Saturday, April 15, 2006

The unity of the organism

This is another "building block" post setting up a point of philosophy for later use.

Look around you and find a book near you. For simplicity I'm going to say it's Kiln People by David Brin, because that's what's near me. It's definitely a book, right? What would have to be different to make it not a book? If it were an audiobook, would it still be a book? Large print? PDF? Galley drafts being sent to the printer? Initial manuscript, or author's outline notes?

Suppose I tore off the covers. Is it still a book? Suppose I poured ink on it so all the pages were a uniform shade of black. Suppose I tore out a page. Suppose I tore out a lot pages. Suppose I tore out half the pages. All but ten of the pages. All but one page. All the pages. Where did it stop being a book?

Logic tends to assume that a definition is a set of criteria to determine whether an object fits into a class. In mathematical terms, in the space of all possible objects, "book" defines an area in that space, an area whose precise borders are determined by the definition.

But in real life, language doesn't work that way. Definitions are actually more hueristic and "fuzzy". We define book not by the borders of its space, but by its center: an idealized, "Platonic form" of what a book is. We know that at that center, this is definitely a book; and somewhere far away, this is definitely not a book. But we don't know where, in any given direction, we cross the unseen line between "book" and "not book". Worse yet, not only will this vary from person to person, it'll vary for a particular person from minute to minute. Most people haven't even considered any of the "border" questions I asked before, and if you get them to, they'll come up with an answer that they will not like. Any criteria you develop that uniquely determine book, I can turn around and find a situation where there's something your criteria will call "book" that you don't think is a book, and one where something you think is a book fails the tests you've come up with.

It gets worse, though -- extend it into another dimension. That's not just a book, it's Kiln People by David Brin. Is the audiobook the same book? How about the large print edition? How about a slightly abridged edition? How about a heavily abridged edition? Suppose an early printing had a small error and subsequent printings fixed it. Are they the same book? Suppose a second edition had substantial changes. Suppose someone printed only the first half of it. Suppose someone printed a book that contained Kiln People, Otherness, and Glory Road in one set of covers. Suppose someone did that, but interleaved the chapters of all three books.

And it's not just a book, and just Kiln People; it's a particular instance of the book that came off a particular printing press at a particular moment. Suppose I had two copies from the same day's print run, and I started cutting pages out of copy A and replacing them with the same pages from copy B. At what point do I stop having copy A and start having copy B?

Let's not even get into the fact that in the two days since I popped this book out of its box, billions of atoms that were part of it are gone, and billions of other atoms are now part of it. Because then we'd have to talk about living organisms that might change out all their atoms many times over the course of their lives.

So many arguments about ethics, philosophy, even politics come down to the fact that we all pretend there is some precise defintion that could address all these questions, but there isn't, and thus we don't all agree on what that definition is, but we think we do. We actually go so far as to attach ethical imperatives to these non-existent precise definitions. We have significant ethical rules that differentiate what you're allowed to do to a human, or a person, or a sapient organism, or a sentient organism, or a non-sentient organism, or a non-living thing. Yet none of these classes are precisely defined, and advances in medicine and technology keep producing objects that sit in the fuzzy area near the border -- over the border for some and not for others. This is only going to get more and more so as transhumanism advances.

You can't apply the rules of logic to address these questions unless you're willing to create precise definitions. And then you're forced to accept that those precise definitions are arbitrary. And then your justification for calling these ethical rules anything but arbitrary becomes ever more tenuous.

The cop-out is to try to create a gradiated system: instead of saying whether an object is a book as a yes-or-no question, you assign a score to it, its relative bookness. Seems like a solution, but it's not. Any set of rules you create to calculate an object's bookness will, again, produce scores that disagree with your intuition, and even more, with other people's intuition. And it will be just as arbitrary as the boundary definition, so attaching ethics to it ("this action's badness score can be calculated proportional the victim's personness score") still ends up just as arbitrary.

We don't have a Cosmic Dictionary that includes revealed truth. None of the world's holy books really do any better at giving the kind of precise definitions than would an eight-year-old who's just learned words like "person". We cannot define what makes this a book, let alone Kiln People, let alone this particular printing of Kiln People, in a rigorous and yet universally acceptable way. How can we hope to do so for what makes this a person, let alone a male human person, let alone Frank?

No comments: