Chapter 5: Aggression: Stability and the Selfish Machine

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, where we take complex sources, your curated articles and research, and provide you, the diligent learner, with the crucial insights you need to get fully informed.

Hello again.

Today, we are undertaking what I think is a really foundational mission, diving deep into the, well, the surprisingly logical world of animal aggression.

And we're looking at it through that cold calculating lens of the selfish gene.

That's right.

We're continuing our exploration of how evolution programs individual animals, these survival machines, to execute behavioral policies that, you know, at the end of the day, maximize the representation of their genes in the next generation.

And our focus today is all on conflict.

Specifically, we need to get our heads around this profound paradox.

What's that?

If every animal is fundamentally programmed to be selfish,

why isn't the natural world just engulfed in constant lethal chaos?

Why do they so often seem to fight with, well, with restraint?

That restraint, yeah, that's the whole issue.

And to really appreciate the dilemma, I think you first have to accept the initial premise the source material establishes.

To a survival machine, every other survival machine, and we'll put a pin in close relatives for later, is essentially just a moving obstacle or a potential resource.

Right, like a rock or a river.

Exactly.

Just like a desirable piece of territory or a water source.

But with one crucial distinction, and this is the whole thing, really, that particular resource or obstacle is itself an entity programmed for its own selfish genetic success.

And therefore.

And therefore, it is inclined to hit back.

That simple fact defines the entire strategic calculus of conflict.

The text draws a really important line between the two types of rivalry, and it's important for you to keep them separate in your mind.

First up, we have interspecies conflict.

Right, interspecies.

So this is rivalry between two completely different types of survival machines.

Think a mole and a blackbird both going after earthworms.

They're competing, but they might never even meet.

Exactly.

The competition is often indirect.

The mole's success directly reduces the food available for the blackbird and vice versa, but they're not, you know, having a showdown.

And these relationships can be incredibly complex.

I mean, we're talking about predator prey dynamics, parasite host interactions.

Or even highly specialized forms of exploitation.

I loved the example the text used of flowers and bees.

Oh, that's a perfect one.

It's pure exploitation.

The flower is using the bee as a pollen carrier, a flying taxi service for its genes.

And the bee gets a little reward.

It's incidentally rewarded with nectar, yeah.

It's all about one survival machine using or avoiding another to further its own genetic agenda.

But the conflict we are most concerned with today, the one that really, I think, tests the limits of the selfish gene theory,

is interspecies conflict.

Blackbird versus blackbird.

Lion versus lion.

And this kind of conflict is inherently fiercer.

It's much more direct.

And that's because the competitors are fighting for the exact same resources.

They live in the exact same niche.

They have the exact same way of life.

They are direct total rivals competing for food, shelter, and most fundamentally of all for mates.

And the text stresses that this struggle is overwhelmingly concentrated among males competing for females.

Because access to fertile females is almost always the limiting factor for reproductive success.

OK, so if the logical, purely selfish policy is to just eliminate the rival male entirely, kill him, maybe even eat him to get the calories back, why don't we see that?

Why doesn't that total uncompromising aggression just dominate the natural world?

And that brings us right back to the paradox of restraint.

Exactly.

And this observation, it was famously championed by ethologists like Conrad Lorenz in his book On Aggression.

Lorenz argued that animal fighting, it often looks like a formal tournament, right?

Like it's governed by ritualized rules.

Almost like a civilized duel.

He saw animals using bluff, threat displays, and they often seem to, you know, pull their punches.

They recognize and respect gestures of surrender.

He had that very poetic description of it, didn't he?

Fighting with gloved fists and blunted foils.

And to an untrained observer or maybe to someone who believes in group selection, that restraint looks a lot like altruism.

Yeah, it looks like it's being done for the good of the species, to keep the population healthy or avoid wiping out the whole gene pool.

But the text is very quick to caution against two things here.

For one.

First, Lorenz probably exaggerated how pure this ritual is.

There are plenty of examples where animals do kill their own kind.

And homo sapiens is certainly not unique in that.

No.

And second, and this is the most important part, this restraint has to be explained without resorting to altruism.

The selfish gene theory has to provide a better, more robust explanation for why avoiding conflict is often the best selfish strategy.

And the explanation, which really sets up our entire deep dive today, it all lies in a cold, hard cost -benefit analysis.

The restraint we see is just the result of avoiding the lethal risk of retaliation.

It's the ultimate calculation.

Is the potential reward really worth the potential cost of being severely injured or even killed?

This complex calculation, which was later formalized by game theory, explains everything.

Okay, let's unpack that.

Let's get into this complex cost -benefit idea in section one.

We're gonna focus on the costs of just outright pugnacity.

The text makes a really powerful point, that the costs aren't always what you'd think.

We think of the obvious ones, energy, time, risk of injury.

But there's a fourth, really subtle cost.

And that is the indirect benefit you might be handing to a third party.

Explain that.

Okay, imagine three males.

Let's call them A, B, and C.

A is our subject, B is his immediate rival, and C is a third rival to both of them.

Okay.

If A spends a huge amount of energy and takes a massive risk to kill,

B, sure, A has benefited, but A has also just removed one of C's rivals.

Right, so rival C is now unopposed by B, and he might suddenly be elevated to a position of strength.

He could become a far more formidable threat to A than B ever was.

So A's initial act of pure selfish aggression, it inadvertently benefits a dangerous long -term competitor.

It's so counterintuitive, isn't it?

The best strategy isn't just to wipe out the competition indiscriminately.

No, not at all.

It mirrors a real -world scenario that pest control officers often face.

It's a wonderful analogy used in the text.

Tell us more about that.

That really clarifies this non -obvious cost.

Well, sometimes pest control programs find that by, say, eradicating one type of pest, let's call it insect A, they inadvertently remove the main competitor of insect B.

And what if insect B is worse?

Exactly, what if insect B is even harder to control or causes more damage than insect A ever did?

The outcome of the whole pest control effort is now worse than the initial problem.

So the lesson is indiscriminate violence, whether it's against pests or rivals, it's just not good business.

Because you risk creating a vacuum that benefits your worst enemies.

You have to be selectively pugnacious, and even that is a risky gamble.

Right, think about fighting a really valuable rival, like a domine elephant seal bull who's guarding a harem of 100 females.

That bull, let's call him bull B, he's dominant for a reason, is because he's an excellent fighter.

He's already survived and won countless high -stakes battles.

So to even engage him is to risk a major injury, maybe a broken flipper or a permanent scar that just reduces your mobility forever.

And if you suffer a serious injury, even if you win the ultimate prize, the harem, you might be crippled.

You might not be able to defend it, or even fully exploit the reproductive opportunities.

Which leaves you vulnerable to the next challenger, maybe rival C, who just watched the whole thing from a safe distance.

The conclusion is just undeniable.

The decision to fight has to be preceded by an extremely complex, though entirely unconscious,

calculation of all the risks and all the rewards.

And this necessity, this need for strategic cost -benefit analysis, is what led to the formalization of the problem using game theory.

And this is where we introduce the foundational concepts of Jay Maynard Smith, who worked with G .R.

Price and G .A.

Parker.

They adapted economic game theory to model evolutionary conflicts, giving us the mathematical tools to actually predict stable behaviors.

Let's start with their definition of strategy.

Right.

In this context, a strategy is a pre -programmed behavioral policy.

And it's critical for you to stress that this is not a conscious plan decision by the animal.

The animal is a robot.

The animal is a robot, and we, as observers, are just describing the genetic programming that guides its actions.

We might write it out as a clear set of instructions, like attack opponent, if he retaliates, retreat.

If he flees, pursue him.

It's almost like a computer program for behavior, and the power of their work is in defining the evolutionarily stable strategy, or ESS.

What makes an ESS fundamentally different from just the best strategy?

The defining characteristic of an ESS is its stability against invasion.

If an ESS is adopted by most of the population,

no alternative strategy, particularly a new mutant one, can successfully invade and replace it.

And the key insight, what makes this game theory, and not just simple optimization, is that the best strategy for an individual depends entirely on what the majority of the other individuals are doing.

That's it.

So, if most people are playing strategy X, and I decide to switch to strategy Y,

selection will quickly penalize me and push the population back to X.

That's why it's stable.

Selection punishes deviation.

Once an ESS is achieved, selection just stabilizes that behavior pattern, because any gene that tries to code for an alternative strategy will systematically suffer lower payoffs and be removed from the gene pool.

Okay, that's a perfect setup.

Let's move into section two and apply this immediately with the clearest, most essential model, the hawk and dove model.

This is the cornerstone of the whole theory.

It is, and we need to walk through the math here slowly so the logic becomes crystal clear for you, the listener.

Absolutely, so the model assumes a symmetric contest.

Two rivals, they're fighting over a resource of a fixed value, and they can't recognize each other or know the opponent's strategy before the fight begins.

And we introduce two highly simplified contrasting strategies.

First, the hawk.

The hawk represents the purely aggressive, unrestrained approach.

A hawk always escalates immediately into a serious fight and only backs down if it's seriously injured.

And then the dove.

The dove is the restrained player.

They only rely on conventional threats and ritualistic displays,

the gloved fist.

So if a dove is attacked.

It immediately backs down to avoid injury.

The opponent gets the resource, sure, but the dove lives to fight another day.

Okay, to quantify the results, the source uses an arbitrary points system.

And these points, they represent genetic fitness, the potential payoff in terms of offspring.

Let's just confirm the key values.

Okay, we'll assign plus 50 points for winning the contest.

You get the resource, we'll assign zero points for losing.

We apply a big penalty of minus 100 points for being seriously injured.

And there's one more.

Right, because contests take time, we apply a minor penalty of minus 10 points for wasting time in a long ritualistic fight.

Excellent.

Now let's run the numbers on the two unstable scenarios, just to show why a mixed strategy is the only thing that works.

Scenario one, imagine a population made up entirely of doves.

In an all dove population, every contest is dove versus dove.

They engage in this long drawn out but completely harmless ritual tournament.

They just posture and threaten.

Until one eventually just pires out and gives up.

Injury is avoided, so the only cost is the time they wasted.

Okay, so a winning dove gets plus 50 points for the win, minus 10 points for wasted time.

That's a net of plus 40.

And a losing dove gets zero for the loss, minus 10 for the time.

A net of minus 10.

And since any given dove has a 50 -50 chance of winning or losing, the average payoff for a dove in this lovely peaceful society is plus 40 plus negative 10 divided by 2.

Which is plus 15.

Doves are doing well.

They're doing very well, but the population is completely unstable.

Now, a mutant hawk gene arises.

When this lone hawk meets a dove, what happens?

The hawk always escalates immediately.

The dove, following its programming,

immediately backs down.

The dove gets zero points for the loss.

But the hawk instantly wins.

It gets plus 50 points and suffers no injury and no time penalty.

The hawk's average payoff is a massive plus 50.

Just let me just say that again because the difference is huge.

The peaceful cooperative doves are averaging plus 15.

The selfish treacherous mutant hawk is scoring plus 50.

The hawk genes are going to spread like wildfire.

They'll take over the population and completely destroy that stable dove society.

That's the nature of treachery against a strategy that benefits the group.

It's not resistant to invasion.

But as those hawk genes become common, the dynamic flips entirely.

Which leads us to scenario two, a population that's all hawks.

Right.

Now every single fight is hawk versus hawk.

They escalate until one is seriously injured, that's minus 100, and the other wins, that's plus 50.

So the average payoff for a hawk in this total bloodbath is the average of plus 50 and minus 100.

Which is minus 25.

Ouch.

Hawks, in their pure aggressive state, are doing terribly.

They're constantly injuring each other and just rapidly lowering the overall fitness of the whole population.

And this sets the stage for the counter treachery.

A mutant dove shows up in this disaster zone.

Now when this dove meets a hawk, the dove always loses, it gets zero points.

But crucially, it immediately retreats.

And it avoids that massive minus 100 injury penalty.

So the dove's payoff is a solid zero.

And zero is dramatically better than the hawk's average payoff of minus 25.

So now dove genes spread rapidly through the hawk population, pushing things back toward the dove strategy.

And this raises a key question.

Does the population just oscillate back and forth forever?

Hawk, then dove, then hawk again.

And the answer is?

No.

That's the brilliance of the mathematical solution.

It doesn't oscillate, it finds a stable ratio.

The system reaches an equilibrium where the selection pressure on both strategies just stops because their average payoffs become exactly equal.

And for the specific point values we use, plus 50 for a win, minus 100 for an injury, that mathematically stable ESS ratio is 7 12th hawks and 5 12th doves.

At that precise mixture, if the population drifts a little bit towards more hawks, doves suddenly gain an advantage and their genes spread, pushing the ratio back.

And if it drifts towards more doves?

Hawks gain the advantage, pushing it back.

It is stable and it is persistent.

Okay, now we need to clearly differentiate the two ways the stability can actually manifest in the genes.

First, there's the stable polymorphism.

This means that individuals are genetically fixed.

You are either a hawk or you are a dove.

And selection maintains the ratio of hawk genes to dove genes in the overall gene pool at exactly seven to five.

You're born one or the other.

Right, and that ratio is what stabilizes the whole population.

The second way, which I find even more fascinating, is the behavioral mixture.

In this scenario, every single individual is capable of playing both strategies.

The ESS dictates that the individual adopts a mixed strategy.

In each contest, they essentially flip a metal coin.

And they decide to play hawk with a seven out of 12 probability.

And dove with a five out of 12 probability.

So the animal isn't fixed, it's choosing its aggression level on the fly based on a genetic predisposition to favor the hawk move just a little more often.

And this is key, that choice has to be truly random.

It must be unpredictable.

Think about what would happen if it were predictable.

A hawk would exploit it instantly.

Of course, if a hawk could detect that you always play dove after three hawk moves, it would just wait for your dove move and pounce.

Selection strongly favors unpredictability.

The opponent has to have no way of guessing which strategy you'll use in that moment.

This foundational hawk dove result brings us immediately to section three and the really profound difference between what is stable for the individual and what is optimal for the group.

We established that the ESS state that seven 12 hawk mix is stable.

But let's calculate the average individual payoff in that population.

When you average out all the payoffs in that seven 12 hawk, five 12 dove population, the average individual payoff comes out to only 6 .25 points.

That is painfully low.

Especially when you compare it to the first scenario we calculated.

The all dove conspiracy.

Yes, the all dove conspiracy where everyone cooperated through ritualized combat yielded an average payoff of plus 15 for every single individual.

So we have a perfect illustration here.

Evolution does not favor the most successful state.

The one that yields the highest group benefit.

It favors the most stable state.

The one that's immune to selfish exploitation.

So if the all dove group is more than twice as successful for every single individual,

why doesn't evolution just maintain it?

Like a group selectionist might argue.

It's because of the problem of treachery from within.

The all dove conspiracy, while it's highly profitable for everyone in it, is fundamentally unstable.

It's wide open to abuse.

In that plus 15 environment, a single mutant hawk is scoring plus 50 every single time.

That difference from plus 15 to plus 50 is a colossal selective advantage.

The moment that first hawk gene appears, there's nothing in the dove population that can stop it from spreading.

The gene pool is just too rich for that kind of treachery to be resisted.

The hawk thrives, it multiplies, and it rapidly drags the entire population down from that lucrative plus 15 state back towards the tough, resilient 6 .25 state.

So the ESS definition gets a little refinement here.

It's stable not because it makes life easy or pleasant for the participants, but because it's the only strategy that can successfully resist being invaded and overthrown by a superior selfish mutant.

It's immune to treachery.

It's like nature's insurance policy against cheating.

That's a great way to put it.

And we as humans, we can use our conscious foresight to try and implement these group beneficial strategies, but even we struggled to maintain them.

The text uses a fantastic analogy.

Price fixing.

Price fixing.

It's a perfect human parallel to the all dove conspiracy.

Imagine a small town with a few independent garage owners.

They realize that if they all collude to fix petrol prices artificially high, they all make a fantastic profit in the long run.

This is their version of the plus 15 dove society.

Exactly.

But the pact is constantly teetering on the verge of collapse.

Why?

Because of the irresistible short -term temptation for one individual to cheat.

One owner realizes he can get a massive immediate advantage, a quick killing, but just cutting his prices slightly below the fixed rate.

He instantly steals all the business.

His short -term profits go through the roof.

And the other owners, they see this cheating and to survive, they have to cut their prices too.

Which leads to a destructive price war that just obliterates the beneficial pact.

And then they have to use their human foresight, their intelligence to try and negotiate a new pact.

If conscious, intelligent humans constantly struggle to maintain these beneficial pacts against the immediate lure of selfish cheating.

Imagine how impossible it must be for wild animals whose behaviors are controlled by blind, non -conscious genes.

They have no mechanism for agreeing to or enforcing these pacts.

So they're forced to rely on the stable ESS state, the one that's resistant to treachery, even if it is suboptimal for everyone.

Exactly.

Okay, let's move into section four and refine the model a bit.

The simple hoktov split is useful, but real animal behavior is rarely that absolute.

We need to introduce conditional strategists.

Right, the hoktov model assumes a strategy is fixed no matter what the opponent does.

Conditional strategies introduce flexibility.

Your behavior is now contingent on the opponent's previous move.

Which is a far more accurate picture of reality.

You often see this restraint that just vanishes the moment it's provoked.

And the most important of these new strategies is the retaliator.

How does the retaliator operate?

The retaliator is sort of restrained by default.

They start a contest just like a dove using conventional ritualized threats.

What?

But if the opponent escalates into a real dangerous fight,

if the opponent plays hok, the retaliator switches immediately to hok behavior and fights back with full uncompromising savagery.

Okay, so a retaliator meeting another retaliator results in a harmless, drawn out, dove style posturing contest.

It minimizes injury for both of them.

Right, but a retaliator meeting a pure hok results in a vicious hok versus hok fight.

Okay, now let's look at a couple of other important conditional strategies that were tested in the simulations.

We've got the bully.

The bully is all bluff.

They immediately escalate, playing hok very aggressively, but only up to the point of retaliation.

The moment the opponent hits back.

The bully immediately runs away.

They do great against doves, but they crash and burn spectacularly against hoks or retaliators.

And then there's the prober retaliator.

This one's a bit more complex, a more nuanced version of the retaliator.

The prober retaliator is fundamentally a retaliator, but they occasionally try a brief small scale escalation, a probe, just to test the waters.

If the opponent ignores the probe.

They might persist in a slightly higher, more hok -like behavior, but if the opponent pushes back, they revert immediately to the cautious dove display, and they will still retaliate fiercely if the contest becomes a full blown attack.

Okay, so Maynard Smith and Price, they threw all five strategies, hok, dove, retaliator, bully, and prober retaliator into a computer simulation.

And the results were incredibly clear.

What came out on top?

Only the retaliator emerged as the sole, true, evolutionarily stable strategy.

The prober retaliator was almost stable, but just slightly less successful.

This is a profound theoretical victory, isn't it?

It explains Lorenz's observation perfectly.

The gloved fist isn't group altruism.

It's the ESS.

It's the ESS restraint as a default policy, backed up by immediate, fierce retaliation when that restraint is betrayed.

Precisely.

If you're a retaliator, you avoid wasting energy and risking injury in these meaningless fights against other retaliators.

But you're not exploitable by a sneaky hawk.

Hawks, bullies, doves, they all fare worse against a population of retaliators.

We have to remember though that the level of aggression,

the viciousness of the retaliation, is determined entirely by the payoff values in that specific ecological context.

Think back to the elephant seal bowl.

The massive harem.

The payoff for winning a near monopoly over a vast harem is astronomical.

So the fight has to be vicious.

The enormous reward justifies the high cost and the high risk of injury.

The ESS in that high stakes environment will favor massive escalation.

Now contrast that with a small insect eating bird, like the great tit.

Its whole life demands constant high frequency resource gathering.

The text noted that a great tit feeding its nestlings needs to catch prey every 30 seconds.

In that scenario, the penalty for wasting time, that minus 10 cost in our simple model, is hugely magnified.

If a fight, even a non -injurious ritual, takes two minutes, that bird has potentially sacrificed four meals for its young.

So the fitness cost of time wasted is higher than the fitness cost of a minor injury.

Right.

In their specific ecology, the ESS will favor avoiding even short ritualistic fights altogether.

The context determines the stable level of aggression.

Okay, let's shift gears a bit to a completely different type of conflict model.

This one's known as the war of attrition.

Here, the conflict is settled entirely by endurance.

Posturing, conventional threat, there's no physical fighting, and crucially, no risk of injury.

This is a pure test of will.

The competitors are basically bidding time against each other.

To win the resource, you just have to persist longer than your opponent.

The currency is solely time, and the resource is only worth so much time investment.

So what happens if everyone adopts a fixed strategy?

Say, everyone knows the resource.

A patch of food is worth five minutes of display, so everyone agrees to persist for exactly five minutes, and no longer.

That strategy is inherently unstable.

If everyone bids five minutes, they all simultaneously stop at the five minute mark.

And nobody gets the resource, or they split it.

Right, if that happens, the strategy of giving up immediately bidding zero seconds suddenly becomes highly advantageous because you suffer no cost.

And if everyone starts giving up immediately to save time, then a mutant who waits just one second suddenly wins the resource instantly for a minuscule cost.

So this theoretical oscillation from immediate surrender to maximal persistence is what your common sense might predict.

But once again, the mathematical analysis proves that oscillation is incorrect.

An ESS exists,

and it's achieved by removing the one single factor that creates the instability,

predictability.

So what is the ESS strategy for a war of attrition?

The ESS dictates that each individual has to resist for an unpredictable amount of time, which across the whole population, averages out to the true value of the resource.

So if the resource is worth five minutes, an individual might randomly persist for eight minutes or three seconds or exactly five minutes.

The essential element is that their choice of persistence time must be drawn randomly from a specific distribution, ensuring the opponent has no way of guessing when they'll quit in this specific instance.

And this unpredictability has huge evolutionary implications for display behavior,

specifically the evolution of the poker face.

Oh, absolutely.

If a competitor could detect any subtle sign, a trummer in the threat display, a flicker of a whisker, a slight lowering of the hackles that reliably indicated their opponent was about to quit.

That sign would be instantly exploited.

Ruthlessly.

Yeah.

Selection would penalize any gene that coded for betrayal of future behavior.

So evolution quickly selects for the unreadable, neutral poker face during a protracted staring match.

But why doesn't selection just favor bluffing?

How do you mean?

Well, if everyone knows that raising your hackle signals tenacity, why don't animals just always raise their hackles to bluff their way to a win?

Because neither telling the truth nor telling a lie is stable in the long run.

Let's say the vast majority of animals truthfully raise their hackles only when they genuinely intended to persist for a long time.

Then selection would favor immediate surrender when the hackles go up.

Right.

But then liars, those who bluff by raising their hackles but quitting early would thrive.

They'd achieve easy wins against the fearful majority.

And once the liars dominate the population, the effectiveness of the bluff just collapses.

Selection then favors those who ignore the display and call the bluff.

Which destabilizes the lying strategy.

The only stable outcome is the unreadable poker face.

The surrender, when it comes, must be sudden and unpredictable because any signal of intent is immediately exploited whether that signal is true or false.

We spent a good amount of time on symmetric contests where the rivals are, for all intents and purposes, identical.

Let's transition now in section five to asymmetric contests.

This is where differences between the rivals, even arbitrary ones, fundamentally change the nature of the ESS.

Right, we can break asymmetry down into three types.

Okay.

First, obvious differences like size, age, or physical equipment.

Second, differences in the value of winning.

Yeah.

So maybe an individual that already has 10 kids values a potential mate less than a younger rival who has none.

And the third type, which is incredibly powerful, is arbitrary asymmetry.

The example the source uses is the distinction between the resident and the intruder.

The fact that one individual arrived earlier than the other is a difference, sure, but it doesn't grant any inherent physical advantage in a fight.

It's arbitrary,

like flipping a coin.

Yeah, this arbitrary asymmetry can be seized upon by selection to create a simple, stable convention that settles disputes instantly.

And prevents injury and wasted time.

The most obvious, sensible convention would be if resident, attack.

If intruder, retreat.

And that is a theoretically stable ESS.

A population that sticks to this convention settles all disputes immediately.

Individuals avoid the high costs of fighting.

Any deviant, like a pure hawk who attacks a resident, runs the risk of serious injury.

And consequently achieves a lower average payoff than the majority who follow the convention, so it's weeded out.

Now, the text notes there is a second theoretical ESS convention, the reverse or paradoxical convention.

If resident, retreat.

If intruder, attack.

It sounds completely backwards.

It is.

And Maynard Smith labeled it paradoxical for a reason.

While a population could, in theory, settle into this reversed ESS if it happened to achieve a majority first, in real life, it's highly self -destructive.

Now, why is it self -destructive?

Because if the intruder always wins, the entire population would be selected to constantly avoid being categorized as a resident.

Everyone would be in ceaseless, pointless motion.

Just trying to secure intruder status.

Right, the whole concept of holding ground territoriality would cease to exist, making the resident category meaningless.

Ah, so this confirms a sensible strategy.

The convention resident wins, intruder retreats provides a massive advantage to holding ground because it makes you automatically the winner in future contests.

Which explains the evolutionary drive toward territorial defense.

The act of holding territory is an ESS that arises from the arbitrary asymmetry of arrival time.

And there's a classic illustration of this from Nico Tinbergen's work with stickleback fish.

It's such a clean experiment.

It's beautiful.

Tinbergen created a setup where he kept two rival male sticklebacks in separate glass tubes.

Both had established territories and built nests at opposite ends of a main tank.

So he then moved the two tubes and held them near male A's nest.

And since male A was the resident, he attacked fiercely.

And male B, the intruder displayed submission.

But here's the critical part.

When he moved the exact same two tubes and held them near male B's nest.

The roles instantly reversed.

Male B now attacked and male A retreated.

That's just beautiful evidence.

The fish weren't playing pure hawk or pure dove.

They were playing the precise conditional strategy.

If resident, attack.

If intruder, retreat.

The outcome was dictated entirely by location, a completely arbitrary asymmetry.

Now let's consider the non -arbitrary asymmetry of size or fighting ability.

If the larger animal always wins the contest, the simple, sensible ESS is obvious.

Run away from larger opponents.

Pick fights with smaller ones.

But again, this concept of a paradoxical ESS returns.

If the risk of injury is very, very high, the text notes the theoretical possibility of the paradoxical strategy.

Pick fights with people larger than you and run away from people smaller than you.

It just sounds completely absurd.

It does, but let's play it out.

Imagine a population made up entirely of these paradoxical strategists.

If a small individual meets a large individual, the large one runs away.

Nobody gets hurt.

If two small ones meet, they fight savagely, but the payoff is still zero injury for the majority of the population.

Now a sensible mutant appears, say an average sized one, who tries to pick on smaller rivals.

That smaller rival being a paradoxical fighter fights back fiercely.

Resulting in a serious injury risk for the sensible mutant.

So the paradoxical majority suffer zero injury while the sensible mutant risks severe loss.

Therefore, the paradoxical strategy is resistant to invasion by the sensible one.

It proves it's theoretically stable under the right high risk conditions.

That's fascinating.

And the text provides a real world example that seems to align with this bizarre dynamic.

The Mexican social spider, Icobius cuvitas.

This tiny spider shows sequential displacement.

When it's disturbed,

a spider looking for refuge will enter another spider's home.

And the resident spider, the established owner,

doesn't fight to defend it.

Instead, the resident darts out and seeks a new refuge, which might be yet another spider's home.

And it initiates a chain reaction of displacement.

So the one who should theoretically be able to defend its home, the resident instead retreats and becomes the intruder somewhere else.

The owner is effectively playing the paradoxical strategy, retreating in the face of invasion.

This complex behavior leads us to the final element of aggressive strategy.

The role of memory in shaping these ESSs and creating social structures.

Memory as an entirely new dimension to the conditional strategies.

The text distinguishes between general and specific memory.

Let's start with general memory.

This is where the animal remembers the outcome of recent fights, but it doesn't recognize the specific opponent.

The cricket is the classic example here.

A cricket that has recently won a fight gets a sort of boost in its perceived self -worth and becomes more hawkish in its next encounter.

And conversely, a cricket that's recently lost becomes more dubbit.

It lowers its expected success rate, yeah.

And this was shown by Artie Alexander's experiment where he used a model cricket to simulate a beating.

The real cricket was essentially tricked into losing.

And sure enough, it became significantly more dubbish and went on to lose several subsequent real fights.

The fighting policy is constantly being updated based on this self -assessment.

So what happens when individuals with this general memory policy are kept in a closed group?

They naturally sort themselves out into a stable rank order, a dominance hierarchy, even without individual recognition.

Winners keep winning and losers keep losing.

And it just reinforces the assessment.

And the benefit of this.

The group's hierarchy, which is purely a byproduct of these individual ESS policies, incidentally reduces the number of serious costly fights over time.

Once the crickets have learned their place, there's less need for violent high stakes contests.

And this contrasts with specific memory, which you find in animals that can recognize individuals like hens or monkeys.

For these animals, the ESS strategy is much more targeted.

Be dovish toward that specific individual who beat you last Tuesday.

But be hawkish toward everyone else.

Exactly.

Hands introduced to a new group will fight fiercely against all the new opponents until they've established a rank order relative to every other individual.

Once that dominance order, the famous pecking order, is established, fighting dramatically decreases.

Okay, we need to pause here to reinforce a key principle of the selfish gene theory.

The text warns us against thinking that the dominance hierarchy itself has some kind of biological advantage or function.

Right.

A dominance hierarchy is a property of the group.

We have to resist the temptation to assign function to group properties.

The real evolutionary drivers are the individual conditional strategies.

If I lost to X, retreat.

If I beat Y, attack.

Those are the true ESS that are selected for at the gene level.

The hierarchy is just the visible manifestation of all these stable individual strategies interacting.

Okay, so now we move to section six, the final transition.

We're gonna briefly look outside the species and then apply the ESS concept at its most fundamental level, the gene pool itself.

Let's start with interspecies conflict.

We established earlier that conflicts between species, say a lion and an antelope, are fundamentally different.

They're not competing for status or territory.

They're competing over the use of a resource,

primarily the antelope's body.

And this circles us back to the question, why don't lions hunt other lions?

Why is cannibalism among adult carnivores so uncommon?

For the same reason the pure hawk is unstable,

the danger of retaliation.

A lion is just too well equipped to hit back.

The cost of injury from fighting an equal rival, a lion versus a lion, is too high relative to the nutritional gain of the meat.

It's simply not an ESS.

They specialize on prey that is less likely to retaliate effectively.

And the antelope running away from the lion instead of standing its ground and fighting back.

That behavior is stable because the built -in asymmetry between predator and prey is just so vast that two species have evolved as specialized opponents.

The antelope that tries a stand -and -fight strategy will have lower fitness than the antelope that runs over the horizon.

The ESS between species quickly diverge, solidifying one as the pursuer and the other as the pursued.

And this sets up the grand finale, applying the ESS concept not just to bodies, but to genes within the gene pool.

The ESS concept is universal.

Wherever there's conflict it applies.

And the gene pool itself is an arena of conflict or at least intense selection.

A good gene isn't selected in isolation.

No, it's selected based on how well it works against the existing background of the other genes already present in that gene pool.

They have to be compatible.

They have to be complementary.

The text uses that fantastic analogy of a rowing crew to illustrate how individual selection leads to collective harmony.

It's a great one.

Imagine a large pool of oarsmen available for selection.

If the crew needs perfect coordination through speech and the pool is dominated by English speakers.

A single German speaker, even if he's a technically gifted rower, causes communication to break down and the crew loses the race.

So the coach, who selects blindly based only on the individual's merit, which is just winning races, will inadvertently canalize the German speaker.

The selection isn't against the German language group.

It's against the individual German speaking gene because it's incompatible with the majority background.

Selection favors conformity, leading to a stable ESS of pure English or pure German crews.

Now take the case where you need complementarity.

Okay, imagine the ideal crew needs exactly four right -handers and four left -handers for optimal balance in the boat.

If the existing gene pool is dominated by right -handers, a left -hander is suddenly at a massive selective advantage.

That left -hander is the equivalent of a dove entering a population of hawks.

Exactly.

Any boat that a left -hander joins will immediately become more stable and therefore more likely to win, which makes the left -handed gene look like a good gene.

So the coach's blind selection for individual winning merit will naturally lead the population toward the balanced four and four ESS crew.

Which appears to have been selected as a complete harmonious unit.

But we don't need to invoke selection for the whole crew or the whole group to explain that harmony.

Not at all.

It's achieved purely by selection, acting on the individual merits of the genes relative to the existing gene pool.

The long -term environment of the gene is the gene pool itself, and this pool will evolve into an evolutionarily stable set of genes.

A set so balanced and complimentary that it cannot be invaded by any new mutant gene.

And this leads to the ultimate conclusion of this chapter, which resolves the paradox of organization within a selfish system.

The stability of this ESS set of genes is what creates the illusion that the body is a single, organized, harmonious, self -regulating unit.

That organization isn't programmed from the top down.

No, it's produced by ruthless selection, acting at the level of the selfish gene.

Resulting in a set of mutually compatible and complimentary genes working together to build a successful survival machine.

So evolution may not be a steady progressive climb, but more a series of discrete quick jumps from one stable plateau, one successful ESS to another.

Whenever a mutation manages to invade and shift the entire set into a new successful equilibrium.

And while we focused our discussion today on these large scale visible interactions, the posturing of the dove or the ferocity of the hawk.

The vast majority of these ESS interactions happen invisibly within the cells of the developing embryo between the genes themselves, ensuring the stability and successful integration of the survival machine.

That brings us to the end of this incredibly detailed deep dive into aggression, stability and the evolutionarily stable strategy.

We took the highly counterintuitive world of game theory and showed how restraint in nature is driven, not by altruism, but by stability.

Ruthlessly favoring the policy that is immune to treachery.

We learned that stability often trumps group optimization and that even simple arbitrary asymmetries, like just being the resident, can be leveraged by selection to create predictable social structures like territorial defense.

And that the retaliator is the most successful social policy in a competitive world.

And as we mentioned right at the beginning, we explicitly excluded one enormous factor from all our ESS calculations today.

And its inclusion fundamentally alters all the math.

That's right.

The entire model of conflict and stability we discussed, it just breaks down when we consider close relatives.

Because they share a significant proportion of their genes.

Exactly.

If a selfish gene has to divide its loyalties and calculate costs and benefits across different bodies, its own, and those of its family, the entire ESS shifts dramatically.

Which raises the essential question for our next exploration.

What happens to conflict, to cooperation, and to the very concept of individual self -interest when kinship and shared genes are factored into the equation?

That is the ultimate test of the selfish gene theory.

Thank you for joining us for this deep dive into the evolutionarily stable strategy.

And a warm thank you from the last minute lecture team.