Friday, July 15, 2005

Limits Of Nature: 1D Replication

I think that there is not always enough emphasis placed upon the massive constraints that evolution has been forced to work within in order to create the amazing ecosystem of life and culture that we (humans) now participate in.

For example, wouldn't Lamarckian evolution (direct inheritance of lifetime experiences) have been a lot more effective to use than Darwinian evolution? Who wants to start over from scratch with each new generation? Why didn't Nature choose to work in this manner?


Well, there are a lot of good reasons why this didn't "naturally" occur at first; mostly having to do with the extreme difficulty of making accurate copies of information, even in only a single dimension. The medium of choice for information transfer had, until recently, been DNA. However, Nature has now discovered a few neat ways to circumvent these problems, namely through the transmission of memes.


In humans, the evolution of memes has continued to rapidly accelerate, starting with the usage of language, then writing, printing, radios, telephones, the internet, etc... (Note that cultural transmission of information is not strictly limited to animals with linguistic capacities; birds do it too.)


Interestingly enough, all of these newer methods of information transfer and copying still essentially revolve around the sequential transmission of a finite set of tokens, i.e. languages; once again, linear chains of information, be they English or ASL or TCP/IP. So nature hasn't really circumvented the difficulties of working in information mediums of greater than one dimension, it's just that it has recently stumbled upon some new one-dimensional mediums to work in!


Of course, these new mediums do not replace the old mediums, but rather build on top of them in a hierarchy of ever increasing abstraction. Entropy must always increase, but paradoxically complexity appears to always increase as well.

Suprisingly, even as the form of the information under consideration becomes progressively more and more abstract, the manner in which that information is dealt with tends to follow a somewhat universal pattern:

Complex adaptive systems tend to make strong use of linear sequences of tokenized information; they tend to use languages.