First published in Rapport, journal of the Association for NLP (UK) Issue 39, Spring 1998
and the Emergence of Background Knowledge
Penny Tompkins and James Lawley
We are please to report that our article Symbolic Modelling in the last issue of Rapport has generated a lot of feedback. In particular, a number of readers asked for more explanation of the following paragraph:
"A subtle distinction we make is between (a) going to Second Position with the person being modelled and what they are doing; and (b) going to Second Position with the information and the way the information is input, processed and output by the person being modelled. The first approach is typical of Sensory Modelling while the second approach is the preferred mode of Symbolic Modelling."
In reply, this article will address the distinction between the two ways of Second Positioning. To do this, let's start with an out-of-the-ordinary example. We'd like you to go to Second Position with the person giving the following description of being an aeroplane, ie imagine you are experiencing what they are describing:
I am an aeroplane about to take off. As I start to move, slowly at first, and then ever quicker, I feel the air pressing against my body as it whistles by, drowning the sound of my whining engines. As I strain to increase my speed, my wingtips vibrate and the buildings on either side become a blur, and then ... my nose lifts, my undercarriage follows and I am soaring up, up, up. The buildings below appear to be shrinking, other planes look like toys. The mosaic of the countryside becomes clear as I climb higher and higher into the sky.
If you associated into this description you will probably have formed pictures in your mind's-eye, heard sounds in your mind's-ear and felt feelings in your body. Chances are, if we compared your Visual, Auditory and Kinesthetic representations to the describer's, they would have had some similarities. If you had been able to match your body movements, eye accessing, breathing and voice qualities to those of the describer, your "sense impressions" would be even more like their VAK representations (although your 'meta responses' might well be different). This is the process of "going to Second Position with the person." Your aim is to be so like the describer that you have a similar internal experience.
Mirroring to Model
Going to Second Position with the person usually entails some form of "mirroring" their external behaviour. This practice is based on the belief that "what happens on the outside of a person is a reflection of what happens on the inside." Thus it is possible to gather information about their internal experience by matching their external behaviour -- and if you are any good at it, it works surprisingly well. What is not normally made clear is that the most important aspects to mirror are the changes in behaviour (eg. changes in location, rate and depth of breathing, changes in eye movements, gestures and other body movements, and changes in voice qualities) as these are likely to correspond to a change in internal state and/or processing.
In modelling you will probably want to mirror all the behaviours in real time and in context, if possible. This process is called "shadowing." And because full-blown shadowing can be intrusive there is a subtler form of mirroring using "micro-muscular movements." Here the modeller matches the behaviours of the subject with very small equivalent body movements which can hardly be observed. This is a very useful and convenient information-gathering process because you still get sufficient VAK sense impressions from which to build your model
Mirroring for Rapport
In NLP, mirroring is often taught as a way of building rapport. The difference with mirroring to build rapport and mirroring to model is that (a) The modeller is not directly involved in the process of the person being modelled (they are not having a conversation for example); and (b) Not all behaviours need to be mirrored when the aim is to respectfully build rapport. For example Michael Grinder says only 50% of body movements need to be matched. [ref. 2 page 65].
There can be unwanted side-effects of mirroring and matching. For example, a term gaining common parlance is "carer-fatigue." This refers to those in the helping professions who experience unpleasant symptoms after being with people in pain and suffering for too long. It is quite likely that the helpers have built enough rapport to have associated into their client's feelings (and patterns) and have "taken on" their emotional malaises.
As an antidote to this we have developed a motto: As long as one person in a therapeutic relationship has access to the unpleasant feelings and any information contained therein, that's good enough!
Second Positioning the Information
Going to Second Position with the information presented by a person is somewhat different to the above. There is no requirement to feel the feelings of the other person, hear their sounds or see their pictures. What is required instead is to code the patterns contained within the information. One marvellous way to do this is through metaphor.
When you go to Second Position with information, you do not expect to get the corresponding VAK representations you would if you went to Second Position with a person. What you create instead is a "perceptual map" of the patterns of the following relationships:
- Spatial (the relative location and movement of 'components' of the experience);
- Temporal (the sequence in which 'events' happen);
- Functional (what the components are there to do, what 'attributes' they possess and how that influences the construct).
These aspects of a person's subjective experience are generally not in the foreground of their awareness. You can't easily see, hear and feel patterns of relationships. Instead the information is in the background: 'behind' the picture, 'between' the words and 'beneath' the feelings.
Going to Second Position with the information allows you to model the Background structure which "holds together" the Foreground experience. The model will have a structural correspondence to the person's experience but it will not be a model of their representations. This is because, by definition, we do not ordinarily form representations of our background knowledge as we are only aware of the Foreground. So when you have constructed a model of a Background process, it will be your metaphoric description of their background knowledge. Thus can be modelled the structure of unproductive symptoms or behaviours of excellence. As Mark Johnson says:
The key word here is 'structure,' for there can be no meaning without some form of structure or pattern that establish relationships. [ref. 4, Page 75]
Background Knowledge is the very
fabric of our reality. Mark Johnson maintains it has been woven from
the "dynamic pattern of our perceptual interactions and motor programs
that gives coherence and structure to our experience" [page 2]. Johnson
refers to one key form of Background Knowledge as "schemata." He
acknowledges Immanuel Kant as the original source and defines these as:
schemata are not rich, concrete images or mental pictures. They are
structures that organise our mental representations at a level more
general and abstract than that at which we form particular mental
images. [Pages 23-24]
schemata will have parts and relations. The parts might consist of a
set of entities (such as people, props, events, states, sources,
goals). The relations might include causal relations, temporal
sequences, part-whole patterns, relative locations, agent-patient
structures or instrumental relations. Normally, however, a given schema
will have a small number of parts standing in simple relations. [page 28]
can use the idea of schemata to get a clearer understanding of the
intrinsic Background Knowledge embedded within a person's experience.
Let's examine how to go to Second Position with the information
presented by a single sentence from a client:
"I can't approach women I find attractive"
In order to model the structure
of his presenting problem we first need to identify the components of
this experience. In this case, there are two or possibly three symbolic
- The "I" that can't approach
- The (category of) attractive women
- The "I" that finds these women attractive
everyday conversation we would presume the two "I's" allude to the same
reference point. Yet if we look at the sentence symbolically, this need
not be the case, as the two "I's" might signify two different persona,
parts, states, times, etc. If this were so, it is likely the client's
non-verbal communication would 'mark out' the difference between the
two "I's." While we cannot be sure until the client further describes
his experience, we can already infer something about the relations between the symbols:
- Can't approach = a desire to 'go near' which cannot be enacted
- Find Attractive = to 'discover or consider' a 'force which draws towards'
Given the metaphors, "approach" and "attract" used by the client, his
can be characterised structurally, as two 'forces' pulling together two
entities. These are kept some distance apart by a 'mechanism' of (at
least) equal 'power' which prevents the entities from coming together.
This sequence of relationships can be diagrammed as:
The spatial relationships and sequence of events depicted in the figure are a figment of our
imagination. And yet this model, premature as it is after only one
sentence, is likely to correspond to the client's experience. If not,
then we know the client has other meanings associated with the words
"attract," "approach" etc. which we would want to further model.
purpose in diagramming this likely set of relationships is to
demonstrate (a) the amount of structural information that can be
embedded in a seven word sentence; and (b) the sort of
structure that must exist in the Background of the client's experience
for their sentence to make sense. It also informs the therapist
(modeller) where to address questions in order to construct a model
which is more accurate, complete and useful. For example, it would be
pertinent to discover:
- How do "attract" and "approach" function?
- Are the "I's" the same or different and if different, what is their relationship?
- How did the whole construction come into being?
- What happens before and after this sequence of events?
- What is the nature of that which prevents approaching?
a strategic therapeutic viewpoint, probably the most valuable
information will come from finding what restrains (or obstructs,
limits, prohibits etc.) him from overcoming that which prevents him
from approaching women he finds attractive!
Given that we want
to preserve the structure of the information in order to model it, we
would ask Grovian Clean Language Questions [ref. 3 and 5], to provide a
context for this information to emerge. A selection of possible
And you can't approach women you find attractive. And when you can't approach women you find attractive ...
... is there anything else about the "you" that finds women attractive?
... what kind of "I" is that "I" that can't approach these women?
... that's "can't approach" like what?
... what happens next?
... what happens just before you "can't approach women you find attractive"
... where could the "can't" of that "can't approach women you find attractive" come from?
Positioning the person gives the modeller VAK sense impressions. These
allow a model of the form and sequence of Foreground representations to
be constructed. On the other hand, Second Positioning the information
enables construction of a model of the Background structure which, by
it's very nature, is metaphorical. And as Gregory Bateson noted:
"Metaphor, that's how the whole fabric of mental interconnections holds together." (ref. 1, page 29)
© 1998 Penny Tompkins and James Lawley
examples of modelling a metaphor see our articles:
And for more on emergence see our article What is Emergence?
1. Combs G & Freedman J, Symbol, Story & Ceremony (1990)
2. Grinder M, Righting The Educational Conveyor Belt (1991)
3. Grove D & Panzer B, Resolving Traumatic Memories (1989)
4. Johnson M, The Body in the Mind (1987)
5. Tompkins P and Lawley J, Articles in Rapport 33, 35, 36 and 38.