Is NLP a Pseudoscience? An Honest Answer

Is NLP a pseudoscience? By the standard definition, yes, in its explanatory layer. A pseudoscience is a body of claims that borrows the vocabulary and authority of science without the testing, and that keeps teaching its claims after the results come back negative. Neuro-Linguistic Programming made several testable assertions about how the brain processes information, those assertions were tested and were not confirmed, and much of the field went on selling them anyway. That is exactly the pattern the label describes.

The second half of the honest answer matters just as much. The label is aimed at a claim to scientific status, and says nothing about whether a week of practising rapport, precise questions and state management leaves you better at conversations. Below: what the term requires, which NLP claims failed, where the label is applied too broadly, and what it should change if you are considering a training.

Question	Short answer
Does NLP meet the definition of pseudoscience?	In its explanatory claims, yes. Testable assertions were made, tested, not confirmed, and not withdrawn
Who applies the label?	Wikipedia’s article on NLP, and reviewers including Michael Heap (1988) and Tomasz Witkowski (2010)
Does that make every NLP technique worthless?	No. Several overlap with practices that carry solid evidence in other fields, gathered by those fields
Is NLP a therapy?	No. It is not a clinical qualification and should not be treated as a substitute for one
Does the label matter when choosing a training?	Yes, mostly as a test. How a trainer answers the question tells you most of what you need to know

What does “pseudoscience” actually mean?

It is not a synonym for “wrong”, and it is not a synonym for “unproven”. Plenty of ideas are unproven and perfectly respectable; they are simply young, or hard to test. Pseudoscience describes a specific failure pattern rather than a specific error.

Three markers are usually enough to identify it.

• Scientific costume without scientific method. Technical vocabulary, brain talk and diagrams are used to borrow credibility that no study supplied.
• Claims that avoid testing. Either they are framed so loosely that no result could count against them, or they are simply never put in front of a control group.
• No correction after disconfirmation. This is the decisive one. Sciences change their teaching when results contradict it. A pseudoscience keeps the syllabus and changes the subject.

Worth saying that philosophers of science have argued for decades about where exactly the line sits. Karl Popper proposed falsifiability as the test, and Larry Laudan argued in 1983 that no single criterion has ever held up, which is why “pseudoscience” is better used as a description of behaviour than as a verdict handed down from a committee. The behaviour, in NLP’s case, is documented well enough that the argument over demarcation does not rescue it.

Which NLP claims failed testing?

Four in particular, and they are the four that made NLP famous. Each was specific enough to test, which is to NLP’s credit, and each came back negative, which is the part that was never absorbed.

• Preferred representational systems. The idea that each person has a dominant sensory channel you can detect from their word choice and then match to build rapport. Christopher Sharpley reviewed the studies in 1984 and again in 1987 in the Journal of Counseling Psychology and found no reliable support for either half of the claim.
• Eye accessing cues. The claim that eye direction reveals which internal channel someone is using, or whether they are lying. Richard Wiseman and colleagues tested the lie detection version across three studies in 2012 in PLOS ONE and found no relationship. Our glossary entry on representational systems in NLP explains what survives of the original observation.
• Learning styles. Not strictly NLP’s invention, but the visual, auditory and kinaesthetic model was carried into schools on NLP’s back. Harold Pashler, Mark McDaniel, Doug Rohrer and Robert Bjork reviewed it in 2008 and found almost no studies built to test it properly, and no support in those that were.
• Health outcomes generally. Jaqui Sturt and colleagues searched the literature for experimental studies of NLP applied to health in 2012, published in the British Journal of General Practice, and concluded there was little evidence of effectiveness. Tomasz Witkowski, analysing NLP’s own research database in 2010, found the majority of empirical results negative.

There is also the name. “Neuro” promises a neurological account, and no neurological account was ever produced. A model of how people talk and how they picture things is a useful model. Calling it neuroscience was a marketing decision, and it is the single largest reason academics reach for the word pseudoscience when NLP comes up.

Where is the label applied too broadly?

In two places. The first is the jump from “these claims failed” to “everything in the field is nonsense”. Most NLP techniques have never been tested at all, which makes them unevidenced rather than disproven. Those are different states, and treating them as identical is its own kind of sloppiness.

The second is a category error. NLP began as modelling: watching Fritz Perls, Virginia Satir and Milton Erickson closely and copying what they did. Copying an expert is how apprenticeships work, and it produces skill without producing a theory. Judged as a craft handed down from practitioner to practitioner, NLP is doing something ordinary and defensible. It only invited the pseudoscience charge by insisting it was doing something else. The pillar page on what NLP is and how the model works keeps that distinction visible.

None of that is a defence of the field’s behaviour. A craft that had stayed a craft would never have been in this argument. NLP claimed the authority of science, was checked, and did not update. The wider question of why the field keeps attracting this fight is worth its own read: see our piece on why NLP is so controversial.

Should the label stop you doing an NLP training?

Not automatically, but it should change what you think you are buying. Book an NLP training as a structured skills week and it can deliver. Book it as a therapy qualification, a scientific credential or a cure, and you have been sold something that does not exist.

Four questions separate the two, and you can ask all of them before you pay.

1. “What does NLP not do?” A trainer who cannot answer this quickly, in specifics, has never thought about the limits of their own subject.
2. “What do you make of the research?” You are listening for familiarity with the critical reviews, not for a rebuttal. Someone who has never read a critical paper about their field is not a reliable guide to it.
3. “Do you still teach eye accessing cues as fact?” The honest version is to teach it as history, name the negative results, and move on.
4. “How many people are in the room?” Correction while you practise is the entire product. It does not survive a group of sixty.

Those questions are also the reason our groups stay small. To see how that plays out in practice, the Society of NLP Practitioner certification we teach in English over seven days in Lisbon puts most of the week into supervised practice rather than demonstrations from the front.

My own position, after training more than 300 people: the label is largely earned, and arguing with it is a waste of everyone’s time. What I teach is a set of communication skills with a fifty year old sales story attached, and the skills survive the story being removed. Trainers who cannot say that out loud are the reason the question keeps getting asked.

Carolin Mallmann, Licensed Trainer of The Society of NLP, Lisbon

One limit is not negotiable. If you are dealing with anxiety, depression, trauma or any other clinical condition, NLP complements professional care and does not replace it. A Practitioner certificate is not a clinical qualification, and a trainer who implies otherwise has told you what kind of training it is.

Frequently asked questions

Keep exploring

• Why is NLP so controversial?, for the reputational history behind the label, including the founders’ own legal fights.
• Does NLP work? An honest look at the evidence, for the study by study version of what does and does not hold up.

---