Sunday, June 19, 2016

Definitions Considered Meaningful

My monster comment may give you the impression that handling definitions are important to doing philosophy. This is because they are.

Philosophy getting tricky on you? Not sure what's up with chicken eggs? Try definitions! Definitions make life easy.

I have catchphrases. One of them is, "If you have a definition, the answer is trivial, if not, the answer is undecideable."

Definitions are the logical equivalent to mathematical coordinate systems. The underlying facts - whether your triangle has three sides or whether it's isosceles or not does not depend on the coordinate system, even though all the numbers describing the triangle do depend on the coordinate system, and it's impossible to describe a particular triangle without said numbers. Similarly, the facts of chicken existence don't depend on the definitions, but whether you say 'the chicken came first' does depend on your chicken definition.

First, let's take evolution as true. Let's say 'chicken' means 'viable organism whose equilibrium is a chicken.' The equilibrium bit is so young chickens are still chickens. In this case, the egg and chicken arrived at the same time, because the chicken embryo is a chicken. If instead we use 'organism that lays eggs that grow into chickens,' then chickens came first, because the last proto-chicken laid a chicken egg, and was thus itself a chicken. If we decides chicks are too different to really be chickens, then chicken eggs appeared before chickens.

Now let's take creationism as true. God goes 'parp' and there's a chicken. Chickens came first. Or maybe God goes 'parp' and there's some eggs. Eggs came first. Programmers can exactly simulate this with spawn_chicken(), making pixel images resembling chickens. Perhaps they instead use spawn_chickens_and_eggs(20), and then eggs and chickens came at the same time. Twenty of them, specifically.

Trivial.

If a tree falls in a forest, does it make a sound? Well, do you mean 'soundwaves' or do you mean 'audio cortex recognition'? Again, the answer is trivial.

The only real difference is dire apes tend to get attached to particular definitions, but not to particular origin points.

A bowl is supposed to hold soup. But then there's no bowls in space, or else gourds are now bowls. Cracked bowls aren't bowls.

There's an attachment to the bowl-shape definition. This is not normally a problem for bowls, but is a problem for selfishness. However, it's a problem for bowls too, because there no useful non-fuzzy definition of 'bowl.' We can't go down to the quantum level, because a bowl will instantly stop being a bowl after a thermal fluctuation, but not going down that far means the definition is necessarily imprecise.

In practice dire apes have an archetype 'bowl' which is precise within natural instrumentation limits, plus a scalar 'similarity to' function. Bowls which are more than a little dissimilar get a modifier, like 'tall' or 'shallow.' A broken bowl is called 'broken,' and so on. It's not a bowl when there's a different archetype with a higher similarity scalar. If it's about equal it gets called something to the format of 'bowl-cup'.

Dire apes don't seem to like this system of theirs, however. They want precise boundaries, rather than appreciating the scalar as it is. This is the sorites paradox. Further, philosophers straight-up have to use precise definitions. Though, for philosophers at least, the answer is simple. The soros is never a heap, because there's no such thing as bowls. Definitions, like coordinates, are arbitrary, and so while there's no spoon, there is a definition 'spoon' which particular clusters of sense-data match sufficiently.

However, this means philosophy does a thing backward to the intuitive way. Having proven an implication of 'bowls' or spoons or heaps or George Washington or chocobos, the proof doesn't actually apply to bowls. It applies to the particular definition, which we happened to call 'bowls.' Let X = bowls. We prove X has a bunch of properties. X may be very bowllike, but since it is necessarily precise and bowls are necessarily imprecise, they won't be quite the same thing. The proofs will have a domain of validity, defined by the definitions used. If I prove a bunch of things about flatware for eating soup out of, it will not be valid for bowls in space, except by coincidence.

This is why I wish academics were still performed in Latin. It would mean philosophers could get on with it without disrupting native meanings of words like 'selfishness.' Since Latin is now out, for preference I'd re-start with ancient Greek, it being strictly more tasteful.