How do you respond to the comment "Neuro-Linguistic Programming is not empirically proven."? by Daryll Scott
Before entering into this debate, it's important to distinguish between not empirically proven and empirically disproven. The underpinning paradigms of NLP and 'normal science' such as psychology differ so fundamentally that the current process of research leading to what is considered to be empirical proof is unlikely to produce a favourable result in addressing the activities described as Neuro-Linguistic Programming (NLP).
The University of Surrey is conducting research into NLP and holds its first academic research conference into NLP later this year1, and the conference organiser Dr. Tosey approaches the study with the reservation: "We may need to stay open to the possibility that NLP may not work whilst adding to the research." Ironically - the criterion upon which every coded pattern of NLP is judged is "does it work in all cases; yes or no?" If it does not work it is never coded. Remember that NLP is not theory put into practice; it is the explicit coding of a tacit ability that already works in practice. (By the way, the activities being researched are inadequately defined at the outset of the project - if this is not the case, I invite someone from the research team to define what activities are being addressed here, distinguishing clearly between (1) NLP modelling (2) The coded patterns discovered as a result of the NLP modelling process (3) The coded patterns that are unique to NLP).
From conception the field of NLP set out to avoid the corruption, bias and limitations of the activity called Normal science - the mopping-up operations of making deductions from the current paradigm and then running experiments to verify the predicted phenomenon and refine the measurements involved in an attempt to force nature into the preformed and relatively inflexible box that the paradigm supplies2. In other words, normal science sets out by creating a theory from which hypotheses are generated and then attempts to prove or disprove them. If you set out with a defined hypothesis, it will come from within the realms of your current thinking. You then set out to prove or disprove the hypothesis, and that's precisely what you get. Therefore, the aim of doing research is to discover what is supposedly known in advance.
When the outcome of a research project does not fall into this anticipated result range ('significance' is not obtained), it is generally considered a failure and therefore not published. Only studies that support what is expected are published and this serves to support the underpinning paradigm or theory. By contrast, in NLP - the exceptions to the rule are the studied phenomenon.
There are no theories in NLP; it is pure sensory observation of stimulus and response. Instead of working from a theory, the NLP agent will utilise resourceful and flexible, operational presuppositions. These presuppositions are not considered to be 'true', they are evaluated and employed on the basis that acting 'as if' they are true can often provide the agent of change with a resourceful frame of reference (intention and attention direction). In the absence of theory, psychologists may begin to theorise, and in doing so we are no longer studying NLP, we are now studying psychology. The underpinning paradigms are so fundamentally different that a psychologist is likely to convert observations from the field of NLP into psychological theory in order to study or 'prove' it. (If any researcher has been able to overcome this I invite you to explain how the research will be conducted in the absence of theory or hypothesis.)
One of the things a scientific community acquires with a paradigm is a criterion for choosing problems that, while the paradigm is taken for granted, can be assumed to have solutions. The intrinsic value of a research question is not a criterion for selecting it, the assurance that the question has an answer is the criterion. "Normal-scientific research is directed to the articulation of those phenomena and theories that the paradigm already supplies. To desert the paradigm is to cease practicing the science it defines. The man who is striving to solve a problem defined by existing knowledge and technique is not just looking around. He knows what he wants to achieve, and he designs his instruments and directs his thoughts accordingly". Thomas Khun.
The challenge is that, to be accepted, new assumptions and paradigms require the reconstruction of prior assumptions and the re-evaluation of prior facts. This is difficult, time consuming and typically strongly resisted by the established community as education within the established field will have a deep hold on their thinking. We would be interested to know what anyone intending to research NLP would do to overcome this.
In NLP we recognise that there are very few processes within the human system that can be isolated in order to be tested individually. Most of the phenomena we observe seem to have multiple observable physiological shifts occurring simultaneously. Creating the conditions that isolate one element of a pattern in order to study it is likely to remove an element of the pattern that is, for the individual, necessary to achieve the result. It's like studying a pair of scales and just paying attention to one side of the fulcrum, whilst changes could be happening at any time to the other side of the fulcrum which will cause the results on the side you are studying to vary wildly.
If you would like to take this investigation further then read on. We first need to distinguish between three areas of NLP: NLP modelling, the application of patterns coded as a result of the NLP Modelling process and the patterns that are unique to NLP3.
Modelling is the mapping of tacit to explicit knowledge. The intention is to discover the difference that makes the difference between a genius and an individual of average performance in the same niche. At the outset we have moved some distance from sciences that use statistical tools to discover that which is average or 'normal'. We are studying the individual behavioural quirks that produce extraordinary results, something that is generally elusive to the masses.
Rather than theorising about what the individual may be doing (how can we know? and given the extraordinary nature of the model, best guesses could easily be way off) the NLP modeller will use their own mirror neuron system to replicate and assimilate the behaviour of the subject with no attempt to consciously understand. The criterion for evaluation of this process is real performance - when the modeller can achieve the same results as the modelled genius in the same time or less. The modeller then makes the assumption that the implicit patterns of behaviour have been captured and begins to test each definable unit of behaviour by isolating it and removing it to see if it negatively effects the results - if removal of a pattern has a detrimental effect on the activity it is therefore significant, rather than an idiosyncratic and irrelevant behaviour of the subject, and is coded.
Rather than building the model one brick at a time (and guessing what the bricks might be - with no method of testing them because the isolated pattern alone may not achieve the overall effect) the NLP modeller captures the entire model - all of the bricks, and then sorts through them to identify the ones that are significant with the ability to test them against the overall results of the modelled behaviour.
NLP modelling is however an imperfect process. Although it captures information more quickly than theorising, and allows the discovery of what's really happening rather than testing what someone may think is happening, at the point of coding the model is subjected to the distortions of the language and explicit knowledge of the modeller.
In terms of modelling what's really happening, with no distortions or confirmation bias on the part of the modeller, NLP modelling is almost perfect, but it falls at the final hurdle.
-
To attempt to empirically prove the application of coded patterns discovered using the NLP modelling process is questionable - these patterns are not NLP. People were employing these patterns before NLP was conceived. NLP is simply the modelling technology employed to discover them. The geniuses from whom these models were coded had backgrounds in more conventional fields, and the patterns have been demonstrated to work, why else would they have been identified as geniuses in their field? What difference would it make if they were empirically proven or not?
Most well formed models of NLP offer a systematic approach to what is normally an existing intangible process. The models are presented as numbered 'steps' with a visible check at each stage to provide the agent of change with a precise and reliable methodology.
This is a discrete process. At each stage the question can be asked: Did it work; yes or no? If yes, continue, if not - go back to step 'n'.
Incidentally, the process of creating a discrete model that allows clearer calibration, testing and ecology checking, does not detract from the continuous nature of calibration as the subject with whom the agent of change is involved in this dance of continuous 2-way feedback will continuously vignette from one state to another often accessing multiple processes and states simultaneously. Unfortunately for normal science, a live subject is never frozen in time, responding to just one stimulus with one isolated process. It's an explosion of activity with multiple processes running at any one time.
As for the patterns that are unique to NLP, perhaps there is a legitimate case for research into and empirical study of these new patterns, however, there is a degree of skill required in consistently and reliably eliciting the responses required to effectively demonstrate them. In isolating a pattern and constructing a test that removes bias the subject may not demonstrate the pattern, and the observer may not have the observational competence to recognise it when they see it.
Imagine that Gordon Ramsay cooks a great dish of food. The process is coded (into a recipe) and given to 100 amateur cooks. If 50% of them over or under cook the food - does that make the recipe questionable, or is it that the subjects used to test it do not have the basic competence that is required to execute it? My question is; how can you use a competent NLP practitioner and remove bias?
- We welcome any intelligent answers to the questions posed above.
- www.som.surrey.ac.uk/nlp2008
- The Structure of Scientific Revolutions - Thomas S. Kuhn 1970
- For a full and clear definition of the activities encompassed under the umbrella of 'NLP' and the activities of coding them, see Whispering in the Wind - Grinder & Bostic - 2001.
