Speaking of the unspeakable

Why do traditions that declare every positive statement about God false still go on speaking about God at length?

A tradition announces that its subject exceeds every name we have, that no predicate we possess applies, that our best assertions are further from the truth than our silences. And then it produces libraries. Maimonides devotes chapter upon chapter of the Guide of the Perplexed to what cannot be said; the writer we call Pseudo-Dionysius composes one treatise on the divine names and another on their removal.

Taken as a plain instruction, the doctrine issues in a blank page. It never does. So either these writers are inconsistent, or "cannot be said" is doing something other than forbidding speech.

Start with a single negation and ask what it delivers. "Not a body" rules out a great deal. "Not ignorant", "not composite", "not in time" rule out more. Maimonides argues that each denial is a real gain: you know more about an object when you learn it is not a mineral, more again when you learn it is not flat. He means the comparison seriously — negations accumulate.

But push on it. Accumulate toward what? Elimination converges on a description only when the field being eliminated from is bounded. Here it is not, by the doctrine's own claim. So an endless series of exclusions never closes on a point, and negation cannot be a slow road to a positive account. Whatever the practice achieves, it is not that.

Then ask who the sentence is addressed to. A denial does no work on its object — nothing is altered by being denied. It works on the hearer, and specifically on something the hearer already has. You cannot remove an image nobody formed. So the negation needs a prior affirmation as its material: it is parasitic on the very speech it cancels. Which is exactly why Dionysius writes both books, and in that order — the names first, the unsaying after. The affirmative discourse is not a lapse before the real doctrine arrives. It is the doctrine's raw stock.

That reframes the length. If the aim were to convey a proposition, one sentence would carry it and everything after is padding. If the aim is to dislodge a habit — the mind's reflex of picturing — then a single correction is useless, because the reflex re-forms the moment the sentence ends. The volume is not rhetorical excess; it is the mechanism.

Now the difficulty every writer in this line sees coming. "No positive statement about it is true" looks like a positive statement about it. If that sentence succeeds, something has been said after all; if it fails, why record it?

Dionysius makes a third move: the denials are themselves denied. Not darkness and not light, not one and not many — the negations are not privations reporting an absence but markers of excess, which is why his vocabulary reaches for "beyond" rather than "not". Maimonides answers more soberly: the negations are claims about the reach of our intellect and language, not descriptions of their object, so nothing has been predicated at all and the self-refutation never arises.

Is that reply secure? Here is the objection with teeth in it. To know that our terms fail of a thing, you need to know enough of the thing to know they fail — so a blanket claim of unsuitability looks like knowledge of the target smuggled in disguised as modesty about ourselves. And if no term applies, "God is good" and "God is cruel" are false alike, and the practice the doctrine was protecting loses the object that made it intelligible.

That is where Aquinas parts company, and his alternative is the strongest rival on the field. In the Summa Theologiae he argues that terms like "good" apply neither in the same sense as they do to creatures nor in a wholly unrelated one, but by analogy: what the word signifies is truly affirmed, while our manner of signifying it falls short. The apophatic writer's reply is that analogy quietly assumes a common measure between the creature and what has no measure — and the argument has not closed.

Three refinements, and they matter more than the headline.

First, "apophatic" names a family, not a doctrine. Maimonides is a rationalist making a semantic claim about predication; Dionysius describes an ascent ending in unknowing; Gregory of Nyssa holds the pursuit endless because its object is infinite. The neti neti — "not this, not this" — of the Upanishads and the Buddhist analysis of emptiness get grouped with these, but the latter is not about a supreme being at all; the resemblance is structural, and treating it as doctrinal agreement is a common overreach.

Second, the doctrine is rarely as total in practice as in statement. The same authors who deny that any attribute applies go on prescribing liturgy and distinguishing pious language from crude — distinctions that need some purchase on the subject to be drawn at all. That tension is what generates the third move.

Third, the reading offered here — that the negations are addressed to the speaker's imagination rather than to their object — is one interpretation among several, and whether the strict version can be stated without collapsing is genuinely open. What turns on it is substantial: if it can, the discourse is a discipline and its length is coherent; if it cannot, something like analogy was doing the work all along and the denials are emphasis.

• Whether Aquinas's analogy genuinely avoids the problem or relocates it.
• How a tradition that forbids images of its subject still permits elaborate verbal description of it.

Every term the collection defines is gathered in the glossary.