coding addressees

[A Cristia]

Dear Chibolters,

We would like to classify utterances as addressed to the key child, to other children, or to adults, to later analyze separately addressed, overheard, etc., within CLAN and outside (e.g. applying word segmentation algorithms). There is already a standardized format for this mentioned in the manuals, but I wonder whether there are less costly alternatives (we are coding 2 continuous hours of each child).

The chat manual (p. 79) describes:
 

Addressee Tier %add:

This tier describes who talks to whom. Use the three-letter identifier given in the par- ticipants header to identify the addressees.

*MOT: be quiet.

%add: ALI, BEA

In this example, Mother is telling Alice and Beatrice to “be quiet.”

And the CLAN manual (p. 128):

9.21.1Including and Excluding in MLU and MLT

Researchers often wish to conduct MLU analyses on particular subsets of their data.

This can be done using commands such as:

kwal +t*CHI +t%add +s"mot" sample.cha +d | mlu

This command looks at only those utterances spoken by the child to the mother as ad- dressee. KWAL outputs these utterances through a pipe to the MLU program. The pipe symbol | is used to indicate this transfer of data from one program to the next. If you want to send the output of the MLU analysis to a file, you can do this with the redirect symbol, as in this version of the command:

kwal +t*CHI +t%add +s"mot" sample.cha +d | mlu > file.mlu

Although this is useful, this requires having an %add line for every transcription line. When there are only two partners talking to each other, it would be easier to define them throughout (only marking deviations) rather than repeating the information. Does anyone have any tips as to how this would be accomplished? For instance, could this be done combining a long scope event (gem-style) with local switches, as is done for example with languages (eng defined by default, switches marked with [- spa] or @s:spa)?

Thank you in advance,

Alex Cristia

[Brian MacWhinney]

Dear Celia (and Alex),

     Many corpora use the current system.  However, I agree that it is cumbersome.  If there is a big block of adult speech that you would like to exclude, then using GEM markers would certainly be a lot easier.  It would be 

@Bg: adult_talk

…..

@Eg: adult_talk

And then you use the GEM program to exclude these.  Alternatively, you could focus on using inclusion.

Also, you can handle inclusion and exclusion through postcodes on individual utterances.

—Brian MacWhinney

--
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 post to this group, send email to chib...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/chibolts/5c6c2fb2-964e-492f-98b0-1a7a8122ab9c%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[celiaro...@gmail.com]

Dear Brian, 

Thank you very much for your answer. We will try with GEM.

Best,

Celia