Another look at namespaces

James Tauber jtauber at
Mon Sep 20 05:59:08 BST 1999

> >> We seem to have fallen into linguistics.
> >
> >And the problem with that is what exactly? :-)
> In computing discussions, it usually signals that an enormous
> communications breakdown either has occurred or is occurring.
> it could mean that we're on the edge of a massive breakthrough, but I
> think that's true in this case.

I think you meant something different by "linguistics" than what I was
thinking. I realise now you probably meant "discussions about language" and
not the actual field of study.

> I think you're both (James and Paul) overstating the distinction between
> natural languages and formal languages

All I was trying to do was point out that there is a definition of
"language", which is the one normally used by computer scientists, more
accurately called a formal language, which requires a grammar. This is the
sense of the word I think Paul meant. I think it is entirely acceptable for
him to use the term in that sense. Most introductory CS books and I would
guess almost all books on compilers and parsing would use the word
"language" in that way.

As an aside, I'm a linguist by training (with an emphasis on formal
syntactic theories) and find it refeshing to be accused of *overstating* the
distinction between natural and formal languages! :-)

> I think I have to suggest that there is a difference between HTML 4.0 (the
> formally specified, designed-by-experts version) and HTML (the one that's
> used in real life) - and do what we can to improve HTML, the real-life
> version.  If XHTML wants to design changes, they'd do well do consider
> the public will actually adopt, not try to force an odd notion of formal
> grammars mapping to namespaces upon that public.

"Odd notion of formal grammars"? What is odd about the EBNF in XML 1.0? What
is odd about a DTD? What is odd about a DDML document? What is odd about an
XML Schema Description?

These are all called "grammars". They all define things called "languages".
They are not "grammars" and "languages" the way Noah Webster used those
words or the way you seem to use them. They are "grammars" and "languages"
the way computer scientists use those words.

Paul wasn't overstating any distinction between formal and natural
languages, he was merely using the words "grammar" and "language" in their
formal sense. Rick and you were using those terms in perhaps a different
sense. No one is more right that the other, you were just using different
senses of the words.

> I think you're making too strong a claim about XML vocabularies
> as formal languages.

Note: I never said XML *vocabularies* were formal languages. I said

(1) XML itself is a formal language (do you disagree?)
(2) DTDs define formal languages. They are grammars (do you disagree?)

I don't equate the term "XML vocabulary" with "DTD".

> >* XML is a language defined by a grammar.
> XML provides a foundation - much like letters give English and alphabet to
> create words with.  (Tim Bray: "XML is ASCII, etc.")

I think you missed my point. All I was saying was that XML is an argument
against your contention that languages aren't defined by grammars. XML *is*
a language defined by a grammar.

> >* The XML REC clearly says that a DTD is a grammar for a class of
> >In a formal language view, this class of documents is a set of
> >utterances---a formal language.
> Except that in many cases there are multiple DTDs (for testing different
> claims), or, in fact, no DTD at all.  Without a DTD, are we truly
> 'grammar-less'?  I don't think so.

Again, I think you missed my point. DTDs are grammars (do you disagree?)
They define languages (do you disagree?) That was my point. DTDs are another
example of grammars being used to define languages.

> >> Whether or not _The SGML Cookbook_ makes this claim that a language is
> >> constrained to a single grammar, I'll be making that claim in my next
> >book,
> >
> >Depending on what you mean by "not constrained to" and what type of
> >languages you mean, I either completely agree or completely disagree with
> >you :-)
> I guess you'll have to buy the book and find out! (Gross sales pitch, eh?)

You know I plan to buy the book anyway, Simon! :-)

If by language you mean a formal language in the sense meant by computer
scientists and the L in XML, and by "not constrained to" you mean that it is
not possible, given a language, to produce a single grammar that generates
that language then I'm afraid you are wrong.

I don't actually think you mean "language" in this sense otherwise this
thread wouldn't have started.

But Paul did, and there is nothing wrong with that sense.

James Tauber / jtauber at /
Maintainer of :, and
<pipe>Ceci n'est pas une pipe</pipe>

xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at
Archived as: and on CD-ROM/ISBN 981-02-3594-1
To (un)subscribe, mailto:majordomo at the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo at the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at

More information about the Xml-dev mailing list