CHAT/CLAN help: add to %mor, missing items, %err line, post codes

[Katerina Tetzloff]

I am currently undertaking a project that requires automated language analysis and have chosen to use CLAN. I have consulted the CHAT and CLAN user manuals but am left with a few questions, hopefully some of you can answer. 

1)      -  Is there a way to add more detail to the %mor tier and run analyses on these things that are not mentioned in the CHAT manual? I am wanting to code argument structure of verbs. I tried adding an extra classification on my %mor tier like v|give-3S-PAST-ditransitive, but it gives me an error that it does not recognize that code. I am not sure how to edit the %mor file to adjust this, if it is even doable.

2)     - If I code missing items, like 0det for missing determiner, how can I run commands just involving these absent items?

3)     - Can I run commands on the error tier? I have coded semantic errors and would like to run an analysis on these if possible.

4)    - I also would like to use post codes (e.g., [+gram] is mentioned in the CHAT manual) at the end of lines, but the manual doesn’t say on which tier this code needs to be. I have been putting it on the annotation tier, which does not cause any errors, but I am unable to run commands based on the post-code. 

Thank you!

[Brian MacWhinney]

Dear Katerina,

I couldn’t replicate any problem directly.  Probably I don’t understand what you are doing.  Which program is giving you the error message?  To understand this, I would need to see your command line and a one-line file in which you enter the code and get an error message. 

 However, the right way to do this would be to add verb class codes to all the verbs in the v.cut file in the English MOR grammar.  That is a big job, but it would be the right way to tackle the issue. 

  Regarding searches for 0det and such, you would use +s”0*” in FREQ, KWAL, etc.

  Yes, you can run commands on the %err tier.   You use +t%err.  However, we encourage people to do error coding on the main line.  The manual has a section on this.

    Postcodes go on the main line. 

 

--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/67443d6b-755a-4007-ad95-5bcfd47a88f4%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Katerina Tetzloff]

Thank you for your response!

-Regarding the first question, I must have had some sort of typo, because I am not getting any errors at the moment. I will try editing the v.cut, as you said this would be the best way to do that. I don't have a huge corpus, so it should be doable for the tokens I have.

-I did, however, get an error when trying to code errors on the main tier. For example:
PAR*:     The fish [* s:r] caught the man [* s:r].

This gave me error messages until I replaced the error indication on the main tier to just [*] and specified the [* s:r] on the error tier. Should this work just on the main tier?

-I will put my postcodes back onto the main line! Thank you! However, how do I run commands searching for lines with a certain postcode? I was thinking

Thank you!!

-Katerina

[Brian MacWhinney]

Dear Katerina,

Yes typos can cause problems.  If you have a small corpus, hand entry of these codes will be much easier.  At some point, we intend to enter in the v.cut lexicon the various codes derived from VerbNet. 

  Regarding your problem with error codes on the main line, I would again suspect a typo.  For example, the line you sent me was

PAR*:     The fish [* s:r] caught the man [* s:r].

But this has to be

*PAR:     The fish [* s:r] caught the man [* s:r].

 

In regards to searching for postcodes, there is a distinction between finding the postcode and finding the material marked with the postcode.  Please just read section 10.7 of the manual.

To view this discussion on the web visit https://groups.google.com/d/msgid/chibolts/f2f09500-8a6a-4667-a43c-712a129d5680%40googlegroups.com.

[Katie Tetzloff]

Oops, yes. Thank you.

I will consult the manual regarding the postcode commands; thank you for the reference! However, if I use the post code *PAR:     words words words. [+gram] it gives me an error saying, "Symbol is not declared in the depfile.(11)" for every main tier with a post code, and [+gram] is highlighted in red. This is why I was wondering which tier they are meant to be on, because I am getting this error. 

So sorry for all of the questions, and thank you so much for your help! I really appreciate it!
Katie

--
You received this message because you are subscribed to a topic in the Google Groups "chibolts" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/chibolts/EIkW8xPmVwg/unsubscribe.
To unsubscribe from this group and all its topics, 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/44F5FF67-6135-4FBC-B46F-EC95203F8C7C%40cmu.edu.

[Brian MacWhinney]

Dear Katie,

You need a space after the plus in [+ gram]

 

--Brian

To view this discussion on the web visit https://groups.google.com/d/msgid/chibolts/CAGBqum8c8s8ANs7%2BOYqSH8ZZta7pRxBGNmeUCBx%3DHF3PVjQLjQ%40mail.gmail.com.

[Katie Tetzloff]

Perfect. Thank you so much for your help!!!

To view this discussion on the web visit https://groups.google.com/d/msgid/chibolts/54CDE36E-EEF9-4C10-B2E8-621C36B1C96F%40cmu.edu.