W3C home > Mailing lists > Public > www-style@w3.org > February 2011

Re: "phonemes" property in the CSS3 Speech module

From: Ambrose LI <ambrose.li@gmail.com>
Date: Fri, 4 Feb 2011 12:43:49 -0500
Message-ID: <AANLkTinfa4XhstD7k7QVyPLQv6amCCz48xrxJNKsMRNn@mail.gmail.com>
To: Mark Kenny <beingmrkenny@gmail.com>
Cc: www-style@w3.org
2011/2/4 Mark Kenny <beingmrkenny@gmail.com>:
> To my mind a certain amount of this discussion seems beyond the scope of
> markup and styling. The user agent ought to determine things like the
> difference between present tense "read" (rhymes with reed) and past tense
> "read" (rhymes with red) based on context. If Google Translate and Microsoft
> Word can understand human language grammar, there's no reason a browser
> couldn't do it.

First, let us get this straight: Google Translate does NOT understand
human language grammar (in fact Google just acknowledged that it does
not); believing that it does would be a grave mistake. There are
simple sentences that when “translated” give results that carry the
exact opposite meaning (in multiple language pairs), among other known
serious problems.

And I always turn Word’s grammar checks off because it frequently
fails to understand grammar that’s perfectly fine.

I share Koji’s position that this is fundamentally the same problem as
ruby. In the general context, this is exactly what ruby is about: a
way to specify pronunciation, for the purpose of disambiguation and/or
allowing the author to specify unusual pronunciations.
-- 
cheers,
-ambrose
Received on Friday, 4 February 2011 17:44:22 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:37 GMT