2003-08-11 Distance, shape and expectations Okay, so here it is. Distance is: . the distance between abstraction layers . the halflife of causality . the measure over which the effect of a cause becomes less probable When I say "distance is" I'm saying that there is a thing that distance was, a property belonging to the universe that we have adapted to, have accreted expectations regarding. (Expectations are not just conscious things, and they're not just things that exist inside the "conscious mind" (that is, the thing that is emergent from/ abstracted from the physical brain): expectations are the surface shape we present the universe. Our ethics are such-and-such because they have an expectation that the game theory equations of a large bulk of people are solved in such-and-such a way. Our eyes are sensitive to light because we expect there to be photons with such-and-such frequencies around, and for the world to be developed in a way to support that. We cup our hands to catch a ball because we expect the ball to be round, and a roundness fits in a cup.) Distance was was it always was, but over time we have set up an expectation of distance. This expectation has two sides: an understanding of what distance is (the cup shape of the hands), and an understanding of how distance behaves (the ball fits in the cup). (Actually, distance is curious because in the real world we encounter places where useful distance slips from physical distance: wormholes (like roads, friction free surfaces of rapid transit, or shouting), and mazes (a small physical distance, but a large gulf to be crossed. A distance has been created by form: distance creates expression and expression creates distance. The two can't be separated. (Distance in this context sounds like form.))) Other properties fit in the same shape distance does - we can catch other things apart from balls - and they fit better if they have the same sort of properties distance does: distance is generally at the most basic level of a system, it's mostly insurmountable; distance affects the transformation of effects; etc. If something acts like distance, we respond to it like distance. If we respond to something like distance, we wear away at its edges, make it still more like distance, and so on. [TODO: meaning and transformations. meaning abates with distance] [TODO: distance as the universe's solution to fairly distribute side-effects for maximum possible reuse] I talk a lot about expectations and surface shapes. This is how proteins interact: they move close to one another and shapes fit into shapes. There are a number of important properties with these shapes: . they're universal. Things can't help having shape . shape, that is, information can't help being communicated . shapes can be approximate We've already seen this with distance. There are a number of other measures which act pretty much like distance. Our expectation to sex (our responses/reaction to it and so on) is such a shape. It can be approximately filled with pornography. But pornography cannot help transmitting the information that it is something-other-than-sex -- and so there can be different expectations for that, more precise ones. (Transmit is the wrong word. A shape is not transmitted, it just is.) (With expectations being noticed there is implicature. I assume my shape is recognised. I can also assume you are recognising my shape, and act accordingly.) (The shape is the face. There are evolutionary pressures to lie with the face, but at a certain point we've reached a stable equilibrium where the arms races doesn't continue and we have a thing we can both understand, involuntarily communicate with maximally (even the absence of any part of communication is important), and draw implications from. The face. Shapes are the universe's solution to this same problem.) Mechanical devices interface with shape. They use cogs and axles to transform motion. It's not quite the same as proteins because these devices can complete hide what is inside with their external faces (and this fact and led us down a blind alley to the container metaphor). But the external faces - the handle and the end of the axle - are surfaces/shapes/interfaces that are properties of the machine itself: they are not APIs. (This is what "being true to the medium" is all about.) Expectation is the level of well-fittedness of one shape to another. But shape must include another property: the distance of the shapes from one another. Not only physical distance but all distances, abstraction layers and so on. A cog emulated in a computer program cannot mesh with a cog of brass. There is a larger shape, the ur-shape, which includes all the properties which two shapes need to fit into one another: the distance between them, their concentrations, the probability of meeting (across time, which is another matter), the distance between their abstraction layers. (This is being more specific. We gain specificity at the cost of generality.) (Each abstraction layer is a level of expectation. Distance is used to let side-effects dilute, so a mechanical universe can come out of a probabilistic, quantum, one. A city with roads and buildings and taxes can come out of a place with weather and tribes.) We have certain expectations of how the universe works, derived by evolution over a long time. For complex systems we call these things like folk physics, folk psychology, folk systems theory, folk economics. Folk game theory is called ethics, or manners. Possibly these folk sciences are sometimes wrong. Sometimes not. (There are folk systems - rules of thumb - that are built up consciously too. This is like rhetoric as a method for presenting persuasive arguments. Sometimes the folk and the reality overlap: folk grammar - universal grammar - is grammar itself.) When we're trying to formalise our understanding of the universe we can get things wrong, but it's useful to formalise the understanding because it can be handled, communicated, worked on together -- and we can derive some kind of implicature from it. Perhaps we can game it. In our folk systems theory we have a problem in how to divide things into parts. Dropping a stone from the top of a tower. [TODO. There's something to say about affordances here. Consider a balloon blown up that's burst, the inside disappears. Turn the balloon inside-out then burst it. Where's the inside? The rest of the universe doesn't disappear. The inside of the balloon, the fact is bursts: these aren't properties of the balloon, they're properties of topologies, the universe, physics, all overlapping.]