HERE'S A FEW MORE FACTS
Randall Sparks · Linguistic Inquiry · 1984
Nathan correctly points out that Dixon's phonological explanation for this phenomenon less than adequate and makes some further observations and suggestions.' Specifically, Nathan claims that number neutralization with be (henceforth NN) is much more widespread than Dixon suggests and proposes some restrictions on its occurrence. Consideration of some additional examples will show, however, that NN does not occur as widely as Nathan claims in certain contexts and that it does occur in other contexts in which Nathan's restrictions would prohibit it. First, consider the claim that NN much more widespread than Dixon originally suggested. Whereas Dixon demonstrated its occurrence only with where-questions, Nathan claims that it may occur in any wh-question: