Sunday, December 19, 2004

Genes, Brains and Behaviour

In my recent post on Conscious Robots, CR commented:
[I]f there is a behaviour that is generally seen in most humans, then at some point in our evolution, that behaviour must have increased the survival chances of our genes.
(See also the email quoted previously.)

It seems to me that CR's comments are based on an impoverished understanding of evolutionary biology. It simply isn't true that every feature of an organism is an adaptation. Everything has evolved, for sure, but natural selection is not the only evolutionary force. Genetic drift, for example, can have very significant effects, especially on smaller populations.

Stephen Jay Gould has written extensively on the importance of evolved non-adaptive traits (what he calls 'spandrels'):
All organisms evolve as complex and interconnected wholes, not as loose alliances of separate parts, each independently optimized by natural selection. Any adaptive change must also generate, in addition, a set of spandrels, or nonadaptive byproducts. These spandrels may later be "co-opted" for a secondary use. But we would make an egregious logical error if we argued that these secondary uses explain the existence of a spandrel. I may realize someday that my favorite boomerang fits beautifully into the arched space of my living room spandrel, but you would think me pretty silly if I argued that the spandrel exists to house the boomerang. Similarly, snails build their shells by winding a tube around an axis of coiling. This geometric process leaves an empty cylindrical space, called an umbilicus, along the axis. A few species of snails use the umbilicus as a brooding chamber for storing eggs. But the umbilicus arose as a nonadaptive spandrel, not as an adaptation for reproduction. The overwhelming majority of snails do not use their umbilici for brooding, or for much of anything.
...
The human brain must be bursting with spandrels that are essential to human nature and vital to our self-understanding but that arose as nonadaptations [...] The brain did not enlarge by natural selection so that we would be able to read or write.
...
In summary, Darwin cut to the heart of nature by insisting so forcefully that "natural selection has been the main, but not the exclusive means of modification"--and that hard-line adaptationism could only represent a simplistic caricature and distortion of his theory.

CR continues:
The generally accepted belief is that we humans are somehow ‘more than evolution’: that something about our brains has allowed us to break free and start doing ‘what we like with our lives’. [...] We need to ask ourselves how we can be more than evolution. And we need to approach it from a scientific point of view, not from your personal experiences of life. That’s very important!

I don't know what 'more than evolution' is supposed to mean. Regardless, there's a methodological issue here, in that CR seems to be advocating dogmatic adherence to his favoured theory over the raw data (of personal experience). But theories are beholden to the data, not the other way around. If a theory doesn't fit the known facts, then so much the worse for that theory!

(Of course, there are times when science shows us that common-sense is spectacularly mistaken. I'm not denying that. But even (especially!) revolutionary theories must fit the known facts - it's just our conceptual interpretations of these that change.)

It's also important to note that psychology is the science of human behaviour. If we want to translate this field into the language of biology, we need to do so in a way that maintains the integrity of the primary scientific findings. If someone proposes a biological account of human psychology that is inconsistent with our best psychological science, then the attempted secondary account has clearly failed. We need a biological explanation that complements, rather than contradicts, the primary data of psychology.

As a striking example of this, when I pointed out that people can act from a variety of motivations, CR responded: "How could we? nothing has programmed us to act from any motivation other than the maximisation of our genetic survival chances." But this is foolish, because it's obvious - a known fact - that we really can be motivated by all sorts of different reasons. Any theory that denies this is simply wrong. Similarly, it is a known fact that the human brain is a very complicated organ, capable of (some degree of) logical thought and open-ended decision making. We can make bad decisions - bad for us, even bad for our genes. It happens, there's no denying any of this.

Of course, evolutionary theory doesn't actually deny any of this - it's CR's misinterpretations of the theory that cause these problems. But I still have to show that evolution really is consistent with the facts I mentioned above. As CR writes:
What you're suggesting is that natural selection can create a brain that is capable of doing something that wont enhance the survival chances of its genes. But you're going to have to explain how that's possible.

This is something that I began to touch on in the original post. As I wrote then:
I find it helps to think of genes as biological blueprints: instructions for how to build a body (and brain). However, those instructions give rise to extremely versatile structures, that can respond appropriately to a wide range of environmental stimuli. That is, we have a huge capacity for learning.

The key thing to note here is the value of flexibility. Our brains evolved to be capable of very general/abstract reasoning and open-ended decision making, presumably because this allows us to solve a broader range of problems - a crucial skill for organisms in our particular niche. (Not so for others, however!)

This is not to say that our genes aren't important, or don't influence us. Of course they do - they're the blueprints that describe how to build these wonderful structures in the first place! But one must realise how very indirect their influence is. So long as the resultant structure as a whole facilitates genetic success, the underlying genes will be passed on. This in no way requires that our every thought be explicitly directed towards achieving evolutionary success.

As an analogy, consider the paradox of hedonism: if you try to be happy, you're sure to fail. The goal of happiness is much better served by indirect means: you should pursue friendships and hobbies (etc.) for their own sake, and happiness will likely result.

Genetic success is kind of like that. Our genes build brains that have various goals, motivations, and modes of thought. These serve as indirect means of successfully spreading the genes, even though none of our cognition is explicitly aimed towards that goal. Overall, this 'strategy' works well, as the success of our species demonstrates. But it's not perfect: although our actions tend to correlate well with genetic success, there will be some times when they end up orthogonal or even in strict opposition to the evolutionary ideal. That's the cost our genes pay for an indirect approach, but it's well worth it - otherwise we wouldn't be here!

Update: The Ethical Werewolf has more:
There's lots of things in human psychology for which an adaptationist explanation makes some sense... The capacity for laughter and the appreciation of music, however, seem more difficult for adaptationists to explain... My guess is that if you throw together a bunch of complicated, evolved modules in the brain -- language processing, emotions, auditory processing -- you get weird emergent capacities on top of those things that weren't specifically selected for.

5 comments:

  1. The key point of this is comment is going to be ‘motivation’.

    Let’s be careful about the use of spandrels. Three important rules: Spandrels can’t be expensive, and they can’t be complex. Third rule concerns how spandrels are used (motivation) – see below.

    Let’s look at the snail spandrel briefly: firstly, it’s not a cost – it’s a free by-product, as Gould is careful to point out. Secondly, it’s not a complex variation – you couldn’t suggest that the peacock’s tail is a spandrel, just as you couldn’t suggest that the human eye is a spandrel – complex variations have to arise by gradual steps, each of which confers a selection advantage. If the snail carried on its back a large flag that not only took many steps to create, and was also a cost, we would know that it couldn't be a spandrel, we’d know that it must confer a selection advantage.

    So, let’s apply these ‘spandrel rules’ to human behaviour.

    Reading and writing clearly couldn’t have arisen by natural selection. I’d add ‘use of right foot to press brake pedal’. Clearly, our feet didn’t evolve for that purpose. But equally clearly, the foot isn’t capable of performing any tasks that weren’t an advantage for walking/running/climbing. Car makers develop brake pedals to fit with naturally-selected foot abilities, not vice versa.

    So, now to writing and reading. It’s easy to think that somehow our brains have ‘moved beyond’ natural selection because we can see that ‘reading and writing’ are complex tasks and could have had no possible use on the plains of Africa while we were doing most of our evolving.
    But we do know that we must apply the rules of spandrels. The ability to read and write must have been a free by-product of naturally-selected abilities. It could not have arisen de-novo in the same way that complex abilities such as sight or hearing arose.

    So we have a naturally-selected brain activity that has been co-opted for secondary use – that secondary use being reading and writing.

    Question is – why has it been co-opted? Because this is the crucial point. Why are we motivated to do anything – why are we motivated to read and write? Let’s look at the humble snail. Some snails, Gould tells us, use their free chamber to store eggs. Note the motivation. It’s to maximise the survival chances of their genes. If they used the umbilicus to store deadly bacteria, they’d have died out. If they used the umbilicus to store tiny pretty pebbles that they found in their travels, they’d be wasting a lot of time and energy that they could be using to maximise their genes – and again, they’d have died out. But hold on! We’re talking about snails here. Why would snails be motivated to store pretty pebbles? They wouldn’t because (critical point coming up) such a motivation would itself be a complex variation. Can I underline those previous nine words?

    Doing something that doesn’t help you maximise the survival chances of your genes is a complex variation. And such a complex variation has to arise gradually, and each step has to confer a survival advantage. Motivations to do things that don’t maximise our survival chances are not free by-products. They are not spandrels. The ability to read and write are spandrels but the use of those abilities cannot be.

    So let’s look to see whether reading and writing improve the survival chances of our genes. Well, reading and writing allow the spread of knowledge accurately allowing us more effectively to stand on the shoulders of giants. Without reading and writing, would we have conquered smallpox? Would we have invented the pesticides required to feed our enormous population?

    But, you might say, how can reading the gossip columns or writing blogs confer a selective advantage? And immediately, you’re cheating. We’ve just shown that they must – a non-selective-helping-motivation is a complex variation that could not arise…

    Which brings us to:


    “It's obvious - a known fact - that we really can be motivated by all sorts of different reasons.”

    I refer to the problem of how motivations could have arisen that were not ‘adaptations’. Gould is the patron saint of spandrels – and uses non-human animals to justify the existence of spandrels. But I don’t believe he has shown the existence of a spandrel that is actually used by an organism in a way that doesn’t confer a selective advantage. As I pointed out above, the umbilicus of the snail is a free by-product, but it is only used in way that maximises the survival chances of the genes.

    You’re claiming that humans do use these spandrels for reasons that don’t help their genes. But you haven’t shown how such motivations should arise. All you say is that there are a variety of motivations. I would simply say that all those variety of motivations are helping us maximise our genes. But I can’t prove it, just as you can’t prove that it isn’t.

    I can’t prove that when you write your blog you’re simply trying to make yourself more attractive to the opposite sex or more attractive to future employers who will pay you large sums of money which will make you and your family safer and your children more sexually attractive. I can’t prove that’s your motivation. But equally, you can’t prove any other motivation – (although I’d be interested to hear what you consider your motivation to be, anyway – or any other ‘motivated by all sorts of reasons’ examples that you have)
    So given that neither of us can prove a motivation, we need to turn to our understanding of evolution to arbitrate. Psychology can’t help us, because it can’t prove motivation, only observe actions. Only the mechanism of creation can help us here.

    I repeat – a motivation that isn’t genetically motivated is a complex variation that could not have arisen by the mechanism of evolution. So what’s your alternate theory to evolution?

    “This in no way requires that our every thought be explicitly directed towards achieving evolutionary success.”

    You comment that the brain is flexible. That’s fine. So’s my foot when it presses the brake pedal.
    But you haven’t touched on motivation anywhere. So I think that my above suggestion of the impossibility of ‘other motivations’ is valid here as well.

    “Our genes build brains that have various goals, motivations, and modes of thought. These serve as indirect means of successfully spreading the genes, even though none of our cognition is explicitly aimed towards that goal.”

    “None of our cognition is explicitly aimed towards that goal:” do you mean 'none of our action'? If so how does that differ from saying that we spend all day knitting a jumper and end up having built a house? I don’t accept it. If I run a business that is aimed at making a profit, then I have to spend all day doing things that will end up making a profit. Although it might look like I’m making a loss when I refund a customer, or spend time cleaning the toilet, it all adds up to the whole job of making my business a better place to buy from. But if I spend all day cleaning my neighbour’s toilet, my own business is going to suffer.
    If you mean, we're not consciously aware that what we're doing is going to help spread our genes, then I agree. But the way we're motivated (via feelings) makes this unnecessary. 

    Posted by consciousrobot

    ReplyDelete
  2. Since your here CR I will have you on - not that I entirely disagree with your argument.

    > Let’s look at the snail spandrel briefly: firstly, it’s not a cost – it’s a free by-product, as Gould is careful to point out.

    It may be free in an absolute sense but not in an opportunity cost sense - it is likely the adition of another gene could hve neutralized it or a slightly different gene might have removed it. But othere genes are added that are dependant on it and so it becomes fixed (ie if it changes the othere genes fail).

    > Secondly, it’s not a complex variation – you couldn’t suggest that the peacock’s tail is a spandrel

    In the end most complex forms are initially hte result of a spandrel. the reason for htat is that somthign that is benificial in its final form (eg the human eye to use an often quoted example) may well be disadvantageous in its earlier forms. It is difficult for natural selection to follow it all the way through BUT if natural selection has added a big disadvantage and is slowly covering up for it that gives much more latitude for experimentation.

    > complex variations have to arise by gradual steps, each of which confers a selection advantage.

    As noted above I think this is seriously flawed. The odds of a complex variation to at every step involce a selection advantage is miniscule. In fact natural selection should lead to "dead ends" for example with a bacteria that is perfectly evolved for breeding and surviving - why would it evolve into a "higher life form"? every step it took in that direction would make it less able to compete against its ancestors. the only reason is via a massive drop in the competitive pressures (like the birds on islands) OR spandrels

    > If the snail carried on its back a large flag that not only took many steps to create, and was also a cost, we would know that it couldn't be a spandrel, we’d know that it must confer a selection advantage.

    OK imagine step one was a big colourful eat me sign and it required only one gene to make it (by some fluke) that gene also gave immunity to herbacide. They probably start to use the sign to identify other members of their species for breeding. In the sprayed patch of lawn the snails start to do well.. but they keep getting eaten by birds - so slowly they cover the "eat me" sign with a more complex flag never becoming better at surviving than the other snails who found other ways around the herbacide but still surviving.

    > They wouldn’t because (critical point coming up) such a motivation would itself be a complex variation. Can I underline those previous nine words?

    Your assumption is a high level of competitive forces. Actually the level of competition fluctuates - in humans right now for example the competition level is very low. A snail in a similar sitaution could develop as odd a mixture of genes as he likes.

    I also note there is a second way a variation can hitchike on a useful gene. imagine lets say that blue eyes are very desirable and green eyes are a very simple and easy mutation from blue eyes. Evolution might select for blue eyes and then drift into green eyes.

    > But I don’t believe he has shown the existence of a spandrel that is actually used by an organism in a way that doesn’t confer a selective advantage.

    1) now we have an issue with "what is a selection advantage" for example one might claim that bacteria are the most sucessful organism and that more complex animals are just degenerations of that early and advanced form (for survival and breeding). Considering that it cant all be natural selection glory between us and them.

    2) The second problem is "once you have a spandrel why not use it?" Ie a snail is unlikely to use a spandrel i na way designed to reduce its survival potentail it will build survival enhancing aspects onto it bringing us to (3)

    3) are you using evolutionary arguments to explain why you think things are survival enhancing? for example do you say a snail has a shell the shell must increace survival how could that be? it makes it harder to eat... therefore it is survival enhancing.
    BUT what if you are fully aware that slugs survive better than snails in most environments? (this may or may no be the case - but htee would be lots of examples) your assumption lead to your conclusion so it is logically flawed. 

    Posted by GeniusNZ

    ReplyDelete
  3. take for example a poisonous frog. Being poisonous has almost no benefits to the individual survival because to find out you are poisonous they have to eat you first. Not only that but if oyu are a frog then lots of things will eat you - you need to teach a reasonable proportion of them that you are poisonous (maybe by natural selection).
    It will take some time to do this particularly since odds are you look almost identical to your ancestors without the poison (who can then freeload on your reputation).
    During that time UNLESS there were very low selection pressures or the poison was beneficial in another way it would make sense that it would be dropped as a strategy.

    The adaptation in the long run may continue to be detrimental to individual survival while helping species survival - o it may even continue to be detrimental and just due to low selection pressures anda bit of luck then survive, possibly a better gene will develop in that population and after many generations other genes built upon developing htat sort of frog ma turn it back into a superior survivor (or at least one with a niche). 

    Posted by geniusNZ

    ReplyDelete
  4. Natural selection, genetic drift and spandrels:

    the question is - which is the most likely explanation for human behaviour - specifically for human behaviour that exists in the majority of humans. For the behaviour to be in most humans, the most likely explanation is that it's been there for thousands of years - from a common ancestor. Otherwise, it's got to have arisen by chance more than once, and it's alternative selected out.

    If it's neutral or negative for selection (as it must be if it's created by genetic drift or spandrel) then it's been an opportunity cost at very minimum for thousands of years. Unlikely. Unlikely that life has been so easy for humans for so long. More likely that it's a selective advantage.

    You can have spandrel and drift as a 'possible' explanation, but only when 'natural selection' simply doesn't work as an explanation. Our assumption has to be 'natural selection' for the majority of human behaviour.



    Poisonous frog: evolution is about gene survival not individual or species survival. If I'm a frog and a single mutation makes me poisonous and then I have 500 children, half of which are poisonous, then it's going to be very good for the gene that made me poisonous. 

    Posted by consciousrobot

    ReplyDelete
  5. "But I don’t believe he has shown the existence of a spandrel that is actually used by an organism in a way that doesn’t confer a selective advantage."

    What do you mean by 'used'? The whole point of spandrels is that they arose non-adaptively. The fact that some are later co-opted for adaptive purposes does not change the fact that the organism originally contained non-adaptive features.

    The upshot is this: Upon seeing some characteristic, one cannot automatically assume that it is an adaptation. It's possible (and indeed fairly common) for non-adaptive byproducts to also evolve.

    Do you really think that susceptibility to cancer, alzheimers, alcoholism, etc. are all adaptations? :p 

    Posted by Richard

    ReplyDelete

Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)

Note: only a member of this blog may post a comment.