Editor’s note: Gerald Rosen and Gerald Davison coined the term purple hat therapy as a metaphor for treatment packages such as eye movement desensitization and reprocessing (EMDR) that combine essential elements (cognitive and behavioral techniques) and nonessential elements (eye movements). Wikipedia now has a page dedicated to this concept.

In the 1980s, several novel psychotherapeutic techniques were proposed for the rapid cure of post-traumatic stress disorder (PTSD). At the forefront of these “power therapies” was Eye Movement Desensitization and Reprocessing (EMDR), a method developed by Francine Shapiro (1948–2019) and still commonly used today. Shapiro reported a 100 percent success rate treating trauma memories with multi-saccadic eye movements, and she assured clinicians who read her report that they could use eye movements to “achieve complete desensitization of 75–80% of any individually treated trauma-related memory in a single 50-minute session” (Shapiro 1989, 221). Shapiro then offered workshops that introduced her treatment to thousands of clinicians. Over time, EMDR came to enjoy the sun of scientific endorsement.

We offer a less than sanguine view of EMDR, believing that Shapiro’s claims and methods often took on the imprimatur of science without meeting the required demand of Carl Sagan (1996), namely that extraordinary claims require extraordinary evidence.

 

Questionable Origins

How a narrative begins influences what we subsequently come to understand, and thus we consider how EMDR first emerged. According to Shapiro, the therapeutic powers of rapid eye movements were discovered in May 1987, while she was walking in a park.

The effect of saccadic eye movements was discovered accidentally by the author when she noticed that recurring, disturbing thoughts were suddenly disappearing and not returning. Careful self-examination ascertained that the apparent reason for this effect was that the eyes were automatically moving in a multi-saccadic manner while the disturbing thought was being held in consciousness. … The author then began to make systematic use of these movements to study the effect and later proceeded to generate the saccades in a large number of volunteers and clients. (Shapiro 1989, 201)

Shapiro’s account is inherently suspect, because people are generally oblivious of eye movements, particularly saccadic movements. Any such awareness is almost exclusively limited to situations of voluntary intention to shift the eyes from one point to another (Brindley and Merton 1960). There also is a deeper problem if one stops to ask why Shapiro focused on eye movements in the first place; after all, feelings of calm can be evoked by walking in a park because of the fresh air, flowers, the benefits of outdoor exercise, and being away from the bustle of life. As it turns out, Shapiro already had a reason to be thinking about eye movements: prior to her now-famous walk, she was a devotee of neurolinguistic programming (NLP). In a 1985 interview (Bonasia 1985), Shapiro expressed the view that NLP was “the most powerful vehicle for personal and social change in existence.” In an article, Shapiro (1985) advanced the idea that NLP’s “super-achievers” technologies empowered people to tap into unlimited sources of personal power so they could, for example, walk across burning coals.¹ Shapiro’s article also made clear that she understood the high importance that NLP theory placed on eye movement patterns:

One of the findings of the Neuro-Linguistic Programming research is that all people cross-culturally (with the exception of the Basque nationality) show how they are thinking by the way their eyes move … Even without their saying a word, if you watch their eyes carefully, you can determine whether they are seeing a picture, hearing, or feeling something. As a further refinement, you can tell if they are remembering something or constructing it.

Primed by her acceptance of NLP theory, Shapiro was ready to advance a belief in the therapeutic powers of eye movements. At the same time, she avoided acknowledging this influence, perhaps wanting to distance herself from those who viewed NLP as pseudoscience or perhaps hoping to obscure her intellectual indebtedness to others.²

 

Shapiro’s Dissertation

Shapiro applied her claimed discovery of eye movement patterns to a doctoral dissertation that she completed in 1988. Here it is useful to provide more context to appreciate Shapiro’s circumstances and her thoughts about science. From 1981 to 1988, she was enrolled at the Professional School of Psychological Studies, an unaccredited program in San Diego described as a diploma mill (Carroll 2004). During those seven years of working on her advanced degree, Shapiro launched New Age Health Services to sell wellness products, the Human Development Institute through which she offered NLP workshops, and MetaVox, whose lofty mission was to recruit the country’s most powerful speakers. In her 1985 article, Shapiro promoted her NLP workshops, enthusiastically writing that “In NLP, the key is that since people share the same neurological system, responses are predictable, verifiable, and repeatable. In other words, Neuro-Linguistic Programming is scientifically rather than merely theoretically based.”

Bonasia’s 1985 newspaper article quoted Shapiro as claiming: “NLP is absolutely the most effective because of its wide range of applications.” Shapiro conflated theory with scientific evidence and advanced the incorrect notion that the scope of problems to which a method is applied provides a measure of its effectiveness. In actuality, methods whose claimed cures have no boundaries have historically been associated with quackery instead of efficacy (see, e.g., Lawrence 1910; Walsh 1923).

As to the question of NLP being scientifically based, the National Academy of Sciences examined the potential value of these emerging claims. In their analysis, the Committee paid considerable attention to the fact that NLP had many enthusiastic supporters. They noted that people are highly influenced by testimonials of success, but following the scientific method is essential for winnowing the real from the illusory. After reviewing decades of human enhancement research and their own studies, they concluded that “there is little or no empirical evidence to date to support either NLP assumptions or NLP effectiveness” (Druckman and Swets 1988, 142).

In the context of her business enterprises and academic pursuits, Shapiro conducted a dissertation study in which twenty-two patients diagnosed with PTSD received the newly discovered eye movement procedure or a control condition while imagining a trauma related memory (Shapiro 1988). Change was measured with pre- and post- ratings of subjective stress. Shapiro reported that 100 percent of participants who engaged in eye movements were completely desensitized to the targeted trauma memory, whereas control subjects reported little improvement or increased stress. Shapiro’s study was so poorly designed that one is left wondering if it would have qualified her for a PhD at an accredited university. One major flaw concerned allegiance effects, namely that innovators are biased to support their own work, which can influence outcomes. Shapiro’s dissertation was exemplary in illustrating this intrusion. Entirely on her own, Shapiro conducted the assessment interviews, made the group assignments, administered the procedures (some of which were at her own residence), and quantified the pre- and post- subjective ratings. The control group in Shapiro’s study checked all the boxes discussed by Cuijpers and Cristea (2016) in their article “How to Prove That Your Therapy Is Effective, Even When It Is Not: A Guideline.” Participants in the treatment group engaged in sets of eye movements while imagining a trauma image, holding a positive cognition, and taking deep breaths. Participants in the control condition engaged in a frank discussion of the traumatic event absent any other treatment component or provision of support. Perhaps not surprisingly, six of the eleven control subjects experienced an increase in distress after recounting their traumatic experience.

Shapiro believed she had conducted a well-designed randomized controlled trial (RCT), and she submitted her study to the Journal of Traumatic Stress. The primary reviewer pointed out various flaws to the journal’s editor, who repeatedly asked Shapiro for revision and the reviewer for reconsideration (Anonymous personal communications, January 19, 2021). Eventually, the reviewer “just gave up,” and Shapiro’s paper was published (Shapiro 1989), fatal flaws and all. In her discussion of her findings, Shapiro rejected the influence of experimenter bias based on “the sheer magnitude” of observed effects. An equally reasonable conclusion would be that the sheer magnitude of reported findings was the very proof of extraordinary experimenter bias and demand effects. This second explanation is particularly compelling considering a recent meta-analysis that questions if eye movements contribute anything to treatment outcome beyond allegiances and placebo (Cuijpers et al. 2020).

 

Shapiro’s Responses to Negative Findings

A pattern emerged of Shapiro rejecting clinical and research findings as multiple publications failed to support her extraordinary claims of success. When several case failures were reported (Metter and Michelson 1993), Shapiro’s response was to blame the clinicians for not having obtained required levels of training. She cautioned, “clients can be at risk if untrained clinicians attempt EMDR” (Shapiro 1993, 420) and revised her earlier assurances that clinicians could succeed merely by reading her 1989 article. Shapiro’s stance was particularly amusing given that two of the failed cases reported by Metter and Michelson involved her own demonstrations at a training seminar.

A year later, Jensen (1994) found that EMDR was not effective in treating veterans diagnosed with PTSD, after which Shapiro criticized him for being half-trained. As explained by Shapiro (1996), Jensen had attended a Level I workshop but not Level II. Shapiro observed: “Such a lack of concern for standards of validity—both by the researcher and the published research reviewers—do a great disservice to the profession.” When leveling this attack, Shapiro failed to take into account that at the time of Jensen’s study, only one workshop was the standard: there were no Levels I and II. In fact, Shapiro had sent Jensen—and, we assume, other workshop participants—a congratulatory letter that thanked him for caring enough to take the specialized training. Here Shapiro had created an impossible situation for researchers: demanding treatment fidelity on the one hand while changing training requirements on the other (Rosen 1999).

Shapiro also changed procedures and theories, at times conveniently accommodating negative data. For example, Pitman et al. (1996) assessed the contribution of eye movements for the treatment of veterans with PTSD. Participants were assigned to standard EMDR or to an “eye-fixed” procedure that included finger tapping as a control for activity. Contrary to the central role Shapiro gave to eye movements, Pitman and his colleagues reported no significant between-group differences. At about the same time Pitman was collecting and analyzing his data, Shapiro (1994) began to propose that multiple forms of bilateral stimulation, rather than eye movements per se, accounted for EMDR’s effectiveness. Thus, therapists could snap fingers, tap on a patient’s knees, or employ bilateral audio stimuli, and they would still be practicing EMDR. In what represents a most curious twist of logic, Pitman et al. had simply compared EMDR to itself; eye movement desensitization and reprocessing no longer required eye movements!

 

Shapiro’s Search for an Underlying Mechanism

Shapiro strived to find a science-based explanation that could lend credence to the notion that eye movements were an active ingredient contributing to treatment outcome. At first, she emphasized the notion of adaptive information processing and the “opening of an information processing ‘block’ caused by neural pathology” (Shapiro 1988, 130). After proposing alternative forms of bilateral stimulation, Shapiro advanced the idea of interhemispheric communication. In a stinging rebuke of these theories, Lohr (1996) declared that Shapiro was advancing “little more than jargon” and “an embarrassing form of cognitive-neuro-pseudo-science.” As was the case with NLP, ill-conceived neuropsychological concepts did not assure scientifically sound theory.

A more recent attempt to explain the presumed therapeutic power of eye movements has focused on the hypothesis that dual tasks will tax working memory, thereby reducing the emotionality and vividness of imagery. While this explanation sounds science-based, the application of a working memory framework to EMDR rests on weak studies that have assessed nonclinical participants who focused on a single upsetting memory (e.g., Mertens et al. 2021). These analogue studies also yield inconsistent results. For example, Englehard et al. (2019) predicted that passive dual tasks that barely tax working memory (e.g., listening to beeps, finger tapping) should obtain fewer benefits than active tasks (e.g., eye movements). The authors cited publications in support of this prediction but ignored other studies that found no significant differences between eye movements and less taxing conditions (e.g., Merckelbach et al. 1994; Pitman et al. 1996). These findings are even more problematic when one considers that actual treatment studies with diagnosed patients have found that eyes-moving and eyes-fixed conditions yield comparable outcomes (e.g., Renfrey and Spates 1994; Sack et al. 2016). How can working memory possibly account for claimed EMDR treatment effects when equivalent outcomes are obtained with a nontaxing task?

 

Concluding Observations

Despite its controversial history, numerous studies supported EMDR, and it came to be recognized as an effective and evidence-based treatment (e.g., by the American Psychiatric Association and the International Society for Traumatic Stress Studies). These endorsements are viewed not as testament to the therapeutic powers of eye movements but as indicants of just how deficient current criteria are for defining what it means to be an evidence-based treatment. Here the core problem is that the profession of psychology remains vulnerable to anyone who claims a new method, trademarks an acronym, conducts the required number of studies—however weak their designs may be—and gains recognition for devising an accepted method. This scenario has become a reality as evidenced by an ever-expanding list of competing eye movement therapies, each with its own preferred pattern and claims of cure. In the spirit of perfecting how best to tax working memory, proponents of these methods have even gone so far as to shout “Whoosh!” after a set of eye movements. The antidote for this developing madness is a science-based framework in which the burden of proof falls on those who promote novel claims (Lilienfeld et al. 2018; Rosen et al. 2024).

EMDR appeared during a time when PTSD was a relatively new diagnosis, remission of post-traumatic symptoms was difficult to achieve, and enormous resources were being spent on treatment (Shalev et al. 1996). These circumstances likely combined with Shapiro’s strident efforts to promote eye movements as a rapid cure for PTSD. Looking at those efforts, it is almost as if Shapiro were applying the very NLP super-achiever technologies about which she had written back in 1985. Under the banner of a breakthrough discovery and pseudoscientific jargon, Shapiro packaged a nonessential novel technique (e.g., sets of bilateral stimulation) along with techniques that were already shown to be effective—exposure and imagery-based methods. Rosen and Davison (2003, 304–305) called this type of treatment package a “purple hat therapy.”

Hypothetically, a doctor could ask clients with driving phobias to wear a large purple hat while applying relaxation and cognitive coping skills to in vivo practice. The practitioner places a band of magnets in the purple hats, claiming that particular algorithms for positioning the magnets are determined by age, sex, and personality structure of the client. When properly placed, so the practitioner claims, the magnets reorient energy fields, accelerate information processing, improve interhemispheric coherence, and eliminate phobic avoidance. The inventor might call his method “purple hat therapy” (PHT) or “electro Magnetic Desensitization and Remobilization” (eMDR), conduct a single RCT against no treatment, and apply for listing as an [empirically supported treatment].

Shapiro’s purple hat therapy had great success. She then directed the narrative: she pushed extraordinary claims, shifted her theories, taught ever more complicated methods, and insisted on treatment fidelity that required higher levels of training. Whatever accounts for EMDR’s acceptance to this day, there is still no compelling reason nor science-based mechanism to justify therapists waving fingers in front of their patients’ eyes. The extraordinary evidence just isn’t there.

References

 

Bonasia, J. 1985. Success: Why it eludes some of us and how to obtain it. La Costan (January 10). Online at https://www.nlp.ch/pdfdocs/Historie_EMDR_Wingwave.pdf.

Brindley, G.S., and P.A. Merton. 1960. The absence of position sense in the human eye. Journal of Physiology 153: 127–130.

Carroll, R.T. 2004. Diploma mill. The Skeptic’s Dictionary. Online at https://skepdic.com/diplomamill.html.

Cuijpers, P., and I.A. Cristea. 2016. How to prove that your therapy is effective, even when it is not: A guideline. Epidemiology and Psychiatric Sciences 25: 428–435. Online at https://doi.org/10.1017/S2045796015000864.

Cuijpers, P., S.C. van Veen, M. Sijbrandij, et al. 2020. Eye movement desensitization and reprocessing for mental health problems: A systematic review and meta-analysis. Cognitive Behaviour Therapy 49: 165–180.

Druckman, D., and J.A. Swets (eds.). 1988. Enhancing Human Performance: Issues, Theories, and Techniques. National Academy Press.

Englehard, I.M., R.J. McNally, and K. van Schie. 2019. Retrieving and modifying traumatic memories: Recent research relevant to three controversies. Psychological Science 28: 91–96.

Jensen, J.A. 1994. An investigation of eye movement desensitization and reprocessing (EMD/R) as a treatment for posttraumatic stress disorder (PTSD) symptoms of Vietnam combat veterans. Behavior Therapy 25: 311–325.

Lawrence, R.M. 1910. Primitive Psycho-Therapy and Quackery. Boston, MA: Houghton Mifflin Company.

Lilienfeld, S.O., S.J. Lynn, and S.C. Bowden. 2018. Why evidence-based practice isn’t enough: A call for science-based practice. Behavior Therapist 41: 42–47.

Lohr, J.M. 1996. Analysis by analogy for the mental health clinician. Contemporary Psychology 41: 879–880.

Merckelbach, H., E. Hogervorst, M. Kampman, et al. 1994. Effects of “Eye Movement Desensitization” on emotional processing in normal subjects. Behavioural and Cognitive Psychotherapy 22: 331–335.

Mertens, G., M. Lund, and I.M. Englehard. 2021. The effectiveness of dual-task interventions for modulating emotional memories in the laboratory: A meta-analysis. Acta Psychologica 220: 1–14.

Metter, J., and L.K. Michelson. 1993. Theoretical, clinical, research, and ethical constraints of the eye movement desensitization reprocessing technique. Journal of Traumatic Stress 6: 413–415.

Pankratz, L. 1988. Fire walking and the persistence of charlatans. Perspectives in Biology and Medicine 31: 291–298.

Pitman, R.K., S.P. Orr, B. Altman, et al. 1996. Emotional processing during eye movement desensitization and reprocessing therapy of Vietnam veterans with chronic posttraumatic stress disorder. Comprehensive Psychiatry 37: 419–429.

Renfrey, G., and C.R. Spates. 1994. Eye movement desensitization: A partial dismantling study. Journal of Behavior Therapy and Experimental Psychiatry 25: 231–239.

Rosen, G.M. 1999. Treatment fidelity and research on Eye Movement Desensitization and Reprocessing (EMDR). Journal of Anxiety Disorders 13: 173–184.

———. 2023. Revisiting the origins of EMDR. Journal of Contemporary Psychotherapy 53: 289–296.

Rosen, G.M., and G.C. Davison. 2003. Psychology should list empirically supported principles of change (ESPs) and not credential trademarked therapies or other treatment packages. Behavior Modification 27(3): 300–312.

Rosen, G.M., W. van der Does, B.A. Gaudiano, et al. 2024. Commentary: EMDR, RCTs, and the proliferation of trademarked acronyms. Journal of Contemporary Psychotherapy. Online at https://doi.org/10.1007/s10879-023-09606-6.

Sack, M., S. Zehl, A. Otti, et al. 2016. A comparison of dual attention, eye movements, and exposure only during Eye Movement Desensitization and Reprocessing for posttraumatic stress disorder: Results from a randomized clinical trial. Psychotherapy and Psychosomatics 85(6): 357–365.

Sagan, C. 1996. The Demon-Haunted World: Science as a Candle in the Dark. New York, NY: Ballantine Books.

Shalev, A.Y., O. Bonne, S. and Eth. 1996. Treatment of posttraumatic stress disorder: A review. Psychosomatic Medicine 58: 165–82.

Shapiro, F. 1985. Neuro-Linguistic Programming: The new success technology. Holistic Life Magazine Summer: 4–43. Online at https://www.nlp.ch/pdfdocs/Historie_EMDR_Wingwave.pdf.

———. 1988. Efficacy of the Multi-Saccadic Movement Desensitization Technique in the Treatment of Post-Traumatic Stress Disorder. Ann Arbor, MI: UMI Dissertation Services.

———. 1989. Efficacy of the Eye Movement Desensitization procedure in treatment of traumatic memories. Journal of Traumatic Stress 2: 199–223.

———. 1993. Eye movement desensitization and reprocessing (EMDR) in 1992. Journal of Traumatic Stress 6: 417–421.

———. 1994. Alternative stimuli in the use of EMD(R). Journal of Behavior Therapy and Experimental Psychiatry 25: 89.

———. 1996. Errors of context and review of eye movement desensitization and reprocessing research. Journal of Behavior Therapy and Experimental Psychiatry 27: 1–5.

Walsh, J.J. 1923. Cures: The Story of the Cures That Fail. Boston, MA: D. Appleton & Co.

Notes

1. For a science-based explanation on how embers of a fire do not efficiently transfer heat, see Pankratz 1988.

2. For further analysis of this history, see Rosen 2023.