Re The Politics of Schemas

[Cc'ed to my usual public mailing list, so that if we continue this
conversation, others might benefit. -- KGC]

>>>>> "david" == David Steinberg <[email protected]> writes:

  david> Hi Kendall I have just got to reading your two-part article.
  david> In my view, it is a thoughtful and thought provoking piece -
  david> congratulations.

Thanks.

  david> Here are a few comments:

  david> 1. I found your approach somewhat static in that I sensed an
  david>    assumption that interests were known, well understood and
  david>    fairly static e.g.. see the para "Schemas
  david>    formalize...world-chunk.".  

I think I assume, in a kind of abstract way, that interests are
*knowable*. More importantly, what I need to make my argument work,
and it's hard to find any disagreement about this, is that interests
*conflict*, and that one side is getting to determine, unilaterally,
that its interests are the ones which structure the way interactions
between machines and people happen.

I'm not really sure I understand what you mean by static, but I take
it as some kind of general complaint; i.e. dynamic things are better,
mutatis mutandis, than static things.

Maybe so, maybe not. I don't know.

            When any major change comes -
  david>    electricity, railroads, telephone, computer, fall of the
  david>    Roman Empire etc. the long term implications take time to
  david>    be discovered.  Opportunities and constraints, not
  david>    compatible with people's current worldview/paradigm/life
  david>    experience, are hard for people to imagine let alone
  david>    accurately predict their secondary, tertiary etc. impacts
  david>    which may be pervasive and extremely decisive.  

Sorry, but I don't really follow any of this.

            Previously
  david>    axiomatic assumptions need to be challenged, new interests
  david>    discovered and the practical implications of shifting
  david>    understandings of group and individual interests must
  david>    repeatedly be thought out, challenged and reconsidered.

Uh... okay. But what relation does that have to my piece? Note that I
don't offer my claims as a comprehensive accounting of the
possibilities of the Semantic Web. I offer it as a description and
analysis of one (fairly obvious to me) problem, which needs to be
addressed, but isn't being.

  david> 2. your understanding of "schema" also comprehends "ontology"
  david>    recently defined as "..ontology is a document or file that
  david>    formally defines the relations among terms. The most
  david>    typical kind of ontology for the Web has a taxonomy and a
  david>    set of inference rules" (The Semantic Web, Scientific
  david>    American, May 2001, Tim Berners-Lee, James Hendler and Ora
  david>    Lassila
  david>    http://www.scientificamerican.com/2001/0501issue/0501berners-lee.html).

I chose "schema" as a catchall term for what some people call
"ontology" and for what others call "vocabulary" or "language"; that
is, there are rational orderings of these terms into a coherent
system, but what I had to say is independent of any such
ordering. Regardless of what you *call* it, what I call a "schema" is
happening, indisputably, all over the place. That's the important
thing.

(Actually, that's not quite right; I say in more detail what I mean by
"schema" in the piece, it's just that I don't have much more to add to
it. Irrespective of whatever public formalism one uses to express
one's understanding of some world-chunk, the resultant formal
understanding is what I call a "schema"; that is, I don't really care
whether it's formalized by a DTD or symbolic logic or Java classes!)

  david> 3. In my view, Topic Mapping etc. can be a useful approach
  david>    for providing a lens for recasting data in ways reflecting
  david>    different interests.  XSLT might also have a role in this

Well, yes and no. Does Topic Maps provide a way to look at a data
collection in a different way? Yes, it seems to.

Is that the problem I describe in my paper? No.

Take an application that makes heavy assumptions about privacy and
privacy rights and personal information. As real-world legislation
continues to lag behind technology (and as politicians begin to
understand the massive jurisdictional issues involved), cutting-edge
applications, widely-deployed and widely-used, especially in
monopolistic markets, become a kind of legislation unto themselves.

Topic Maps is great, but my bank, say, won't let me use Topic Maps to
instantiate a different privacy policy, one that reflects *my*
interests, to replace the understanding of privacy instantiated in
their banking application.

In cases where various parties, w/ conflicting interests, share equal
access and control over a common data collection, then Topic Maps
might be relevant. The problem, of course, is that case isn't the
common case, as far as I can tell.

  david> 4. I found a kind of either-or approach in your article
  david>    e.g. -

  david> "With regard to the Internet's infrastructural technology
  david> that will power the Semantic Web, to say nothing of XML
  david> itself, the recent overweening concentration on commerce, to
  david> the exclusion of other fields of human endeavor, is part of
  david> larger cultural and historical trends that place the
  david> stateless corporation at the center of human life as lived in
  david> the West (Schiller 1991). And yet the earliest promise of the
  david> Web (as well as its precedent hypertext forms) was always for
  david> a richer, more collaborative, less commercial environment
  david> than we face today. No one should be surprised by that
  david> outcome."

  david> The fact that commerce takes a lead does not mean that the
  david> complex, variegated Semantic Web will not also provide a
  david> context for "...a richer, more collaborative, less commercial
  david> environment than we face today."

But my claim, which I can warrant in any number of ways, is *much*
stronger than "commerce takes a lead". My claim is that commerce is
driving, as it always has (well, together with the urge to kill our
fellow human beings in greater numbers more efficiently; i.e.,
commerce and war are the two main engines of technical achievement --
together with a differently-situated impulse to *understand* the
world), the growth of this technology, and that the technology is
being driven in directions that seen inimical to what I take to be the
early and interesting and humane promise of the Web.

Non-commerical, human concerns definitely take a backseat to the
demands of commerce in almost all of the research done at the W3C and
elsewhere about the future of the Web. Just read the W3C's site to see
that this is the case.

  david> 5. You correctly state

  david> "Schemas are often created by institutions whose sole purpose
  david> is financial profit. Institutions create private schemas to
  david> use internally, within the institution itself; and they
  david> create public schemas to use externally. The number of public
  david> schemas increases daily. Schemas are being created that will
  david> structure exchange between people, between people and
  david> institutions, and between institutions, including all levels
  david> of government, corporations, and non-governmental
  david> organizations."

  david> This is not a new issue.  it is highly analogous to the
  david> paradigms earlier imposed by government forms, the decisions
  david> on how government services are to be subordinated to various
  david> departments and agencies, the background of staff hired etc.

The difference, of course, is that at least *formally*, I have some
input into decisions about the allocation of government services and
resources, at least, "I" do insofar as I live in a State that is so
structured.

With Semantic Web schemas, though, I don't have any assurances of
input or even access, not even sterile, formal assurances that have no
de facto value.

That seems new to me, and even if it isn't, it's very troubling.

In short, I tried to show that 1) schemas are shot through with
political considerations, 2) that these political considerations are
of a kind that citizens in democracies ought to be fully interested in
and engaged with, 3) that citizens are completely in the dark about
all of this, and, therefore, 4) corporations and schemas are forming a
kind of "unacknowledged legislation", to steal a term of Shelly's, to
which citizens have neither input, nor access, nor control.

That's the nastiest kind of politics of all.

Best,
Kendall Clark