Thematic unit and code-switch codes

[Ran Li]

Hello,

I'm trying to analyze some discourse data for bilingual patients with aphasia, and here're two specific code-related questions:

1) When the frequency of thematic units is calculated ("@e:tu"), CLAN doesn't pick up those following a phonological paraphasia, e.g., irl@e:tu for the target girl. Is there a way to make CLAN to recognize these codes?

2) Specifically for bilingual patients, I've been using "@s" to code cross-linguistic errors. But in cases where the response has a phonological error (e.g., target language is English, patient's response is in Chinese: bao@s for [pao] - means running in Chinese): (a) how should I code the thematic unit? and (b) can I use replacement [: ] so that MOR can process it as a real word? If so, should I replace it with the word in the target or the non-target language?

Thank you,

Ran

[Brian Macwhinney]

Dear Ran,
See below.

> On Sep 13, 2021, at 11:28 AM, Ran Li <rl...@bu.edu> wrote:
>
> Hello,
>
> I'm trying to analyze some discourse data for bilingual patients with aphasia, and here're two specific code-related questions:
>
> 1) When the frequency of thematic units is calculated ("@e:tu"), CLAN doesn't pick up those following a phonological paraphasia, e.g., irl@e:tu for the target girl. Is there a way to make CLAN to recognize these codes?

Please send me a sample that has these codes, along with the version of the FREQ command you are using and identifying which number in the output is wrong.

>
> 2) Specifically for bilingual patients, I've been using "@s" to code cross-linguistic errors. But in cases where the response has a phonological error (e.g., target language is English, patient's response is in Chinese: bao@s for [pao] - means running in Chinese): (a) how should I code the thematic unit? and (b) can I use replacement [: ] so that MOR can process it as a real word? If so, should I replace it with the word in the target or the non-target language?
>

If the focus of your work is on thematic units, it would be best not to try to jam all of this information in the special form marker. So, using the Chinese replacement form [: pao@s] would be the best approach.

—Brian MacWhinney
> Thank you,
> Ran
>
> --
> You received this message because you are subscribed to the Google Groups "chibolts" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to chibolts+u...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/chibolts/fe343c70-9b1d-40f3-be83-b539917037fan%40googlegroups.com.

[Ran Li]

Dear Brian,

Thank you very much for your quick reply! Below is a sample that includes the codes. The FREQ command I'm using is: freq +s"*e:*" +t*par +d2  @

For the output, there is no column listed for TUs with phonological errors (i.e., dirl@e:tu for girl, ndog@e:tu for dog).

@Begin
@Languages: eng, zho
@Participants: INV Investigator, PAR Participant
@ID: eng, zho|NounTxPost_cat|INV||female|||Investigator|||
@ID: eng, zho|NounTxPost_cat|PAR||female|||Participant|||
@Transcriber: Ran
@Media: PostN_EN_cat, video
@Date: 09-DEC-2020
*PAR: irl [: girl] [* p] [//] boy [* s] &-um <no> [e] . [+ gra] [+ g3]
*INV: okay that's okay .
*PAR: XXX +/. [+ gcu]
*INV: &-hmm ?
*PAR: cat@e:tu . [+ g4]
*INV: okay .
*PAR: dirl@e:tu [: girl] [* p] . [+ g4]
*INV: yeah .
*PAR: &-mm ao@s [* p] [//] &-mm hat [: cat] [* p] . [+ g4]
*INV: &-mhm .
*PAR: gou@s [//] &-um &-um ndog@e:tu [: dog] [* p] . [+ g2]
@End

Thank you,

Ran

To view this discussion on the web visit https://groups.google.com/d/msgid/chibolts/CF6A60EB-28BC-48DB-8384-FA38E0F1676E%40andrew.cmu.edu.

[Brian Macwhinney]

Dear Ran,
It seems that you are doing a fine job of coding thematic units on the words with @e:tu and then marking the phonological error with [* p]. You can just run the FREQ command you give here for the “e:tu” forms and then another one for [* p] as in
freq +s"[* p]" *.cha

Best,

— Brian MacWhinney

> To view this discussion on the web visit https://groups.google.com/d/msgid/chibolts/CAG8jtt_vf_y9DMYhK2x97OtRNpMarBq0TFZFDQgz3WBMpsLRSg%40mail.gmail.com.



— Brian MacWhinney
Teresa Heinz Professor of Cognitive Psychology,
Computational Linguistics,
and Modern Languages, CMU

[Ran Li]

That's an easy solution, thank you!

Ran

To view this discussion on the web visit https://groups.google.com/d/msgid/chibolts/3A7646FD-C0E0-4EC9-BDA9-3AE8ABE18E9A%40andrew.cmu.edu.