Between you and me, I keep finding that just when I'm beginning once again to be prepared to trust etymologists, they go and do something that makes me want to forget about it again. I can definitely believe that worm and crimson have a common source: there's the crucial [r] and [m] as the main consonant sounds in the root, and we know that crimson pigment was made from worms so there's a meaning connection, and I'm just beginning to develop some trust... and then they tell me that rhapsody is also a derivative of the same Indo-European root (really, it does say that under wer-2, and also under wed-2, in the American Heritage Dictionary appendix on Indo-European roots — Mark didn't just slip that one in to see if you were paying attention this morning).
Rhapsody. Yeah, right. Worm is related to rhapsody because of the [r], and rhapsody is related to saxophone because of the [s], and saxophone is related to onyx because of the [o] (and hey, there's an x too!), and onyx is related to asininity because of the [n]... Suddenly trust has evaporated and it all seems suspicious again, and I'm remembering the satirical remark of Mark Twain's that the name Middletown is derived from Moses (by loss of -iddletown and addition of -oses, don't you see). I don't do etymology. For me (and I suspect many linguists) it is the topic that first made me think I might be interested in linguistic study, when I was a kid, but then I found it was entirely different things about language that really made linguistics exciting for me. I never went back to word origins, and I doubt I ever will (far too much specialized knowledge of ancient languages is called for) — even though it is perhaps the thing linguists are most likely to be asked about by members of the general public.
[Update: I have started getting emails from philologically very competent people (thanks in particular to J. S. Bangs) who normally begin by asking me if I am serious (the answer is no: language is much too fascinating to be entirely serious about) and then, as if losing faith, start very solemnly trying to convince me that rhapsody really is related to worm.
This is not something that I really doubt. In brief, the -od- bit is a stem meaning "song" (and of course we have the Greek-derived word ode from that), and the rhap- part (in which the -h- doesn't matter; it's just a Greek thing, like the connecting -s) is from a root meaning something to do with folding or bending or sewing or stitching or wrapping that may actually be related to the English word wrap (in which the w- doesn't matter; it's just a Germanic thing), so all that remains is to find a way to tell a plausible story about how rhapsody originally meant "sewn or stitched song" (don't ask me to explain this; I've already told you more than I know)... I don't want to mislead you: etymology for Indo-European is highly developed, and they're not telling lies. I'm not saying that I really truly think they are making stuff up. I know these things they say are true. That is not my problem. My problem is that I sometimes find I cannot make myself feel like I think they are true. You see the difference?
Just to increase my discomfiture, John Anderson (thanks a lot, John) has insisted on reminding me of the following facts about the staggeringly widespread Indo-European language family, which more than a millennium ago had already established itself from Iceland in the north-west to Ceylon (now Sri Lankaa) in the south-east, and was once found as far east as Chinese Turkestan: the English word head is actually related in ancestry not only to capital but also to chapter; precocious is historically connectable to apricot; the word hound is related to cynic; and weird is related to rhombus (and worm and rhapsody and stalwart and vertebra and wrath and wrong and wrestle and briar, of course). I repeat, I know perfectly well that these things are true. I just can't make it feel like they are.]Posted by Geoffrey K. Pullum at December 8, 2006 12:12 PM