polarity in English

Clausal negation in English attaches to the finite auxiliary and not to the lexical verb. The negator has a full form and a reduced one, and the reduced one is a clitic fused to the auxiliary tightly enough that some of the resulting words are irregular in shape. Where the clause has no auxiliary, the language inserts do for the negator to attach to, exactly as it does for a question, and the lexical verb reverts to its plain form. That insertion is the reason this record reads explicit rather than periphrastic: the exponent is a bound element on a word, and the language will manufacture a host for it rather than leave it unattached.

Beside the clausal negator sits a set of words that incorporate negation lexically, covering a negated person, thing, time, place and determiner. These are single words, not phrases, which is the sense in which this dimension is marked at the level of the individual word rather than the clause.

The standard variety allows exactly one negation per clause, so a clause with a negative determiner cannot also carry the clausal negator on the same reading, and a speaker who does combine two is understood as cancelling one against the other rather than as reinforcing them. This is a real point of variety internal divergence and is recorded here rather than split into a separate node: many nonstandard varieties of the language use negative concord productively, where a second and third negative element reinforce a single negation instead of cancelling it, which is what a great many other languages do as their standard. The record describes the standard variety and says so.

The affirmative is unmarked, with one exception worth naming: emphatic affirmation uses the same inserted auxiliary under stress, so the language has a way of asserting a positive polarity out loud when it is at issue. The obligatoriness is recorded as obligatory because a clause that means a negation must carry the marking, and no clause can leave polarity undetermined.