posting over in L words

the linguistics puzzles discussion appears to have been relocated in the L words blog. i may come back here, e.g. to comment further on earlier posts. for the most part, though, this blog is effectively retired.

coming soon to the L words blog: my rant about why i don't believe in empirical induction; my theory of generics; and the intersection of these two things. maybe i'll throw in something on the existence of god just for kicks.

modality in "future"

(1) is true, if utterered by me.

(1) christine is my former future roommate

it's true because back in june, i had some concrete plans and a commitment to sharing a place with christine.

but since then, the plan fell apart, and i have no expectation that we will ever live together in the future. i would be lying if i uttered (2).

(2) christine is my future roommate

in general, (3) is false in the present and in all points past, if uttered by me.

(3) christine is (currently) my roommate

suppose former and future are operators that shift times, without any additional modality. then (1) can be represented as in (1'), or in terms of time operators past and future, as in (1'').

(1') former[future[christine is my roommate]]
(1'') past[future[christine is my roommate]]

if the time of the utterance is t1, then (1) would mean that there is are points in time t2 and t3, such that t2 precedes t1 and t3 follows t2, and christine is my (current) roommate is true (when uttered by me) at t3.

but given that (1) is true when uttered by me, and given that (3) has never been and will never be true, this analysis appears to be wrong.

why? i blame the word future, which appears to include an element of modality, specifically modality of prediction or expectation, not unlike the future modal will.

my problem with this analysis is, it doesn't seem to me that (4) is true when uttered, if it happens to be a plausibly surmised but factually wrong statement about the future.

(4) jane is my future wife

[notice, though, that if i will propose to jane tonight, it is true that she is my future future wife).

in any case, the facts above are not limited to the word future, but also apply to other (apparent) adjectives and nouns:

my soon-to-be roommate
the president-to-be

are all futures in natural language really modal?

gaps with "have" in relative clauses

well, i sure abandoned this blog for a while, didn't i?

here's a puzzle:

(1) i have a brother
(2) *the brother that i have
(3) the only/tallest brother that i have

also, compare the clearly ungrammatical (2) with the not so bad (2').

(2') (?)the book that i have

is the contrast between (2) and (2') a result of the well-known distinction between relational nouns like brother and nonrelational nouns like book? it seems not. rather, it appears to be the distinction in the kind of relation expressed by the possessive construction: (2') is grammatical because the possessive can be interpreted with an ownership interpretation, while (2), which cannot give rise to this interpretation, is ungrammatical as a result. presumably this contrast shows that where an empty category is the complement of have, the ownership relation is the only acceptable interpretation. the pattern in (4) supports this presumption.

(4) Q: what do you have in the house?
A: *a brother/*a husband/*a problem/a table/a bed

it is not surprising that the ownership interpretation is grammatically significant. partee and borschev and luca storto have both shown that this interpretation plays a role in the grammar of adnominal possessives.

what about the contrast between (2) and (3)? in general, the relevant pattern appears to be that a modifier imposing uniqueness allows a post-have gap. however, not all uniqueness-imposing modifiers have this effect. favorite, for instance, doesn't.

(5) *the favorite brother that i have

intro

as part of my ongoing project of uploading absolutely everything, i hereby dedicate this blog. hope i don't forget my password.