NLP, Science, and Pseudoscience

A small number of vocal individuals have orchestrated a campaign of negative PR against NLP by calling it “pseudoscience.” The definition of pseudoscience is something that attempts to gain the legitimacy of rigorous scientific inquiry and testing by feigning scientific credibility. Notably, there is no scientific criteria for what constitutes a “pseudoscience” — this term is primarily a pejorative word that indicates the opinion of the speaker of the subject matter. It is equivalent to saying “I think X is not legitimate, real, or authentic, and should not be taken seriously.”

Lots of Things That Work are Non-Science

Personally I think NLP is a non-science, and thus cannot be rightly called a pseudoscience. NLP practitioners, in their work with individuals, do not generally conduct rigorous research into the workings of human psychology in general, but operate out of a set of loose models, using whatever pragmatic approaches they come across to facilitate changes in unique individuals. A specific NLP method or psychological research finding might apply to many people, but if a particular approach doesn’t work for the client in front of me, my job is to be flexible until I find something that does work.

In addition, the methods of modeling in scientific psychology and NLP are different. Scientific psychology looks at many individuals and attempts to create generalizations that apply to most or all people, and throws out statistical outliers. NLP seeks out outliers, people who can do things radically better than the norm, and attempts to discover what they do differently inside that allows them to get those results so that those ways of thinking and behaving can be taught to others.

Not only is NLP a non-science, so is the practice of psychotherapy, psychiatry, and indeed medicine. Biological medicine is supposed to be based in a scientific understanding of human biology (but often corrupted by insurance company billing and direct marketing of prescription drugs to consumers), and medicine can be studied scientifically, but the practice of medicine itself is not science. Similarly, psychotherapy is supposedly based in a scientific understanding of human psychology (quite often it is not), and psychotherapy can be studied scientifically, but the practice of therapy itself is not science.

Both medicine and psychotherapy are professional services, and perhaps arts, but not sciences. Both are scientifically-informed to various degrees, and depending on the practitioners. The fact that medical doctors will sometimes prescribe antibiotics for what is likely a viral infection to please the patient is evidence enough that doctors are not scientists.

Or perhaps a clearer example, a fantastic IT or tech support person does not need to be a computer scientist, because tech support is not a science. When I worked in tech support, I had literally no idea why many of the solutions worked, including rebooting (something to do with caches or the BIOS maybe?), but they did, and I basically just kept trying stuff until it worked. Know-why is very different than know-how.

Pseudoscientific Auto Repair

To fix a car, a mechanic does not need to know a lot about physics, chemistry, or anything at all about Bayesian statistics, despite the fact that cars function precisely according to very well established models from physics and chemistry. Auto mechanics is not a science. Even if your local mechanic started calling his services “Automobile Physical Systems Repair for Homo Sapiens Locomotive Transport” it wouldn’t make it any more of a science, nor would it make fixing cars into a “pseudoscience” that should be dismissed as ineffective for copying the unnecessary jargon of scientific journal articles. The methods of fixing cars that work, they would continue to work even if people dismissed them as merely pseudoscience. Reality doesn’t care about PR.

NLP is similarly informed by sciences to various degrees but itself is not a science. As far as I know, nobody in the field of NLP claims that NLP is a science (despite the subtitle of Robert Dilts’ early book NLP, Volume 1: The Study of the Structure of Subject Experience — one can study something without claiming one’s study is rigorously scientific, and scientific psychology long ago abandoned introspection as a method although cognitive psychology may be bringing back some of that from the behaviorist turn). Of course, psychological and communication sciences are also “softer” than chemistry and physics and more subject to popular trends and politics (and the history of physics itself shows that even hard sciences are not immune to these non-scientific influences). The fact is that we don’t know all that much for sure about human beings, so it is best to be cautious in our claims, but also open-minded and curious enough to be willing to try out low-risk experiments.

What is True? What Works?

I am deeply, sincerely interested in both psychological science and pragmatic models and methods for helping people to make changes. When people claim that the models I and many other practitioners use and get great results with regularly are “pseudoscience,” what I would like is if those individuals would explain to me what models they think are more scientific and yet just as effective if not more effective, with as many practical tools for facilitating change as I have found in NLP. So far, the only response I have seen is the claim that Cognitive Behavioral Therapy is the only useful modality for change, a claim for which I have some doubts (and I’m not the only one, as the reputation of CBT has been shown to be inflated due to publication bias).

Please note however that I have taken rigorous college-level coursework in psychology at an accredited and competitive university, from introductory classes to developmental psychology, cognitive psych, statistics, and research methods (I was just a couple classes away from a double major in psychology when I ran out of money for school, graduating with a single degree in Philosophy). I am familiar with all of the major schools of psychotherapy, from CBT, REBT, ACT, psychoanalysis, hypnotherapy, humanistic, gestalt, etc. I also read psychology journal articles regularly, scored over 100% in my symbolic logic class, have studied philosophy of mind and philosophy of language, and am intimately familiar with the importance of skepticism, cognitive biases, and how science works in general. In other words, I’m not an astrologer or some woo-woo new age thinker here. If there is a better way to do things, or a more accurate model for understanding human beings, I’m all ears! In fact, I enjoy busting my own cherished beliefs about how humans work.

What is NLP, Anyway?

I have not encountered many ways in which the basic models I use in NLP (as taught by Steve and Connirae Andreas specifically) to facilitate personal changes are contrary to any established psychological science. The science I have read tends to confirm what has been taught in NLP for decades (and I regularly seek out things that challenge I believe to attempt to account for confirmation bias, which I acknowledge is always a possibility). I strive to keep an open and yet critical mind, and would rather revise the models of NLP when the science shows they are flawed than preserve them like a museum piece. I think this is completely in line with the spirit of NLP and that of science, to evolve one’s understanding as new data comes in.

On the other hand, NLP is taught very differently by different trainers, some of whom mix NLP with models which make claims contrary to that of established psychological science. Perhaps my experience is unusual in the field of NLP, because of the two individuals I’ve learned the most from, one has a PhD. in psychology and the other and M.A. in Chemistry (a “hard” science), and they tend to be skeptical of unfounded claims and magical thinking.

Some of the original founders and trainers made absurd claims however about what the excellent methods of NLP could do for people. Some NLP practitioners continue to do so, or combine NLP with nonsensical woo like aromatherapy, crystal healing, and “The Law of Attraction.” While this has done nothing positive to establish NLP as a science-based modality, it does not prove that NLP is pseudoscience but that there is no singular body of knowledge and practice known as NLP. Your car mechanic might play the lottery (“a tax on people who are bad at math”) but that doesn’t matter as long as he can fix your engine.

On every NLP forum I’ve been a part of there is an ongoing debate about what is NLP and what isn’t. Even the founders of NLP can’t agree on a definition, let alone the scope of what is included under the banner of NLP (there are hundreds of models and thousands of methods that people claim as NLP which are not necessarily all compatible). Some practitioners question the “presuppositions” of NLP, arguing that some or all are false, others say they are true, and still others counter that the presuppositions were never supposed to be statements of fact but just useful perceptual biases or injunctions to get people to notice certain things or adopt certain attitudes.

In any case, I support the idea of scrapping models that do not conform with the most well-established and rigorously tested models of scientific psychology without throwing out methods that work in clinical practice. As practitioners of change, we must respect the data at least as much as our own eyes and ears and the followup reports of our clients.

Hasn’t NLP Been “Debunked” Though?

I realize that it is dogma within skeptics’ circles that NLP is “discredited,” but I’m skeptical of that dogma. There are approximately 3 well-designed studies that discredit certain claims of a minority of NLP practitioners and trainers. Those claims are…

1. “People have a primary representational system – either visual, auditory, or kinesthetic – and they represent things primary through that representational system.”

1a. “Primary representational systems can be determined though either verbal predicates (e.g. ‘I see what you are saying’ = visual vs. ‘that rings true’ = auditory), or eye movements.”

2. “You can tell when someone is lying because they move their eyes up and to their right, which indicates visual constructed internal processing.”

Firstly, these are claims that no respectable NLP trainer, and no NLP trainer I’ve ever studied with of NLP makes, precisely because they are false and misleading!

I think people absolutely do process information internally using visual, auditory, and kinesthetic representational systems, but people definitely do not have a primary representational system, and certainly not one that is easily determined through predicates or eye movements. Tony Robbins basically teaches “primary representational systems,” but Tony Robbins also references the non-existent “Yale Goals Study.” He also long-ago renamed his version of NLP to NAC. And he also happens to be a conman…but that’s another story.

While we can learn to observe little “tells” that people have, there is no reliable way to tell when a person is lying, period. Even if we grant that looking up and to their right is an unconscious signal of visual constructed processing, it may be that a liar will keep their eyes steady in order to mask their tell, or they might look up and their left showing visual remembered processing which may be not a sign of truth-telling but remembering their version of the events which they came up with previously. Or they might have a “creative memory” like some people are “creative spellers,” a poor strategy for both memory and spelling which is to create anew what the scene looked like (or what the word looks like in their mind). Or one of a million other things. Or their eye accessing cues might be reversed left/right, which seems to be the case in some percentage of people. Or their cues might not fit the model at all, which also seems to be the case for a number of people.

Debunking straw man claims and then generalizing to many thousands of other models and methods of NLP is perhaps an easy way to survive the “publish or perish” pressures of academic research, but it is not good science.

I’ve read through quite a few of the “debunking” studies and the experimental design has often been laughably bad, like using a method that only works with phobias on anxieties or guilt which has a completely different internal structure. One study from Richard Wiseman gave evidence that NLP practitioners teach the “you can tell when someone is lying from their eye movements” idea from an obscure YouTube video with almost no views that I myself couldn’t find with a search of YouTube. (Ironically, Wiseman wrote a whole book about a method taught in NLP trainings, that of the act as-if principle.)

Part of the problem is that NLP trainings are far too short to train competence, let alone mastery. Skill development takes years, not days. Another part of the problem is that some of the founders of NLP were coke-addicted sociopaths who (in some ways rightly) were very publicly critical of psychiatry and psychological science, but also made absurd claims about what their (often very good) methods could do. Yes, I myself have facilitated helping people overcome phobias in minutes, but it doesn’t necessarily always go that easily by any means.

Funnily enough, one of the most well-researched and supported treatments for PTSD is EMDR. According to various anecdotes in the NLP community, NLP co-founder John Grinder made up EMDR on the spot for Francine Shapiro when she asked about a method that could work for a friend who had sexual abuse trauma. Shapiro later claimed that the method spontaneously came to her when out walking, or something like that. But people who were present at the conversation between her and Grinder say otherwise. Some people in NLP claim that both could be true — that Grinder gave her the idea and then she forgot about it and it came back to her but she didn’t realize where it had come from originally. Recently a similar thing happened with the “Reconsolidation Traumatic Memories” technique from NLP (aka the Fast Phobia Cure, aka the movie theater technique, aka the rewind technique). It seems that NLP works in the lab as long as you don’t call it NLP! ;)

There are a number of other eye movement methods in NLP that I think are actually more effective than EMDR. Connirae and Steve Andreas (whom I work for) came up with EMI (Eye Movement Integration) long before EMDR became popular. Andrew Austin came up with another acronym IEMT (Integral Eye Movement Therapy) which also works really well for a number of conditions. EMDR seems to rely on people being in a terrible feeling state for a long period of time, whereas both EMI and IEMT don’t, so that is a very significant advantage.

I’ve thought about going through every point of the “debunkers” and debunking the debunking, but that project would take hundreds of hours and probably not convince a single person. Or I could put those hundreds of hours into working on helping more people to make personal changes. I think that is a better use of my time. (Plus Richard M. Gray has already done it here.)

Think For Yourself

So when I have read persuasive essays masquerading as authoritative and encyclopedic arguing that NLP is a pseudoscience, I can’t help but recognize the irony. Persuasive essays are not science but mere rhetoric (even if they are published on Wikipedia — which is not a peer-reviewed scientific journal). There is no scientific criteria for determining whether something is or isn’t pseudoscientific. To claim something is pseudoscientific by citing poorly designed studies and opinions of “experts” as evidence, making sure that contradictory information is restricted from the conversation through political means, seems to me itself pseudoscientific.
Skeletor Wikipedia He-Man

In the mean time, practitioners of NLP who care both about helping people and about rigorous research are busy securing funding for scientific research into various specific methods of NLP (a single pilot study requires many hundreds of thousands of dollars). If anything, NLP methods are defined so clearly in a stepwise fashion and with a solutions-focus that they are more easily testable than other methods of personal change like gestalt, psychoanalysis, or CBT. I fully support this kind of research for both getting more positive PR for the field and for improving our ability to understand how to effectively help people.

If you’d like to try NLP, you can get a free 15-minute consultation with me by clicking this button:

Schedule Now


Here’s a long document that lists many references to studies that support models and methods of NLP, compiled by the fine folks behind the NLP Research and Recognition Project and the NLP Wiki.

And here’s a list of some key concepts in the scientific psychology of goals, compiled by me.

Also see Richard Bolstad’s excellent book RESOLVE: A New Model of Therapy which contains many journal article references.

And here’s an article from my friend Mike Bundrant at the iNLP Center on The Truth About the NLP Wikipedia Page.