December 7th, 2008
|08:34 am - Oh Superman|
Here's a piece of research which seems as usual to have an obvious flaw.
(Study of war veterans)... those who performed better on intelligence tests tended to have more - and more mobile - sperm.... people with robust genes might be blessed with a biological "fitness factor" making them fit, healthy and smart. Lead researcher Dr Rosalind Arden said: "...our results do support the theoretically important 'fitness factor' idea. We look forward to seeing if the results can be replicated in other data sets, with other measures of intelligence and other measures of physical health that are also strongly related to evolutionary fitness."
But the question about intelligent people isn't why they are so reproductively fit: it's not why there are so many, but why there are so few. Cheetahs are specialised for running, and all cheetahs run very fast. The fast cheetahs have greater 'evolutionary fitness' so there aren't any slow ones any more. That is manifestly not the case with human intelligence. If intelligent men have such a reproductive advantage over unintelligent men - why are they so rare in human populations?
My guess is that sperm quality and brain development are both very dependent on maternal health and childhood nutrition, rather than being joint markers of genetic superfitness. There may also be a link between nutrition and more educated mothers etc. However, there are obvious political and social reasons why people don't want to think about that, and prefer to say there is a class of super-fit spermaniacs.
(To be fair the article quotes a guy at Sheffield university who says "The fact that it's possible to detect a statistical relationship between intelligence and semen quality in adult men probably says more about the co-development of brain and testicles when the man was in his mother's womb" Well, yeah.)
|Date:||December 7th, 2008 09:08 am (UTC)|| |
There is also an issue that in humans (who knows about other species?) we seem to breed towards a norm: very stupid parents frequently have quite smart children; very smart parents have children who are "averagely bright". There is no "onward and upward", or downward spiral towards degeneracy.
That's not an explanation, it's a mathematical feature. If a population has an average, then if you take samples, they are more likely to be found near to the average. But it doesn't explain why the average is where it is.
|Date:||December 7th, 2008 09:16 am (UTC)|| |
I didn't say it was an explanation, I said it was an "issue".
Some people talk as if averages pull populations towards them - the causality the wrong way round. I know you weren't saying that.
It's a classic case of mistaking correlation for causality.
Yes. I expect the silliest reports (and I'm expecting this in the British press) will be that intelligence causes more sperm: 'do sudoku and get more sex-power', and then the next silliest will be that sperm and intelligence are both 'caused by' some hypothetical 'fitness factor', which somehow exists in the platonic realm.
|Date:||December 7th, 2008 10:47 am (UTC)|| |
No, no, no - clearly more sperm cause greater intelligence. That's why there are so few women in top jobs.
|Date:||December 7th, 2008 11:04 am (UTC)|| |
But of course!
Oooh, a magical 'fitness factor'. Yes, I'm sure it's that and not one of many scientifically plausible reasons for the data looking like this.
Boffins here in Bristol have studied thousands from before birth to early adulthood, and things rarely seem so clear cut. For example: differences in IQ that look like a straightforward, physical effect in the womb, become much less clear when they turn out to be associated with fathers' habits just as strongly as with pregnant mothers themselves. A recently published analysis found this to be the case with moderate use of alcohol and tobacco.
I also think that people show more or less intelligence according to whether it is socially safe and appropriate for them, and if they feel confident, if they feel they have to take responsibility as opposed to being passive, and all of this varying between different circumstances
Some cheetahs are faster than others, but cheetahs as a species are faster than anything else around. Similarly, some humans are smarter than others, but humans as a species are smarter than any other animal.
(Of course "intelligence" is a fuzzier concept than speed, and not so straighforward to measure.)
Not really, Cheetahs are all pushed up one end, they don't have a bell curve. Also, the limiting factor on cheetah speed is physical - the rapid heating of the muscles. They are pushed right up against the physcial limits of what's possible.
Some animal features are distributed in bell curves, and that's when there's a trade off. For instance large bulls (or walruses or whatever) will have an advantage in a fight, but they need more food so they die off in famines. In such a case there is more than one factor - one limiting the other (or a complex of factors, even more likely).
If a factor is a simple matter of inheritance, and an unequivocal advantage, then it will become widespread. We are used to imagining that an IQ of 100 is the norm - but why should it be? If, as the first researcher claims, what we call 'High IQ' is a simple marker for superior reproductive fitness, then why isn't it common? Why is it abnormal if it is simply a matter of inheritance, and an unmixed advantage?
My argument of course is that it is not simply a matter of inheritance.
Edited at 2008-12-07 02:03 pm (UTC)
Such data as I've been able to uncover suggests that cheetah top speeds are approximately normally distributed - i.e. a bell curve - with a mean around 60 mph and a standard deviation around 3 mph.
Tradeoffs have nothing to do with bell curves. The reason why bell curves are so often seen in random processes in nature is to do with the Central Limit Theorem.
There are doubtless limiting factors for intelligence just as for speed. We don't know what they all are, by any means, but among ape species the level of intelligence is correlated to the ration of brain mass to total body mass. This introduces a limiting factor in that a larger head increases the risk of fatal complications in childbirth. This is why human newborns are at an earlier stage of development than newborns in other ape species.
An IQ of 100 is the norm by definition: IQ test results are normalised to a gaussian distribution with a mean of 100 and a standard deviation of 15, if I recall correctly.
No, you misunderstand the point I'm making, which isn't about the theory of statistics. Sampled points cluster around a mean, in a regular way, because the mean is an emergent property of the population. The population isn't forced to vary in a particular way by the statistics.
The issue in this case is why does a species have a particular average size, or average number of offspring, or average antler size. It's that which I am talking about. That is what is determined by multiple factors. Obviously not the stats, but what the stats describe.
The point is - why is that the average. Why does (say) the size of warthogs vary around a particular point? Because there is an optimum size to be a warthog. Why does it vary a lot or a little around that point? Because the warthog population is living in a variable environment. If a bunch of warthogs got stranded on an island the average would probably shift down, because the optimum would change.
Now, it clearly isn't the case that IQ 140 is rare because people with IQ 140 have bigger heads. So - why is it rare? Saying 'because it's far from the mean' isn't a reason. It's just a different way of describing the same thing.
Are you talking about IQ, or intelligence? The former is a very imperfect proxy for the latter.
As for statistical distributions, well, you were the one who started going on about bell curves. The central limit theorem is really interesting. Roughly speaking, it says that if a quatity is an average of a whole lot of randomly distributed quantities, then it will be normally distributed regardless of how the quantities that make it up are distributed. (There are technicalitites, but that's the gist.) This is when normal distributions - "bell curves", if you must, though I hate that term - are so prevalent in nature.
On the issue of why humans exhibit a certain level of intelligence, there are two questions: why is it so high, and why isn't it higher?
Why is it so high? No none knows, although these kinds of freakishly exaggerated characteristics are often the product of biological arms races. In this case, we might conjecture that, as humans are social animals in which inclusive fitness is related to success in a complex social environment, intelligence - particularly in communication, strategy, and machiavellianism - will be selected for in the positive feedback loop characteristic of an arms race as success is relative to the rest of the society.
Why isn't it higher? Again, no one knows, but here are some ideas:
1. It is actually getting higher on evolutionary timescales, we just haven't noticed because we haven't been studying it that long.
2. There are tradeoffs between intelligence and other essential biological functions, so that increasing intelligence beyond a certain level hits diminishing, then negative, returns.
3. There is some physical limiting factor that cannot be overcome without a step change in development, and that step change hasn't happened yet. This may be something quite crude, like the infant head size issue that limits the intelligence of other apes, or it might relate to the fundamental engineering of the brain.
The bottom line is, the biological limiting factors are simply not well known. It's a difficult and complex topic. Research into questions like whether there is a statistically significant link between intelligence (however measured) and sperm quality is about figuring out tiny little bits of the jigsaw. In time, hopefully, a fuller picture will emerge.
I'm taking about intelligence and using IQ for a shorthand to avoid these comments getting incredibly long, same goes for lots of other verbal shorthand I'm using.
there are two questions: why is it so high, and why isn't it higher?
Yes, and the latter is what I'm talking about. The final part of your comment, with the three hypotheses, is the thing, specifically the second hypothesis about biological trade-offs. If (for instance) 'there are trade-offs between intelligence and other essential biological functions', then intelligence is not a 'sign of the underlying fitness factor'. Instead it is one of several competing factors which add up to a range of fitness strategies.
On your third hypothesis I sometimes wonder if there's an association between something that causes high intelligence and something that causes neural problems, but it's just a guess. In any case, none of these guesses detract from the point which I was making - that 'being very intelligent' isn't a simple fitness optimum, like immunity to smallpox would be.
Not really, Cheetahs are all pushed up one end, they don't have a bell curve.
Cheetahs may be a bad example, since isn't it the case that they have very little genetic variation? Didn't someone hypothesize that cheetahs nearly became extinct and that the current cheetah population are descended from a small initial population? So talking about the lack of variation in Cheetahs may be not relevant to this discussion, since there are different factors involved with Cheetahs than humans.
Well, that's a fair comment that cheetahs are a genetically anomalous species, because of the genetic bottleneck. However, there's no need for any species to show wide variation in phenotype.