I didn't write this talk because syntax: A syntactic analysis of because NOUN

I didn’t write this talk because syntax

The novel construction because X is illustrated in (1), where because appears with a bare noun complement and no linking of.

(1) I wore my skeleton leggings on Wednesday because Hallowe’en.

In previous work the complement of because has been analysed as a propositional ‘non-sentential’ element (in the sense of Progovac 2006) rather than part of a true subordinate clause, selected and fully integrated into the main clause. This raises an interesting question about the behaviour of negative sentences with this construction.

The presence of a because adjunct normally provides the possibility of ambiguity about the scope of the negation, as illustrated in (2) and (3) (and as noted many years ago by e.g. Lakoff (1970) and Linebarger (1987):

(2) I didn’t wear my skeleton leggings on Wednesday because of Hallowe’en.

= negation scope over the VP

= I didn’t wear my skeleton leggings on Wednesday, and the reason was Hallowe’en (perhaps I’m saving them to wear on Thursday, when it is Hallowe’en).

(3) I didn’t wear my skeleton leggings on Wednesday because of Hallowe’en.

= negation scope over just the because adverbial

= I wore my skeleton leggings on Wednesday, but not because of Hallowe’en (perhaps they were the only clean thing I had to wear that day).

When the because X construction is used, the ambiguity seems to disappear, leaving only the VP negation reading. This talk describes a study testing this intuition on 74 respondents to an online grammaticality judgement survey.

Participants were presented with sentences like those in (1) and (2), both with and without of, and asked whether the meaning corresponded to VP negation, adverbial negation, both, or neither (this allowed respondents to indicate that a sentence was ungrammatical, as this construction is not possible for all speakers).

As expected, sentences with a standard CP or PP complement were almost never judged ungrammatical, and just 7% of the because X type were. Of the remaining responses, VP negation was overwhelmingly favoured for both types of sentence, with just under 5% of responses selecting adverbial negation in both cases. However, the crucial question was whether the rate of ‘both’ options fell under the because X condition, as predicted by our hypothesis that the adverbial negation reading is not possible with this sentence structure. This was supported by the data, with ‘both’ responses falling from 29% to just 7% across the pairs, and the ‘VP negation’ option correspondingly increasing along with ‘ungrammatical’.

The experiment therefore demonstrates that, without intonational cues,

a) VP negation is the unmarked reading;

b) Adverbial negation is only possible as a second option if VP negation is also possible;

c) The adverbial negation reading is strongly dispreferred in because X sentences.

This result in fact answers one of the most fundamental questions about the construction: if one adopts a syntactic analysis of neg-raising (e.g. Collins & Postal 2014), it shows that it cannot be a surface phonological effect, with of simply unpronounced but still present in the syntax. If this were the case, this difference in behaviour is unexplained. In this talk, we argue that the because-adverbial is prevented from either having the option of a high or low adjunction site as generally assumed, and instead being restricted to low adjunction to VP, or from having the inherent focus assumed by Kawamura (2008) that permits the ambiguity of reference. This new, ‘internetese’ usage of English thus provides insight into the scope of negation, focus and adverbial attachment.