Wednesday, June 27, 2012

Back from Cambridge Ma combining work with admiring new granddaughter

Back from almost a week in Cambridge, Mass combining work with admiring lovely new grand-daughter. Had dinner with Martin Nowak at Harvard Faculty Club discussing amongst other things the article  Brian Josephson and I are writing for Prospect about the Dawkins/EO Wilson controversy. Finally started reading The Telltale Brain - terrific. I didn't realise Ramachandran was also a Trinity man!

Nice long email from Bill Newsome who has kindly looked through the Appendix B on Neuroscience of Questions of Truth. We must meet and discuss.

Saturday, June 23, 2012

Congratulations to Brian Josephson on the 50th anniversary of the publication of his stunning paper  - "Possible new effects in superconductive tunnelling," Physics Letters 1, 251 (1962) work for which he won the Nobel Prize. There is a conference at Cambridge today in his honour. We're collaborating on an article about evolution.  He wrote this paper when he was 22.
I've been very busy so only just seen that Martin Nowak's latest paper on the Evolutionary Dynamics of Cancer was published in Nature on the 17th. Note the very nice agreement of theory and experimental data - none of these wishy-washy correlations with an R^2 of about 0.3.

I've also only just seen the Obit of Andrew Huxley who died on 31st May. I knew him a bit, and vividly remember taking him for dinner at Midsummer House in Cambridge with my son Rupert and with John Polkinghorne. He was in great form, and remaked that he had started training as a doctor but dropped out. I couldn't resist remarking: "just think, if you had qualified, you might have won two Nobel Prizes".

It seems that Jerry Coyne is something of an intellectual coward. Having been hopelessly trounced in his attempts to pick holes in Questions of Truth, he has decided not to allow any comments from me on his blog to point out (politely and rigorously) some of his errors. Ah well. I approached Prof Coybe about this and he seems now to be allowing my comments again. So I take this assertion back. Sorry Jerry.

Friday, June 15, 2012

Evolution of Mathematics and the Microbiome

Piqued by our interaction on the Prospect article, poor old Jerry Coyne bought Questions of Truth and then tried to take pot-shots from his blog (and here, here, here). Of course he could only do this by taking responses completely out of context, thus making himself look rather stupid/dishonest. Eventually he resorted to blocking my comments since he was hopelessly losing the argument.

The interesting issue that came up was the evolution of mathematical ability. Polkinghorne suggests in QoT that "conventional Darwinian theory is unable to explain [the human ability to study noncommutative algebras and to prove Fermat’s last theorem], which requires for its understanding the belief that our environment is not limited to the physical and biological but must also include contact with a noetic realm of mathematical ideas, into which our ancestors were increasingly drawn."

Crude estimates of number of microbial species
and genes at various sampled sites.
From Nature News and Views article
Jerry pretended this was a "maths proves Jesus" argument  (but it is in response to the question "what makes humans more significant than animals")  But of course we're simply suggesting that conventional theory needs to be extended to deal with it.  It's clearly somewhat related to the evolution of language, which is obviously not just a matter of "kin selection" - languages exist and evolve in groups of speakers and presumably thrive to the extent that the provide survival advantages for the group as a whole. This is not just a matter for humans. Birdsong is (of course) not simply genetic but learned  (see this fascinating paper in Nature which elucidates some of the mechanisms). It would be very interesting to start to sketch out a theory of the evolution of complex mathematical ability: this would surely have something to do with the group-selective advantages of deeper mathematical understanding and something about how mathematical ideas are communicated, assimilated and rewarded.

Fig 4 of Yatsunenko et al.
Distances between Fecal microbiota.
Click to enlarge
Another new-ish area which completely demolishes the kin-selection fundamentalists is of course the Microbiome and some fascinating papers came out in Nature and elsewhere about this. Self-evidently the microbes in our bodies (about 10x as many as there are human cells) do not particularly share our genes but this symbiosis is essential - indeed we depend on our bacteria for our health. Part of the point of kissing, eating and drinking together etc.. is presumably to share microbes and indeed the counterpart which is the state of our immune systems.  All of this is done in groups and not just in kin, and does not depend on genetic relatedness. (PS: indeed Yatsunenko et al have lots of fascinating data about this, and find that the fecal microbiota of mothers of teenage US twins "were no more similar to their children than were those of biological fathers, and that genetically unrelated but co-habiting mothers and fathers were significantly more similar to one another microbially than were members of different families."  The bacterial communities evolve as people age, and diversity is significantly greater amongst children than adults. All fascinating stuff, and well beyond "kin selection")

Of course we can't expect people like Dawkins to understand these issues properly: people forget that he was never a professor of science at Oxford, merely Professor for the public understanding of science - a chair bought for him by a rich admirer.  And as a very distinguished scientist explained to me earlier this week, it was perfectly clear that his FRS was as a "General" candidate ie “on the basis of their wider contributions to science, engineering or medicine through leadership, organisation, scholarship or communication” and not as a “Mainstream Candidate” ie for “contributions to knowledge and understanding in science, engineering or medicine.”

Sunday, June 10, 2012

The difference between predictions of Evolutionary Dynamics and Kin Selection

Fig 2 of Field et al.
(click picture for larger version)
Very happy family event has rather interfered with blogging. And of course wonderful Jubilee!  But I'm doing some more work on this Kin Selection business.  The more I look at the predictions of Kin Selection the clearer it seems that they are pretty weak.  I'm trying to look through any papers published in Nature but at present the best I can find is one in Phil Trans B which is referenced by Stratssmann et al in their response to Nowak.  They reference this for the claim that "Other experiments manipulate costs and benefits and show that kin selection is predictive". Hmm... By eye that's an R-squared of about 0.2.

By comparison this is the kind of fit to data you get from evolutionary dynamics. I know this is a bit unfair because Field et al are reporting experiments whereas this paper is based on observations.  But nevertheless ... spot the difference.
a, The evolution of 177 verbs from Old English (green) over time, through Middle English (red) and Modern English (blue). The fraction remaining irregular in each bin decreases as the frequency decreases... b, The regularization rate of irregular verbs as a function of frequency. The relative regularization rates obtained by comparing Old versus Modern English (green) and Middle versus Modern English (red) scale linearly on a log–log plot with a downward slope of nearly one-half. The regularization rate and the half-life scale with the square root of the frequency.
Fig 1 of Lieberman & al 2007