LINGUIST List 4.567

Mon 26 Jul 1993

Qs: Compounds, Russian, ARTFL, Mandarin

Editor for this issue: <>

Directory

1. Beard Robert E, Request for information on compounds
2. "STEVE SEEGMILLER", The Great Russian Vocabulary Hoax?
3. , ARTFL
4. Allan C. Wechsler, Phonology of Mandarin

 I am working on a project to determine the modifier-head order in
nominal compounds in a GB-like framework without resorting to movement.
The problem is seen in English compounds like _lion-hunter_ which seem
related to _hunt lions_ but whose modifier(adjunct, complement,
specifier)-head order is just the reverse of the phrase. The working
hypothesis of this project is that at least some languages maintain a strict
modifier-head order which holds across syntax and the lexicon, and is most
predictably reflected in the relation of adjuncts to nouns in NPs, e.g.
_red car_ = M + H.
 If this hypothesis is true for some languages, conflicts arise
between the Hoeksema-Stump 'head-application default' and the availability
of prefixes and suffixes to mark compounds. The head-application default'
is the preference affixes show for the heads of compounds, e.g.
_understand_ : _understood_ and _mothers-in-law_. The conflict arises in
languages whose modifier-head order is M + H if the language is
predominately prefixing and those with H + M order if the language is
predominately suffixing. I predict that if a language is compelled by
default modifier position to locate the compound modifier in that position
normally occupied by the affix if the derivation were not compound,
affixation will be forced by the head-application default to an internal
position (_mothers-in-law_) or to be omitted (_pick-pocket_). The first
method seems very unpopular for derivation, probably because it violates
Lexicalism; the second appears in French and other Romance language
(_coup-eur_ "cutter" but _coupe-fil_ "wirecutter"). This variation in noun
compound affixation, in other words, is predicted by the default order of
the single adjunct modifying a noun in a NP: A + N in English, N + A in
French. (_Pick-pocket_ compounds are, of course, unproductive in
English).
 This principle holds in a substantial number of IE, Turkic, Finno-
Ugric, Bantu, and Paleosiberian languages but the grammars seem to
indicate that Amerindian languages like Kiowa, Navajo, Dakota, and Koasati
do not follow this principle and often have opposite modifier-head order
in NPs and compounds, e.g. Kaosati _isa hatka_ = house white "a white
house" but _alahko-bit-li_ = gourd-dance-Suf "gourd-dancer (moth)". I am
also having difficulty finding data from Turkana, Hausa, Somali, Djirbal,
Diyari, or related Australian languages, some of which also seem to be
problematic. I would deeply appreciate two bits of information on these or
related languages or any other language with an interesting pattern
between single NP adjunct and noun compound order:
 1. The 'normal', 'default', or 'most common' order of Adj+N and V+Obj
sequences in the language, and
 2. some typical examples of synthetic V+N, A+N, and N+N nominal compounds
with meanings like those of _lion-hunter_, _lion-hunting_, _lion-huntery_,
_left-hander_, _(an) egg-head_, _(a) long-hair_.
 Any other advice on the languages mentioned above or any other which
reflect or fails to reflect this parallel would also be appreciated.
Please address all responses to me, <rbeardbucknell.edu> and I will
summarize the results and share them later on.
 --Robert Beard, Linguistics Program, Bucknell University

The July 27, 1992 issue of The New Yorker contains an article by
Robert Cullen entitled "Siberia" in which the following sentence appears:

 Clumps of birch and evergreen trees broke up the vista of white
 rolling fields, populated by occasional flocks of black-and-white
 magpies or red-winged finches called _sneguri_ -- a name derived
 from on of the dozens of Russian words relating to snow.

Well, it's been a long time, but I studied Russian for a number of
years, and the only word relating to snow that I ever encountered was
_sneg_ (along with a coue of clearly related forms like _snez^nyj_.
This is also the only word listed in my (admitteldy poor) Russian
dictionary. Did I miss something important, or are we seeing here
an incipient myth comparable to the one about Eskimo words for snow?

Steve Seegmiller
<seegmillerapollo.montclair.edu>

I would like to know if anyone has used the ARTFL corpus for syntactic
analyses?

L. Lapierre
Department of French
Dalhousie (NS, Canada)

How big is the vowel repertoire of Mandarin Chinese? This
question is prompted by the bizarre mapping of graph to segment
in the Pin1-yin1 orthography. I also have a possibly garbled
memory of someone claiming that the vowel repertoire was in
fact much smaller than one would guess from the phonetic
range on the surface; did someone claim there were only
two vowel phonemes? Can anyone reconstruct the reasoning for
me?