RSS
Science needs the check of experiment and observation because we aren’t smart enough to generate very long sequences of inferences from first principles without walking off into fantasy land. For example, check out this exchange between Freeman Dyson and Richard Dawkins. Dyson is a physicist who has made some serious contributions to our body of knowledge. He is a realized genius; the promise of his brilliance was kept. Richard Dawkins on the other hand has made his mark as a popularizer of science. With a second class degree from Oxford, he has admitted difficulties with differential calculus (see A Reason for Everything). Dawkins is a bright enough fellow when measured against the mean, but in terms of g I have no doubt that Dyson stands head and shoulders above him by ~1.5-2 standard deviations. Nevertheless, Dyson says the following:
First response. What I wrote is not a howler and Dawkins is wrong. Species once established evolve very little, and the big steps in evolution mostly occur at speciation events when new species appear with new adaptations. The reason for this is that the rate of evolution of a population is roughly proportional to the inverse square root of the population size. So big steps are most likely when populations are small, giving rise to the ”punctuated equilibrium” that is seen in the fossil record. The competition is between the new species with a small population adapting fast to new conditions and the old species with a big population adapting slowly.
I don’t do the “fisking” thing, and in this case it is too easy. I just want to fix onto his formal allusion; Dyson is basically conflating random genetic drift with evolution itself. He should look up Neutral Theory and its prediction of a constant rate of substitution. And then there is Fisher’s Fundamental Theorem of Natural Selection. But even if you reject these models, extrapolate from the assertion he makes to the world around you: do species with the smallest effective populations evolve the fastest???
To think with speed and clarity is critical in modeling the world around us. But to know facts is also important. Freeman Dyson has a excellent inferential mental engine, but he obviously isn’t able to derive the insights of evolutionary genetics from first principles on the fly. He should read a book, otherwise he’ll have to deal with being schooled by minds of far lesser rate.
Note: Dyson reports that his response to Dawkins’ critique was presented verbally at John Brockman’s farm in the presence of biologists. That would be George Church & Craig Venter at the least. Perhaps I’m a retard from bizarro world and everything I’ve learned is turned upside down, but I’m a bit mystified as to why Dyson wasn’t immediately corrected. OK, actually I’m not. It seems likely that Dyson was around molecular people would couldn’t respond with a sentence as to why the rate of evolution isn’t dependent on population size in the way he believes.
English
Arabic
Chinese (Simplified)
Dutch
French
German
Hebrew
Hindi
Italian
Japanese
Korean
Persian
Polish
Portuguese
Russian
Spanish
Swedish
Turkish
Vietnamese
That reminds me of the time David Deutsch was pwned here by a local.
do species with the smallest effective populations evolve the fastest???
interesting. That would be my intuitive assumption also, though you’re clearly stating it’s false and quoting stuff like Fisher and whatnot that I’ve never read (and am unlikely to read unless someone helpfully prepares a brief on basic genetic drift vs evolution for me.)
I’m likely 4 standard deviations below him, but Dyson and I apparently have the same inferential engine when it comes to population dynamics. Is that because of our training in physics and math?
/begin arguing out of arse unsupported by any actual due diligence
In a nutshell, I’d intuitively assume that evolution in a smaller population would be faster than in a large one simply because there is a smaller pool, so changes in genome would be transmitted throughout the pool faster. Granted, there would be numerically more mutations/changes in a larger pool, but the relative impact would be small. In a sense it’s like a mini-blogsphere; its easier to be “known” and get links and traffic from your peers in a small group than in a gigantic one. It is simply easier to be an authority in a small pond.
/end arguing out of arse unsupported by any actual due diligence
question, has anyone ever related any aspect of evolutionary theory to a Zipf distribution?
azizhp
Yes, I thought Dyson’s comment made sense too. But then razib said “extrapolate from the assertion he makes to the world around you” and it occurred to me that the champions of rapid evolution are the viruses and bacteria that mutate to defeat our vaccines and antibioticsm, and they surely have colossal populations.
So maybe (relative) simplicity of the organism is the key to rapid evolution?
Certainly puts his recent anti-global warming/climate-modeling critique in perspective.
Aziz, here’s a hint as to why that intuition is wrong: the equation (derived by Kimura) for number of generations to fixation for a positively selected allele is (2/s)ln(2N), where s is the selection coefficient and N is population size. Assume the number of novel mutations per generation is equal to N and think about how this scales. Also note that the capacity of selection to overpower drift on nearly-neutral loci increases with N (just think of it as analogous to an increase in statistical power from having a larger dataset), so ceteris paribus there will be more loci that undergo sweeps rather than drifting randomly.
keypusher,
It’s not so much about being simple as having a really short generation time, but other than that you’ve got the right idea. In one sense a simpler organism may actually evolve “more slowly”* (all other things equal, which they aren’t) than a more complex one, because more complex organisms have much more elaborate designs with many more axes along which changes can be made. A bacterium has fewer degrees of freedom than a bear.
* Scare quotes because this isn’t really an accurate description, but I just got up and can’t think of a better one. The point is about how many possible moves are open to an organism in design space, and that given a constant mutation rate a population of more complex organism will move in more directions per generation.
let me put it this way: dyson comes close to not even being wrong. matt covers some of the fundamental issues, but ultimately these are empirical scientific questions and you need to explore the various parameters. intuition is a place to start, but it can’t replace experimentation, observation and the grinding working out of the formal models.
so changes in genome would be transmitted throughout the pool faster.
i think this a major issue, as i said on my other weblog. we are using a physical metaphor, a “gene pool,” and so it just seems like something could “get around” easier in that pool. i think some of the intuitive traps could be removed by taking a “gene’s eye view” as dawkins often suggests, but if you read the rest of dyson’s response it is clear he doesn’t want to do this.
I hesitate to say anything because I’m not clear about the actual issues, but it seems that geniuses jumping from one discipline to another sometimes revolutionize the second discipline, and sometimes just look stupid. This can be true even when they jump from a harder discipline to a softer one.
I think that the movement of physicists and mathematicians into social science has had very mixed immediate results, though the long-term net result will certainly be positive since the negative results will wash out. (A second factor is that many guys jumped only after they found that they couldn’t make it in the tougher science.)
ultimately these are empirical scientific questions and you need to explore the various parameters. intuition is a place to start, but it can’t replace experimentation, observation and the grinding working out of the formal models.
well, of course. Though Matt’s quoting a derivation by Kimura suggests that he (Kimura) approached it from first principles, ie intuition, rather than just plotting the data out and fitting a curve. Is that correct?
empirical results are at their most powerful when they are applied to disprove/prove an intuited theory. I alluded to this in my comment by asking whether anyone has observed Zipfian distributions in any aspect of evolution; the Zipf has been observed empirically all over the place in information theory (for example, the internet). Now, Matt has mooted that question by pointing out the relation actually scales with log N, but i wonder if that is just as empirical. If not, then what counter-intuition – or, if you will, the layman’s argument – is operative to make Dyson close-but-wrong where instead he would have been right?
also, i think the “gene pool” metaphor has a literal interpretation; the number of people in the population (N) multipled by the number of genes. It’s much more likely for a gene of mine to meet a gene of yours in a downstream individual if we are both patriarchs in a small community. I am thinking specificlaly of my Bohra community here; I’ve got only about 3 degrees of freedom from anyone else whereas about 6 for the global population (and probably 4 or 5 from you Razib, even though we are both from the subcontinent).
As far as what Dyson himself is thinking, I dont much care 🙂 im asking all these questions to figure out what i am thinking now, though Dyson was useful in leading me to this line of inquiry. Though I dont doubt i am well below him in general g/iq whatever, I also think he’s highly fallible, and i trust my intuition more than I do his on any matter outside of his physics expertise. Computer modeling included; see this.
Razib:
It seems to me (another amateur, and no Freeman Dyson either) that neutral evolution ought to work faster in a small isolated population, but that mutations arise as a function of the size of the population. Would this lead to the fastest neutral evolution happening when you had a large population that was split up into a lot of isolated subpopulations most of the time? I’m thinking partly of early humans, spread across the globe in usually-isolated populations that occasionally would mix.
How do we measure “speed” of evolution? By the number of new, conserved allels as a percentage of all conserved allels in a given species, per unit of time (or generation?), or by the absolute number of new, conserved allels? This strikes me as a poorly defined metric.
Sorry, Aziz, the Bohra are not a small endogamic community, they number over a million and are a missionary sect. You are not on the way of speciation.
[i don’t need a water engineer’s authority on questions of evolutionary biology -razib]
Luke, I think the per-generation temporal measure makes the most sense, because it sort of normalizes across the different longevity of species. Otherwise the bacteria have too much of an advantage, the little buggers.
J, I wasnt suggesting that Bohras are on their way to speciation, however intriguing that would be 🙂 I was just pointing out that as N increases, teh degree of connectivity decreases.
but it seems that geniuses jumping from one discipline to another sometimes revolutionize the second discipline, and sometimes just look stupid.
well, dyson has noted that he thought crick was making the wrong move to go from physics to biology 😉 in any case, this isn’t a case of a physicist becoming a biologist, this is a case of dyson being ignorant and pig-headed.
Though Matt’s quoting a derivation by Kimura suggests that he (Kimura) approached it from first principles, ie intuition, rather than just plotting the data out and fitting a curve. Is that correct?
here is his 1968 paper. he makes a reference to lewontin & hubby’s results. in short, they found the extent of polymorphism (molecular level variation) to be far higher than expected. neutralism seems to have filled that breach. jim crow (one of kimura’s collaborators during the period) told me in a 10 Q that it wasn’t that surprising if you thought about it, but it seems obvious that the empirical results really forced a more serious debate about the possibilities of neutral evolution.
Now, Matt has mooted that question by pointing out the relation actually scales with log N, but i wonder if that is just as empirical.
re: zipf, i think it might have some relation to a few mutations of large effect vs. many small. i think this is more of an empirical question since biological populations are subject to a lot of contingent constraints (i’m more with wright here than fisher).
It seems to me (another amateur, and no Freeman Dyson either) that neutral evolution ought to work faster in a small isolated population, but that mutations arise as a function of the size of the population. Would this lead to the fastest neutral evolution happening when you had a large population that was split up into a lot of isolated subpopulations most of the time? I’m thinking partly of early humans, spread across the globe in usually-isolated populations that occasionally would mix.
you are thinking something like sewall wright’s shifting balance. not neutral evolution. as matt alluded to, the rate of substitution in hard neutralism is constant.
mutation rate = 2 X pop size X mutation rate
pro of fixation of mutation = 1/(2 X pop size)
the pop size and 2 (cuz of diploid) cancel out. so you are just left with the mutation rate.
This strikes me as a poorly defined metric.
yes, i think that dyson doesn’t even know what he’s talking about. in terms of ‘rate of evolution’ i think on a molecular level the rate of substitution on a nucleotide is a good enough measure. by this, i mean the rate at which a gene, locus 1, goes from being more 99% of allele A to 99% of allele B.
p.s. the nearly neutral theory does come close to what dyson is saying…but i don’t think that’s what he meant.
Luke, I think “number of substitutions per genertion” gets pretty well at the heart of what we mean by speed of evolution, but the differentiation between rate of *adaptive* evolution versus *neutral* evolution needs to be kept in mind at all times.
Aziz, H. Allen Orr wrote a 1998 paper that might speak to your question about power laws. He built on earlier theoretical work by Fisher and Kimura to come up with a model of the distribution of allelic effect sizes during bouts of adaptive evolution, finding that the phenotypic effect size of substitutions should follow a power law. He checked this against some computer simulations and found that the distribution did indeed look log-normal.
re: orr, i blogged an earlier paper of his with coyne which touched on the topic of mutations of large effect.
Also, distribution of effect size for skin color loci in humans looks log normal if I’m not mistaken.
Also, distribution of effect size for skin color loci in humans looks log normal if I’m not mistaken.
yes.
a few issues need to be fleshed out
1) dyson is right of course that the deviation in allele frequencies is highest in very small populations. stochastic forces get really strong at such small N’s.
2) that being said, what about the longer time period? fisher didn’t think it mattered much. i think he is as far as anyone is willing to go, but it isn’t as if small populations stay isolated forever. inbred groups are often reabsorbed by other groups.
3) there is the point about speciation. the idea that a population is isolated and changes certainly existed. but there are lots of debates about allopatric vs. sympatric speciation and the population sizes under which it occurs. very small populations are often subject to extinction.
4) species itself is a fuzzy concept. dyson seems to hew strong to the BSC, though i don’t think he’d term it that.
the bigger point that the ability to speak about a science is proportional to immersion within that science. dyson shouldn’t have popped off the way he did because he was making vast general and authoritative statements on very complicated issues.
Is there any possibility that Dyson is of the generation where Physicists knew that they were Gods and that Biologists were a pretty poor sample of mortals?
Is there any possibility that Dyson is of the generation where Physicists knew that they were Gods and that Biologists were a pretty poor sample of mortals?
this is certainly still true of the average physicist vs. the average biologist in terms of pure g. the typical biologist couldn’t think their way out of a box, let alone tell you what a hypercube is. but if dyson has that arrogance, it is subconscious. see here, and jump to ‘bad advice to a young scientist.’ he knows that biology has the ripe fruit right now.
a few issues need to be fleshed out
1) dyson is right of course
[you regularly speak ex cathedra on topics which you don’t know in great depth. you have a history of this in many fields (e.g., “iranians can’t be called aryans since aryans are nordic”); so you know why i come down hard on you. if it bothered you would have left long ago since i regularly delete your comments that i feel go over the line in terms of supportability -razib]
I’m sure I could straighten out Dyson in five minutes.
Also, keep in mind that Dyson is 83. He wasn’t at Los Alamos, but he worked on statistical modeling of RAF bomber runs over Germany, and he accompanied Richard Feynman on a cross country drive around 60 years ago. His most important contribution to physics came in the first half of the 20th Century. In other words, he deserves to have a little slack cut him.
“John Emerson
I hesitate to say anything because I’m not clear about the actual issues, but it seems that geniuses jumping from one discipline to another sometimes revolutionize the second discipline, and sometimes just look stupid. This can be true even when they jump from a harder discipline to a softer one.
I think that the movement of physicists and mathematicians into social science has had very mixed immediate results, though the long-term net result will certainly be positive since the negative results will wash out. (A second factor is that many guys jumped only after they found that they couldn’t make it in the tougher science.)”
Hm, that is interesting. Do you know of any books that document anything about this trend?
And is a lot of this research related to complex adaptive systems, the research we see at the Santa Fe Institute?
And is a lot of this research related to complex adaptive systems, the research we see at the Santa Fe Institute?
way before. paul samuelson was an acolyte of a man who was originally a physicist who consciously wanted to introduce those techniques into economics. so it goes back way before santa fe. and of course crick was a physicist. r.a. fisher had a degree in mathematics and did post graduate work in thermodynamics.
Just curious, what is the point of this blog entry? Physics/Math envy? Freeman was a top notch mathematician and world class physicist. He accomplished much more than providing Feynman’s amusement during a cross-country drive!!!!!!!!!!! There are some who speculate that the primary reason he was not awarded a Nobel prize was his personal conflicts w/ Oppenheimer, a very influential force on the Nobel committee–although his Freeman’s age is a valid excuse/consideration.
Just curious, what is the point of this blog entry?
geniuses can be stupid too.
razib,
I disagree.
Possibly I misunderstand but you seem to be saying that Dyson is
stupid because he came up with the wrong hypothesis when he posed
himself a question.
The problem with that idea is that I believe that any one of us
when faced with a new question is highly likely to come up with
the wrong answer.
Do you remember what you thought when you first thought about this
issue?
Is your thinking today really your own invention? Or could it be
that you read the work of others and were persuaded by their logic?
It might be argued that Dyson is stupid because he did not adequately
familiarize himself with the field before opening his mouth.
The problem with that reasoning is that it’s an absurdly non-trivial
task. It isn’t even well-defined what one should know. There’s also
an issue of different cognitive styles. One approach would be
to study and imitate what the experts say. Even if they conflict.
Another would be to try to first use one’s own head.
Even knowing this means one will be frequently, or even usually, wrong.
I’d guess that Dyson at 81 is not at all what he was at 25. But it
wouldn’t surprise me at all if the latter approach was his, i.e.
trying to solve the problem for himself first, at that earlier date.
Just curious, what is the point of this blog entry?
geniuses can be stupid too.
hey-when Heathen Imperialist used to post here, I thought that was one point of this blog!
“It isn’t even well-defined what one should know.”
Oh come on. There are fairly canonical texts that will give you what you need to know about population genetics, just like with physics. I have three on my shelf that would suit the purpose. (Only gotten around to reading one so far, of course…)
It might be argued that Dyson is stupid because he did not adequately familiarize himself with the field before opening his mouth.
No, it’s because he did not listen to corrections after he opened his mouth.
And besides, if his idea were true (everything depends on population size), ancient species should start evolving rapidly whenever their population sizes fall. Species age should have very little to do with it.