Here’s a question that came up at BAPHLD the other day, to which I might now have an answer. It’s impossible to interpret (1) as (2) – it has to be read as (3)

(1) Three or four jobs interest me.
(2) Three interests me or four jobs interest me
(3) Some jobs, three or four to be more precise, interest me

This point is made by Thomas Hofweber in Number Determiners, Numbers, and Arithmetic, and he uses it to make a point about the possible meanings of ‘three’. I thought there should be a simple explanation for this. (More precisely, various theoretical claims I believe commit me to there being a simple explanation, and I believe the consequences of my theoretical beliefs, even when this is hard work.)

I was trying to argue that (2) was a genuine disambiguation, and it just wasn’t preferred for pragmatic reasons. Normally pragmatic explanations are a little ad hoc, but this one looked like a no-brainer. Among the possible pragmatic explanations are

(a) Who’s interested in the number three anyway?
(b) If you were interested in the number three, wouldn’t you put that by sayings “The number three interests me”, not “Three interests me”?
(c) Under what circumstances would you be unsure whether you were interested in the number three or four jobs?
(d) How do the number three and four jobs fit into a common enough group that they could be conjoined?

Well, maybe I could go on, but you see the drill. The problem (allegedly) is that a pragmatic explanation is too weak. All of these are reasons why (3) should be the default interpretation, not the only possible interpretation.

Fortunately there’s a simpler explanation, one that I managed to miss the first time around. Note that to get from (1) to (2) I had to change ‘interest’ to ‘interests‘. If we interpret (1) as something like (2), there will be a plurality disagreement. So (3) is the only acceptable interpretation, I think.

More generally, “three or four” is an interesting phrase. It would be decidedly odd to say “Three or five jobs interest me”, although one could imagine circumstances under which that is true. It’s not that “three or four” is an idiom, but there is something odd about the construction. Maybe one for the Language Loggers, though it’s probably wrong of me to suggest that since I haven’t responded to their last philosophy challenge.