The Binding Argument
Jason Stanley claims that
the following kind of argument is generally sound.
(1) It is possible to interpret In every room in Johns house, every bottle
is in the corner as being true iff in every room in Johns house, every bottle
in that room is in the corner.
(2) Hence the logical form of Every bottle is
in the corner includes a quantifier domain restrictor, as it might be the
property being in this room, which is bound by the outermost quantifier
in the longer sentence.
This is a move in a dispute between those
who say that Every bottle is in the corner expresses the proposition that
every bottle whatsoever is in the corner, or perhaps expresses no proposition
at all, and only conveys the proposition that every bottle in this room is
in the corner. Jason thinks that the restriction to bottles in this room is not
pronounced, but it is articulated – it is a genuine feature of the syntax of
the sentence, and of the proposition that sentence expresses.
This proposal has come in for some criticism
recently, especially in Ernest Lepore and Herman Cappelens paper in the most
recent Analysis.
Let me add my own little objection, mostly due to a conversation with Europa
Malynicz. (The examples are very close to ones she suggested, as I guess is the argument, but the usual disclaimer that all the faults are mine applies.) (3) is a well-formed sentence, and the quantifier here does not feel
like a null quantifier. It is just the kind of thing that could be said by a
rather cautious traveller.
(3) Everywhere
I go, I only drink bottled water.
Jasons argument is that the quantifier here
is not a null quantifier, so it must be binding something, so the embedded
sentence must have syntactic element that is bindable. Hence I guess (4) really
expresses the proposition (5).
(4) I
only drink bottled water.
(5) I
only drink bottled water at place x.
But it is quite implausible that this is
what (4) means. I dont think this is implausible for the reason Lepore &
Cappelen think it is implausible, that it puts too many variables into (4).
Maybe they are there, maybe they arent. I think it is implausible because it
suggests, falsely, that one could use (4) when at place x to express the
proposition that one only drinks bottled water at that place, even though you drink
tap water anywhere else. So I could use (4) here and now to express the
proposition that in Providence I only drink bottled water.
This argument needs one qualification. In
some contexts it is possible to use (4) to express the proposition that I only
drink bottled water in Providence. If I have just been asked a question about
Providence and especially its water quality, and I respond with (4), it will be
natural to view my utterance as being elliptical for the claim that I only
drink bottled water in Providence. But what Jason needs (I think) and what he
cant have (I also think) is a situation where I could use (4) to express that
proposition without it being simply elliptical.