Question about tagging prompted language in a transcript

[Amanda Huensch]

Hi All-

We are attempting to code some language in our transcripts such that it is ignored in a freq count. We have narrative data which includes language that was prompted by written text in the stories (e.g., “every morning was the same”) that we don’t want to be included in some analyses. Sometimes the language is an entire utterance and sometimes it is a partial utterance (examples below).

 

The current way we are coding this language to be ignored is by using angle brackets to select the language to be ignored and adding [@g] at the end of the line. So, for example:

 

13    *400:            <so this is the story of Natalia and her cat Pancho> [@g] .

14    *400:            <every morning was the same> [@g] .

15    *400:            Natalia &euh read to her toys .

…

23    *400:            and they went home together .

24    *400:            <until one day> [@g] when Pancho left the house he was scared away by a dog .

 

Then, using this command

 

freq @ +t*1* +r6  -s"<@g>"

 

we get a word count without that language.

 

While this seems to work for us at the moment, my concern is that when I conduct an esc+l, I get the error message:

 

([@g] highlighted): Symbol is not declared in the dep file

 

My questions then are: (1) Is there a better way to accomplish this goal? or (2) Is there a way to use this method but not get the error message?

 

Thank you for your help,

Amanda

 

PS If it is the case that we can/should continue to code this way, do you foresee complications with phrases that already have angle brackets (the count still appears to work). So, for example:

 

*400:     and she said <<did he say France> [= r]> [@g] ?

[Brian Macwhinney]

Dear Amanda,

     If what you want to do is to exclude a whole utterance, then the usual method is to insert a postcode, such as [+ exc] and then you would use the command with the -s”[+ exc]” switch.  I think this is a good method for whole utterance exclusion.  However, you also want to exclude strings within utterances and for clarity you should use a different method for that.  The only problem there with your [@g] code is the actual syntax of the code.  It should instead be [% g] and then you also need to add the second switch -s"<% g>”.

—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/9df5e374-d731-4329-8752-d45fbc3b1143%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Amanda Huensch]

Dear Brian,

Thank you very much! That works perfectly. 

Amanda

To view this discussion on the web visit https://groups.google.com/d/msgid/chibolts/F54AE7BD-C4DE-4408-841D-F55DBB93209F%40cmu.edu.