On the Origin of Objects is a difficult book to read; it presupposes a good familiarity with both computer science and hard-core metaphysics (knowledge of current issues of representation, ontology, epistemology, inscription errors, and a passing understanding of the history of philosophy over the past two hundred years will help). That said, this book is far more accessible than it might be, because Smith uses every trick in the book to get his ideas across, including cartoons, while avoiding complicated predicate logic or equations. And what he has to say is interesting indeed.

OtOoO (a great acronym!) is nothing less ambitious than an attempt to define the basic nature of the universe. This is unfashionable; the last person to really take a stab at it was Hegel, and he did such a monumental job--and most people would say, failed so categorically--that the whole enterprise has become suspect. But Smith embarks on his project for very pragmatic reasons: a computer scientist, he is interested in nailing down the boundaries of computation, in order to make it easier to build both computers and computer programs. His 25-year quest to do this led him (perhaps kicking and screaming) into metaphysics because he discovered early on that questions of semantics (meaning) cannot be separated from questions of ontology (being). And computers are all about semantics.

So the quest begins, and it goes down numerous paths and into strange countries; Smith uses every analogy he can lay his hands on to get his ideas across and for the most part succeeds admirably. In the end, having dismantled our current metaphysics (even those of us who like to think of ourselves as above such things have one) he reaches a very odd, but very powerful idea: the idea of registration.

Object-hood, according to Smith, is not a given, it is something achieved. In fact, a lot of ongoing work has to happen for an object to have a stable existence for any observer (computing system, person, etc.). This continuous work of stabilizing the flux of essentially undifferentiated reality is what's not acknowledged in current theories of reality; but it's this exact work that computer systems and, by extension, human brains, have to engage in in order to engage with the world at all. The act of co-constructing an object with the collaboration of the world is called registration; this word is very carefully chosen. Registration is not exactly like perception, understanding or being; it is, if you will, the activity that happens prior to any of those things.

Armed with the idea of registration, we can make inferences and new leaps of understanding in many areas, ranging from physics to cognitive science to computer science. Not insigificantly, on the way we're invited to understand our own human place in the world in a new, and much more comfortable, way. This books sweeps away the old cathedrals of cold thought--the monuments to Being, Monads, Platonic Forms etc. that have contributed so much to our modern feelings of alienation from the world. Smith's world is the exact opposite of alienated--it is participatory to the core, and scrappy, hard to pin down--alive.

Ultimately, Smith insists that studying formal systems leads us inevitably to the conclusion that the world itself is not a formal system. I agree: the world is music, the world is a house of mirrors in which even the mirrors are only reflections... but the world is not a romantic heap of vague poetry either. In the middle ground between ambiguity and precision Smith has found an idea, in fact a full theory of the world, that is not only reassuring and wise, but useful.

<  |  >