Psychology Wiki

NLP modeling (or modelling) is a process used in neuro-linguistic programming to discover and codify patterns of excellence as demonstrated consistently by top performers in any field, ideally via direct experience (although modeling from books, historical records of people's words, or video is not unknown). It can be thought of as the process of discovering relevant distinctions within these experiential components, as well as relevant sequencing of these components necessary to achieve a specific result.

A "modeling project" involves spending time studying and observing in depth, discussing, and imitating and practicing many different aspects of the subject's thoughts, feelings, beliefs and behaviors (ie, acting ""as if" the modeler is the expert) until the modeler can replicate these with some consistency and precision. Once this has been achieved, the modeler then refines the target skills by removing certain features to eventually discover the essential features distinguishing average performance and top performance, thus building a learnable/transferable model, and tests it by seeing if it can be taught.

The NLP theory behind modeling does not state that anyone can be Einstein. Rather it says that know-how can be separated from the person, documented and transferred experientially, and that the ability to perform the skills can be transferred subject to the modelers own limits, which can change, and improves with practice.


In a way, modeling is the core skill in NLP, which was itself created based primarily on models by John Grinder and Richard Bandler of three famous psychiatrists:

NLP has a broad view of what constitutes a "skill", and in this sense anything people do can be construed as a skill, since they have developed an ability to do it. In this context, "modelable skills" have included: Olympic shooting, Trance induction, waking up without an alarm clock, motivation, legal advocacy, creative thinking, and even clinical conditions such as schizophrenia or depression.

(To imagine how the last two count as "modelable skills" in this sense, one would start by asking questions such as, "how does the person manage to do this internally? How do they maintain it? How do they generate those behaviors?")

Analytic Modeling means applying the distinctions which are often used in NLP, such as the meta model, meta programs, values, beliefs, representation systems, etc in order to analyze the behavior of a model. When combined with research methodology from psychology, analytical modeling can be used for scientific research. (eg see the NLP inspired modeling research methodology at

According to John Grinder (the co-founder of NLP) there are 5 steps to the modeling process:

  1. Identify an expert (outstanding performer) that you want to model. Create a well-formed outcome and life lines for this project.
  2. Arrange to observe and mimic while the expert is in the field of expertise (direct access is the best option, then video, then high quality audio).
  3. Without presence of the expert, the modeler practices until the same set of responses can be elicited of a similar quality in a similar context. The modeler removes certain features to discover the essential elements while practicing to find out what works, and what can be dropped out. At this stage the modeler may require to return to step 2 to observe and mimic some more before continuing.
  4. Codify the model in some way via video or some other medium.
  5. Teach the model to someone else. If the student can then elicit the same set of responses with the same or similar quality as the original expert, then this is a workable model. If not, the modeler may have to return to step 2 or 3.

Another means of modeling comes from the work of Milton Erickson, and is known as deep trance identification, where a profound hypnotic trance is induced and within that, a person is led to recollect everything they have ever known or seen of a person, and then to "step into" and identify with them completely, experiencing their world completely. In such a state, it is said, the unconscious mind creates its own rich understandings by sidestepping the need for conscious analysis.

Note that modeling refers to the internal processes and external behaviors. It does not state that everyone will be Einstein, or an olympic champion (although equally, it does not set limits on what may be possible). Rather, it states that in principle:

  1. What one human being can do, in principle any human being can learn a similar approach.
  2. There are more (and less) efficient, and more (and less) well designed ways to do any task.
  3. People develop their own ways to do things, and their own beliefs what can be done, so some people will have developed more effective ways than others, and this can be changed by learning better ways.
  4. The actual know-how can be transferred as an experience, not just a theory.
  5. The ability to perform the skills can be transferred subject to limits set by external circumstances such as environmental or physiological factors, and improves with practice.
  6. Any limits which may exist on what can be learned or improved are unknown, may be unexpected, and can change.

Detail and examples

Typically a "modeling project" might cover the following sources of behavior:

  • Beliefs
  • Values
  • Internal strategies
  • Outcomes
  • Sensory perceptions and submodalities
  • Physiology (body movement and body language)
  • Language patterns
  • Fall-back strategies ("what if it isn't working")
  • Conscious and unconscious communications
  • Perceptual positions
  • Locus of consciousness (ie where ones attention is)

Each of these is individually a deep and rich field; there is no point where one knows everything, but as a process of replication, the goal is met when the modeler has enough parts of the puzzle to piece together and document how the subject seems to be doing his competent skills.

Many trainers stress that fully identifying with the model ("embodied" modeling) is an essential part of the modeling process. However, when NLP practitioners do any modeling at all in practice, it is often Analytic Modeling.

Ideally, the result is that the modeler feels that the information of "how the skill is done" is sufficient, and the rest is practice or external limitation rather than understanding of the process.

Other uses include:

  • Modeling how a therapeutic client maintains and engages in their "problem behavior", with the intent of learning enough to change it for the better
  • Modeling famous or dead people to gain a sense of how they did what they did, and their views and beliefs which allowed them to do so. (Robert Dilts is a proponent of this process, having described models of notable people such as Jesus of Nazareth, Sherlock Holmes, Albert Einstein and Nikola Tesla)

Criticisms of modeling

Some NLP proponents have criticised modeling from other than a living person, and believe that with no access to the model (nor quality video), it is not possible to test whether the model is accurate.

Similarly, scientists have criticised NLP modeling, stating that it is impossible to determine a "correct" model, and that applying one particular model to everyone is over-simplistic, and will be no substitute for hard earned expertise (Carroll, 2003). NLP proponents reply that in common with most forms of heuristic approximation, there is not intended to be "one correct way", but only more and less effective and transferable models.

Referring to "The Map is not the Territory", one of NLP's main presuppositions, NLP proponents reply that in common with most forms of heuristic approximation, there is not intended to be "one correct way", but only more and less effective and transferable models.

Other information

  • Robert Dilts and John Grinder argue for a distinction between Analytic Modeling and "NLP modeling" that requires unconscious uptake via imitation. (announced October 17 2005-[1])

External links

This page uses Creative Commons Licensed content from Wikipedia (view authors).