A Talent, Not a Rule

Practicing reflective judgment in Raumschach — and why no one, including this essay, can simply teach it to you

International Raumschach Federation  ·  2026

I. Two Arguments, and the Question They Leave Open

Two essays already on this site make the case that Raumschach must be played from principle rather than calculation or memorization, and they make it from different directions. The first argues from Kant’s mathematical sublime: a 125-cell lattice with a 26-directional Queen and triagonal Unicorns exceeds what the imagination can hold in a single intuition, and the proper response to that failure is not despair but the discovery of principles that don’t depend on the faculty that just failed. The second arrives at the same practical conclusion without invoking Kant at all: with roughly 61 pseudo-legal moves available to White on move one, against chess’s twenty, the game tree outgrows human calculation so quickly that heuristics are not a fallback but the only available strategy at any human time horizon — and Raumcapa, which selects every move from named positional principles and never searches an opponent’s reply, exists partly to demonstrate that this is enough to produce a coherent opponent.

Both essays win the argument that you ought to play this way. Neither one tells you how to start. A reader convinced by both, who has never in their life reasoned about a position from bare geometry rather than recalling a pattern someone else already named, is left exactly where they began — knowing the destination and nothing about the first step. This essay tries to close that gap, with an honest warning, stated now rather than buried later, about why it can only be closed partway.

II. Why This Cannot Be a Checklist

Kant made a claim about judgment in general well before either the sublime or the third Critique’s account of reflective judgment, in the Critique of Pure Reason. The understanding, he says, is the part of the mind that holds rules, and it can be furnished with them by ordinary instruction. Judgment is different in kind. Judgment is the capacity to recognize whether a given case actually falls under a given rule — and no further rule can be supplied for that recognition without itself requiring another act of judgment to apply it, a regress with no floor. Kant’s conclusion is blunt: judgment is a special talent that cannot be taught but only practiced, and what ordinary language calls stupidity — a standing deficiency in judgment rather than in knowledge — is not a gap that any syllabus closes.¹

That is already true of judgment generally, including the everyday, determinant kind that simply applies a rule you already have. Reflective judgment inherits this limit and adds a second one on top of it. Determinant judgment at least starts with a rule in hand and only has to decide whether it fits; reflective judgment, as the third Critique defines it, faces a particular with no rule given for it at all and must find one. If ordinary judgment can’t be installed by instruction, judgment that doesn’t even have a candidate rule to check against certainly can’t — there is, by construction, nothing yet to teach.

This bears on the present essay’s ambition directly enough that it is worth stating the consequence before going any further. If anything below reads like five numbered steps for discovering Raumschach principles, that is a failure of the essay, not a feature of it. A five-step method for finding new principles is itself a rule, and applying it correctly to a given position would require exactly the judgment it was supposed to replace. So: no checklist. What follows instead is a way of recognizing the kind of moment that calls for this faculty at all; two reconstructions of principles already documented elsewhere on this site, done from nothing, the way a player without a textbook would have to do them; and a short list of habits — not steps — that exercise a capacity Kant says can only be exercised, never installed.

III. Two Kinds of Position, and a Boundary That Moves

The practical question worth carrying into a game is not is this position hard but do I have a rule for this, or only the bare geometry. Most tactics a chess player already knows transfer as determinant judgment on day one: a position is of a recognized type, and the learned response follows. Other shapes have no two-dimensional precedent to recognize at all. The Piece Monographs document one directly: a Knight planted on the board’s center can, in a single leap, command a piece two levels below it and another two levels above it, on the same file — a fork shape that no flat board could ever produce, because flat boards don’t have a second piece two levels above anything. The first player to notice that this was possible had nothing to subsume it under. That was reflective judgment, run cold, with no help available.

It does not stay that way. Once the pattern has a name and a checking procedure — scan all five levels before every Knight move, not just the current one — recognizing it is determinant judgment for every subsequent player who has read the page. This is exactly how Kant expects reflection to behave: a reflective judgment that successfully finds its universal does not remain provisional forever, it becomes available as a rule for the next case that resembles it, and using it then is no longer reflection but subsumption. The frontier of things only discoverable by reflection recedes as a body of theory grows. It does not close. A game whose opening theory, per this site’s own account, remains largely uncharted guarantees there will be positions on the far side of that frontier for a long time yet, and being able to tell which side of the line you are currently standing on is most of what this essay is trying to teach.²

IV. Two Things Found From Nothing

What the Center Has to Be

Suppose you are given only the movement rules and nothing else: a 5×5×5 lattice, a Queen riding six orthogonal directions, twelve face-diagonals, and eight space-diagonals at once, and two Unicorns per side confined to those eight space-diagonals alone. No opening theory, no named squares, no Axiom I. Where would you expect the board’s center of gravity to sit, and on what basis?

Start with what "center" could even mean on a cube, which has no single square in the middle the way a flat board does. Level C sits exactly between the two armies’ home levels — that much is counting. The more interesting fact concerns the piece that exists only because there is a third dimension at all. A Unicorn’s eight directions are the eight space-diagonals of the cube; follow any one of the cube’s four long diagonals end to end and all four meet at exactly one cell. That cell is Cc3, and it is therefore the unique square lying on every triagonal simultaneously — the single most geometrically connected point on the board, for the piece that has no existence outside three dimensions, independent of anything resembling chess judgment yet. The Rook’s new axis, the column, treats Level C as a literal midpoint between both home territories; the Bishop’s new vertical diagonals do the same. A pattern, not yet a rule, starts to assert itself: every piece that gained something by the move from two dimensions to three gained it specifically by running new lines of force through that one cell.

That is the whole argument, and nobody had to hand it to you. You noticed that an unusual number of independent geometric facts pointed at the same place, and you trusted, without yet being able to prove it, that the convergence was not an accident worth ignoring. That trust is what Kant calls purposiveness: not a fact about the board, but a working assumption you adopt about it before you have evidence a unifying principle exists, because refusing to look guarantees you never find one even if it is there.³ The published version of this conclusion is Axiom I — Level C is the Fulcrum, and the chain running from it — development order, the hierarchy of opening moves, most of the named openings — is set out at length in Volume II. What matters here is not that conclusion but the path to it. Every step above was available to a player who had been told the rules and nothing else.

The Square No Unicorn Can Reach

A second case, smaller and stranger. Suppose your two Unicorns are on their starting squares and you find yourself wondering, idly, whether either can ever reach Cc3 — the square the argument above just established as the board’s single most valuable point. There is no rule to consult. Nobody has told you the answer is no.

Notice that a Unicorn’s move changes level, file, and rank by exactly one step each, in some combination of plus and minus. Sit with that for a moment and a consequence falls out: each move flips whether every one of those three coordinates is odd or even, all three at once. Starting from any square, a Unicorn’s combination of odd-or-even across the three axes alternates between exactly one pattern and its precise opposite, forever; it can never settle anywhere else. Two starting squares give two such pairs, covering four parity classes between a side’s two Unicorns — and Cc3, whose level, file, and rank are all odd, belongs to none of the four. No Unicorn, of either color, starting from the array this game has used since 1907, can ever stand on the board’s own center. Not eventually. Never.

This is reflective judgment doing something the first case did not: finding a law nothing about ordinary chess intuition would lead you to expect, by following a parity argument wherever it goes regardless of whether it feels like it should matter. It is also, once stated, exactly as determinate and teachable as any rule in the game, which is the point made in Section III about reflective discoveries not staying reflective for long. The strategic upshot — Cc3 is effectively a Queen’s square, and each Unicorn’s real ambitions lie on its own nearby outposts instead — is recorded on this site as settled fact. So is the parity argument that gets you there. But running it yourself, from the bare movement rule and nothing else, is the only way to find out whether you can do the thing this essay is actually about, and it costs about four lines of arithmetic to check.

V. What Is Actually Worth Practicing

Kant’s own remedy for undeveloped judgment is not instruction but exercise on cases — sharpening through repeated use, not installation through a manual.¹ The following are habits in that sense: things to do across many games, not steps to execute within one.

VI. Conclusion: The Difficulty Was Never the Obstacle

The sublime essay’s argument was that the imagination’s failure in front of 125 cells is not a defect to be engineered around but the very occasion on which reason discovers a power adequate to the case. The same shape recurs here, one level up. The fact that no essay — this one included — can simply hand a reader the rule for finding rules is not a flaw in the pedagogy. It is the same dialectic applied to teaching itself: if reflective judgment could be installed by direct instruction, it would not be the faculty Kant is describing, and an essay that succeeded in installing it that way would have proven, by its own success, that it had been writing about something else.

Capablanca’s own description of his sense for a position — that when he was right, he simply felt it — was not mysticism. It was practiced judgment operating too quickly to narrate, the destination this kind of practice actually points toward, and exactly as unreachable by a checklist as Kant said any talent would be. There is no shortcut to it on offer here, only the suggestion that the next unfamiliar position you meet is the material the practice runs on. Don’t look it up first. See what the geometry tells you, and then go check.