Friday, October 20, 2006

Dawkins Again

Here's a very interesting video clip with Richard Dawkins being "interviewed" on American TV. So many questions spring to mind...

  • What kind of dumb programme is that? The host seemed to spend more time high-fiving members of the audience than he did interviewing Dawkins. He (the host, not Dawkins) also appears to be a few cans short of a six-pack.
  • What on earth is Dawkins doing appearing on a programme like that, especially when he refuses to debate McGrath? Is he trying to reinforce his stereotypes of Christians as stupid or something? (possibly he's trying to appear intelligent)
  • What does it say about the quality of both people when they just stoop to insulting one another?
  • I note the programme was on Comedy Central. Were any bits of that funny? (I honestly can't see the humour there...)
  • Why does Dawkins say it's not due to random chance? "Darwinian natural selection is the exact opposite of random chance." Ummmm... I thought the mutation had to be random...
  • "Nothing in nature looks random" - really, even nuclear decay?
  • I'm not sure about the natural / artifical distinction Dawkins makes.
  • "That's just so easy. If God is outside time, you can explain anything." Well, quite. What's wrong with being easy?

19 comments:

Anonymous said...

Time is limited so I will look at the video clip at a later date.

Just to respond to a couple of points: my understanding is that mutations are indeed random. However, it is not random for a particular mutation to become dominant within a gene pool, as a consequence of 'natural selection'. My favourite Dawkins book (the Blind Watchmaker) is excellent in explaining this.

Nuclear decay: I remember my mind being boggled thinking about this some time ago. Assuming a particular sample of atoms came into existence at the same time, then what causes some atoms to decay before others? Is it a result of 'random' fluctuations in the quantum energy levels of the constituents that make up each atom? My physics isn't up to an explanation, but from what little I understand I would have to say it seems random to me too.

AC

Anonymous said...

The Steven Colbert show is one of the highest profile comedy shows in the US. He as well as Jon Stewart often have authors on who are pimping their new books. You would have to watch the show for a while to get the humor, and even then you might not...it's nothing like British humor. I actually thought it was a pretty good interview given the forum.

Dawkins mentions McGrath in his new book, The God Delusion. If I have time later I'll post what he says. Or you could pick up a copy and read it yourself. You might find it entertaining.

AC hits the nail on the head in explaining natural selection. Try reading Dawkins for a better understanding of how it all works.

What is wrong with being easy is that it actually explains absolutely nothing. You have no proof, only assertions that God is what you say and does what you think he does - which is to say, anything that humans have not yet found a natural explanation of. I could just as well assert that the fairies in my garden are doing it all. It's easy, isn't it? The only difference is that I don't have the weight of 2,000 years of delusion on my side.

John said...

I'm fine with natural selection not being random - that's obvious.

My point is that Darwinian evolution has three main components: random mutation, inheritance and natural selection.

Of those, one (random mutation) is completely random and one (inheritance) is party random. So saying that evolution is "the exact opposite of random" is quite a long stretch.

God being outside time as an "easy answer" - true, there are many things it doesn't explain. But it does explain why God didn't have to have a creator, which is all it needed to explain for it to be an adequate answer to the question.

Thanks for the info about Steven Colbert.

John said...

Oh yes - nuclear decay.

It's the classic example of something which really does seem random in a quantum sense - the decay probability of a nucleus of a given isotope doesn't appear to be affected by anything.

The only real description that seems to work is that it has a probability p of decaying in time t (or equivalent).

I guess there's also the quantum Copenhagen-type description that if it isn't being observed it is a superposition of decayed and not decayed, with the weight of the two components changing with time, but if you then observed it it would go to being a random situation. Schroedinger's cat and all...

Anonymous said...

You seem to be mistaking the expectance of humor for the expectance of humour.

Anonymous said...

"God being outside time...does explain why God didn't have to have a creator, which is all it needed to explain for it to be an adequate answer to the question."

Actually not much of an explanation in the sense of "explain how rain happens". It's just an assertion without proof. You have no evidence that God is "outside time" and you can't show that "He" "exists" there or anywhere else.

Anonymous said...

Dawkins' remark about McGRath:

...after his admirably fair summary of my scientific works, it seems to be the only point in rebuttal that he has to offer: the undeniable but ignominiously weak point that you cannot disprove the existence of God. On page after page as I read McGrath, I found myself scribbling 'teapot'* in the margins. ...McGrath says, "Fed up with both theists and atheists making hopelessly dogmatic statements ont he basis of inadequate empirical evidence, (T.H.) Huxley declared that the God question could not be settled on the basis of the scientific method." McGrath goes on to quote Steven Jay Gould in a similar vein...Despite the confident, almost bullying tone of Gould's assertion, what, actually, is the justification for it? Why shouldn't we comment on God, as scientists?...As I shall argue in a moment, a universe with a creative superintendent would be a very different kind of universe from one without. Why is that not a scientific matter?

Dawkins' point is that science is much, much better equipped to answer the question of whether a God exists than theologians or philosophers.

Re: why he declined a debate with McGrath he says elsewhere in the book that he regularly declines invitations to debate. DK whether he never accepts such an invite.

*teapot being a shorthand reference to the Bertrand Russell analogy of claiming a tiny teapot to be in orbit between the Earth and Mars...I'm sure you're familiar with this.

John said...

"God being outside time...does explain why God didn't have to have a creator, which is all it needed to explain for it to be an adequate answer to the question."

Actually not much of an explanation in the sense of "explain how rain happens"...


It's more of an explanation on the level of "explain why there doesn't seem to be an atmosphere on this lump of rock we can see in the sky" - "it's not got enough mass". Yes, it's an assertion, but it's a plausible assertion that answers the argument perfectly well.

Dawkins' point is that science is much, much better equipped to answer the question of whether a God exists than theologians or philosophers.

I'd take the line that if we use the Judaeo-Christian-Islamic conception of God as outside the universe, then he isn't scientifically investigable.

On the other hand, if he has chosen to reveal himself, whether by a book as in Islam or by a person as in Christianity, then the grounds for investigation shift heavily.

Anonymous said...

It's more of an explanation on the level of "explain why there doesn't seem to be an atmosphere on this lump of rock we can see in the sky" - "it's not got enough mass". Yes, it's an assertion, but it's a plausible assertion that answers the argument perfectly well.

This is a scientifically testable assertion.

I'd take the line that if we use the Judaeo-Christian-Islamic conception of God as outside the universe, then he isn't scientifically investigable

There is a perfectly formed teapot in an elliptical orbit between the earth and mars. I would show it to you but unfortunately it's too small for our most powerful telescopes to discern. However, I have a number of ancient documents that refer to it in great detail. They were all inspired by exalted Divine Cups of Tea and they tell us all about the attributes of the teapot.

What, don't you believe me?

John said...

Scientifically testable? It really depends where the lump of rock is. If it's orbiting a distant star, and its gravitational effect on the star is not great enough to cause a noticeable wobble, then it might not be scientifically testable within the lifetime of the observers.

Teapots

You're just being silly now. The key point, as I'm sure should be clear to you, is ability to know. If someone claims that a relative of mine had an accident, then I'd take them potentially seriously if there was a possiblity they had a way of knowing. If someone claims they know that there is an omnipotent God because that God revealed themself to them, that's a plausible way of knowing. (of course, there might be other explanations too).

Steven Carr said...

Wasn't it a comedy programme? A bit like Ali G?

Steven Carr said...

'If someone claims they know that there is an omnipotent God because that God revealed themself to them, that's a plausible way of knowing.'

And Paul said he got the Gospel from revelation, not from any man.

The New Testament makes clear that Paul and Peter were the sort of people to believe what they saw in dreams and trances.

But a dream is just a dream.

And when Paul saw a man from Macedonia in a trance, he did not see a real live flesh-and-blood person, no more than the Yorkshire Ripper really had instructions from God to kill prostitutes?

Why should we believe Paul's visions , dreams and revelations?

Why should we not follow the example of other early Christians, such as the Corinthians, who scoffed at the idea that God would choose to resurrect a corpse?

John said...

Ali G was funny... (makes note to go and see Borat)

The New Testament makes clear that Paul and Peter were the sort of people to believe what they saw in dreams and trances.

But a dream is just a dream.


Lets be clear.

The New Testament makes it clear that Peter and Paul believed that God could speak to them in dreams and trances if he so wished. It's also clear that they tested what they saw against what Jesus had said to other people (see Galatians 2, for example) and Jesus had promised rather special conditions for the first leaders of the church.

You make it clear that you do not believe that God can speak in dreams and trances. On what grounds do you think that an omnipotent being couldn't do this? I'd be interested to know.

Iain said...

Custard,

You mention that mutation is completely random. I think this is actually not true, or at least a simplification of the facts. Errors that occur during replication of DNA are random events, but the mutations themselves are NOT randomly distributed across the genome. In other words mutations are concentrated in some hot spots and not others. A classic example of this is when your immune system fights off infection. A part of the genome that encodes antibody molecules mutates rapidly until an antibody arises that can bind on to the invading pathogen.

If you want to know more about this, I would suggest reading "Darwin in the Genome" by Lynn Caporale. It is a fascinating study that answers a lot of questions commonly raised as objections to evolution. Also, refreshingly, Caporale has no quarrel with religion and doesn't do the tedious Dawkins thing of continually taking swipes at religious believers.

This was refreshing to me as a Christian believer - Dawkins has too much religion and not enough science - Caporale just gives you the science, as is proper for a popular science book.

John said...

Now that's interesting, Iain. I wasn't aware that genetic mutations were non-random in location (but biology really isn't my field). Any idea what causes the localised susceptibility to mutation? (can see part of the answer, and it's nifty but I'm not sure it's completely Darwinian).

I guess there's probably not an agreed definition of "Darwinian evolution". The one I'm used to from what little reading I've done is "random mutation + natural selection", as opposed to, for example, Lamarckian evolution, where there's "directed mutation + inheritance + natural selection".

Iain said...

Hi, Custard!

To answer your question about what causes localised susceptibility to mutation would take too long for a comment on a blog post! I think your best bet would be to read Caporale's book - be prepared for quite a tough amount of detail - she delves into the science MUCH more deeply than Dawkins ever does (though it's still a popular science book).

Some ideas to get you started.
(1) The chemical strengths of bonds between A-T and between C-G are different, and so the probability of error is different.
(2) Local sequences may also be more likely than others for copying errors. Caporale gives the elegant metaphor of a badly constructed staircase where the step sizes are all different - meaning that certain combinations may make you more likely to trip up than others.
(3) Formation of "hairpin loops" in DNA. DNA sequences have a number of "mirror palindrome" sequences, where you see a given sequence and soon after the "compliment sequence" in reverse order. During copying, sometimes these pair up even though they are on the same strand, to form "stem loops". The points were there is NO match are much more likely to mutate than the points where they are matched up. This is all explained much more clearly than I can in one of the early chapters of Caporale's book.

If you do a Google on "Koshland science museum" you can find a "virtual exhibit" on DNA (Putting DNA to work) where you can probe the sequence of the entire human genome for number of occurrences of three or six letter sequences. Try a six letter one, e.g. AAGCCG. It gives back that there are 116571. You now type in the complement sequence (A/T interchanged and C/G interchanged) and in reverse, ie CGGCTT, and you find that there are 116346 occurrences (if there were a uniform random distribution you'd expect around 700,000. Note how close the two figures are, 116346 and 116571 differ by less than 300, or about 0.3%. This seems true of just about any sequence. What this seems to tell me is that most of it is governed by mirror palindromes like this, and these therefore indeed cause highly localised hotspots of mutations (ie where the symmetry is broken). Furthermore I think there is a very simple explanation as to how this mirror palindromicity formed naturally, but I'll leave that as an exercise for the reader! (I'm sure it's known about)

John said...

Interesting, particularly as that would then provide a potential partial evolutionary solution for the punctuated equilibrium problem in the fossil record.

Is that still considered Darwinian though?

I was always taught that random mutation + natural selection + time = Darwinian evolution.

John said...

D'oh! I'm being stupid. Of course the mutations are still random (and directionless), it's just that some sites have a lower probability for mutation and others have a higher one.

Iain said...

Yes, there does seem some confusion over what is meant by "random mutation". This is one small criticism that I'd give of the Caporale book, in that she suggests that mutations aren't really random. What she means by this is that the probability of mutation isn't uniform over different sites - but it is still random even if not uniformly distributed.

However, you should be careful also what you mean by "directionless". Because of the (probabilistic) constraints placed on where mutations occur, this implies that it is less "directionless" than if there was complete uniformity of distribution. Before I realised about this non-uniformity, I thought there was a big problem for evolution as one is exploring a search space of huge dimension, and one is subject to the so-called "Curse of Dimensionality" (Look it up on Wikipedia for more details). However, if mutations are effectively limited to a much smaller number of sites, then in fact one is searching a space of much smaller dimension. What it means is that if there is a uniform probability of mutation, then it's very unlikely that a random mutation will "go in the right direction", but with a smaller dimension, it is much more likely.

There is a great deal more to it than this, but I hope that this makes it clear that there is a great deal more to it than just "random mutation + natural selection". Caporale's book has the subtitle "molecular strategies for evolution" - an array of strategies, that she argues have themselves been selected for in the course of time.