IA Summit 2006 Opening Plenary — David Weinberger

Speakers – ASIS&T 2006 Information Architecture Summit

Here are my notes from David Weinberger’s opening plenary.

What’s Up With Knowledge??

David Weinberger, March 2006 IA Summit

Everything is Miscellaneous

bizarre artifact of publishing on paper that you have to be *done* with something

we’ve been org’ing ideas using same principles we use for physical stuff
a limitation we don’t need due to digitizing, so we now need new ways of organizing.

imporant changes to knowledge

Data -> Info -> Knowledge -> Wisdom … to my mind, you have to ask why is this an appealing formulation

looks too much like the old formulation of evolution, apa up to man…Jay Gould saying this is a fallacy.

we generated all this information and it wasn’t helping us, still a lack

stripped out context to get stuff to fit into databases (!!!)

how do we get to knowledge if we stripped out the context???

this idea that we can go from info to knowledge is very new in our culture … for 2500 years our culture has thought knowledge had to do with a lot else.

Hume: Impressions, sense data/perceptions, then three relationships we order them through, resemblence, contiguity, cause, then the filter of reason…then *maybe* we get knowledge

Facts, experience and wisdom — what we need to make it work

Wisdom is what guides knowledge, not a product of it. (traditonally)

This has the causality backwards, this new formulation.

You need knowledge to figure out what info you want.

They very prototype of wisdom in our culture was socrates who was wise because he said I know I don’t know anything.

“informationalization” — the reduction of all experience, knowledge, feelings and awareness to information. It’s permeated what DNA is — everyone seems to know it’s information. We depict DNA as a nice clean info view — the double helix

DNA is actually clumps of atoms … twisty and irregular. Not information. Important we can address it as info…but we don’t usually confuse the magical landscape.

Seven properties of knowledge:

1 Knowledge … started as connection between knower and known.
2 Same for everyone — only one knowledge. world is very confusing in its appearance, but knowledge is the ‘true’ view. it’s simpler than the world. because of that, most things aren’t knowledge.
Doesn’t matter who says it if it’s true.
Bigger than we are. Francis Miksa — we believed “there ixists a realm of knowledge that grows through individual contributions and is transmitted from generation to generation such that its existence is thought to be continuous and is capable of being exasmined.” like gardeners and garden
same descrip works for physical libraries
7 Orderly: a system. Hierarchy/tree is the main way of ordering. We don’t have to think explicitly but we can if we need to by going through the tree…compresses a lot of info and highly valuable knowledge artifcat.

(Same for all, one, independent of knower, outlasts us, orderly)

These are formal properties.

We’ve had to collect and save knowledge in physical objects. Can’t be in 2 places at once. have to have primary categories. Paper — have to explain things in 2-D. Our knowledge largely has been shaped by the nature of books, paper, and econ of publishing.

Traditional knowledge view under attack:

Postmodernism, Wittgenstein, etc.

(in some ways tagging an absolute fulfillment of posmodern ideas — readers and taggers are better at saying what the b ooks are about, like what postmoderns say)

Rush (?) showed it’s not how we think … a robin is a much better example of a bird than a penguin. Not an ostrich. Your parents pointed at robins and said “birdie!”

don’t argue about what is tagging: point at it! delicious, flickr. some other things are close, but delicious etc are the prototypes.

Digitizing everything:

Three orders of order — gross simpl.

First Order: Organize things themselves

Second Order: metadata — organize that (huge reduction of knowledge to come up with that second order — lots of utility, and we’re used to it)
(what constitutes metadata is simply the thing you use to find the other stuff — including the thing itself — line between metadata and data ends up being an artificial construct necessitated by limitations of paper)

Third Order:
What can you do easily digitally that the real workld makes really hard?
1. leaf on many branches
– photographic equipment –put it in as many categories as possible to sell it to lots of people, makes it messier traditionally having a clean order was 2. ? sign of virtue, crossing over said your work needed to be redone. digital world: messy is better — lots of associations.

3. Unowned order:
you see UNC’s faceted classification system… allwos users to dynamically select trees that suit them, always leap from a branch.

tagging also allows users to control org of info

Amazon: tagging system that nobody uses.

Tree designed by experts for use by others.

Excludes what we don’t need to know.

Pile of leaves: each highly imbued with metadata of all sorts, including tagging but not just that. The old value of exclusing the noise gets turned on its head, has more value the more it includes. Excluding stuff from the pile has negative value. You can afford to include everything because storage is cheap and users can filter on the way out, post-filtering.

Knowledge’s properties:
one and the same, siple, impersonal, bigger than we, filtered, orderly, has a know-er

Possibly most authoritative journalist is a comedian, john stewart — because of his point of view and how he says things.

Editors at a newspaper: filtering — old version — editorial function has already moved to the web. Digg, etc.

Everybody here reads the newspaper through the web, i guess. You may get the paper paper, but it all happened yesterday, I read it for other reasons…i’m online most of the day, but maybe more importantly i learn news from blogs and things. Socially filtered sites, mailing lists, this web is a huge recommendation engine. Also a huge distraction engine.

Authority: beyond its utility, it gives you social standing, aggregates power for institutions, can control conversations, personal virtue (we’ve been told in our culture) a fulfillment of species destiny–disastrous idea, but it’s there. And money.

wikipedia doesn’t have authority! not the way britannica does — they have a board of advisors, etc. that’s how they earn their credibility and authority… so why do we believe the article on the JFK assassination. We will believe it if it’s a major article more than if it’s a minor one. In part because if it’s a major article we’re more likely to know something about it already.
Trust it more if lots of edits and discussion.
Another thing that give sit reliability — also posts these metadata notices saying we should question the article. The fact wikipedia is willing to tell us where it’s weak increases our trust in it!!

A list of ways encyclopedia articles can go wrong. Neutrality disputed, e.g. Trust article because the conditions are understood. That notice will NEVER show up in the NYT or Britannica.

Traditional sources for all their value, for truth vs. authority, opt for Authority!!!

One reason: printed on paper.

Publicly Negotiated Knowledge:

Tomato page — how to pronounce! Publicly Negotiated Knowledge.

The greatest expert in the world, if she’s unwilling or unable to engage this public negotiation, is OUT of wikipedia.

Knowledge WITHOUT a Knower!

Individual authorities don’t count unless can negotiate position in public.

One thing about wikipedia makes me nervous: imagine 10 years from now, most of the major and minor topics have settled down. Used to their being fact boxes. Knowldge is a commodity as well, wikipedia representing the knowldge we agree on (some of we…nature of we problematic, of course)

Wikipedia not a single thing, more of them in other languages, and others who do their own. Lots of wikipedias all based around it.

Does this mean we have separate knowledge? Baseline from which controversy emerges? Re-calcification of knowledge? Knowledge alliances? Fragmentation or reflection?

[my note: don’t we have this now, and always have?]

Point 2: new infrastructure (“getting vaguer and vaguer and speaking faster and faster”)


Heidegger: Being & Time

hammer — to understsand the hammer you have to understand other things like nails, and wood, and thus lumber yardsa nd trees and forests and economy and sun and earth, etc. Everything. It’s referential context. Which is what Heidegger called “The World” — this is “MEANING”

relationships among things … we cannot make it explicity, it’s the unspoken that allows us to speak.

we have a tendency to externalize parts of our consciousness — libraries meaning, calculators arithmetic — are we in fact seeing the externalization of meaning??

Big problem with this — if we were to TAG a real jar of jam. we’d put a lable on it, seeing strawberry in context on jar on a stand, if we saw a picture at the same jar art flickr, cann’t rebuild the context of the real jar. Too rich in the world. We’d have to rebuild the world. Can’t happen.

Can’t make things explicit.

My kids: if i could make everything explicit about them, it would mean we have a very shallow relationship.

Start button: not raised so nobody could see it was a button, looked like a label. had to add a sign, “press start b utton” (on gas pump)

Quotes morville: the inability of folksonomies to “handle equivalence, hierarchy , and other semantic relationships cause them to fail miserably at any significant scale”

if the task is to rebuild the world, certainly it’s true… scaling is part of the solution though, won’t rebuild the stuff entirely, but scale a way of getting there. Over a hundred million photos tagged at flickr, can ccluster cat noses vs. dog noses, work sbetter with bigger set.

will never get to perfect hierarchies — not with tagging, like we did by hand… i don’t want aircraft control or much of science to be handled with tagging.

I’m happy with good enough — beer on a hot day example.

We just need good enough info.

“cool, local and refreshing”

We’re going to keep going with plalists, tags, etc. we’re going to do our own front pages, get really godo at them, and have social groups based on what we’re making … at a pace like we cannot imagine. we’ll draw from ti what we want. It is not the case that the king is dead: no more hierarchy…of course not. But something is going on… the king isn’t dead but has fallen in love.
In love with this rich context of meaning, to have knowledge and hierarchies.

Author: Andrew Hinton

I use information to architect better places for humans. More at andrewhinton.com.