W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > January to March 2005

RE: A note about the definition of "structure"

From: Mike Barta <mikba@microsoft.com>
Date: Mon, 17 Jan 2005 15:11:36 -0800
Message-ID: <7DF35A0B5F67E84B9095C21C8A97641803F15CD3@RED-MSG-33.redmond.corp.microsoft.com>
To: <w3c-wai-gl@w3.org>

I would agree with Michele here; structure is the 'arrangement and interrelation of elements' and not the elements themselves.  Thus the text of an element is not structural it is merely the thing of itself; the text as a thing in part of a whole is structural and this structure may be expressed semantically in markup, e.g. <h1/>.  

I wonder if we might want to go a bit further and note that the expression of the structure in markup is not the structure itself but a representation of the structure of the content in the output.  That is that we are suggesting here that the authors use the technology they choose per spec to 'represent' the structure, e.g. header markup in HTML, rather than the more nebulous 'separate structure from presentation' as the HTML itself is considered by many to be a presentation layer already.


η ελευθερία της ομιλίας είναι ουσιαστική στη δημοκρατία

-----Original Message-----
From: w3c-wai-gl-request@w3.org [mailto:w3c-wai-gl-request@w3.org] On Behalf Of Michele Diodati
Sent: Monday, January 17, 2005 2:58 PM
To: w3c-wai-gl@w3.org
Subject: Re: A note about the definition of "structure"

John Slatin wrote:
> You're correct: I do intend the term "structure" to encompass 
> "content."  But this is not an arbitrary extension of the concept. It is 
> consistent with the definitions listed below, from Merria-Webster's 
> 10th Collegiate Dictionary (...)

I read all the definitions you reported. Sincerely it seems to me that
these definitions point out a meaning of the word "structure" very
similar to that I am trying to propose. In particular you refer to
definitions 4a 4b and 5 as examples of structure that "encompasses"
content. Doing to my imperfect knowledge of English, I probably used
the verb "to encompass" in a wrong way, but I meant that structure
surely refers to contents (it organizes them), but it is a _different
thing_ with respect to contents.

According to this idea, definition 4a says that structure is: "the
arrangement of particles or parts in a substance or body ". Structure
is "the arrangement", that is the particular organization of particles
or parts in a body, but it is not the various parts and particles.
Definition 4b seems to me even clearer: "organization of parts as
dominated by the general character of the whole". The accent is on

Even when a definition refers without doubt to a material thing, i.e.
to a content, it seems to me rather a figure of speech. For example,
when "structure" is used in the meaning of "something (as a building)
that is constructed", it seems to be a synecdoche, where the part (the
structure) stands for the whole (the building).

Ultimately, it isn't my intention to quibble over this definition. I
just made a proposal that you are free to accept or refuse.

> For me, the critical difficulty with this definition is that it seems too 
> narrowly focused on a single "resource" such as a "delivery unit," 
> when I think it will probably have to be capable of handling 
> collections of delivery units (e.g., entire sites as well as single 
> pages within those sites).

I agree with your opinion, but I wouldn't forget the necessity to
distinguish when structure refers to markup code in a web page from
when structure refers to objects in the real world.

Received on Monday, 17 January 2005 23:12:10 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:33:52 UTC