laurence.horn at YALE.EDU
Wed Jan 30 12:25:54 UTC 2002
At 7:59 PM -0500 1/30/02, James A. Landau wrote:
>In a message dated 01/30/2002 5:24:59 PM Eastern Standard Time,
>einstein at FROGNET.NET writes:
>> There are a number of well-established noun adjuncts like "baby carriage,"
>> gas station," "government agency" which may be used as the models for "can
>> food drive," "L. I. ice tea" and the like. They are gaining acceptance
>> although the individual ones like "box set" or "bottle water" may strike us
>> as novel.
>I think that there are three separate processes that can contribute to the
>dropping of the past participle suffix.
>1) phonetic. There is a limit to what even native speakers of English can
>accomplish. In "boxed set", even by English language standards, the phonetic
>combination /ksts/ is preposterous. Try saying it out loud. You will either
>say /bahkset/ or have to insert a noticeable pause before the /s/ of set.
>Similarly with "iced tea". To pronounce it as /aisdtee/ requires you to quit
>voicing a stop (plosive) halfway through, which I suppose can be done but
>which is not a normal phonetic pattern in English.
You're right about the phonetics, of course, but there's a stage of
reanalysis needed. I would always have referred to those
three-in-one Vox Boxes some of my fellow ancients may remember as
"boxed sets" in writing, but I'm sure I'd have pronounced them as
"box sets" for the reason Jim gives.
>3) occasionally there is the influence of a similar-sounding expression. For
>"box set" the speaker may unwittingly draw an analogy with "box seat".
Good point, as is the one not copied here about the semantic grounds
for reanalysis. I'm just saying that there are speakers who delete
the /t/ in production but still have the participle in their mental
(and orthographic) representation. Incidentally, I think the
undeleted version would have to be /aistti:/ with a geminate, not
/aisd/ with voicing: compare "I'd like my coffee iced", which can
only be /aist/.
More information about the Ads-l