[Sigia-l] IA system components - add to the list!

[Arno Reichenauer]

<Eric Scheid> wrote:

>On 15/3/03 12:10 PM, "Boniface Lau" <boniface_lau at compuserve.com> wrote:
>>> Why can=92t we first of all just collect what each one of us is
>>> contributing to a working website or intranet and then use the sum
>>> of it as a starting point not for defining IA but for defining IA
>>> methodology, i.e. the process you have to go through to obtain these
>>> contributions?
>>=20
>> If I understand you correctly, what you call "IA methodology" is
>> really a process for drawing information architects' contributions
>> into the development of various web site components. It is a misnomer
>> to call that IA methodology.
>
> No, you didn't understand correctly. He's not saying that the sum of
> components =3D=3D a methodology, he's saying the sum of components is just
gris=
> t
> for the mill, the process, of *developing* a methodology. I'd expect part
o=
> f
> the process of developing that methodology would be an examination of the
> relevance and value of each submitted component, and also an analysis of
> dependencies and inherent flows so as to ascertain where there are gaps in
>the corpus (ie. components that were not submitted).

Exactly, Eric. The web site components that come from IA are an analytical
starting point for developing a "IA methodology" (better: IA process
framework, see below). A process usually is developed by starting with the
desired outcome. Then you can deduct what your process has to cover.

>> The collective result [...] is a web site, not an information
architectur=
> e.
>
> A difference in interpretation could be because I'm also not limiting
> "contributions" to meaning "artifact present in the final product". Maybe
> Arno was only thinking of the latter (or maybe not - Arno?), but that
> original position is inconsequential: if we were to use the wider meaning,
> in the process, could we develop a methodology?
>

Of course, IA is more that the artifacts present in the final product, lots
of intermediary stages etc etc. But we don't have to make a myth of it.
There's the desired result, that's what you have to do to get there. IMO,
the quality of a process is measured in terms of its end results. Nobody
cares how you got there if the result satisfies their needs. What's
necessary to get there, indeed is one step deeper. But you can only decide
if you know what the outcome should be.

> ps: counter to the whole idea of "grand unified methodologies", I am
> actually partial to the notion of a toolbox of techniques and
> understandings. Christina W. writes about this idea in Blueprints,
refuting
> in the process the underlying assumption that a "methodology" is actually
> useful and valuable.

True, Eric, I also read Christina's book and conced the point to her that a
fixed methodology is of no use in the field. But in any project, there have
to be achieved at least some of what I call IA system components. And these
can not be achieved each in isolation, as they are dependent on each other
(e.g. navigation, labels and metadata). Therefore, my goal is a process
framework for IA which shows how the basic process parts have to be
interconnected to result in a high-quality IA system. Remember, a process
FRAMEWORK: adjustable to the specific needs of the project at hand, no fixed
methodology.

Arno