The OED offers, as one of two meanings for the lovely nineteenth-century coinage falsism, the definition "a platitude that has not even the merit of being true," noting further (just in case it might not be evident) that "the word owes its meaning to the antithesis with truism."
There are an unfortunate number of falsisms about language in current circulation, and in a recent post at Language Log, Mark Liberman and his correspondent Robert Neal Baxter demolish one of them. Citing an informative essay by Victor H. Mair, Liberman points out that, contrary to what some non-sinophone armchair etymologists assert, the Chinese word for 'crisis' is not a compound of the words for 'danger' and 'opportunity', and then quotes Baxter's analogy showing why the falsism, even if it were a truism, would still be a thoroughly pointless sort of platitude:
Would it be fair to assume that English has no word for what the French refer to as 'papillon', resorting instead to a compound made out of the words 'butter' and 'fly'. What would such a statement, even if it were linguistically valid - which it isn't - show about the language or the speakers of the language in question? Probably very little. In fact it's about as likely that a Chinese speakers using the word 'crisis' made up of whatever morphemes it happens to be made up of is to be aware of this secondary reading as an English speaker would be to think that butterflies are some sort of 'air-borne grease balls'.
To this Liberman adds, "Didn't Michel Foucault once point out that butterfly expresses a fundamental contradiction in anglophone culture: the libertarian urge to take wing, subverted by the consequences of a diet too rich in animal fat?" (I know just enough about Foucault to have no idea whether Liberman is joking. At any rate, German-speaking peoples must be implicated as well, for the root of Schmetterling is cognate with smetana, which is a Czech word that denotes sour cream as well as the composer of The Bartered Bride (right).)
At any rate, Baxter is quite right in pointing out that an English speaker is no more likely to consider the word butterfly in terms of its component morphemes than a German speaker is to think of a nipple as being a wart on the breast (Brustwarze), if I may borrow an example from Douglas Capitalcitizen. Whatever first prompted someone to call a butterfly a butterfly is now thoroughly opaque to contemporary English speakers, and though we can be provoked to see the word anew if our attention is called to it (as described in Viktor klovskij's essay "The Ressurection of the Word"), this is not something we ordinarily do.
Still, butterfly is a funny thing to call a butterfly, isn't it? It's also not obvious exactly what the compound means—okay, so it's probably right-headed, and therefore refers to some sort of flying insect. But what, exactly, is the relation between the 'butter' part and the 'fly' part? (OED sez: "The reason of the name is unknown," but offers some speculation, to which I return below.) There are several possibilities. I hope that the Language Loggers will forgive me for saying this, but Sanskrit has at least four words for 'compound', and I intend to use them here to illustrate the multiplicity of possible meanings of butterfly.
- Butterfly could be a karmadharaya compound, with the word 'butter' describing some attribute of the particular sort of 'fly' denoted by the whole. The 'airborne grease-ball' reading potentially fits into this category, provided it is not meant literally, but it is far from the only possibility: within this same general category of interpretation, a butterfly could also be a butter-coloured fly, or a fly that smells like butter, or a fly that habitually alights on toast. The number of ways in which an insect may resemble butter is limited only by the power of the oberver's imagination to fabricate analogies.
- Butterfly could be a tatpurusha compound, in which the relation is one of interaction rather than resemblance. For example, a butterfly could be an insect that eats butter, in which case one would have to wonder, as Alice did of the bread-and-butter fly, how it could possibly survive without human intervention. Or it could be quite the reverse—an insect that shits butter, as suggested by the OED: "Wedgwood points out a Dutch synonym boterschijte in Kilian, which suggests that the insect was so called from the appearance of its excrement." Trouble is, as A World for Butterflies points out, butterflies don't shit. (Caterpillars do, though, and apparently there is one species that, thanks to a diet of yellow flowers, does emit appropriately coloured frass. (Yes, caterpillar shit is called frass, and yes, it's derived from fressen. The OED defines frass as "the excrement of larvæ; also, the refuse left behind by boring insects," and although I'm sure Nabokov (whose birthday was Earth Day) would have insisted that there are no boring insects, I will not.))
- Butterfly could also have been a dvandva compound, in which the two parts of the compound have equal status (as in bittersweet or fax-copier). I have trouble imagining what something that is both butter and a fly would look like, or how a taxonomist might cope with the intersection of the lepidoptera with the dairy products, but it's certainly a semantic possibility if not a pragmatic one.
- Finally, we could give up on the notion that butterfly must denote a fly at all, and say that it's a bahuvrihi compound, like pickpocket or bluestocking. It could denote a person who has so much butter that they leave it out for the flies, or perhaps someone who transports butter by airplane. With a little bit of imagination, you could even get it to mean 'a sexually promiscuous person', the conceit being that such a person slips out of his or her trousers so readily as to suggest that the zipper is greased.
All of this, of course, brings us no closer at all to understanding the origin of the word butterfly, but by this time I've typed the word often enough that it no longer looks as if it means anything, and so I arrive at the wood where things have no names.