Apophatic Language in Mystical Experience — Epoche C1
When the Bṛhadāraṇyaka Upaniṣad, the oldest and longest of the Upaniṣads, says of the self that it can be indicated only by neti neti — "not thus, not thus" — it has plainly said something, and the difficulty this essay exists to address is what. The standard reading of such utterances, which I shall call the linguistic-limits reading, takes them to report a failure: ordinary descriptive language is not fit for a transcendent object, and the mystic is announcing that fact. An earlier version of this essay rejected that reading on the ground that apophatic statements — statements of what the divine is not, from Greek apophasis , denial — instead positively assert the transcendence and inaccessibility of the divine. That objection is right about what is missing from the linguistic-limits reading and wrong about what should replace it, and showing why requires first distinguishing three things that "negation" can mean here. Three kinds of negation, only one of which does the work Take the sentence "God is good" and consider what its denial might amount to. There are three logically distinct options, and apophatic writers use all three without always marking the difference. Predicate negation. "God is non-good" — the subject is credited with a positive property, the property of falling outside goodness. No theologian intends this, since it makes God defective rather than transcendent. Sentential negation. "It is not the case that God is good." This is a straightforward truth-claim, and if true it tells us something: that goodness, as we understand it, is not among God's properties. This reading keeps the negation informative. Category refusal. Neither "God is good" nor "God is not good" is true, because the predicate has no application here — as neither "the number seven is blue" nor "the number seven is not blue" is true. On this reading the negation is not a claim about God at all but a refusal of the question. The distinction matters because the earlier version's thesis requires the second reading and the tradition it appeals to requires the third. If a mystic who calls God unnameable is stating a truth about God's nature, he has described God — namely as possessing the property of unnameability — and the apophatic programme has failed in the act of being stated. That is not a debating point; it is the problem the major apophatic writers explicitly address, and their solutions differ from one another in ways worth setting out. Neti neti: what the text does, and what its commentators made of it The Sanskrit formula settles part of the question by its grammar. Neti is a contraction of na iti , in which iti is the quotative particle, the marker that closes a stretch of reported speech — roughly, closing inverted commas. Na iti therefore means "not thus", where "thus" refers back to what has just been said. The negation is anaphoric: it rejects the description just offered, rather than predicating anything of the self. The formula appears at several places in the Bṛhadāraṇyaka, including the passages in which the teacher Yājñavalkya, pressed for a description, offers this and states that there is nothing beyond it, and it always comes after a description, never on its own. Grammatically it is a retraction, which is the third of the three options above, not the second. The original version of this essay described neti neti as "a systematic process of negation designed to strip away all conceptual superimpositions". That is an accurate description of what the formula becomes in the Advaita Vedānta tradition, where it acquires a technical name — adhyāropa-apavāda , deliberate superimposition followed by rescission, in which a provisional description is offered precisely so that it may afterwards be withdrawn. It is not, however, a description of what the Upaniṣadic text does, which is considerably terser. The conflation of a scripture with its commentarial elaboration is worth flagging because the commentary supplies the very thing — a repeatable method — that the argument needs, and it is better to credit it than to read it back into the older text. Moses Maimonides, working in Judaeo-Arabic in the twelfth century, built an explicitly systematic version of the same procedure, and his treatment is the clearest case of an apophatic writer who unambiguously takes the second reading. In the Guide of the Perplexed he argues that no positive attribute can be predicated of God except attributes of action, and that every apparent positive predicate must be construed as a negation: to say God is powerful is to deny powerlessness, not to ascribe a quality. He is emphatic that this is informative rather than merely defensive, and he defends the claim with an analogy. A man who has never seen a ship is told successively that it is not a mineral, not an animal, not a plant, not a homogeneous body, not a single flat plane, and so on through a series of exclusions; at the end of the series, Maimonides says, he has come near to the notion of a ship, purely by elimination. Negations accumulate towards knowledge. The analogy also shows where the accumulation stops. Elimination narrows a field only when there is a field to narrow — the ship is located within the space of artefacts, and each negation excludes a region of it. Maimonides holds that God shares no genus with anything, and so there is no space within which the successive exclusions converge. He draws the consequence himself in holding that knowledge of God consists in knowledge of one's incapacity to comprehend Him, and that silence is the highest praise. The method is real, and it terminates in something other than a description. The self-refutation problem, and the move that answers it This brings the argument to the point where the original essay's central claim has to be revised. Its first proposed function for apophatic language, "epistemic assertion", holds that in calling God unthinkable the mystic affirms a fundamental ontological distinction a