LINGUIST List 7.121

Fri Jan 26 1996

Qs: Serbo-Croatian, 'Swarm', Judgments, Transcribe

Editor for this issue: T. Daniel Seely <dseelyemunix.emich.edu>

Directory

1. Nick Ourusoff, Serbo-Croatian corpora and dictionary sought
2. Svetlana Vasina, 'Swarm' alternation
3. Michael Wescoat, English judgments solicited
4. Marrion Ward, Transcription

The Computing Research Laboratory (CRL) at New Mexico State University is 
trying to obtain in machine-readable form both a large (several million 
words), non-fiction Serbo-croation corpus and a Serbo_croation/English 
dictionary, preferably with the part of speech designated for each entry 
to support projects in machine translation of languages.

 CRl is a non-profit, research organization with a focus on machine 
language translation. Further information about CRL can be found on our 
web page:

 http://crl.nmsu.edu/index.html

 We will greatly apprecite any information.

	Sincerely,

	Nicholas Ourusoff (CRL)

 I am collecting data on the 'swarm' alternation in different
 languages. I am trying to test a couple of hypothesis: (i) the
 alternation occurs only in a set of closely related (syntactically,
 morphologicaly, historically) languages, (ii) the alternation is
 semantically motivated. So, I am asking for help from people
 familiar with different languages to supply examples. The kind of
 examples that I am interested in are exemplified in English by the
 following:

 (1) a. Bees are swarming in the garden.
 b. The garden is swarming with bees.
 (2) a. Stars are glittering in the sky.
 b. The sky is glittering with stars.
 (3) a. The sound echoed in the hall.
 b. The hall echoed with sound.
 (4) a. Garlic reeked on his breath.
 b. His breath reeked with garlic.
 (5) a. Fish aboud in the pond.
 b. The pond abounds with fish.

 This kind of alternation has been discussed extensively by Salkhof
 (1983), "Bees are Swarming in the Garden", Language Vol59.2, Dowty
 (1995) "The _Swarm_ and Middle Alternations, Predicate Semantic
 Transfer, and Thematic Role Alignment", talk given at Paris Syntax
 and Semantics conference, among others. Dowty (1995) has classified
 the predicates of this class into 5 semantic groups, corresponding
 to the five examples above.

 They are:

 I. Simple movements: crawl, drip, bubble, dance, drible, erupt, foam,
 froth, jump, shake, swarm, pulsate, etc.
 II. Simple sounds: hum, buzz, twitter, chirp, creak, fizz, rustle,
 echo, resonate, etc.
 III. Light emission: beam, blaze, brighten, glow, flicker, flash,
 glisten, glitter, light up, etc.
 IV. Smells and Tastes: reek, smell, taste, be fragrant, etc.
 V. Degree of abundance: abound, be rich, rife, etc.

 This type of alternation is found in English, French and possibly
 other Romance languages, Russian, Serbo-Croatian, possibly other
 Slavic languages.

 I would very much appreciate well-glossed examples from other
 languages than the ones mentioned. I need pairs of examples, just
 like (1-5), showing that what is a location adverb in a) examples can
 be a subject in b) examples; and what is a subject in a) examples is
 expressed in PP (with) phrase (instrumental case) in b) examples. The
 verb has to be intransiive without a transitive counterpart like in
 the case of 'flood' in English. The form of the verb has to stay the
 same with respect to valence changing morphemes such as passive,
 applicative suffixes, middle suffixes (like the reflexive morpheme
 _se_ in Slavic) etc. Agreement morphemes are different of course for
 languages that have them.

 As usual, I will post a summary if there is enough interest. Please,
 send responses directly to me at the following address:
 vasinaling.ohio-state.edu


 Thanks,
 Svetlana Vasina
 The Ohio State University, Linguistics Department

I am interested in seeing what a cross-section of English speakers
feel about the acceptability of the following sentences. If you
have the time and inclination, would you kindly rate them on a
four-point scale? If the sentence seems perfect, please give it
a 4. If it is unacceptable, it should get a 1. Intermediate
levels of acceptability can be indicated with a 2 or 3. Please,
no decimal points!

1. We weren$BUt(J aware of the fact that anyone had left.
 YOUR RESPONSE [4 (=good), 3, 2, or 1 (=bad)]:

2. Students in any danger of failing will be warned in advance.
 YOUR RESPONSE [4, 3, 2, or 1]:

3. Every student with any hope of passing has already finished
 that assignment.
 YOUR RESPONSE [4, 3, 2, or 1]:

4. All the candidates with any chance of being elected will be
 participating in the debate.
 YOUR RESPONSE [4, 3, 2, or 1]:

5. Any person with a penny to his or her name should be concerned
 about the new tax provisions.
 YOUR RESPONSE [4, 3, 2, or 1]:

I will happily report the results of this little survey after
getting the responses, and I will reveal the issue that motivates
my seeking these data at that time. Until then I would prefer
not to prejudice your judgments.

Many thanks,
Michael T. Wescoat
wescoatlisa.lang.osaka-u.ac.jp

I would appreciate some advice on dealing with data for my dissertation.
I have taped conversations between 16 dyads of paired expert-novice
children playing a computer game to my criterion. I want to show how they
co-construct meaning, as exemplified by their discourse, to accomplish the
goal of reaching my criterion. I have between 4 and 5 hours of tape for
each dyad. My question is this: I am buried under transcriptions of
varying degrees of reliability! Is there a way to deal with the mountains
of data without having to transcribe verbatum each tape? Is there a
precedent set for transcribing a certain percentage of tapes, do coding,
and then only transcribe discourse that matches my coding for the
remainder of the tapes? I would greatly appreciate any guidance you can
offer. Marrion Ward

* Marrion W. Ward Work: (704) 262-2995 *
* Public School Partnership *
* Edwin Duncan Hall, Room 303G Home: (704) 963-7709 *
* Appalachian State University *
* Boone, NC 28608 E-Mail: WARDMW *