-----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of [EMAIL PROTECTED]
[snip] > I believe the best comparison in electronics terms would be multiple crosslinked >(parallel?) > processors working simultaneously. Nick (for example) could probably >define it better. :-) Maybe. It touches on a metaphor that has been bouncing around inside my head for quite a while -- the notion that people and brains organize themselves similarly. But it is still rough, so I'll offer it only as a discussion point. As we become more aware of social structures and deliberately create increasingly sophisticated ones to bring about order, as in government, I suspect that we are discovering and inventing patterns that already exist in nature. What seems to be invention is often recognition of a higher order of structure than we previously could see. Perhaps the simplest way to look at social structure is that it's a big game of "telephone." I tell you something, you tell somebody else, etc. This view ignores feedback and errors. Perhaps that's a hunter-gatherer society, where there are few resource constraints, so that the consequences of feedback and errors are relatively small. When resources are limited, the knowledge that comes from feedback and the penalty for error (which are interrelated, of course) become far more significant. The power to create order becomes centralized. This looks like feudalism to me, a bunch of highly independent small networks. I think that's the structure that resembles a bunch of interconnected processors; they can accomplish more because they can work in parallel, but they limited to the "intelligence" of any one of them, because the interconnection is weak. As the information that flows among these largely independent groups becomes smarter, a structure like the computer science notion of a neural net emerges because the groups begin to have greater influence on the decisions of the others. (At this point, it's probably obvious that this metaphor has fractal qualities: it applies within feudal city-states as well as between them.) The printing press, by increasing the sophistication of the interconnect, metaphorically transformed western civilizations from a parallel computation cluster (raw horsepower, which is at least a pun, given the role of knights in feudal times) into a distributed computation system, in which the components perform complementary, rather than duplicate, tasks. Recognition of the role of feedback led to development of social structures that prevent undue centralization of power -- democracy as we know it now, with its emphasis (worship, I'm afraid) on the free flow of information. Although the most successful democracy retained aspect of the city-state system through strong states' rights, we tend to discount that as a pragmatic compromise to the ideal of a monolithic network in which everyone has the same information and the majority rules. But we are discovering, as I think nature did, that a system that deliberately creates information boundaries among structures that watch over one another, is the next level of structure. More than just a distributed computing system, it is a network of networks, in which information is processed at multiple levels, so that the communication among higher-level structures is so sophisticated that to call it feedback would be a vast oversimplification. The "multiple crosslinked (parallel?) proc! essors" model fails because it implies that the communication is taking place at the lowest level of data representation, when just the opposite is taking place. In fact, we are pushed, hard, to invent high-level representations because we are otherwise overloaded with data -- "information overload," as we call it. Along the way, corruption is always lurking around the corner. We learn to communicate with greater abstraction using higher-level representations, but the evil-doers (sorry, had to say it) will figure out how to hide their misdeeds in the shortcomings of the data structure. For example, consider the SEC's disclosure documents as the data structure for economic communication. Enron figured out how, for a while, to present consistent, believable data in that structure, hiding misdeeds, pushing us to reconsider both the data and the oversight that is supposed to stop those who would take advantage of weaknesses in the structure. So I think we sit at a point where we have recognized the value of networks of networks, their ability to structure and route information so that the system is most efficient. This raises a couple of questions for me. Is there a structure beyond this? Recognizing it, will we inevitably invent government and social structures that deliberately embrace it, as modern democracy embrace feedback. In such a society, efficient information representation and routing would be as vitally important to protect as free flow is in ours. Government's role might be to protect information standards and pathways from corruption. My favorite example of such an issue is Microsoft's attempt to corrupt Java with a proprietary version. Sun stopped it on a trademark principle, but it seems to me that a higher principle needs to be articulated, so that such protection is broad and deep. Imagine the effect such a principle would have on the current debate over digital music formats. Of course, I have wandered far from your question, but these thoughts have been rambling around in my head for a long time and it was time for them to spill out. Now that I put all this together (somewhat coherently, I hope), I'm thinking this is worth polishing into a real essay. Nick
