Working on my FAQ which deals with articles by critics of whatever they think of as "NLP", I have now read over 30 articles, etc., by academics and the like.
Except that there aren't really 30+ genuinely separate studies, more like over 30 subsets of the misinformation described here.
In practise the same quite small list of criticisms has been repeated over and over again. But a lack of originality would not be a fault if the criticisms in question were actually justified. The problem is that the criticisms are invariably based on earlier misinterpretations or lack of reliable information. Or both.
The first two errors are, in most cases, implicit rather than explicit. That is to say, they appear to be presuppositions rather than overt claims so that it is usually unclear whether the authors know that they are making assumptions and not basing their claims on genuine facts.
The "Man with a Hammer" error
The "Man with a Hammer" error is summed up in the saying that, "To the workman whose only tool is a hammer, everything looks like a nail".
In the context of criticism of NLP this works itself out as, "Every academic critic describes the term NLP in a way that coincides with his or her speciality". Thus psychologists/therapists such as Heap and Sharpley claimed NLP was a form of psychotherapy; psycholinguists such as Levelt have criticised "NLP" for trying to be like psycholinguistics (and/or neurolinguistics); sociologists dealing with global trends, such as Boussebaa and Spicer criticised "NLP" for failing to fulfil an imaginary promise to bring the Earth to a new state of calm and harmony; and those interested in the sociology of religion, such as Hunt, claim that even if NLP isn't actually a religion, it is at least infused with "religiosity".
For what it's worth, according to the entry on Wikipedia:
Religiosity, in its broadest sense, is a comprehensive sociological term used to refer to the numerous aspects of religious activity, dedication, and belief (religious doctrine). Another term that would work equally well, though is less often used, is religiousness. In its narrowest sense, religiosity deals more with how religious a person is, and less with how a person is religious (in terms of practicing certain rituals, retelling certain myths, revering certain symbols, or accepting certain doctrines about deities and afterlife)."
Quite how NLP fits any of these description is a mystery to me. In an attempt to understand Hunt's claim, for example, I e-mailed Mr Hunt, well over a year ago, with a polite request for clarification. So far he still hasn't responded to my enquiry.
It seems, then, that whilst these critics may all be experts in their own fields of study, when it comes to authentic NLP and the FoNLP, they wouldn't know the real thing from a hole in the ground.
The "Assumption of Expertise" error
It is commonplace for journal articles to include numerous citations of other pieces of the literature - articles, papers, books, etc. - which cover the same area of investigation, usually because the earlier work supports some or all of the current author's claims.
The basis for this practise is, of course, an assumption that since the previous author(s) has/have a firm grounding in the subject under discussion so their comments can be assumed to be based on relevant experience and expertise.
This same "assumption of expertise" is clearly evident in articles critical of whatever the authors think of as "NLP". Thus Sharpley, in his second review (1987), argued that the research he had reviewed must be valid because:
It is very difficult to accept (as Einspruch and Forman, 1985, suggested) that all of these researchers were guilty of the "methodological errors" (p. 590) that they claimed leave the total research on NLP [sic] to date inconclusive and "trivial" (p. 594).
(Sharpley, 1987. Page 105)
And later on the same page:
The basic tenets [sic] of NLP [sic] have failed to be reliably verified [sic] in almost 86% of the controlled [sic] studies, and it is difficult to accept that none of these 38 studies (i.e. those with nonsupportive, partial or mixed results) were performed by persons with a satisfactory understanding of NLP (or at least enough of a satisfactory understanding to perform the various procedures that were evaluated).
(Sharpley, 1987. Page 105)
It is unclear what "controlled studies" Sharpley is referring to, here. After all, in his previous article he commented that:
A series of controlled studies using reliable indicators of change in clients' behavior (rather than their perceptions of counselors, which may not be correlated with problem dissolution by clients) is called for.
(Sharpley, 1984. Page 247. Italics added for emphasis)
But whatever Sharpley meant, a review of the research literature provides plentiful evidence that (a) qualifications and experience in psychology aren't at all the same thing as viable knowledge of NLP or the FoNLP, and (b) that many of the errors were in fact based on the same kind of misplaced "assumption of expertise" regarding previous researchers that undermined the validity of Sharpley's own articles.
The "Lack of an Accurate Definition" error
If one is going to discuss something, favourably or otherwise, it would seem to be an obvious pre-requisite to supply an accurate and adequate definition of what it is you are discussing. Yet few, if any, of the critics of whatever it is they (individually) think of as "NLP" provide anything of the kind. Thus the literature is full of claims such as:
Furthermore, there is no evidence that NLP is an effective therapeutic technique (Sharpley, 1984).
(Elich et al, 1985. Page 625)
Did Elich et al really mean
"there is no evidence that non-analytical modelling is an effective therapeutic technique (Sharpley, 1984)"?
It seems highly unlikely, since Sharpley gave no indication in either of his reviews that he had the slightest idea that NLP was actually a modelling technique.
So whilst the authors may have thought that their comment was a damning indictment of "NLP", it was in fact merely a concrete demonstration that - just as they indicated, several times over, on the first page of the article (p. 622) - they had no idea what either NLP or the FoNLP are really about.
In practise, most critics address no more than one or two concepts/techniques which, as in Elich et al (1985), they then refer to as though they constituted NLP (i.e. the modelling process) and/or the whole of the FoNLP.
In a majority of cases the result of this lack of correct information involves just four elements of the FoNLP - eye movements, representational systems, preferred representational systems and the "predicate matching" technique - are dealt with as though they represented the whole of the authentic FoNLP. Thus we find comments like:
To evaluate NLP [sic] without testing for the presence of the PRS and it's usefulness in communication and treatment [sic] appears to be impossible.
(Sharpley, 1987. Page 103.)
In the real world, however, detecting someone's PRS and using it in the "predicate matching" technique is just one of a number of rapport-building techniques and both could disappear overnight without making any significant difference to the rest of the FoNLP.
Or again:
... it appears that with 44 experimental reports on the effectiveness of NLP, a conclusion can indeed be drawn regardless of the degree to which the various researchers were experts on NLP.
(Sharpley, 1987. Page 105. Italics added for emphasis)
This comment is especially notable since virtually none of the 44 experimental reports dealt with anything other than the four elements cited above: eye movements, representational systems, preferred representational systems and the "predicate matching" technique.
And less than a handful of the reports dealt with these or any other elements of the FoNLP with even a hint of understanding of the genuine claims made by the co-creators of NLP and the FoNLP.
The "Reification of NLP" error
To be fair, this error is not limited to the writings of critics of the FoNLP.
In brief, "reification" is the process of treating something abstract - an idea/concept, for example, as though it were a tangible "thing" (a process sometimes known as "thingifying"). as Einspruch & Forman (1985) pointed out, in response to Sharpley's first review article (1984):
these authors [i.e. the ones whose work Sharpley had reviewed] focus on the primary representational system (PRS) and reify the term, another major mistake. The danger of reifying terms is that one may be easily led to mistake a construct for reality
(Einspruch and Forman, 1985. Page 590)
Typical examples of the thingifying of NLP include: "NLP says", "NLP claims", "according to NLP" and so on.
The practical reasons why this is an error include:
- This is, in effect, an example of a "missing referential index". That is to say, instead of naming a specific source we are offered "NLP", which is simply a particular modelling process (see above), as though it were something that could make valid pronouncements,
- Like the "Proponents claim" error (see below) it implies that any comment on NLP and the FoNLP is valid, regardless of who made it and whether they knew what the heck they were talking about.
The "NLP as Therapy" error
Most of the literature which is critical of whatever each author thinks of as "NLP" includes some form of the allegation that this "NLP" is a type of psychotherapy. Indeed, so pervasive has this myth become, when Norcross et al (2006) carried out their poll of "doctoral-level mental health professionals", which asked them to rate 55 "alternative" treatments - including "Neuro-Linguistic Programming (NLP) for treatment of mental/behavioral disorder" (Norcross et al, 2006, page 518) - only 26.7% of the original respondents admitted to being "not familiar" with this "treatment".
Which says a lot about both the research team and the so-called "experts" who replied to the question, since:
- The researchers failed to provide a standard description of any of the "treatments" they had listed so they had no idea what respondents were actually rating,
- There was no indication of which particular "mental/behavioural disorder[s]" were being supposedly being treated, and
- There is in fact no specific technique or group of NLP-related techniques which fit this description.
It is interesting to compare attitudes on "NLP" as therapy with attitudes to a similar field of study known as General Semantics, established by Alfred Korzybski in the 1930s and one of the leading influences on the development of NLP (the specific, non-analytical modelling technique) and the wider FoNLP.
In his book Language Habits in Human Affairs: An Introduction to General Semantics (1941, pages 9-11), Professor Irving J. Lee listed a number of sources that demonstrated the breadth of applications of General Semantics. The list included Dentistry, English and Speech, Philosophy, Physics and Sociology, with two lengthier sections on General Education and Psychiatry.
The section on psychiatry included items such as A Preliminary Report on the Psychotherapeutic Application of General Semantics, Congdon & Campbell (1938) and Preliminary report of Two Cases of PsychoPathic Personality with Chronic Alcoholism Treated by the Korzybski Method, John K. Lynn (1938).
Which begs the rather obvious question - given that General Semantics is still alive and well - as to why it was not included by Norcross et al in their survey, and why it is not routinely subjected to the kind of misleading claims about it being a form of psychotherapy in the way that routinely occurs in relation to the FoNLP.
(Incidentally, General Semantics is no more a form of psychotherapy than NLP or the FoNLP are. Just like the FoNLP it merely includes elements which can be applied to enhance the psychotherapeutic process - and education, and business, etc., etc.
Though having said that, it seems that some psychologists are now in the process of calling anything that moves "psychotherapy", and painting the rest white (based on an old British army joke).)
Here, then, for anyone still not clear on the subject, are five reasons why it makes no sense to talk about either NLP or the FoNLP as forms of psychotherapy:
- Both Grinder and Pucelik have told me, with no equivocation and in the presence of other people, that at no time were they trying to create a form of psychotherapy, not in the 1970s, 1980s, 1990s nor in the 21st century. Likewise, as recently as November 2010 Bandler told an interviewer for BBC Radio 4: "NLP is not therapy, and it is not self-development".
- Neither the UKCP (United Kingdom Council for Psychotherapy) nor the EAP (European Association for Psychotherapy) have accepted NLP as a form of psychotherapy. On the contrary, a group calling themselves the NLPtCA (Neurolinguistic [sic] Psychotherapy and Counselling Association) has been set up, entirely outside the authentic NLP-related community, which has gained recognitions as a form of therapy allegedly based on constructivist principles.
This would obviously be unnecessary if NLP and the FoNLP were already forms of psychotherapy.
- Although critics have claimed that "NLP" claims to know how people's brains work, in practise the FoNLP does not include a concept of the "Self", and far from claiming to know how to "make people right" one of the NLP-related presuppositions explicity states that, "People aren't broken and don't need to be fixed".
- The true focus of the development of NLP and the FoNLP was not to create yet another form of psychotherapy (even in the 1970s the field was already swamped), but to discover what communication techniques made Perls, Satir, Erickson, etc., particularly successful.
Despite claims that "NLP" started as a form of psychotherapy and then branched out into other fields (see Eisner, 2003, for example), there is abundant evidence that the techniques have been available for use in a wide range of contexts - business, education, parenting, etc. from the late 1970s onwards. That is, even before the initial stage of the development of the FoNLP was complete.
As Bandler and Grinder explained:
NLP begins in the early 70s when we found ourselves in possession of a set of extremely powerful and effective communication models. We had originally developed these models for use in the psychotherapeutic context. It quickly became apparent that they could be easily generalized to other areas of human communication - specifically business (sales and negotiation), law and and education.
(John Grinder and Richard Bandler, in the Forward to Dilts, Grinder, Bandler and DeLozier Neuro-Linguistic Programming: Volume 1 (1980))
- The National Center for Biotechnology Information, U.S. National Library of Medicine, which should presumably reflect the difference between a form of psychotherapy and anything else describes NLP thus:
Neuro-Linguistic Programming
A set of models of how communication impacts and is impacted by subjective experience. Techniques are generated from these models by sequencing of various aspects of the models in order to change someone's internal representations. Neurolinguistic programming is concerned with the patterns or programming created by the interactions among the brain, language, and the body, that produce both effective and ineffective behavior.
The "Three Ways of Identification" error
This is possibly the oldest of the common errors, dating back at least as far as a 1977 dissertation submitted by Lee F. Owens at Ball State University .
The error is a mistaken belief that a person's PRS (preferred representational system) can be determined in three ways - by watching their eye movements, by self-reporting, and by listening to their use of sensory predicates - or any combination thereof.
If Owens (and everyone else who has replicated the error in their own research) were correct in thinking that this was/is an authentic claim made by the co-creators of NLP/the FoNLP then by implication we should expect to find that using all three tests at the same time would produce congruent results across the three methods. Yet the results of such experiments very seldom produce congruent results, a fact that is frequently taken as proof that "NLP doesn't work", as in the conclusions of one of the more frequently cited articles:
... the findings of recent [i.e. recent in 1985] systematic research on NLP [sic] ... bring into question the veridicality of the [undefined] theory of NLP ...
(Gumm, et al, 1982. Page 330)
In reality, however, Bandler and Grinder never made the claim attributed to them by Owens et al, and it should come as no surprise at all that the related experiments seldom, if ever, produce positive results.
What Bandler and Grinder actually "said" was:
In order to identify which of the representational systems is the client's most highly valued one, the therapist needs only to pay attention to the predicates which the client uses to describe his experience.
(Bandler & Grinder, The Structure of Magic II (1976. Page 9)
In short, as far as I know, no one making this claim has ever produced any evidence that Bandler and/or Grinder and/or Pucelik ever made the "three ways" claim. And, perhaps even more telling, no research I know of has ever mentioned Bandler and Grinder's instruction that listening to someone's use of verbalized sensory predicates is the only thing needed to spot a person's "most highly valued/preferred representational system".
Not surprisingly, then, numerous researchers have wasted their time looking for agreement where no agreement was ever implied.
The "Unproven Representational Systems" error
Several critics have seized upon the "representational systems" concept and referred to it as though it was, or at least might be, lacking in adequate "scientific support". But since no one ever seems to be willing to explain what they mean by that it simply isn't possible to know what any particular critic means when they raise this objection. On this basis it seems that this error is little more than a red herring - an "empty" criticism that is offered merely for it's nuisance value.
However, if anyone will be kind enough to provide an explanation as to why psychologists think the representational systems concept is problematical this section will be expanded accordingly.
It does occur to me that some people might have difficulty with the idea of our senses being "representational" systems, so let me explain that a little more fully.
According to John Grinder, the processes of mentally perceiving and dealing with sensory information goes something like this:
- Information reaches our senses from both the external and internal world in a purely experiential form. This he calls "first access".
- However, our senses are all limited to some extent - we can only detect certain frequencies of "sound", certain wavelengths of "colour" and so on, Therefore there is an immediate transformation, at point F1, where the original experiential information is whittled down - where it is distorted and deleted - by the perception process.
- The raw perceptions then pass either into our conscious or unconscious.&nbs; If to our conscious awareness, then, or anytime later, it seems that we still have various kinds of images - visual, auditory, etc. - but in order to relate them to each other we turn the images into language - at point F2. This, Grinder suggests, results in further deletions, distortions and generalisation as we try to apply our comparatively limited language skills to this translation process, allowing that we also have a tendency to want to make everything new fit in with we have already learned from previous experience.
It is this whole process, by which original experience is "represented" to our conscious awareness, that is the basis for calling the senses "representational systems". It is one way of identifying the functionality of the sesnses, and is not meant to exclude any other valid description.
The "Unproven Preferred Representational Systems (PRS)" error
This is a very simple criticism to answer since the proof is in the definition.
A person's DPRS (default preferred representational system) is the one they use most often over time - measured over months or years. It can be determined by counting the number of sensory predicates from each modality - visual, auditory, kinaesthetic, olfactory, gustatory. However the DPRS is really only of interest if one or two representational systems are clearly preferred over the others so counting likely to be unnecessary.
The importance of detecting a person's DPRS is to find out how skilfully a person is using their representational systems across a variety of contexts.
A person's CPRS is readily identified by their latest verbalized sensory predicate. The CPRS indicates which representational system a person is consciously aware of at any time (though all sensory systems are operating all of the time at a sub-conscious level).
The "Static Preferred Representational Systems (PRS)" error
Sharpley (1984) included brief details of an experiment by Laura Birholtz (1981) in which she tested what she referred to as "the implied assumption that [the] preferred modes of representation are stable". Though she didn't directly acknowledge that the assumption was "all her own work".
In practice no such assumption is justified, a fact concealed, to a certain extent, by very poor experimental design.
The "Proponents claim" Error
Many of the academic articles critical of "NLP" feature one or more examples of the "Proponents claim" error.
This is a manoeuvre in which the author(s) ignore(s) relevant comments made by the obvious authorities on a given subject in favour of "quoting" what are usually unspecified, non-authoritative sources who may or may not exist. This allows the author(s) to insert "evidence" which supports their point of view, possibly from a real source; possibly one devised the author(s) and used as an excuse to express their own opinion without openly acknowledging it as such.
This is an important error, not only because it is frequently used to introduce inaccurate information but also because it demonstrates the kind of illogical thinking that supports criticisms of whatever the author(s) think of as "NLP".
In brief, going by the material I have investigated so far, all of the critical academic articles refer to Neuro-Linguistic Programming as having been created and developed by Bandler and Grinder. And one or two writers also acknowledge the involvement of Frank Pucelik. On this basis we are entitled to require an end to the nebulous term "proponents" in favour of precise references. After all, if the quote is genuine then the person citing it must know where it came from.
We are also entitled to expect that if an author chooses to refer to the authentic FoNLP created by Bandler and Grinder then they will restrict their quotes to comments made by Bandler and Grinder, or at least to comments/claims which accurately reflect the claims and views of Bandler, Grinder and Pucelik with, at most, only minor, non-crucial variations.
It is not a satisfactory practice to refer to unspecified "proponents" since:
- This does not allow us to check that these "proponents" actually exist,
- It doesn't allow anyone to check whether the quotes/claims attributed to these hypothetical proponents are genuine or accurate, and
- It does not allow anyone to check whether the claims made by the "proponents" are congruent with genuine claims made by any or all of the co-creators.
In short, far from having any place in allegedly "scientific" criticisms, the "Proponents claim" ploy is no better than mere rumour-mongering.
It is a fact that Bandler, Grinder and Pucelik - jointly and individually - have written enough books, and are sufficiently well represented on YouTube, that quoting third parties on the subject of NLP and FoNLP is seldom if ever necessary. Indeed, in my experience of critical articles/reviews/etc., a quote from a third party, especially from an unidentified third party, is, almost by default, a guarantee that what is being offered is in fact misinformation.