Thursday, November 28, 2013

Berlin with Yundi and Trust with Rowan

Illuminated Christmas Statues
in Berlin last night
Back from Berlin where I went to have dinner with Yundi who is over to record with the Berlin Phil.

Just before dinner there was a crisis and when I arrived Yundi's highly experienced CEO met me and explained the problem. I suggested a solution, which was also one of the possibilities they had been considering, after some discussion they have decided to go for it!

I can't explain more at this stage though I should be able to later, and it will probably become clear in time what I'm talking about. I'd have loved to stay and see the recording, but had to catch a 7am flight back to London for an 11am meeting.

On the way there and back, in addition to working, I've been reading Rowan Williams' wonderful book "Tokens of Trust".  The first chapter (about "I believe in God the Father almighty") is called "Who can we trust" and begins "A few years ago the British philosopher, Onora O'Neill, argued...that our society was suffering from a crisis of trust."  Amongst many gems...
  • "I  believe in God..." isn't the first in a set of answers to the question, 'How many ideas or pictures do I have inside my head?' as if God were the name of one or more doubtful thing like UFOs or ghosts...It is the beginning of a series of statements about where I find the anchorage of my life,"
  • The word [God] has made is designed to become a reconciled world.
  • God can't have a selfish agenda, because he wants nothing for himself
  • There is no situation in which God cannot be relied upon
  • [God's] power is made clear in his patience and in his capacity always to bring something fresh into a situation.

Tuesday, November 26, 2013

Nowak and colleagues skewer Inclusive Fitness

Allen at al Fig 2 (see text below)
Martin Nowak sends me a new PNAS paper called "Limitations of inclusive fitness" which he as co-authored with EO Wilson and Benjamin Allen. This shows that:
inclusive fitness is ... a limited concept, which exists only for a small subset of evolutionary processes. Inclusive fitness assumes that personal fitness is the sum of additive components caused by individual actions. This assumption does not hold for the majority of evolutionary processes or scenarios. To sidestep this limitation, inclusive fitness theorists have proposed a method using linear regression. On the basis of this method, it is claimed that inclusive fitness theory (i) predicts the direction of allele frequency changes,(ii) reveals the reasons for these changes, (iii) is as general as natural selection, and (iv) provides a universal design principle for evolution... all of [these claims] are unfounded. If the objective is to analyze whether mutations that modify social behavior are favored or opposed by natural selection, then no aspect of inclusive fitness theory is needed.
 Strong words, but the text supports this. They point out that additivity seldom holds, and once you are in the realm of linear regression you confuse correlation with causation.  In particular they offer three compelling mathematical counter-examples:
  1. The Hanger-On (purple)who finds a high-fitness partner to interact with. The regression recipe yields B > 0, C < 0 misinterpreting this behaviour as beneficial cooperation.
  2. The Jealous Individual (red) who attacks an individual of high fitness. This attack reduces the recipient’s fitness from 5 to 4, and the attacker's fitness from 1 to 0. The regression recipe yields B ,C > 0, misinterpreting this attack as costly cooperation.
  3. The Nurse (blue)who helps an individual of low fitness. This aid increases the recipient’s fitness from 0 to 0.5 (representing a 50% chance of having an offspring), and decreases the nurse’s fitness from 1 to 0.5. The regression recipe yields B < 0, C >  0, misinterpreting this aid as costly harming or spite. 
It will be interesting to see what the Inclusive Fitness crowd make of it. I expect they will simply question the provenance of the authors (two Catholics and an old man) because the arguments seem unanswerable.

PS Though I'm informed that one of the alleged "Catholics" isn't in fact a Catholic, although he works at a Catholic University. Of course Mendel was a Catholic and Fisher was a staunch Anglican and the whole point about science is that the conclusions do not depend on the person of the authors.

Tuesday, November 19, 2013

The Second World War by Anthony Beevor

Been very busy with work and also a family celebration so little time to blog.

Reading Anthony Beevor's outstanding history The Second World War which gives a fascinating and masterly overview of this terrible conflict. He's perhaps a little harsh on Montgomery who he says was over-cautious and too conservative - probably true but anyone with experience of WW1 wanted to be very careful with his men. Solders trusted Monty to look after them and it may be better to have a small victory when you could have had a bigger one than to have risked a major defeat.

There are so many gems in this book that I can't quote them all.  Getting an overview of the whole conflict is well worth the inevitable over-simplification.

Saturday, November 09, 2013

Earth size planets and MaxHELP

Back from a flying visit to San Francisco and LA. Stayed with one of our dearest and most delightful friends who lives in Silicon Valley, then to LA for meetings and back (on an A380 - excellent) for a dinner party for 12 at home - the oldest guest was 50 years older than the youngest and it was a terrific evening, but I'm glad I was on LA time not Beijing!

Fig 1 from Petugura et al, showing observed planets
and calculated survey completeness.
Just before I left on Mon news broke of Petigura et al's PNAS paper "Prevalence of Earth-size planets orbiting Sun-like stars"  This suggests that the fraction of Sun-like stars which harbor an Earth-size planet with orbital periods of 200–400 days is about 6%. They also make somewhat more speculative arguments for a 22% occurrence rate of Earth-size planets in habitable zones of Sun-like stars.

Of course an earth-sized planet with a reasonable level of energy flux doesn't make it a Habitable Earth-like Planet within my definition, since we need sufficient stability (amongst other things) which almost certainly requires a Jupiter-style planet and a Moon-sized moon. Petigura et al suggest that only 1.6 ± 0.4% of Sun-like stars harbour a Jupiter-size (8− 16 R⊕) planet with P = 5–100 d and given that there seems to be a log-linear distribution of orbital period. Jupiter has an orbital period of over 4,000d and if we assume that the Jupiter-sized planet needs to have an orbital period of say 500-10,000 d in order to induce the necessary stability then perhaps about 0.2% of these stars will have Jupiters. And no-one quite knows how the Earth got such a large moon let alone what the prevalence was. Nevertheless it is at least suggestive of support for MaxHELP.

I sent a short email to Don Page and Martin Rees and this triggered a fascinating series of emails between them with additional contributions from Bill Unruh and Andrei Linde.  I need to reflect on these fascinating ideas.

I did need to point out respectfully that my hypothesis relates to Habitable Earth-like planets. I suggested that this might entail:
  1. stable average temperatures somewhere between say 280-350K
  2. Not too much ionising radiation or
  3. bombardment by large meteors
  4. M/MEarth greater than 0.1 and stability for at least (say) 2bnYr, and
  5. an abundance of key elements (esp H,O,C,N) within a factor of 2 of their observed values.
I suspect that (3) may require a Jupiter-sized planet with a relatively long orbital period, an a large enough Moon. This makes the hypothesis a bit less falsifiable than simply counting Earth-like planets in the sense of Petigura et al.

I agree of course that the hypothesis may well be false but I hope it is sufficiently interesting to warrant further investigation. Certainly if we knew enough to calculate the partial derivatives of E[HELP] wrt each of the fundamental parameters we would know an enormous amount about solar system formation than we do at present!

Maybe by 2015 there would be some point in having a seminar to explore this? In the meantime much would be learned from any comments they can give.

Sunday, November 03, 2013

A Secular Requiem

To the Cadogan Hall where a composer friend of my sister's had the London Premiere of his new piece, A Secular Requiem, performed by the Three Spires Singers. I was delighted that our old friend Catherine Wyn Rogers was one of the soloists.

First we had Strauss's Four Last Songs with an excellent Sop called Sarah Fox. Unfortunately the conductor, who was mainly a choir and chorus conductor, wasn't really able to control the orchestra who are mostly London session musicians so the balance was jarringly wrong. However it is a great work.

The Secular Requiem started well with Donne's famous "No man is an island). The librettist was a retired medical Prof (who had worked for our friend Sir Keith Peters at one stage) and he'd done a good job selecting poems and arranging them in the classical five stages of bereavement. The choir and soloists sang very well. But the work suffered from a refusal to engage with the reality of death, and the fortissimo triumphant conclusion was a travesty: even when you are reconciled to the death of a loved one and come to accept that (at some level) this is part of the cycle of life you are not exultant - not even if they are a deep believer and you are convinced of their salvation (Vide the first Fool/Olivia dialogue in Twelfth Night). I had expected that a Secular Requiem would be more attuned to the harshness of death and bereavement than the Christian ones - but no.  In this matter, far from the Devil having all the best tunes, he doesn't even have plausible words.

I did ask both the composer and librettist whether they had thought about reprising the Donne at the end - they both said they had but the composer had rejected it as being too Christian.