Monday, April 16, 2007

How do things change? Ontology, Perdurantism, and Endurantism

Consider a famous philosophical problem, called "the ship of Theseus". This particular ship was used heavily by the Athenians. Each time it was in need of repair, they would replace bits of wood planking (or whatever was necessary). After years and years of use and repair, most if not all of the ship had been replaced.

The question is, "is this ship still Theseus' ship - or something else"? Furthermore, if we had taken all of the old material that was being replaced, and had built a second ship with the castoff material, wouldn't that ship be Theseus' ship?

A related question is "Locke's socks" (from the philosopher John Locke).

John Locke proposed a scenario regarding a favorite sock that develops a hole. He pondered whether the sock would still be the same after a patch was applied to the hole. If yes, then, would it still be the same sock after a second patch was applied? Indeed, would it still be the same sock many years later, even after all of the material of the original sock has been replaced with patches?


This is an important philosophical question. In the human body, skin cells, organ cells, just about everything is being sloughed off and renewed. Over the course of many years, there are few if any cells in your body that are the same cells as those you are born with. In other words, you have been materially replaced many times over the years. Are you still you? How did you retain the essence of what makes you you, even though you've been replaced?

This is a question of ontology. Ontology is primarily a sub-field of philosophy, but I'm interested in it for its impact on a computer's ability to model reality. (See this link for way more than you want to know about ontology)

How do things change? How can we understand the process by which they change? Surprisingly, this is still an open question. I think it's one of the reasons why people and computers are still lousy at modeling reality - because we can't agree on how to conceptualize it in a way that makes sense in all situations.

People who adhere to "perdurantism" (i.e. objects "perdure" over time) believe that objects are in essence 4-dimensional. The 4th dimension is time. You remain you, despite your cells being replaced, because your cells are temporal aspects of the substance you. At different points in time, you may have different characteristics, just as a mathematical function may have different values of x for certain values of y, yet still be the same function.

People who adhere to "endurantism" (i.e. objects "endure" over time) believe that "material objects are persisting three-dimensional individuals wholly present at every moment of their existence". From the ontology site:
(...) That is, only the present exists. The past and the future do not exist. Thus, for a substance to exist at all is for it to exist at the present moment. This view is also called "presentism." But since substances never come into existence, every substance must have existed at every past moment in the history of the world. And since they never go out of existence, every substance will still exist at every future moment in the world’s career. In other words, substances are identical through time: each substances that exists now is identical to some substance that existed or will exist at every other moment in the history of the world.

Based on my limited understanding, I interpret the definitions of endurantism to mean that you are no longer you, because you've changed. This seems more compatible with a "multiverse" outlook. You at this moment in time are a certain substance, and that substance will endure for eternity, and has always existed. The moment one of your skin cells is replaced, that's a separate substance which itself has endured for all eternity and will continue to endure. I may however be misunderstanding the arguments about this viewpoint - this is the best I can figure. Note that in any case, endurantists focus on present - they really do seem to deny the existence of a past or a future.

Pretty strange, huh? Here's where it gets sticky - to my reading, the perdurants are right, or at least their viewpoint makes the most sense to me. In the computer realm, when modeling reality through the use of ontology it turns out both tacks have certain practical benefits. Depending on whether you are modeling a process or an object, taking the wrong philosophical perspective can cause you to reach a situation where some aspect of reality can't be modeled. That's bad, because we want to create flexible upper ontologies that can accommodate any aspect of reality. In fact, from what I've seen the argument between endurantists and perdurantists is one (of many) reasons why philosophers despair of finding a single upper ontology that everyone can agree on.

If you thought you were going to get a definitive answer to this, I hate to disappoint you but I haven't got one. One of the subtle errors philosophers have to avoid in this situation is confusing their model of reality with actual reality. We attempt to fit models as best we can to our empirical observations about reality. In most branches of science, when you can't articulate a model that covers all of the bases, (similar to physic's lack of a grand unification theory) that's reality tipping you off that you have missed something or misunderstood some aspect of reality. If we can't cover our bases with either endurantism, perdurantism, or some mix of the two -- we may not be seeing the whole picture.

And if you followed all of that, my friends, I'll leave you with this conclusion which has nothing to do with philosophy or computer science:

These "gaps" in our models are exactly why you don't have to be religious or even spiritual in order to feel a profound sense of wonder about the workings of the universe.

0 comments: