repetitions: tokens and %mor analysis

[Cynthia Audisio]

Dear chibolts,

We have transcribed natural interactions with children using [/] to indicate repetitions (e.g., mamá [/] mamá [/] mamá). When we use Freq command to calculate types and tokens, repetitions aren't count as different tokens. Thus, mamá [/] mamá [/] mamá yields "1  Total number of items (tokens)". Is there a way to include these repetitions in the token count?

The same with the %mor tier. %mor disregards repeated words. Instead we'd like to have one morphological analysis for every word (whether or not repeated) in the utterance.

Thanks!

Cynthia

[Cynthia Audisio]

+r6 switch appears to solve the token issue when using freq.

Now, I'd only need to know how to get different morphological analysis for repeated words in the utterance.

Thanks

[Brian MacWhinney]

Dear Cynthia,

   MOR can't do that.  I'm not sure why you would want to include repeated words in a MOR analysis.

--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/16e1089c-e29b-40f8-8caf-1ab1cf4f286b%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Patricia Sorensen]

Cynthia, 

It sounds like the easiest solution may be to just go through your sample and remove the [/]. If you copy the file into Word and use the find and replace function, it wouldn't take long to remove them. 

Patti

--

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/16e1089c-e29b-40f8-8caf-1ab1cf4f286b%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Patricia Sorensen, M.Sc., M.Ed.

Registered Speech-Language Pathologist

PUF ECS/Early Education and Family Wellness Centre

Wetaskiwin Regional Public Schools

Telephone: 780-352-0224 ext. 229/cell: 780-335-4471

[Brian MacWhinney]

Yes, that would work.  Or, more simply, you could make a copy of your corpus folder and then use CHSTRING to remove all the [/], as in 

chstring +re *.cha +s"[/]" "" +1

No need to cut and paste into Word.

--Brian MacWhinney

To view this discussion on the web visit https://groups.google.com/d/msgid/chibolts/CACDP4aeiS%2BL1UMDPAufm6NQpScj4i_RwhHXf0xujsh1oTF%3DW_Q%40mail.gmail.com.