Wolfgang Schwarz

I guess that the thought behind (3) is that "the actual number of the planets" changed its meaning because "planet" changed it's meaning. Putting my reference-magnetism hat on, I'd be a little sceptical that the committee really managed to pick out the unique magnetic kind in the vicinity. (if I've got my facts roughly right, why should "planet in orbit around sun, not a satellite" (~classical planet or dwarf planet) be any less magnetic than "planet in orbit around the sun, not a satellite, and which has cleared it's orbit of stray debris" (~planet)?)

The idea that these extensions were equally eligible, and so (perhaps) "planet" had an indeterminate extension till just now seems equally plausible to me, reference-magnetically speaking. (Perhaps you'd then regard the latest decision as the scientists changing meaning in the sense of changing usage so as to resolve the indeterminacy).

But if a case could be made that the reference-magnet facts stacked up in the right way, I'd be happy to go for (2).

Here's one thought that does seem natural from the ref-magnet point of view: the options aren't that "TANOP" used to rigidly designate 8, or used to rigidly designate 9. Rather, the options are that it used to rigidly designated 8, or used to rigidly designate n where n is whatever the number of classical planets +dwarf planets is.

Last remark: isn't a dwarf planet a planet? (or is it like a rubber duck?) If so, surely what was resolved was that the actual number of planets is really really large!

I've just read the two Wikipedia articles now linked to in the main post. It's interesting how astronomers always tried to keep the definition of "planet" in line with what, on their current view, constitutes a natural class of objects: first, any heavenly body that is not a fixed star counted, including the sun and the moon, but excluding the earth. Then, with the heliocentric view, a planet was taken to be any non-satellite orbiting the Sun, which by the mid 19th century included Ceres, Juno and others. Then, when too many of those were discovered, the definition was restricted to sufficiently large of those objects, apparently without a clear borderline. Now they realised that if Pluto counts, then at least three, and probably dozens, of others will have to count as well, so they looked for some reasonably natural borderline in the vicinity, and they found two such borderlines, even though both are still quite vague.

By opting for (3), I didn't mean that until last week, "planet" determinately meant something of which there are 9. In fact, now that I've read that 2003 UB_313 is larger than Pluto, it's clear that whatever "planet" meant until last week, it certainly didn't comprise Pluto but not 2003 UB_313. So the number of planets definitely wasn't 9. (And Quine *did* get the astronomical facts wrong.)

In fact, I now believe that things were just as you suggest: "planet" used to be indeterminate between planets and planets+dwarf planets, or rather (as "planet" and "dwarf planet" are still vague) between many precisifications of those two terms. That fits how competent speakers would have classified various possible objects. So it isn't reference magnetism of any strong kind.

What I find objectionable (and what I meant by option 2) is that "planet" always (perhaps even back when the sun and the moon still counted as planets) meant what the IAU now decided (or even some precisification of the new definition), even though nobody realised this and nobody would have classified possible objects accordingly.

Oh and yes, "dwarf planet" seems to be like "rubber duck".

Interesting post. I have a thought about the second option that relates it to Putnam's twin earth article. In that article he said that "water" always denoted water even before anyone was able to discern the things that would make water differ from twater. It seems somewhat similar to this case where "the actual number of planets" actually denoted 8, even before astronomers were put into a position where they had to very carefuly scrutinize the boundaries of the class of planets. We were always talking about the number 8 even though we couldn't discern this fact. This sort of argument seems like it would only go through if the planets formed a natural kind. I'm just playing devil's advocate here. I think that referential theories of meaning could easily have this result, but I think that the third option is more sensible.

What should we say about whether a dwarf planet is a planet? A rubber duck isn't a duck, but a dwarf rabbit is a rabbit.

Hi Shawn. Yes, I think many philosophers hold views that would seem to commit them to the second option. On these views, as soon as you introduce a term like "planet" for certain objects (Mars, Venus, Jupiter, Sun and Moon, say), the term automatically comes to denote the most natural kind in the vicinity, even if that differs widely from the theoretical role associated with the term and even classifies some of the exemplars differently. But in the case of "planet", this seems very implausible. Conversely, I think if one accepts that "planet" slightly changed its meaning in 2006, one shouldn't be too afraid of saying that "water" and "fish" slightly changed their meanings (mainly by getting more determinate) when it was decided that heavy water is water and dolphins aren't fish.