Q: `workaholic'

[bwald]

----------------------------Original message----------------------------
Larry Trask writes:
 
>...Is there an accepted name for the slightly peculiar process
in which a piece of a word is somewhat arbitrarily ripped out of it
and then used as a kind of affix for forming new words?
 
>I'm thinking of cases like these:
alcoholic --> -(o)holic --> workaholic, chocoholic, shopaholic,...
Marathon --> -(a)thon --> telethon, bikeathon, danceathon,...  [..., e.g.,
talkathon]
panorama --> -(o)rama --> washorama, launderama,...
>Watergate --> -gate --> Irangate, Contragate, Whitewatergate,...
 
I'm not sure, but maybe this is the type that many dictionaries (and other
sources) refer to as
"STUMP COMPOUNDS".  Current dictionary examples (not all assured longevity)
include:
infomercial, infotainment,  docudrama, docutainment, hairtician, sitcom
 
The idea of "stump" seems to be that they were not originally cut according
to a historical morpheme boundary but according to some other (metrical or
rhythmic) principle.  Hence, "workaholic", like "gasahol" (or is it
"gas*o*hol"?), does not presuppose a morphemic cut
"alc" + "ohol"  Meanwhile, the term "compounding" suggests that one or more
"stumps" have grown semantic lives of their own and have come to have
meanings that allow recombination, or at least allow analysis into separate
meaningful elements, related to the morphemes they are stumped from.
(True, compounds are usually considered to be formed from independent
words, in current English, but the "stump" takes the precaution to suggest
that each element is the "stump" of a "word".  NB: the whole "gate"
syndrome is a secondary phenomenon, since it has nothing to do semantically
with the word "gate", but with political scandal since "Watergate", a
proper name.  Hence, Larry is quite right to include that type with the
others.)
 
If all the examples listed above deserve to be called "stump compounds",
then they are of various types.  "sitcom" is a compound of two stumps,
neither of which occurs as an independent word, cf.  or "fizz ed" (i.e.,
phys ed) -- but isn't there "Board of Ed"?.  But in such cases, the
BEGINNING of both words are stumped.  In contrast, "docutainment"
resembles a BLEND, where the stump preserves the beginning of the first
word but the END of the second. Next, "HAIRtician" (an aesthetically
displeasing word), like most of Larry's examples, e.g., "WORKaholic", has
an INDEPENDENT word followed by a stump.  Thus, it combines an independent
word as a first element with a stump as second element.  Nevertheless,
Larry has double stumps in "chocoholic" and maybe "telethon" and/or
"launderama".  (True, "tele" is an isolable root, but in "telethon" it
seems to be directly related to "television/televised/etc", rather than
directly to the "tele" root.  Similarly, "launderama" stumps at "laund",
not the word "launder", though the "er" here can be considered an overlap
blend shared by "laundER" and "ERama")
 
The "info(mercial/tainment)" words are different again, because "info"
exists independently as a stump, cf. "math(s)" (true, there is the morpheme
"math" as in "polymath", but isn't "math" a stump for "mathematics", the
British form "maths" saving the final -s of "...icS").  Thus, it combines
the features of double stumps with WORD+stump.
 
As anticipated above, a BLEND seems to be a specific kind of stump, where
the first element stumps to the BEGINNING and the second stumps to the END.
I can understand why some might then suggest that where the first element
is a monosyllabic word (not stump-worthy) but the second stumps to the end
the term "blend" is appropriate.  Meanwhile, there is an unusual *semantic*
step in "workaholic" that Larry alludes to by saying:
 
>I don't think `workaholic' can reasonably be regarded as a blend of `work
>alcoholic'
 
The step is generalisation of "alcoholic" to any kind of "addict", hence,
"workaholic" quite
reasonably becomes "work addict".  However, it does not seem that
"alcoholic" as "addict" in
general exists as an isolated entity.  Are there any other cases like this?
(It would not surprise since a humorous intent is obvious that would not
appear in the straight-forward "work addict"; "gasohol" is similarly
humorous, betraying the degree of seriousness with which it should be
taken).  Thus, the problem seems to be one of separating various processes
that are collapsed in "workaholic", one involving semantic change of an
element (for humorous purposes), and the other a purely formal stump
compounding (or blending?)  which retains the meanings of the unstumped
sources.
 
Now are the following stumps or blends, or both?:
 
stagflation, workfare, gues(s)timate
 
Somebody may have worked out a more detailed typology/terminology of all
these kinds of words.  If so, let's hear it.  Also consider the types of
"dognap", "carjack" etc.
 
Now, as long as we're dealing with "compound" related phenoms (another
independent stump), I have a question.  It begins with the observation that
in Middle English -er became an increasingly common type for certain
purposes as earlier suffixes with the same meaning became phonetically
disfunctional, cf. grasshopp-e became grasshopp-er by morphological
replacement.  Now it seems things are going "back" in the opposite
direction.  What's the "term", and, in fact, the phenomenon by which the
derivational suffix -er has been getting lost or deleted in compounds so
that we have such compounds as:
barkeep, nosewipe, sodajerk, bellhop, carhop, bedsit,  bellpull, bellpush,
cocktease, keypunch,
etc.
 
where a final -er might be expected (and might even be attested earlier)?
Does the process also apply to the case of "doorstop" and even "shortstop",
where at least for "doorstop" we might assume it is related directly to
"stop" as an alternative to "stopper" (hmm. maybe also "tease" for
"teaser", cf. "he's such a tease(r)/jerk(*er)")   -- Benji