Dr. Nardi ([MENTION=14444]AncientSpirits[/MENTION]) —
You've long subscribed to the Harold Grant function stack, which is the one that says the typical INTJ stack is Ni-Te-Fi-Se and the typical INTP stack is Ti-Ne-Si-Fe — with the result that INTJs and INTPs, despite sharing three out of four dichotomy preferences, typically have
no (preferred) functions in common. INTJs and ESFPs, by contrast, are viewed as sharing the same four functions, albeit in a different order, making them similar in ways that have led you and Linda Berens to endorse the same type foursomes that are also reflected in the socionics quadras.
At a reddit AMA in February 2013, you said that the brain activity you were seeing in midlife INFJs and ISTPs was so similar — because there was so much
tertiary Ti and inferior Se activity in the INFJs, and
tertiary Ni and inferior Fe in the ISTPs — that you "could hardly tell" which type was which.
But as I'm forever pointing out in forum threads, the Harold Grant function stack is inconsistent with Jung, inconsistent with Myers, and has never been endorsed by the official MBTI folks — and more importantly, and unlike the respectable districts of the MBTI, that stack has no substantial body of evidence behind it.
You've mentioned "type historian" Peter Geyer a couple times, and as you may know, he wrote an article on the occasion of the 20th anniversary of
Image to Likeness: A Jungian Path in the Gospel Journey — the 1983 book that introduced the Grant function stack to the world. And in that article, Geyer pondered why the Grant stack had gained so much traction, given that it had never been validated to any respectable degree. And he noted that a substantial part of the answer seemed to be that, at a particular moment in the MBTI's history, several influential members of the MBTI community felt the need for a "framework" that was "amenable" to a particular kind of "training environment," and Grant enjoyed a high "personal status" as "one of the earlier users of the MBTI" — with the result that "a specific and influential group of actors picked it up and ran with it."
They "picked it up and ran with it"! They certainly did. And they're still running today, even though the notion that an INFP has "tertiary Si," and will therefore tend to have "Si" aspects of personality in common with a typical ISTJ that ISTPs tend
not to exhibit, is a typological assertion that — more than 30 years after
Image to Likeness, and like all assertions that
crosscut the dichotomies in that counterintuitive way — continues to have no more respectable support for its validity than the notion that two people born at the same time will have aspects of personality in common because they're both Capricorns.
And Geyer, to his credit, viewed the Grant stack's popularity with some serious misgivings, given the lack of empirical support. He said:
Finally, there's the question as to whether appropriate research, particularly crossdisciplinary
research has been attempted i.e are the models under discussion compatible with other research into what is known about human beings?
This is an extremely important question for the validity of anything, but particularly in the field of personality where constructs are necessarily correlated. All too often, and one sees this in the Jungian as well as the type community, there's a sense of being right about what one is doing, as opposed to the desire to investigate this rightness and see whether it succeeds or fails.
One thing that has long struck me as curious about the Grant stack is that it reflects a highly counterintuitive element that I like to call the
Harold Grant Double Flip. The Grant model says that, with respect to the function-related dichotomy pairs — i.e., S/N with J/P, and T/F with J/P — if you start with one type (e.g., an INTP), and you flip
two preferences, the resulting type (e.g., in this example, ISTJ) will be
more like the original type in various ways than the two types you arrive at by only flipping one of those preferences (in this example, ISTP and INTJ).
So, as in your reddit AMA example, FJs and TPs supposedly have "Fe" and "Ti" stuff in common that isn't characteristic of FPs and TJs, and NJs and SPs supposedly have "Ni" and "Se" stuff in common that isn't characteristic of NPs and SJs.
And of course, there's nothing
necessarily wrong with a theoretical element that appears strangely counterintuitive, because sometimes it turns out that reality works in counterintuitive ways. But on the other hand, the more counterintuitive a theory is, the more you'd expect the theorists to feel that it was incumbent upon them to subject their theory to some reality-testing and see, as Geyer put it, "whether it succeeds or fails."
It's maybe also worth noting that Geyer's an INTP — Ti-Ne-Si-Fe under the Grant model — but pointed out that he experiences his sensing function as Se rather than Si. And when he wrote that article, he was well into midlife — that time when you say you've observed so much tertiary
double-flippiness in some of your subjects' EEGs that you can "hardly tell" the INFJs and the ISTPs apart.
In any case, returning to Geyer's issue of performing "appropriate research"... the personality psychology field has been
validating theoretical type groupings for decades now, without any need for brain scans of any kind, and everybody understands how you go about doing that. You come up with an instrument that does a reasonably good (not perfect) job of typing people, and then you see if FJs and TPs (for example) actually do exhibit personality characteristics in common that they don't share with FPs and TJs. And there are countless ways that personality characteristics can be exhibited, from questionnaire responses to hobby choices to behavior in lab exercises (like your workshops).
And here's the thing that I find most curious about your EEG-dominated approach. Brain regions don't have "Fi" and "Se" and similar labels attached to them when you peer inside the skull, obviously, and as I understand it, the way you're determining that the human brain tends to activate in particular ways when the subject is doing "Fi" stuff (for example) is that you're registering the relevant activations happening
at the same time that the subject is doing some
externally observable thing that you consider to be an "Fi" thing. Right?
One of your "brain maps" points to various regions that you identify as associated with things like "admit novelty," and "place personal value," and "attend to literal details," and "aesthetic recall," and "notice errors," and "categorize and define."
So... when you tell us that the "brain activity" of midlife INFJs and ISTPs in your lab is barely distinguishable because they're both lighting up those Ni, Fe, Ti and Se regions (and/or patterns), I assume that that means that those INFJs and ISTPs are, in addition to (and in conjunction with) the EEG readings,
also exhibiting
externally observable Ni and Fe and Ti and Se reactions/responses/behavior/etc. that your INFP and ISTJ subjects tend
not to exhibit (or not to the same extent) when you put them through the same exercises. Is that correct?
And if you're conducting those exercises with those types, and their
externally observable responses are putting them in groups that line up with those Harold Grant Double Flips, then it seems to me that you and your fellow HaroldGrantians should have little trouble validating the Grant function stack
the old-fashioned way — i.e., by correlating the associated type groupings (e.g., FJ/TP vs. FP/TJ) with personality-related responses that show up
outside the subjects' skulls.
As one example, your written materials describe a "verbal creativity task" that asks the subjects to "provide a coherent sentence for each phrase below," and you explain that "subjects with trans-contextual thinking" — which you say distinguishes Ne types from Ni types — "craft sentences from [certain] phrases more quickly, coherently, and creativity [sic] than other subjects."
So... if NPs and SJs are "Ne" types, and NJs and SPs are "Ni" types, then you should be able to provide some EEG-free validity for the Grant stack by having a suitable number of subjects perform that task, and by having the results reflect the fact that not only are the sentences of the NPs (on average) meaningfully different from the sentences of the NJs, but also — and here's where the Harold Grant Double Flip comes in — that, contrary to what Reynierse (and reckful) would expect, the SJ sentences are more like the NP sentences than like the NJ sentences, and the SP sentences are more like the NJ sentences than like the NP sentences.
If it's true — and as I understand it, some serious questions have been raised — that you've come up with a brain scan technique that's capable of respectably registering "cognitive functions" (assuming they exist), then congratulations, and all power to you, and in that case it could turn out that your EEG readings provide
another layer of evidence for the functions (and the Grant stack). But given the existing state of the art, and the fact that your EEG technique hasn't yet established its
bona fides in a way that's likely to make any associated "validity" claims for the functions widely credible, it's a holy mystery to me why you (and your fellow HaroldGrantians) wouldn't be inclined to
first validate the Grant stack with the same kinds of correlational studies that personality psychologists have been conducting for decades.
Buuut separate and apart from that puzzlement... here's another thing I have trouble wrapping my head around, Dr. Nardi. I also don't understand why there should be a need to conduct
any further studies at this late date in order to provide at least
some respectable level of empirical support for the Grant stack. People have been correlating MBTI types with just about everything under the sun — internal/attitudinal stuff, external/behavioral stuff, you name it — for over 50 years now. The 1985 MBTI Manual referred to over 1,500 studies (according to a 1990 review of MBTI research), and was full of correlational data. The 1998 Manual reflected the results of many more studies. And you've linked us to the "boat loads" of articles, etc. in the MILO database.
And the correlational patterns that show up in an MBTI data pool
are what they are, and
don't depend on what patterns the people gathering the data may have been testing, or expecting. So
every MBTI correlational study has the potential to end up providing some validity support for the Grant stack — or any other possible MBTI-related model — if it turns out that one or more of the associated type groupings end up being reflected in the resulting correlations.
Assuming that the eight "cognitive functions" have impacts on someone's personality that make "Ne" types significantly different from "Ni" types (and "Fe" types different from "Fi" types, and so on), and assuming that who's an Ne type and who's an Ni type generally reflects the Harold Grant model, it seems inconceivable to me that in all those thousands of existing MBTI data pools, there aren't any where one or more of Fi, Fe, Ti, Te, Ni, Ne, Si and Se are the most significant MBTI-related influence on whatever the study involves, with the result that the Grant model is reflected in the correlations — e.g., with FJs and TPs on one side and FPs and TJs on the other (if it's an Fe vs. Fi thing, or a Te vs. Ti thing, or an Fe/Ti vs. Fi/Te thing).
But as I understand it, when Reynierse published his articles in the official MBTI journal, and effectively pronounced the Grant stack (among other aspects of "type dynamics") an emperor with no psychometric clothes, not a single HaroldGrantian came forward and defended the Grant stack by pointing to any respectable body of data.
Instead, you wrote that APTi Bulletin article (
"The Case FOR Type Dynamics"), which (no surprise) failed to point to
any respectable empirical support in the form of correlational data — and instead, as previously noted, pointed to your "pilot study" EEG lab results, and had the chutzpah to characterize those results as providing "strong and unmistakable" "quantitative support for the 8 Jungian functions."
Strong and unmistakable quantitative support for the 8 Jungian functions.
And that was in 2009, Dr. Nardi, after you and Linda Berens had spent over 20 years (between you) peddling the Grant stack — and describing it in your writings
not as a tentative, unofficial, yet-to-be-validated offshoot of the MBTI that you'd decided to
run with, but instead as if it was simply
how the MBTI works — or "what the four-letter code really stands for" (to quote Linda Berens).
And if the only (purportedly) empirical-supporty response to Reynierse that you (or any other APTi Bulletin contributor, or HaroldGrantian) could manage to come up with in 2009 was your surprising (there's a gentle word for it) characterization of those lab results, then I'd say that arguably gives anyone all the background they need to gauge whether the Grant stack has a bright future ahead of it.
And the rest of your "response" to Reynierse was pretty much just a hand-waving, straw-manny non-response.
On the validity front, you asked why Reynierse's article didn't "mention the existing validated assessments that are designed to tap the 1 Jungian functions." But if a "validated assessment" only "taps" a single Jungian function, and it types somebody "Si" (for example) by virtue of assessment items that tend to be more characteristic of MBTI SJs than other types, then that doesn't contradict anything Reynierse said — or Reynierse's assertion that all MBTI-related personality characteristics result from one of the four relevant preferences and/or simple additive effects (e.g.,
no double flips) of two or more of the preferences.
And on the other hand, if a "validated assessment" taps a single Jungian function, and the assessment items that get somebody typed "Si" are
not things that tend to be more characteristic of MBTI SJs than other types, then that's a "Jungian" assessment that involves a Jung-related typology that
isn't the MBTI, and Reynierse's article doesn't speak to those. Reynierse's objections to "type dynamics" essentially involve theoretical elements that (1) are inconsistent with (or go beyond) preference multidimensionality, (2) have no respectable empirical support,
and (3) are purportedly associated with the subject's MBTI type.
And #2 is really redundant with #1, since as Reynierse notes, and as far as I know, there is currently no respectable empirical support for
any aspect of "type dynamics" that is inconsistent with (or goes beyond) preference multidimensionality.
And that
emphatically includes the Harold Grant Double Flip, which in turn is the basis for the forum-famous "tandems" (or "function axes"), not to mention the socionics quadras and that exciting new Berensian lens you've mentioned.
================================
It's worth emphasizing, partly for the benefit of readers unfamiliar with the applicable history, that the MBTI owes much of its tremendous success, not to mention its claims to intellectual respectability, to the fact that Isabel Myers devoted a substantial chunk of her life to putting Carl Jung's typological concepts to the test in a way that Jung never had, and in accordance with the psychometric standards applicable to the
science of personality.
And Myers' many years of data gathering demonstrated that Jung had gotten quite a lot wrong in terms of the ways in which many of the aspects of personality he described actually cluster in real people. As McCrae and Costa (the leading Big Five psychologists) have explained:
Jung's descriptions of what might be considered superficial but objectively observable characteristics often include traits that do not empirically covary. Jung described extraverts as "open, sociable, jovial, or at least friendly and approachable characters," but also as morally conventional and tough-minded in James's sense. Decades of research on the dimension of extraversion show that these attributes simply do not cohere in a single factor. ...
Faced with these difficulties, Myers and Briggs created an instrument by elaborating on the most easily assessed and distinctive traits suggested by Jung's writings and their own observations of individuals they considered exemplars of different types and by relying heavily on traditional psychometric procedures (principally item-scale correlations). Their work produced a set of internally consistent and relatively uncorrelated indices.
And those scientifically respectable "indices" Myers produced were the
dichotomies, rather than the functions. Or rather, you might say that Myers' psychometrically valid categories included the
four functions, but not the
eight functions — and with the four so-called "perceiving" and "judging" functions really being the two sides of two of the MBTI dichotomies.
And the 30-plus years of data that's been gathered since Myers died reinforces the notion that the reason it makes sense for the official MBTI indicator to test people based on the dichotomies is not because the dichotomies are a "convenient fiction" for that purpose, but rather because those four dimensions are actually the core, substantially-genetic building blocks of Jungian/MBTI type — while the "cognitive functions" are appropriately characterized as (in Reynierse's words) a "category mistake."
In any case, and more specifically, I know of no passage in
Gifts Differing or any of Myers' other writings where her type descriptions reflect any
double flip effect. Regardless of what dichotomy pair is involved — non-function-related (e.g., NF and ST), or function-related (e.g., SJ and NP), Myers' writings consistently reflect the unsurprising notion that if there's a personality characteristic associated with two preferences, then flipping both preferences will lead you to a type that exhibits the
opposite of that characteristic. If SJs are the
most whateverish types, then you can pretty reliably expect the NPs to be the
least whateverish.
Gifts Differing includes a Type Table that looks like this:
ISTJ | ISFJ | INFJ | INTJ |
ISTP | ISFP | INFP | INTP |
ESTP | ESFP | ENFP | ENTP |
ESTJ | ESFJ | ENFJ | ENTJ |
And Myers explained that "the Type Table is a device for seeing all the types
in relation to each other. It arranges the types so that those in specific areas of the Table have certain preferences in common and hence share whatever qualities arise from those preferences."
And Myers specifically noted, as one example, that "the more resistant types, the thinkers at left and right and the judging types at top and bottom, make a sort of wall around the Type Table; the 'gentler' FP types are inside. The types with both of the resistant preferences, the tough-minded, executive TJs, occupy the four corners."
Double-flip (and
function axes) fans think of TJs and FPs as the
Fi/Te types, but Myers consistently described preference-pair-based type groups with opposite preferences as
opposites when it came to
all of the personality characteristics that she associated with those preference pairs.
And assuming that aspect of Myers' writings reflects reality — and as Reynierse pointed out, that was the state of the data when Myers died, and remained the state of the data in 2009 — then NJs and SPs don't have "Ni" or "Se" things in common, and NPs and SJs don't have "Ne" or "Si" things in common, and FJs and TPs don't have "Fe" or "Ti" things in common, and FPs and TJs don't have "Fi" and "Te" things in common.
And if you can point us to a respectable body of correlational data to the contrary, Dr. Nardi — with respect to
any of the eight functions, and
any aspect of personality — then please do so. And if the large body of existing MBTI data pools is as lacking in support for double flips as Reynierse says, then unless and until you and your fellow HaroldGrantians can manage to produce some kind of respectable support, why are you OK with continuing to peddle those double flips as if they had any more claim to respectability than the zodiac?
You conducted a "pilot study" that was exploratory, involved a very small sample, and (by your own admission) shouldn't be viewed as "proving"
anything. And that's all fine as far as it goes, but... I would have expected a "world renowned ... expert in the fields of neuroscience and personality" (as you've described yourself) to wait until they'd gathered significantly more data — a quantity that, you know, offered some kind of respectable level of support for whatever conclusions they might be moved to draw — before deciding it was time to publish a book with "brain-savvy insights" in its subtitle, or time to tell prospective purchasers of the book that it provides "first-hand scientific knowledge" that they can use as a "practical guide" to help them "improve their work-flow and learning," and "identify people's struggles and stress areas," or time to claim that their lab results provided "strong neurological validity" for Jung's eight types.
================================
I suspect Dr. Nardi may not be interested, but any other reader who's made it this far, who's unacquainted with my much-linked-to (by me) posts about what I call the Real MBTI Model, and who's open to a hefty helping of reality-based input on the relationship between the dichotomies and the functions, the place of the functions (or lack thereof) in the MBTI's history, and the tremendous gap between the dichotomies and the functions in terms of scientific respectability — not to mention the unbearable bogosity of the Grant function stack — can find a lot of potentially eye-opening discussion in these three posts:
The Real MBTI Model
The bogosity of the "tandems"
The dichotomy-centric history of the MBTI
(The third post replaces the old INTJforum link at the end of the first post.)
Gluttons for punishment should note that the end of the second linked post talks about Berens and Montoya's new "Intentional Styles" grouping — which they were then calling "Cognitive Styles," and which Dr. Nardi prefers to call "Growth Styles," and which I like to call
the apotheosis of the double flip. It matches the socionics quadra groups, and it's one of the models that Nardi's recommended (in this thread) as a source of potentially useful "data points" for typing purposes.
Dr. Nardi has also recommended Berens' more well-known "Interaction Styles" model, and anyone interested in reading a long analysis/takedown of that particular typological
muddle (an apter word, in this case) will find it in
this PerC post.