james - with regard to your points about irreducible complexity, I have a couple of additional comments.

Firstly, humans have "fairly reasonable" eyes in evolutionary terms, but remember that they are poor compared to say a Sparrow Hawk. A Sparrow Hawk may think that you or I only have 50% of an eye as their vision is that much better. So potentially our eyes are "reduced" versions of better eyes.

Likewise, many animals have less of an eye that humans, squid I believe have eyes very similar to ours but without a lens - their eyes are actually full of water! So in this example they have eyes that are "reduced" versions of our own.

So talk of eyes being in anyway irreducible is kind of nonsense because their are many eyes of many level of sophistication and complexity. Eyes can be reduced in complexity or indeed increased in complexity and still serve their owners perfectly well.

Which brings me onto my second point which is how eyes - or anything else with any level of dependent complexity - got that way. If you see, for example, a biological system with 3 components (A, B and C) that all appear to be part of a complex system - each part "relying" on the other, you may well wonder how it got that way. But remember this:

Part A in a more primitive state may serve a function of some sort - lets take a primitive tail for example - it wags weakly and with limited range but helps to propel our imaginary creature along. It manages to feed and reproduce well. Now a later natural variation causes a new component to appear, let’s say a larger socket for the tail. The creature with the larger tail socket has a larger range of motion in its tail, it swims faster than its rivals (with small sockets) and gets to food quicker, it is stronger and therefore healthier. It also gets to females first and as such it is highly successful in the evolutionary game.

Quite soon all these creatures have larger sockets in their tails, the decedents of our first “large socket” version have become the norm. Now a further variation may introduce a new variation to the socket - let’s say a muscle that attaches to the socket and the tail – it directly provides much more force to the tail. In this way A (the tail) may now appear reliant on B (the socket) and C (the muscle), it’s a well designed system for driving our creature through its habitat. A-B-C one without the other is useless.

Component A does a job, component B helps A do this job, but later A becomes reliant on B because this configuration is better than then previous non reliant setup. Component C then comes along and helps both A and B do their jobs better, A & B evolve to become reliant on C.*

The system as we now see it appears to be "irreducibly complex", and in a way it is - take away A or B or C and the whole system stops working. But that’s not to say that it was always that way!

(In this example for ease of explanation I "introduce" new components but remember that they evolve over many thousands or millions of years. The socket very slowly become more rounded each step aiding the tail further, only after many, many years the tail relies 100% on its socket.)

Irreducible complexity is NOT unexplainable complexity.

* In fact you can probably explain it in more simple terms that even that: A does a job, B helps A; but later, A becomes reliant on B.