In the Roger Scruton on altruism thread, some commenters have expressed confusion over the evolutionary explanation of altruism in ants. If workers and soldiers leave no offspring, then how does their altruistic behavior get selected for?
The answer is simple but somewhat counterintuitive. The genes for altruistic behavior are present in both the workers/soldiers and in their parents. Self-sacrificing behavior in the workers and soldiers is bad for their copies of these genes, but it promotes the survival and proliferation of the copies contained in the queen and in her store of sperm. As long as there is a net reproductive benefit to the genes, such altruistic behaviors can be maintained in the population.
Selfish genes, altruistic individuals.
Let’s dedicate this thread to a discussion of other counterintuitive evolutionary truths. Here are some of my favorites:
1. The classic example of sickle-cell trait in humans. Why is a disease-causing mutation maintained in a human population? Shouldn’t selection eliminate the mutants? Not in this case, because only the unfortunate folks who have two copies of the allele get the disease. People with one copy of the allele don’t get the disease, but they do receive a benefit: improved resistance to malaria. In effect, the people with the disease are paying for the improved health of the people with only one copy of the mutation.
(Kinda makes you wonder why the Designer did it that way, doesn’t it?)
2. In utero cannibalism in sharks:
Shark embryos cannibalize their littermates in the womb, with the largest embryo eating all but one of its siblings.
Now, researchers know why: It’s part of a struggle for paternity in utero, where babies of different fathers compete to be born.
The researchers, who detailed their findings today (April 30) in the journal Biology Letters, analyzed shark embryos found in sand tiger sharks (Carcharias taurus) at various stages of gestation and found that the later in pregnancy, the more likely the remaining shark embryos had just one father.
(Kinda makes you wonder why the Designer did it that way, doesn’t it?)
3. Genetic conflict between parents and offspring. Here’s a great example from a 1993 paper by David Haig:
Pregnancy has commonly been viewed as a cooperative interaction between a mother and her fetus. The effects of natural selection on genes expressed in fetuses, however, may be opposed by the effects of natural selection on genes expressed in mothers. In this sense, a genetic conflict can be said to exist between maternal and fetal genes. Fetal genes will be selected to increase the transfer of nutrients to their fetus, and maternal genes will be selected to limit transfers in excess of some maternal optimum. Thus a process of evolutionary escalation is predicted in which fetal actions are opposed by maternal countermeasures. The phenomenon of genomic imprinting means that a similar conflict exists within fetal cells between genes that are expressed when maternally derived, and genes that are expressed when paternally derived.
(Kinda makes you wonder why the Designer did it that way, doesn’t it?)
Can readers think of other counterintuitive evolutionary truths?
Addendum
4. Mutant organism loses its innate capacity to reproduce and becomes a great evolutionary success. Can anyone guess which organism(s) I’m thinking of?
keiths,
Genomes in Conflict by Burt and Trivers. Very comprehensive.
It’s bad for their specific copies, but good for the portion of genes shared by the queen. It’s a dynamic trade off balancing many different factors; energy lost for productive work for the colony, chance of being caught, number of queens, number of mates the queen takes, external threats, etc.
It’s amazing the collective still exists. The young don’t have the commitment of their elders, not even a proper respect for the Queen. (May She be fertile forever.) Those darn liberals!
I think I understand DNA_Jock’s position now: he doesn’t see reproductive altruism as a type of behaviour, and any additional altruistic behaviour doesn’t affect the reproductive odds for the physical copies of genes in the bodies of sterile workers. Therefore, according to his thinking, the statement that ‘altruistic behaviour is bad for the copies of genes in sterile workers’ is false. According to his definitions, he’s right. It’s irrelevant to the models under discussion, but still right, according to his definitions.
Well, they shouldn’t. They’re using her as a breeding factory, after all. Can’t be fun, being pregnant all the time.
Gralgrathor,
Yes, that was my only point. Reproductive altruism is very interesting, but it was not the form of altruism discussed in the OP, which came from Roger Alton’s statement
He was writing about the kind of altruism that gets people all teary-eyed. Which, unfortunately 😉 , does not include reproductive altruism.
I was trying to make an ironic point to keiths about the difficulty of admitting an error, no more. I may have broken my own irony meter 🙂 .
Heresy! Heresy! Sisters! Sisters! Bring out the Queen’s guards!!
The way I read Scruton’s quote, he’s offsetting ‘moral altruism’, genuine informed and intentional self-sacrifice by humans, against biological altruism. He starts off with the definition of biological altruism, for comparison, then switches to the other kind. His ‘moral argument’ has little to do with the altruistic behaviour described in the first few lines, and little to do with the discussion we’re having here – which is about biological altruism.
By the way, I think Scruton is wrong: I think there are genetic factors in play in moral altruism; genetics can very well explain why some people have a greater tendency to display altruistic behaviour than others. In humans, genetics necessarily is only part of the explanation, but it’s still a part.
keiths,
Hahahah!
Gralgrathor,
Gralgrathor, you write, By the way, I think Scruton is wrong: I think there are genetic factors in play in moral altruism; genetics can very well explain why some people have a greater tendency to display altruistic behaviour than others. In humans, genetics necessarily is only part of the explanation, but it’s still a part.
Is there something in Scruton (whom I’ve not read much of) that suggests that he would deny that genetics plays some part in moral altruism, i.e., that some people have a greater tendency to it than others? (Incidentally, if he does say that, I agree with you that he’s wrong. I’m just surprised that he would–although, again, I don’t know his views well.)
Joe Felsenstein,
Joe, I think that phoodoo is helped in muddying the water by some of the silly and irrelevant quibbling that’s been going on in this thread. There’s a moral here which I probably don’t need to make explicit.
There’s an additional point to be taken from phoodoo’s diatribe. If the term “genetic explanation” is made too broad or is not properly defined at all, the bar will be either raised too high for scientists or a contradictory story will be required of them. Obviously, that situation can only help the anti-science crowd.
No, he conveniently states his arguments in the form of hypotheticals (‘if…’). Nevertheless, and without having actually read Scruton’s book, that’s what I’m getting from the part quoted by Keith. The quote ends with the sentiment that “evolutionary explanations reduce to triviality, when the thing to be explained contains its own principles of persuasion”, which I read as an argument for explaining human altruism entirely in moral terms.
But of course I may be wrong about that too. Not interested in Scruton enough to buy his book, when there’s all this fascinating biology to read about.
How did his comment (paraphrased as: ‘you people are confused about some aspects of selfish gene theory so evolution must be wrong’) achieve anything? How could one do anything but ignore it as the irrelevance it is?
DNA_Jock,
Now that is indeed ironic!
Gralgrathor, I agree that that remark of phoodoo’s was pointless. Much of his post was. Just not all of it. There’s a moral there too!
That’s how ridiculous your whole premise is Joe, you actually think a computer simulation can somehow demonstrate a random Darwinian process for developing altruism. There are 1000 reasons why a computer program could never show this. You put into any simulation program exactly what you need to come out. Sometime try programming a computer to just copy a number a thousand times, and sometimes let it make a mistake, and see how long it takes to actually come up with something meaningful. See if it can build a small hut for example.
Sorry to be offensive Joe, but if this is the level of University thinking in America, its no wonder America is becoming so stupid.
You don’t have the faintest idea how the social structure of ants came to be, little yet, the mechanisms which worked to build this organism and make it do what it does. You shroud the problem in complete bullshit claims to bogus computer simulations, and then claim I muddy the water.
In the meantime,, China has just surpassed America as being more advanced in scientific research. And you are part of the reason Joe.
phoodoo,
Can we anticipate a busy afternoon ahead listing them?
Indeed, ‘doo; models and simulations are good for nothing, that’s why nobody ever uses them.
You mean because one reason why a computer simulation could never show this would never be enough to shake you from your faith?
I don’t know much about evolutionary theory, and even less about how to do computer simulations, but, FWIW, about a hundred years ago, when I was reading a bunch of Samuel Butler’s Lamarckian criticisms of Darwin, I made a crude spreadsheet using random number generation to simulate the random variation of mutations, and was kind of surprised actually, that the randomness proposal of Darwin (whose posse was incredibly mean to Butler, incidentally, and who did make unacknowledged use of some Lamarckianism that found its way into some of his writings) held up perfectly fine against Butler’s complaints when put to a mathematics test–at least the baby test to which I was competent to make for it.
phoodoo,
You claimed 1000. One will do for starters. Something other than your opinion.
well, just don’t tell Joe; his livelihood depends on it.
Can’t edit my post for some reason, but I wanted to fix “mathematic” and “who’s” (and all the other typos in there that I haven’t noticed).
I just gave you one Allan, how could you have missed it? Any computer program you write requires that you program in the result you are looking for, before you start. But evolution (supposedly) has no purpose, not even to survive, so without pre-programing the need to reach something, you program falls apart. It is the crux of every program there is, rubbish in means rubbish out. You tell it what to keep and what not to, and it simply listens to what you say. the opposite of what you claim evolution does-which is that it never knows what to keep and what to discard, and what the goal is, it just screws up until it gets lucky.
walto,
If there is even a hint of Lamarckism in evolution, your whole theory goes out the window. And there is.
False. If you already had the result, then the programme wouldn’t be required. Especially for complex models involving genetic and neural algorithms, computer simulations can be used to obtain results that cannot be easily found without a great many calculations.
Where?
phoodoo,
You are wrong, phoodoo. The purpose of an evolutionary algorithm is to explore a particular fitness function. You have to give it something to chew on – there’s no point writing a program with no purpose whatsoever. But the fitness function merely generates a survival differential among the strings thrown up during the run of the program (they are DATA). Which is exactly the same as Natural Selection.
You think all professors are thick?
Wrong. Yesterday I delivered a updated simulation to a client that predicts the future behavior of a specific market. Some of the results are surprising…
Suppose I wrote a computer model of a sieve. I supply the mesh size, and randomly generate 3D particles. Each is evaluated according to the mesh and placed in one of two piles. Is that an invalid model of ‘real’ sieving? If so, why? If not, why’s it different from any (other) fitness function? The extra layer in an evolutionary version is that surviving particles reproduce with variation. This is of course a very simple fitness function with no inheritance and you wouldn’t actually bother. But it illustrates the principle – you are simulating a sieving process, which is what selection is.
If evolution has no purpose, why would you need to write a purpose into the program? Why not make one that supposedly does exactly what you claim evolution does?
I know why, because it would fail every time.
Nothing about evolution is like sieving. So your premise is flawed from the start. Sieving does absolutely nothing to add something to the mix that didn’t already exist.
You’re not thinking, ‘doo. Evolution has no purpose – but our research does. The purpose of our research is to better understand evolution (or the effects of particular fitness landscapes). That’s the purpose written into the program.
False. Natural selection is like a sieve – one in which the size and shape of the holes in the mesh is determined by environmental factors.
Right, and in your program you tell it what a market is, you tell it what conditions to accept or reject, you give it values, you do everything that is the opposite of random failures adding up to organized functioning.
Gralgrathor said:
Unless you know what you are looking for in terms of a set of parameters, you cannot create a program that is effective in acquiring your goal even through “random” variants of configurations. For instance, a computer program can be created to find the best configuration of a certain pool of resources that will result in the most aerodynamic and effective wing design for a certain kind of airplane. However, you have to know what you are looking for, in terms of those parameters, in order to hope to construct an algorithm or program to find any specific successful configuration.
If there is no purpose or goal provided, there is no way to write a program to achieve anything. If the program code itself is randomly generated and changed without any limiting or constructive reference whatsoever to any purpose or goal, you’ll never have a functioning program that does anything beyond the trivial in the first place.
Oracle information is in every so-called evolutionary computer simulation that acquires significant targets because it is required to make any program or algorithm whatsoever work.
That is the simplicity of your thinking, and why you are so quick to accept something you have no evidence for. You think if you just keep discarding things you don’t want, you will end up with something useful. You never will.
The useful things need to exist first, before you can sieve for them.
walto,
His views on that are very odd and confused. From the passage I quoted in the other OP:
Gralgrathor,
That doesn’t answer why you can’t write a program that has absolutely no purpose at all, and have it come up with something meaningful.
You can set the parameters for a particular fitness landscape, but that does not mean that you’ll know the outcome of the simulation.
No, it’s merely an analogy. I said that natural selection is, in specific ways, like a sieve, and that therefore the concept of the sieve may be used as an analogy to illustrate a specific aspect. I did not say that natural selection is identical to a sieve.
It seems like your insisting that analogies be exactly like the thing itself in everyway is somewhat simplistic.
Do you read your own comments before posting them?
Its not a question of knowing the outcome, its a question of knowing what an outcome is. In your little theory, there is no identification of what an outcome could or should be, Without knowing what an outcome should look like, the program can do nothing.
In evolution, the outcomes that count are any outcomes that show up. There is no such thing as a good or bad outcome, simply that it is outcome is all that matters. No computer program in the world can do that, and have it mean anything.
I am not the first person to point out the logical problem that you can never overcome. The problem exist, but every week a new evolutionist comes along and thinks the problem is just a trivial distraction. Same old story.
But the problem with the simulation of such a “sieving process” is that it is not programmed to be blind at all, but rather programmed to provide a particular sieving process that serves the target the programmer is attempting to acquire.
IOW, in the real world, the sieving process is not trying to acquire a configuration of materials that will be effective at staying out of reach of predators by flying and building nests and living in trees. The sieving process in the real world is blind; the sieving process in the lab or in a computer simulation has a goal and is configured to have at least a chance (if not the best chance) at acquiring it.
For example, Liz’s natural selection algorithm took the best performers in terms of the stated goal and removed the worst, which assumes that the sieve is set to acquire the specific goal. One is assuming that the real-world sieve happens to be configured in a way that would happen to acquire the specific target the programmer designs his/her program to acquire.
One is simply assuming that the physical environment can happen to provide a sieving process that, over millions of years, is capable of acquiring such targets. One is not demonstrating that it can; one is only assuming that it can. IOW, such evolutionary simulations simply assume that blind natural selection can do the same thing that otherwise is known to require intentional programming with a goal or purpose.
But a computer program is useless as an analogy. Its supposed to be a simulation, not an analogy. Analogies are precepts of the mind (anyone can believe their own analogy is convincing to them) , simulations are supposed to be recreating what you want to explain.
So if you want to explain evolution, try to make a computer that does what you believe evolution does-make bad copying mistakes, and see if it can turns into something you haven’t pre-programmed it to become.
That too is false. Many such simulations have arbitrary halting conditions. For instance, you could tell the simulation to stop at 20.000 generations, so that the state may be examined. The halting condition itself has nothing to do with the results expected or produced.
You were talking about the sieve as an analogy for natural selection, not about computer programs simulating natural selection.
Are you intentionally trying to confuse issues?
phoodoo,
That’s precisely what people do, and it doesn’t fail every time. F’God’s sake phoodoo – do you habitually turn up at people’s workplaces and tell them they’re doing it all wrong?
Of course, a computer program does not make a real eye. Is that the problem?
As WJM understands, but you apparently do not, the simulation gets some parameters – stuff that we do know about. It gets lots of stochastic rules. After millions of stochastic failures and successes, it reports results. Which are often surprising and counter-intuitive. The client pays gobs of money for this tool, which would be rather stupid of them if they were only getting results they already knew…
is hilarious. The program has no idea “what a market is”.
But thank you for telling me how my model works.
phoodoo, you are doing it wrong.
Darwinian evolution (of which evolutionary algorithms are an example) is not just natural selection. It is cumulative natural selection. Not only does a program discard things it doesn’t like; it keeps things it does like.
What you have described in that paragraph is a random search. You pick something, it doesn’t quite work, throw it away, begin the search anew. Of course that way it will take an eternity to find something useful.
But that’s not how an evolutionary algorithm works. It makes a small change to a genome; if that change makes things worse, the change is discarded; if it improves things, it is kept. As time goes on, digital organisms get better. If the fitness landscape is sufficiently smooth, this method quickly takes you to a fitness peak.
William J. Murray,
Congratulations, you have noted that practical applications of evolutionary computing serve a practical purpose. This has no relevance to the claim that they are doing exactly what evolution does, sifting random variation in a differential manner. If there is variation, and a consistent differential in success, the favoured variant will increase and the less favoured will decrease. In the lab, in the wild, in the computer, even in a test tube.
Sure you do. You know that whatever survives (if anything) will be more fit in terms of your intentionally specified fitness measure than that which doesn’t survive. That you don’t know the exact configuration your specified sieve process will produce on its attempt to acquire your fitness goal is irrelevant to the point.