atw: Is structured authoring really a paradigm shift?
- From: "Geoffrey Marnell" <geoffrey@xxxxxxxxxxxxxx>
- To: <austechwriter@xxxxxxxxxxxxx>
- Date: Sun, 14 Sep 2008 23:24:54 +1000
Hi austechies,
The mostly informative and often passionate responses to my claim on
austechwriter last week that knowledge *of* XML (as opposed to knowledge
*about* XML) was not necessary, and is unlikely to be necessary, for anyone
to gain sustained employment in technical writing gives me good grounds to
think that a related topic-is structured authoring really a paradigm
shift?-will also be of interest to some of you. In many ways, the two topics
are related: those who have been misled into thinking that a knowledge *of*
XML is fundamental to a career in technical writing are likely to think that
they also need to embrace an entirely new authoring paradigm, to wit,
structured authoring. They are likely to think this because it has become a
common claim these days, even by those who are usually level-headed about
XML-based publishing and not caught up in the hype and hoopla. For example,
the usually level-headed Sarah O'Keefe (from Scriptorium Publishing
Services) calls the move to structured authoring-the move to content-driven
authoring as opposed to format-driven authoring-"paradigm-shifting" (see
INTECOM, Society for Technical Communication, July/August 2008, p. 27).
Oh no it's not.
Many of us have heard this claim before. In fact, it's been spruiked for
years, ever since at least 1979 when SGML first appeared: almost 20 years
before XML, it's child, appeared on the scene. (Yes austechies: structured
authoring has been around for a lot longer than XML and DITA.)
But does theory match practice? Does any writer (can any writer?) really
adopt a format-driven authoring methodology to the exclusion of a
content-driven authoring methodology? I suspect not. Suppose, for example,
that I am writing a scientific paper, reporting the results of my
experimental research. Do I really say to myself : first, I shall start
with some heading 1 text, move on to a heading 2 text, choose a
smaller-than-normal body text format (maybe with some left and right
indents) for the next paragraph, then add some more heading 2 text, and then
some standard body text, and so on and so on. Of course not. No, we
*naturally* think in terms of content: I start with a title, then I write
the authors section. Next I add an abstract (introduced with its own
heading), followed by the introduction, then the materials and methods
section, followed by the results section, the discussion section, the
acknowledgements, appendixes and finally the list of references. The whole
paper is written in content chunks, not format chunks. We will certainly
format the content chunks (and the sub-chunks: headings, lists and so on).
But we primarily think of the paper we are writing as composed of chunks of
content or topics. Formatting is always secondary.
It is exactly the same with technical writing (and with any form of
declarative writing). Formatting is always an after-thought. The
fore-thought is the content types that will be the building blocks of my
document. I don't say to myself as I am about to begin the steps in a
procedure that I am choosing a list format. No, I say to myself that I am
about to begin a procedure. This is a content type, not a format type. I may
format it in a particular way-as a numbered list-but that is not the primary
consideration. The primary consideration is that I am about to set out the
steps to describe how a specific goal can be achieved. When I write a
warning or caution, I am not saying in my mind that I am about to enter some
bold text with a danger symbol attached. No. I say I am about to enter a
warning or a caution. This is a decision about content. When I am writing a
trouble-shooting section, I don't say that I am choosing a particular format
type. No, I am choosing a particular content type. And so on.
So, in a fundamental sense we have *always* been engaged in structured
authoring, despite the relatively recent appearance of structured authoring
tools. Far from being a paradigm shift in how we author declarative
documents, structured authoring methodologies are actually doing little more
than mirroring, at last, the way we have always authored. It is the tools
that have changed; it is not how we author that needs to change. The
structural components (or elements) that we see in DocBook, DITA and the
like are just a reflection of the way we naturally chunk our writing: how we
build a document from the blocks that are its necessary constituents. If
it's always been that way-and it has-then authoring that way cannot be a
paradigm shift.
Strictly speaking, how we author in a structured authoring environment is a
little different, but the difference is not such that a would-be or novice
technical writer needs to be especially concerned. In the structured
authoring we naturally do but with a non-structured authoring tool (such as
Microsoft Word), we are free to construct the structure as we please and are
free, too, to apply any format to any paragraph (and any character) that we
type. In the structured authoring we naturally do but with a structured
authoring tool (such as Structured FrameMaker), we are free to apply
whatever content type is appropriate wherever we have our cursor. This is
determined by the content rules in the associated schema, DTD or EDD.
Formatting is another step: it is either applied via an associated style
sheet or set out in the format rules specified in the associated EDD. But
the principal difference is that with an unstructured authoring tool, you
apply styles (aka formats) directly to the text you enter, whereas with
structured authoring tools you directly apply content types (aka elements
and their qualifying attributes) to the text you enter. Format is another
step. With an unstructured tool, you select a style to apply to a chunk of
text; with a structured tool you select a content type to apply to a chunk
of text.
And what is so smart about this is that the structured authoring methodology
exactly mirrors the way we author, and the way we have always authored. This
is no paradigm shift or quantum change. Rather, the tools are catching up
with us. It is is not us who has to catch up with the tools.
So don't be alarmed if you are new to TW and are confused by all the hype
and hoopla about the need to learn XML and the need to embrace a new model
of authoring. The model that many are apparently moving to is not new. It is
a "model" (for want of a better word) that truly reflects the way we have
always authored. Indeed, if you have written anything at all-with quill,
crayon, chalk or Microsoft Word-you have more than likely engaged in
structured authoring: content first; format second. What *is* new is that
modern authoring tools can enforce a particular overall structure. And what
is also new is the potential side-benefits of the infinitely customisable
content-rule-cluster known as DITA. But the authoring methodology you will
apply to any particular instantiation of DITA is as old as the quill (nay,
as old old as speech itself): content first; format second. Did I really
need to repeat that?
Cheers
Geoffrey Marnell
Principal Consultant
Abelard Consulting Pty Ltd
T: (+61 3) 9596 3456
F: (+61 3) 9596 3625
W: http://www.abelard.com.au <http://www.abelard.com.au/>
Other related posts:
- » atw: Is structured authoring really a paradigm shift?
