Q: Pedagogical use of a students' writings corpus

Gerard Gautier (ggautier@speedy.grolier.fr)
Thu, 15 Aug 96 09:26:23 CST

Hello,

I follow the discussion on CORPORA since a good time, but it is quite the
first occasion for me to "talk". I have teached French for 3 years in
Taiwan (I am due to come back to France next September), and I am specially
interested in teaching _writing_. This has slowly led me to a corpus
approach :

1- I have first collected a certain number of erroneous sentences from
the compositions of my 4th year students.

2- Then, as many errors kept repeating themselves over years and
different classes, I used a little software to generate corrections for
them (basically, the teacher reads the composition, and when he spots a
mistake of a certain type, he selects this type from a prepared file, and a
matching comment is printed).

3- For pedagogical purposes (incitating auto-correction by students) this
was perfect. But now I want to use the data I have to spot the most commmon
mistakes and help designing instruction materials. Idea is : knowing what
this type of public (students of chinese mothertongue) _do_ write in 4th
year may help designing what to _teach_ them in 2nd and 3rd year...

Seeing that without keeping the _context_ of the types of mistakes, I was
not sure if a specific occurence was to be encoded under THIS or THAT
category (problem of repeatability), I started photocopying the
compositions to keep them altogether...

--
So now I have about 345 textual units (from one paragraph to three pages
length), and a base of 249 "most frequent" errors (the ones which happened
in 95-96 in whatever I corrected), and some correspondances (in my mind)
between French and Chinese way of saying things, which may help to explain
what happens.

Now, during the International Congress of French Teachers in Tokyo end of August, I plan to propose to the interested teachers to continue the work by keeping as well their compositions and analysing them in common. Obviously, my interest is primarily in the students of chinese mother tongue... The texts as well as any improved verwsion of the errors database could be put on the net on a ftp / www site.

I was wondering if anyone on this list would care to give any methodological remarks about the whole idea, and if there were people "out there" who were or had been involved in any similar activity and with whom I could exchange. Then, if this project ever happens to start really (ie, if other teachers in the Congress appear to have enough interest in it), would anyone be interested to host the thing on a net site ?

Thanks in advance and sorry for this long mail.

Gerard GAUTIER French Teacher in Taiwan.

------------------------//----------------------------------------- Gerard \| \| // | \| . ggautier@ Gautier " \ | . . club-internet.fr _/\__) O__\ \ | \___) \ ----------(__)''----__/ --------- _/ _/ .. _/ ------------------