Meta programs are the unconscious filters that decide how someone sorts information before they decide anything, and the clearest meta programs examples are not personality labels but phrases: “I want a team where nobody shouts” (away from), “I just know when it is good” (internal reference), “I followed the steps” (procedures). You hear the pattern in how an answer is built, not in what it is about. The glossary entry on meta programs in NLP covers where the idea came from; this article is the listening practice.
Below are the six patterns that carry the most weight in a real conversation, the question that brings each one out, and the sentence that gives it away. Then the part most articles skip: what a meta program cannot tell you about a person.
| What they are | Habits of sorting: what someone attends to before content enters the picture |
| Where you hear them | Sentence structure, verbs, what gets mentioned first |
| What they need | One open question and the patience to listen to the whole answer |
| What they are not | A type, a diagnosis, or a stable trait across every context |
| Best use | Phrasing your next sentence so it lands, instead of repeating yourself louder |
What is a meta program in NLP?
A meta program is a sorting habit that runs above content. Two people can want the same promotion and arrive at it from opposite directions: one talks about the influence it brings, the other about finally leaving a manager behind. Same goal, opposite filter, and anything you say to motivate them has to be phrased differently.
The idea came into NLP through Leslie Cameron-Bandler in the late 1970s and was turned into something testable by Rodger Bailey, whose LAB Profile set out which question reliably brings which pattern to the surface. Shelle Rose Charvet made that work readable for people outside the field in “Words That Change Minds”. If you want the research-shaped version, start with Bailey rather than with the quizzes.
One thing to hold on to from the start: a meta program describes behaviour in a context, not a person. Someone can run strongly toward goals at work and strongly away from risk in money. Asking “what is their meta program” is already the wrong question. The useful one is “which pattern is in the room right now”.
What do the six most useful meta programs sound like?
They sound like ordinary sentences, which is why they are easy to miss. The table gives the question that elicits each pattern, the phrasing that signals it, and the way to answer so the other person can actually use what you said.
| Meta program | Question that elicits it | What you hear | How to answer |
|---|---|---|---|
| Toward / away from | “What do you want in your next role?” | “More ownership” vs “No more firefighting” | Name the gain, or name the problem that disappears |
| Internal / external reference | “How do you know you did a good job?” | “I just know” vs “My clients tell me” | Ask for their judgement, or bring evidence from others |
| Options / procedures | “Why did you choose this one?” | Reasons and alternatives vs how they got there, in order | Offer a choice of three, or one clear sequence |
| Sameness / difference | “How is this year’s work related to last year’s?” | “Much the same” vs “Completely different” | Stress continuity, or stress what is new |
| Specific / general | “Tell me about the project.” | Dates and names vs the one-line purpose | Match their altitude before you change it |
| Proactive / reactive | “When you decide, do you act or study it first?” | Short sentences, active verbs vs conditionals and “we should look into” | Give a first step, or give time and context |
Two of these do most of the work in practice. Toward and away from changes every piece of motivation language you use, and internal versus external reference decides whether praise lands as support or as pressure. Telling someone with a strong internal reference that the whole team agrees with them can read as noise, or worse, as an attempt to outvote them.
How do you elicit a meta program without interrogating anyone?
Ask one open question, stay quiet, and read the shape of the answer rather than its content. Six to practise this week, one per day:
- “What is important to you about this?” The first three words usually reveal direction. Something wanted, or something avoided.
- “How do you know when you have done it well?” Watch for whether the evidence sits inside them or arrives from outside.
- “Why this one and not another?” Reasons point to options. A story about how they got there points to procedures.
- “What is the connection between this and what you did before?” Sameness, difference, or the sameness-with-exception answer that most people give.
- “Give me the picture in one sentence.” Then notice whether that was easy or whether they needed the detail first.
- “What happens next?” A proactive answer contains a verb and a date. A reactive one contains a condition.
Write down the exact words, not your interpretation of them. “She is away from” is already a conclusion; “no more firefighting” is evidence you can check tomorrow against a second sample. Pair this with the representational systems self-check and you have two independent readings of the same conversation, which is how you catch yourself hearing what you expected.
The mistake almost everyone makes in week one is asking the question and then answering it themselves, in their own pattern, before the other person has finished. The pattern you end up noting is your own. Carolin Mallmann, Licensed Trainer of The Society of NLP, Lisbon
What do you actually do with what you hear?
You change your own phrasing, and nothing else. That is the whole legitimate use. A reader who sorts away from problems does not need your vision statement, they need to know which specific mess goes away. A colleague with a procedures pattern asked to “be creative with the format” is not being difficult, they are being handed a task with no sequence in it.
Three places this pays off quickly. In writing, one sentence of each kind near the top catches both directions of filter. In hiring and delegation, a procedures pattern is excellent news for work that must be repeatable and bad news for an unbuilt role. In difficult conversations, matching chunk size stops the argument where one person keeps zooming out and the other keeps producing examples.
Where it stops being legitimate is clear enough to state as a rule: you use a pattern to make yourself understood, not to make a decision harder to refuse. Reading someone’s filter and then building a pitch around their fear is the version that earns NLP its reputation, and it is the reason these patterns sit inside a broader set of NLP techniques taught with an ecology check attached. The seven days of practitioner training in Lisbon spend more time on that boundary than on the patterns themselves, because the patterns take an afternoon and the judgement takes longer.
Where do meta programs get overstated?
At the point where a pattern becomes a label. There is no peer-reviewed instrument behind these categories, the published work sits in trade books rather than in validation studies, and the number of meta programs in circulation ranges from six to more than sixty depending on who is selling the training. That spread alone should tell you the taxonomy is a working tool and not a finding.
Three honest limits worth keeping in view. Patterns shift with context and with stress, so a reading taken in a tense meeting describes the meeting. One answer is one data point, and people produce mixed answers more often than clean ones. And the observer is part of the measurement: if you have just learned about away-from language, you will hear it everywhere for a fortnight.
None of that makes the practice worthless. It makes it a listening discipline with a short reach. The claim “this person is externally referenced” is overreach; the claim “in this conversation about her work, every piece of evidence she offered came from someone else” is a description you can act on without pretending to know her.
FAQ
How many meta programs are there?
There is no agreed number. Rodger Bailey’s LAB Profile works with fourteen categories, other trainings list six, and some list more than sixty. The useful answer is that six patterns cover most of what changes your phrasing in an everyday conversation, and learning those properly beats memorising a longer list badly.
Are meta programs the same as personality types?
No. A personality type claims something stable about a person across situations. A meta program describes how someone sorted information in one context, and the same person often sorts differently at work, at home and under pressure. Using them as types is the most common misuse.
What is the difference between meta programs and the meta model?
The meta model is a set of questions for recovering what a sentence left out, such as who exactly, compared to what. Meta programs are patterns in how people habitually filter before they speak. You often use meta model questions to elicit a meta program, which is why the two get confused.
Can you change your own meta programs?
You can widen them rather than swap them. Someone who filters strongly for difference can practise naming what stayed the same before listing what changed, and the practice holds while they keep doing it. Expect a wider range in a given context, not a rewritten default.
How long does it take to hear these in real time?
Most people can spot direction and frame of reference in a recorded conversation within a week of deliberate practice. Doing it live, while also holding your side of the conversation, takes longer and improves fastest when someone else is listening to the same exchange and comparing notes with you afterwards.
Is it ethical to read someone’s meta programs without telling them?
Listening carefully to how someone speaks is what any attentive person does, and adjusting your own wording so it lands is ordinary courtesy. The line sits at what you do next. Phrasing an offer in their language is fine; engineering pressure out of a pattern you spotted is not, and in coaching it belongs in the conversation openly.
Keep exploring
- Calibration exercises, for the non-verbal half of the same skill, where the evidence is posture and breathing rather than sentence shape.
- The representational systems test, if you want to know which channel your own filtering runs through before you start reading other people.
Carolin Mallmann is a Licensed Trainer of The Society of NLP, trained by Richard Bandler. She has trained more than 300 people and runs English-language NLP trainings in Lisbon, in groups of no more than ten.
