Neuro-linguistic programming was assembled in the mid-1970s at the University of California, Santa Cruz, by Richard Bandler, then a student, and John Grinder, a linguistics professor. Their method was to model excellence: they studied the therapists Fritz Perls, Virginia Satir and Milton Erickson, extracted what they took to be the patterns underlying the therapists' effectiveness, and packaged them as a transferable system.
The system's core claims are specific and testable. That people have a preferred representational system — visual, auditory or kinesthetic — detectable in their vocabulary, and that matching it improves rapport and influence. That eye movements indicate which internal process is running, the model demolished in chapter 53. That anchoring a state to a touch or gesture allows it to be re-triggered. That specific language patterns produce reliable effects.
The claims have been tested, repeatedly, since the early 1980s. Reviews of the empirical literature — including Passmore and Rowson's evidence review and Tomasz Witkowski's analyses of the published research — find no support for the preferred representational system hypothesis, none for eye-accessing cues, and no evidence that matching predicates improves outcomes. Heap's earlier review of the 1980s literature reached the same conclusion. It is one of the more comprehensively falsified frameworks in applied psychology, and it is taught, right now, in sales organizations, coaching certifications, management courses, police training in some jurisdictions, and seduction manuals.
The interesting question is not whether NLP works. It is why practitioners sincerely report that it does.
The answer is that NLP contains real techniques wrapped in a false explanation. Mimicry increases liking — that is Chartrand and Bargh's chameleon effect from chapter 41, and it is well supported. Vague, generalizable statements are personalized by the listener, which is the Barnum effect from chapter 27. Confident authority framing raises compliance, per chapter 7. Attentive listening and matching someone's vocabulary do build rapport, for reasons that have nothing to do with representational systems. The Milton model in the next chapter is a genuine set of linguistic patterns whose effects are real even though the theory around them is not.
So the practitioner applies a bundle, observes an effect, and attributes the effect to the branded mechanism rather than to the generic ingredient. Since the branded mechanism is what was paid for, this attribution is not examined.
This makes NLP the clearest available example of a pattern that recurs throughout the influence industry, and it is the reason this chapter is in the guide. A commercial persuasion product typically consists of a small number of genuine effects, drawn from ordinary social psychology and available free, packaged with an elaborate proprietary theory that explains them incorrectly. The theory is what makes it sellable — a technique everyone already knows cannot be certified — and the theory is also what makes the whole thing unfalsifiable in practice, because a failure is attributed to incorrect application rather than to the model.
The audit is chapter 53's: what was the controlled test, who ran it, and what happened. For NLP the tests exist, and they were negative, forty years ago.
Cross-ref: Part 2 mirroring and rapport