Model A gives you eight aspects in eight positions. Two independent extensions push the count to sixteen: Ibrahim Tencer's Model A2 and the Model L framework of Kimani White and Aleesha Lowry. The easy version of this comparison says they agree on the number and nothing else. The easy version is wrong. They agree on far more than the number — and the real disagreements sit somewhere less obvious than the count.
A convergence deeper than the count
Both systems take the eight aspects and split each one in two. That alone would be a shallow coincidence; sixteen is what you get almost any time you halve eight. But look at how each splits, and the coincidence stops being shallow.
Tencer divides each aspect into a positive and a negative version — +Telos and -Telos, +Laws and -Laws — and the two halves are not the same thing in different lighting. They are distinct elements. He says it directly: the four domains are the underlying information, but each of the sixteen represents information to the extent that the sixteen differ from one another. Crucially, the same-aspect pair stands in a named relationship. In Model A2, +Laws and -Laws are Comparative, or Kindred, elements. The split produces two genuine elements bound by Kindred.
White and Lowry do the same thing and name it the same way. Each aspect resolves into two distinct monadic elements, and those two stand in the Kindred relation. Habitus is Kindred to Intellect; the two readings of Laws are Kindred partners, not a parent and its shadow. Model L's sixteen are distinct elements that pair within each aspect by Kindred.
So the convergence is not "both reach sixteen." It is "both split each aspect into two distinct elements, and both bind the resulting pair by the same intertype relation." Two researchers, working from different premises and different vocabularies, landed on the same architecture: not eight aspects with a modifier, but sixteen elements organised into eight Kindred pairs. That is a striking result, and it deserves to be stated plainly before we go looking for what divides them.
Where the elements actually come from
If the output architecture matches, the generators do not. The first real difference is a difference in derivation rather than ontology.
Tencer works top-down from meaning, then down again into structure. He starts by defining the metaphysical essence of each aspect — aspects are processes that seek states, carrying both information and energy — and he derives the sign from the Process/Result distinction laid over a double ring of progression and regression. The positive and negative versions correspond to a type's progressive and regressive orientations through its values. The split, for Tencer, falls out of what an aspect does as it moves toward or away from what a type wants. The signs line up with the Reinin positivism/negativism dichotomy, but the engine underneath is semantic: he can tell you why +Laws limits and -Laws unifies, because he has a story about what Laws fundamentally is and what happens when you run it forward or backward.
White and Lowry work from structure outward. The sixteen monadic elements are entailed by applying strength dichotomies to the eight aspects within Model L's block architecture, which extends Model A's four mapped blocks to eight. The Radial positions — the B and C capacity blocks — are required by the full Reinin dichotomy space even though classical Model A never renders them. The qualifier on each element identifies the family of its block partner. The elements are not interpreted into existence from a theory of essences; they are forced by the combinatorics of the block structure.
Both methods land on eight Kindred pairs. But they buy different things on the way. Tencer's semantic derivation explains why a pair feels the way it does — its content comes with a rationale attached. White and Lowry's structural derivation explains that the elements must exist, with less hostage given to any particular reading of what an element means. One approach is richer in interpretation and easier to challenge on whether the interpretation is right. The other is sparer and harder to fudge, but leaves more of the element's character to be filled in after the fact. These are different bets about where a typology's rigour should live, and they expose the two systems to criticism at different points.
What the pairing does in practice
A second real difference shows up at the typing table, in what each system asks the Kindred pair to do.
For Tencer, the positive and negative versions of an aspect are tools a type deploys at different rates, and he is unusually specific about it. Within a single type, the progressive and regressive forms of the same function get used to differing degrees, and he catalogues the preferences — which signed element a type leans on, which it reaches for as a defence, which it neglects. The pairing is dynamic: it describes how one psyche moves between two readings of the same aspect across its mental and vital rings. The same machinery is what he uses to distinguish types that share a leading function in Model A — the LIE who likes set limits where the LSE does not, because of which signed logic each prefers.
For White and Lowry, the pairing is structural first. The two monadic elements occupy defined positions in the block architecture, and the Kindred relation between them is a fact about that architecture before it is a fact about behaviour. The framework's discriminating power is meant to come from the positions — especially the Radial ones that Model A cannot see — rather than from a catalogue of usage rates.
This is a genuine and probeable difference, and it does not require settling whose metaphysics is correct. Tencer predicts intra-type dynamics: a single person should show measurable, patterned movement between the positive and negative readings of one aspect. Model L predicts structural placement: the discriminating information lives in which positions an element occupies, including positions with no classical correlate. One theory says the action is in how a type oscillates within a Kindred pair; the other says it is in where the pair sits.
Epistemological stakes
TetraTypes applies a Popperian standard, where corroboration — surviving a risky test — is the reward, and the conventionalist stratagem of protecting a theory with ad hoc exceptions is the central failure mode. Measured against it, the two systems are closer than their reputations suggest.
Tencer is more falsificationist than the "irrefutable mathematics" label implies. Read his working notes and he states that nearly all the element-level content — the usage preferences, the behavioural readings — is empirically determined and can be revised without touching the fundamental theory, and that the framework rests on a single axiom: the progressive first function as the overriding motivation, from which everything else should follow. That is not a man building an unfalsifiable fortress. That is a man marking off one load-bearing assumption and inviting revision of everything above it. The irrefutability ambition, where it appears, applies to the formal skeleton, not to the empirical flesh — and the empirical flesh is explicitly offered up for refutation.
So the honest epistemological picture is not irrefutability versus falsifiability. Both systems expose claims to test, and both have a soft spot. Tencer's soft spot is that his semantic derivations are interpretive: when a behavioural reading fails, the theory can in principle migrate to a different reading of the same aspect without anyone being able to say a rule was broken. White and Lowry's soft spot is the Radial positions: structurally entailed, but thin on distinct behavioural signatures, which means a claim that has not yet been made fully refutable. Different exposures, comparable seriousness. Neither is the unfalsifiable caricature, and neither is home free.
What would move the needle
Be fair to both, because both earned it by arriving at the same eight Kindred pairs from opposite directions, and that convergence is the most interesting fact in the whole comparison. The differences that remain are real but downstream: where the sixteen come from, what the pairing is asked to do, and where each system is most exposed to refutation. A full head-to-head is also still limited — Tencer's quadra-element treatment is incomplete in published form, with Beta and Alpha developed and Gamma and Delta not yet out, so any verdict resting on the full set is provisional on material that does not yet exist.
But the systems already make one cleanly opposed, testable prediction, and it exploits the difference in what the Kindred pairing does. Tencer predicts intra-type oscillation: a single confidently typed subject should show patterned, measurable movement between the positive and negative readings of one aspect — the same person using +Laws to impose structure in some moments and -Laws to simplify and unify in others, at rates his catalogue predicts from their type. Model L makes no such prediction from the pairing itself; its discriminating signal is positional, and the Kindred relation between the two elements is a structural fact, not a behavioural oscillation to be clocked.
So here is the test. Take confidently typed subjects and, blind to the hypothesis, have experienced observers track whether a single subject shows systematic switching between the two signed readings of their leading aspect at the rates Model A2 predicts. If that oscillation shows up, patterned and type-predictable, it corroborates Tencer's dynamic account of the Kindred pair and asks Model L to explain a behavioural regularity its structural derivation does not anticipate. If the oscillation does not appear — if what distinguishes the subjects is better explained by positional structure than by within-pair movement — that corroborates White and Lowry and leaves Tencer's usage catalogue looking like description without predictive teeth. Either way the result does not depend on first settling whose derivation is more elegant. It turns only on whether the Kindred pair is something a type moves through or something a type is built from.
That question has a wrong answer, which is what makes it worth asking. Two systems that agree on sixteen elements and agree on Kindred pairs can still disagree about what those pairs are for — and that disagreement is one an experienced socionicist can actually put to the test.