Talk:Neuro-linguistic programming and science

POV Fork
After reading the main NLP article and then this one, I noticed that a huge amount of material is repeated, and in particular that this article represents (at least in its introduction and first half) a MUCH more positive view towards NLP. In particular the claim in the introduction that "There is significant evidence, both in research and anecdotally, that NLP does something significantly more than a placebo" Seems specifically contradicted in the last part of the article. There is, in fact, significant evidence that it does not.

It also must be rememebered that evidence that certain techniques 'claimed' by NLP practitioners to be NLP are in fact banal truisms or established counselling techniques, and therefore that evidence that they specifically work should NOT be presented as evidence for NLP as a whole.

Finally, a telling criticism of NLP is that, if it works, then why don't the people who claim to be masters at it use their skills to become fabulously wealthy. It would be easily done; go to a gambling even, use NLP techniques to convince the bookie you have won, collect winnings, repeat.

---

"..if it works, then why don't the people who claim to be masters at it use their skills to become fabulously wealthy" < --

The other side to this idea is that perhaps it tells only of the fact that "masters" of NLP have morals. Looking at it from another angle reveals to us that some people project this particular idea onto others because they can see that they themselves would actually do something like that (given the knowledge and a context) and therefore believe NLP masters would do the same too.

However, for some people, using their NLP skills to simply acquire large sums of money by manipulation and continuing to do so with complete moral disregard is a greedy and narrow minded pursuit/application of such knowledge, hence why you never see or hear of "masters" (such as Bandler, Grinder, Derren Brown, Bolstad etc) abusing their position as NLP practitioners.

- JC —Preceding unsigned comment added by 203.211.79.237 (talk) 07:41, 15 October 2007 (UTC)

Edits I am making
Just a quick explanation. I am basically going through the article and making it more encyclopedic. I removed the "What is NLP?" section because it simply isn't needed. We're an interactive encyclopedia. All you need to do is to click on a link to find out what something is. There is no need to have a What is section when NLP is linked above. --Woohookitty(meow) 08:34, 13 June 2006 (UTC)
 * As with anyone, sometimes I write decent draft articles. Sometimes its good. Other times it needs or benefits from others review and some redaction, cleanup and copyediting. I think thats a common thing. WP:OWN says something about it, I'm sure :)


 * By collaboration and discussion, we seem to get decent quality information :) I'll head off to the other tab to go read the edits. I'm sure the edits are good and experienced ones. If there was by chance anything major lost by them, then I'm sure it'll get talked round. FT2 (Talk) 10:55, 13 June 2006 (UTC)


 * Cool. Yeah I'm just trying to make it more encyclopedic. If I removed something     that you think is needed, we can discuss it. --Woohookitty(meow) 11:13, 13 June 2006 (UTC)
 * Please -- and much appreciated. Sometimes when one tries to add info or get more in, one don't always balance the wood and the trees so well, and its really good to have others watching to double check it doesn't start going down the wrong track. It needs to be kept cited and balanced such (even if not all cites are in and some more Fact tags should be added at times)... so its valued :) feel free to add Fact liberally: it helps. FT2 (Talk) 14:23, 13 June 2006 (UTC)

Text chunk snipped from old NLP article, to consider using in this article

 * "Some critics of NLP assert that the majority of methods taught as part of NLP have not been scientifically verified and some even classify it as a pseudoscience."


 * "In many traditional senses of the philosophy of science (see for example writings by Carl Hempel), this could be an apt claim: many of NLP's propositions are functionally untestable in the empirical predictions they seem to make, and those that are testable have often not been tested or have been tested with mixed results (see for example peer-review studies on the effectiveness of NLP modality matching techniques in psychotherapy). Correlational studies supporting NLP's ideas are limited or open to alternate interpretations and laboratory causal studies are almost nonexistent in the literature."


 * "In assessing the criticism of NLP as "unscientific," however, it must be noted that numerous popular psychotherapeutic traditions lack such clear, peer-reviewed scientific support as well. Psychodynamically oriented schools of therapy are criticized today for similar reasons, and Humanistically oriented schools of therapy often are as well. As the field's techniques stand today, scientificially evaluating the effectiveness of various talk-based therapies often tends to be a very problematic undertaking."


 * "Furthermore, in an Anthropological or Kuhnian sense, criticism of NLP as "unscientific" may be related to the aforementioned divorce between NLP's development and traditional institutions of science and psychology. In this vein, NLP is often hastily dismissed without earnest consideration by even non-scientist psychotherapists because the claimed efficacy of some NLP techniques seems "unrealistic" or "impossible.""


 * "By its nature, Science is skeptical of unproven claims, and anecdotal evidence - even accumulated over time and people - is not sufficient to establish a scientific fact. Thus, until controlled research is carried out, the claims of NLP practitioners that "in my experience it works" will not persuade scientists, for precisely the same reason that scientists do not often believe in astrology."


 * "The method of proof in NLP is different than that used by scientists; NLP places little emphasis on prediction. (However, note that some sciences lend themselves to predictive theories more than others.) While NLP models which have repeatedly been found useful may be regarded as generalizations which will usually (but not always) be useful, practitioners do not usually refer to these generalizations and do not extrapolate them into predictions for experimentation over extended periods of time. Rather, such extrapolation typically occurs within a single session by a practitioner working with a subject. The "hypothesize - predict - test - verify" cycle is performed in minutes and repeated many times during a session, on an observational basis."


 * "While NLP makes heavy use of the scientific method in the small, it lacks fundamental characteristics of science in the large, such as carefully controlling experiments, and subjecting them to peer review in refereed journals. It must be noted that those NLP practitioners who do conduct experiments and write up the results may find that the recognized mainstream journals have a policy to be overly skeptical of their results."


 * "Unlike formal scientific research, NLP does not have truth as a primary goal. Rather, it seeks to do things effectively and efficiently. Some have argued that NLP might be more properly classified as an engineering or technology discipline rather than a science. It has, in fact, been frequently marketed as a "technology" or "instruction manual" for the mind. It is worth noting, however, that many traditional definitions of technology define the word in terms of useful entities or ideas that are based on scientifically established empirical or theoretical principles."


 * "Because NLP does not possess empirical or theoretical claims that have been validated in the traditional scientific senses discussed above (predictivity, proof by experimental design seeking to establish causal relationships, correlational studies, etc...see Karl Popper and logical positivism for more on one method of approaching the philosophy of science), it is questionable whether or not NLP can even be properly called a "technology" in this sense."

FT2
I'm giving you a couple of more days and then I'm going to need to do some major editing here. We cannot let this article stand in this state (with the empty sections) for much longer. --Woohookitty(meow) 09:19, 19 June 2006 (UTC)

Need for article expanding
Hello, guys, can you please start working on this article? We desperately need this article in Russian Wikipedia with sources, links etc. Eli the Barrow-boy 20:03, 5 November 2006 (UTC)

From NLP article on 'Theory'
This section was moved from the main article.
 * Robert Dilts in Roots of NLP (1983 p61) explains neural functioning in relation to the adding of newconnections,Hebbian cell assemblies (Hebbian engrams), causal loops, and digital circuitry. From his observation of the work of scientist Konrad Lorenz, Dilts states that when learning experiences occur in our life, new neural networks are imprinted in our brains recording events and their associated meaning. Basing his conclusions in part on Timothy Leary's 8-Circuit Model of Consciousness, Dilts states that these imprints "established at neurologically critical periods," could be later re imprinted or reprogrammed. (Dilts, 1990, p76,77). Practitioners such as Derks, Singer, and Goldblatt theorize that NLP processes can be explained through the neurological concepts of programming and reprogramming engrams. According to Derks, NLP anchors are conditioned stimuli which work by activating engrams which are proposed "to give a patterned response which has been stabilized at the level of unconscious competence".

It appears to be mainly original research. --Comaze 10:30, 30 November 2006 (UTC)

Some Comments
I should be honest and state that I find the evidence for NLP weak. However, I'll try to be bipartisan in pointing out a few issues with the article... Is "The Irish National Center for Guidance in Education" really a credible and authoritative source for anecdotal evidence? The article basically reads like it wants to say "Look, systematic reviews have failed to find much evidence for NLP but basically a lot of people [mainly NLP trainers] want to say it works and that's backed up by some comments in a leaflet by an obscure Irish charity." Surely, even the fans of NLP don't think that looks credible and balanced?

I think if you want to appeal to anecdotal evidence then you need to discuss in detail the established problems with doing so. This is a well-known minefield, that's why most researchers avoid it like the plague. There are notoriously high placebo and spontaneous remission rates for the kind of issues NLP techniques like the fast phobia cure claim to treat. That's one obvious reason why NLP practitioners, and other therapists, have been accused of tending to over-estimate their success. Beecher's (1955) famous study found the average placebo control group to achieve 35% improvement, some studies on anxiety show rates as high as 60% for sham therapies. So if you were to use a totally ineffective technique with 100 people you would still get 35-60 testimonials, or pieces of anecdotal evidence. That's why it's not worth the paper it's written on unless backed by independent research. This is a well-known (major) problem in psychotherapy research but NLP practitioners tend to ignore it, and this article seems to gloss over it.

Worse, anyone familiar with psychotherapy research knows that the seminal meta-analysis by Smith et al. in the 1980s showed that many different models of psychotherapy, using conflicting theories and techniques, tended to achieve similar effect sizes. All had fans who passionately proclaimed their effectiveness, but were unaware that they would have been equally effective doing anything else because what success they observed was almost certainly down to the notorious "non-specific" factors that all therapies had in common. That is, receiving attention, basic care, warmth, listening, etc. We've known since then that any psychotherapy can achieve success rates of about 60% due to these non-specific factors, basically a glorified placebo effect, rather than the specific techniques or concepts being employed. In other words, the NLP advocates who claim their therapy works in practice need to show more, that it works better than if they were to do something else instead, e.g., that rewinding a movie clip works better than fast-forwarding it, or just relaxing (as in systematic desensitisation). Anecdotal evidence fails to take account of comparisons with alternate methods, and so it tends to conflate genuine efficacy with the success achieved by non-specific factors, the placebo effect, and spontaneous remission. The appeals to anecdotal evidence in this article are seriously pseudo-scientific unless they address this obvious problem.

Finally, and I do apologise for the length of this comment but I has to be said, it creates a well-known problem if you depend upon anecdotal evidence, and one especially troublesome for for Wikipedia. That is, if the only "evidence" for a statement is the fact that "such and such" said so then its credibility obviously turns on the credibility of the person providing the testimony. That means criticism tends to have to focus upon "ad hominem" points, questioning the personal authority, or even honesty of the individual or groups whose "anecdotal evidence" is cited. For instance, if we are told to believe that NLP is spectacularly effective because NLP trainers claim it is, the obvious response would be to question whether these are intelligent, trustworthy and honest people. Given the well-documented controversies surrounding the personal lives and business activities of many leading lights in the NLP world -many of which probably abound elsewhere on Wikipedia- that potentially opens up a whole can of worms! HypnoSynthesis 18:55, 20 February 2007 (UTC)

Hypnosynthesis. This is one of many daughter articles to the main article, some of which appear to have been written out of desperation given what was happening on the main article. The main article has been under attack for over 18 months by a long term abuser and multiple socks, all banned, although the most recent manifestation has only just been banned. The POV pushed was that NLP was a mad, evil cult with links to Scientology etc. Other editors were (and still are) the subject of vicious personal attacks. Citations were invented and/or grossly misrepresented and all references and cites have had to be checked again. The editors battling against this on the main site are currently putting it in order. You should look at the main site (NLP) and read the discussion page and references there first.eg, the opening sentence of the section on psychology now reads:
 * The broad judgement of the evidence-based psychology community is that NLP is scientifically unvalidated as to both underlying theory and effectiveness.[18][5][19]Amidst pleas for further research there is much criticism of the failure of proponents to undertake adequate research.[34][35]Some classify it as a pseudoscience citing that many proponents claim a scientific basis that is not supported by research or current scientific knowledge.[36][35] [3]Others have suggested that any claimed effectiveness results from its reliance on a range of therapeutic techniques gleaned from other therapies rather than any new theories or techniques.[5].....etc.

The plan would be to deal with the daughter articles (of which there are too many and which are out of date) once the main article has been cleaned up.Fainites 17:56, 21 February 2007 (UTC)


 * Fair enough. I understand what it's like having to deal with e-vandalism and stuff.  Won't score me any popularity points for saying this but, to be honest, I think there is at least a grain of truth in the idea that specific NLP organisations/courses, without naming names, may resemble a kind of personality cult.  I mean the way people are manipulated into accepting an ideology wholesale and suspending any critical thinking, incorporating Huna and other strange New Age beliefs, and uncritically idolising figureheads as if they were gurus.  I'm sure NLPers won't see it that way but to an outsider it does often seem a little cult-like.  If it helps to soften the blow, the same criticism has very credibly been laid at the door of psychoanalysis by several academic authors.  172.207.163.158 09:36, 22 February 2007 (UTC)

Yes. People who are motivated to create cult like groups will use any number of persuasion techniques and NLP is no exception. There is quite a long thread on the NLP talk page about this. Have you read Singers book 'Cults in our Midst'? She defines what makes a cult a cult as opposed to merely some weird therapy. She doesn't actually mention NLP as a cult or even hardly as used by cults (although she's pretty damning about NLP in her other well known book 'Crazy Therapies'). Charismatic figureheads certainly feature. However, cults existed before NLP was invented. Some cults have arisen from the abuse of older forms of 'therapy'. As alot of NLP is based on hypnotic techniques it is inevitable that cult type organisations will use it and it may be that some culty type affairs have arisen from NLP groups. Tony Robbins may be borderline here. However, that is very different to saying NLP itself is a cult. It's far too diffuse, it has no worshipped charismatic leader (they all fell out!) and no concerted programme or aim. We can't even agree what the hell NLP is on the Wiki entry! There is an entry dealing with cults on the NLP page. The trouble was, there were a total of 9 citations put in by various abusers and socks to the effect that NLP was a cult. All were examined and all except one were false. The only one that wasn't and did actually name NLP as a psychocult was a Russian Archpriest called Novopashin. This didn't seem enough by itself to call NLP a cult! We have quoted Langone who expresses concern about the use made by cults of NLP. Of course exit counsellors who rescue people from cults also use some NLP techniques. If you have any details of specifically NLP cults I'd be most interested to see them. You say 'without naming names' but we have to have verifiable, credible sources, ie name names or cite investigations thereof.Fainites 17:51, 22 February 2007 (UTC)

Why Can't NLP Be Tested Scientifically?
I find the section "NLP And Science" very problematic. It is basically a lengthy arugment that NLP cannot be tested as a science due to various factors, but I find some of these factors nonsensical, while others have long ago been recognized and addressed in other fields that are verified scientifically.


 * "In NLP a 'hypothesis' relates not to human processes in general, but to the inner processes of a given person at the present time, the relationship between the observable exterior and the unobservable interior, and the presence of other potential processes and inner structures which may be evidenced by deduction or suspected from prior experience." This seems to be saying that different people may have different belief systems at any given moment. Fine. But these are not the hypotheses that would need to be tested in the study of NLP. Rather, the overall hypothesis would be something to the effect of "the practice of NLP produces visible results," which should certainly be externally verifiable.


 * "The methodology of NLP has therefore been compared to an engineering discipline, in that it seeks what works, rather than what is theory or true in a testable sense." Sure, but then the truth we are trying to test by scientific study is, again, "NLP works."


 * "A notable difference is that the subject-object or observer-observed barrier is explicitly removed in NLP." Standard science already addresses this problem at great length. See for example Objectivity (science).


 * "it is necessary to ask whether NLP is expected to act like a science, or whether it acts more like a black box in which the only effective measure is to statistically evaluate quality of output for a given input." Again, why can't we test the claims made for the "black box"?


 * "However, a high degree of variability in each individual trial..." This is hardly unique to NLP but exists in pretty much all social sciences, and medicine too. That's why we use large numbers of randomly selected individuals in a controlled trial; indeed, much of modern statistics was developed to deal specifically with this problem. See also experimental design.


 * "the vagarities of human whim and craftsmanship, and the existence of multiple optimal solutions, are inherent in NLP's structure." Yes, but evidence-based medicine also has this problem, yet it is studied scientifically.


 * "Overall, NLP's results are broadly at some level, metaphorical tools which are believed to have an unusual ability to indirectly manipulate neurological structures, to obtain subjectively beneficial ends (rather than natural entities that exist or do not)." Psychological research has exactly the same problem, in that the end states are subjective. This has led to the development of sophisticated measuring tools for internal states and attitudes. See for example the Beck Depression Inventory or even the good-old-fashioned IQ test. This is not an easy thing to do, to be sure, but we have reason to believe that such tools can be effective; these reasons are far beyond the scope of this comment.


 * "Used appropriately in situ they will be found effective or non-effective rather than true or untrue". Fine, but then we can simply test this effectivenes, as above.

Now, I am not an NLP expert, though I do have an interest. However, I am a scientist, and feel well qualified to comment on the techniques of scientific study. Thus it would be my inclination to clean up the points I have made above and incorporate them into the main article. In the interest of fairness, politeness, and in the awareness of my own possible ignorance, I first await a response from the NLP community.

--Jonathanstray 16:08, 17 July 2007 (UTC)

I've rewritten the "Comparison With Science" section to clarify the difference between being a science (which NLP is not) and being scientifically testable (which it is.) --Jonathan Stray 20:23, 19 July 2007 (UTC)

what does poor scientific appreciation mean?

 * "Equally (as scientific researchers have pointed out), attempts have also been greatly obfusticated by other factors, not least of which are poor scientific appreciation of the subject being researched, failure to fully consider, control and understand all key variables, unrealistic claims by some practitioners, and often, lack of high quality experimental design."

... "poor scientific understanding of the subject being researched".

I know that one of the criticisms of NLP studies is that they didn't really understand NLP... as such I assume the article saying "poor scientific understanding of the NLP process being researched" is one of several reason most studies are not clearly supportive (or unsupportive) of NLP?

Just wanting to clarify. Greg 04:19, 20 October 2007 (UTC) ps. "obfusticate"?

NLP and "scientific" research
Before making my point may I just congratulate everyone involved in creating the page as it now is. Even in those parts which I personally don't agree with there is, it seems to me, a far more constructive approach and a seeming desire to reach a useful and accurate understanding of the subject. Thanks to you all for that.

Next, if there are any "grannies" out there who think I am trying to "teach them to suck eggs" (i.e. stating the blindingly obvious), please excuse me. This is meant for those readers who may not be familiar with research as applied to "pure" psychology.

And so to my point - why the field of NLP is not readily amenable to scientific investigation, leastways, not right now.

The process known as the "scientific method" is based on certain presuppositions.

Firstly it is assumed that what is to be tested will be consistent. If I hold a genuine magnet low enough over a pile of iron filings some of the filings will become attached to the end(s) of the magnet. This won't just happen 10% of the time, or 64.2% of the time - barring any kind of interference it will happen EVERY TIME.

Secondly it is assumed that if I give someone else accurate instructions on how to carry out my experiment, and they follow them accurately, then they will get the same results that I got. If I tell you how to obtain a genuine magnet and genuine iron filings, and I tell you how high to pile the filings, and at what height to hold the magnet, then not only should your magnet pick up filing, it should pick up roughly the same amount of filings as I picked up in mine.

Thirdly, it is assumed that all important variables can be controlled.

Thus we typically have a "dependent variable" - DV -(the one we want to test), an "independent variable - IV - (the one with wich we intend to influence the dependent variable) and other variables including those which we recognize beforehand could interfere with the results of the experiment - extraneous variables (EV) - and influences which aren't recognized until after the event - confounding variables (CV).

In the case of experiments in pure psychology the number of possible EVs and CVs is very large and experiments must be so fenced about that researchers end up with a totally unnatural situation, the results of which aren't really of any use to anyone.

Iron filings and magnets don't have arguments with their spouses, hangovers, bad hair days, or paranoia. They aren't suddenly hit by a lack of confidence or low self-esteem. Nor do they daydream or otherwise lose concentration/focus. And they don't misunderstand the experimenter's instructions whilst saying that they know exactly what is expected of them.

It must be recognised that the whole "scientific testing" business is based on an incredibly simplistic and wholly unsupported premise: that EVERYTHING can ultimately be explained in terms of physical basics - molecules and atoms, etc. - and that therefore everything can be explained in purely physical (i.e. consistent, controllable) terms and according to the standard physical laws, so far as we know them.

As I say, this belief has not been verified beyond reasonable doubt, and there is thus no reason to suppose it must be true. Indeed, our experience, whilst by no means an infallible guide, suggests that we are genuinely capable of acting outside of any neat and tidy laws. Thus human beings themselve can be the EVs and the CVs who screw up nice, neat little "scientific" tests literally just by being there.

Under such circumstances, then, it really doesn't make sense to assume that EVERYTHING can be meaningfully/usefully tested in the scientific manner.

And it makes even less sense to assume that a set of experiments by students still waiting to achieve a master's degree, with little or no accurate training in, or knowledge of, the field of NLP (a description that applies to a substantial majority of the researchers quoted, especially those cited by Heap and Sharpley), could somehow come up with incontrovertable evidence of a scientific nature that accurately reflects on the validity or otherwise of NLP.

So whilst this page has become substantially more civilized than it once was, the discussion still seems to be largely concerned with chasing theoretical hypotheses (or hypothetical theories) instead of getting down to a useful discussion of where NLP and NLP-related techniques and applications can be useful, and where they can't.

NLP's Resistance to Science
I must say I find some of the comments above to be quite disheartening.

For example:

The process known as the "scientific method" is based on certain presuppositions.

Firstly it is assumed that what is to be tested will be consistent. If I hold a genuine magnet low enough over a pile of iron filings some of the filings will become attached to the end(s) of the magnet. This won't just happen 10% of the time, or 64.2% of the time - barring any kind of interference it will happen EVERY TIME.

The individual who wrote this is simply wrong. The scientific method doesn't assume that what is to be tested will be consistent. In fact, an important part of science is the realization that many things aren't consistent at all. Part of the purpose of statistical analysis in science is to put a number on how consistent things are. And imagine if his statement were true! If medical science depended on absolute consistency in its results in order to move ahead, nothing would get done. Tell the creators of Chlorpromazine that their drug needs to work 100% of the time, like a magnet over iron fillings.

In the case of experiments in pure psychology the number of possible EVs and CVs is very large and experiments must be so fenced about that researchers end up with a totally unnatural situation, the results of which aren't really of any use to anyone.

While the issue of mundane realism in psychological experiments is important to keep in mind, the comment here seems rather short-sighted, particularly the conclusion that such research isn't "of use to anyone." The value of such research shouldn't be weighed by the usefulness of individual studies. Scientists don't spend all their time coming up with life-changing discoveries that will benefit lots of people, at least right away. Science is usually a lot less glamourous, and less immediately applicable than that. Some typical examples might be researchers who spend their time studying a single brain protein for years without even knowing what kind of cell it comes from; or studying changes in brain activity when specific stimuli are introduced in a lab; or the behaviour of a particular sub-atomic particle at 99.9% the speed of light.

Science doesn't have to be "useful" to be worthwhile. It allows us to understand the world as it truly is. As an added bonus, it also allows us to cure diseases, fly around the globe, and do pretty much everything we do on a daily basis. But those benefits are the result of a slow accumulation of data, from hundreds of thousands of scientists, who all work on a little piece of the puzzle.

NLP is in an interesting point in its history. Those who practice it make some very interesting claims as to what it can do. Some in-depth empirical investigation of concepts such as representational systems, as well as techniques such as "anchoring", or the "swish", could add quite a bit to our understanding of cognitive processes. Unfortunately though, it seems like every discussion of scientifically studying NLP is padded with statements explaining why it wouldn't be possible, or "useful".

...poor scientific appreciation of the subject being researched," [i.e., "they just don't get it"]

...it is necessary to ask whether NLP is expected to act like a science, or whether it acts more like a black box in which the only effective measure is to statistically evaluate quality of output for a given input.

Some even resort to blatant anti-science comments: "It must be recognised that the whole 'scientific testing' business is based on an incredibly simplistic and wholly unsupported premise: that EVERYTHING can ultimately be explained in terms of physical basics - molecules and atoms, etc. - and that therefore everything can be explained in purely physical (i.e. consistent, controllable) terms and according to the standard physical laws, so far as we know them."

I think I understand the reason for this resistance to science in the NLP community. Isn't it just an underlying fear that the cold, skeptical, scientific endeavour will just dilute or undermine NLP? It seems like a lot of the resistance to science is from the perspective that the researchers' primary aim is to "prove" or "disprove" NLP, rather than solidify our understanding of it. Sadly, this attitude will probably just widen the gulf between NLP and science, and it will be an all-out pseudoscience before long (if not already). Meanwhile the cognitive sciences will continue to uncover the machinations of the human mind. Those findings that are consistent with NLP will be trumpeted as scientific evidence for NLP (there's an abundance of this sort of thing in the List of Studies on NLP), but any scientific findings that contradict NLP will likely be passed off as somehow invalid. And with that, NLP will have joined the ranks of crystal healing, homeopathy, and UFOlogy.

Sad, really.

--PNB (talk) 04:57, 12 December 2008 (UTC)


 * Sorry for adding nothing useful to the discussion, but I just read this, and I had to say, Great Post. Very balanced and well-spoken. Axlrosen (talk) 02:48, 3 March 2009 (UTC)

Psychology is NOT a "science" in the sense that physics, chemistry, biology, etc. are sciences. The reason is simple: psychology (unless it includes some physiological endeavour - neuropsychology, etc.) is dependent on being able to control the variables in a given context. Now most of the so-called "scientific" experimentation on alleged "NLP" claims has been to do with some aspect of the PRS concept and the predicate matching technique. Yet as far back as 1987, two Harvard psychologists, Monica Harris and Robert Rosenthal, examined the research reviewed by Sharpley and found that it was impossible to conduct either a single- or double-blind experiment to test the predicate matching technique (see http://www.bradburyac.mistral.co.uk/nlpfax32.htm) for precisely the reason I have just given.

Set that alongside the fact that nearly all of the experimenters were using the wrong definition of the PRS concept, and it is obvious that the research reviewed by Sharpley and Heap, which is recycled over and over again in subsequent criticisms, and it quickly becomes clear, to anyone who understands what they are reading, that the academic criticisms are founded on quicksand rather than concrete, so to speak. See http://www.bradburyac.mistral.co.uk/nlpfax28.htm. NLPers do NOT resist scientific investigation - they simply note the bias and flawed methodology of the research rhat is presented as scientific - but isn't.

Any rebuttal of these statements would do well to cite actual evidence rather than simply regurgitating those errors yet again. —Preceding unsigned comment added by 81.156.163.198 (talk) 07:19, 3 August 2010 (UTC)

Redirected
Per comments above, and discussion with others. This article duplicates everything that is in the NLP and science section of the article on [Neuro-linguistic programming]. Hope that helps. Peter Damian (talk) 18:28, 26 December 2008 (UTC)


 * The main article is meant to be a summary of what is on this page. Action potential t c 12:50, 30 December 2008 (UTC)

NPOV
We really need to rewrite this article for NPOV. It is currently just a collection of the negative findings and negative criticism. We really need a balanced overview of the research to date. Action potential t c 05:45, 3 January 2009 (UTC)
 * Yes but all criticism would be negative, wouldn't it. As mentioned on the NLP talk page, this is an encyclopedia, not a promotional vehicle for some dubious therapy.  If you want the 'positive' view of NLP, just Google the term.  You see this article now stands out in a class of its own, thankfully. Peter Damian (talk) 20:08, 3 January 2009 (UTC)
 * There are conflicting perspectives that must be presented fairly under wikipedia NPOV policy. There are many errors and omissions in this document. Action potential t c 04:24, 4 January 2009 (UTC)

In support of NLP: http://www.inspiritive.com.au/nlp-research.htm Critical of NLP: http://donaldclarkplanb.blogspot.com/2007/03/nlp-no-longer-plausibe.html

Cheers.--Loodog (talk) 04:50, 2 March 2009 (UTC)

I've made a first attempt at merging what was at main article about NLP and science into this article. We have more space on this article so we can expand flesh out some of those points. Action potential discuss contribs 10:45, 16 May 2009 (UTC)

What does NLP claim?
An article titled "X and science" is pointless unless it begins by outlining what X claims on scientific grounds. After reading a fair amount of stuff, here and elsewhere, I feel no closer to clearly understanding this. A clear statement of claim, by some primary contributor or proponent, would go a long way to making this easier treading.

Lacking the statement of claim, the only traction I was able to manage came from this quote: NLP may be seen as a partial compendium of rather than as an original contribution to counseling practice and, thereby, has a value distinct from the lack of research data supporting the underlying principles that Bandler and Grinder posited to present NLP as a new and magical theory.

From the main article: "found little or no empirical basis for the claims about preferred representational systems". Nowhere can I find an explanation of what specific claims PRS made/makes. Am I the only person who thinks that a page that functions under the title "the compendium of NLP and science" has a lot of 'splainin to do? &mdash; MaxEnt (talk) 12:26, 5 January 2010 (UTC)
 * I agree. AJRG (talk) 10:34, 27 February 2010 (UTC)
 * We could nominate the article for deletion. -- Snowded TALK  10:37, 27 February 2010 (UTC)
 * Before that we could take Heap's own advice and read Andrew Bradbury's article in Skeptical Intelligencer. AJRG (talk) 11:23, 27 February 2010 (UTC)

For background, Andy Bradbury is devastatingly accurate. I spent some years reading and applying NLP techniques, and, indeed, the techniques "worked." I was very aware of what Bradbury points out, over and over, that Bandler and Grinder made a crucial distinction between actual practices and effectiveness and the conceptual models that were presented by "master therapists," such as Perls, Satir, and Erikson, and this would obviously apply to NLP as an approach as well as anything else. My own experience is, of course, anecdotal. But the results were little short of astonishing. That does not establish NLP as a "science," but as more of an "art." The articles on NLP should reflect this understanding, based on reliable sources, and care should be taken about avoiding defining NLP so as to set up straw man arguments. --Abd (talk) 23:00, 27 February 2010 (UTC)


 * Which is the point. If NLP is making scientific claims then it is subject to scientific tests.  If it is just claiming the "people report it works" and saying that this is the only appropriate form of validation then its a pseudo-science and/or something like Dowsing where the explanation, while it might convince followers, does not stand up to experimentation.  The same type evidence supports many a belief system.  -- Snowded  TALK  06:25, 28 February 2010 (UTC)
 * It might help to separate effect from explanation. For example, western science acknowledges the effect of acupuncture but so far rejects the explanation. AJRG (talk) 06:44, 28 February 2010 (UTC)
 * Acupuncture is an interesting one - its effect that be physically verified. Its not a purely reported effect.  So if NLP purely relies on the "it worked for me" argument then it is the same category as people at a religious revival meeting.  -- Snowded  TALK  06:58, 28 February 2010 (UTC)
 * The Eye Accessing Cues Model bears comparison. Since well before NLP there have been attempts to unravel the fact that observers can extract information on mental processes from eye movements, and that eye movements affect quality of recall.  None of the explanations (NLP or otherwise) has yet stood up to scrutiny, yet the effects remain. AJRG (talk) 07:49, 28 February 2010 (UTC)
 * Possibly, although the question of non-verbal and visual clues is exploredin cognitive science and know that there are aspects of electro-magnetic and chemical transfer which also impact. Also the issue here is what does NLP claim qua NLP-- Snowded  TALK  08:29, 28 February 2010 (UTC)
 * And in many other branches of science. Eye accessing cues are part of a much larger canvas, but researchers study them nonetheless.  As to what NLP claims, in its own words, it's probably better to restrict that to the publications of Bandler and Grinder. AJRG (talk) 09:10, 28 February 2010 (UTC)

Since I (Andy Bradbury) find I have been cited several times on this page I hope you will excuse me for leaping directly into the discussion with a few observations:

1. Snowden writes: "If NLP is making scientific claims then it is subject to scientific tests. If it is just claiming the "people report it works" and saying that this is the only appropriate form of validation then its a pseudo-science".

This is not, however, the description I believe is used by genuine scientists. In fact, if I am correct, a pseudo-science is something which claims to be scientific but does not conform to the standards shared by physics, chemistry, etc. Which means that whole field of psychology, except those elements which have a neurophysiological basis, are in fact pseudo-scientific.

But does it even make sense to try to apply the same standards to both the "hard" sciences and to psychology?

A magnet, is a magnet is a magnet. It doesn't refuse to work when it has a hangover. It doesn't work especially well if it is in a good mood, or because you offer it a pay rise. It doesn't do what it does because of peer pressure (from other magnets in the same room), or because it believes that it is God's will that it should pick up iron filings. And scientific experiments don't show results like: "50% of the magnets functioned consistently, 30% only worked in a relaxed environment, and 20% didn't work at all".

A genuine magnet, unless it is subjected to unusual external conditions, works in a clearly defineable, consistent manner.

Now try applying the same characteristics to EVERY human being.

In fact psychology and science are clearly two quite distinct fields of investigation - though many psychologists are desperate to be thought of as genuine "scientists". Not least because it's a whole lot easier to get research grants for projects which promise to produce "hard" results.

2. It should be noted that although Dr Heap has done the largest review, covering over 60 papers against Sharpley's initial 15, for example, still I have Heap's personal assurance that he based his articles on the abstracts, not on the full papers/articles. Some of these abstracts are ridiculously short and/or uninformative and this seems *to me* like a rather unsatisfactory approach. Not least because from the abstracts alone I calculate that it APPEARS that a significant proportion of the reports are partly or wholly favourable to the NLP-releted position even though they are based on very confused accounts about the FoNLP (the field of NLP - NLP itself (a specific modelling technique), the NLP-related techniques and their applications, and training in NLP and/or the NLP techniques).

Sharpley, on the other hand, has assured me (in early 2009, and I have no reason to doubt him) that he read the whole of each article he reviewed. Thus Sharpley was making his evaluation on the basis of much more in depth information than Heap had access to.

3. In regard to the claim that critics of NLP often do not understand NLP, let me cite Sharpley's first article (1984) as a perfect example.

Sharpley treated just two topics in that review - predicate matching and eye accessing cues. Yet he treated them as though they were essentially all there was to the FoNLP. At one point (page 246), for example, he says:

"... if NLP [sic] is suggesting that counselors who demonstrate high levels of reflection and empathy will be more effective than those who do not, then little new is being said. If NLP [sic] seeks to promote empathetic responses from counselors, then scales designed to measure empathy ought to, and do, show this is (e.g. Hammer, 1983). Although this is a worthwhile procedure for counselors, it does not justify NLP as a separate theoretical position ..."

Now look at what Sharpley was actually doing, as compared with what he seems to have thought he was doing:

(a) He confirmed that the emphasis on creating rapport was an important element of effective communication, at least within a counseling set up. Thus supporting rather than refuting a genuine NLP-related claim (though he doesn't seem to have been entirely sure whether it is a genuine NLP-related claim or not).

(b) He assumed that NLP and the NLP-related techniques are being offered as something new - even though he made it perfectly clear, elsewhere in the article, that he knew that Bandler and Grinder based at least some of their ideas on observations of Satir, Erickson and Perls (Sharpley, 1984, page 238). So isn't it sheer commonsense to expect that their ideas will to some extent reflect existing practices?

In other words, far from offering negative evidence on the subject, Sharpley is actually confirming a degree of agreement between the FoNLP and "conventional" psychology.

Having said that, the reference to Hammer, though I'm personally not quite clear what Sharpley was trying to say there, is another excellent example of an experimenter and a reviewer not understanding the thing they were investigating.

Most of the original researchers (in Sharpley's 1984 review) apparently assumed that someone's Preferred Representational System (PRS) was pretty much fixed. Which in 1976 was indeed what Bandler and Grinder were saying (see, for example, The Structure of Magic II, Bandler and Grinder; and Changing with Families, Bandler, Grinder and Satir), both published in 1976).

Hammer, however, was smart enough to recognise that this "fixed PRS" didn't seem to work, and so he tried an alternative approach - tracking and matching. That is to say, instead of assuming that subjects would use the same PRS throughout the period of the experiment, Hammer tracked each use of a sensory predicate and responded to it in the same modality - regardless of what had gone before. This, he reported, brought better results.

Strangely enough, both Hammer and Sharpley treated this as evidence that the "NLP" claim regarding PRSs was in error. Totally ignoring the fact that Bandler and Grinder had themselves observed the same thing and, by March 1978 at the latest, had changed the emphasis on PRSs in exactly the way indicated as more effective by Hammer's findings (see, for example, Frogs into Princes, Bandler and Grinder, 1978/1979. pages 34 and 36).

In other words, in 1983, when Hammer's results were published, and Sharpley first submitted his review for publication, they were both at least 5 (five) years out of date in their understanding of what Bandler and Grinder were really saying about the way to use the predicate matching technique.

By the same token, Sharpley insisted, several times, that "NLP" offered three ways of determining someone's PRS - by watching their eye movements, by listening to their use of sensory predicates, and by self-reports (see, for example, Sharpley, 1984, page 242). Yet in both The Structure of Magic II (1976) and in Frogs into Princes (1978/1979), Bandler and Grinder clearly stated that in order to determine a person's PRS you need ONLY listen to their use of sensory predicates. Moreover, Sharpley made this claim despite his previous acknowledgement that:

"They [eye movements, use of sensory predicates, and self-reports] are not interchangeable, with verbalizations emerging as the only procedure that shows any reliability."

(Sharpley, 1984. page 246)

(Interestingly enough, Sharpley arrived at this correct assessment on the basis of an experiment which claimed to have demonstrated that a person's PRS was fixed! Though Hammer's results also supported "the use of ongoing predicate matching over prior determination of clients' PRS" (Sharpley, 1984. page 245).)

For a far fuller exploration of the way that academic psychologists have misunderstood and misinterpreted/misrepresented the FoNL over the last 30 years see http://www.bradburyac.mistral.co.uk/nlpfax28.htm.

STOP PRESS: Indeed, there is an item in the latest (at the time of writing) issue of Scientific American Mind (March/April 2010) which criticizes the idea of identifying students' "learning styles" - and confirms an authentic NLP-related claim in the process:

"... studies show that students' learning styles are difficult to reliably identify, because they differ greatly across situations.",

(in the ironically entitled Busting Big Myths in Popular Psychology, Scott O. Lilienfeld, Steven J. Lynn, John Ruscio and Barry L. Beyerstein. In Scientific American MIND, Vol. 21, No. 1. March/April, 2010.)

The title is ironical because at least three of these authors - Lilienfeld, Lynn and the late Barry Beyerstein have all criticised some aspect(s) of the FoNLP on the basis of misinformation. And are now asserting yet another myth of their own making.

In this particular case the authors presumably do not know that it is a basic concept in the FoNLP that the current PRS (CPRS) will be influenced by the context, and by any changes in the context, and provides a method for tracking and responding to those changes (see above)! —Preceding unsigned comment added by 86.146.93.207 (talk) 15:29, 28 February 2010 (UTC)

It should also be noted that, in response to an enquiry from myself regarding the article, Professor Lilienfeld categorically denied that the comments had anything to do with NLP or the FoNLP. Andy Bradbury.


 * Andy, resorting to a purely phenomenological approach to verification is not science. The same claims could be made by any cult.  -- Snowded  TALK  10:06, 3 August 2010 (UTC)
 * I was under the impression that the leading proponent of the phenomenological approach was the Large Hadron Collider project at CERN. Are you claiming that isn't science, or did I miss something?  AJRG (talk) 10:56, 3 August 2010 (UTC)
 * In fact, the phenomenological approach goes back at least as far as Werner Heisenberg's 1942 "S-Matrix theory" (a phenomenological approach in quantum electrodynamics) and continues through Shigeo Minami's Phenomenological Approach to Explain the Behavior of the Diffraction Peak in High-Energy Scattering Phys. Rev. 135, B1263–B1269 (1964) and Phenomenological approach to hadron interactions at high energies Phys. Rev. D 14, 1874–1882 (1976) and on to Yinon Arieli and Yoel Rephaeli Dark matter profiles in clusters of galaxies: a phenomenological approach New Astronomy Volume 8, Issue 6, August 2003, Pages 517-528 whose abstract says "In this paper we explore a purely phenomenological approach to determine dark matter density profiles that are more consistent with observational results." AJRG (talk) 08:43, 7 August 2010 (UTC)
 * Different use (and it is a much abused word), you are really talking about abductive or hypothesis generative approaches in the material above. In this case we are talking evidence which is entirely reliant on reported effects.  The same approach (as has been pointed about) would validate any conversion experience to a cult as "scientific.  -- Snowded  TALK  08:48, 7 August 2010 (UTC)
 * When two particle beams collide, the result is not repeatable. It is, however, measurable and a sequence of results can be statistically analysed.  In some ways it resembles a series of "Big Brother" where different personalities collide.  Again, the interaction is observable and can be analysed.  There are legitimate concerns about rigour, but the phenomenological approach is definitely part of science.  AJRG (talk) 10:25, 7 August 2010 (UTC)
 * In that sense yes AJRG, but not in the sense that it is used in the University of Surrey reference. Different disciplines and areas of study use words in different ways (ontology would be another).  -- Snowded  TALK  10:46, 7 August 2010 (UTC)
 * I can't see any reason why this sense wouldn't apply perfectly well to research into NLP, so why do you think they're using it differently? AJRG (talk) 11:20, 7 August 2010 (UTC)
 * Its not my word AJRG, it was used in a UoS reference. In that context they were placing reliance on self-reported effects from individual humans and as I said earlier that would make a "mercy seat" session scientific.  I can't see how the methods you mention could be applied to something like NLP, but always open to ideas.-- Snowded  TALK  11:36, 7 August 2010 (UTC)
 * NLP started with the analysis of taped therapy sessions. I don't see a problem with self-reported effects as long as the statistics are done properly - it's no different to an opinion poll or a customer survey, both of which can be mined (with appropriate caveats) for useful qualitative and quantitative data.  If you combine video of what happened with reports of what everyone experienced, and repeat the process enough times, how is that different from conventional phenomenology?  AJRG (talk) 12:22, 7 August 2010 (UTC)
 * An opinion poll makes no claim to "science" and you haven't responded on my cult point which is key here. Otherwise let me look at a paper which examines NLP with the same rigor as the physics you mention above.  -- Snowded  TALK  12:44, 7 August 2010 (UTC)
 * Opinion polls do make claims to science (Q5), but not to exactness. Last time I looked, statistics was an entirely respectable part of science. I didn't respond to your "cult" point because it's just a straw man.  AJRG (talk) 14:11, 7 August 2010 (UTC)

Recent edit by AJRG‎
The above user has failed to abide by WP:BRD with his/her recent change. AJRG (a largely single purpose editor around NLP articles) has been trying to insert this material into the main NLP article in order to cast doubt on Heap's work. The source there was an article in a collection published by a NLP promotional organisation. Having failed on that page (and broken AP:AGF and a bit of forum shopping with a spurious COI report they have now brought the change here. Regardless of the value of the change, it should be debated here first if it is disputed.  -- Snowded  TALK  10:51, 27 February 2010 (UTC)
 * Heap (1988) (http://www.mheap.com/nlp1.pdf - ref 25) is quite clear about the nature of his references : "NB the abbreviation DAI refers to Dissertation Abstracts International." AJRG (talk) 11:04, 27 February 2010 (UTC)
 * Happy to agree that, but I really don't see what value the addition makes here. Come to think of it, the number 60 is not really needed either.  Content issue aside you really need to learn to act in a collegiate way,  WP:BRD is there for a purpose and you have blatantly ignored it. -- Snowded  TALK  11:06, 27 February 2010 (UTC)
 * Tu quoque... A proper assessment of what scientific research has to say about the actual claims of NLP is appropriate to Wikipedia. Partisan defending of entrenched positions is not. AJRG (talk) 11:13, 27 February 2010 (UTC)
 * Heap himself (here http://www.mheap.com/nlp.html) recommends Bradbury (2008) Neuro-linguistic programming: Time for an informed review. Skeptical Intelligencer, 11, 14-27 (here http://www.bradburyac.mistral.co.uk/response.html) as a rebuttal of his papers. AJRG (talk) 11:21, 27 February 2010 (UTC)
 * (ec)OK please add to your reading list WP:OR (as well as WP:AGF, WP:BRD and WP:COI). Heap's article makes a series of statements which are supported from his paper.  In this edit (taking your comments at face value)  you decided to extract from the article a split between sources.  That is OR.  On the NLP page, you did not break WP:OR in introducing the point but used a source instead.  That avoids OR.  However the source was disputed as coming from a publication by an organisation devoted to the promotion if NLP.  You might want to strike the "Partisan" point as its another case (on top of many) of you attacking the editor rather than dealing with content issues.  -- Snowded  TALK  11:26, 27 February 2010 (UTC)
 * On your Bradbury point, please note this is a web site. If there is reliable third party material that is a rebuttal of Heap then it should of course be included.  However I strongly suggest that you use the talk page in future if an edit is reverted-- Snowded  TALK  11:26, 27 February 2010 (UTC)
 * Skeptical Intelligencer is the journal of the Association for Skeptical Enquiry, of which Michael Heap is Chairman. Not an NLP publication. AJRG (talk) 13:29, 27 February 2010 (UTC)
 * If a citation references apples and pears, it's not OR to mention them separately. Since this is supposed to be about science, it matters whether studies have been peer reviewed. AJRG (talk) 11:35, 27 February 2010 (UTC)
 * And you are choosing the relative merits of Fruit over types of fruit. The point is that you making that selection is OR.  -- Snowded  TALK  11:37, 27 February 2010 (UTC)
 * I didn't make the selection - Heap is explicit about it. AJRG (talk) 11:45, 27 February 2010 (UTC)
 * All Heap does is to say at the top of the references "DAI denotes Dissertation Abstracts International" which is a source guide. He says nothing about the value of that material, and makes no mention of the distinction at any point in the text (If I have missed something tell me).  You are the one deriving a specific meaning here - and that is OR -- Snowded  TALK  11:55, 27 February 2010 (UTC)
 * You missed off the "NB" at the beginning of the statement. Claiming that a caveat on the references isn't part of the text is artificial at best. Heap is careful to make the distinction. AJRG (talk) 12:09, 27 February 2010 (UTC)
 * Still your interpretation of what "NB" means, if it was a caveat in the sense you are using it then it would be normal practice to have said as much in the main text of the article and he doesn't. Look, you really need to get your mind around this.  It may or may not be material to the article, but even if I agreed with you that it was, then it could not stand.  You need to find a source that says it is relevant, and that source needs to be reliable etc. etc.  -- Snowded  TALK  12:13, 27 February 2010 (UTC)
 * NB stands for Nota bene meaning "note well". AJRG (talk) 13:18, 27 February 2010 (UTC)

"The Irish National Center for Guidance in Education"
Googling on the above yields only the following links:, , ,. Of these, the first is a Wikipedia page on NLP, the second is a (now deleted) page in wikipedia userspace, the third seems to be a mirror of the main wikipedia page and the fourth is a separate wiki. Dropping the initial "The" gives only two more links: and. The former seems to be a mirror of wikipedia and the latter may be another. In short, the alleged organisation has no web presence, a rather odd thing, making me wonder if the organisation is: Autarch (talk) 18:15, 14 April 2010 (UTC)
 * Very obscure.
 * Defunct.
 * Working under a different name.
 * Otherwise non-existent.
 * Look here for an explanation. AJRG (talk) 20:09, 14 April 2010 (UTC)

Condensed intro
The intro reiterates what already is stated in the NLP article. I suggest a direct introduction into the NLP-science argument and a summary of such. —Preceding unsigned comment added by Clay Juicer (talk • contribs) 05:34, 21 June 2010 (UTC)