LINGUIST List 4.342

Thu 06 May 1993

Qs: PS Trees, Sicilian, Field recording, Brown Corpus

Editor for this issue: <>

Directory

1. "Karen A. Mullen", Phrase-Structure Trees
2. , grammar of Sicilian
3. , field recording equipment
4. "Amy Uhrbach", q: Brown Corpus

 Dept of English, University of Louisville, LOU KY 40292
PHONE: 502/588-5901 or -7158
Does anyone know of any software that will generate phrase-structure
trees? I am teaching a course on the Structure of Modern American English,
using Noel-Burton's text *Analysing Sentences* as the students' source of
information on representing the structure of sentences. It follows the
usual conventions in its representation of tree structures. (The students
are native-speakers of the language--most usually English majors, most
usually seeking a teaching certificate.)

In the course, I ask students to analyze sentences according to their
structure. I want to try to develop computer-assisted tutorials for
students who are having difficulties doing the analysis and representing
that analysis via tree-diagrams.

In the first phrase, the tutorial would ask questions (e.g., from Chapter 4):

 What is the subject of the sentence?
 What is the predicate of the sentence?
 What is the verb-group of the sentence?
 Are there any complements to the verb-group?
 If so, how many?
 (If one), what is the complement?
 (If one), what is the function of this complement?
 (If two), what is the first complement?
 (If two), what is the function of the first complement?
 (If two), what is the second complement?
 (If two), what is the function of the second complement?
 What is the sub-category feature of the verb-group?

 If the students answer incorrectly, some response would be given
 to guide students to the rethink their answers. For example, if
 they said that the complement of the verb was a direct object, but
 they indicated that the sub-category feature of the verb-group was
 "intensive", it would say something like "You indicated earlier
 that the complement of "x" was "y" and that "y" functioned as a direct
 object. "Intensive" verbs have subject complements, not direct objects.
 What kind of verb has a direct object?"

In the second phase, the tutorial would ask them to draw a tree-diagram
to represent this information.

 The program would ask them what the label for the root node is and
 put it on the screen. It would then ask them how many branches they
 want to come off the root node. The program would draw these branches
 Then the program would ask them how they want to label these branches,
 one by one, and would label them accordingly. Then the program would
 ask them how many branches were to come off each of these nodes, and
 would draw them. Then, the program would ask them how they wanted to
 label these branches, one by one, and label them accordingly. And so

 Naturally, the program would have information about the sentence
 already, would be able to tell if the student is answering
 correctly, and would be able to provide responses to lead the
 student in the right direction, if they answer incorrectly.

Given this format, then, in the second phrase, the program would have to
have a way of:
 a. reading in instructor-provided sentences, gleaned from *Newsweek*,
 *Time*, *The New Yorker*, etc.
 b. reading in instructor-provided answers to the questions that
 have to do with constructing the tree (or perhaps there are
 are programs that will parse instructor-provided sentences
 automatically!?)
 c. putting a node on the screen, finding it later, drawing branches
 from it, and labelling the branches.

The tutorial should be usable on IBM-compatible machines of something less
than a 386 or 486 vintage, and should be PC-based.

Given this lengthy introduction, my questions are two:
 1. Does anyone know if tutorials like this already exist?
 2. If not, does anyone have any idea of the appropriate software to use
 to generate such tutorials? (I COULD start from scratch and write
 them in BASIC, but I think there must be a more efficient
 way to do this!!! I have programming experience and know computers
 rather well, so I can learn almost anything, if it will do what I want
 it to do.)

I would appreciate any advice to design (or re-design) this project and
implement it with the technology currently available.

I am doing some research on the Sicilian language with native-speaking
respondents. I would also like to look at some grammars of Sicilian.
Can anyone recommend grammar sources for this language? Thanks in advance.

I would like to hear about people's experiences with recording and
transcribing equipment in the field. I'm particularly interested in
hearing about experiences with providing electricity via solar panels
in areas with unreliable or no electricity. what is involved?

also, what models of tape recorders, conference mikes and
transcribers have people used and had good experiences with (ie
minimal or no breakdowns even under heavy use and good quality
sound). what experiences have people had with camcorders? do they
provide good enough audio to not need an additional audio recording
or not?

the types of projects I have in mind are ones which would involve
taping and transcribing naturally occuring interactions.

Please respond directly to me. If there is interest, I will summarize
for the net.

Is the Brown Corpus available anywhere on the net, ftp or otherwise?