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

Re: [css3-lists] [css3-speech] Purpose of the module

From: Andrew Thompson <lordpixel@mac.com>
Date: Thu, 02 Jun 2011 15:52:15 -0700
Message-id: <103A998C-F0B1-4D16-9931-A21C7287DE4A@mac.com>
Cc: Daniel Weck <daniel.weck@gmail.com>, www-style list <www-style@w3.org>
To: timeless <timeless@gmail.com>




On Jun 2, 2011, at 1:30 PM, timeless <timeless@gmail.com> wrote:

> 2011/6/2 Andrew Thompson <lordpixel@mac.com>:
>> I assume the intent here is that the letters should be pronounced using the names most familiar to the reader of the document, e.g. To avoid speaking English letter names in an otherwise French document.
> 
> "names most familiar to the reader of the document" would be "user's
> language" as with numbers...
> 
> I can read a Spanish document or with difficulty a Italian, French or
> Portuguese document, and if there's a speech agent involved the
> reading part isn't my problem anyway. I certainly have been to Italy,
> France and Portugal and have heard Italian, French and Portuguese
> spoken -- that doesn't make their alphabets the "most familiar" to me.
> 
>> But the numbers will be read differently,
>> 1 being one in an English document,
>> eins in a German document
>> and ichi or hitotsu in Japanese.
> 
> No, they won't. The spec says:
> 
>>> decimal
>>> This list item style corresponds to numbers beginning from 1.
>>> These numbers are spoken as-is by the speech synthesizer, in the user's language.

Indeed you are correct and I have reading comprehension issues :) But in that case the spec makes no sense to me. If a document is written in Spanish but my user agent preferences happen to say I am primarily an English speaker would I really want 1 pronounced 'one' rather than 'uno' in the middle of an otherwise exclusively Spanish document?

In most synthesizers switching the language implies switching to a different voice, which is jarring (think font substitution but much worse) and potentially very expensive in memory and time. 

Surely the spec should say either 'in the element's language' or 'with the synthesizer's current settings' (meaning withe the current voice and settings in effect for the list)?

I take your point documents are often poorly marked up with lang attributes, but if a speech synthesizer has failed to determine the correct language for an element then generally speaking the entire output will be unintelligible anyway so counters would be the least of your problems. 

> 
>> The trouble is that numbers map into every language practically considered and Latin letters do not.
>> 
>> Given the visual presentation of the letters will be a,b,c,d,e… does it make sense to try to map them for other scripts?
> 
> I would say it doesn't make sense to try to pronounce them randomly
> based on document language (whose tagging is likely to be poor
> anyway).
> 
>> One could say 'They are spoken as-is by the synthesizer, using the document script's names for  the letters, when said script is derived from the latin alphabet, otherwise the English letter names are used.'
> 
> I'd rather "user's language or if not covered, user's language's
> interpretation of English". I'm not being centrist, the letter set is
> "a b c d e f g h i j k l m n o p q r s t u v w x y z", and doesn't
> have ñ or ç or any other interesting characters.
> 
> *If* the list set is not that standard 26 letter sequence, it should
> still be read in the user's language where possible, or some
> translation to that language in that voice.
> 
>> But
>> 
>> Do we always know the script?
> 
> no
> 
>> Where is there a definitive list of which scripts derive from the Latin alphabet?
> 
> I tried finding that when I wrote my earlier letter. I gave up.
> 
>> Are there any derived alphabets which omit any of the 26 Latin letters in modern usage?
> 
> There are some tables showing evolution of scripts. I'll need to
> bookmark them (i.e. write an email somewhere so I can find it later
> using gmail) so that I can find them in the future...
> 
> I offered Greek because while not derived, it's related enough to show
> the problem.
> 
>> Is English really an acceptable fallback?
> 
>> Maybe one of our Japanese speakers can comment on whether it would be better to try to map, something like
>> 
>> a -> ア
>> b -> ビ
>> c -> シ
>> d -> ディ
>> ...
>> t -> ティ
>> 
>> What would a Japanese speaker do faced with a list with Latin counters?
> 
> I look forward to hearing an answer to this question.
> 
> I can't speak for Japanese, but for Hebrew (as a non native speaker),
> I wouldn't want:
> a -> א
> b -> ב
> c -> ג - gimmel is closer to gamma from Greek
> d -> ד
> ...
> v -> ת
> z (the Hebrew alphabet only runs to 22 characters)
> 
> (See other threads for how Hebrew counting works)
> 

fantasia and Charles Belov's other thread on this topic points out this is going to be a general problem, eg for Greek lists the English pronunciation would be alpha, beta, gamma (though US  and British speakers differ on beta) whereas the French would be "/ɑ/, /be/, /se/, ". 

Fundamentally it seems this is best left to the expertise of the synthesizer vendors. CSS should simply specify that the generated characters a, b, c or 1, 2, 3 should be passed unadorned (*) to the synthesizer to do with what it will? 

(*) when I say 'unadorned' I don't mean to preclude the very reasonable idea of prefixing the counter qith audio icons or words like 'item' as discussed in the other thread. But that's a separate issue from how '1' or 'a' should be pronounced. 
Received on Thursday, 2 June 2011 22:52:57 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:41 GMT