TIRCIA · EMDR
Working memory theory
Why eye movements soothe a memory: what the studies really show
By Marie-Agnès Thulliez · Author · Publishing Director11-minute read
Why would following the therapist's fingers with your eyes make a memory less painful? Of the explanations put forward, working memory theory is the only one to have been tested in detail in the laboratory. It is robust on some points and more fragile on others. An overview for therapists.
At a glance
Recalling a memory while performing a demanding task, such as following eye movements, makes it slightly less vivid and less emotionally charged. The more demanding the task, the stronger the effect. This effect, measured mainly in students, is modest and sometimes short-lived: it sheds light on part of how EMDR therapy works, while leaving the rest still to be explained.
1. The principle
Working memory is the mental space that allows us to hold and manipulate information for a few seconds, such as a phone number. Its capacity is limited (Baddeley and Hitch, 1974; Baddeley, 2000).
When the patient brings up a distressing memory while following eye movements, the two activities compete for these limited resources. The memory is then recalled less vividly and less intensely, and is thought to be "stored back" in this attenuated form (van den Hout and Engelhard, 2012).
This hypothesis complements Francine Shapiro's clinical model, described in our article The AIP model: it proposes a precise mechanism that can be tested experimentally.
2. What the experiments show
The first studies showed that eye movements reduce the vividness and emotional charge of unpleasant memories, and that other demanding tasks, such as counting backwards, produce a similar effect (Gunter and Bodner, 2008; van den Hout et al., 2010). What matters, then, is the load placed on working memory, more than the movement of the eyes itself.
A meta-analysis of 44 laboratory experiments confirms the effect, while showing that it remains limited: after a dual task, the vividness of a negative memory falls on average by about 11 points and its emotional charge by about 11 points, on a 100-point scale, compared with almost no change in vividness and about 4 points for emotional charge with recall alone (Mertens et al., 2021). The effect also applies to positive memories, which confirms that it is a general memory mechanism.
3. Eye movements, tones, tapping: all equal?
If the effect depends on the load placed on working memory, not all forms of stimulation are equivalent. In a series of experiments, alternating tones taxed working memory much less than eye movements did, and reduced the vividness of memories less (van den Hout et al., 2011).
A small study of 12 patients with PTSD, during a first EMDR session, points the same way: eye movements outperformed tones, and it was not even certain that tones added anything compared with simply recalling the memory (van den Hout et al., 2012). Tapping, for its part, has been studied too little to draw any conclusion.
Tones and tapping remain valuable when eye movements are impossible or poorly tolerated. In the light of these data, however, they are better seen as an alternative than as an equivalent.
4. Is there an "optimal" load?
It is often said that the effectiveness of EMDR depends on an "optimal" level of load: a task that is too easy would be ineffective, while one that is too difficult would prevent access to the memory. This "inverted U" hypothesis has not been confirmed. The studies that tested it found, on the contrary, that the more demanding the task, the greater the effect (van Veen et al., 2015; Littel and van Schie, 2019).
Should the speed then be adapted to each patient, according to their working memory capacity? One study tested this: slowing down the eye movements for people with low capacity brought no benefit, and fast movements remained the most effective for everyone (van Schie et al., 2016).
This is the idea that inspired EMDR 2.0, which makes the stimulation more demanding. The only published trial comparing it directly with standard EMDR, conducted in the laboratory with 62 volunteers without a diagnosis, found equivalent effectiveness with fewer sets, for a comparable session length; the authors see this as partial support for the efficiency hypothesis (Matthijssen et al., 2021). A trial in patients with PTSD is under way (the ENHANCE trial; Alting van Geusau et al., 2023).
5. Does the memory need to be kept vividly in mind?
In EMDR therapy, the patient is asked to "hold in mind" the image of the memory during stimulation. A recent study sought to strengthen this activation by showing participants an image taken from the distressing film they had watched. This visual reminder made no difference: all the conditions involving a dual task gave the same results, better than recall alone (Nuijs et al., 2025). The authors suggest that continuous, intense activation of the memory may not be necessary, which echoes the intuition behind the Flash technique.
And reconsolidation?
Working memory theory assumes that a recalled memory becomes modifiable again and is then re-stored in its attenuated form: this is the process of reconsolidation, well described in animals. In humans, and in EMDR in particular, this step remains a hypothesis: we cannot yet demonstrate that it is indeed this mechanism that makes the effects last.
6. The limitations
- Mainly non-clinical participants: the vast majority of experiments involve students, with memories that are unpleasant but not traumatic. A systematic review found only five studies conducted with people suffering from PTSD, often small and without a control group (Wadji et al., 2022).
- An effect that can fade: in one study, the benefits obtained with eye movements had partly disappeared 24 hours later, whereas those of simply recalling the memory were maintained (van Veen et al., 2020).
- Some negative results: after a distressing film, dual tasks did not reduce the vividness or unpleasantness of the images, and their effect on intrusions was not replicated (van Schie et al., 2019).
- Subjective measures: most studies rely on ratings that participants give themselves.
In clinical sessions, the data are more encouraging: a meta-analysis of clinical trials found that eye movements provide an additional, moderate benefit compared with recalling the memory alone (Lee and Cuijpers, 2013).
7. What this means in practice
Working memory theory provides a few useful pointers:
- The stimulation should genuinely engage the patient's attention while remaining manageable for them: the speed and length of the sets are adjusted to what they can follow.
- Tasks other than eye movements can be used when these are difficult or uncomfortable.
- To date, no study shows that finer adjustment of the stimulation, whether manual or digital, improves clinical outcomes.
Above all, this mechanism works alongside the rest of the therapy: assessment, preparation, choice of targets and the therapeutic relationship remain at the heart of EMDR work.
An example in session (fictional case)
Claire, 38, is working on a car accident. The target image is the windscreen shattering; her disturbance is at 8 out of 10. After a few sets, she says: "I can't follow your fingers any more, it's too fast." The therapist slows down slightly, without returning to a slow pace: the aim is for the task to remain demanding while still being manageable.
Later, Claire reports that the eye movements are giving her a headache. The therapist switches to tapping, knowing that it probably taxes working memory less, and monitors the change in disturbance closely.
This case illustrates how laboratory data can guide concrete choices. It is offered as an illustration, outside any validated protocol.
The main studies at a glance
| Study | Participants | Main finding |
|---|---|---|
| Gunter and Bodner, 2008 | Students | Eye movements and other tasks reduce vividness and emotionality |
| van den Hout et al., 2011 | Healthy volunteers | Tones tax working memory less and have less effect |
| van den Hout et al., 2012 | 12 patients with PTSD | Eye movements superior to tones |
| van Veen et al., 2015 | 72 students | Fast movements more effective; no "inverted U" |
| van Schie et al., 2016 | 66 students | Adapting speed to individual capacity does not help |
| Littel and van Schie, 2019 | 44 participants | The more demanding the task, the stronger the effect |
| van Veen et al., 2020 | 100 students | Effect of eye movements partly lost at 24 hours |
| Mertens et al., 2021 | Meta-analysis, 44 experiments | Real but modest effect |
| Matthijssen et al., 2021 | 62 volunteers | EMDR 2.0 not superior, but fewer sets |
| Wadji et al., 2022 | Review, 11 studies | Plausible hypothesis, few studies in patients |
| Nuijs et al., 2025 | 136 participants | Boosting memory activation adds nothing |
Frequently asked questions
Should I go as fast as possible?
Studies show that a more demanding task has a greater effect, but they involve volunteers and very short sessions. In clinical sessions, the speed should remain one the patient can follow and be compatible with their emotional tolerance.
Do tapping and tones work just as well?
Probably less well, according to the available data, which remain limited. They are a useful alternative when eye movements are not possible.
Should I add mental arithmetic, as in EMDR 2.0?
This reduces the number of sets needed, but no study has yet shown a better final outcome in patients. The results of the ENHANCE trial, which compares standard EMDR, EMDR 2.0 and the Flash technique in patients, are awaited (Alting van Geusau et al., 2023, Frontiers in Psychiatry, 14, 1278052, link).
Does working memory theory call the AIP model into question?
The two work at different levels: working memory theory describes a precise mechanism, while the AIP model provides an overall clinical framework. They complement each other.
What is established, and what is not yet
- Established in the laboratory: a dual task modestly reduces the vividness and emotional charge of a memory, all the more so when the task is demanding.
- Not confirmed: the existence of an "optimal" level of load.
- Yet to be confirmed: how long these effects last, and the exact contribution of this mechanism to the effectiveness of EMDR in patients.
In a nutshell
Working memory convincingly explains why bilateral stimulation softens a memory, though it cannot yet, on its own, explain why EMDR heals.
It is one important piece of the puzzle, among others.
References
- Baddeley, A. D., & Hitch, G. J. (1974). Working memory. In G. H. Bower (Ed.), The psychology of learning and motivation (Vol. 8, pp. 47-89). Academic Press.
- Baddeley, A. D. (2000). The episodic buffer: A new component of working memory? Trends in Cognitive Sciences, 4(11), 417-423. Link
- van den Hout, M. A., & Engelhard, I. M. (2012). How does EMDR work? Journal of Experimental Psychopathology, 3(5), 724-738. Link
- Gunter, R. W., & Bodner, G. E. (2008). How eye movements affect unpleasant memories: Support for a working-memory account. Behaviour Research and Therapy, 46(8), 913-931. Link
- van den Hout, M. A., Engelhard, I. M., Smeets, M. A. M., et al. (2010). Counting during recall: Taxing of working memory and reduced vividness and emotionality of negative memories. Applied Cognitive Psychology, 24(3), 303-311. Link
- van den Hout, M. A., Engelhard, I. M., Rijkeboer, M. M., et al. (2011). EMDR: Eye movements superior to beeps in taxing working memory and reducing vividness of recollections. Behaviour Research and Therapy, 49(2), 92-98. Link
- van den Hout, M. A., Rijkeboer, M. M., Engelhard, I. M., et al. (2012). Tones inferior to eye movements in the EMDR treatment of PTSD. Behaviour Research and Therapy, 50(5), 275-279. Link
- Mertens, G., Lund, M., & Engelhard, I. M. (2021). The effectiveness of dual-task interventions for modulating emotional memories in the laboratory: A meta-analysis. Acta Psychologica, 220, 103424. Link
- van Veen, S. C., van Schie, K., Wijngaards-de Meij, L. D. N. V., Littel, M., Engelhard, I. M., & van den Hout, M. A. (2015). Speed matters: Relationship between speed of eye movements and modification of aversive autobiographical memories. Frontiers in Psychiatry, 6, 45. Link
- van Schie, K., van Veen, S. C., Engelhard, I. M., Klugkist, I., & van den Hout, M. A. (2016). Blurring emotional memories using eye movements: Individual differences and speed of eye movements. European Journal of Psychotraumatology, 7, 29476. Link
- Littel, M., & van Schie, K. (2019). No evidence for the inverted U-curve: More demanding dual tasks cause stronger aversive memory degradation. Journal of Behavior Therapy and Experimental Psychiatry, 65, 101484. Link
- Matthijssen, S. J. M. A., et al. (2021). The effect of EMDR versus EMDR 2.0 on emotionality and vividness of aversive memories in a non-clinical sample. European Journal of Psychotraumatology. Link
- Wadji, D. L., Martin-Soelch, C., & Camos, V. (2022). Can working memory account for EMDR efficacy in PTSD? BMC Psychology, 10, 245. Link
- van Veen, S. C., van Schie, K., van de Schoot, R., van den Hout, M. A., & Engelhard, I. M. (2020). Making eye movements during imaginal exposure leads to short-lived memory effects compared to imaginal exposure alone. Journal of Behavior Therapy and Experimental Psychiatry, 67, 101466. Link
- van Schie, K., van Veen, S. C., & Hagenaars, M. A. (2019). The effects of dual-tasks on intrusive memories following analogue trauma. Behaviour Research and Therapy, 120, 103448. Link
- Alting van Geusau, V. V. P., de Jongh, A., Nuijs, M. D., Brouwers, T. C., Moerbeek, M., & Matthijssen, S. J. M. A. (2023). The effectiveness, efficiency, and acceptability of EMDR vs. EMDR 2.0 vs. the Flash technique in the treatment of patients with PTSD: Study protocol for the ENHANCE randomized controlled trial. Frontiers in Psychiatry, 14, 1278052. Link
- Nuijs, M. D., Theunissen, T. P. M., van Schie, K., Brouwers, T. C., de Cort, K., Schruers, K. R. J., & Matthijssen, S. J. M. A. (2025). Boosting memory activation to target aversive memories in dual-task interventions. European Journal of Psychotraumatology, 16(1). Link
- Lee, C. W., & Cuijpers, P. (2013). A meta-analysis of the contribution of eye movements in processing emotional memories. Journal of Behavior Therapy and Experimental Psychiatry, 44(2), 231-239. Link
This article summarises the scientific evidence available at the date of publication. It is no substitute for clinical training, supervision, or the advice of a health professional.
