Friday, May 25, 2007

The Blind Watchmaker - chapter 6

As I said earlier, the question of "Where did the first DNA replicating machine come from?" is just as intractable as "Where did God come from?". In this chapter, Dawkins sets out to disprove this statement by completely evading the issue. Instead, he focuses on how big the universe is and whether or not there might be some unknown (and probably unknowable) chemistry that was the true precursor. The behavior of crystals is alluded to, but I don't see where this leads to anything useful. Did the first microprocessor wash up on a beach due to waves rearranging silicon crystals in sand? Or was it designed? Oh, I forgot - design is impossible, in spite of the neighbor who works at Intel!

In the earlier chapter, Dawkins notes that the Histone H4 protein is common to both peas and humans. We can deduce that the sequence (or something very close) of amino acids is vital to Eukaryotes. This protein has 102 amino acids. If only one 102 amino acid sequence is needed, then the chances of this occuring are one in 5x10132, which is 5x1021 times greater than our absolute upper bound for the number "astronomical" as derived earlier. What is the minimum set of proteins for the first replicating machine? What else is required, since the only replicating machines we know of include proteins, RNA and DNA? Even if all the pieces do happen to exist, what is the chance of them actually being assembled correctly? How many other problems are there that I didn't consider that are the minimum requirements for life? Not only does this chapter not provide any answers, but it doesn't even begin to acknowledge the size of the problem. Again, the only thing that can be said on behalf of the theory of evolution here is that no competitors are permitted, so it is guaranteed to come in first! Survival of the fittest?

Links: Chapter 1 & 2, Chapter 3, Chapter 4, Chapter 4b, Chapter 5,


Bunc said...

The Theory of evolution isnt actually a theory about how life emerged. Its not a theory which sets out to explain how the first replicating "machines" were formed.

I have less problem with Deists who argue that God (the Intelligent designer) may have caused the first spark of life to appear.

The fossil record by its nature can tell us nothing of the very irst life. Indeed if you read on this issue you will see that it is quite possible that the very first life is not the common ancestor of all current life.(basically because life could theoretically have got going a couple of times during early earth history and then beeen wiped out only to have to start again).

There is little or no direct observational evidence available to us to help us devide how life strted or arrived on earth. The issue I have with Intelligent designers is with their attempts to refute the extrememly strong evidence for natural selection processes as the driver of evolutionary developemnt - after life had first emerged.

So speculation about "designers" clearly has more scope as an "explanation" in an area where there is little or no evidence one way or the other.

Bunc said...

looney, your use of probablities in the way you do I think betrays a fundamental misunderstanding of what the theory of evolution actually implies. It is mots definately not a purely chance process as your use of probabilities implies. It is the combination of chance, reproduction and selection that characterises darwinian evolution.

I would suggest you have a look at

Looney said...

Bunc, the link doesn't address the mathematics I used with respect to the first replicating machine.

The issue is frequently framed as to whether or not Chistians (IDers) can refute evolution. This is the wrong question. A better one is does evolution refute technology? If the answer to this is yes, then science has an insurmountable problem.

Bunc said...

Looney, your last comment is very strange. In science we select and assert those theories which appear to the current state of knowledge to best explain all the observations made. We also select those theories which continue to remain robust in the face of new findings and which also ideally make useful predictions which can then be tested.

Evolutionary theory fits these requirements as strongly as almost any other theory in science. Intelligent design makes no scientific predictions, has no explanatory usefulness and is simply used by its proponents as a way of subtley attempting to undermine evolutionary theory in the minds of the average person.

It is not for darwinian evolutionary theorists to explain away Intelligent Design ( whatever that is - I would be interested to hear you explain how it works?) It is for Intelligent design proponents to actually propose mechanisms which amount to a scientific theory. We have all been waiting a long time to hear anything that meets that test.

As for your "matrhematical " critique it is in fact relatively trivial in terms of similar attempts to attack darwinian natural selection and genetic theory. I suggest that you read

which addresses some of the issues and have a good look around that site. You will find all of your issues more than adequately addressed.

I would be pleased to hear what youo think the "theory" of intelligent design actually involves and what predictions it makes. And in the process just who you think is doing all this intelligent design and at what stage of the normal biochemical/physical processes this intervention by the Spaghetti monster / God or some engineer in the sky takes place.

I would also be interested to hear how you explain the fossil record. I assume you are not so literal a Bible reader that you are a "young earther".

Looney said...

Bunc, regarding the GA effectiveness, I would point you to an engineering optimization web site:
or a more specific:

The fact that GA works (i.e. a GA/ID hybrid can tweak an ID design) misses the key issue: The engineer wants to know what the most effective method is. GA never shows up in these comparisons nor does it even show up in the optimization text books. The reason is that they are embarrasingly bad compared to what we already have. We are left with two hard facts:

1) GA is always worse than non-GA in effectiveness.

2) All of the ID is worked out before the final optimization algorithm does its tweaking, so there is zero evidence that GA can diminish the role of ID, even when it does succeed in the final design optimization role.

The subject of how ID works is wonderful. I have to run now. Have a good day.