Nokia, Ontology, and Legal Research
John Wonderlich
Nokia, Ontology, and Legal Research
September 3rd, 2007 by John Wonderlich · No Comments
Among my favorite blogs lately has been Future Perfect, apparently a Nokia researcher's personal brainstorming repository. I enjoy it so much because it reminds me of how I take notes, through either practical categorization or an off-the-cuff to-be-revisited kind of personalized conceptual banter that "Future Perfect" offers frequently. I find this refreshing because I'm a fan of unfinished thoughts, which are often be more productive to share than polished ones, if only because they arise so much more frequently.
The Open House Project's success has come from just such a public conceptual space, a precondition for real collaborative analysis and constructive advocacy. If the societal influences that affect democratic decision making are fundamentally changed by the Internet and the broad public access that it promotes, it makes sense that a project promoting exactly that kind of public access would be among the first to demonstrate success. In a certain sense, the kind of public access that we're advocating for is a precondition for the civically-empowered society we'd like to see, and the accountable transparent government that is built on it.
In the spirit of loosely connected observations potentially generating progress, here are some things I learned reading this weekend.
I finally got around to reading The Legal Research Manual, which a former employee of mine gave me (a pre-law student). The first few chapters were particularly relevant for me, since they provided a broad practical view of government information sources. I didn't know about the US Government Manual, which is a yearly document by the GPO outlining the organization and function of the entire government.
The book also helpfully distinguishes between administrative (executive), statutory (legislative), and common (judicial) law, depending on its source, and further distinguishes between digests that organize this law based on either subject matter or chronology. All of the above are organized in both ways, leading to the existence of both an authoritative compendium and a running update. (For example, the Federal Register prints regulations from federal agencies as they're passed, among other things, and the Code of Federal Regulations lists all of them according to topic.)
The Manual also explains judicial organization in great detail, with special attention payed to judicial jurisdiction (that construction may be a tautology) and practical approaches to researching case law and legal citations. Supreme court decisions are all apparently published in the United States Reports, which was good to see. Legal information, however, seems to suffer from an ad-hoc organizational system combining both official public prints and authoritative private services. For example, West publishes a "US Code Congressional and Administrative News" update, and digests of court opinions from throughout different levels of the Judiciary branch.
The linguistics appreciator in me loves that lawyers rely on a book called "Words and Phrases," which catalogs legally articulated semantic distinctions. This would be a sort of legal ontology, I suppose.
Ontology, in philosophy, refers broadly to the study of being, in the same way that ethics is about action or epistemology is about knowledge. Ontology often gets brought up in discussions of structured data, and I've usually taken it to refer to the way that markup languages choose to recognize certain qualities or attributes as existent in certain settings. According to my understanding, a philosophical outlook leads to an ontology, or the ontic properties, those of existence, in the same way that the context of a given programming language leads to certain attributes. The radical doubt of solipsism leads to the lonely existence of the self, or the medieval metaphysics of Aquinas yields a complex hierarchy of angelic beings, just as RSS standards lead to attributes of title or publication time, or the context of HTML makes tags for italics or headers existent.
What got me thinking about all of this was Clay Shirky's Ontology is Overrated, which is a great read. He outlines categorization schemas, and concludes that the imprecise statistical organization that arises from crowds organizing on the Internet is more useful than the top-down methods of organization like card catalogues (depending on the context).
Intellectual interaction's electronically magnified potential will have considerable consequences for democratic governance. I assume that the effect will be a good one. In terms of corruption and conflict of interest, I wonder what the ideal conditions are for it to grow? I assume that a low information sharing environment is one of the basic conditions for administrative abuse. I wonder what else helps corruption to thrive? Is this studied systematically?
--
John Wonderlich
Program Director
The Sunlight Foundation
(202) 742-1520 ext. 234
Clay Shirky
> the same way that ethics is about action or epistemology is about
> knowledge. Ontology often gets brought up in discussions of
> structured data, and I've usually taken it to refer to the way that
> markup languages choose to recognize certain qualities or
> attributes as existent in certain settings. According to my
> understanding, a philosophical outlook leads to an ontology, or the
> ontic properties, those of existence, in the same way that the
> context of a given programming language leads to certain
> attributes. The radical doubt of solipsism leads to the lonely
> existence of the self, or the medieval metaphysics of Aquinas
> yields a complex hierarchy of angelic beings, just as RSS standards
> lead to attributes of title or publication time, or the context of
> HTML makes tags for italics or headers existent.
>
> What got me thinking about all of this was Clay Shirky's Ontology
> is Overrated, which is a great read. He outlines categorization
> schemas, and concludes that the imprecise statistical organization
> that arises from crowds organizing on the Internet is more useful
> than the top-down methods of organization like card catalogues
> (depending on the context).
Glad you liked the piece - one quibble, but possibly a relevant one:
my position is that tagging's advantage isn't that its more useful
but that it's more _practical_, which is to say that, in moderate
data sets with a moderate number of users, it generates good results
at lower cost, and with large-scale data sets and populations
(datasets and userbases many times the size of the Library of
Congress) it gets better while professional efforts degrade.
There's a lot to like about professional classification, except for
its inutility at post-paper scale.
You could see this in action, brilliantly, with the 'Find out who was
editing Wikipedia by organization' effort -- the creator couldn't
look through the db himself, but by opening it up, he got distributed
work which was lower quality in the individual case but high quality
in aggregate. Political datasets are similar -- we'll see a lot more
corruption uncovered when people can do the looking themselves.
> Intellectual interaction's electronically magnified potential will
> have considerable consequences for democratic governance. I assume
> that the effect will be a good one.
I wouldn't assume that, for historical reasons outlined below.
> In terms of corruption and conflict of interest, I wonder what the
> ideal conditions are for it to grow?
My vote for most important condition is the ability for the public to
take direct action.
The great repudiation of the idea that openness is per se good is the
rise of the lobbying industry after the Watergate reforms. Prior to
1974, lobbying was a bit like visiting a shaman -- the lobbyist would
take your money, and assure you that they would pass your message on
to the relevant committee members. A vote would be taken in
committee, and you’d see the results. Critically, though, the two
steps couldn’t be linked; like waving chicken bones, some action was
taken, and some outcome reached, but any causal link was fairly
speculative. The committee votes were secret, so neither you nor your
lobbyist could say for sure what your money had bought, and there was
always the risk that the members of the committee were hashing out a
compromise different from the outcome you’d paid the lobbyist to
influence. The congressman could always emerge from committee
assuring your lobbyist that he’d fought hard for your position, but
that in the end, he was overruled, and there was nothing for the
lobbyist to do but take him at his word.
Transparency changed all that. Once all deliberations and votes were
on the record, the lobbyist could actually monitor compliance. Each
member of the committee could be watched, both for what they said or
didn’t say, and for how they voted or didn’t vote. Special interests
are the bug in large democracies, so it's hard to say whether special
interests had more or less leverage post-1974. It is easy to say,
however, that the massive increase in transparency occasioned by the
last round of sunlight didn't lead to either a massive decrease in
the leverage of special interests.
When the millions of people paying higher prices for sugar, say, can
take on the thousands profiting from the current agricultural policy,
we'll know something has changed.
-clay