The Presence Cube

Recently I discovered a new model that is more meaningful than previous models we've discussed.

Any cube model requires three basic traits. The classical Model A cube uses mental/vital, valued/subdued, and evaluatory/situational. The issue is, other than valued/subdued, these dichotomies are not visible in practice. The other variations on the cube all depend on unobservable dichotomies like this.

The most visible traits ⁠in Model A are strength, values, and boldness ⁠— the presence traits. If we use the presence traits as the basis we get the presence cube:

The presence cube for IEI (image credit: Andrew Joynton)

It turns out that this cube gives some very interesting results. We can directly see dimensionality as a projection of the cube onto a certain axis. And there is a directly complementary projection which I call Priority, which defines the most preferred functions in practice: functions 1 and 6, then 2 and 5, then 3 and 8, then 4 and 7. The leading and mobilizing function (the 4P functions) tend to be very visible in someone's preferences, while the suggestive function and creative function are less visible as values due to being cautious (somewhat related to triads). The demonstrative function and role function are similarly used somewhat more due to being bold, although maybe not clearly less than the suggestive function. The 4P and 1P ends do seem to be clear in practice though, much like 4D and 1D functions — so we can think of them as trichotomies, with two extremes and one ambiguous middle region. The middle functions can in fact be lined up by rotating the cube appropriately. So we have two trichotomies and three dichotomies.

The presence cube viewed according to the Priority trichotomy for "Strategic Democrat" types: LIIs and SEIs prioritize Si and Ti and LIEs and SEEs prioritize Se and Te.

The trichotomy for values is linked to the inert/contact (aka stubborn/flexible) dichotomy: the functions at the extremes are stubborn, as their priority does not change easily, and the functions in the middle are flexible: if you rotate the cube slightly the 2nd and 5th functions may fall below the 8th and 3rd ones in priority. Likewise for the evaluatory/situational dichotomy which has the most extreme strengths and weaknesses (4D + 1D functions) on one end, and the medium ones on the other.

The vertical axis in the diagram is sort of a combination of dimensionality and priority, the sum of all three traits which we may call Level of presence, so we get 1L, 2L, 3L, 4L, with the leading function being at the top as the sole 4L function. This roughly describes how much each function is present in someone's cognition and behavior. The 3L functions are the producing functions which are all directly connected with the leading function in some way. The connection the 1st function has with the 4th function is "severed" or veiled.

The LII cube, labeled

The main issue with the presence model is that there is a separate cube for Democrats and Aristocrats (in the Reinin dichotomy sense):

You can reflect the cube in any way to get all the other types with the same Democracy/Aristocracy trait, i.e. either the same quadra or opposite quadra. Rotations are actually unnecessary here; we have 2x2x2 = 8 transformations coming from each reflection / inversion of a trait.

If you project directly onto one of the faces, then you pair elements either as the standard blockings for the given quadras (ignoring boldness), or with one of their "skew blockings" along the benefit ring, ignoring either strength or value. These represent the most common element pairs that we see in practice, since typically rational elements work with irrational ones and vice versa.

The presence cube ties together some fundamental observations from socionics practice and is another significant step towards a meaningful, mathematical model of socionics.

Model G and How To Fix It

Jack Aaron and Ben Vaserlan recently had a heated debate over the relative merits of Model G and Model A: why Model G was created and if the reasons really hold up.

Aside from a few minor quibbles (such as the supposed difficulty of telling apart the vulnerable and role functions), I completely agree with Jack's criticisms of Model G. In short, it adds essentially nothing to Model A, is steeped in vague jargon (like "long range" and "short range", the social vs. personal spheres, etc.), and is at odds with how the types actually work (in particular how it conceives of the suggestive function as being somehow "high energy" or more prominent than the mobilizing function — or whatever they're called now).

Model G does, however, have a few selected insights. One is the greater emphasis on the benefit rings, or rather, the bold/cautious dichotomy.

Another is the idea of "energy." For some reason Ben flounders in this video when asked to define energy, while he had previously connected it to Jung's concept of libido (not Freud's), which Jung defined as a kind of "life force". This is a sound idea in itself: Model A addresses information processing (information metabolism) but it does not address the obvious limitations of resources that apply to each function's processing, in particular the strength and boldness traits.

Unfortunately, beyond this very basic outline, the details of Model G seem disconnected with the reality of the types — including which functions are supposedly maximum energy, etc.

That's a very brief take on the semantics — the details are really not that interesting and are addressed in the video. Another interesting idea (which precedes Model G) is the signed elements, but I'm not going to get into that here either. The real point of this post is the structural deficit which I pointed out and Jack later mentioned in the debate. It's very obvious if you look at Andrew's diagrams (and translate the names and numbers accordingly):

As you can see here, some of the standard socionic dichotomies are presented asymmetrically in Model G. Gulenko apparently does not assign any meaning to the left and right sides of the model, he still uses the standard dichotomies such as strong/weak, etc., albeit with different names, as displayed on the right here. Strong/weak becomes "master/slave" and valued/subdued becomes "values/tools". (correction: It is said that Gulenko does consider the left-hand functions to be "better" in the "sense [...] of energy allocation to the function and the degree of freedom of behavior afforded by this." Again, more jargon which does not seem to apply to the suggestive function, or if not actually wrong, at least is not as clear as the existing Model A dichotomies.)

How can we fix this? The obvious thing to do would be to simply switch the suggestive and ignoring functions (or "manipulative" and "control", numbers 6 and 8). Then the left side is strong and the right side is weak. And the valued functions are "outside" and subdued "inside", which is at least as good as it is in Model A. And we still have the benefit rings proceeding horizontally: NeTeSeFe... and TiNiFiSi....

In my opinion this clearly shows that the Model G blocks are defined wrong. The issue is that Gulenko wanted to have a benefit loop of types, but he represented them using the standard Model A ego blocks (as the columns of Model G). So for ILE we have NeTi, then TeSi, then SeFi and FeNi for the ILE and its Process-Extrovert ring. Instead we should express the types entirely in terms of the benefit ring (in analogy with how it is in Model A and the supervision ring) with each represented using a consecutive pair of elements in the ring, ILE being NeTe, LSE being TeSe, etc. Then the types can be thought of as "edges" between the IM elements, and they interact at their shared points. This makes much more sense if Model G is meant to show energy flow, does it not?

Ben made the interesting point (possibly the only one he made in the entire video) that Model G includes not only benefit rings, but also supervision rings if you extend it vertically:

(Written in to the left of the #4 heading.)

Note that this property still holds if we switch the suggestive and ignoring functions. There are literally only two ways to make a grid like this, and Model G does not use the right choice.

Aside from the semantic issues with Model G, this is a very obvious structural flaw. Fixing it might be the first step to salvaging the model.