VAK is the most famous and the most abused idea in NLP. Visual, auditory, kinesthetic: the claim that people process experience through three main sensory channels, that most people have a "preferred" channel, and that you can spot it and use it. The pop version of this idea, learning styles, has been tested and failed. The underlying observation, that people express their experience in sensory language and that you can work with that, remains useful. This post separates the two honestly.
What the model claims
The VAK idea started as a description, not a diagnosis. Bandler and Grinder noticed that people describe experience in sensory terms: "I see what you mean" (visual), "that rings a bell" (auditory), "I can't get a grip on this" (kinesthetic). Their model of representational systems said the brain encodes experience in sensory modalities, that people differ in which modality they favour, and that a skilled communicator can notice the favoured modality and join it: speak "seeing" to a visual person, "feeling" to a kinesthetic person. In its original form it was a rapport observation: match how the person builds their world.
Why the learning-styles version failed testing
Somewhere between NLP and the school system, VAK became "every child has a learning style: teach the visual kids with pictures, the auditory kids with sound, the kinesthetic kids with movement." That version got tested, and it failed.
The most influential review, by Pashler and colleagues in Psychological Science in the Public Interest in 2008, examined the evidence for matching instruction to learning styles and concluded that there is no adequate evidence base to justify putting learning-style assessments into general educational practice. A broader review by Coffield and colleagues in 2004 catalogued 71 different learning-style models and found most of them unevaluated. Decades of searching, and the core claim still lacks support: classifying people into sensory types and teaching to the type does not reliably improve learning.
Two things are worth being precise about. People do have preferences; you may genuinely prefer diagrams or discussion. But preferences are not the same as a learning advantage, and the evidence says the advantage does not reliably show up. And the idea of a stable dominant sensory channel has not held up; people are flexible and use all channels, context by context. The original NLP "preferred representational system" claim quietly fell over under testing, and honest NLP teaching says so.
What remains useful
Strip away the failed claim and a genuinely useful observation remains: sensory language is how people express and access experience, and attention to it is a real communication skill.
Pacing with sensory language. If a client says "this whole situation feels heavy", you can either correct them with facts or join their language: "heavy like what, and where do you feel it?" The second response meets them where they actually are. That is rapport, not brain science. You are matching their words, not proving their neurology.
Making communication vivid. Concrete sensory specifics are memorable. "Picture the finished report on your desk, hear the quiet, feel the weight of the folder" is instruction someone can actually simulate; "please complete the deliverables" is not. This works regardless of anyone's supposed "type".
Noticing sensory verbs in your own speech. Your own language is a mirror of how you are processing a problem. If everything in your head is "I can't see a way through this", you are stuck in a visual frame; switching to "let's get a feel for the options" can unstick thinking. Sensory flexibility is a thinking skill.
How to use it honestly
- Use sensory language to build rapport and clarity. Fine.
- Label children or colleagues as "visual" or "kinesthetic" and teach them accordingly. No; the evidence is against it.
- Claim you can "read" a person's thinking from their language or their eyes. No; that claim has failed testing too (see the eye accessing cues post).
- Ask what would make something clearer, and listen to the answer. Always.
The honest frame: VAK is a language awareness tool with a failed theory attached to it. You can keep the tool, drop the theory, and lose nothing. That is exactly how we treat it in our teaching, alongside the other core techniques.
Why honesty here matters
The learning-styles myth is a perfect example of why NLP's reputation is mixed. A useful observation got inflated into a diagnosis, sold to schools and companies, failed its tests, and left the whole field looking foolish. The fix is not to defend the inflation; it is to name it, drop it, and keep what works. An approach that can name its own failures is a discipline. One that cannot is a belief system.
If you want to learn the useful version, with the failed claims named in the syllabus, see the Spirit of NLP programme. The free webinar is a good first look.