Subject–Auxiliary Inversion

AbstractIn subject–auxiliary inversion in English, the typical declarative word order subject–auxiliary‐verb is instead realized as auxiliary‐subject‐verb. This inversion occurs in a number of environments, namely, matrix questions, conditionals, blessings and curses, comparatives, exclamatives, negative imperatives, and in environments where certain elements have been fronted (so/as/nor, negative phrases, phrases withonly, and certain others). Most theoretical analyses of subject–auxiliary inversion treat all of these instances uniformly, but some recent work indicates that there are actually two distinct types of inversion, one which can affect multiple auxiliaries and one which cannot. The most prominent theoretical account of subject–auxiliary inversion, head movement of Infl (or T) to C, accounts best for the type that can only invert one auxiliary, but is incompatible with the type that inverts multiple auxiliaries. This type may involve a lower‐than‐usual position for the subject, instead. Subject–auxiliary inversion is subject to a strict adjacency requirement, which requires that the fronted auxiliary be adjacent to the subject it has inverted with. Many theoretical proposals struggle to account for this requirement, as well as for the possibility of subject–auxiliary inversion in embedded clauses. Subject–auxiliary inversion is also one of the contexts fordosupport, and any theoretical account needs to explain why this is so and howdosupport comes about. Finally, subject–auxiliary inversion has figured prominently in debates regarding learnability (the “poverty of the stimulus” argument), and it is also important for the question of whether grammatical roles like “subject” are primitive notions of grammar, or not. It appears that subject–auxiliary inversion refers only to structure, and not to notions like “subject,” suggesting that grammatical roles are not primitives.