Neuro Linguistic Programming and Its Place in the Origins of BDD

Behaviour driven development grew, by its author’s own account, out of a thought experiment based on techniques from neuro linguistic programming. The specific borrowed idea is that the words used to describe an activity influence how people think about it.

That is the whole of the connection, and it is worth stating plainly in both directions, because the association is sometimes used to inflate NLP’s standing and sometimes used to dismiss BDD.

The borrowed idea

Coaching developers into test driven development, Dan North found that the objections he met clustered around a word rather than a practice. Programmers said writing tests was somebody else’s job. Testers asked why programmers were doing their work. Managers asked about the new testing phase in the schedule.

None of those objections concern the activity. Replacing the noun, from a test to an example of desired behaviour, dissolved most of them, because nobody argues that a team should avoid agreeing on what it is building.

The prediction that the rename would change the reception is the part traceable to NLP. It was borne out well enough that a practice was built on it.

What that does not license

NLP’s broader claims, about representational systems, eye movements, and modelling excellence, have not held up. Systematic reviews have repeatedly found the core model unsupported, and it is commonly characterised as a pseudoscience.

Those two judgements sit together without difficulty. That vocabulary affects framing is an ordinary and well-attested observation, studied under other names long before NLP existed, and it does not require NLP’s model to be correct. BDD borrowed a true observation from a body of work that is mostly not.

The honest summary is that the origin story runs through NLP and the method does not depend on it. Nothing in writing a scenario, running a three-amigos conversation, or automating an example rests on any NLP claim.

Where the influence is still visible

In two places, both to do with wording.

The insistence that the specification be phrased in the domain’s own words rather than a technical equivalent is a bet that the phrasing changes what people notice. Getting the words right is treated as method rather than style for that reason.

And the attention paid to how a question is asked in the conversation, covered under powerful questions, carries the same assumption: that an open question and a closed one about the same rule will produce different information.

Both are defensible on their own terms and neither needs a theory of mind behind it. Which is the sensible way to leave the connection: recorded, because it is historically accurate, and load-bearing for nothing.

Why the association is worth stating plainly

There is a temptation to quietly drop it. NLP is a liability to cite, the connection is thin, and no part of the practice depends on it, so a reference page could omit it without anyone noticing.

That would be a small dishonesty about where an idea came from, and this domain is a poor place for one. The founding account names the connection; readers who go to the source will find it, and finding it unmentioned here would reasonably make them wonder what else was tidied.

The better handling is the one above: record the borrowing, say exactly which claim was borrowed, and be equally clear that the borrowed claim stands on its own evidence rather than on the framework it came from.

The general point about renaming

The mechanism BDD relied on is not exotic. Changing what an activity is called changes who believes it is theirs, and that has been observed in far more rigorous settings than NLP under names like framing.

It is also reversible, which is the part worth remembering. Calling BDD scenarios “the acceptance tests” in a planning meeting hands them back to whoever owns testing, and the conversation the practice depends on stops being everyone’s business. Teams that have adopted the practice successfully tend to be quite careful about the noun years later, and that care is the clearest evidence the original observation was sound.