More on the SignTel Interpreter

[Angus B. Grieve-Smith]

        SignTel (www.signtelinc.com) has put more information up on their
web site -- see the FAQ page in particular.  It turns out there's no
translation whatsoever involved -- it's just concatenative synthesis
combined with a speech recognizer.  But they have an option that strips
out all the grammatical content, 'cause you know, ASL users enjoy that.
And they imply that subsequent versions will produce ASL, as though that's
a minor add-on, like mail merge or something.

        I have to say I'm genuinely disappointed.  I actually believed
that they were really doing some kind of translation.  I should have known
something was up when I emailed them and the guy said that they hadn't
published anything in peer-reviewed journals because they wanted to patent
the technology first.

        What bugs me is that they found some people at the Texas School
for the Deaf to go "wow!" and put it on their testimonials page.
Apparently it's being discussed on a list in Colorado.  I wonder how many
companies, schools and government agencies are going to cough up $4000 for
it.

        Another thing I noticed is that they got an endorsement from the
"International Center for Disability Resources on the Internet."  This
center apparently collects "best practices in areas related to disability
and accessibility issues."  Maybe someone should tell them that any "best
practice" relating to sign linguistics has to use a true sign language...

--
                                -Angus B. Grieve-Smith
                                Linguistics Department
                                University of New Mexico
                                grvsmth at unm.edu