The Usual Questions
Two of the search questions regularly asked by visitors to this site are:
Both of which leave open the important question: What do you mean by "NLP"?
What is "NLP"?
And in most cases I doubt very much if that is what most visitors wanted to ask about.
The FoNLP is essentially a set of concepts and techniques which cover a range of communication skills, mainly derived from the use of the modelling technique developed by Richard Bandler and Frank Pucelik in the early 1970s (1972-73?). The wider FoNLP was jointly developed at first by Bandler, Grinder and Pucelik (from the mid-1970s (1974?) onwards), and after 1978 by just Bandler and Grinder. Bandler and Pucelik had pretty much completed their work on the modelling process when Bandler recruited Grinder to help Pucelik and himself to put what they had learnt into a form that could be taught to other people.
1. Does NLP Work?
The answer to the first question, then, is that it depends which part of the FoNLP the question refers to.
2. Do the Rest of the Elements in the FoNLP Work?
As for the rest of the elements - concepts and techniques - that make up the FoNLP, those work too, but many (most?) of them are "context related", to a greater or lesser degree. That means that few of the techniques will work in ANY context, for ANY practitioner, with ANY subject ALL of the time. So, to get a desired result in a given context, the practitioner must recognise what concept(s) apply, select the appropriate technique(s) and apply it/them with genuine skill.
If a technique does not work (get the required result) then the practitioner must be sufficiently knowledgeable and flexible to apply other techniques until they find one that does get the looked for result. As one of the NLP-related presuppositions suggests:
If you go on doing what you're doing now, you're likely to go on getting the results you're getting now.
This approach has sometimes been criticised, by the late Professor Margaret Singer, for example, who described it as "a fail safe argument" (Singer & Lalich, 1996, page 172). Presumably Singer believed that if a technique doesn't work first time we should apply the "talk louder and slower to foreigners" technique: Just keep battering away with the same technique until it does - if ever - produce the required result.
How Does NLP Work?
Talking of criticisms of whatever it is the critics think of as "NLP", it is worth noting that reports of experiments which have supposedly "disproved NLP" were in the vast majority of cases (allegedly) examined just two concepts - preferred representational systems (PRSs) and eye accessing cues - and a single technique - predicate matching.
Unfortunately - for the researchers - this finding tells us nothing of any significance about any part of the FoNLP.
Not only that, but we've just seen that NLP itself is one specific modelling technique, nothing else. Everything else that most people think of as NLP consists of NLP-related concepts, techniques and applications. Most of the techniques which make up the authentic FoNLP were arrived at through modelling successful behaviour of people who excelled in certain skills, observing what people actually do, as distinct from what they say and think they do.
Thus authentic NLP and genuine NLP-related techniques work because they are pragmatically-based - on real life behaviour - rather than on hypotheses and theories about what "ought" to work. They are techniques used by, or based on the behaviour of, people judged by their peers to be excellent in whatever skill/ability was being modelled.
What about replication of results - Or: What About "Scientific" Testing?
To put it another way, the NLP-related techniques are primarily based on codified models of behaviour that successful people naturally use to achieve their outstanding results in various situations and contexts as observed by the modeller(s). This is quite different from a genuinely scientific approach where, if applied in this context, would require that the modellers created a hypothesis, based on their observations, which they then subjected to some form of testing. And that alone raises a serious problem.
The "scientific method", as it's called, is very useful in situations where inanimate material is being investigated. It can also be applied quite effectively to a limited range of "very simple" animate creatures.&nbs; But by the time we start trying to apply the scientific method to more complex life forms, especially human beings, the notion of replication of results is only possible if we ignore individuals and look for purely statistical results based on behaviour in groups.
Compare this to the NLP-related approach where a model is based on a specific individual. So we immediately know that any resulting technique has been used successfully by at least one person. It is then down to other individuals to decide - preferably from personal experience of using the technique - whether it is sufficiently useful to them, as an individual, for them to make use of it in their own lives.
In short, the core NLP technique - modelling - is completely unlike the kind of situations where genuinely scientific testing is applied. Where Science is about "What ought to work - in a consistent fashion", NLP modelling is about "What has worked for someone and might work for others". There is no assumption in regard to NLP-related techniques that every technique will work every time in every situation.
If NLP Works - Prove It
That last comment is one which seems to confuse numerous critics of "NLP" who are concerned with the question of whose responsibility it is to "prove" that "NLP" works.
A response from many NLPers, whenever the question of "proof" comes up, is something like: "If you think that something to do with NLP needs to be proved, you go ahead and test it."
This is often, and wrongly, interpreted as meaning that NLPers think that it is the critics' responsibility to prove or disprove the validity of the various NLP-related techniques. But that really isn't the point at all. The real message is:
There is, moreover, one important qualification. Whoever is doing such research should only say they are testing claims made for any part of the FoNLP if that is indeed what they are doing. They need to do the necessary background research to discover which are the claims made by the co-creators of NLP and the FoNLP, and which have been simply made up by "wanna be" NLPers, free loaders, or even by other researchers. Otherwise, according to their own standards, they have no basis for claiming that they are researching NLP and/or the FoNLP as created by Bandler and Grinder and should make that clear when they report their findings.
In other words, it isn't a question of "passing the buck," but rather a matter of questioning whether any genuine "buck" exists.
Another important issue to remember here is this: Not being able to validate a given NLP-related technique does not prove that either (a) the particular technique is invalid, nor that (b) the FoNLP as a whole is invalid. As all genuine scientists know, even in the most rigorous, and appropriate scientific experiments the results do not prove that the hypothesis is true or false - they demonstrate a level of probability that the hypothesis being tested is true or false.
The Importance of Modelling
As explain above, NLP itself is actually a specific modelling technique - capturing the elements of how a particular person carries out a specific function to create a model that can be used by the modeller, and possibly taught to other people. (By implication, then, NLP is primarily designed for modelling the behaviour of people who have some outstanding skill or talent which other people might want to duplicate.)
So, if you wanted to help people to become really effective estate agents (realtors), you would not, from an NLP perspective, waste time sitting them round a table and getting them to imagine what qualities it might be useful to have in that job. Rather you would either model a top estate agent and teach the model to your students, or you would teach them how to model and then put them in a position where they could "shadow" a top realtor for themselves.
That is "genuine" NLP, as distinct from techniques which someone thinks ought to work, or might work if the wind is blowing from East Nor East and there's an "X" in the month, etc.
Everything else - virtually everything that most people think of as NLP - is a collection of techniques, methods and applications which can be used in the modelling process - and many other contexts.
In the end, short of trying to analyse NLP and every one of the NLP-related techniques, then, the answers to the two questions are the answers given at the end of the "Quick Answers" section:
In a nutshell: We know that the authentic FoNLP techniques work because, if they hadn't already been observed at work they wouldn't have been included in the FoNLP in the first place.