new CLAN program for phonology

[Brian MacWhinney]

Dear Info-CHILDES,
  Leonid Spektor has written a new CLAN program for phonological analysis.
The command is MAKEMOD, which stands for "make a new %mod or model
phonological tier."  This program uses the CMU Pronouncing Dictionary to
automatically insert the adult phonological targets for words.  This %mod
line can then be analyzed in conjunction with a %pho using the MODREP
program.  I can also imagine other uses.
  The CMU Pronouncing Dictionary is huge and CLAN requires a lot of memory to
load it.  So, if you are using CLAN on a Mac, you will want to increase its
memory to about 20MB from the default of 10MB.  On Windows, I guess the
operating system figures this all out for you.
  We translated the CMU Pronouncing Dictionary into SAMPA format (close to
UNIBET) for this purpose.  The %mod output is in SAMPA, but we have added
underscore characters to clearly delimit segments.
  I would be happy if people interested in phonology could try this out and
tell me what they think and whether we should modify any details.

--Brian MacWhinney