In the Kent Bach/Jeff King/Jason Stanley love-fest in Pasadena, there was a bit of difference of opinion whether the names in the following sentence are used literally:
(1) Leningrad became St. Petersburg.
What strikes me as odd here is the use of "become." I would have thought that "become" meant "started to be," so it's natural to use it with adjectives, common nouns, or descriptors. Hence the following are perfectly straightforward:
(2) Leningrad became the capital of Russia.(3) Leningrad became a world center of trade.
(4) Leningrad became isolated.
They're interpreted as, for some time t:
(2a) At some time t' before t, Leningrad was not the capital of Russia; after time t it was.(3a) At some time t' before t, Leningrad was not a world center of trade; after time t it was.
(4a) At some time t' before t, Leningrad was not isolated; after time t it was.
But the corresponding
(1a) At some time t' before t, Leningrad was not St. Petersburg, after t it was
is weird, at least on the view that true identities are metaphysically necessary; if Leningrad was St. Petersburg after time t, it was at t'. It seems that the pressure here is to say either that "St. Petersburg" is not used literally here or that "St. Petersburg" is not used to refer, not to say the same thing about "Leningrad." I hadn't realized this when I started typing.
[UPDATE: Added the quantification over t'.]
Also--does anyone know why "becomes" is sometimes used to mean "suits," as in the title of this post? I just can't think of a half-plausible story, although perhaps it's obvious from the OED.
(There won't be any APA quotes here, because all the decent ones I remember are ones that someone might not want published, and that someone is me.)
Posted by Matt Weiner at March 31, 2004 01:07 PMI expected you to go in a different direction with this one.
(1a), read literally, fails for reasons unrelated to the point that you're making. It is false that "Before t, Leningrad was not St. Petersburg", but for historical rather than logical reasons. Before being 'Leningrad', the city was 'St Petersburg' (with 'Petrograd' in the interim.)
Is the example just meant to be cute this way?
PS: Matt, I enjoyed the chance to meet you in Pasadena.
Posted by: P.D. at March 31, 2004 04:34 PMIt certainly seems as though there is pressure to say either that 'St Petersburg' is not used literally or that it is not used to refer.
Here's a suggestion. Is it most natural to interpret 'Leningrad became St Petersburg' as having an element of disguised semantic ascent? That is, should the statement be interpreted as 'the city referred to at t by 'Leningrad' became referred to after t by 'St Petersburg''. By this reading, the term 'became' operates on the practice of referring to, as opposed to the city itself. I find this interpretation not only intuitive with regard to how one naturally parses the content of the sentence, but also possessed of the benefit of avoiding the problem that you mentioned regarding the necessity of true identity statements. Does it have any semantic plausibility?
Posted by: Will Davies at March 31, 2004 04:47 PMP.D.--Gah, I suppressed another quantifier--make that "At some time t' before t." I have my doubts about the truth of (3) as well, but the intended interpretations are pretty clear. (It was nice meeting you too, BTW--I'll send you an e-mail about one of the sillier discussions we had.)
Will--I think it's very plausible, and probably what those who thought "St. Petersburg" wasn't used literally had in mind. So the interpretation becomes, "Leningrad became known as 'St. Petersburg'", or even "The city that had been known as 'Leningrad' became known as 'St. Petersburg"--though there's no need to opt for the second reconstrual over the first.
As to the semantic costs and gains in these locutions, I don't feel qualified to pronounce. I think that it tells slightly against the idea that 'St. Petersburg' is used non-literally that there's no felt asymmetry between the names as used in the sentence (at least I don't feel one), but I'm not sure that should carry any weight. Maybe there's someone lurking around who has an opinion....
Posted by: Matt Weiner at March 31, 2004 05:54 PMI think Will is right that there is a hint of the metalinguistic in "StP became Leningrad". Here are some similar examples:
"Taber St. becomes Gano St. when you cross Waterman St."
"The Missouri becomes the Mississippi somewhere north of St. Louis." [Please do not assess this sentence for geographical correctness!]
"Saul became Paul on the road to Damascus."
This last especially seems to be metalinguistic, except that the guy undergoes a pretty serious change, too. In each case, to me there is at least a *whiff* of the idea that a soul has left a body and been replaced by another.
Feh! What you call the Mississippi is really the Allegheny. I can back this up with a bit of Googling.
Sorry, I know you asked me not to do that. I've got to go teach but what you say does seem sensible. I think there are some interesting questions about identity conditions in play here. Go to it, please, and I'll step back in when I'm back in the office.
Posted by: Matt Weiner at March 31, 2004 07:10 PMOK, a little bit of more serious response here. For most of the objects under discussion, I (and I think we) are unsure about identity criteria. For instance, I'm at least half-tempted to say that to be Taber St. is just to be called 'Taber St.', and that Taber and Gano are therefore two different streets. On the other hand, I'm also half-tempted to say that a street is the same street if it keeps going continuously and contiguously in the same direction--which presumably makes Taber and Gano the same street in different directions.
(This is reminding me of whether an author is identical with her pseudonym.)
That makes it easy to both that there is one street here, and that that one street is Taber on one side of Waterman and Gano on the other side. It doesn't sound metalinguistic because there really is a sense in which the street you're on is not identical to Gano on one side of Waterman--with just a little equivocation, you can say "The [one] street I'm on is not identical to Gano on this side of Waterman and is identical to Gano on that side of Waterman." But the second sense of identity turns out to depend on linguistic facts--to say "It's identical to Gano" just means "It's called 'Gano'."
That accounts for why your last example seems particularly metalinguistic--there's just no sense in which we think that changing someone's name is automatically sufficient to make him a different person. Take also "Jane Roe became Jane Doe on April 2, 2004" when Ms. Roe got married and changed her name. (Is it odd to say "Jane Doe became Jane Doe on April 2, 2004"?)
I like the geographic nature of the examples--obviously my definition in terms of time misses some cases.
Posted by: Matt Weiner at April 2, 2004 09:06 AMI think that it tells slightly against the idea that 'St. Petersburg' is used non-literally that there's no felt asymmetry between the names as used in the sentence (at least I don't feel one), but I'm not sure that should carry any weight.
I think it should, but the second reconstrual ("The city known as 'Leningrad' became known as 'St. Petersburg'") wouldn't have this problem.
Having not been there, what was Bach's take on it? I don't know how his views have evolved since Thought and Reference, but if he still thinks that 'St. Petersburg' is semantically equivalent to 'the bearer of "St. Petersburg"', the sentence just literally means (reading 'became' as you did):
At some time t' before t, the bearer of 'Leningrad' was not the bearer of 'St. Petersburg', after t it was.
I would think Bach would say we use this sentence non-literally (because for all we know there are multiple bearers of each name, e.g., St. Petersburg, FL) but he's pretty much committed to all our sentences involving names being non-literal in that sense. In any event, if that's what the sentence means, you wouldn't have the necessity of identity problem.
Posted by: Geoff at April 2, 2004 09:19 AMBach thought we use (1) literally, King thought not--at least that "St. Petersburg" isn't being used in the same sense here as it is in cases in which it is used to refer. Here is Kent's paper; this example comes up on p. 21, where he says that the name is being used as a predicate (and that it seems perfectly literal).
The point is supposed to be that singular terms (in this case, names) can be used in non-referential ways--eventually leading to the conclusion that (maybe) most of the time names don't refer semantically.
BTW, At some time t' before t, the bearer of 'Leningrad' was not the bearer of 'St. Petersburg', after t it was is false, isn't it? After t the bearer of 'Leningrad' is not the bearer of 'St. Petersburg', since that city has ceased to be the bearer of 'Leningrad'. Perhaps that's what you mean when you say that this is meant non-literally.
Posted by: Matt Weiner at April 2, 2004 01:26 PMI've refined my crazy view, the Nominal Description Theory, since Thought & Reference. Check out "Giorgione was so-called because of his name" (Phil. Perspectives 2002), posted at: http://online.sfsu.edu/~kbach/Giorgione.htm
In that paper there are some nice examples involving Eminem and P Diddy, among others, but let's stick with the one I used in Pasadena.
(1) In 1991 Leningrad became St. Petersburg.
The likely intended use is in my view a literal use, but that view says that a name expresses the property of bearing that very name, and that when occurring as a standalone NP (without a determiner) a name is semantically like a definite description. So (1) says that in 1991 the bearer of 'Leningrad' became the bearer of ' St. Petersburg'. (As P.D. pointed out, Leningrad also used to be St. Petersburg.)
Of course, if you assume that proper names are Millian, you'll naturally find the likely intended use of (1) not to be literal, because metalinguistic. I don't assume that. Instead, I argue that proper names only SEEM to be Millian because, like most definite descriptions, especially incomplete ones, they are standardly used to refer. And there's a reason for that.
BTW, At some time t' before t, the bearer of 'Leningrad' was not the bearer of 'St. Petersburg', after t it was is false, isn't it? After t the bearer of 'Leningrad' is not the bearer of 'St. Petersburg', since that city has ceased to be the bearer of 'Leningrad'.
I guess it depends on what it means to be the bearer of 'N'. Speakers can still refer to that city readily using 'Leningrad', even though that's not its official name.
In any event, if the sentence does means that (or what KB above says it means), then it's literally false, since there is more than one bearer of 'St. Petersburg' and 'the bearer of ...' implies uniqueness. So on this view of the meaning of a name, I would think that the sentence (though not the name 'St. Petersburg') is standardly used non-literally.
That's what I meant earlier, though KB's comment appears to imply that on his view the sentence is standardly used literally, and since I was trying to figure out what he would say, I must have made a mistake.
Posted by: Geoff at April 5, 2004 05:25 AMPeople think I'm nuts to believe this, but I believe that uses of incomplete definite descriptions are standardly nonliteral. Ditto for shared proper names. In both cases, we make a temporary pretense of uniqueness (see "Descriptions: Points of Reference," esp. Point 12).
One reason people think I'm nuts, at least about this, is that they don't have their usual intuitions of nonliterality about such cases. But in various places I've tried to explain the difference between familiar kinds of nonliterality, which drive these intuitions, and other kinds. We don't have such intuitions about pragmatic regularities, like standardized nonliterality and, to mention the best known case, generalized conversational implicature.
As for reference to individuals by names they no longer possess, don't we use locutions like "the former Cassius Clay"? A bit trickier is "the artist formerly known as 'the artist formerly known as "Prince"'"!
Posted by: Kent Bach at April 5, 2004 10:16 AMSo (1) says that in 1991 the bearer of 'Leningrad' became the bearer of ' St. Petersburg'.
I'd think you can make this come out true (modulo the multiple St. Petersburgs) by making it an existential quantification: There is a unique x s.t. before 1991 x was the bearer of 'Leningrad' and not the bearer of 'St. Petersburg', and after 1991 x was the bearer of 'St. Petersburg'. As in:
(5) On July 4, 1776, a motley collection of colonies became a nation.
Kent, if I'm following correctly you think that 'St. Petersburg' need not be used nonliterally in (1) because it's a predicate, and so it doesn't require uniqueness; is that right? So we could say "...the bearer of 'Leningrad' became a bearer of 'St. Petersburg." In any case any nonliterality springing from the shared proper name would be very different from the nonliterality Jeff King* ascribed to predicative uses.
*A great disappointment as an overall no. 1 draft pick, but apparently a quite good Iditarod racer.
Yes, Matt, I think you got me right, esp. about nonliterality. In my view, standalone proper names occurring in argument position are like incomplete singular definite descriptions, which imply uniqueness (notice that this uniqueness is not encoded in the meaning of ‘the’, which implies totality, but is implied by the combination of singularity and totality). Jeff was claiming a different sort of nonliterality. He maintained that it is not literally true that in 1991 Leningrad became St. Petersburg - it was all along - whereas I claimed that it came to acquire the property of bearing 'St. Petersburg' and that (1) expresses this.
Btw, notice that 'becomes' is almost as big a "disgrace to the human race" (in Russell's famous phrase) as 'is' is. It can mean, roughly, come to have a certain property or come to be a certain other thing. Few metaphysicians think that the latter is possible -- except, perhaps, where fusion, as in Matt's example, or fission is involved. But you have to be careful how you put this, since you don't want to say that two things can each come to be identical to one thing or that one thing can come to be identical to two.
Posted by: Kent Bach at April 7, 2004 01:02 PM