check errors and adding own tags

Thu Jun 23 16:58:52 UTC 2011

Hello ChiBolts

We have run into some interesting issues in my lab today, so this discussion
is particularly timely.  We have transcripts from bilingual children (SPA
and ENG) that we need to run multiple analyses on, including MLT and MOR and
POST.  Of course, even though transcripts are supposed to be ENG or SPA,
there is a lot of code switching going on throughout.  Our issues are
outlined below.

First, compound words (french_fry) do not check.  They end up in the output
file.  Does this mean that they would not be included in subsequent MOR
analyses?  If we want them to be part of the MOR analyses, how do we get it
to check?  We checked the depfile, and *_* was there under the %syn line.
We are hesitant to change the depfile unless we need to; can someone tell us
how to get compound words into the MOR analyses?

Second, non-words such as ‘mmhm’ also do not pass check.  We attempted to
run MOR even with these in the main tier, and the %MOR tiers were not
created. Any idea why not?

Third, our collaborator needs to run MLT and TTR in both English, Spanish,
and ‘mixed’ (utterances that include code switches) utterances.  She has
been using pre-codes to identify the language used in the utterance, and
post-codes to identify utterances that she wants omitted from the MLT
analyses.  However, we cannot get transcripts with pre-codes to run in MOR.
Is there a way to calculate MLT that does not involve the use of precodes?

Fourth, more generally, does CLAN treat pre-codes and post-codes similarly?
Or are they unique in how they influence subsequent analyses?

Finally, we have been able, with Leonid and Brian’s wonderful assistance, to
get MOR to run in both SPA and ENG when precodes are deleted from the
transcripts.  I have a set of directions for this, and can share if folks
want them.

 Shelley Brundage,

George Washington University

On Thu, Jun 23, 2011 at 11:46 AM, Leonid Spektor <spektor at andrew.cmu.edu>wrote:

> Rasmus,
>
>
> 1. The form "bird+house-s" is incorrect. If the manual says otherwise, then
> it needs to be changed. The correct form is "bird+houses".
>
> 2. We don't support this feature at this time. We are in a process of
> implementing a generic "@x:" tag which will allow users to specify anything
> they want to. Your example below, then will look like "Toronto at x:geo".
> This should be implemented in a day or so.
>
> 3. We discourage people from editing the "depfile.cut". Adding anything to
> that file without additional changes to other supporting files and/or
> programs will cause problems with other programs like mor and post. We also
> no longer support the "depadd.cut" file. Check just simply ignores it.
>
> We no longer distribute the "depadd.cut" file. You must have a very old
> installation of CLAN and new updates do not delete any old files present
> there.
>
> Leonid.
>
>
>
>
> On Jun 23, 2011, at 09:20, RSteinkrauss wrote:
>
> > Hi,
> >
> > I have some questions regarding checking files in CLAN and would be
> > very happy if someone could help me. I am using the Win version from
> > 30-Apr-2011.
> >
> > 1. If understood the manual correctly, English compounds should be
> > transcribed with a + sign. Also, plurals in compounds should be marked
> > with a dash-s. However, when i run CHECK on my transcript, a form such
> > as bird+house-s gives me an "undeclared suffix" error. Same with bird
> > +house-s at s if it is a codeswitch in an otherwise German text.
> >
> > 2. We would like to use some @-tags in our transcripts that are not
> > specified in the manual (such as @geo for place names, e.g.
> > Toronto at geo). I understand that this is not longer possible using a
> > depadd-file, so I wondered if the way to go is to change the
> > depfile.cut-file.
> >
> > 3. Assuming that this was indeed what should be done, I changed the
> > depfile.cut so that it would include the tags we need. After that,
> > CHECK worked fine (with the exception of the errors described in
> > question 1). However, if I now change the depadd.cut file again, CLAN
> > will not react to the changes. E.g., if I remove a previously allowed
> > @-tag such as @geo from my depadd.cut, CHECK will not complain when it
> > encounters that tag in a transcipt. Strangely, that behaviour persists
> > even after de- and reinstalling CLAN. I have set the lib path
> > correctly, to a folder containing my changed depadd.cut. Also after
> > changing the lib path back to the original depadd.cut that came /
> > installed with CLAN, CHECK still allows my own @-tags.
> >
> > Does anyone have an idea how to go about this? Any feedback would be
> > greatly appreciated!
> >
> > Thank you in advance,
> > Rasmus
> >
> > --
> > You received this message because you are subscribed to the Google Groups
> "chibolts" group.
> > To post to this group, send email to chibolts at googlegroups.com.
> > To unsubscribe from this group, send email to
> chibolts+unsubscribe at googlegroups.com.
> > For more options, visit this group at
> http://groups.google.com/group/chibolts?hl=en.
> >
> >
>
> --
> You received this message because you are subscribed to the Google Groups
> "chibolts" group.
> To post to this group, send email to chibolts at googlegroups.com.
> To unsubscribe from this group, send email to
> chibolts+unsubscribe at googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/chibolts?hl=en.
>
>

-- 
Shelley B. Brundage, Ph.D., CCC-S
Associate Professor and Graduate Program Director
Board Recognized Specialist and Mentor-Fluency Disorders
Speech and Hearing Science department
George Washington University
2115 G St NW Suite 201
Washington, D.C. 20052
(202) 994-5008 office
(202 994-2205 lab
(202) 994-2589 fax

-- 
You received this message because you are subscribed to the Google Groups "chibolts" group.
To post to this group, send email to chibolts at googlegroups.com.
To unsubscribe from this group, send email to chibolts+unsubscribe at googlegroups.com.
For more options, visit this group at http://groups.google.com/group/chibolts?hl=en.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/chibolts/attachments/20110623/d83bb3d0/attachment.htm>