Chapter 5 Chapter 4 The Practical Functions of God
The minister who wrote to me, mentioned in the previous chapter, discovered God from a wasp; and Darwin lost his faith with the help of another bee: I could not convince myself, Darwin wrote, that mercy And Almighty God would have purposely created Hermites with the stated intention that they would obtain food inside living caterpillars.In fact, Darwin originally maintained his religious beliefs to cater to his religious wife Emma. The reason why he gradually lost his beliefs was more complicated.What he said about the jibe was just an epigram.The dreadful habits of the pheasant-bee, to which he speaks, are also shared by the genus wasp, a relative of the pheasant-bee.We covered this in the previous chapter.Not only does the female wasp lay her eggs inside a caterpillar (or locust, or bee) so that her larvae can feed on the caterpillar, etc., but, according to Fabre and others, the female wasp carefully keeps her stinger Inserts into every ganglion of the prey's central nervous system, anesthetizing but not killing it.This keeps the meat fresh.
It is not clear whether this anesthesia acts like a general anesthetic, or merely incapacitates the prey, as nuxy seeds do.In the latter case, the prey may realize that it is being eaten alive from the inside, but be unable to take any action.This sounds cruel, but as we are about to see, nature is not cruel, it is just ruthlessly impartial.This is one of the hardest lessons for humans to learn.We cannot admit that things can be neither good nor bad, neither cruel nor kind, but just insensitive to all suffering, without purpose.
Our human brains are intentional.We find it difficult to do anything without asking what it is for, what is the motive for doing it, or what the purpose is.To be morbidly obsessed with purpose, to seek malevolent motives out of sheer chance and bad luck, is paranoia.Yet this is only an exaggerated form of a common delusion.Show us any thing or process, it is hard for us not to raise the question of why and ask what is it for?It is natural for animals living in environments surrounded by machines, artwork, tools, and other artifacts of human design to be eager to know motives everywhere; moreover, animals' waking thoughts are governed by their own purposes .A car, a can opener, a screwdriver, a pitchfork, etc., all give us legitimate reasons to ask: what is it for?Our pagan ancestors may have asked the same question about thunder and eclipses, rocks and streams.Today we pride ourselves on being free from this primitive animism.If there were a stone in a stream that we could conveniently step across, we would regard its use as an accidental gift, without any real motive.But when disaster strikes, the old temptations come back with a vengeance. Indeed, the word blow is an echo of animism: Why, oh, why is cancer (earthquake, hurricane) befalling my child?When it comes to the origin of things or the fundamental principles of physics, the same temptation is often celebrated, culminating in the existential question of nihilism: Why is there such a thing instead of nothing?I can't remember how many times, after I give a public lecture, someone in the audience stands up and says something like: You scientists are very good at answering how questions.But you must admit that you are helpless when you encounter the question of why.Prince Lip listened to my colleague Peter.Dr. Atkins (Peter Atkins) expressed this view after his speech in Windsor.Behind this question there is always an unspoken but never proven meaning: the why questions that science cannot answer, there must be some other sect qualified to answer them.Of course, this implication is highly illogical.
I'm afraid Dr. Atkins won't show any good looks to His Highness Why.It's a simple fact: just because you can think of a problem doesn't make it sound or sensible.There are many things, and you can ask questions about it, such as what is its temperature?or what color is it?But you can't ask how hot or what color envy or prayer is.Likewise, your why question about a bicycle fender or the Kariba Dam is valid; but in any case you cannot expect an answer to a why question about a boulder, a catastrophe, Mount Everest, or the universe .No matter how sincere the questions are, they can be quite inappropriate.
Between windshield wipers and can openers on the one hand, rocks and the universe on the other, there is a living creation somewhere in between.Unlike a rock, although a living body and its organs are also objects, it appears to have motives written all over it.Of course, the notorious idea that living beings have motives has always dominated classic creationism, adopted by theologians from Aquinas to William.Paley (William Paley) continues to be cited by modern scientific creationists.
This real process of giving birds wings, eyes, howls, nesting instincts, and everything else about life has been given the strong illusion of design motivation.The real evolutionary process is now generally understood and it is Darwinian natural selection.It is astonishing that we did not begin to understand this until a century and a half ago.Before Darwin, even educated people, though, had abandoned the why question of rocks, streams, and eclipses.Yet implicitly accepts the plausibility of the why question about living things.Now, only the scientifically ill do it.However, just those two words conceal the unpleasant fact that it refers to the vast majority of people.
In fact, Darwinists do conceive a kind of why question about living things, but in a special metaphorical sense.Why do birds sing and what are their wings for?A modern Darwinist would jot down such a question and give a plausible answer based on the natural selection of the bird's ancestors.Illusions about motivation are so powerful that biologists themselves use well-designed hypotheses as research tools.
As we saw in the previous chapter, long before the epoch-making research work on bee dancing, Carl.Feng.Frisch found that some insects did have color vision, in spite of powerful orthodox objections.What prompted him to do the decisive experiment was a simple observation that bee pollinators went to the trouble of making pigments.If bees are color blind, why are they doing this?Here, the motivational metaphor is more precisely the assumptions involved in Darwin's natural selection being used to make a powerful inference about the world.If Feng.Frisch was quite wrong when he said that flowers are colored and therefore bees must have color vision.But here's what he said: Flowers are colored, so it's at least worth my effort to do some new experiments to test the hypothesis that bees have color vision.He is absolutely right in saying this.When he looked into the matter, he discovered that bees have excellent color vision, but they can see in a different spectrum than we do.They cannot see red light (they may call what we call red light yellow).However, they can see shorter wavelengths of ultraviolet light, which is a vivid color for them and is sometimes called bee purple.
When Feng.When Frisch realized that the bees were seeing the ultraviolet part of the spectrum, he reasoned again using the motivational metaphor.What, he asked himself, do bees use their ultraviolet vision for?His thoughts turned a full circle, and he returned to the flowers.Although we cannot see ultraviolet light, we can make photographic films that are sensitive to ultraviolet light, and we can make filters that pass ultraviolet light but block visible light.According to his hunch, Feng.Frisch took some ultraviolet photos of the flowers.He was delighted to see patterns of spots and stripes that had never been seen by the human eye before.What we see as white or yellow flowers are actually dotted with patterns of ultraviolet light, which often serve as guideposts to guide bees to the nectaries.The hypothesis with an obvious motive has come back to us again: the flower, if it had been designed, would take advantage of the fact that bees can see ultraviolet light.
Feng.In Frisch's old age, his most famous experiment, that of bees dancing which we discussed in the previous chapter, was inspired by the American biologist Adrian C.Questioned by Adrian Wenner.Fortunately Feng.Frisch lived long enough to see his work replaced by another American, James. L．James L. Gould (now at Princeton University) proved right with one of the most brilliantly conceived experiments in biology.I'm going to touch on this succinctly because it's relevant to my point of view on assumptions like design.
Wehner and his colleagues don't deny that bees do dance.They didn't even deny Feng.All the information that Fritsch says is contained in the dances, what they deny is that other bees can read these dances.Yes, the vertical angle of straight travel in the waggle dance is related to the angle between the direction of the food and the sun, but other bees don't learn that information from the dance, Wiener says.Yes, the speed of the various movements in the dance can read information about the distance to the food, but there is no solid evidence that other bees read this information.They may have overlooked it.Skeptics say that von Frisch's evidence is flawed, and that when they replicate Frisch's experiments with appropriate controls (that is, pay attention to other ways bees use to find food), the experiments no longer support von Frisch.Frisch's dance language assumes.
Jim.It is at this point that Gould comes into the story with his virtuosic experiments of genius.Gould draws on a long-known fact about bees that you read about in the previous chapter.Although bees usually dance in the dark, using the vertical up direction to represent the sun's projection on the horizon, if you turn on the lights in the hive, they will effortlessly switch to a direction that may be more original method.They will completely ignore gravity, and use the light bulb instead of the sun as a reference for the direction of the dance.Fortunately, when the dancing bees switched their reference from gravity to the light bulb, the other bees were not misunderstood.Other bees watching the dance also changed direction, so the message of the dance remained the same.Other bees still flew in the direction pointed by the dancer in search of food.
Now, let's take a look at Jim.Gould's ingenuity.He applied black shellac to the eyes of a dancing bee so that it could not see that the light bulb was lit.So it still dances according to the normal rules of gravity.However, other bees watching the dance were not blindfolded and were able to see the lighted bulbs.They see the dance as if the rules of gravity were dropped and replaced by the rules of the lightbulb sun.Watching bees measure the angle of their dance steps relative to a light source, while dancing bees adjust their angles according to the lines of gravity.In effect, Gould was forcing the dancing bees to lie about the direction of the food, not in general, but in specific directions that Gould could precisely manipulate.Of course, he didn't just use a blindfolded bee for the experiment, but selected a sample of bees according to statistical requirements, and conducted the experiment from different angles.The experiment worked.
Feng.Frisch's hypothesis about the language of dance was triumphantly proved correct.
I don't tell the above story for fun.I want to illustrate the negative and positive sides of assumptions about good design.When I first saw Weiner and colleagues' challenge paper, I publicly laughed at them.Even if it turns out that Vena was wrong, it's not good.What I'm laughing at is all based on the assumption of good design.
After all, Weiner did not deny the bee dancing, nor did he deny Feng.Frisch said the dance contained all the information about the distance and direction of the food, Wiener simply denied that the other bees read the information.This is something that I and many other Darwinian biologists cannot tolerate.The dances of bees are so complex, so varied, and so attuned to the apparent motive of the dancers to inform other bees of the distance and direction of food.It seems to us that no other route but natural selection could have brought about such a perfectly harmonious expression.In a sense, we fall into the same trap that creationists fall into when they meditate on the miracle of life.The purpose of the bee dance is simple: something useful must be done, which is to help other bees find food.Again, even the aspects of the dance in perfect harmony relate to the angle and velocity of the dance in relation to the direction and distance of the food: something useful has to be done.Therefore, we believe that Vena must be wrong.I have no doubts that even if I were smart enough to think of Gould's blindfold experiment (which I didn't, of course), I wouldn't bother to do it.
Gould not only thought of doing this experiment creatively, but also took the trouble to do it.This is because he is not led astray by good design assumptions.We're walking a tightrope, but I guess Gould and vonLike Frisch (doing color research), one already has enough good-designed hypotheses in mind to believe that one's own good experiments have a decent chance of success and are therefore worth the time and effort.
Now, I'm going to introduce two technical terms: reverse engineering and utility functions.In writing this section, I was inspired by Daniel.Darwin's Dangerous Ideas by Daniel Dennett.Reverse engineering is a reasoning technique, and it's done like this.You are an engineer who discovers an artifact without understanding how it works.You make some assumptions about what it does: it's designed for such and such a purpose.To figure out what problem this artifact solves, you dissect it and analyze it: If I wanted to make a machine to do such-and-such, would it be made like this?Or is it better to interpret the object as a machine designed to do that?The slide rule, until recently the talisman of the venerable engineer's profession, has fallen into disuse in the electronic age like a relic of the Bronze Age.Future archaeologists might marvel at a slide rule and notice how easy it is to draw straight lines or butter bread.But to set either of these two uses as its original use violates the economic premise.Just use as a ruler or butter knife without having a sliding part in the middle of the ruler.Furthermore, if you look at the marking lines on the ruler, you will find that it is a precise logarithmic scale.The graduations are printed with great detail, so it cannot be an accidental thing.Archaeologists will come to think of it as a clever way to do multiplication and division quickly in the pre-computer age.The mystery of the slide rule was solved with the help of reverse engineering, using clever and economical design assumptions.
Utility function is not an engineer's term, but an economist's term.It means maximum.Economic planners and social engineers are very much like architects and real engineers in that they try to enlarge something.Utilitarians strive to maximize the greatest happiness for the greatest number of people (btw: this sounds more sensible than it is).Under this umbrella, utilitarians may prioritize long-term stability more or less at the expense of short-term happiness, and they differ in how they measure happiness: or in terms of monetary wealth, Or according to job satisfaction, or according to cultural life satisfaction, or according to interpersonal relationship.Others blatantly seek the greatest possible individual happiness at the expense of the public good, and they glorify their egoism with a philosophy which asserts that if everyone takes care of themselves, the public can achieve the greatest happiness.By observing the behavior of individuals throughout their lives, you should be able to reverse engineer their useful function.If you reverse-engineer the behavior of a country's government, you might assert that what the government maximizes is employment and universal welfare.For another country, its utility function may be the right to re-election of the president, or the wealth of a ruling family, the size of the harem, the stability of the Middle East, or the stability of oil prices, and so on.The problem is that you can envision more than one utility function.It is not always clear what individuals, large corporations, or governments are trying to maximize, but it may be safe to assume that they are maximizing something.This is because humans are a deeply purpose-driven species.This principle holds true even when the utility function is expressed as a weighted sum, or in complex functions with many inputs.
Let's go back to living organisms and try to infer their useful functions, there may be many kinds, but they all boil down to one.A good way to dramatize our task is to imagine that the living creature is created by a fairy engineer, and then try to reverse engineer what the fairy engineer is trying to maximize: God What is the utility function of a cheetah, the cheetah gives people all kinds of signs that it is brilliantly designed for something, and it is extremely easy to reverse engineer a cheetah to figure out their utility function.It appears that they were carefully designed to hunt antelope.Their teeth, claws, eyes, nose, leg muscles, spine, and brain remind us that God designed the cheetah to cause the greatest possible death toll of the antelope.Conversely, if we reverse-engineer the antelope, we find equally impressive evidence that the antelope was designed for exactly the opposite purpose: the antelope survives, and the cheetah starves.As if the cheetah was designed by a god, and the antelope was designed by the opponent of this god.In other words, if there is only one Creator, and he created the tiger and the lamb, the cheetah and the gazelle, what did he intend?Is he a sadist who enjoys watching bloodshed?Is he trying to avoid overpopulation of African mammals?He is trying to maximize David.Attenborough (David Attenborough) TV ratings?These easy-to-understand practical functions may all prove to be true.In fact, of course, these are all wrong.Now, we have a very deep understanding of the only practical function of life, which has nothing to do with the above statements.
This point of view has been prepared for the readers in the first chapter, that is: the real practical function of life, the practical function that is maximized in nature is the survival of DNA.However, DNA is not free to drift, it is locked in living organisms, it must also make most of the major functions at its disposal.The DNA sequence in the cheetah's body relies on driving the cheetah to kill the gazelle to maximize its own survival.The DNA sequences in gazelles maximize their own survival by promoting development in the opposite direction.
In both cases, however, it is the survival of the DNA that is maximized.
In this chapter, I'm going to reverse engineer several examples so that the reader can see how everything can be explained once you assume that the survival of DNA is extended to the maximum.
Sex Ratio The male to female ratio in the wildlife world is usually fifty:fifty.But there are many species of haremism, where one male possesses multiple females.In these species, this ratio seems to have no economic sense at all.In a population of elephant seals that was studied, 4 percent of the males accounted for 88 percent of the mating volume.No one ever noticed how unfair God's utility function was to most elephant seals who became bachelors.To make matters worse, God, bent on reducing costs and improving efficiency, is obliged to note that the 96% of male elephant seals that are deprived of mating rights consume half of the food resources of the entire population of elephant seals (actually more than half, Because adult male elephant seals are much larger than females).The extra bachelor elephant seals have nothing to do but wait for their chance to replace the lucky four percent harem master.How could the existence of these plethora of bachelors like seals be justified?Any practical function that slightly considers the economic efficiency of the group will eliminate these bachelors.Instead, there should only be as many males as there are just enough to impregnate females.However, once again this apparent anomaly is explained beautifully and succinctly when you understand the practical function of true Darwinism, which is to maximize the survival of DNA.
I will discuss the sex-ratio example in a little more detail, because its utility function fits subtly into an economic treatment.Darwin admitted his own perplexity, saying: "I used to think that it would be in the interest of the species to produce equal numbers of both sexes, in accordance with the laws of natural selection."But now that I see the complexity of the issue, it's safer to leave it for the future.As usual, the great Ronald.Sir Ronald Fisher stepped up after Darwin.
Fisher's reasoning is as follows.
All individuals are born with only one mother and one father.Therefore, considering the situation after several generations, the total number of births of all males must be equal to the number of births of all females.I don't mean every male and every female, because it is clear and important that some individuals reproduce more than others.I'm talking about a comparison between all males and all females.All offspring must be assigned to individual males and females not equally, but shared.The birth cake shared by all males must be equal to the birth cake shared by all females.Therefore, if there are more males than females in the population, the average cake per male must be less than the average cake per female.That is, the ratio of average male births (expected births) to female average births depends only on the ratio of males to females.The average number of offspring of the minority sex is larger than that of the majority sex.Only when the sex ratio is equal and there are no minorities can the two sexes have an equal expected number of children.This extremely simple conclusion is purely the result of theoretical logical reasoning.This logical reasoning does not depend on any empirical fact other than the basic fact that all children are born with a father and a mother.
Gender is usually determined during pregnancy, so we can assume that no individual has the ability to determine their own gender (this time the verbosity is not a formality, but a necessity).Let us join Fisher in hypothesizing that one of the parents might have the ability to determine the sex of their offspring.Of course, this ability does not refer to the conscious or deliberate use of power, but to the fact that a mother may have a genetic predisposition to produce a chemical in the vagina that somewhat repels the sperm that bears sons, but not the sperm that bears daughters .
Alternatively, a father may have a genetic predisposition to produce more sperm for daughters than sons.Regardless of what it actually does, imagine yourself as a parent and try to decide whether to have a son or a daughter.Again, we are not talking about conscious decisions, but about the effect of genes on the generational selection of organisms on the sex of their offspring.
If you want to maximize the number of your grandchildren (including grandchildren, granddaughters), should you have a son or a daughter?We have seen earlier that the child you should want, regardless of gender, should be a minority in the group.
In this way, your children can expect a larger birth share, and you can expect more grandchildren.If neither sex is more than the other In other words, the sex ratio is already fifty:fifty You can't benefit from choosing this or that sex, whether you want a son or a daughter It doesn't matter anymore.Thus, a fifty:fifty sex ratio is called an evolutionary stable point, a term coined by the great English evolutionist John G.Maynard.Created by John Maynard Smith.Your predisposition towards sex selection will only pay off if the existing sex ratio is not fifty:fifty.The question of why an individual should have as many offspring as possible need not be asked.The genes that cause an individual to produce the most offspring are the ones we expect to see in the world.The animals we are studying inherit the genes of their successful ancestors.
It is tempting to state Fisher's theory in terms of fifty:fifty being the optimal sex ratio, but it is certainly wrong.If males are a minority, the best sex choice is boys; if females are a minority, girls.If both men and women are in the minority, then there is no optimal choice, and well-planned parents do not care about whether to have a boy or a girl.Fifty:fifty is said to be a stable sex ratio in evolution because natural selection dislikes any tendency to deviate from it, and if there is any deviation, natural selection will favor a tendency that restores balance.
Moreover, Fisher recognized that, strictly speaking, the number of males and females is not controlled by natural selection in a fifty:fifty ratio, but by what he calls the amount of parental expenditure on children.Parental spending is all the hard-earned food that goes into one child's mouth, and all the time and energy that goes into caring for that one child that could be used for something else, like caring for another child.For example, imagine a species of seal parents who spend twice as much time and energy raising their sons as they do their daughters.Male seals are much larger than females, which makes it easy to believe this (although it may not actually be accurate).Let's think about what it means.The real choice parents face is not the question of should I have a son or a daughter, but should I have a son or two daughters?This is because you can raise two daughters with food and other things that raise one son.The evolutionarily stable sex ratio would then be two females to one male, measured in number of individuals.However, the evolutionary stability sex ratio is still fifty:fifty when measured in terms of parental spending (rather than number of individuals).Fisher's theory is to balance spending on both genders.The result often happens to be the same amount that balances the two sexes.
As I have said, even among seals, the total amount of parental expenditure spent on sons does not appear to differ appreciably from the total amount spent on daughters.The big difference in weight presumably didn't show up until after the parental spending was done.So the decision facing parents is still, should I have a son or a daughter?Even if the total cost of raising a son to adulthood is much greater than the total cost of raising a daughter to adulthood, as long as the additional cost is not borne by the decision makers (parents), that is considered in Fisher's theory.
Fisher's rule about balancing spending still applies when one sex has a higher death rate than the other.For example, suppose male infants are more likely to die than female infants.If the sex ratio at pregnancy is fifty:fifty, by adulthood females will outnumber males.Males would then be the minority sex, and we would naively expect that natural selection would favor those parents who would produce sons.Fisher would expect this too, but only to a certain extent, and to a precise point.He would not have expected parents to conceive a surplus of sons to compensate for higher infant mortality and achieve sex balance in mating groups.No, at pregnancy the sex ratio should be slightly more males than females, but only so much that the total cost of raising the sons will equal the total cost of raising the daughters.
This time, the easiest way to think about it is again to put yourself in the position of the parent making the decision and ask: should I have a daughter who will probably survive, or a son who will die in infancy? ?If you decide to get grandchildren by sons, you may have to spend more resources on some extra sons, replacing those who will die.Think of it this way: every son who survives carries the ghost of his dead brother.He carried them on his back, which meant that the decision to rely on sons for grandchildren resulted in parents spending extra money wasted on the baby boy who would die.Fisher's ground rules still apply.The total amount of food and energy invested in the sons (including feeding the aborted son until the moment of death) will equal the total amount invested in the daughters.
What if, instead of higher male infant mortality, males had higher mortality after parental spending ends?In fact, this is also a common thing, because grown men often fight and hurt each other.This condition can also lead to a surplus of females in mating colonies.At first glance, it might seem that parents of single sons are in favor, which would give them an advantage in mating groups where males are absent.On reflection, however, you realize that this reasoning is absurd.Parents have a decision to make: should I have a son (brought up, who is likely to die in the war, but who, if he lives, will bear me exceptionally many grandchildren), or should I have a son? To have a daughter (she will certainly bear my fair share of grandchildren)?You can expect to have as many grandchildren from a son as you can expect to have on average from a daughter.
The cost of making a son is still the cost of feeding and protecting him until he leaves the nest.The fact that he was likely to be killed after he left the nest did not change the calculations.
Throughout this reasoning, Fisher assumes that parents are the ones who make the decisions.However, this calculation changes if someone other than the parents makes the decision.For example, suppose an individual is able to influence its own gender.Again, I don't mean conscious will to influence gender.My hypothesis is that the gene that drives an individual toward femaleness, or toward maleness, is oriented in a direction that depends on the cues it is given by the environment.In accordance with our usual conventions, and for the sake of brevity, I shall speak of the deliberate choice of the individual; in this case, the deliberate choice of its own sex.If an animal with a harem, such as a seal, was granted this flexibility of choice, dramatic consequences would occur.Every elephant seal aspires to be a male with wives, but if they lose the fight for a female, they would rather be a female than a single male.The sex ratio of the population will then become substantially more female than male.Unfortunately, elephant seals cannot reconsider their sex, which is determined during gestation.However, some fish are able to change sex.Male bluehead wrasses are large and brightly colored, and they have dull-colored females as wives.Some females are larger than others and form a dominant group.If a male dies, his place is quickly taken by the largest female, who quickly becomes a brightly colored male.This kind of fish is blessed by nature, and it will not suffer anywhere.Instead of wasting their lives as bachelors, waiting for a male fish with wives to die, they spend their waiting time as fertile females.The bluehead wrasse sex-ratio system is rare, where God's utilitarian function coincides with what socioeconomists consider shrewd.
Therefore, we recognize that parents and themselves are the decision makers.Are there any other decision makers?In social insects, input decisions are largely made by sterile workers.Workers are usually big sisters (and in termites, big brothers) of the nursed larvae.Among social insects, honeybees are a class we are more familiar with.If any of my readers are beekeepers, they probably already know that the sex ratio in the hive does not, on the face of it, match Fisher's calculations.First, it must be noted that worker bees should not be considered female.嚴格地說，它們是雌性，但它們不生育。因此，根據費舍爾理論調整的性別比，應是雄蜂數與蜂房培育出的新蜂王數之比。在蜜蜂和螞蟻中，存在著一些特殊的技術上的緣由（這個問題已在《自私的基因》中討論過，這裡不再重複），可以期望性別比保持在三：一，雌性占多數。養蜂人都知道，實際情況遠非如此，而是雄性占多數。一個興旺的蜂巢在一季裡只能產生五六隻新蜂王，但是卻會產生成百甚至上千隻雄蜂。
How is this going?像在現代進化理論中常有的情況一樣，我們從牛津大學的w． D．漢密爾頓（W．D．Hamilton）處得到了答案。它揭示並概括了費舍爾關於性別比的全部理論。解開蜜蜂性別比之謎的鑰匙，就在蜜蜂成群地飛離蜂巢這一現象中。一個蜂巢在很多方面就像是一個個體。它發育成熟，繁殖後代，直到最終死亡。蜂巢繁殖出來的產品就是蜂群。盛夏，當蜂巢最興旺的時候，它就放出一群女兒一個蜂群。對蜂巢來說，生產出蜂群就是繁殖、生育的同義詞。如果把蜂巢比喻為一座工廠，蜂群就是最終產品，它們攜帶著群體的寶貴基因。一個蜂群有一隻蜂王，還有數千隻工蜂。它們作為一個整體全都離開了原蜂巢，聚集在一起形成密集的蜂群，掛在大樹上或者岩石上。這就是它們的臨時宿營地，同時它們也在尋找永久性的新家。在幾天之內，它們找到一個洞，或者一棵中空的樹安下新家。或者，在今天更常見的情況，它們被一個養蜂人捕獲（也許就是原來那個養蜂人），裝入一個新蜂箱中。
放出子群是一個興旺蜂巢的職責。為此要做的第一件事是造一個新蜂王。通常是造六隻左右蜂王，但是能存活的註定只有一隻。第一隻羽化的蜂王把其他幾隻都刺死（或許，多餘的幾只是出於保險的目的而造的）。蜂王與工蜂在遺傳上是可以互相轉變的，只是蜂王是被放在專門的蜂王室裡撫育（蜂王室掛在蜂巢下部），享用特別富有營養的膳食。蜂王膳食中包括王漿。小說家戴姆．Barbara.卡特蘭德（Dame Barbara Cartland）把她的長壽，以及女王般的舉止，都羅曼諦克地歸功於王漿。工蜂住在較小的巢室中，這些巢室以後用來存放蜂蜜。雄蜂在遺傳上不同，它們是未受精的卵孵化而來的。值得注意的是，一個卵究竟是變成雄蜂還是雌蜂（包括蜂王和工蜂），這由蜂王決定。蜂王只在一次婚飛中交配，那是它剛剛成年的時候。它把精子在體內儲存一生。當卵經過產卵管產出時，蜂王也許會從儲存中放出少量精子使卵受精，也許不放出精子給卵子授精。因此，是蜂王控制著蜂卵的性別比。然而，看來是工蜂掌握著隨後的所有權力，因為它們控制著對幼蟲的食物供應。例如，如果從工蜂的觀點來看，蜂王所產的雄蜂卵過多，工蜂就可以讓雄幼蟲挨餓。在任何情況下，都是工蜂們決定著是把一個雌卵變成一隻工蜂呢，還是變成蜂王，因為這完全取決於養育條件，特別是飲食。
現在，讓我們回到性別比的問題上來，看看工蜂所面臨的抉擇。止像我們已經看到的那樣，工蜂們不像蜂王那樣選擇生兒還是生女，它們要決定的，是生產弟弟們（雄蜂），還是生產妹妹們（年輕的蜂王）。現在讓我們回到我們的謎語上來。因為實際上性別比是雄性嚴重過剩，這從費舍爾的觀點來看是沒有意義的。讓我們更仔細地研究一下工蜂面臨的決定。我們剛才說過這是在弟弟和妹妹之間的選擇。但是且慢，決定撫育一個弟弟，實際上僅僅是把蜂房裡的食物和其他資源撥出一些供撫育一隻雄蜂用。然而，決定撫育一隻新蜂王，則意味著必須提供遠比一隻蜂王所需營養資源多得多的東西。作出撫育一隻新蜂王的決定，等於承擔起建造一個蜂群的職責。一隻新蜂王本身的花費微不足道，僅僅是它所吃的少量王漿和其他食物。然而它所包括的大部分花費，是製造數以千計的工蜂。蜂群離開後，蜂巢也就失去了這些工蜂。
對於性別比中那種明顯異常的雄性過多現象，這幾乎肯定是個確切的解釋。對於我在前面談到的性別比，這是一個極端的例子。費舍爾的規則聲稱對雄性和雌性的支出量必定相等，而不是雄性與雌性個體數目相等。為一隻新蜂王所支付的花費，也包括對工蜂的龐大支出，這些工蜂如果沒有新蜂王本來不會離開這蜂巢。這很像我們對海豹群體數量的假設。假設養育一種性別的花費是飼養另一種性別的兩倍，得到的結果是，這一種性別的數量是另一種性別的一半。對於蜜蜂來說，一隻蜂王的花費是雄蜂的數百乃至數千倍，因為它承擔著新蜂群所有工蜂的花費。因此，雄蜂數目比蜂王多數百倍。這個奇怪的故事還有一個極具諷刺性的結尾：令人不可思議的是，當蜂群離去時，它裹走的是老蜂王，而不是新蜂王。然而無論如何，其經濟意義是一樣的。製造一個新蜂王的決定仍然擔負著新蜂群攜同老蜂王遷居新址的費用。
為了圓滿結束我們關於性別比的論述，讓我們回到開始時的後宮之謎：一大群光棍象海豹消耗著近一半（甚至比一半還多）的食物資源，卻永遠不會生育，也做不出任何有用的事，這是一種極其浪費的安排。很顯然，在這裡，全體成員的經濟利益沒有得到最大限度的發揮。What's going on?讓我們再來一次，把你自己擺到決策者的位置上比如一個母親，她試著決定，為了得到最多的孫子孫女，是要個兒子呢，還是要個女兒。乍看上去，她的決定是不公平的：我應該要個兒子呢，還是要個女兒？兒子可能成為光棍而根本不可能給我生任何孫子，女兒可能成為後宮裡的妻妾，給我生下數量不少的外孫。對這位未來母親的適當的回答是：但是，如果你要個兒子，他將來可能擁有許多妻妾，這種情況下，他就會為你帶來遠比靠女兒所能得到的多得多的孫子孫女。為了簡單明瞭起見，假設所有雌性都以一個平均值來生育，而每十個雄性中則有九個根本不能生育，另一個雄性獨佔了所有雌性。如果你有一個女兒，你可以指望得到平均數量的外孫外孫女。如果你有一個兒子，那麼你就有九十％的可能性連一個孫子孫女都得不到，你又有十％的可能性，得到十倍於平均數的孫子孫女。你期望從兒子那裡得到的孫子孫女的平均數，與你期望從女兒那裡得到的外孫外孫女的平均數是一樣的。儘管物種水準上的經濟理由要求雌性占多數，自然選擇仍然偏愛五十：五十的性別比。費舍爾法則仍然適用。
我用動物個體的決定介紹了這些推理，但是要重複一句，這只是速寫。事情的本源是，在基因庫中，能最大限度地增多子孫的基因變得越來越多了。這個世界充滿了經過世世代代成功地流傳至今的基因。若不是靠影響個體在獲得最大數量後代方面的決定，基因怎能成功地穿越世代到達今天？費舍爾的性別理論告訴我們，實現這個最大數量，這同最大限度地提高物種（或全體成員）的經濟利益是非常不同的。這裡也有一個實用功能，但它與出現在我們人類經濟思想中的實用功能相去甚遠。
關於後宮經濟的浪費，可以總結如下：雄性們沒有致力於有用的工作，而是把能量和體力浪費在互相之間的爭鬥上。即使我們用明顯的達爾文主義的方式來定義有用這個詞，例如以撫養孩子作為有用的工作，情況也是如此。如果雄性都把精力轉用於有用的事情，而不是浪費在互相爭鬥上，整個物種就能花費較少的力氣、消費較少的食物，而撫養更多的孩子。
研究工作效率的專家會被象海豹世界的情形驚得目瞪口呆。
下面是個類似的情形。在一個車間裡只有十台車床，因此只需不超過十個工人，車間就可以運轉。管理部門卻決定雇用一百個工人，而不僅僅是十個工人。每天，這一百名工人都來上班並領取工資，然後把一天的時間都花在為搶佔車床而打架上。在這些車床上造出了一些產品，但是不會多於十個人造出的產品數，可能會少一些，因為這一百人都在忙著打架，這些車床並未有效地利用起來。工作效率專家不會有任何遲疑。九十％的工人是多餘的，應該正式宣佈這一點並將多餘工人解雇。
雄性動物們不僅僅在肉體搏鬥上浪費它們的精力在這裡，人們又一次採用人類經濟學家或效率專家的觀點來定義浪費這個詞。許多動物還有選美競賽。這給我們帶來了另一種實用功能，一種雖然沒有直接的經濟意義，卻有可供我們人類欣賞的功能審美。乍看上去，似乎上帝的實用功能有時是沿著競選世界小姐（謝天謝地它現在已不再時髦）的路線展現的，不過是雄性們在展臺上炫耀自己。在鳥（比如松雞和流蘇鷸）的表演場，這種情形看得最清楚。一個表演場是一小塊地，傳統上是雄鳥在這裡列隊向雌鳥炫耀自己。雌鳥們訪問表演場，觀看數隻雄鳥在場地上作器宇軒昂的表演，然後從中選出雄鳥與之交配。在有表演欲的動物中，雄鳥通常都有稀奇古怪的裝飾；在作展示時，它們跳躍、點頭，還發出奇怪的聲音，極令異性矚目。剛才我用稀奇古怪這個詞形容雄鳥的裝飾，當然包含著主觀的價值判斷；雌松雞們大概不會認為一本正經作表演的雄松雞那趾高氣揚的舞姿、那低沉壓抑的聲音是稀奇古怪的，這就行了。某些種類的雌鳥們的審美觀念與我們碰巧一致，比如孔雀或天堂鳥。
夜營的歌聲，野雞的長尾，螢火蟲的閃光，還有熱帶魚身上的彩虹鱗，都將美昇華到最高水準，但是，這還不是或者說這僅僅是偶然給人類帶來歡娛的美。如果我們喜歡這些，那是一份額外的收益，那是個副產品。那些使雄性對雌性具有吸引力的基因，很自然地發覺它們自己沿著數字之河順流而下，從歷史走向未來。只有一個實用功能才令這些美麗有意義，這個實用功能就是解釋象海豹性別比的同一個實用功能；是解釋獵豹與羚羊之間表面上無益的速度之爭的同一實用功能；是解釋杜鵑和寄生蟲，眼睛、耳朵和氣管，不育的工蟻和生育力極強的蟻后，等等現象的同一實用功能。這偉大的萬能的實用功能，數量在生物世界每個角落被孜孜不倦地擴展到最大的這個功能，在每事每物中，都是ＤＮＡ的生存；你試圖去解釋的一切現象都是由它造成的。
孔雀華麗的羽毛成了它沉重的負擔，使它很不方便，嚴重地妨礙它去做有用的事，即使它願意做也不行，總的說，它們並不願意做。雄性鳴禽把大量的時間和精力用在歌唱上。這肯定會給它們帶來危害，不僅是因為這歌聲會引來捕獵者，還因為這消耗了精力，浪費掉本可用於恢復精力的時間。有一個專攻鷦鷯生態學的學生曾經宣佈，他有一隻野生雄鷦鷯就確確實實是唱死的。任何將本物種的長期福利放在心上的實用功能，甚至只是將某個特定雄性個體的長期生存放在心上的實用功能，都會減少唱歌的時間，減少炫耀的時間，減少在雄性中互相爭鬥的時間。然而，因為ＤＮＡ的生存被最大限度擴展了，所以沒有什麼東西能阻止那些除了使雄性在雌性眼中顯得漂亮之外別無所長的ＤＮＡ傳播。漂亮本身並不是一種絕對的優點，但是如果某些基因確實給予一個物種的雄性某些素質，這物種的雌性又覺得它有吸引力，那麼這些基因就會生存下去，不管它們自己願意不願意。
為什麼森林裡的樹都那麼高？很簡單，為了超過競爭對手。一個合理的實用功能會努力使它們長得不那麼高。每棵樹仍能得到等量的日光，而用於粗大樹幹和結實的支撐根基的支出卻要少得多。但是，如果它們都很矮，自然選擇就忍不住要去偏愛一棵長得略高些的變異個體。賭注已經增加，其他賭徒不得不跟著來。沒有什麼力量能阻止賭博逐步升級，直到所有的樹都長到荒謬而浪費的高度。說它荒謬和浪費，只是從有理性的經濟計畫者的觀點看；他們考慮的是最大限度地提高效率。但是一旦你理解了真正的實用功能基因在努力使它們自己的生存達到最大限度，你就會明白它是合理的。Similar examples abound.在雞尾酒會上，你大聲說話，聲音嘶啞了，原因就在於每一個人都在用最大的嗓門喊叫。
只要賓客們達成一致意見，大家都低聲講話，那麼他們就能聽清對方的談話，並且少用大嗓門，少消耗些精力。然而，除非設置員警，否則這個意見是不會生效的。總會有人自私地用稍大一點的聲音講話，破壞了這個意見，於是一個接一個，每個人都不得不隨著提高聲音。當每個人都竭力大喊大叫時，才能達到穩定平衡，這比從合理觀點要求的聲音要大得多。合作的約束一次又一次被它內部的不穩定性所破壞。上帝的實用功能很少是為最大多數的最大利益服務的。上帝的實用功能總是在不協調的自私自利的爭鬥中背叛初衷。
人類有一種相當可愛的傾向，就是認為福利指的是集體福利，利益指的是社會利益、物種的未來福祉，甚至是生態系統的未來福祉。從對於自然選擇的難題與意外事件的冥思苦想中，人們引伸出了上帝的實用功能，糟糕的是它與這類烏托邦的空想並不一致。
確實，有時候基因會由生物體在其自身水準上採取編制非自私的合作程式，甚至是自我犧牲的程式，來實現基因在自己水準上最大限度地增加自私的福利。但是，集體的福利總是一種偶然的結果，並不是一種原動力。這就是自私的基因的意思。
讓我們從一件類似的事情開始，來看一看上帝的實用功能的另一個方面。達爾文主義心理學家尼古拉斯．韓弗理（NicholasHumphrey）曾虛構過亨利．福特一件很顯眼的事情。據說，製造業效率的最高典範福特有一次委託他人對美國的廢車堆放場做一次調查，讓他們尋找T型福特車上有沒有從未損壞的零件。調查員們回來以後報告說，幾乎每一種零件都有損壞：車軸、殺車、活塞等等都發生了損壞。但是，有一個例外情況引起了他們的注意。在這些損壞的轎車裡，所有的轉向軸主銷都還能用很多年，無一例外。
福特以無情的邏輯做出結論，T型車的轉向軸主銷的品質超過了要求，命令以後生產的主銷應該降一個等級。
也許你和我一樣，不大清楚轉向軸主銷是個什麼東西，但這不要緊。它是汽車上用的某種零件。福特的無情斷言是完全合乎邏輯的。另一種做法是，提高汽車其他零部件的品質，使它們都達到轉向軸主銷的水準。但是，如果那樣做，他所製造的就不是T型福特車，而是羅爾斯．羅伊斯了。那可就不是這個故事的初衷了。羅爾斯．羅伊斯是很有名望的轎車，當然T型車也是，但它們的價位不同。問題的關鍵是，整車是以羅爾斯．羅伊斯標準製造的呢，還是以T型車標準製造的。如果你想製造一種雜交車，有些零件達到T型車品質，有些零件達到羅爾斯．羅伊斯的品質，那麼你得到的是兩者最糟糕的結合，因為一旦車裡面最差的零件損壞了，整輛轎車就會被扔掉；花在高品質零件上的錢，則由於這些零件根本沒有時間被用壞而簡單地被浪費了。
把福特的經驗用在活的軀體上比用在汽車上意義更大，因為汽車上的零部件在一定限度內是可以用備件來更換的。猴子和長臂猿在樹上生活，總有掉下來摔斷骨頭的危險。假設我們對猴子屍體做一個調查，統計一下猴子體內主要骨頭摔斷的頻率。假設每一根骨頭都會在什麼時候摔斷，只有一個例外：腓骨（與脛骨平行）在任何猴子屍體裡都沒有被摔斷的先例。Henry.福特的毫不猶豫的處方會是：重新設計腓骨，將它的標準降低一個等級，而這也正是自然選擇所可能做的。於是產生了腓骨被降了級的突變個體，它們的生長規律要求將寶貴的鈣從腓骨中轉移，突變的個體可用節省下來的材料加厚體內的其他骨頭，以達到每一根骨頭易被摔斷的程度都相同這個最終目的。或者，變異個體可以用節省下來的鈣來製造更多的奶，以餵養更多的幼仔。從腓骨上可以安全地削去一些骨頭，起碼可以削到結實程度與略弱於腓骨的那根骨頭一樣。而另一種辦法羅爾斯．羅伊斯的辦法是要使其他的骨頭都達到腓骨的水準，這是難以達到的。
實際的計算並不這麼簡單，因為有些骨頭比其他骨頭更重要。
我想，如果一隻蜘蛛猴的腳後跟發生了骨折，那它比手臂骨折活下去要容易些，因此，我們不能簡單地要求自然選擇使所有的骨頭都同樣容易折斷。但是，我們從亨利．福特的傳說中學到的主要教訓毫無疑問是正確的。對一種動物而言，身體中的某一部分好過了頭的可能性是存在的，我們應該期望自然選擇使這部分的品質降低，直到與身體其他部分的品質相平衡，而不是越過這個平衡點。更正確的說法是，自然選擇應該偏愛均衡品質，從降低品質和提高品質這兩個方向使身體裡所有零件達到適當的平衡。
在考慮生命的兩個相當獨立的方面的平衡時，我們就特別容易評價這種平衡。比如，雄孔雀的生存與它在雌孔雀眼中的美貌之間的平衡。達爾文主義的理論告訴我們，所有生存的終極目的，都是為了傳播基因，但這並不能阻止我們將身體分成這樣一些部分（比如腿），主要關係到個體的生存，以及那樣一些部分（比如生殖器），關係到生育。再如鹿角是用於與對手搏鬥，而腿和生殖器等與對手的存在並不相干。許多昆蟲對它們生命歷程中一些根本不同的階段加以嚴格劃分。毛蟲的任務就是取食和生長。蝴蝶就像它們所訪問的花朵，其任務就是繁殖；它們並不生長，它們吸吮花蜜用作飛行時的燃料消耗掉。一隻蝴蝶成功地生殖的時候，它傳播的基因不僅是為了使之能成為有效地飛行與交配的蝴蝶，而且也是為了能成為有效地進食的毛蟲。蜉蝣幼時為水下若蟲，它們在水裡進食，生長達三年之久後，它們羽化成為能飛的成年蜉蝣。成年的蜉蝣只生存幾個小時。很多蜉蝣都被魚吃掉了，即使不被吞食，它們也會很快死去，因為它們不會進食，甚至連內臟都沒有（亨利．福特或許喜歡）。它們的任務就是一直飛到發現一隻配偶，然後，在交配，留下它們的基因包括能在水下進食三年的有用若蟲的基因之後，它們就死去了。一隻蜉蝣就像一棵經年累月成長起來的樹，花開後僅經歷輝煌的一天就死去了。成年蜉蝣就是那在生命結束而新生命開始之際短暫開放的花朵。
幼年鮭魚沿著它出生的溪流順流而下進入大海，在大海裡進食、生長，度過它一生中大部分時光。當它成年時，它可能靠著嗅覺會找到它出生溪流的入海口，經歷一次史詩般著名的旅行，躍上瀑布、跳過急流險灘，回到河源出生處。它在那裡產卵，一個新的迴圈又開始了。在這一點上大西洋鮭魚與太平洋鮭魚之間存在一個典型的差異。大西洋鮭魚在產卵之後，會回到海洋中去，它們還有可能再重複一次這個迴圈；而太平洋鮭魚在產卵之後幾天內就死去了。
一條典型的太平洋鮭魚就像一隻蜉蝣，但是在其生命歷程中沒有明顯劃分的若蟲階段與成蟲階段。回游時溯流而上需要付出如此巨大的努力，以至它不能做第二個迴圈。因此，自然選擇偏向於這樣的個體，它把每一盎司的資源都投入到重大戰役生殖中去。生產之後，任何剩餘資源都是浪費，這就相當於T型福特轎車超過設計標準的轉向軸主銷。太平洋鮭魚已經進化到在生產之後生存能力減弱至零，節省下來的資源被轉移到魚卵或魚的精子中去了。大西洋鮭魚被引向另外一條路線，也許是因為它們回游的河流比較短，而且這些河流發源於不太高的山，有時一些個體能夠靠著剩餘的資源成功地完成第二個迴圈。這些大西洋鮭魚付出的代價，就是它們不能為產卵付出那麼多的精力。在延續生命與生殖之間必須作出取捨，不同鮭魚選擇了不同的取捨。鮭魚生命週期的一個特徵，就是由回游導致精疲力盡所造成的不連續性。在一個生育季節與兩個生育季節之間不存在輕易的連續。承擔第二個繁殖季節的任務，就會極大地削弱第一個繁殖季節的成功率。太平洋鮭魚的進化朝著這樣的方向前進，即毫不含糊地完成第一個繁殖季節的義務。結果，一個典型的個體在進行偉大的一次性產卵努力之後，隨即就肯定會死去。
在每一種生物中都可以發現同樣的取捨現象，但是通常不那麼具有戲劇性。我們自己的死亡程式或許與鮭魚有某些相似之處，但是不那麼露骨與清晰。無疑，一個優生學家能夠培育出一個特別長壽的人類種族。為了繁殖後代，你會選擇這樣一些個體，他們以犧牲自己的孩子為代價，把大部分資源都投入到自己身上：比如，這些個體的骨骼已加固厚實，很難折斷，但是用於製造奶水的鈣卻所剩無幾。如果你以犧牲下一代為代價養嬌自己，那就很容易活得稍長些。優生學家能夠這樣精心養育，並將取捨法用於所希望的長壽方向。可是，大自然不會用這種方式來嬌養誰，原因在於，對下一代吝嗇的基因傳不到未來。
大自然的實用功能從來都不是為了自身，而只是為了未來的繁殖才重視長壽。任何能生育一次以上的動物（像人類，而不像太平洋鮭魚）都面臨著在現有的兒女（或一窩小動物）與未來的兒女們之間作出取捨的問題。一隻野兔將其全部精力和資源都傾注於第一窩小仔兔，它的第一窩小兔會非常強壯，但是將不會剩下什麼資源來支持她生育第二窩幼仔了。第二窩、第三窩仔兔體內攜帶的、留些東西作儲備的基因會在兔群中傳播開去。這類基因顯然沒有擴展到太平洋鮭魚群體中，因為在一個繁殖季節與兩個繁殖季節之間實際存在的不連續性是那麼不可逾越。
隨著年齡的增長，我們將於明年死去的概率最初逐年下降，然後是一段穩定狀態，最後又開始向上爬坡。在死亡率上升這段長時間內發生了什麼事情？這與太平洋鮭魚的原理基本相同，但是延長到一段時期之內，而不像太平洋鮭魚那樣，在無節制地產卵之後，倉卒而集中地大批死亡。衰老的機制最初是由諾貝爾獎獲得者、醫學家彼得．梅達沃爵士（Sir Peter Medawar）於五十年代早期提出來的，以後傑出的達爾文主義者G． C．威廉斯（G．C．Wil liams）和w． D．漢密爾頓（W．D．Hamilton）又對其基本思想作了許多修正和補充。
基本的論點是這樣的：第一，正如我們在第一章中所看到的，任何遺傳效應都是在生物的生命歷程中某個特定時間內開啟的。
許多基因在胚胎早期就打開了，但是其他一些基因（像亨廷頓氏舞蹈病的基因）則是直到中年之後才打開的。第二，遺傳效應的細節，包括它在何時打開，可以被其他基因所修改。一個帶有亨廷頓氏舞蹈病基因的人可能會死於這種病，然而這種病會使他在四十歲上死去，還是到五十五歲逝去，可能受到其他基因的影響。也就是說，通過修正基因的選擇，特定基因的作用時間可能在進化過程中推遲，也可能被提前。
在三十五歲到五十五歲之間開啟的基因（像亨廷頓氏舞蹈病基因）有足夠多的機會在殺死它的攜帶者之前被傳給下一代。但是，如果這種基因在二十歲時打開，它就只能被那些在很年輕時就生育的人傳給後代，因此它就很容易被篩選掉。如果它在十歲時打開，那麼一般來說它就不會被傳下去。自然選擇會偏愛那些使亨廷頓氏舞蹈病基因延遲啟動的修正基因。根據梅達沃．威廉斯的理論，這就是為什麼這種基因要等到中年以後才開啟的準確原因。從前它可能曾經是一個早熟的基因，但是自然選擇促使它將致死效應推遲到中年。毫無疑問，仍有少許的自然選擇壓力會把致死效應推至老年；然而，這種壓力很輕微，因為很少有人在生育並把這基因傳給下一代之前死去。
亨廷頓氏舞蹈病基因是致死基因的一個特別明顯的例子。還有很多基因本身並沒有致死效應，卻具有提高因其他原因死亡的概率的效應，這種基因叫做亞致死基因。再者，它們的開啟時間可能受到修正基因的影響，因而被自然選擇推遲或提前。梅達沃認識到，老年人的衰弱可能表明致死和亞致死遺傳效應的積累；這些效應在生命週期中被一再推遲，並且僅僅因為它們晚起作用，就得以悄悄通過生育網遺傳給後代子孫。
現代美國達爾文主義者中的老前輩G． C．威廉斯於一九五七年給這個故事帶來一個重要的轉折。他回到了我們關於從經濟角度進行取捨的觀點上。為了理解這一點，我們需要再補充幾個背景事實。通常，一種基因不只具有一種效應，並常常作用於身體上明顯不同的部位。這種多效應不僅是個事實，而且很多都是基因對胚胎發育發揮的效應，而胚胎發育又是一個複雜的過程。所以，任何一個新的突變都很可能不僅有一種效應，而是有幾種效應。儘管在這些效應中可能有一種是有益的，但是似乎不會有多於一種的有益效應。這是因為大多數突變效應都是壞的。除了這是個事實之外，從原理上還可以這麼說：如果你開始製造一台複雜的裝置比如一台收音機把它做得更差些比把它做得更好些，會有更多的方法。
每當自然選擇因為某個基因在年青時的有益效應比如說，讓一個男性青年具有性吸引力而偏愛該基因，就會有不利的一面出現：比如，在中老年時患上某種特殊的疾病。理論上，年齡效應是另外一回事，但是根據梅達沃的邏輯，自然選擇不會因為同一基因在老年時表現出來的有益效應，而偏愛青年時的疾病。再者，我們可以再次引用有關修改基因的觀點。一個基因有數種效應，每種效應（好的效應和壞的效應）都會在後來的進化中改變其開啟時間。根據梅達沃原理，基因的好效應都傾向於在生命早期呈現，而壞效應則傾向於延遲至後期才表現出來。另外，在某些情況中，早期效應和晚期效應之間會進行直接的交換。這在我們對鮭魚的討論中已經是不言而喻了。如果一種動物所消耗的資源數量是有限的，比如說，只夠使身體強壯並且逃脫危險，早消耗這些資源的任何傾向都會比晚消耗它更受歡迎。消耗晚的動物更有可能在有機會耗盡它們的資源之前就已經由於別的原因而死去。我們在第一章中曾介紹過一種倒敘的方法。將梅達沃的基本觀點用這種倒述法來表示，每一個人都是從綿延不斷的祖系傳下的後裔，祖輩們在他們生命中某個時期都曾是年青的，但是他們中許多人卻從未能活到老年。所以，我們繼承下的就是那些年青的東西，而不必要繼承那些衰老的東西。我們傾向於繼承那些使我們在出生之後很久才死亡的基因，而不傾向於繼承那些使我們在出生後早早夭折的基因。
讓我們回到本章開頭的悲觀論點，當實用功能被最大限度擴大的功能是ＤＮＡ的生存時，這並不是獲得幸福的妙方。
只要ＤＮＡ能夠傳下去，在其傳播過程中，不管是誰或是什麼受到損害都無關緊要。對於達爾文的姬蜂基因來說，毛蟲最好是活的，被吃掉時仍是新鮮的，至於這樣會使毛蟲感到怎樣的痛苦，則是無所謂的。基因不在乎痛苦，因為它們不在乎任何事情。
如果大自然是仁慈的，那麼它至少會作出使毛蟲被從內部活活吃掉之前先行麻醉這樣一個小小的讓步。但是大自然既非仁慈，也非不仁慈。它既不反對遭受痛苦，也不贊同遭受痛苦。除非影響到ＤＮＡ的生存，否則大自然對這樣或那樣的痛苦根本不感興趣。
你可以想像一種基因，比如說，當瞪羚在遭到致命一咬時，這種基因能使它們平靜下來，這是很容易做到的。這種基因是否會得到自然選擇的偏愛呢？除非使瞪羚平靜的行動能提高該基因傳播給未來世世代代的機會。很難說清楚為什麼必需是這樣，我們可以因此猜想，瞪羚在被追捕至死亡時承受著可怕的痛苦和恐怖就像它們大多數終將要承受的那樣。自然界中每年遭受磨難的動物，總數大大超過了任何公平的期望值。就在此刻我構思這句話的時間內，數千隻動物正在被生吞活剝；其他一些動物則正在驚恐地嗚咽著逃命；還有一些正被使人焦躁的寄生蟲從內部緩慢地吃掉；數以千計的各種動物因為饑餓，乾渴和疾病而正在死去。肯定是這麼回事。如果時間充裕，這一事實會自動導致動物群體增長，直至饑餓和苦難的自然狀態重現。
神學家們為邪惡問題以及相關的受難問題而煩惱不已。
在我最初寫這一段的那天，英國的報紙全都報導了一輛大轎車並無明顯原因地撞車的可怕事件，車上滿載著一所羅馬天主教學校的學生。所有孩子無一逃生。在倫敦《星期日電訊報》上，一位作者這樣說：你怎能相信充滿愛心的、全能的上帝能允許發生這樣的悲劇？牧師們已經不是第一次為了這種神學問題而突然大發作了。這篇文章繼續引用了一位教士的回答：對這一問題的簡單回答是，我們不知道為什麼非要有一個上帝來允許這些可怕事件發生。但是這一可怕