Socionics Small Groups: All 35 Systems

Socionics names about a dozen ways of splitting the sixteen types into four groups of four. Quadras, clubs, temperaments, the rings. Each arrives with its own theorist and its own vocabulary, and you could be forgiven for thinking the list is open-ended and someone invents a new one every few years.

It isn't open-ended. There are exactly 35, and this page lists every one of them. Of those, 11 have a name and a page on this site. The other 24 have never been described by anybody, and I am not going to be the first to make something up about them.

Why the number is exactly 35

Start with the building blocks. Socionics has fifteen dichotomies: the four Jungian ones (Extraversion and Introversion, Intuition and Sensing, Logic and Ethics, Rationality and Irrationality) plus the eleven Reinin dichotomies. Each one cuts the sixteen types into two halves of eight.

Now take any two of them and ask a third question about a type: does it land on the same side of both, or on different sides? That question is itself one of the fifteen dichotomies. It always is, for any pair. So every pair of dichotomies drags a third one along with it, and the three together form a closed trio.

Here is the one everybody already knows. Take Extraversion and Rationality. Types that are extraverted and irrational, or introverted and rational, fall on one side of the combined question. The rest fall on the other. That combined split is Reinin's Static and Dynamic, and fixing a pole on all three leaves four groups:

Those are the four temperaments. The third dichotomy adds no new cut, which is the whole point: two dichotomies give four groups, and the third tells you which pairs of groups belong together.

So how many trios are there? Fifteen dichotomies make 105 pairs. Each trio contains three pairs, and each of those three pairs produces the same trio, so every trio gets counted three times. 105 divided by 3 is 35. For anyone who wants the formal version: the sixteen types form a four-bit space, every dichotomy is a yes-or-no question about it, and these trios are its 35 two-dimensional subspaces. Every one of the 35 splits the sixteen into four groups of four, and every dichotomy belongs to exactly seven of them.

What the types inside each group have in common

A group of four types contains six pairs, and every pair has an intertype relation. So each group has a relation signature: the set of relations you find if you look inside it.

Take Alpha. ILE and SEI are duals, LII and ESE are duals. ILE and ESE are in activation, LII and SEI too. ILE and LII are mirrors, and so are ESE and SEI. Two dual pairs, two activation pairs, two mirror pairs. Now look inside Beta, Gamma or Delta and you get exactly the same six. That is what I mean by a uniform signature: all four groups in the system are internally built the same way, so whatever it feels like to be in one quadra, the relational texture is the same in the others.

I did not want to assume that held for every system, so the build works it out from the relation table for all 35 rather than taking it on trust. It holds for 31 of them. In the other 4, every one of which includes Rationality and Irrationality, the signature splits cleanly along that line: the two rational groups share one set of relations and the two irrational groups share another, with kindred and business trading places, and semi-duality and mirage doing the same. I have marked those four in the table rather than smoothing them over.

The asymmetric relations, supervision and benefaction, turn up inside 22 of the 35. Where they do, the table gives the direction of every pair, because "A supervises B" and "B supervises A" are different claims and only one of them is true.

All 35 systems

Named systems come first, in the order the small groups index lists them. The rest follow, starting with the ones built from the most Jungian dichotomies. Groups are numbered 1 to 4 within each row so the relations column can refer to them.

No system matches both filters.

The eleven with names

Eleven systems match a page on this site exactly, type for type, and the build checks that match on every deploy rather than trusting the page. One of them carries two names: the temperaments and the bouquet groups are the same four sets, the second a reading of the first through energy exchange.

The names come from different places. Reinin's square groups are the relaxation and blocking groups. Gulenko supplied the romance styles and communication styles, and his cognitive styles and project groups sit on exactly the supervision rings and benefit rings. Each page credits its own sources.

Three of the site's groupings are not on the list at all, and that is correct rather than an oversight. The health groups sort types by the kind of function they lead with. The stress behaviours, drawn from Kretschmer's special dispositions, and the pedagogic needs, my own framework within the SLIDE System, sit on the same four sets of types: the sensing types split by rationality and the intuitive types by logic and ethics. Neither rule comes from a trio of dichotomies, so neither lands on one of the 35. A grouping does not have to be one of these to be worth looking at. It just is not part of this particular closed set.

Why most of them have no name

24 of the 35 have no interpretation anywhere I can find, and I have labelled them exactly that.

There are two honest readings, and I do not know which is right for any given system. Some of them are probably noise. A split of the sixteen types that falls out of the arithmetic is not obliged to mean anything, and a trio built from three weakly supported Reinin traits has no particular reason to describe a real pattern in people. Others may be patterns nobody has gone looking for. Reinin's eleven dichotomies were themselves worked out on paper before anyone went looking for them in people, and several of them have since picked up reasonable support.

What I will not do is write the missing descriptions. It would be easy to produce a plausible paragraph for every row, and every one of them would read well, because descriptions of personality groups always do. That is exactly the problem. The test is not whether a description sounds right but whether people in one group behave measurably differently from people in another, and that needs data. Socion matches people by type, and match data like that is where these 24 splits could eventually be tested.

Until then, the complete list is the useful part: it tells you that the named systems are a small, specific selection from a fixed set, and it shows you everything they were selected from.

Where to go next

• The small groups index has every named system with a full page of its own.
• The Reinin dichotomies article explains the eleven traits these systems are built from.
• The Group Mapper shows every relation inside a real group of people, which is the practical form of the signatures above.