Monday, November 30, 2009

From Spears to Aircraft


A Sophisticated Flying Machine: A series of incremental technological changes has brought this configuration into existence.

This post by GilDodgen on Uncommon Descent gives an indication of the central place that Irreducible Complexity (IC) has in the ID theoretical canon; everything in ID theory seems to be either an aspect of this principle or a detail. This is what GilDodgen says:

The more we learn the more it appears that almost everything of any significance in living systems is irreducibly complex. Multiple systems must almost always be simultaneously modified to proceed to the next island of function. Every software engineer knows this, and living things are fundamentally based on software.

Evolution in the fossil record is consistently characterized by major discontinuities — as my thesis about IC being a virtually universal rule at all levels, from the cell to human cognition and language, would suggest — and the discontinuity between humans and all other living things is the most profound of all. Morphological similarities are utterly swamped by the profound differences exhibited by human language, math, art, engineering, ethics, and much more.


GilDodgen then goes on to suggest that the vast differences between human beings and other primates are evidence of the discontinuity imposed by IC.

In biological terms the crucial concept is structural stability; that is the ability to self perpetuate. But this is only possible if the components of a biological structure serve to promote self perpetuation and they can only do this if they work; that is if they function correctly.

Crucially for ID, GilDodgen assumes that functionality only comes in absolutely isolated islands; that is any change of a structure that goes beyond a certain small threshold entails a loss of functionality and thus a non-viable unstable structure. Thus if IC is true then there are no incremental entry points in or out of a stable structure. Thus IC structures (by definition) have no stable precursor structures incrementally separated from them. Ergo, evolution cannot happen.

The inverse of the notion of IC is Reducible Complexity (RC); RC requires that the domains of functionality/stability are so connected that there exist ways of incrementally modifying a structure and yet a stable structure being the result. The RC conjecture is that there is class of functional/stable structures that form a fully connected set and that a broad range of structural complexity is found in this set; if such exists then a barrier to evolution is lifted.

Of great significance to ID theory is this quote from GilDodgen: “Multiple systems must almost always be simultaneously modified to proceed to the next island of function”. That simultaneous modification is the only way to reach the next island of stability is, in the mind of the ID theorist, sufficient evidence of the complete isolation of islands of stability in a huge sea of non-viability; for to jump the gap between islands of functionality using random changes would require the simultaneous adjustment of several features in just the right way. The probability of these changes coming together is utterly negligible and thus a barrier of improbability isolates the islands of stability. The only thing that we know capable of jumping the space between the islands is intelligence.

Currently I have several issues with the concept of IC. The following points are really areas of research rather than killer arguments against IC.

1) If the only way to reach a near neighbor stable structure is by a very rare simultaneous change of features, then the IC case holds. However, if a route exists to a stable neighborhood structures via a “path” consisting of a series of incremental changes to single features ,each of which results in a stable structure, then we have a Reducibly Complex connected set of structures. However, these "single feature change" pathways are likely to be a relatively rare occurrence in the space of all possible changes, thus giving the impression that they don’t exist.

2) Human technological advance has only been possible because the limited quantum of human intelligence (call it i) can leap the gap between islands of functionality. But, and here is the important point, i is not large enough to leap the gap between stone age spears and GilDodgen’s aircraft in one leap. However, it is clear that the islands of functionality in technological configuration space are close enough so that given the quantum of human intelligence, i, the human mind can leap the gaps in functionality leading up to an sophisticated flying machine. It is conceivable that there are structures out there that can never be reached by the human mind because they are effectively irreducibly complex with respect to the quantum of human intelligence i. Nevertheless, it is clear that given human technology, large parts of technological configuration space are connected enough (i.e they reducible complexity with respect to i) to make technological evolution possible. It seems that IC and RC are not binary opposites but come in degrees. Fortunately for us technological configuration space is reducibly complex with respect to i. But, and this is the 64 billion dolllar question, is biological configuration space sufficiently connected with respect to the blind watch maker?


3) Human Intelligence itself appears to employ a kind of evolution; it makes a series of incremental adjustments to its concepts, concepts that are either selected or rejected; search, reject and select is the general evolutionary algorithm.

No comments: