We based this squib in part on results from the literature. No outside research is required for Syntax II squibs. However, any outside resources must be cited.
First, there seems to be a range of other null subject constructions in English:
| (1) | "Woke up, fell out of bed, dragged a comb across my head,...'' | Diary Drop | |
| (2) | Ames is more devious than (*it) was reported in the press. | Comparatives (see Merchant and Kennedy 2000) | |
| (3) | Ames was a spy, as (*it) was reported in the press. | As-parentheticals |
Second, even if we grant that English has null subjects, they are clearly of a different sort, and subject to different restrictions, than the null subjects of languages that linguists often classify as null subject languages. Italian, Spanish, Bulgarian, Greelandic Inuit, and many (perhaps most) other languages have null subjects in basic clauses. So, for example, the contrast between the following examples is clear and in need of explanation:
| (4) | Parla | italiano | (Italian) |
| speak.3RD | Italian | ||
| (5) | *Speaks Italian. |
Third, we say in the squib that PP's cannot be subjects in English. But this is not quite correct; sometimes one finds such things in [Spec,IP]:
| (6) | In one's closets is/*are a smart place to store one's many pairs of shoes. |
| (7) | "By canoe is the way the first people saw this
place, doubtless." —Bill Roorbach. Temple Stream. Harper's Magazine, December 2001 (p. 56). |
Under what conditions does one find such oddities? What properties do they have?
Fourth, we show in the squib that the PP in locative PP constructions does not determine subject-verb agreement in English. But Bresnan observes (in the paper we cite in the squib) that agreement with the fronted PP is required in Chichewa. So it is an attested pattern, though not an English one. Further cross-linguistic work is essential to figuring out the source of this typological difference.