| |
 |
 |
 |
 |
 |
 |
 |
by John Grinder
I find myself in the somewhat awkward position of responding to the work and writings of a person � Michael Hall - that have nearly no intrinsic interest to me. I will respond to the best of my ability on this particular occasion since he took the time and courtesy to address certain comments made by Carmen Bostic St. Clair and I in our recent book, Whispering in the Wind (see www.nlpwhisperinginthewind.com).�
I have the impression that Mr. Hall and I are working in entirely different fields with profoundly different criteria for presentation, argumentation and evidence, and with significantly different purposes. In Whispering in the Wind (Bostic and Grinder, 2002), we note, for example, that the term Neuro-Linguistic Programming itself - NLP) has become something of a wild card. We offer the suggestion that publications and discussions would be significantly improved in their coherence and quality by the simple device of identifying whether the aspect of NLP involved is NLP modeling, NLP application or NLP training.
Nowhere in his article does Mr. Hall identify which of these domains he is operating in nor is it clear to me that he is operating in any of these. Clearly, NLP modeling is not the issue as there are no comments by Hall that touch on this, the heart of the NLP activity. Hall refers to his own productions as Neuro Semantics -� perhaps this is the point. We are operating in entirely different fields � an observation that would go a long way toward explaining the confusion I experience when reading Hall�s productions.
I offer the following responses: there are five points that I can distinguish that Mr. Hall presents and to which I am willing to make a response. These are:
I did not write Whispering in the Wind (www.nlpwhisperinginthewind.com) � I co-authored this book with Carmen Bostic St. Clair. It is a collaborative effort in the finest sense of the word. I would request that Hall recognize this simple fact..
Korzybski produced a rich piece of work: indeed, one that is capable of supporting multiple interpretations. We had and have no intention of proposing that the man who established the map/territory distinction does not understand the map/territory distinction as Mr. Hall states. Rather what we propose is a more refined representation of the complexities in the mapping from receptor to what has come to be known as mental maps (the term is unfortunately as it suggests mental as a separate category thereby supporting implicitly the Cartesian split). Quoting Mr. Hall,
�FA is John's new terminology. It stands for First Access and refers to how we first access the outside world, not through our language or even representation systems, but prior to that, through the sense receptors of our eyes, ears, skin, and other end receptors.�
�������� ��������� ��������� ��������� ��������� ��������� ��������� Michael Hall
Hall�s characterization of FA is flawed. This is not �John�s new terminology�, it is a proposal crafted by Carmen Bostic St. Clair and John Grinder as part of an attempt to make explicit an important research distinction. In the quote above, moreover, Hall specifically excludes any representational systems representations from FA. How it is possible to have access to any internal representations (FA) without representational systems remains beyond my apparently limited imagination. Carmen Bostic St. Clair and I did not make any such proposal as a perusal of pages 9 � 49 and again on page 57 where the reader will find,
In previous work in NLP, especially by Grinder and/or Bandler (in, for example, Patterns of the Hypnotic Patterning of Milton H. Erickson, M. D. or Neuro-Linguistic Patterning, or Turtles All the Way Down), this privileged level of representation was referred to as the 4-tuple.
����������� ����������� ����������� ����������� ����������� Whispering in the Wind, page 57
The 4-tuple has for decades been the point in the neurological processing where the representational systems first are displayed and become available for consideration both at the conscious and unconscious levels of functioning.
Hall correctly points out that Korzybski posited a series of levels of abstraction. However, and this is precisely the point, these levels of abstraction proposed by Korzybski are mute with respect to the refinements proposed in Whispering. In particular, we distinguish between two sets of transforms: f 1, consisting of all those transforms that occur between receptor and our first access (FA) to the resultant neurological events; and f 2, the set of transforms that occur post FA (for example, linguistic mappings). We argue extensively that these two sets of transforms operate by distinct processes (see especially pages 28 � 40 in Whispering) � what Bateson was fond of referring to as two distinct logics. To fail to recognize this distinction (either in the form we propose in Whispering or some even more refined version of this) will constitute a grave flaw in any research program.� All the quotations from Korzybski offered by Mr. Hall are congruent with this conclusion.
Hall poses an apparent contradiction concerning our critique of his putative expansion of the meta model. He begins by quoting from The Structure of Magic, volume I
"... our Meta-Model covers only a portion of the verbal communication which is possible..." (p. 107)
"... we suspect that some of the research currently being conducted in Generative Semantics ... will be particularly useful in expanding the Meta-Model further." (p. 109)
The first quote refers simply to the fact the meta model was created by Bandler and myself for the express purpose of re-connecting the words presented by a client that fail to refer to the actual reference experiences from which they are presumably derived (through f 2 transforms) with the attendant positive consequences of involving clients in actively expanding their maps and consequently generating new sets of choices in precisely those portions of their maps where they lack choice. This meta model is a meta model for the specification of language.
Bandler and I were well aware of this express purpose of our creation. Further, we were perfectly content to note there exists a rich set of meta models (in the linguistic sense) for achieving other purposes. This was the motivation for the comment we made that Hall quoted. Indeed, the two volume work entitled The Hypnotic Techniques of Milton H. Erickson, M.D. written by Bandler and myself (J. Delozier is a co-author for the second volume) contains precisely such an example � a distinct linguistic meta model whose purpose is to structure verbal communications that enhances responses at the level of the unconscious mind � a point that Dr. Erickson had a complete grasp of as well.
Hall�s selection of the second quote from Magic has an interesting historical flavor. In late 60s and early 70s, I was a member of a group of linguists (Postal, Ross, McCawley, G. Lakoff�) whose purpose was to extend the formal power of the work developed initially by Noam Chomsky in syntax into an arena called semantics with the same rigor that we had succeeded in patterning portions of the syntactic component. A slightly more radical objective in this endeavor was to demonstrate that the line initially drawn by Chomsky establishing a boundary condition between syntax and semantics was artificial and that while this boundary had served an excellent historical purpose it now constituted an obstacle to further development in transformational grammar. The group involved worked under the name of Generative Semantics.
These were heady times, and Bandler and I predicted that an extended formal analysis of semantic issues would reveal additional distinctions worthy of incorporation into the meta model that we had expressly created for the purpose of specifying language forms. Our prediction was not confirmed � in fact, the entire Generative Semantics enterprise collapsed not long after the publication of Magic.
As Carmen Bostic St. Clair and I describe in Whispering, out of the ashes of this fiery collapse, an entire new field has arisen � cognitive linguistics which offers fascinating studies that cross precisely that line of division between syntax and semantics that was the target of the work of the Generative Semanticists. Carmen Bostic St. Clair and I will forego any prediction concerning future contributions of cognitive linguistics to NLP (in any of its aspects) and content ourselves with an invitation for NLP practitioners to note advances in this field in hopes of discovering patterning of utility in our endeavor.
It is prudent to mention to the reader unfamiliar with the history and current status of models in linguistics and related fields that there is no connection between this historical movement in linguistics called Generative Semantics and General Semantics (Korzybski�s work and movement) nor between Generative Semantics and Neuro-Semantics (Hall�s title for his own work).
In his article Hall then poses his contradiction,
Hmmmm, "useful in expanding the Meta-Model further" was Grinder1975 and Hall1997, a point that Grinder2001 now has problems with and argues conflicts against. Now he wants to reduce the model rather than expand it. Of course it is perfectly fine to change one's mind. I have no problem with that. But how fine is it to encourage "expanding the Meta-Model further," provide the justification for it, and then demand that I have to justify it now without reading my justifications for it or remembering that he himself began NLP on the note of expanding it?
����������������������� ����������������������� ����������������������� ����������� ����������� Michael Hall
I have not changed my mind in this matter. I still find myself in full argument with Grinder1975.
Further, Hall is being quite disingenuous here as the material quoted comes from a section in Whispering which also contains specific comments that would resolve his contradiction: namely,
Rather than expand a model already proven effective in securing some outcome, X, the task of a modeler is to attempt to reduce the model consistent with achieving X � that is, to demonstrate that X can be achieved with fewer distinctions or more efficiently�
Therefore, we would propose that anyone who wishes to argue for the inclusion of additional verbal patterning would accept the challenge of motivating their inclusion in the model. More specifically, such motivation would demonstrate that there are useful outcomes in addition to X that the inclusion of these proposed additional patterns allows that strictly speaking are not achievable through a congruent application of the original model. The only other motivation we can imagine would be a proposal to replace some or all of the patterns in the original model by some other set of patterns that are more efficient or more effective in achieving X.
����������� ����������� ����������� ����������� ����������� ����������� ����������� Whispering, page 186
I offer a somewhat larger frame and then respond with more precision to this apparent contradiction. As we argue extensively in Whispering, NLP from its inception has been a technology for modeling that extreme of human functioning called excellence. Of late, it has drifted more than a little, with vastly more attention and time given to applications and training. Of course, NLP application is a worthy and useful activity � indeed, if there were no application work occurring, the justification for NLP modeling would come into serious question.
At its core, then, NLP modeling� is the mapping of tacit, implicit knowledge onto an explicit and transferable model. By definition, all models are reduced versions of the thing they purport to represent. The challenging processes involved in mastering a set of patterning through the unconscious uptake of the patterning of excellence from the initial model (the person displaying the patterning of excellence � Dr. Erickson, for example) through the coding of such patterning once the modeler has demonstrated his or her ability to elicit from the relevant portions of the world of experience the same responses with roughly the same quality and within the same time frames constitutes the fundamental task for the modeler.
In the case of the modeling of the verbal behaviors of Perls and Satir by Bandler and myself, for example, our task was to distill the effective portions of their verbal behavior out of the vast array of verbal productions they used into explicit and learnable patterns of verbal specification. This mapping, clearly a reduction of the set of verbal utterances used by these two famous therapists, had as its purpose the presentation of the minimal effective set of verbal distinctions in a form easily learnable by interested parties.
The question, then, is what would justify the inclusion of additional patterning in the meta model. While the quote from Whispering presented above is to me adequate, apparently it was not effective. I will therefore expand on it as follows: there are two clear methods for justifying such additions to the meta model. Both of them involve the identification of some syntactic pattern, s i, that occurs in the speech patterns of clients but which is not effectively challenged by any one (or any sequence) of the 13 verbal patterns that constitute the meta model. Given the identification of s i, we could subsequently propose a specifying pattern (a challenge to this identified syntactic form that is effective into re-connecting it to the reference experiences from which it was originally derived).
I identify two possibilities that would justify or motivate an argument for the addition of a verbal specifying pattern to the meta model:
a. the strong case: the identification of some syntactic pattern, s i, which when challenged for specificity produces a refinement in the mental maps of the clients in a way that enriches it and generates new choices.� To be fully compelling (the strong case considered here), the argument would have to demonstrate that no application of any pattern or any sequence of patterns already present in the meta model effectively challenges this particular syntactic pattern
b. The weak case: the demonstration that there exists some syntactic pattern, s i, in the language of clients for which the pattern of specification proposed for inclusion in the meta model is superior by some explicit criteria for the specification of s i � more precisely, superior to the application of any pattern or sequences of patterns already present in the 13 distinctions in the meta model.
In this second case, there already exists challenges in the meta model that are applicable to s i but the argument would run that the new challenge proposed for incorporation into the meta model makes the model more efficient or more effective (by some set of explicit criteria). While less compelling than the strong case above, it would be a worthwhile contribution. At a minimum, it would offer practitioners a stylistic option. Remember, being more efficient is ultimately a question of style. Erickson was not particularly efficient when measured by the number or depth of changes per hour or per hundred words uttered or per metaphor, but then he never professed much interest in such efficiency measures � he had a quite distinct style.�
I take it that the criteria for additions to the meta model are now perfectly explicit and invite anyone interested in the challenge to make a proposal.
Now, a critique by Hall demonstrating that the functional consequences of what Bandler and I presented in The Structure of Magic, volume I could be achieved with fewer distinctions would have been interesting. But his movement is, unfortunately, in the opposite direction. I therefore take this opportunity to offer a challenge presented by Carmen Bostic St. Clair and myself in Whispering that is relevant to this process of finding the minimum number of distinctions that deliver the specification so critical to clients in NLP applications:
In our own work, it has become clear to us (Bostic and Grinder) that it is possible to achieve X, the same set of outcomes achievable by the meta model with only two of the original verbal patterns � the noun specifier and the verb specifier. We propose this as the minimum set. Further, we note as we argue in the text (see especially chapter 1, Part III), that there are competing requirements in the modeling of such phenomena � for example, while it is possible (according to our claim) to achieve every outcome that was achievable with the full original meta model with the reduced set of two mentioned above, it may be far more effective for training purposes to include patterns other than the minimal set.
However each trainer decides to approach the presentation of verbal patterning, we leave the challenge before the community: identify an outcome that is achievable with the original meta model that is not achievable with the reduced set proposed here.��
����������������������� ����������������������� ����������������������� ����������� Whispering, page 186/187
By the way, a perusal of the �extensions� to the meta model by Hall (see, for example, appendix B, pages 105 � 108, Advanced Flexibility Training, 2000, �patterns� 14 through 22) by any mildly well-trained NLP practitioner will reveal that all of the examples offered there are well-handled by the set of patterns in the original meta model. The peculiar thing is that insofar as the �extensions� offered are intelligible, many of them explicitly use the original meta model� patterns � thus, it is far from clear what Hall thinks he is doing. Here are a couple of examples from his �extensions� (please note that these are Hall�s examples, not mine):
����������� Pattern 15. Static words (SW)
����������� ����������� Science says that�
����������� The amazing extension by Hall for challenging this sentence is,
����������� ����������� What science specifically?
����������� Pattern 16. Over/under defined terms (O/U)
����������� ����������� I married him because I thought he would make a good husband.
����������� The original meta model challenge would be
����������� ����������� A good husband, how specifically?
����������� which gets you there a quicker than Hall�s
What behaviors and responses would make a good husband for you?
I leave the analysis of Hall�s other alleged �extensions� to the meta model as an exercise for the reader. All of them fail the explicit tests presented above for extending the meta model. Thus, the �extensions� to the meta model proposed by Hall fail to meet either the strong or the weak case versions and are therefore entirely unmotivated.
The challenge to determine whether the minimalist strategy (the two questions identified in the quote above) is also left to the reader.
4. Meta-states:
The proliferation of meta states offered by Hall may or may not be useful for him. In general, I confess that I am somewhat suspicious about proposals that move us further away from FA or direct experience � the only reliable source of correction for our errant musings. However, of far more significance than my personal response to such endeavors, consider the following:
I take it that it is uncontroversial to state that the term meta is roughly translatable as about and that any statement containing the term meta (for example, meta state) will require a specification of what it is meta to. Another (and I hope, equally uncontroversial) way of glossing the term meta is in terms of scope. I will accept Hall�s favorite application of the term meta for purposes of illustration. Thus a meta state is a state that is about X, or equivalently, has X in its scope (it covers X). Presumably, then, in the case where X itself is another state, we have the situation where the meta state, m i, for X is a state about the state X (or again, equivalently, m i has state X in its scope.
Note that m� i� with respect to X in this schema will have some (possibly all) of the features that state X has plus something else (the meta contribution apparently). Now as far as I can determine there are no constraints on what this something else might be. When I actually examine the examples that Hall offers what strikes me is that he appears to be using this meta relationship to sort out various aspects of the original state X � roughly along the lines that Bandler and I proposed in far simpler and more sensory grounded formats in The Structure of Magic, volume II some decades ago.� The �addition� proposed by Hall is to do this is some form of a hierarchy (the about relationship). This inherently hierarchical ordering is as yet undefined. It is clearly not, a logical level hierarchy - it fails the tests presented in Whispering for such hierarchies (see Hierarchical Ordering in Whispering page 285 and succeeding pages and especially Logical Levels beginning on page 294).
Until some explicit mapping is offered by Hall for his about relationship, it is impossible to determine what he is talking about and therefore to intelligently evaluate his proposed contributions � we trust that there will be more substance to this than his claim about �extensions� to the meta model.
There are two statements by Hall in the section about meta states in his article that I find easy agreement with,
�Perhaps this riotous proliferation of higher states is the magic and dynamic of what's emerging in Neuro-Semantics.��
Good, perhaps so - clearly in Neuro-Semantics, there are lots of meta states. May they find a secure home there.
�I show how that meta-states can be sick, morbid, and toxic and the very structure of self-sabotage.�
I sincerely hope that Hall enjoys his continued focus on meta states in Neuro-Semantics.
I found nothing in Hall�s remarks relevant to the request/challenge offered in Whispering so I will content myself with simply repeating,
If there is some serious intention involved here, specification of the terms, psychologically encompassing and impactful is required to allow the rest of the world of NLP to participate intelligently in the discussion.
����������� ����������� ����������� ����������� ����������� Whispering in the Wind, page 347
What Carmen Bostic St. Clair and I were requesting from Mr. Hall in Whispering is an explicit mapping from these undefined terms,� psychologically encompassing and impactful onto some relatively sensory based representation that would allow us and others to appreciate what is being proposed and thereby arrive at an intelligent decision about their utility.
Final Comment
In a recent publication (cited previously as Advanced Flexibility Training, 2000, page 48) Hall states,
The Uncertainty Principle (Heisenberg, 541) This fundamental principle in science enables us to adopt a style for more comfortably living with change.
For those readers unfamiliar with Heisenberg�s excellent work, his Uncertainty Principle refers to a fundamental limit to measurement. Crudely put, one cannot measure with precision both the location and the energetic state of a particle. How Hall gets from this basic finding in physics (the material, non-living world) to his above rendition is so far beyond my willingness to imagine, that I will simply pass.
The gloss by Hall of Heisenberg�s work has about as much value as his comment on page 8 (Advanced Flexibility Training) that,
����������� � the grandfather of NLP, Count Alfred Korzybski�.
Is Hall seriously proposing that all the modeling of excellence and the ensuing explicated� patterns of excellence that Bandler and I coded in creating NLP were somehow already in Korzybski�s work? Korzybski coded a powerful perception � the map-territory distinction. Congratulations and full stop!
I have no interest in pursuing additional conversations with Mr. Hall (I am wary of his use of #1 in the title of his article, Response to John Grinder #1) as I am presently of the opinion that NLP (in all its aspects) has a minimum overlap with Neuro-Semantics.
I wish him well in his endeavors and respectfully request that he clearly distinguish in the future between these two endeavors � only one of which I wish to pursue.
John Grinder
Co-creator of Neuro-Linguistic Programming
Note: Thanks to Carmen Bostic St. Clair, Jeisyn Murphy (www.Got-NLP.com) and Michael Carroll (www.nlp-academy.com for their helpful comments on this article.