Re: [Gnu-arch-users] the buzzword paper

[Thomas Lord]

ralf:

> There is a technical term "ontology" that has emerged in recent years,
> which has nothing to do with the philosophical term.
> http://en.wikipedia.org/wiki/Ontology_%28computer_science%29

Thank you.

Yes, I know that (and it's obvious from the context of the original
paper).  While writing my long post I actually made the conscious 
decision to not confuse my point by bringing that up, assuming 
that someone would bring it up in a reply :-)

The technical definition given to "ontology" is redundant: the word
for that definition is "taxonomy".   At best, the poor word choice
is naive.  More realistically, it's a dangerous naivete.

The taxonomic vs. ontological distinction is, already, long before
the use of either in software engineering, a deep philosophical
and ethical concern:

Taxonomies are socially active (because the rules of how the
power of institutions is operated are expressed in terms of
taxonomies).  Get classified as a witch in 17th century Salem, MA.
and then see what happens to you.  Get classified as a "three-strikes"
criminal in CA see what happens.   Power organizes, reproduces, and
deploys itself using taxonomies as a cornerstone of the foundation.

Taxonomic vs. Ontological confusion is exactly where our institutions
go badly wrong: it's where systematic injustices happen; it's 
where systematic oppression happens.  So deep are the institutional
and social effects here that it is where conceptions of Self often
come from (Thus ends your quick introduction to Foucault through
Tom's eyes. :-)

There is a wealth of serious thinking about the social responsibility
that comes with being placed in an institutional position to define
a taxonomy.   A simple example can be found in library science where
taxonomic decisions determine and shape people's access to and
perceptions of available information about the world.   In some 
other fields, people take this distinction as important -- central
even -- to the benefit of us all.

Screwing up the language this way in software engineering cuts
practitioners off from that accumulated wisdom and discourages them
from contemplating otherwise well-understood ethical concerns.  Witness
the casualness with which, for example, the authors of the paper we 
are discussing leap from an artifact in their taxonomy to the conclusion
of a *hypothetical* "social problem" among programmers -- which
connection they will now, it seems, set about to *"prove"*.

In software engineering, we are now tinkering with unprecedented
feedback signals where the signals in question are large essences of
human culture and one of the components in the circuit is large 
groups of people in consequential circumstances.   The new feedback
circuit is beyond human scale but its personal consequences for many
individuals is at a very human scale.   Power is reconfiguring, and
far from obviously in any of our interests.  (Witness me and the history
of this project.)

-t