Hello to everyone,
I'm using CLAN for a research project.
In the file I'm working on, an erroneous sentence such as "they has" is tagged as follows:
*CHI: they has [: have] [*] +/.
%mor: (1)pro:sub|they (1)v|have +/.
I would like to search for sentences with agreement errors based on the %mor tier, but it appears that the tags in the %mor tier are assigned based on text replacement (i.e., [: have]) rather than actual production (i.e., has).
So I'm wondering if it is possible to ignore text replacements when searching in the %mor tier. According to the manual, the command +r5 eliminates any text replacements, but it does not seem to work for the %mor tier.
Thanks for your help,
Takayuki Kimura