Friday, August 08, 2008

Google and the limits of knowledge

"Reports that say that something hasn't happened are always interesting to me, because as we know, there are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns -- the ones we don't know we don't know." Image: A young Donald Rumsfeld as (meta?)physicist. Image found here.

"Now if only we could google sites Tyler Cowen has not linked to."

--a comment in response to this post in which Cowen calls out, but does not name, "The Most Obnoxious Blogger in the World"

Those of you not terribly fond of deconstructionist thought may be pleased to realize that Google is its antithesis in two senses, at least. The first: No absence-is-presence shenanigans figured into its mysterious algorithms. Nope: No matter how tenuous the actual link between/among terms in your search, if a document contains them, Google will (or is supposed to) find it. Presence is presence. Dammit.

The other is that Google models, every time you use it, a firm rebuttal to Derrida's well-known declaration, "Il n'y a pas de hors-texte." "Nothing" is a pretty big claim, and Google slices and dices it whenever a search does not list a document (read: a text that Derrida alleges nothing is beyond). Ha! Wrong yet again, Frenchie-boy. One trillion pages on the Web, but this lists only 28 million hits.

All the above is foolishness, of course, but as I thought about the commenter's comment at Cowen's blog, it drove home to me (yet again) the idea that empirical knowledge, valuable as it is, can take us only so far when it comes to the question of knowing. To go beyond knowing a fact to knowing its ultimate value or worthlessness is not in the domain of empiricism. That is the domain of wonder, of the imagination. The hypothesis--the wondering--must come before the experiment(s) to test it.

Like the rest of us, Google can know only what it knows--in its case, the fact of the presence of a word or words in a document. Except. Google doesn't "know" anything, and it certainly doesn't wonder about anything. Google treats texts as the matrices within which it finds the crystals it seeks, context be damned. It only detects SiO2; it can't tell us whether what we have is quartz, or amethyst, or citrine. It is the ultimate word-processor--a term which, as I once said elsewhere, makes me think its job is to turn language into lunch meat.

None of the above is a criticism of Google--only a critique. It can do only so much, just as the imagination has to have something concrete to begin with before it can do what it does. The two are ultimately symbiotic rather than opposed to each other.

And all the while, the unknown unknowns recede? multiply? morph into some other form as we get closer to knowing them? Are these even close to the right questions to be wondering about them?

5 comments:

Anonymous said...

Ever read "The Black Swan"?

John B. said...

No, but it keeps popping up in things that I read. I'll be heading to a bookstore soon; I'll have a look at it.

Anonymous said...

To go beyond knowing a fact to knowing its ultimate value or worthlessness is not in the domain of empiricism. That is the domain of wonder, of the imagination. The hypothesis--the wondering--must come before the experiment(s) to test it.

Like the rest of us, Google can know only what it knows . . .


Damn right.

Anonymous said...

Off topic: Via Ariel comes this.

Unknown said...

Nice post, I wish I had written it.

I want the words "useful idiot" written on my tombstone.