ERR; BUM

[Laurence Horn]

At 4:53 PM -0500 2/4/01, D. Ezra Johnson wrote:
>Regarding "err" -- does the American Heritage Dictionary really have it
>pronounced "air" as one of the pronunciations? I know some people have a
>merger here, but I would have thought the dictionaries would have the open e
>sound, as in the first syllable of "error".
>
>Regarding "bum" (verb), as in "Can I bum a smoke?" -- does anyone think the
>below is possible? I am taking a Haitian Creole course, and "give me" is
>"ban-m" with a nasal vowel plus [m]. It reminded me of the English "bum" --
>and the dictionary I looked in derived one meaning of "bum" (that relating
>to a homeless person) from a German term meaning to wander or stroll around,
>but it allowed that other senses of "bum" could have "other origins". Could
>the Kreyol "ban-m" possibly be one of them?
>
>Daniel
>_________________________________________________________________
The RHHDAS regards the verb "bum" as a back-formation from the
agentive-looking noun 'bummer' ('a shiftless person, as a beggar,
vagrant, sponger of drinks'), which in turn is taken to derive from
the noun"bum", although there's a clear misprint here, since it
derives "bummer" from its "bum-1" ('the buttocks, anus, rectum;
ass'), where clearly what is intended is its "bum-2" ('a contemptible
or despicable person').  Under the latter, we are told 'orig.
uncert.', but referred to subsense 2a ('a lazy, shiftless person who
does not or will not work; a beggar, tramp, or esp. a habitually
drunken derelict'), which is seen as bearing a close resemblance to
an early Scots word, 'itself of unknown origin' that 'seems never to
have been in widespread use'.  But given the extensive range of
citations from the British isles, including the predictable cite in
Farmer & Henley (1888), I suspect the resemblance to the Haitian
Creole form may be just fortuitous.

larry

P.S.  I don't see any particular reason why the verb meaning 'to
mooch' couldn't have been a simple zero-derivation from the noun
"bum-2" (not "bum-1"!), with the agentive "bummer" derived from that
in the usual way:  bum (n) > bum (v) > bummer (n).  Both the
chronology of the cites and general word-formation principles would
seem to support this over the vacuous agentive noun and back-formed
verb assumed in the bum (n.) > bummer (n.) > bum (v.) derivation the
RHHDAS favors.  True, "bummer" OUGHT to be blocked by "bum" (with the
same meaning), but that consideration applies to both stories, and
blocking isn't really an absolute.