Socionics Supervision Rings: Four Loops

Every other small group on this site is a set. A supervision ring is a sequence, and the sequence is a direction: ILE supervises LSI, LSI supervises SEE, SEE supervises EII, and EII supervises ILE, closing the loop. Say it the other way round and you have described four relationships that do not exist.

That is why this page is built the way it is. The direction of every step comes out of the relation table and is checked against socionics-core on every build, including the order the rings are printed in, because four right types in the wrong cyclic order is the same error as a wrong type and reads just as smoothly.

Each arrow runs from supervisor to supervisee, and the last member of every row hands back to the first.

How the four rings are found

Nothing needs to be assumed. Take any type, look up the type it supervises, go there, and repeat. The chain closes on the type you started from after four steps, and it does that whichever of the sixteen you start from. Sixteen types, chains of four, four rings, no remainder.

Three things fall out of that walk which are worth more than the membership lists.

The first is that a ring never visits a quadra twice. Every supervision ring takes exactly one Alpha, one Beta, one Gamma and one Delta type, and exactly one member of each club. Two of the four rings walk the quadras in the order Alpha, Beta, Gamma, Delta, which is the sequence an idea travels in the quadra lifecycle; the other two walk it backwards. So half the supervision in the socion runs with the direction ideas are supposed to develop in and half runs against it, which is at least a suggestive thing to notice and is not, on its own, evidence of anything.

The second is that the two types facing each other across a ring, the first and the third or the second and the fourth, are always a super-ego pair. So a ring is not only a chain of supervision: it has two of the socion's flatly uncomfortable pairings built into its diagonals, and they are there in all four rings.

The third is that the attitude alternates at every single step. Extravert supervises introvert supervises extravert, all the way round, in all four rings. The benefit rings do the opposite and never change attitude at all, and that difference turns out to be the visible edge of a much sharper one.

Where a supervision step actually lands

This is the part I would keep if I had to throw the rest of the page away, and it is arithmetic rather than description.

Take any of the sixteen supervision pairs and ask where each partner's Ego block falls in the other's stack. The answer is the same for all sixteen:

Positions one, two, three and four. Every point of contact in a supervision relation sits in the conscious half of Model A, the Ego and Super-ego blocks, and none of it reaches the Super-id or the Id. The benefit ring is the exact complement: all four of its points of contact sit in positions five to eight and none in the conscious half. Between them the two asymmetric relations divide Model A down the middle, and I have not seen that stated anywhere, which is odd, because it explains the thing everybody notices about both.

What it explains here is why supervision is loud. The supervisor's best function is aimed, without any intent behind it, at the one position the supervisee has no equipment for. The supervisee answers with their whole identity, their leading function, and it lands on something the supervisor handles flexibly and does not especially value. Neither person is doing anything. The mismatch is in the wiring, it runs both ways, and it is entirely above the waterline, which is why supervision is the relation people can usually describe without knowing any theory.

The four rings

Causal-Determinist

Reasons forward in chains: this caused that, which caused the next thing. An account is finished when no link in the sequence is missing. The ring closes from EII back to ILE.

Holographic-Panoramic

Takes one subject from several fixed angles and lays the views over each other. An account is finished when the separate projections stop disagreeing. The ring closes from SLE back to LII.

Vortical-Synergetic

Thinks in loops that feed themselves, where a small push compounds until something tips. An account is finished when it names what did the amplifying. The ring closes from IEI back to ESE.

Dialectical-Algorithmic

Works from a contradiction and grinds it down by procedure, step after step. An account is finished when the opposition has been turned into a rule. The ring closes from LSE back to SEI.

Gulenko's cognitive styles, and what the match does and does not prove

The four names in the rows above are not mine. They are Viktor Gulenko's, from his work on forms of thinking, and the four groups he describes have exactly the membership of the four supervision rings. That is the interesting claim, and it is the first thing this page checked: the published lists were compared against the rings derived from the relation table, and all four match, type for type. The descriptions in the rows are my own compressions of what each label means rather than anything translated, and they are deliberately short, because a form of thinking described in a sentence is a label and I would rather not dress it up as more.

Gulenko grounds the four in three Reinin dichotomies: Static against Dynamic, Positivist against Negativist, and the trait he calls Evolution against Involution, which this site calls Process against Result. Each ring is indeed constant on all three, which the build checks against the trait lines on the dichotomies page so the two cannot drift apart.

There is an arithmetic point to make about that, though, and it is not a criticism so much as a correction of emphasis. Three two-way splits would carve the socion into eight groups, not four. Static against Dynamic crossed with Positivist against Negativist already produces exactly these four rings on its own; Process and Result is then fixed by the other two and adds nothing. So "defined by three dichotomies" is loose. Two define it and the third comes along for the ride, which is worth knowing if you ever try to use the third as an independent check on a typing.

The pattern people notice about duals

A claim that circulates on the forums: every dual pair consists of one Causal-Determinist and one Dialectical-Algorithmic, or one Holographic-Panoramic and one Vortical-Synergetic. All eight dual pairs, no exceptions.

That one is true, and it is worth saying why rather than leaving it as a curiosity. Duals share a quadra and share rationality, and differ on attitude, on intuition against sensing, and on logic against ethics. Flipping the attitude while holding rationality flips Static against Dynamic. Positivist against Negativist flips too. So a type and its dual always sit in diagonally opposite cells of the grid, which is precisely the pairing the claim describes.

Notice what that means. The pattern is a consequence of how the dichotomies multiply, not independent support for the cognitive styles. It would hold just as neatly if the styles described nothing at all.

The transmission claim, and what would test it

The observation attached to these rings since long before the cognitive styles were attached to them is that information travels round the ring in one direction: each type can take what the one behind it produces and hand it on, so the socion has four closed circuits of transmission running through it.

Half of that is simply true. The ring is a one way cycle, it closes, and it does not run backwards. Nobody supervises their own supervisor.

The other half does not survive the contact map on this page. Every step of a supervision ring puts the sender's leading function onto the receiver's vulnerable fourth position, which is the worst landing site in Model A for anything you want retained rather than resisted. If a supervision ring transmits, it transmits under pressure, and pressure is a good way to make somebody comply and a poor way to make them understand. My own reading is that what actually circulates round a supervision ring is not information but correction, and that the two get confused because they look identical from outside.

What would settle it is not complicated, and its absence is the point. Give a non-trivial claim to the first member of a ring, have them pass it to the second in their own words, and so on round the loop, then measure how much of it survives. Run matched chains of four types that do not form a ring as a control. Type everybody blind, and have somebody who does not know the hypothesis score the output. Nobody has done this. Until somebody does, the transmission claim is a description of a structure that plainly exists, plus an assertion about what happens inside it that has never been measured.

Practical use

Recognition, not assembly. You will not build a supervision ring on purpose and you would not want to.

Where it earns its keep is in reading a chain you are already in. If you find yourself consistently corrected by one particular colleague and consistently correcting another, and the two are not the same person, you are standing in the middle of a ring segment rather than having a run of bad luck with personalities. The useful move is almost never to resolve it. It is to stop routing anything that needs understanding, as opposed to compliance, along that line. The group mapper will show you the pairs, and the Supervision relation page covers what to do inside one.

Honest limits

The rings themselves are as solid as anything on this site: they fall directly out of the relation table, they are checked on every build, and there is nothing to believe or disbelieve about them. The same goes for the contact map, the quadra walk and the alternating attitude.

Everything in the second layer is weaker, and the weakness is worth naming precisely. The cognitive styles are one theorist's framework, circulating in English mostly through forum translations, with no published protocol or replication behind the descriptions. The membership match is real, but a match between two four-way partitions of sixteen types is not as impressive as it first looks: they agree because both are built from the same dichotomies, so the agreement is arithmetic rather than evidential. What would be evidence is a blind test in which people's thinking is classified from their output and lands in the ring their type predicts, and I have not found one.

Typing error does its usual damage and slightly worse than usual here, because a ring is a chain. One wrong member does not just add noise, it breaks the loop, and the person who is supposedly supervising them is now the odd one out in every description you read.

The other half of the asymmetric socion is the benefit rings, which are the same idea run on the relation that lives in the unconscious blocks. For the groupings built on sets rather than sequences, start at small groups.