Jeezum… from the blog about stuff Integral (aka Ken Wilber et al) Mystery of Existence,
Dangers of models
In writing the aqal review of local organizations, and also talking with a friend yesterday who’s very much into integral things, I am reminded of the dangers of Spiral Dynamics, and of any map, framework or model.
As with any map, or sets of ideas, it can be taken as a relative or an absolute truth, it can be used in service of shadow projections, and it can be used with more or less heart and empathy.
Relative and absolute truths
The clearest danger is in taking it as an absolute truth, to mistake the map for the terrain, to put more faith in and emphasis on what the model says rather than what the terrain is doing.
Seeing any map as a relative truth, it becomes a tool of temporary and practical value, an aid for navigating and functioning in the world. There is nothing absolute about it. Just a tool that works more or less well in any situation. A tool with no inherent value, which can be modified and discarded as needed. It remains secondary to the terrain, to life itself.
“Just a tool that works more or less well in any situation.” No, a model works for exacting reasons in only situations for which those reasons hold. Always the question begged by a model has to do with how these reasons are commensurate with a precise situation where the utility of the model is demonstrable.
The model is not the thing modeled. A model describes and may depict operations. If you could start up a model of an internal combustion engine it would not be a model.
The description is not the thing described. The danger implicit in any model comes about from their reification. This turn immediatly causes a category error. It does so even if the utility of a model lends itself to the description of predictable consequences of the operations of the thing modeled.
Models may be more or less accurate but they cannot be absolutely true. Their accuracy refers to the accuracy of their description, and given predictive utility, their predictions. But, models are by their categorical (or domain) nature are always, by definition,always reductive.
Every model’s utility is constrained for many reasons but one of the primary reasons is that for every model the implicate description is not complete, nor can it be completely accurate. It is okay to qualify the truthfulness of the description compilation and the model itself, but absurd to say a model is itself and otherwise relatively or absolutely true. A model is true relative to a qualified frame of reference. If you can think of a universal model unhinged from a relation to a frame of reference, please let me know!
And, now I must go on at length about a very pet peeve.
The AQAL model rests on assumptions and all of its assumptions are necessarily constrained. Although it is purported to be a comprehensive model, the decisive class of questions about its comprehensiveness has to do with what it leaves out, and why stuff is left out. Most complex models are in this situation.
(Simple models may highly correspond with the thing modeled, even be analogous, isomorphic. Some models, including complex ones, are highly predictative because their operational characteristics are well understood and this understanding includes their rate of error, etc., and includes also where their utility is counter-indicated.)
AQAL is, for lack of a better term, a human relations and actions model. It also promotes and aligns with Spiral Dynamics. SD is also a human relations and actions model, and, it is a typological model too. (I use here ‘actions’ in lieu of the term, behavior.) Typological models are often deployed via a pernicious pseudo-utility, presuming in their use to both capture deterministic characteristics and decisive operations.
To unlock my criticism a digression is required. An accurate model compiles accurate constructs so a good model is an accurate constructive model. But, this accuracy is always constrained for many reasons. A good reason comes up when constructs are accurate at describing what happens but not accurate at describing what structures underpin what happens.
For example, one can describe what happens when an internal combustion engine is running and translating heat into mechanical force without describing the structure of the mechanism. This provides for a model of the engine yet it’s clear that to blend structural constructs into the description will increase its descriptive accuracy.
Another example closing in on the problem. Consider psychology and its models. Consider too the informal models of the psyche people carry around in their heads. Throughout the history of psychology, (in fact this history could be told in these terms,) various models have been utilized to describe and represent mental and behavioral phenomena. But, although it is possible and valuable to come to agreement on what the constructive elements of those models describe and indicate, it is only with the interest–given to cognitive psychology–in correlation to physiological structures, that the structural guesses of the preceeding depth psychology has been largely deposed because, as models go, they are not very accurate or correlated.
This doesn’t mean, for example, psychoanalysis or analytic psychology aren’t valuable. But it does mean that their constructs are no longer very good descriptions of structural aspects of human behavior. For example, the term “ego” is a valuable construct when it is smartly qualified, but it does not correspond to a coalescence of physical phenomena in the brain from which it can be deduced the term “ego” is the construct correlated with the structure, “ego”. There is no structural “ego”. The best that can be said for ego is that there is a proscribed utility to classifying behaviors as being ego-like.
The problem of reification in this case comes about from presuming that a mechanized model accurately describes the dynamic relationship between constitutive parts. Those presumed parts were named in early depth psychology because vitalism had given way to models which implicated a kind of causal hydraulics, or mechanics. The features of the presumed mechanics allowed for plugging named parts into structural models. Better descriptions have mitigated the utility of those precursor models.
Alas, if a person believes intensely in those old models, it’s easy to project them on one’s own mind, and onto other minds as if they remain accurate. There’s a paradoxical utility that remains workable too. Therapy, for example, can use models of behavior and be effective even when the structural constructs don’t correlate or even exist. From this it’s easy to understand how reification is mistaken where the structure is presumed to exist.
A model of mental function may implicate ego-like behavior, be predictative about the consequents of thse behaviors, yet not in any way require an accurate structural description also be a part of the same model. So, we navigate a world in which people instantiate big ego-like behaviors, yet, such people do not possess big ego (parts). Still, it is common to presume that such big parts underpin the big behaviors.
This mistake is clearly revealed when it is fused to the attribution error. This error holds that the person with the big egoistic behavior is always walking around with the concomittant big ego part. They are never relieved of having this big ego!
This error of attribution is pervasive in well-known circumstances. (Circle back to the fundamental reductive assumption of any model from which it, unfortunately, follows.) For example, there are any number of assessments and instruments used to identify characteristics, propensities, features of the personality. Many people working in professional organizations have to take MBTI, 360’s, Learning Style, and other kinds of evaluative tests. Each of these tests is based in a constrained model having to do with the class of behaviors being assessed and typified.
The kicker is that their utility is also always wed to the attribution error! In fact their very presumed and mistaken utility requires this error. I test as an ENFP (MBTI II) and it is presumed I will go to bed an ENFP, show up for work an ENFP, and, not only this, but that I be surrounded with people who’s own typologies effectively and mutually reinforce the positive features of my being an ENFP.
It’s important to understand the predictation of positive results in this example follows from the utility implicate in the model just as it is not possible that the model provides a temporally holistic accurate description of the aspects it evaluates. Holistically accurate means at all times.
Extroversion is, per force, reductive because, aside from its evaluation, there is no robust means for determining what combinations and blends of mental characteristics constitute the rich set which reduces to the less rich classification, ‘Extroversion’. It is a revealing experiment to do: ask any person who wishes you to take an assessment, “What are the working assumptions and descriptions of the model?” and you will quickly learn they do not have the foggiest notion of what are the actual components of the model.
Descriptive categories are often used to whack people. To this day I have yet to learn from either a credentialed or lay Jungian what the heck they mean by “intellect” and “feeling”. This is true given that Dr. Jung’s own discernment of the differences and distinctions is no longer in the least bit robust, and, as wonderful as his |Psychological Types| is, it contains little about how his bluntly dichotomized descriptions are themselves constrained. This same problem is ripened in the widely used MBTI.
It is an aspect of the rich flux of relational life that everybody possesses folk psychological working models of behavior and these are relentlessly projected upon other people. I have long grouped people into the two classes of either doing this thoughtlessly or thoughfully, and, doing this well aware it is error prone. Furthermore, many people can be found (out!) presuming that there own model is widely, if not, comprehensively shared. This is always a mistake and always cannot help but be a mistake.
The AQAL Model is good only as much as it provides for some kind of accurate description or prediction. The former objective always implies the ways in which the model’s descriptive utility is constrained or blinded or inappropriate.
Among the principle deficits of both AQAL and Spiral Dynamics, just taken for what each is, a model, is that their categories tend to rest on collective and situated approximations, and they’re, in turn, lent to highly reductive qualifications of what are in effect, “types”. Neither strikes me as being a very robust model. Still, the questions I would argue on behalf of are the ones about why one descriptive class is justified and another is not.
AQAL or Spiral Dynamics can be misappropriated in service of a utility unable to be validated. In the latter case it already is well known that adherents see the world memetically and see the world full of Red and other SD Meme personalities.
The aqal model is often presented as only of interest to second tier folks, or maybe even only understandable by second tier folks.
An operational feature is reified in this remark. Incidentally, the author disagrees this is true and i would go farther and suggest anybody who thinks it is true is being ridiculous. Were it true, then the AQAL Model supposes a description of cognitive capacity that holds that short of obtaining this capacity, the very description of the capacity would not be understood. The operational aspect presumes the description exists only due to the operation of a non-normative understanding.
This is, to my mind, the conceit of esotericism. This is different than a physicist noting their are levels in a physical model which suppose high level mathematical skills is a requisite. That requirement is easy to demonstrate. Whereas I would propose the esoteric (and self-referential appeal to a capacity of consciousness,) qualification supposes no real demonstration whatsoever.
There’s no model of the mysterium. Models become more robust the more their constraints, blind spots, incapacities, and descriptive and predictive deficits are understood. Models become better when they become more refined. Often their utility decreases as their robustness increases, which is to say the better the model is the better its applications are, even if those applications have to shrink.
Following from all this, models evolve and are subject to conceptual revolutions, paradigm shifts and the dynamics of increases in knowledge. The user of thrilling and edgy models is no less in the situation of more modest model users in that their use should rest on a sound argument on the models behalf as well as an allowance for the model itself changing, even changed to the extent of being overturned, vanished.
Models should be judiciously used, not mistaken for the thing modeled, and, at the beginning of every day, each of should laugh at the certain prospect that we will injudicously use our own pet folk psychological models as if the this model was the think modeled!