Reprinting vaguely philosophical comments elsewhere to have something on this blog. I also have a post at my other blog that's kinda sorta related to the epistemology of testimony, but it's also rather scurrilous, so there it stays.
Matthew Yglesias wrote, under the title "The Transitivity of Timetables":
So if McCain likes Maliki's timetable, and Maliki likes [Obama]'s timetable, then logically McCain has to like Obama's timetable.
A commenter going by "pedant" responded:
The "likes" relation is not transitive, nor are timetables (whatever that could mean). What we have here is the following inference:McCain likes Maliki's timetable.
Maliki's timetable = Obama's timetable.
So, McCain likes Obama's timetable.Whether or not the inference is good depends on whether McCain knows Maliki's timetable = Obama's timetable (he does), and whether McCain is rational enough to draw the relevant conclusion (questionable).
Taking that pseudonym as a personal challenge, I replied:
It doesn't even necessarily depend -- it depends on whether in "McCain likes Obama's timetable" the term 'Obama's timetable' appears de dicto or de re. In "John flew to X" X is always de re; if John flew to Hesperus then he flew to Phosphorus, even if he doesn't know that Hesperus and Phosphorus are the same thing (Venus). In "John intends to fly to X" then X can be de dicto; if John wants to go to Hesperus but wants to avoid Phosphorus (not realizing they're the same), then we can say that he intends to fly to Hesperus but that he doesn't intend to fly to Phosphorus.If "John likes Obama's timetable" is a de re ascription then the inference is good whether or not McCain knows that it's good. He may not think he likes Obama's timetable, but it's Obama's timetable he likes. I actually think that after 'likes' it has to be de re. I'm not sure we can make sense of "John likes aubergines, John doesn't like eggplants, but aubergines and eggplants are the same thing." Aubergines/eggplants are a vegetable, and either John likes that vegetable or he doesn't.
For more on de re and de dicto see here; as McKay and Nelson point out, nobody actually agrees on exactly what these things are.
I actually think that the problem of how to report speech is very important in politics, and the de re/de dicto distinction (and indirect discourse in general) is an important part of that. The interesting part does not extend to whether one should fix Yglesias's mind-blowing typos.
Posted by Matt Weiner at July 26, 2008 03:57 PMIsn't your own example of John wants to go to Hesperus but wants to avoid Phosphorus a case of John liking Hesperus but not liking Phosphorus?
Posted by: Richard at August 5, 2008 09:33 AMThis discussion would be even better if it was illustrated with the "Gaahhhh! Kreplach!" story from Gravity's Rainbow.
Posted by: Ben at August 5, 2008 07:33 PMI explain the de re/de dicto distinction to myself with this example: Jimmy says, "I hate the person who drove over my tricycle in the driveway." But it's his mother who drove over his tricycle in the driveway, and he doesn't mean to say "I hate my mother." It reminds me of Oedipus, who spends a lot of time condemning the source of the corruption in Thebes...and it turns out that HE is the source of the corruption.
Ooh, considering that I'm Matt's mother, maybe these examples are too fraught.