Benjamin Zimmer bgzimmer at RCI.RUTGERS.EDU
Mon Oct 17 04:33:09 UTC 2005

On Sun, 16 Oct 2005 17:08:38 -0700, Arnold M. Zwicky wrote:

>On Oct 16, 2005, at 2:04 PM, Wilson Gray wrote:
>>> From Letters in today's NYT Magazine:
>> "[NP] may have caused more than one clerk a reprimand, if not a job."
>> Presumably, what the writer was reaching for was:
>> "[NP] may have cost more than one clerk a reprimand, if not a job."
>that's just as bad as the original.  the problem is that if X causes
>Y NP, Y receives NP, but if X costs Y NP, Y gives (up) NP.  so "may
>have caused one clerk a reprimand" is fine (the clerk gets/receives a
>reprimand, i.e., is reprimanded), but "may have cost one clerk a
>reprimand" is not, at least if understood in the historical sense
>(the clerk does not relinquish a reprimand) -- though it's possible
>that some people have extended "cost" to mean 'be negatively
>affected', in which case it might be possible for the second object
>to be understood as denoting the negative consequence.  the question
>is whether some people can say things like "My one mistake cost me a
>final grade of D" 'the cost to me was (that I had to accept) a final
>grade of D'.  for me, that's just impossible, but i could see how it
>could happen.
>but there's nothing wrong with "may have caused one clerk a
>reprimand".  there *is*, of course, a problem with "may have caused
>one clerk a job".
>in any case, this looks like blendish rather than eggcornish.

So, leaving aside the coordination with "caused X a reprimand", you'd see
"caused X his job" as a blend of "caused X to lose his job" and "cost X
his job"?

I've put this in the database, marked questionable. Comments welcome.

--Ben Zimmer

More information about the Ads-l mailing list