[iDC] Re: Re: Everything is Miscellaneous

[iDC_fwd]

from Ryan Shaw:
ryanshaw at sims.berkeley.edu

I think Weinberger identifies some important issues: specifically that
digitalization frees information architecture from the constraints of
physical location, and that multiple organizational schemes can overlap
and possibly complement one another. Insofar as he brings these insights
to a popular audience, his book is good.

Unfortunately, his book also typifies uncritical cheerleading of
technological trends. Instead of closely examining the rhetoric
surrounding "bottom-up" organization to help people decide whether, e.g.
Amazon is really more "open" or "free" than your local library, he just
echoes that rhetoric. As a result the book comes across like a religious
tract. And in many cases he is clearly uninformed about how the systems
he discusses actually work (as Prelinger notes in the thread on iDC).

Here's some specific examples of the problems I have with Everything is
Miscellaneous (quotes are taken from Weinberger's recent presentation at
Harvard):

"...it turns out when you have enough tags (for example, Flickr) there
is so much data there, just in the tag set itself, that they are able to
cluster photos without knowing anything about the photos except the tags
they are using. And it's remarkably precise. So when you have enough
tags (despite what common sense would say) you don't necessarily end up
with chaos. You may end up with actually more meaning and quite precise
meaning."

Flickr has decent (not great) precision and extremely low recall (only a
small percentage of photos are tagged). Weinberger focuses on the former
and ignores the latter. Low recall is not a problem for Flickr, because
photo enthusiasts rarely need to see every picture taken at a certain
place or with a certain subject. But low recall is a huge problem for
scholars, researchers, patent attorneys, and investigative journalists.
Creating information systems that don't serve these needs is not
"sticking it to the man"--it's screwing ourselves over.

"The world's greatest expert doesn't matter because he refuses to engage
in a public negotiation of knowledge (for example, Wikipedia). Which is
what happens when the authority vanishes and we are only left with each
other and we engage with one another. This is how we get to the best
truth we can manage. This is through the public negotiation of
knowledge."

Weinberger's "us vs. the man" rhetoric blinds him to the very real
concentrations of power and authority that exist in the new digital
order. Believing that authority has somehow "vanished" is stupid and
dangerous. Just look at Weinberger's beloved tagging systems: it's
become clear that different design choices (tagging interfaces,
algorithms for indexing, suggesting, and aggregating tags) can radically
affect how tags are used. Those design decisions are not reached through
public negotiation--they are imposed by people in positions of
authority. And Wikipedia hasn't made authority disappear, either--it has
simply divorced authority from institutions and the credentials they
provide.