Hofstadter’s GEB has a thread – his eternal golden braid I guess – on strange loops from the outset, mathematical, visual and musical at the obvious level of his title, and clearly from the content, he’s leading to the idea of emergence of “intelligence” from multi-layered (recursive, cyclical) patterns of complexity.
I noted a clear strategic / tactical game-theory angle in the “evolutionary” cycles of development. Whether it’s the mathematical quest for a model to completely contain all others, or a record player to play all records designed to break it (GEB shows its age there), or a model that humans know more than any formal system can, or the game between advertisers “honesty” and their target audiences, or a bio-evolutionary battle for survival, they are all battles of (metaphorical, anthropomorphic) wits. The key point that at any given “level” of current strategy, it’s a matter of trading tit for tat tactics, until either side finds a different strategy, and moves the game on a level. The other side may intially cry “foul”, but must find a new tactic that acknowledges the new strategy. Strategy continuously leap-frogs tactic, until it finds itself to be just another outmoded tactic in the face of a new strategy. It’s an inevitable driver for evolution.
The aspect that creates the drive is the desire for a “response” that is the response to beat all responses. ie it accomodates all previous responses, but for good measure claims to include the defintion of (or response to) itself, so that no further response can trump it. It really is the Russelian set of all sets vs Godel’s incompleteness. You can’t win, but you can’t step out of the game either. (NB out of the “frame” … frame analysis here, and a Pirsigian connection too, and for me the original metaphysical “bootstrapping” problem.)
Two other corollaries …
(1) The problem is when you give the concept, “the idea of the mother of all responses” a name, a definition. because you are always then back in the linguistic game. Any conceptual high ground is lost. (Hence the attraction of Zen, at least as a concept, if not a solution ;-))
(2) Talking of mothers, it is also in fact a manifestation of a conundrum that I’ve called the parent / sibling problem in classification / ontologies. Supertype is called “a widget”; a popular subtype is also commonly known as “a widget”; a less common subtype (of the supertype) is also a widget by inheritance, but is usually know as a “special widget”. Special widget gets very confused about the identities of its mother and sister. (Also the “I am my own Grandpa” variation on that theme.)
The really attractive positive conclusion of all this … there can never be a complete formal model of everything or a metaphysics with solid foundations and defences … get used to it … forget that, but do we really have “mechanism” to explain the emergence of intelligence from physical patterns ? Wow.
[Post Note : Of course this links directly to the “learning organisation” and “theory in use” cycle stuff in the original business context in my dissertation – low-level adjustments within the current “model” as loops within higher loops of “strategic shift”. Bullshit baffles brains, but that doesn’t make it wrong.]