Re: Can we have @lang back in XHTML Family?

Hi Richard,

This reply from you rings a bell that you had said that you would suggest  
this to the HTML5 group. I think you speak a lot of sense, and that it  
would be good if we could move to a future where browsers recognised both  
lang and xml:lang.

Still, in the short term, it looks like we do need lang to be available  
for dual-purposing existing documents.


On Fri, 23 Jan 2009 21:33:54 +0100, Richard Ishida <> wrote:

> I'm copying in the i18n WG to this thread.
> I18n folks:  See the following thread, where Dom proposes the  
> introduction
> of the lang attribute to XHTML in addition to xml:lang, so that when  
> people
> serve XHTML as text/html the language information is available.
> I was sure I had written something along these lines and sent to the  
> html5
> WG, but I don't seem to be able to find it.  We certainly had some
> discussion of it in the i18n WG a while back.
> I hear many complaints from authors using XHTML about having to declare
> language twice, once with lang and once with xml:lang for XHTML 1.0, and  
> I
> must say that I also find it burdensome myself.  I encourage people to
> persevere because agents that process text/html don't recognise xml:lang,
> but agents that process the file as XML (eg. using XSLT) only recognise
> xml:lang.
> I would very much like to reach a situation where an author could just  
> use
> one or other of these attributes, and achieve the desired result.
> I was originally thinking, however, that we should ask the HTML5 people  
> to
> write their spec such that future processors of text/html would recognise
> both lang and xml:lang as equivalent.  That way it wouldn't be necessary  
> for
> the XHTML specs to change, and authors of XHTML would be able to use just
> xml:lang, rather than both attributes.
> The idea that it might be possible to introduce lang to XHTML 1.1 etc was
> interesting, but I think that the problem would be that, if people don't
> continue to use both attributes, xml processors would have to also be
> changed to recognise that lang is equivalent to xml:lang (eg. so that the
> XPath lang() function would still work in XSLT or XQuery).  In fact, I  
> think
> that that would ultimately mean changing the XML spec, and the
> Canonicalisation spec, XML Schema, etc.  I think that many people would  
> only
> use lang when writing XHTML, and then we'd have the opposite problem from
> the one we're trying to fix, ie. that XHTML doesn't work as it should as
> XML.
> I can't say what level of acceptance the idea would have with the HTML5
> folks, but it seems to me that moving text/html processors to accept
> xml:lang as equivalent to lang would be more effective, and perhaps  
> easier.
> Ok, so what am I missing?
> RI
> ============
> Richard Ishida
> Internationalization Lead
> W3C (World Wide Web Consortium)
>> -----Original Message-----
>> From: Dominique Hazael-Massieux []
>> Sent: 21 January 2009 08:24
>> To:
>> Cc:;
>> Subject: Can we have @lang back in XHTML Family?
>> Hi,
>> The to-be-published version of the XHTML Media Types note allows for any
>> XHTML Family document to be served as text/html:
>> But as was discussed in this very list [1], this is problematic since
>> the lang attribute (the only one interpreted as a language annotation on
>> documents served as text/html) is not allowed by the XHTML DTDs (but the
>> XHTML 1.0 one).
>> Could the lang attribute be added to the relevant DTDs so as to enable
>> properly lang-marked up XHTML documents to be served as text/html?
>> FWIW, I'm fairly confident I could get formal support from the Mobile
>> Web Best Practices Working Group on this proposal if this is of any
>> help, since this impacts negatively on the deployment of their mobileOK
>> specification.
>> Thanks,
>> Dom
>> 1.

Received on Wednesday, 28 January 2009 14:42:06 UTC