LINGUIST List 2.120

Saturday, 6 Apr 1991

Misc: Clitics,Syntactic complexity,Banned lgs,Poetry, Software

Editor for this issue: <>

Directory

1. "JOHAN ROORYCK", query: imperative + clitic ordering
2. Koenraad De Smedt, syntactic complexity
3. Bernard Spolsky, Banned languages
4. , Poetry
5. Jonathan Mead, Renumber Software

It is well known that Romance languages have enclitic ordering
in imperatives even if the unmarked order in the language is proclitic:
Donne-le-lui
Give it to-him'

Je le lui donne
I it to-him give'

The descriptive generalization also seems to hold for completely unrelated
languages such as Albanian and Modern Greek (Rivero 1988).
My question is the following: is anyone aware of languages which arguably
have clitics, but nevertheless proclitic ordering in imperatives? If so,
please give me some references. I would also be interested in languages
which confirm the general pattern of enclitic ordering with imperatives.
I pretty much checked the various Romance dialects myself, but it would 
be interesting to know any examples confirming or falsifying the observation.


Johan Rooryck

In a recent posting, Frederick Newmeyer wrote:

%% It's quite true, I'm
%% sure, that (functional) pressure on the parser explains why in V-O
%% languages heavy constituents tend to appear at the right (Hawkins)

Now I'm very interested in a psycholinguistic account of syntactic
complexity; so, does anyone have anything to say about this or has
pointers to the literature about this? I have myself built a
computational model for language production (rather than parsing as
Hawkins presumably does) that is based on the assumption that if several
constituents can be produced in parallel, then complex constituents will
tend to be syntactically completed (and thus uttered) later even if
their content was given sooner than that of shorter constituents. It
would be nice to see if constraints on production and those on parsing
tend to converge on this issue. It would also be nice if there was
psycholinguistic work defining factors of syntactic complexity in some
measurable way (Should one count in milliseconds per syllable, per word,
or per node in the syntactic structure?:-)

For some work on production that I mentioned, see my contribution in the
forthcoming book edited by Adriaens, G. & Hahn, U. 'Parallel natural
language processing' (or something similar) to be published soon by
Ablex.

Koenraad de Smedt

Does anyone know a list of languages that are legally banned in various
parts of the world? I would be interested in cases where the use of a
named language is specifically prohibited (as opposed to cases where the
ban is implicit in the requirement to use anothrer language).

Bernard Spolsky <F24030Barilvm.Bitnet>
Department of English Telephone: +972-3-531-8239
Bar-Ilan University Home: +972-2-282-044
52 100 Ramat-Gan Fax: (office) +972-3-347-601
Israel

If you want to contribute poems to the next volume of poetry by linguists,
send 3 copies of up to 10 poems by May 1, to NAPOLICAMPUS.SWARTHMORE.edu.
If you want your poems back, use regular mail and enclose an SASE.
Thank you.

[Moderators' Note: The following is an advertisement for shareware,
 and is offered here solely as a service to the subscribers
 of LINGUIST. We have no personal knowledge of the software
 in question, and have no idea of its effectiveness or its lack 
 of it.]

RENUMBER for the IBM PC and Macintosh

RENUMBER is a program that renumbers both examples and
references to examples in a linguistics text. It takes the
misery out of adding or deleting examples or finding two
examples with the same number. RENUMBER allows sequential
numbering of separate files as well as the use of labels
(e.g. ECP) instead of numbers. It is also clever enough to
change only the example (i.e. 14a --> 15a). Best of all,
RENUMBER is fast: a large document can be renumbered in a
few minutes.

The current IBM version of RENUMBER (1.3) is compatible with
many popular word-processors including WordPerfect (4.0-
5.0), Microsoft Word (4.0-5.5), Wordstar (3.3), FinalWord II
and Borland Sprint and any program that can generate ASCII
files.

The current Macintosh version (1.2), works with any word
processor that can save to Rich Text Format (RTF) files
including WriteNow version 2.0 and Microsoft Word versions
3.0 and 4.0.

The price of RENUMBER is $20.00. Students can purchase the
program at the special price of $15.00. To order a copy or
get more information, write to:

 Jonathan Mead
 356 N. Spaulding Ave.
 Los Angeles, CA 90036

Email inquiries can be sent to izzyt09uclamvs (BITNET).

Name:____________________________________________________
Address:_________________________________________________
City:______________________ Province or State:_________
Country:___________________ Postal Code:_______________
Email Address:___________________________________________

Version: PC _____ Diskette Size: 5.25 ____ 3.5 ___
 Mac_____

[End Linguist List, Vol. 2, No. 120]