally metabolizes experience, integrating new events into existing memory networks so that their useful lessons are retained and their distress fades. Trauma, on this account, overwhelms that system, so the memory is stored in an unprocessed, state-specific form — locked with the original images, sensations, and beliefs intact — and it is this frozen storage, rather than the event itself, that produces later symptoms when the memory is triggered (Shapiro, 2018).

On this model the eight-phase procedure works by resuming the stalled processing. Bringing the memory into awareness under safe, dual-attention conditions is said to let the adaptive system make the links it could not make at the time, so the memory loses its charge and is re-filed with the ordinary autobiographical record. The model is deliberately broad, and it is best understood as a clinical heuristic that organizes practice rather than a specified cognitive mechanism, which is why much of the experimental literature has looked elsewhere for a testable account of the eye movements (Landin-Romero et al., 2018).

Why the Eye Movements? The Working-Memory Account

The most tested explanation of the eye movements comes not from the adaptive information processing model but from cognitive psychology, in the working-memory account first set out by Andrade, Kavanagh, and Baddeley. Its premise is Baddeley's model of working memory as a limited-capacity system with a distinct visuospatial component, the sketchpad, that holds and manipulates mental images. Recalling a vivid autobiographical memory draws on that sketchpad, and so does tracking a moving target; because the resource is limited and shared, doing both at once means neither is done fully, and the recalled image is held in a degraded, less vivid form (Andrade et al., 1997).

The clinical payoff follows from a second step. A memory recalled while degraded is, on this view, re-stored in that weaker form, so that after the dual task the image is not only momentarily blurrier but durably less vivid and less emotionally intense when next brought to mind. Laboratory studies support the pattern: recalling a negative memory during eye movements reliably lowers its rated vividness and emotionality relative to recall alone, and the effect appears with other tasks that tax the same resource, which is exactly what a working-memory account predicts and an eye-movement-specific account does not (van den Hout & Engelhard, 2012).

The account also predicts a dose relationship shaped like an inverted U. A concurrent task that is too undemanding leaves the memory almost fully resourced and changes little; a task so demanding that it crowds the memory out of mind entirely gives nothing to reprocess; the largest reduction comes from an intermediate load that competes with the image while still letting it be held. Evidence for this central-load prediction is one of the stronger pieces of support for the theory, and it reframes the therapist's twice-a-second sweep as a calibrated demand rather than a mystical stimulus (Gunter & Bodner, 2008).

The working-memory account is best understood against the eye-movement-specific alternatives it competes with. One such account holds that the lateral eye movements elicit an investigatory orienting response, the reflexive shift of attention toward a novel stimulus, which dampens arousal; another, a neurobiological account, likens the movements to the rapid eye movements of REM sleep and proposes that they engage a sleep-like memory-processing state (Stickgold, 2002). Both tie the effect to the specific act of moving the eyes, which is exactly what the working-memory account denies: if competition for a limited resource is what matters, any sufficiently taxing task should serve and eye movements should hold no privileged status. That contrast is what much of the dismantling and task-comparison work is designed to adjudicate.

Origins: Shapiro's Discovery

EMDR began with an observation Francine Shapiro reported making on a walk in 1987, when she noticed that deliberately moving her eyes back and forth seemed to lessen the grip of a distressing thought. She developed the observation into a procedure, first called eye movement desensitization, and tested it in a controlled study of trauma memories, reporting large drops in distress after a single session (Shapiro, 1989a). A companion paper set the method out as a treatment for post-traumatic stress disorder and positioned it within the behavioral tradition of desensitization (Shapiro, 1989b).

The early reception was sharply divided, and for reasons that still shape the field. The reported effects were unusually large and the proposed mechanism unusually novel, an unfamiliar pairing that drew both rapid clinical uptake and pointed methodological criticism. Reviewers noted that many early studies lacked adequate controls, and a recurring charge was that the eye movements might be an inert addition to a treatment that worked through ordinary exposure, a suspicion that motivated the dismantling studies of the following decade (Maxfield & Hyer, 2002). Over time the method was renamed to add reprocessing, its protocol was standardized and manualized, and its evidence base was rebuilt on better-controlled trials, so that the therapy now recommended by guideline bodies is a more disciplined descendant of Shapiro's first procedure (Shapiro, 2018).

The Evidence Base

Two questions have to be kept apart when weighing the evidence for EMDR: whether the whole treatment reduces trauma symptoms, and whether the eye movements specifically contribute to that reduction. On the first, the evidence is now strong. Controlled trials and their syntheses find EMDR clearly more effective than waitlist and routine-care controls for post-traumatic stress disorder, and a Cochrane review of psychological therapies for chronic PTSD placed it among the treatments with the best support (Bisson et al., 2013). A meta-analysis of randomized trials confirmed medium-to-large reductions in PTSD symptoms and, in several comparisons, an advantage over some control conditions on speed of response (Chen et al., 2014).

That evidence has carried EMDR into first-line status. The World Health Organization's guidelines for conditions specifically related to stress recommend EMDR alongside trauma-focused cognitive behavioral therapy as a treatment for PTSD in adults and children, an endorsement that marks the method's move from contested novelty to guideline standard (WHO, 2013). Reviews have also examined its use beyond PTSD, in conditions from anxiety to depression to chronic pain, where the trial base is thinner and the conclusions more tentative but the direction of effect is generally favorable (Valiente-Gomez et al., 2017).

The harder question is the eye movements, and here the literature genuinely disagrees. An early and influential meta-analysis concluded that adding eye movements to the procedure produced no measurable benefit, implying that EMDR worked through its exposure component and that its signature feature was inert (Davidson & Parker, 2001). A later meta-analysis reached the opposite conclusion, finding a moderate additive effect of the eye movements, larger in the laboratory studies of memory vividness than in the clinical trials, and argued that the earlier null reflected underpowered dismantling designs rather than a true absence of effect (Lee & Cuijpers, 2013). The disagreement is not fully resolved, and it maps onto the two levels of the theory: the clinical efficacy of the package is settled, while the causal role of the eye movements — the very thing the working-memory account exists to explain — remains an open empirical question (Cuijpers et al., 2020).

EMDR in Motion

The three demonstrations below make the method and its central dispute manipulable. The first runs the working-memory account, showing how the load of a concurrent task degrades a recalled image's vividness along the inverted-U the theory predicts. The second steps through the eight-phase protocol, tracking the distress and belief ratings across a course of desensitization sets. The third makes the evidence base concrete, translating a meta-analytic effect size into the share of comparison patients the average treated patient exceeds, and showing how that figure collapses as the comparator is made more demanding.

Working-memory taxation: why a moderate load works best

Taxation = 4 × load × (1 − load) = 1.00, so the degraded vividness is 80 × (1 − 0.5 × 1.00) = 40. A load near 0.5 competes most with the held image; a load too low barely touches it and one too high would crowd it out, the inverted-U the working-memory account predicts. Values are illustrative.

The working-memory demonstration puts the eye-movement mechanism under the reader's hand. Setting the vividness of a recalled memory and then varying the load of the concurrent task computes the degraded vividness the working-memory account predicts, and the emotional intensity that is assumed to track it. It makes the inverted-U concrete: too little load barely touches the image, too much would crowd it out, and an intermediate load produces the largest reduction — the calibrated competition the therapist's moving hand is meant to supply.

Walk the eight-phase protocol

Phase 1: History and planning

Take history, select target memories, and judge readiness for reprocessing.

The reprocessing EMDR is known for is phase four, one bounded stage inside a longer structure. Running successive sets drives the distress rating down and the belief rating up, the pattern a course of desensitization is designed to produce.

The protocol demonstration walks through the eight phases in order. Advancing through the sequence shows what each phase does and, once desensitization begins, tracks the Subjective Units of Distress falling toward zero and the Validity of Cognition rising toward seven across successive sets. It shows that the reprocessing the method is known for is one bounded stage inside a longer structure of preparation, consolidation, and review.

Effect size by comparison condition

EMDR vs untreated waitlist: illustrative d = 1.00, so the average treated patient does better than 84.1% of the comparison group.

The treatment is identical across the three buttons; only the comparison changes. Against an untreated waitlist the effect looks large; against another active trauma therapy it shrinks toward the 50% mark that means no difference; and the dismantling contrast of eye movements against eyes held still gives the contested moderate value one meta-analysis found and another did not.

The efficacy demonstration makes the evidence base tangible and its interpretation cautious. Choosing the comparison condition — an untreated waitlist, another active trauma-focused therapy, or the dismantling contrast of eye movements against eyes held still — sets a standardized effect size and converts it into the percentage of comparison patients the average treated patient outperforms. It shows both why EMDR is called effective and why the size of that claim depends entirely on what it is measured against.

Worked Example

Begin with the working-memory account the first demonstration builds. A patient brings a memory to mind and rates its vividness at 80 out of 100. The concurrent task is set to a moderate load of 0.5 on a 0-to-1 scale. Model the competition for the visuospatial store as an inverted-U in load, taxation = 4 × load × (1 − load), which peaks at 1.0 when load is 0.5 and falls to 0 at either extreme: here taxation = 4 × 0.5 × 0.5 = 1.0, the maximum. Let the degraded vividness be the original scaled by the taxation, vividness_after = 80 × (1 − 0.5 × 1.0) = 80 × 0.5 = 40. The recalled image, held at 80, is degraded to 40 while the dual task runs. Emotional intensity is assumed to track vividness rather than to be addressed directly, so an initial emotion rated 70 is scaled by the same factor: 70 × (40 / 80) = 35. (The coefficients are chosen to show the arithmetic, not to prescribe how any real memory should be scored.)

Now shift the load to an extreme to see the inverted U. At a very low load of 0.1, taxation = 4 × 0.1 × 0.9 = 0.36, so vividness_after = 80 × (1 − 0.5 × 0.36) = 80 × 0.82 = 66 — the image is barely touched. At a very high load of 0.9, taxation = 4 × 0.9 × 0.1 = 0.36 again, giving the same 66: a task so demanding it crowds the memory out reprocesses it no better than one too easy to compete. Only the intermediate load produces the large drop to 40, which is the working-memory account's central and testable prediction.

Now the evidence base. A standardized effect size (Cohen's d) can be read as U3, the fraction of comparison patients the average treated patient exceeds, where U3 is the standard normal cumulative distribution evaluated at d. Against an untreated waitlist an illustrative d = 1.00 gives U3 = 0.841, so the average treated patient does better than about 84% of untreated controls — a large, headline effect. Against another active trauma-focused therapy an illustrative d = 0.05 gives U3 = 0.520, about 52%, barely above the coin-flip value of 50% that marks no difference. Against the dismantling contrast of eye movements versus eyes held still, an illustrative d = 0.41 gives U3 = 0.659, about 66% — the contested moderate effect that one meta-analysis found and another did not (Lee & Cuijpers, 2013). The treatment has not changed across the three rows; only the comparator has, which is why an effect size for EMDR is uninterpretable until the condition it was measured against is named.

Discussion

EMDR occupies a peculiar dual status: it is an established, guideline-recommended treatment whose signature ingredient is still of disputed value. On the clinical question the verdict is settled and favorable. The trial base for EMDR in post-traumatic stress disorder is substantial and consistent, it is recommended by the World Health Organization and other guideline bodies, and it takes its place among the trauma-focused therapies with the strongest support (Bisson et al., 2013). A patient offered EMDR is being offered an evidence-based treatment, and nothing in the mechanistic debate undercuts that.

The unresolved question is what the eye movements contribute. Here the adaptive information processing model and the working-memory account do different jobs. The first is a clinical heuristic, useful for organizing practice but too broad to specify a mechanism; the second is a genuine cognitive-psychological hypothesis, and its central prediction — that taxing visuospatial working memory during recall degrades a memory's vividness and emotionality along an inverted-U in load — has drawn real laboratory support (van den Hout & Engelhard, 2012). Yet the step from that laboratory effect to the clinical outcome is exactly where the dismantling meta-analyses disagree, one finding the eye movements inert and another finding them moderately additive (Davidson & Parker, 2001); (Lee & Cuijpers, 2013).

The wider interest for cognitive psychology lies in that gap. EMDR is a rare case in which a clinical technique and a basic-science model of memory make contact, so that a laboratory finding about working memory bears directly on a treatment recommendation and a treatment's success would count as evidence about the architecture of memory. Whether the eye movements ultimately prove to be an active ingredient or an inert companion to exposure, the effort to settle the question has produced some of the cleanest experimental work linking a psychotherapy to a cognitive model of the mind (Landin-Romero et al., 2018).

Current Directions

The most active line of current work extends the dismantling debate from a yes-or-no question into a search for boundary conditions: when, for whom, and under what task parameters the eye movements add to outcome. Meta-analytic reviews now separate the strong and replicable laboratory effect on memory vividness from the smaller and noisier clinical effect, and treat the discrepancy between them as the phenomenon to be explained rather than an embarrassment to be argued away (Lee & Cuijpers, 2013). This program is the direct descendant of the earliest methodological criticism, converted from a complaint into an experimental agenda.

A second direction is the broadening of the working-memory account beyond eye movements. If the mechanism is competition for a limited resource rather than anything specific to lateral eye motion, then other taxing tasks — counting, tapping, drawing — should reproduce the effect, and mapping which tasks work and how strongly has become a way of testing the theory against an eye-movement-specific alternative and of freeing the clinical procedure from the particular stimulus Shapiro happened upon (van den Hout & Engelhard, 2012).

A third strand concerns reach and delivery. Trials have extended EMDR beyond post-traumatic stress disorder to a range of conditions in which distressing memories or images figure, and have tested intensive and group formats that compress or scale the standard protocol, widening access while raising fresh questions about how much of the method's effect survives changes to its structure (Valiente-Gomez et al., 2017); (de Jongh et al., 2019). Across all three strands the common movement is from asking whether EMDR helps toward specifying the process through which its most distinctive component is supposed to help (Cuijpers et al., 2020).

Common Misconceptions

The eye movements hypnotize the patient.

EMDR is not hypnosis, and its dual-attention structure is in some ways its opposite: the patient stays fully alert and oriented, aware of both the memory and the present room, and the concurrent task divides attention rather than narrowing it into a trance (Shapiro, 2018).

EMDR erases or removes traumatic memories.

The memory remains; what changes is its vividness and emotional charge. On the working-memory account the recalled image is degraded and re-stored in a less intense form, so the event can be recalled without acute distress, not forgotten (van den Hout & Engelhard, 2012).

Because EMDR works, the eye movements must be doing the work.

The efficacy of the whole treatment does not by itself establish that its signature component is the active ingredient; dismantling meta-analyses have reached opposing conclusions about whether the eye movements add anything beyond the exposure the procedure also contains (Davidson & Parker, 2001).

It is a quick fix that resolves trauma in one session.

Although early reports emphasized rapid single-session gains, standard practice embeds the reprocessing in an eight-phase course of history-taking, preparation, consolidation, and review, and complex or multiple traumas typically require many sessions (de Jongh et al., 2019).

Glossary

Adaptive information processing model.

Shapiro's clinical theory that trauma is stored as an unprocessed, state-specific memory and that EMDR resumes the innate processing that integrates it with existing networks.

Bilateral stimulation.

The alternating left-right sensory input — eye movements, hand taps, or tones — paired with memory recall in EMDR; the concurrent task the method is built around.

Cohen's d.

A standardized effect size expressing the difference between two group means in units of pooled standard deviation, used to summarize how much a treatment outperforms its comparator.

Desensitization.

The reduction of a conditioned emotional reaction through graded, controlled contact with what provokes it; the broader family of treatments under which MeSH classifies EMDR.

Dismantling study.

A trial design that adds or removes a single treatment component — here the eye movements — to test whether it contributes to outcome beyond the rest of the procedure.

Dual attention.

The condition of attending simultaneously to a recalled memory and a concurrent task, the structural feature that distinguishes EMDR from focused exposure.

Exposure therapy.

A trauma treatment in which the patient concentrates fully on the feared memory or situation until distress subsides; the focused-attention procedure against which EMDR's divided-attention structure is contrasted.

Eye movement desensitization and reprocessing.

A structured eight-phase psychotherapy for trauma that pairs recall of a distressing memory with sets of guided eye movements or another bilateral stimulus.

Post-traumatic stress disorder.

The trauma-related condition, marked by intrusive memories, avoidance, and hyperarousal, that is EMDR's primary and best-supported indication.

Reprocessing.

The shift in a memory's content, meaning, and emotional charge that occurs across successive desensitization sets, from which the method takes the second half of its name.

Subjective Units of Distress scale.

A 0-to-10 self-report rating of how disturbing a target memory feels, tracked across sets and expected to fall toward zero during desensitization.

Validity of Cognition scale.

A 1-to-7 self-report rating of how true a preferred positive belief feels, strengthened during the installation phase.

Visuospatial sketchpad.

The limited-capacity component of Baddeley's working-memory model that holds and manipulates mental images, the shared resource the eye movements are said to tax.

Working-memory account.

The cognitive theory that eye movements reduce a memory's vividness by competing with recall for a limited visuospatial resource, degrading the image that is then re-stored.

Key Researchers

Jackie Andrade. University of Plymouth psychologist and co-author of the 1997 working-memory account of EMDR, the paper that first explained the eye movements as a taxation of visuospatial working memory. ORCID - Wikipedia - Wikidata - Google Scholar - Faculty page

Pim Cuijpers. Emeritus professor of clinical psychology at Vrije Universiteit Amsterdam whose meta-analyses supply much of the field's quantitative evidence base, including the dismantling analysis with Lee and a broad review of EMDR across disorders. ORCID - Wikidata - Google Scholar - Faculty page

Iris M. Engelhard. Professor of clinical psychology at Utrecht University who leads the Experimental Psychopathology Lab and co-developed the working-memory theory of EMDR with van den Hout. Faculty page

Marcel A. van den Hout. Emeritus professor of clinical psychology at Utrecht University whose experimental work with Engelhard established the working-memory taxation account of why recall plus eye movements degrades a memory's vividness and emotionality. Wikipedia - Wikidata - Faculty page

Ad de Jongh. Professor at the University of Amsterdam and prominent EMDR researcher and trainer whose intensive-treatment and mechanism studies helped establish the therapy as a first-line PTSD treatment. ORCID - Wikipedia - Wikidata - Google Scholar

Christopher W. Lee. Clinical psychologist and adjunct associate professor at the University of Western Australia whose meta-analyses and process studies address whether the eye-movement component adds to EMDR's effect. ORCID - Google Scholar - Faculty page

Francine Shapiro. Senior Research Fellow at the Mental Research Institute, Palo Alto, who originated EMDR in 1987 and formulated the adaptive information processing model that frames its theory. Wikipedia - Wikidata

Frequently Asked Questions

What is eye movement desensitization and reprocessing?
It is a structured eight-phase psychotherapy for trauma in which a patient recalls a distressing memory while performing sets of guided eye movements or another bilateral task, repeated until the memory can be brought to mind without acute distress (Shapiro, 1989a).

What is EMDR used to treat?
Its primary and best-supported indication is post-traumatic stress disorder, for which it is a first-line treatment; it has also been applied more tentatively to anxiety, depression, and other conditions in which distressing memories figure (Bisson et al., 2013).

Why does EMDR use eye movements?
The leading cognitive explanation is that the eye movements tax a limited visuospatial working-memory resource that recalling a vivid image also needs, so competing for it degrades the memory's vividness and emotional intensity (Andrade et al., 1997).

Do the eye movements actually matter?
This is genuinely contested: one influential meta-analysis found that adding eye movements produced no benefit, while a later one found a moderate additive effect, so their specific contribution remains an open question even though the whole treatment works (Lee & Cuijpers, 2013).

Is EMDR an evidence-based treatment?
Yes. Controlled trials and reviews find it effective for PTSD, and the World Health Organization recommends it alongside trauma-focused cognitive behavioral therapy as a first-line treatment (WHO, 2013).

What is the adaptive information processing model?
It is Shapiro's clinical theory that trauma is stored as an unprocessed, state-specific memory and that EMDR resumes the mind's natural processing so the memory is integrated with existing networks and loses its charge (Shapiro, 2018).

How is EMDR different from ordinary exposure therapy?
Exposure asks the patient to concentrate fully on the trauma memory, whereas EMDR deliberately divides attention between the memory and a concurrent task, and it is that divided-attention structure that its mechanistic debate turns on (van den Hout & Engelhard, 2012).

Does EMDR erase traumatic memories?
No. The memory remains accessible; what changes is its vividness and emotional intensity, so the event can be recalled calmly rather than being forgotten or removed (Gunter & Bodner, 2008).

References