cosmic objects and linguistic objects
Well, I've been scooped on this whole current "definition of planet" issue and its linguistic interest, which is a shame given that I'd intended to post about it last February but was too unmoved to do so at the time. But piloklok is not about scoops, so I'll still add my thoughts.
Notably, two things written recently by Geoff Pullum make it clear how this debate is relevant to linguistics: first, indeed, this issue is not a matter of changing the universe; instead it's simply a matter of lexicography. It's a question of using definitions that provide a set of criteria to determine whether a particular entity (e.g., an individual heavenly object) fits into a larger category (e.g., planet).
Second, Pullum agrees with a correspondent that the general public's likely dissatisfaction with a definition that denies "planethood" to Pluto is similar to a prescriptive and by-rote approach to grammar instruction. In short, people do not want to admit that what they learned in school, such as the planetary status of Pluto, is no longer current (and perhaps never should have been). Likewise, the same people do not want to admit that what they learned regarding grammar (take your pick - singular-referring they/their, stranded prepositions, whatever) never should have been current.
Here's another parallel to consider: this is not only a discussion of definition and lexicography; it closely mirrors the linguist's choice of how to define linguistic categories. For example, here are some possible criteria that could be used to distinguish planets from other big things in outer space: (a) Made of rocks not ice; (b)spherical by virtue of its own mass; (c), in the same plane and solidified from the same nebulous stuff as other planets in the same system (and the central object). The first two criteria focus only on properties of the object itself, while the third relies on how that object behaves in relation to other big things in space.
Likewise, in linguistics we are confronted with how to develop suitable sets of criteria to define categories as varied as gesture, feature, phoneme, word, stem, noun, verb, utterance, message, and so on. Take noun: we have the elementary school definition, "a person, place or thing", and the syntactic definition, "word that can occupy subject or object position in a sentence". The first definition looks desperately to properties of the word itself (in this case, what can it refer to?), while the other relies on how this category behaves in relation to other linguistic objects. That this more suitable definition relies on close analysis of the structure of a sentence, rather than a handy memnonic, makes it less likely to be a widely-accepted means of defining nouns.