Chapter 10: You Scratch My Back, I’ll Ride on Yours

Welcome to Last Minute Lecture.

This free chapter overview is designed to help students review and understand key concepts.

These summaries supplement not replaced the original textbook and may not be redistributed or resold.

For complete coverage, always consult the official text.

Welcome back to The Deep Dive.

Today, we are diving into one of the most fascinating paradoxes in the natural world,

cooperation.

I mean, if evolution is really all about selfish genes, genes that just care about making more copies of themselves, then why do we see animals helping each other all the time, often at a huge cost to themselves?

That is the absolute core of it.

We've talked before about parents and offspring or mates, but today we're looking at something different, at all the behaviors that happen within a group.

We're pulling everything from a single chapter.

You scratch my back, I'll write on yours.

Really, our mission is to unpack how these selfish genetic rules can actually drive behaviors that look incredibly selfless.

When you look out there, group living is just everywhere.

You see these massive flocks of birds, huge schools of fish, even mixed herds of zebra and new.

The main takeaway, before we even start, seems to be that for any individual in that group, its genes have to get more benefit out of the deal than they put in.

It looks like teamwork, but underneath it all, the driving force is always self -interest.

Exactly.

Before we get to the really puzzling stuff, the sacrifices, let's just cover the simple selfish benefits,

and they can be very straightforward.

Take hunting, a single hyena, it can only do so much, but a whole pack can bring down a water buffalo.

Even when they share the meal, the net gain for each selfish hyena is way bigger than if it had hunted alone.

Or what about just the physical advantages?

I love that image of emperor penguins huddling together for warmth.

A classic example.

But that huddle is pure unadulterated self -interest.

Every single bird gains because it's exposing less surface area to the cold than if it were standing by itself.

The bird in the middle has the best spot, and the ones on the edge are just desperately trying to push their way in.

It's not about keeping the group warm.

No.

It's about me not freezing.

And we see that same principle, that same exploitation of others when they're in motion.

Fish in a school get a hydrodynamic boost from the fish in front.

Like cyclists in a race drafting.

Or the V formation in birds.

They're constantly trading that hard lead position to share the burden.

It's all self -serving.

Okay, so let's lay out the roadmap for this dive.

We're going to start with the most basic form of group living, just bunching up, and show how it's actually a form of selfish exploitation.

Then we'll hit those really puzzling examples, like the bird alarm calls and the the starting gazelle.

From there, we have to talk about the ultimate altruists, the social insects, and their totally bizarre genetics.

And we'll finish up with the really high -level strategies, the ones you need for delayed cooperation between animals that aren't even related.

Let's start there, then.

This idea that just clustering together is fundamentally selfish.

And that brings us to W .D.

Hamilton's brilliant thought experiment, the geometry for the selfish herd.

The great thing about Hamilton's model is just how simple it is.

And he's very clear from the start.

This is a herd of selfish individuals.

They're only acting for themselves, not for some good of the group.

And to model this, he starts with one very simple,

very elegant assumption about the predator.

And what's that assumption?

That the predator, being smart about its energy, always attacks the nearest prey.

Right.

Why waste energy chasing something far away when you've got a meal right in front of you?

It's just an efficient hunting strategy.

So if I'm the prey, my entire goal is just to make sure that I am not the nearest one.

So how does Hamilton visualize this race to not be the closest?

He introduces this concept of the domain of danger.

So just imagine a field with all the pre -animals scattered on it.

The domain of danger for, say, individual A, is the patch of ground where any point is closer to A than it is to any other animal.

OK.

So if a predator shows up inside my domain.

You're probably the target.

I'm lunch.

Right.

So the goal is to make my domain as small as humanly or animally possible.

Exactly.

Now picture them moving in a nice geometric formation.

If you're an individual and you're surrounded by six neighbors, like in a honeycomb pattern,

your domain of danger is this tiny little hexagon.

You're relatively safe because you're shielded.

But the individual is on the edge of the herd.

That's where the model shows the danger just in splits.

Right.

Precisely.

The animals on the edge are only partly surrounded.

Their domain of danger isn't a neat little hexagon.

It's this huge wide open wedge of space where there's no one else to act as a buffer.

So if a predator approaches from the outside.

That poor guy is definitely the nearest target.

Inevitably.

So the urge for every single individual is the same.

Get to the middle.

Yeah.

Shrink my domain of danger by putting other bodies between me and the open space.

I want a shield of my neighbors.

And this relentless inward pressure driven purely by each individual trying to lower its own risk is what makes a loose group become a tightly bunched herd.

It's like a constant stampede towards the center.

So the bunching itself isn't cooperation.

Not at all.

It is selfish exploitation by each individual of every other individual.

That's a really critical distinction.

The tight herd, which we see as this united front, is really just a byproduct of millions of tiny selfish decisions.

Each one is just using its neighbors as mobile, edible shields.

And that bunching only stops when they physically can't get any closer.

Or it becomes impossible to move or feed.

The selfish herd is just a beautiful example of how complex group behavior can emerge from simple self -interested rules with absolutely no need for group level altruism.

Okay.

So the selfish herd model explains why animals clump together.

But what about when they take an act of risk?

I mean, a behavior that seems to put them in more danger to save others.

This is why we have to tackle things like the bird alarm call.

Right.

The alarm call.

It works as a warning.

A hawk is overhead and the whole flock just freezes or dives for cover.

The problem, as you said, is that making the sound seems like it would draw the hawk's attention right to the caller.

It seems like you're volunteering to be the target.

And we know that danger is real because of the sound of the calls themselves.

P .R.

Marler's work showed that they are these really high -pitched whistles.

They often start and end gradually.

And they're in a very narrow frequency band.

All characteristics that make the sound incredibly difficult to pinpoint by ear.

So if natural selection went to all that trouble to make the call sneaky, it must be because the non -sneaky callers got eaten over a very long time.

The risk is definitely high.

So the selfish gene theory has to find a selfish advantage that's big enough to outweigh that obvious danger of calling out.

And the first sort of default explanation is always kin selection.

Right.

If the flock has your brothers, your cousins, your kids, then the gene for calling can still prosper because it's saving copies of itself in your relatives, even if you sometimes pay the price.

But that's just the safety net.

It doesn't really explain what happens in mixed groups.

We need purely selfish theories.

Okay, let's get into those.

The first one is what the chapter calls the cave theory, from the Latin for beware.

This one works best for birds that use camouflage, the kind that crouch and freeze in the bushes when they see a hawk.

So one sharp -eyed bird spots the hawk and it freezes instantly, but its buddies are still hopping around making noise.

And that could draw the hawk over to the general area.

Exactly.

The hawk might be drawn in not by the silent caller, but by its noisy conspicuous neighbors.

Wait, so the call isn't to save the neighbor, it's to silence the neighbor.

Precisely.

The selfish gain is immediate.

The bird gives a quick, quiet shush.

The cave call to force its companions to freeze and be quiet,

and that reduces the chance that they'll accidentally bring the hawk down on top of the caller's head.

It's self -protection.

That completely reframes it.

It's not an alert, it's a command.

Shut up so you don't get me killed.

Okay, what about the second theory?

The never break ranks idea.

This one applies to birds that fly into a dense tree for cover when they see a hawk.

Now, if one bird sees the hawk and just flies off on its own without warning anyone, it becomes the odd one out, an easy target.

Right.

We know predators like peregrine falcons specifically go for the isolated individual, the one that breaks formation.

The safety of the flock is gone the second you're alone.

Flying off by yourself, even to a safe tree, just massively increases your personal domain of danger for that split second.

It does.

So the best selfish strategy is to fly into cover, yes, but only if everyone else does too.

The warning call is an act of, well, manipulation.

You know, to make sure the whole flock moves together so the caller gets the benefit of cover without losing the protection of the crowd.

So in both of these theories, the bird giving the call is just picking the option that makes it least endangered.

It would actually be more dangerous to stay quiet.

It's just smart risk management that happens to look like altruism.

But now we have to go from that discreet little whistle to the complete opposite.

The, well, the seemingly suicidal high -jumping of the Stodding gazelle.

Right.

Stodding is so flamboyant.

It's this huge bounding leap that burns a ton of energy and they do it right in front of a predator like a cheetah.

This was the example that people like Ardre said had to be for the good of the group.

The selfish gene explanation for this one proposed by Emman Sahavi is, well, it's a really daring piece of thinking.

He said the Stodding is aimed at the predator, not at the other gazelles.

So the cheetah is the audience.

How is showing off for the cheetah less dangerous than just running?

Because the display is an honest advertisement of fitness.

The gazelles that can stop the highest, the most powerfully, are showing off their superior muscle, their stamina.

The message to the cheat is basically, look how high I can jump.

I am in peak physical condition.

Chasing me is going to be a complete waste of your energy.

You should go try and catch my neighbor who isn't jumping as high.

So it's a competition and the weakest jumper loses because that's the one the predator who is already looking for the old and the sick is going to choose.

Exactly.

It turns this moment of terror into a competitive display.

It's not suicidal altruism at all.

It's a high stakes advertisement designed to push the predator's attention onto someone else.

A completely selfish strategy.

Okay, so we've seen how selfish strategies can explain clumping together and even risky behaviors like alarm calls and starting.

But now we have to face the ultimate challenge.

The social insects.

I mean, bees, ants, wasps.

These are creatures where most individuals are sterile workers who will literally kill themselves to defend the colony.

The cooperation in these colonies is just.

It's legendary.

They really do act like a single superorganism.

They can regulate the nest temperature, share information with chemical signals, and even fight off invaders with a precision that's like a body's immune system.

But the absolute key to their amazing altruism, and the key to explaining it with selfish genes, is worker stability.

The workers themselves never reproduce.

That's right.

Their germline, the immortal genes that pass to the next generation, only flows through the queen and the drones.

So in a very real sense, the sterile workers are just like the cells of a body.

Like our liver cells or muscle cells.

Which explains the kamikaze bee.

Its death is no more of a genetic loss than a tree shedding a leaf.

All of that worker's energy is just aimed at preserving its genes by taking care of its relatives, which are being churned out by the queen.

And this division of labor between specialized bearers, the queen, and specialized carers, the workers, is an evolutionarily stable strategy.

But, and this is the crucial part, it can only be stable if the carers are very, very closely related to the babies they're raising.

And this leads us to the unique biological engine behind it all.

The weird genetics of the hymenoptera.

Something called haplodeploidy.

Right, and we need to go through this carefully because this one genetic quirk changes all the math.

So in ants, bees, and wasps, the males, the drones, they develop from unfertilized eggs.

They have no father.

They only have one set of chromosomes.

They're haploid.

And the crucial part is this.

Because they only have one set of genes,

every single sperm they produce is a perfect, identical clone.

Unlike in humans where every sperm is a random 50 % sample.

Exactly.

The females, on the other hand, are diploid.

They develop from fertilized eggs and have two sets of chromosomes.

Okay, so how does that mess with the standard 50 % relatedness that we're used to?

Well, the relationship from a mother to her offspring is still 50%.

A queen passes on half her genes to her sons and half to her daughters.

But now, let's consider two full sisters, two worker ants.

They share, on average, half of their mother's genes, just like we do with our siblings.

But what about their father?

His sperm was identical.

Yes.

So for all the genes they got from their father, they are 100 % identical, like twins.

So, you averaged that out.

50 % relatedness through the mom, 100 % through the dad.

And the total relatedness between full sisters jumps to 75%, three quarters.

And that single fact, the haploid father, means a worker is more closely related to her full sister than she would be to her own daughter.

That's the evolutionary lever.

A gene in a worker for helping to raise sisters gets passed on more effectively 75 % than a gene for having her own kids, which would only be 50%.

The math is inescapable.

Worker sterility becomes the best strategy for the genes.

And this explains why this kind of true sociality with sterile castes has evolved at least 11 separate times in this one group, the hymenoptera, but basically only once anywhere else.

But here's where it gets even more fascinating.

This genetic imbalance creates a huge conflict of interest inside the colony.

A battle over the sex ratio, which was figured out by Trevor's and Hare.

Right.

The queen is equally related to her sons and her daughters, 50 % of both.

So, from her point of view, the best investment is a one -to -one ratio.

Equal resources spent on making new queens and new drones.

But the worker has this massive bias.

She's 75 % related to her sisters, but only 25 % related to her brothers.

So, the worker's ideal investment ratio is totally skewed.

Three -to -one in favor of sisters over brothers.

And there's the conflict.

The queen's body and the worker's body are controlled by the same set of genes.

But those genes are pushing for different outcomes, depending on which body they're in.

So, who wins?

The queen who lays the eggs or the workers who control the nursery?

That's what Trivers and Hare tested.

They went out and collected reproductives from 20 species of ants.

And they didn't just count them, they weighed them.

They measured the total biomass, the actual investment.

And what did they find?

The investment ratio was a dead ringer for the three -to -one female -to -male ratio that you would predict if the workers were in charge.

The workers control over the food supply.

Their ability to neglect or even kill the male larvae allowed them to impose their genetic preference over the queens.

That's incredible confirmation.

But the real clincher, the critical test, involved slave -making ants.

This is the perfect natural experiment.

These ants raid the nests of other species they steal the pupa and they raise them as slaves who then run their nursery.

So the workers raising the brood are completely unrelated to them?

Correct.

So there is zero selection pressure on the slave genes to enforce any particular ratio for this foreign brood.

The slave has no genetic stake in the game.

Which means the slave -making queen is finally free to pursue her own optimal one -to -one ratio without the workers fighting back.

And that is exactly what they found.

In the slave -making species, the investment ratio was basically one -to -one.

It proved that the outcome of this conflict depends entirely on who holds the power in the nursery and has the genetic incentive to use it.

Although the source does mention it's not always that clean.

Honey bees, for instance.

Yes, honey bees are a bit of a complication.

But Hamilton figured out that you have to include the cost of the swarm that goes with the new queen.

I mean, all those attendant workers are part of the investment in females.

When you add their biomass to the calculation, the ratio shifts back towards the worker's preference.

And there's also the issue of queens mating with multiple males.

Right, that lowers the average relatedness between sisters.

Which weakens the whole effect.

But even with these messy details, that core genetic asymmetry from haplodeploidy is still the best explanation we have for the incredible altruism we see in these insects.

Okay, so let's move beyond kin and insect genetics to cooperation between completely different species.

Mutualism and symbiosis.

Here, the cooperation is still selfish, but it works because of a fundamental asymmetry.

The partners have different skills.

And the clearest examples of this involve ants, who act remarkably like human farmers.

They practice both cultivation and domestication.

Let's start with cultivation.

The fungus gardens.

You see this in New World ants, and, completely separately, in African termites.

How does that work?

Well, ants aren't very good at digesting leaves themselves.

So they chew them up into a kind of compost, and then they sow a very specific kind of fungus on it.

The fungus is much better breaking down all the tough plant material.

So the ants eat the fungus, and they get an efficiently digested meal.

What's in it for the fungus?

A guaranteed home and protection.

The ants actively weed the garden, clearing out any other competing fungi.

And they spread their preferred fungus much more effectively than it could spread on its own.

It's a perfect partnership.

The fungus does the chemistry, the ant does the physical labor.

And the domestication side of the analogy is their relationship with aphids.

The ants milk them.

Right.

Aphids are these little sap -sucking machines.

They pump so much plant juice that what they excrete is still full of sugar.

The ants stroke the aphids with their antenna, which makes them release this honeydew, and the ants just drink it up.

So the ants are protecting their little sugar dispensers.

But why doesn't the aphid just leave?

What does it get?

Protection.

Aphids are basically defenseless against things like ladybikes.

But the ants will guard them fiercely.

They'll even move their aphid herds to better plants.

And this relationship has become so specialized that many of these farmed aphids have actually lost their own natural defenses.

They rely completely on the ants for security.

That's an amazing trade -off.

Give up your own armor in exchange for a full -time army.

And so the ant genes for farming aphids thrive, and the aphid genes for cooperating like waiting to be milked also thrive.

The relationship just becomes more and more intimate.

And this leads to one of the most radical ideas in the source material.

What if this symbiosis goes all the way down to the level of our own cells?

Right.

We move beyond something simple like lichens, which we know are a fungus and an alga living together.

And we start looking at our own bodies.

Think about the mitochondria in your cells, the powerhouses.

They look and act a lot like ancient bacteria.

They even have their own separate DNA.

The theory is that billions of years ago, our simple ancestor cells basically absorbed these bacteria, and we are the result of that permanent team -up.

So we are in a way just giant walking colonies of symbiotic genes.

And you can take it even one step further.

What if viruses aren't just invaders?

What if they're rebel genes that broke away from our cellular colonies eons ago?

And now they travel directly from body to body through the air, completely bypassing the normal routes of sperm and egg.

Wow, that's a completely different way to look at biology.

But the core insight is the same.

Whether it's two separate species or genes inside a cell, these mutually beneficial partnerships evolve if each partner gets more out of the deal than they put in.

So we've covered cooperation among relatives and between species where the benefits happen at the same time.

But the hardest problem is delayed reciprocity.

The whole you scratch my back, I'll scratch yours idea, but where's the time lag between the favor and the payback?

Because that delay just creates a massive opportunity for cheating.

I can take your help today and then just refuse to pay you back tomorrow when you need it.

And that's why delayed reciprocal altruism can only evolve in species that can do two things.

Recognize and remember each other as individuals.

If you can't remember who helped you and who stiffed you, the whole systems collapses.

To figure out how this could even get started, Trivers used game theory.

He set up a simple model.

Let's use the example of mice grooming parasites off each other's heads.

And we imagine three different genetically programmed strategies.

First, you have the sucker.

This mouse just grooms anyone who needs it.

No questions asked.

It's a pure altruist.

It pays the cost of grooming and gives away the benefit.

Second, you have the cheat.

The cheat loves to be groomed, but never ever grooms anyone else.

It gets all the benefits.

A clean head with none of the costs.

Now, if your population is only made up of suckers and cheats, the result is a foregone conclusion.

Cheats always do better than suckers.

Always.

They just exploit them mercilessly.

The cheat gene spreads like wildfire, and the sucker gene is driven to extinction.

And then what happens?

Once all the suckers are gone, the cheats have no one left to exploit.

They only ever meet other cheats.

Nobody grooms anybody, and the whole population might just collapse from all the parasites.

Exactly.

The population needs a way to punish the cheats.

It needs memory.

So enter the third strategy,

the grudger.

Okay, so let's define the grudger strategy clearly.

The grudger starts out nice.

It'll groom a stranger on the first meeting, and it will keep grooming anyone who has groomed it back in the past.

But if it grooms an individual, and that individual refuses to return the favor, the grudger remembers.

It holds a grudge and will never groom that specific cheat again.

So how does the grudger do in this environment?

Well, in a population of just the suckers and grudgers, they're identical.

Everyone is nice to everyone, and they all do very well.

But if you drop a few grudgers into a population of cheats, they'll do very poorly at first.

They'll waste all their energy grooming cheats who never pay them back.

If they're too rare, they just get wiped out.

But there's a tipping point.

The critical proportion.

How many grudgers do you need?

If the grudgers can somehow survive long enough to build up their numbers past a certain critical point, their chances of meeting each other become high enough to make up for all the cheats they run into.

Once they hit that threshold, they start getting a higher average payoff than the cheats because they're basically forming a little exclusive club of cooperators.

And once the grudgers get that slight advantage, the cheats start to decline because they're being locked out of all the grooming.

So the grudger strategy, with its memory and its punishment, is an evolutionarily stable strategy.

It is, but, and this is a really important detail from the game theory,

cheat is also an ESS.

A population that's almost all cheats is stable.

A lone grudger or sucker trying to invade would be instantly exploited and would fail.

So cooperation isn't inevitable.

A species could get stuck in one of two stable states.

A highly cooperative grudger society or a miserable low trust cheat society.

The computer simulations really bear this out.

You see the suckers get wiped out first by the cheats, then the cheats have a population boom.

But once the suckers are gone, the grudgers, who are just barely hanging on, can slowly start to gain ground and eventually drive the cheats out.

So the poor suckers are basically the fuel that allows the cheat strategy to work in the short term.

And in doing so, they almost doom the truly cooperative grudger strategy.

You need a critical mass of discriminators for cooperation to take hold.

And this model is just the foundation for understanding cooperation in so many species.

You see it everywhere, like with the cleaner fish.

Right.

Tell us about them.

There are dozens of species of these little fish or shrimp that just pick parasites off of bigger client fish.

And the big fish will just let this little cleaner swim right into its mouth and gills without eating it.

So the big fish is holding back, making what looks like an altruistic choice.

The system works because the cleaners have these special stripes and dances, but mostly it works because of sight tenacity.

Meaning they stay in one place.

Yes.

They have a fixed territory, like an underwater barbershop.

The client fish line up for their turn.

So the benefit to a big fish of having a reliable place to go and get cleaned outweighs the small one time benefit of eating the cleaner.

You invest in the long term service.

But of course, the selfish gene means there are cheats here, too.

There are other little fish that mimic the cleaner's look.

They get close to the big fish and then just take a bite out of its fin and run.

And that constant threat from mimics probably keeps the whole system honest.

But the fact that the real cleaners stay put allows the clients to learn who to trust.

It's the grudger strategy in action, rewarding the good guys and avoiding the cheats you remember.

So let's bring this all back to humans.

If this kind of cooperation requires huge memory and individual recognition, this pressure must have been a major driver in the evolution of our big brains.

Trivers actually suggested that many of our most deeply human psychological traits, things like envy, guilt, gratitude, sympathy, were shaped by natural selection to make us better at playing this game.

How does something like guilt fit into the grudger model?

Guilt could be a preemptive strike, a way to avoid being labeled a cheat in the first place.

If you do something antisocial, the feeling of guilt might make you act remorsefully.

Maybe you overcompensate later on.

It's a way of signaling that you're still a reliable cooperator so you don't get kicked out of the group.

So if I feel guilty, that feeling is pushing me to repair my reputation before I'm permanently blacklisted as a cheat.

Exactly.

And this suggests that our highest intellectual faculties, our ability for mathematical reasoning, might have evolved as mechanisms for social bookkeeping, for being more clever cheats, and, crucially, for getting better at detecting cheating in others.

Money itself is really just a formal token of delayed reciprocal altruism.

It's a powerful, maybe slightly cynical way to look at human society.

It suggests our most complex mental tools might be rooted in the very simple, ancient need to keep track of who owes what to whom.

So this deep dive has really given us a complete framework for seeing cooperation through the lens of the selfish gene.

It shows us that these selfless acts are almost always hiding a self -serving mechanism.

We started with the selfish herd,

showing that just clumping together is really about selfishly reducing your own individual risk.

Then we saw how things like alarm calls and starting are really just disguised selfishness, either silencing your neighbors to protect yourself, or advertising your fitness to push a predator onto someone else.

And the most extreme altruism of all, the sterile workers and social insects.

That comes down entirely to the bizarre genetics of haplodeploidy, where sisters are 75 % related.

That creates the epic 3 -to -1 sex ratio conflict, a battle the workers usually win.

And finally, we saw how cooperation between non -relatives depends on these complex strategies like the grudger, which uses memory and punishment to solve the problem of cheating,

stabilizing everything from cleaner fish to human societies.

I think the essential insight to take away is this.

Stable cooperation is almost never built on simple blind altruism.

It requires a robust mechanism, either a huge genetic imbalance or a clever behavioral strategy like the grudger, that punishes the exploiters and makes sure that the benefits of cooperation only go to those who pay their dues.

So think about the implications of Traver's suggestion.

If our most advanced mental skills, like abstract reasoning, evolved mainly to handle the complex accounting of our social contracts, to track and predict cheating.

How much of your own daily decision -making, even those little flashes of guilt or gratitude, is still being run by that ancient genetic program to detect and avoid being cheated?

Something to think about as you navigate your own social world today.

Thank you for joining us for this deep dive.

We'll see you next time.