Nobody asks to be issued the beliefs they run. They arrive from parents, teachers, bad Tuesdays and one too many films, get installed before the age where consent means anything, and then run for decades as if they were the laws of physics. Here is the liberating and slightly vertiginous consequence of the field's most famous idea: if the map is not the territory, then your beliefs are not descriptions of reality. They are edits. And edits can be chosen.
Walk into any NLP training and at some point you will be handed "the presuppositions": a list of statements the field treats as its working beliefs. They sound profound and are usually delivered with the gravity of scripture. The honest version is more useful than the scripture: presuppositions are not proven truths, they are working assumptions, bets you make so that you can act. Some are pragmatic and testable. Some are philosophy. Knowing which is which is part of learning NLP properly, and the same discernment applies to every belief you hold, not just these.
Where they come from
Bandler and Grinder borrowed freely. The most famous presupposition, "the map is not the territory", comes from Alfred Korzybski, a philosopher who popularised it in the 1930s. The feedback orientation comes from cybernetics and systems thinking. The modelling attitude comes from how Bandler and Grinder themselves worked: they watched excellent therapists and coded what they did. The presuppositions are not original research; they are the operating assumptions that made that coding useful. The guide to what NLP is covers that origin story in detail, and the map idea gets its own piece here.
The list, in plain language
Here are the presuppositions most commonly taught, with the plain translation under each. Treat the list as a toolkit, not a creed.
The map is not the territory. Your experience of the world is not the world. You respond to your internal map: beliefs, memories, filters. Two people can sit in the same meeting and come out with different maps of it. This one is the foundation of everything else, and it is more true than profound.
People respond to their map, not to reality. Consequence: if you want to change how someone responds, you work with their map, not with "the facts". Arguing with a map directly usually strengthens it.
The meaning of your communication is the response you get. Not your intention. If you meant encouragement and the other person felt attacked, the communication failed at the delivery point. This one is pragmatic, testable, and humbling in exactly the right way.
People make the best choice available to them at the time. Not "always the best choice by some absolute standard". Given their map, state and options, people choose what makes sense to them. This is a stance of curiosity rather than judgement. It is also, strictly speaking, unfalsifiable, so hold it as an attitude, not a law.
No failure, only feedback. Outcomes are information. A pitch that flopped tells you something specific about the pitch, if you look. This is a genuinely useful operating rule, and it has a limit: some outcomes are real failures, and pretending otherwise is denial. The useful version is "treat failure as data, not identity".
If you always do what you've always done, you'll always get what you've always got. A reminder that behaviour is the variable you can change. Pragmatic, testable, and mildly annoying in exactly the way good advice is.
Behind every behaviour there is a positive intention. The behaviour itself may be destructive; the intention behind it is usually protective (safety, belonging, comfort). This is philosophy, not fact. Used gently, it stops you from demonising people. Used carelessly, it excuses harm. Keep it as a question, not an answer: "what might this behaviour be trying to do for this person?"
People already have all the resources they need. The claim that no one is broken, just disconnected from resources they already possess. This is the most optimistic and the least provable item on the list. As a working bet it is often useful. As a literal claim about people with clinical conditions, it is wrong and dangerous, which is exactly why we teach scope and referral rather than cheerleading.
Mind and body are one system. States have physical signatures; posture, breath and tension feed thought as much as the reverse. Whatever the philosophical nuances, this one is practical and mostly borne out by how stress works.
Choice is better than no choice. More options, more flexibility. The "four pillars" of NLP are often taught as outcomes, rapport, sensory acuity and flexibility; this presupposition is the ancestor of the flexibility pillar.
What they are for
Presuppositions exist to change where your attention goes. If you assume "the meaning of communication is the response you get", you stop blaming the audience and start adjusting the delivery. If you assume "no failure, only feedback", you actually try things. Their value is behavioural: they make you more curious, less defensive, more willing to experiment.
That is also their danger. A presupposition is a tool, and tools have scope. The moment a list of working assumptions is treated as a set of truths about the universe, it stops being NLP and starts being a belief system. The field has done exactly that at various points in its history, and it is the main reason critics are right to be suspicious of parts of it.
The choice nobody told you was available
Now apply the list's own logic to the beliefs you did not choose. A belief is, structurally, a compression: evidence from a few cases, promoted by emotion, frozen into a rule about the world and about you. I am bad with numbers. People like me do not get taken seriously. If I relax, everything falls apart. Each one was installed by someone, dated to something, and has been running unopposed ever since.
If that is what a belief is, then the move is not affirmations, which are usually just new decorations on an old rule. The move is the audit: what is the evidence, dated? What has changed since installation? What does believing this cost, weekly? What would be a better working bet, given the current data? That is choosing a belief: not pretending, but re-filing an old compression as what it always was, a working assumption, and then testing the replacement the way you would test any new tool. The story you tell about yourself is the biggest one of these, and it usually needs this treatment first.
A warning label, in keeping with this site's habit: choosing beliefs is powerful precisely because beliefs filter evidence, which means a badly chosen belief can hide its own counter-evidence indefinitely. Choose from evidence, and keep the audit running. A belief that cannot survive examination was never serving you.
Our stance at the Institute
We teach the presuppositions as working hypotheses: adopt them, test them, drop the ones that fail. A presupposition that makes you worse at your job is not sacred. What is sacred is the attitude underneath them: curiosity, experimentation and a willingness to be wrong. That attitude is the actual foundation of NLP, and it is the first thing we train.
If you want to learn the method, not memorise the creed, see the Spirit of NLP programme. The free webinar is a no-pressure way to see how we teach it. For the rules-as-rules version, read the rules of the game nobody told you, and for what a better chosen belief does to the inner war, see when you're at war with yourself.