Eric B's Counterpoint: A Defense of Functions

There has been some dissension in the matter of whether the basic elements of type are the cognitive functions or just the four dichotomies of the MBTI type code. Jung spoke of eight "types" defined by a dominant function (sensation, intuition, thinking or feeling) and "introverted" or "extraverted" "attitude". Yet Isabel Myers (in creating the popular Myers-Briggs Type Indicator), while drawing upon Jung's theory, turned the four functions into two dichotomy pairs (S/N and T/F; the middle letters of the type code), dominant attitude into the first dichotomy (I/E), and then recast Jung's final "attitude" pointer, "rational" vs "irrational" into "Judging" vs "Perceiving" (J/P). One difference, was that Jung's dichotomy pointed to the dominant function, while Myers' pointed to the preferred extraverted one, actually deemed a bit more important in our personal interactions, which figured a lot in MBTI's purpose. (One version of Socionics attempts to restore Jung's use, in using a lowercase "j/p" to point to the dominant function).

But our own critic of functions has illustrated the very point of functions in the way he has approached the theory! http://www.typologycentral.com/forums/myers-briggs-and-jungian-cognitive-functions/71276-finally-official-tandem-names.html

The INTP and INTJ both are dominant introverts, who prefer iNtuition and Thinking. The difference in the final J/P notation is much more than a mere "dichotomy" of simply one of them being more "open" while the other is "closed" (or any other "facets" of these).

As an INTP, my whole approach to the theory has been in terms of the mechanics (T) of the systems, with symmetries, factors, etc. determining an individually (i) analyzed sense of what's "true" or "correct". This is informed by turning to the environment (e) to take in the hypothetical data of what could be (N); like the (intangible) patterns I see in the various systems. In this perspective, it's easy to find a "place" for almost everything. Many different concepts can exist side by side, and each one offers a different "lens" or "angle" at looking at the same things. (Which is the purpose of type theorist Linda Berens' models; previously called "Multiple Models", and now renamed "CORE"; which uses the type dichotomies, functions, archetypal stacking positions, temperament, and Interaction Styles models, as all pointing to the same things).

INTJ has the same "individual" (i), "mechanical" (T) and "hypothetical" (N) approach. However, this type starts with the hypothetical perspective of what could or couldn't be, and this is what's individually derived (from unconscious impressions), and what's "correct" is what's determined by the environment (object), which includes a focus on empirical data (such as which approach is more practical, defense of the framework of one instrument, appeals to the various theorists, tests, and "case studies" regarding people's type scores). Hence, what we see, is that the person starts out just "knowing" that functions are "incorrect" and thus "could not" exist (not even simultaneously with dichotomies; it's either one or the other), and that therefore, only dichotomies are "correct". All the external data is gathered to prove the preconception.

We see further here the demand for "real world" evidence, defined as a "body of MBTI data pools". This is Te, and not the same Thinking that I use. It's both "logic" or a judgment of what's "valid" (correct) or impersonally "true". But demanding "data pools" places this judgment on an external "object" or the "environment" in general. (Assuming those sources are always true, and can't miss important data. In order for stuff like that, such as statistics, to be absolutely "true"; they would have to survey every single person in the world, but how many of us have ever been in their studies? So the most we can do is discuss our own experiences, which don't get added to this "data pool").

When someone (like Grant [not sure what type he was] or Beebe) looks beyond that, and proposes a judgment of true/correct/valid outside that data pool, that's obviously "subjective" or based on the individual. So someone who thinks truth should be environmentally determined (or "objective") only will see the other way as "fanciful speculation"; and that's it; hands down, as if the universe said it itself. But it's really the different attitude of a function, comparable to looking east and seeing the external world before you in that direction, as opposed to remembering what you saw in that direction before. On the flipside, we both use iNtuition, but that's where I or someone like Beebe turns to the environment and lets the objects or experience itself, like what I'm appealing to now, determine their own possibilities, and the other perspective rather turns inward to a subjective sense that this just doesn't exist; no matter what; I just KNOW it isn't real no matter what anyone [besides official "researchers"] says or experiences, so again, it's left up to the judgment process only to turn to "the real world", to prove it isn't "true".

This is what another INTJ had once described to me, regarding the dominant Ni perspective, of looking at "what a theory doesn't take into account", and then, from that, (as Nardi's definition would put it), "forecasting" (i.e. "knowing" no one will ever be able to come up with any evidence of functions, from the missing information everyone else has supposedly ignored).
I don't usually think like that. My perspective is Ne. I look at what a theory could take into account. (i.e. looking at what the "object" or theory in the environment could do, where Ni's sense of "could" comes from the "subject", which for iNtuition is the individual's unconscious, itself). I say "hey, this looks interesting, it looks like it could fit, now let's see what happens. I think the theorists are continuously refining the theories, so they probably will come up with more definitions and descriptions, but let's wait for more information [from these external sources] before making a final judgment". Reflecting the J/P dichotomy (which correlations link to FFM's "Openness"), this is how one is more "closed", and the other, more "open".

This shows, what that one different letter (J/P) is telling us, is that my perception function is environmentally oriented, and the other person's judgment is environmentally oriented. Which is what we can see. Thus, for both of us, these functions are auxiliary, since our dominants are individually focused; mechanics for me, and hypothesis for him. Which is what we also see. Hence, NP's (who prefer the function of extraverted iNtuition) are known to toss around a lot of random ideas, while NJ's (who prefer introverted iNtuition) usually take a more skeptical view. If you see the functions as telling us "yes/no", then Ne, which looks to objects says "yes" to just about anything (since most objects imply just about any possibility, and thus judgment or tandem-mate Si is needed to do the "nay-saying"), while Ni is more likely to say "no", based on what comes up from inside the subject. Extraverted Judgment is then used to verify Ni's yes or no with its own yes or no, as we see with the focus on empirical "truth". (We can see now, why NP and NJ also figure in the "informing/directing" pole of the Interaction Styles model). This is what the functional perspectives tells us that dichotomies only doesn't.

It seems people (and "empirical research") are looking for behavioral similarities, in any persons of opposite types, for functions to be true. Dichotomies may be more evident in behavior (and hence why they work so well with temperament; i.e. both Keirsey and Interaction Styles, and also FFM). Functions are about how we divide reality, and while this does influence behavior, it is not always directly obvious.

For one thing the unpreferred functions become more conscious later in life, so no, an ESFJ and INTP child are not going to have any evident similarity. I clearly can see a difference in relating to an ESFJ like my wife, compared to an ESFP. Our types basically grow toward each other as the unpreferred functions develop, but I don't see this with SFP's at all; instead, they and NTJ's would grow together. So it's not a static "this type is always like that type" if there's something in common; it's about what they're growing toward.

According to this "dichotomy only" approach, the ESFP should have slightly more in common with me than an ESFJ, and ISFP should have even more in common, but it seems to be reverse, where those types and I "miss each other" (share no common interest in communication) with the ISFP the least similar of all. All the common P is saying is that either the S or N are oriented externally. This produces a slight commonality in behavior, of "openness" compared to a J type who orients them internally and thus needs "closure". But the way we process S, N, T and F are totally different.

It's also worth pointing out, that an appeal is made to the popular "Keys 2 Cognition" test by Dario Nardi, which measures the eight function-attitudes directly (according to "strengths") instead of the four dichotomies. It's pointed out that many people score high on both Ni and Ne.
But all any N type getting high in both attitudes means that they are overall strong in iNtuition (and likely haven't developed their tertiary or inferior S), so it seems to "spill over" into both attitudes, and remember, the test is not perfect. It's operating on a set of definitions that do not take into consideration the fact that either attitude can do some of the same things. It just uses general sets of behaviors and assumes they indicate a particular function-attitude. Like it assumes "considering others and responding to them" is an interpersonal judgment based on an external standard. (So if you select high on that, it will score a point for "Fe"). But not necessarily! With an internal standard (Fi), you can infer a sense of the other person's need, and then "consider and respond to them".

I myself have been trying to get back to a focus on singular ("natural" or "whole") functions (rather than strictly eight function-attitudes or "processes") because of this. So if you look at it that way, then you have two of your "dichotomies", and the dominant orientation (held independently of the functions) would be a third (or actually, first) dichotomy.

All the e/i at the end of the function letter is telling you is that the functional perspective will be connected to stimulus by the individual or environment, and so the person tends to turn to an inner or outer orientation or standard when engaging the function. (Which is what the fourth dichotomy is indicating). This does create significant differences in perspective.

Both perceptions are iNtuitive products: observation of intangible or "hypothetical" data. One simply looks at the object and sees multiple possibilities, and the other starts with an internal "awareness" used to filter the data. That's all the function-attitudes are.

And as we see, it shapes our perspectives and approach to the issue. My ESFJ wife, though not as geared toward Ne as I am, nevertheless handles data in a similar fashion. (And I clearly experience a similarity of perspective that I don't get with INTJ's, SFP's, or even, in part, STJ's and NFP's. And this is my marriage, so I should know!) An SFP would likely think all of this is a waste of time, but if they had to deal with it, they would probably handle it more like the INTJ. So that's something "TP's and FJ's have in common with each other and not with TJ's", and thus why opposite functions (Si-Ne; Se-Ni, Fe-Ti and Fi-Te) are said to work in tandem.

And this is precisely what the new "Intentional Styles" and tandem terms coined by Berens and associate Chris Montoya were made to address. (i.e. "Inquiring", "Realizing", "Aligning", and "Ordering", respectively. My approach is more of "Inquiring" ⦅gathering multiple emergent intangible connections {Ne}, and measuring them against a storehouse of tangible data {Si}⦆. The INTJ's approach is "Realizing" ⦅from an internal intangible connection {Ni}, along with emergent tangible facts {Se}⦆. The SFJ's and SFP's will take the same respective approaches, but simply place more of an "accent" on the S).

The Empirical data argument

A big part of the argument is that there were over 50 years of studies done with MBTI, and none of them picked up any similarities between the polar opposite types. The whole argument revolves around one person, James Reynierse, who wrote a paper published in CAPT called "The Case Against Type Dynamics", where he characterizes functions as a "category mistake" and lays out these arguments against them.
There's a claim that he "caused a stir" in the type community, though it's hard to find much about him, and searches will generally point to discussions on him on these forums, or this CAPT paper and one other, "An MBTI Model of Entrepreneurism and Bureaucracy: The Psychological Types of Business Entrepreneurs Compared to Business Managers and Executives", (which argues the J/P dichotomy is primary, and the others are subordinate). You can see his credentials here: https://www.zoominfo.com/p/James-Reynierse/77565068

The argument is made that the functions have barely even been studied, because they've never been taken seriously by any significant number of academic psychologists. Of course, the reason any theory wouldn't be taken seriously is because it lacks studies! This creates a tautological fallacy.

So to start, let's look over this claim of Reynierse about a "grand total of eight studies" in the MBTI Manual involving "type dynamics" that he's all discounted for one reason or another. Here's the whole quote that is taken that from:

There are four additional points to be made here. First, six studies were cited that did not demonstrate expected type dynamics predictions. Whether or not the studies cited in the MBTI Manual provide a good test of type dynamics and are in fact weak tests of type dynamics as argued there misses the point. Regardless, they still failed to support type dynamics.

Second, Thorne and Gough's (1991) observer ratings and the reanalysis of that data for the dominant and auxiliary forms of the N, T and F functions does not, as contended, support type dynamics. The MBTI Manual identified the overlap in the 10 most and 10 least adjective descriptors for the dominant and auxiliary forms of N, T, and F for both males and females and then stated, “These percentages range from 0.0 to 13.8, suggesting that independent observers, who did not know the types of people they were describing, clearly did not see much similarity between types having the dominant versus auxiliary forms of Thinking, Feeling, or Intuition” (p. 204). The implication is that there are striking differences between the dominant and auxiliary as predicted by type dynamics. There is a fundamental baseline problem for these comparisons, however, as Thorne and Goughs observers did not find much similarity anywhere, even within the same types—a result that might be expected, as observers made ratings using a rich and diverse pool of 300 adjectives.

Third, the original research based on national sample data, although presented in a cursory fashion and for a very limited set of dependent variables, provided some, albeit limited, support for type dynamics. Dominant extraverted thinking (DET) was greater than auxiliary extraverted thinking (AET) for both health and friendships dependent variables and dominant extraverted feeling (DEF) was greater than auxiliary extraverted feeling (AEF) for a social dependent variable (MBTI Manual Table 9.21, p. 203). At the same time, there were also significant reversals that are incompatible with a type dynamics interpretation and contradict it. Auxiliary introverted sensing (AIS) was greater than dominant introverted sensing (DIS) for an accomplishment-dependent variable and auxiliary introverted intuition (AIN) was greater than dominant introverted intuition (DIN) for a home-and-family-dependent variable (MBTI Manual Table 9.22. p. 204). Support then was about as frequent as disconfirmation and not particularly compelling.

Fourth, the discussion of type dynamics introduced very restrictive conditions for evaluating the efflcacy of type dynamics. e.g., requiring first a significant effect for the E—1 x J—P interaction term, an effect that occurred rarely for the national sample data and in other investigations (Reynierse St Harker. 2001). Such restrictive methodological requirements do not correspond with the generality and ubiquity of type dynamics anclications—applications in which type dynamics interpretations are unrestricted. In my judgment, type theory is better served by embracing both theory and method that are inclusive and where the application applies to all—without restriction—dependent variables.

And that is the full extent of direct support for type dynamics as reported in the MBTI Manual (1998)—six studies that failed, one with a questionable interpretation, and one where contradictory evidence was offered as support.

"Six that failed":
For the first three; from the Manual itself: "However, in all three studies only half or fewer of the people had higher scores for the dominant function than the auxiliary function".

The next three simply found "no significant association" or "any support for making a distinction between dominant and auxiliary functions". I don't know what exactly he means by "questionable interpretation" and "contradictory evidence" for the remaining two, but his fourth point is just his own "judgment" of these "conditions" being too "restrictive". Like I say, this ultimately seems to boil down to mostly just one man's interpretation of the data.

And I notice all of these studies are about the differences between a given function as dominant vs auxiliary. The Manual itself (p.204) addresses this and the problems with the results of those studies:

Although these findings appear to be contrary to the predictions of type theory, it is important to examine these hypotheses tested in these studies and the assumptions that underlie the approaches taken by these authors to testing their hypothesis.

First, all of the analyses described above either tacitly or explicitly assign unintended meaning to high or low preference scores. Even the original hypothesis that drives the analyses—that the person's score on the dominant function should be higher than the score on the auxiliary function—is not consistent with current interpretations of the theory (Myers & Kirby, 1994) and is therefore not a good test of type theory. The magnitude of the preference scores used in these studies does not indicate development, skill, aptitude, maturity or excellence in the use of a function. (The same is true of the preference clarity indexes used in form M [i.e. the % "scores" you get for each preference in a respondent's results]). Yet the meaning ascribed to these scores by the hypothesis and the analyses that derive from it ignores the categorical nature of the preference scales and assumes a quantitative linkage between the properties measured by MBTI and the scores used to identify type preferences.

Second, the item responses that form the basis for classifying people into different preferences are designed to appeal to people of one preference or the other on a given scale. For example, the "S" response to an S_N item was written to appeal to all Sensing types. Measurement at this gross level of the overall preferences does not permit direct inferences about the dominant or auxiliary status of the functions based solely on preference scores.

Third, by its very nature the dominant form of a function is presumed to be different from its auxiliary form. Myers' method of identifying dominant versus auxiliary functions assumed an interaction between E-I and J-P dichotomies The impact of this interaction shows up in its effect on the forms and attitudes of the Judging and Perceiving functions.

The characteristics of dominant introverted Feeling, for example, are different and separate from those facets of Feeling that are shared among all Feeling types. Few authors have spelled out such differences explicitly (an exception is Myers & Kirby, 1994). This is one area of type theory that is open for future development. Nevertheless, at present the theory presumes that such differences are real and can be observed. Studies that simply compare scores of dominant and auxiliary functions ignore these differences.

So I would also say that the differences between the dominant and auxiliary won't be as great as between the dominant and tertiary or inferior, which is the issue that has been raised here with regarding "tandems", and "INFP's and ESTJ's having something in common". It's easy to confuse the dominant and auxiliary, and this will result in I/E uncertainty, which a lot of people have. (Keirsey even said that as a standalone dichotomy, it's the least important). Sometimes, the dominant is so "second nature", it actually becomes "unconscious" in a way. These tests they're using aren't going to pick up all of that.

Reynierse just dismisses this, because it "still failed to support type dynamics", but it touches up upon this related point:

Next, a claim is made:
"Going all the way back to 1985, the MBTI Manual described or referred to somewhere in the neighborhood of 1,500 MBTI studies, and not one of the many study-based correlations reported in the manual were framed in terms of the functions."

So the question is, how would these tests have picked up that A TJ and FP, demand an external (e) criteria for determining what's “correct” (T), while a TP and FJ tend to determine it internally? Were these old studies even geared for that sort of data when they simply looked at T and J (separately), or F and P, T and P, etc.? Did any of them look at how the dichotomies work together?
No; if they're correlating with dichotomies only, then of course, the results will reflect dichotomies only (where they are completely opposite).

While the functions might affect behavior, they themselves are not really “traits” these tests often focus on. (I would also acknowledge Jung's point that the inferior will often come out in an opposite fashion, so that the dom. will favor the function's products, while the inferior will try to avoid it.
Still, when forced to deal with T products, Fi doms will tend to turn to an external focus. And let's also not forget the counterclaim by the larger psychological field that even the MBTI pool of data is not really valid for some reason. So they apparently don't prove as much as one may think).

Then, there's also correlations with "countless aspects of personality as separately measured by lots of other established personality instruments".

Like what else? FFM (NEO-PI)? Those are factors closely correlated with MBTI dichotomies, and thus will also turn out similar results. CPI? (A two-factor model also owned by CPP which is similar to classic temperament). Even less likely to say anything about function-attitudes or axes. FIRO-B? (which I have personally read several correlations of, as my long standing interest was how they might fit with the dichotomies). The same. It's a completely different framework that was not looking for something like that (Though still, I did notice that N and P both did correlate with "high Wanted Inclusion", which I predicted would match Keirsey/Berens' "informing communications", which is a common feature of "extraverted iNtuitive" types! But it would have no way of picking up anything called "Ne" directly, when you presume it must be proven separately from an N+P preference when all you're correlating it with are dichotomies, N, P and the others. And again, the people doing those studies obviously saw that high correlation with those two dichotomies, but if they were not "looking for" (as I put it) the function of "Ne" to correlate with anything, then they're not going to report it).

Splitting of Reality: Function Dynamics expressed in terms of Dichotomies

In reality, dichotomies and function preferences alike are but the ways the ego divides reality. If I prefer N and T, then both S and F are suppressed, and can be seen as "collecting" in a place lower in consciousness. This "dichotomy only" argument makes it look like I have no access to S and F (or E and J for that matter either), and that ESFJ's have no access to I, N, T and P. So yes, when an ESFJ needs to access intuitive products, it will tend to be more "open" (emergent), while an ESFP will prefer it more settled.

So the reason why NTP and SFJ will fall into one group, that would exclude NTJ and SFP, is because of the splitting of those "pairs" of functions. For an NTP, N, T and P are preferred together, and S, F and J are suppressed together. So S, F and J are still "together" in the psyche, even though less conscious. For NTJ, N and T are preferred, but not P. J is preferred instead. S and F are suppressed, but not J. P is, instead. So the NTJ will not have a whole NTP or SFJ "image" (so to speak) in the ego-syntonic part of his psyche.
So you will have some surface similarity (in preferring iNtuitive and logical data), but something will be missing. The orientations (or where the energy is directed, which is what J and P are telling you) will be different. He will have to go into the "shadow" (meaning the unconscious) in order to access one or two of the preferences, to put together a complete SFJ or NTP perspective (and according to another version of the theory, it is various complexes that will put these together. Functions are otherwise really "undifferentiated", outside of the dominant).
I guess it's like a kind of "dissonance" in mixing together preferred and unpreferred elements, so that all unpreferred poles together will be more palatable than a mixture.

It all works together (dichotomies and dynamics), and is not "either/or".

Meanwhile, Berens takes the multiple models as part of a more "holistic" framework.

A couple of things she says in an article:

"Since human nature is so complex, no one typology can adequately describe behaviors, systems, relationships, and meaning making, hence there is no one 'official' typology used in Integral Theory. To this point, no set of criteria seem to have been set forth for what makes one typology more useful than another and what makes a typology more consistent with an integral approach to working with living systems."

"The Cognitive Dynamics model is based on Jungian theory from which the Myers-Briggs Type Indicator instrument was derived (Berens & Nardi, 2004). In actuality, the dichotomies identified by the MBTI instrument were somewhat artificial constructs designed to create an instrument to detect the types predicted by Carl Jung (Myers et al., 1998; Jung, 1921). Presenting them as either/or led people to think they could not access the other side of the dichotomy."

"A Meta-Model for Types: Patterns, Polarities, and Autopoiesis" – Linda Berens Journal of INTEGRAL THEORY and PRACTICE: A Postdisciplinary Discourse for Global Action volume-8-numbers-34