The simplest strategy in designing this CS would be to come up with a table that lists all possible, or at least all relevant, signals in one column and matches them with the appropriate behavior in the next column – the CS as a table look-up, and all behavior as simple reflex. Would this work?
Not for anything other than the simplest of organisms. Because, among other things, the import of the signal needs to be able to vary considerably depending upon its context, and this context is not just the immediate environment, or, rather, the complete set of signals being received moment by moment, but also the state of the organism itself, including "memories", "expectations", and "purposes" (the quotes being just a reminder that these terms are suggestive place-holders at best at this point).
No comments:
Post a Comment