Chapter 8: Communication: Basic Principles

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.

This is where we really get into the weeds of some foundational texts so you can walk away not just with the basics but with the specific details that

and today we are taking on a genuinely huge chapter, one that really sets the table for all complex social behavior in biology.

We are diving into chapter eight of E .O.

Wilson's landmark work, Sociobiology, and the chapter is Communication, Basic Principles.

So Wilson's goal with this whole book is to try and synthesize the just the staggeringly diverse world of social life through a biological and evolutionary lens, and our mission today is to break down his very formal, almost quantitative framework for how organisms talk to each other.

This is where he draws the lines and defines the rules.

And that framework is everything.

I mean this chapter isn't just a list of animal sounds, it's asking a central question.

How can all these different social behaviors, we're talking everything from ant trails to you know, primate politics, how they be explained and measured using the principles of adaptive signaling.

So by grounding it in biology and math right at the start, he basically anchors everything that comes later.

Exactly, it gives it a necessary biological and mathematical reality.

We have a lot to get through.

We're going to start with the very precise biological definition of communication, and we're going to spend a good amount of time on the great divide, really what separates human language from even the most advanced animal systems like the honeybee waggle dance.

And from there we'll get into the core mechanics, things like discrete signals, graded signals, composite signals, and then we'll finish up with the really fascinating but challenging world of information theory, trying to actually measure communication in bits.

It's a real progression then from defining what communication even is to trying to put a number on it.

That's the journey.

Okay, let's unpack this.

So to start, we really need Wilson's definition because in a field that can get really fuzzy with linguistic theory and all that, he just, he cuts straight to the biological reason.

So what is it?

What's communication in this adaptive sense?

So his definition is very specific.

It's any action from one organism.

Or even a cell.

Or even a cell, right, that changes the probability of another's behavior.

And here's the key part, right?

This is the key.

In a way that is adaptive for one or even both of them.

That last part, adaptive to either one or both, that's the entire engine of this whole chapter, isn't it?

It is.

It means we're only looking at signals that have been shaped by natural selection because they somehow improve the fitness of the sender or the receiver or both of them.

It's the only thing that matters.

Communication is really the relationship between the signal and the response.

Exactly.

It only happens if that signal changed the odds of the receiver doing something.

If the behavior was going to happen anyway,

then, you know, nothing was really communicated.

And this definition is powerful because it lets us just throw out a bunch of situations that feel like communication, but don't pass that adaptive test.

Like the classic example, a lion hunting a gazelle.

Right.

I mean, the lion's action definitely alters the gazelle's behavior.

It runs for its life.

Of course.

But it's not communication in this framework.

The lion isn't signaling to build a relationship.

The attack is the endpoint.

And it's the same with simple perception.

You know, an animal just wanders by a hidden predator.

The predator sees it, its own behavior changes.

It gets ready to strike, but the passing animal has no idea.

Exactly.

That's just perception influencing the observer.

It's not communication between the two of them.

Wilson also brings up JBS Haldane's principle, this idea that communication is usually really, really energetically efficient.

Yeah.

Haldane's rule of thumb was basically that a tiny little signal, something that costs almost nothing, should trigger a much, bigger response in terms of energy.

It's a good way to spot a truly evolved signaling system.

And pheromones are the perfect example of this.

Oh, absolutely.

I mean, an ant releases the tiny microscopic puff of a chemical.

The energy cost is basically zero.

And that can trigger a mass migration of thousands of other ants, a huge expenditure of collective energy.

The ratio is just completely skewed.

But it's not a hard and fast law, which is an important thing to remember.

Where does it get physical?

So two animals in a serious fight,

they've kind of stopped communicating and just started fighting.

The energy cost of the signal, the attack and the response, the counter -attack are both really high.

And Wilson gives this great human example, the abrazo.

Yeah, the congratulatory hug.

I mean, lifting your friend off the ground in this big, expressive hug is definitely communication.

It's signaling joy and warmth.

But it takes way more energy from you than the response from your friend, which might just be a smile or a pat on the back.

Exactly.

So we communicate adaptively all the time in ways that violate that simple energy rule.

The lines get even blurrier when we get down to microorganisms.

He calls this the gray zone.

Right.

Take bioluminescent bacteria, photobacterium.

They only light up when there are a lot of them around.

So a density thing.

It's a density thing.

They secrete this activator substance.

And once it hits a certain concentration in the water, it tells all the other bacteria the same strain to turn on their light producing enzyme.

So they all light up together.

So is that communication?

I mean, one cell is altering the physiology of another.

Wilson basically says it's a gray zone.

You can call it communication or not.

It's really just a matter of convenience.

It's a physiological synergism, but it really blurs that line.

And then there's an even more complex example, the grass grub beetle.

Oh, this one is fascinating.

The female beetle releases a sex attractant, but she doesn't actually make it herself.

So where does it come from?

It's made by symbiotic bacteria that live inside her glands.

She then releases the chemical that the bacteria made to attract the male.

So who is communicating with whom?

The bacteria.

The beetle.

And Wilson says asking that question is frivolous.

The real point is that communication is the adaptive relationship between the sender and the receiver, no matter how complex the channel is.

The bacteria are just a necessary part of the machinery.

It all comes back to the adaptive outcome.

So with that biological foundation set, we have to talk about the elephant in the room, our language.

It is the great dividing line in the evolution of communication.

It really is.

I mean, we use human verbal language as the yardstick to measure every other animal system on the planet.

And when you do that, you see just how profound their limitations are.

Our system has all these unique features.

We have an arbitrary vocabulary words that have meaning because our culture decided they do, and we learn them.

It's not genetic.

And that, crucially, leads to the potential for infinity.

We're not stuck with a fixed set of words.

If we come up with a new concept, we just invent a new word for it.

Wilson uses the Google, the number one with the hundred zeros.

We just made up a word for an abstract idea.

No animal can do that.

And on top of that vocabulary, we have syntax.

The order of the words changes the meaning completely.

The man bit the dog is very different from the dog bit the man.

Right.

We also have metal language.

We can talk about our language.

And maybe the most important thing, we can project the unreal.

We can lie, we can tell stories, we can speculate about the future.

To really see how constrained animal systems are, Wilson dives deep into what is often called the peak of animal communication, the waggle dance of the honeybee.

Yeah, this system is a true feat of biological engineering.

A worker bee finds a great patch of flowers.

She comes back to the hive and she needs to tell everyone else where it is.

So she performs this repeated figure eight pattern on the honeycomb.

And all the actual information, the real juice is in that straight part of the figure eight.

That's it, the straight run.

And while she's doing this, she's wagging her abdomen like crazy about 13 to 15 times a second.

And there's this audible buzz from her wings.

What's so amazing is the symbolism.

If she's outside on a horizontal surface, she just points the straight run directly at the food source using the sun to orient.

Simple enough.

But inside the hive, it's pitch black and she's on a vertical wall.

So the sun is gone.

The sun is gone.

So she replaces it with gravity.

She performs this amazing angular transformation.

The straight run now points at an angle away from straight up that is the exact same angle as the food sources from the sun outside.

So if the food is 30 degrees to the right of the sun.

The run is 30 degrees to the right of vertical.

It's this complex symbolic map, but it is completely fixed.

It's hardwired.

And the dance also tells them how far away it is.

It does through duration.

The farther away the food, the longer she does that straight run.

And we have specific data for one type of bee.

A one second run means about 500 meters.

A two second run means two kilometers.

And the other bees are pretty accurate.

They usually land within about 20 % of the correct distance.

It seems so advanced, but Wilson really stresses the limitations.

What are those rules that make it so different from our language?

Well, first, the signal isn't arbitrary.

It's a reenactment.

The straight run is like a miniature version of the flight path.

It's a direct representation, not an abstract symbol we just made up.

And second, the rules are genetically fixed.

The bee can't just decide to use that 90 degree angle to mean danger today instead of direction.

And this is the critical part, the part we can measure.

The messages are not infinitely divisible.

Because of, you know, biological noise and errors, only a tiny amount of information actually gets through.

Okay, let's break that down because it's so important for understanding animal limits.

The math shows only about three bits of information for distance and four bits for direction.

What does that mean in, like, plain English?

Okay, so four bits.

You use the formula two to the power of H, where H is the bit.

So two to the power of four is 16.

It means the bee can effectively signal direction on a compass that only has 16 points.

North, north, northeast, northeast, and so on.

You can't get any more precise than that.

And three bits for distance would be two to the power three, which is eight.

Exactly.

So the bee dance is basically giving directions using an eight level distance scale and a 16 point compass.

It's amazing for an insect, but it is so, so constrained.

There's no room for novelty or abstraction.

So that really sets the stage for how all other non -human signals are structured.

And Wilson splits them into two big categories, discrete and graded, or, you know, digital and analog.

Let's start with the discrete or digital signals.

These are your simple on or off signals, yes or no, present or absent, friend or foe.

And they come about through this process he calls typical intensity.

What does that actually mean in terms of how the signal evolves?

It just means it gets more and more stereotyped, like the intensity and the duration become less variable over time.

No matter how excited or motivated the animal is, the signal always looks the same.

It's built for absolute clarity.

Perfect for recognition or simple binary choices.

What are some of the classic examples?

The three spine stickleback is a great one.

The male gets this steel blue back and bright red belly.

It's an invariable pattern.

It just says, I am a male stickleback.

I'm ready to court.

The signal is either there or it's not.

Like the ritualized preening in ducks.

Exactly.

The male mandarin duck does this thing where he whips his head back and points right at this bright orange patch on his wing.

That specific precise action is the courtship signal.

It doesn't matter how fast he does it.

The signal is just the action itself.

And fireflies might be the clearest example of all.

Oh yeah.

Wilson has this great figure showing nine different species of Photinus fireflies.

Each one has its own unique, completely invariable flashing pattern.

A sequence of light pulses and dark pauses.

The female will only respond if she sees the exact pattern of her species.

There is no middle ground.

And in total contrast to that, you have the graded or analog signals, which evolve specifically to increase variability.

Graded signals are all about scale.

Their intensity, their duration.

It all correlates directly with the animal's internal state.

How motivated it is or what it's about to do.

The meaning of the signal is proportional to its intensity.

We saw a little of this with the bee dance, where the liveliness of the dance signals the quality of the food.

But where do we see the clearest examples of this kind of motivational grading?

Aggressive displays.

They're perfect for this, especially invertebrates.

Look at the rhesus monkey.

A low intensity threat might just be a hard stare.

But as the hostility ramps up, it starts adding components.

So it's not just getting louder, it's actually getting more complex.

Exactly.

It might go from a stare to an open mouth.

Then it adds head bobs.

Then some specific sounds, maybe slapping the ground.

And finally a lunge.

So the receiver can literally read the probability of an attack by seeing how many components are in the display.

And this happens all over the animal kingdom.

Oh yeah.

Squirrels go from a slow tail wave to a violent twitch.

Birds fluff up their feathers to look bigger.

Fish spread their fins.

Lizards inflate their crests.

The message is always the same.

The more intense the display, the more likely the action that follows.

And this can happen by just increasing the power, like volume or more subtly, by adding new elements to the signal itself.

The mobbing call of the European blackbird is a fantastic example of that.

If you look at a spectrogram of its call, you can see that as the situation gets more urgent, as the threat gets bigger, the bird literally adds higher frequency notes to the call.

It's compounding the meaning into a richer, more urgent analog signal.

And that progression from low to high intensity often has a mirror image.

This brings us to a really foundational idea from Charles Darwin, his Principle of Antipasys.

This is one of the most sort of intuitive and elegant ideas in animal behavior.

Darwin noticed that when an animal reverses its intention, say, from aggression to submission, it tends to reverse the physical posture of the signal.

The two displays are literally opposites.

His famous description of the dog is the perfect example.

It is.

Think about an aggressive dog.

Everything is up and rigid, right?

Stiff walk, head high, tail straight up, hair bristling.

Totally tense, ready to go.

Exactly.

And then when it sees its owner and gets friendly.

Everything just reverses.

Completely.

The body sinks down, it gets all wiggly and fluid, tail drops and wags, hair lies flat, ears go back.

It's signaling its opposite intention with an opposite posture.

And we see this everywhere.

Yep.

Goals.

Aggression, stretching the head forward, ready to peck.

Appeasement is turning the head a full 90 degrees to the side, breaking that line of sight.

But Wilson warns us that it's not a perfect rule.

Evolution is an opportunist.

Right.

Appeasement sometimes brings in totally new things, like grooming in primates, which doesn't really have an aggressive opposite.

And you can't assume a posture means the same thing everywhere.

A dog showing its belly is submissive.

Some shrews do the same thing, to signal hostility, because that's how they fight.

It's all about what's adaptive for that species.

Let's shift a little bit from the internal logic of a signal to the external need for

privacy or clarity.

Signal specificity.

Yeah.

In some cases, a signal has to be absolutely unique.

For insects, this can get, well, it gets kind of unbelievable.

The female silkworm moth is the classic example.

She releases one chemical, bombicol, to attract males.

And the potency is just staggering.

It really is.

A male will start searching when there are as few as 14 ,000 molecules of this stuff, a cubic centimeter of air.

And to trigger a nerve cell on his antenna, it only takes a single molecule.

That cell is tuned to literally nothing else.

The male is, like Wilson says, a sexual guided missile.

And that has huge implications for how new species can even form.

It does.

It means a tiny genetic mutation that changes the pheromone molecule a little bit, plus a matching mutation in the receptor, could basically create a new species overnight.

They'd be reproductively isolated.

And there's real evidence for this.

Oh yeah.

Studies on bryo -tofa moths found their sex attractants only differed by the geometric shape, the isomer of a single carbon atom.

A tiny structural flip, but it's enough to keep them separate.

And the ultimate case is using a blend of chemicals.

Yes.

Two other moth species use the exact same two isomers, but in different proportions.

Just changing the ratio of the blend is enough to ensure they don't interbreed.

It shows how tuned these systems can be.

So that's high specificity for privacy.

But what about when privacy isn't the goal?

Then you get low specificity.

The best example is alarm calls.

Ants, bees, termites.

They use all sorts of simple chemicals for their alarm signals.

And these chemicals often work across different species.

So a honeybee's alarm signal can also alarm a nearby ant.

Exactly.

Because there's no evolutionary advantage to keeping an alarm private.

If there's a predator around, it's good for everyone to know about it and panic.

And we see the same pattern in vertebrates, this idea of contextual specificity.

Right.

If the goal is isolation, like preventing hybridization, the signals are really elaborate and unique.

Think of bird songs or courtship displays.

But if the goal is cooperation, the signals get more generic.

Exactly.

The mobbing calls small birds use to drive off a hawk are really similar across lots of species, so they can all team up.

It's the same in monkeys.

They have unique calls for spacing out their own groups, but their general alarm calls are pretty similar.

The signal's complexity always matches its adaptive function.

So we've established that while human language is infinite, animal systems are very constrained.

And this brings us to a really striking idea.

Signal economy.

The fact that most animals have a really small number of distinct signals or displays.

It's a key finding.

Wilson defines a display as any behavior that evolved specifically to send information.

And when you actually count them up for a given species, the number is surprisingly small.

Even for really highly social animals.

Yes.

If you look at the data in his table, even for complex vertebrates, the range is only from about 10 displays in a bullhead fish to a maximum of 37 in the rhesus monkey.

The number only varies by a factor of three or four across all these different species.

And social insects with their huge colonies, they're in the same ballpark.

They are.

Ants and bees usually have between 10 and 20 known signal categories.

It's a really consistent pattern.

So why?

Why the strict economy?

Why not evolve a hundred different displays?

Well, this is where Moynihan's hypothesis comes in.

The idea is that this range, 10 to 40 displays, is probably the most a typical animal brain can handle efficiently in the middle of a fast paced social interaction.

It's a cognitive bottleneck.

After a certain point, the risk of misinterpreting a signal is higher than the benefit of adding a new one.

A cognitive sweet spot.

And you can actually see some interesting parallels in humans, even with our infinite language.

You can.

I mean, our nonverbal signals, gestures, facial expressions, our repertoire is probably in the range of 150 to 200.

It's in the same order of magnitude, just a bit higher.

And the basic building blocks of language too.

That's the parallel to phonemes.

Most languages are built from a set of about 20 to 60 basic sounds.

And that might be the maximum number of simple vocal sounds the human ear can efficiently tell apart.

It's structurally similar to that 30 or 40 display limit in animals.

There are even parallels in how often they're used.

Yeah, it's like Zip's law.

The most complex, most stereotype displays in primates, like a chimpanzee drumming on a tree or a gorilla beating its chest, are also the rarest.

It's the same way that in English, longer, more complex words are used less frequently.

It's all about economy.

So given this strict limit of, say, 10 to 40 displays, how on earth do animals manage the complexity of social life?

They must have ways to enrich the meaning of that limited vocabulary.

They do.

They manipulate the physics of the signal in both time and space.

And a great way to understand this is with the Q over K ratio model for chemical signals.

Okay, let's define the variables, Q and K.

Q is the amount of pheromone the signal releases.

K is the threshold concentration.

So the minimum amount the receiver needs to detect to have a response.

The ratio of Q to K basically defines how far the signal reaches and how long it lasts.

So if you want a signal to be really sharp and quick, like a fire alarm, you want it to fade out fast.

How do you do that with Q and K?

To shorten the fade time, you need a low Q over K ratio.

So you can either release less chemical, that's lowering Q and T, or you can make the receiver less sensitive, so it needs a bigger hit to react.

That's raising K alarm systems and trail markers or like this.

You want the signal to disappear quickly so it doesn't give old false information.

So the fire alarm is a high K, low Q event, very intense, very local, gone in a flash.

Exactly.

Now think about the opposite.

You want to increase the signal distance.

You want to expand what's called the active space, where the chemical is still concentrated enough to be above K.

This is what you want for a sex attractant.

You could just pump out more chemical, increase Q, but that seems like it would be really costly.

It is.

It would require huge glands.

The more efficient evolutionary path is to dramatically decrease K, to evolve incredibly sensitive receptors.

By making K super, super low, the active space can be orders of magnitude larger.

That's why some insect pheromones work with just a few hundred molecules in the air.

And that sensitivity lets the sender use a tiny amount of chemical for a really long And with airborne pheromones, the receiver can then just fly upwind to find the source, which adds even more directional information to the system.

Enrichment also happens by just increasing the signal's duration, making it a continuous broadcast.

Like with anatomical structures, a deer's antlers are always broadcasting its status, or scent posts left by mammals.

The rate at which the scent decays can even tell a visitor how recently it was left.

But maybe the most fascinating enrichment device is what Wilson calls semi -tectonic communication.

What does that mean?

SEMA means sign, and tecton means craftsman.

It's communication through things that are built.

The structure itself is the signal.

So it's signaling through the evidence of work you've already done.

Exactly.

Think of weaver ants.

When one worker manages to pull two leaves together to start a nest, that little physical fold in the leaves becomes the signal.

Other workers see it, they abandon what they were doing, and they come over to help.

The beginning of the nest coordinates the whole group's labor.

Wow.

So the structure replaces a pheromone trail or a touch signal.

And we see it in other ways.

Some wasps build cells for their larva, where the inner wall has a different texture from the outer wall.

The larva uses that tactile information, that built -in sign, to know which way to face before it pupates.

And then there's the male ghost crab, which builds these elaborate sand pyramids next to its burrow.

Right.

They aren't for shelter.

They're pure signal.

They're what Wilson calls petrified display signals.

They repel other males and attract females to that specific mating spot.

The crab literally builds a message out of sand.

Let's circle back to the advantage of graded signals.

Mathematically, a continuous graded system can carry more information than a discrete step -by -step one.

Why is that?

Because the total information is a function of the logarithm of all the discriminable points along that gradient.

A continuously varying signal, like the length of the bee's waggle run, allows for all these subtle shifts in meaning.

A discrete system, like a 16 -point compass, loses all the information that falls between those fixed points.

And that gradation can even lead to a complete shift in meaning.

The harvester ant is a great example.

A low concentration of its alarm chemical just makes other ants move toward the source to investigate.

A slightly higher concentration triggers an alarm frenzy.

Just panic.

But if that high concentration sticks around for a few minutes, the ant's behavior shifts completely from alarm to digging.

The duration of the signal as a whole new layer of information.

So animals enrich their limited vocabulary with physics and architecture.

But the real power comes when they start combining signals, creating composite signals.

Right.

You have three simple signals, A, B, and C.

You can theoretically create up to seven different messages by combining them in different ways.

This is how you get complexity from a small set of actions.

The zebra is a good example of how they mix discrete and graded signals.

Yeah.

Their ear position is discrete.

Ears back means threat.

Gears up means greeting.

Simple.

But the intensity of both the threat and the greeting is shown by how much they open their mouth.

That's the graded part.

So you get a fixed meaning plus a level of urgency.

And they can create a totally new message by mixing things up.

Right.

A mare who's ready to mate will show a threat face ears back, but at the same time, she'll raise her hindquarters.

So it's this mosaic of a hostile signal and a receptive signal that creates a brand new context -dependent message.

Approach me, but for mating.

And we see this in chemical communication, too.

The queen honeybee's mandibular gland is an amazing chemical factory.

It is.

It releases lots of things, but two are key.

The first one acts as an inhibitor.

It stops the workers from developing ovaries and from building new queen cells.

It's for cast control.

The second one is an attractant.

It's what makes the workers cluster around her and guides a swarm when it moves.

So the whole social state of the hive is being managed by the precise ratio of these two competing chemicals.

Exactly.

And fire ants do something similar to create a whole new behavior.

They have one chemical for general alarm and another one that's an attractant.

When a worker releases both at the same time, it creates oriented alarm behavior.

It makes the other ants rush toward the source of the danger.

The composite signal is more than the sum of its parts.

We also see enrichment through what are called orthogonal gradients.

This is where two graded signals, like threat and fear, can intersect to create these really ambiguous but very informative states.

The classic Halloween cat posture is the perfect illustration.

The cat isn't just aggressive and it isn't just scared.

It's showing high intensity threat body raised, mouth closed, and high intensity fear back arched, ears flat at the same time.

So the message is what?

It's confusing.

It's ambiguous about whether it will fight or run, but it's sending a very clear new message.

I am in a state of maximum excitement and internal conflict.

And there's some neurophysiology to back this up.

There is.

You can stimulate different centers in a cat's hypothalamus and get these conflicted behaviors.

So the animal's internal state of conflict translates directly into a distinct visual signal.

And the final way to squeeze more meaning out of a small vocabulary, and maybe the cheapest, is context.

The meaning of a signal changes depending on what's going on around it.

The Eastern Kingbird has this all -purpose call, the kitter.

It uses it whenever it's feeling indecisive.

But the meaning is completely fluid.

So give us the shift.

Okay, if a male is in a new territory, the kitter call serves to attract a female or warn off a rival.

But if he uses that exact same call when he's approaching his established mate,

it works as an appeasement signal.

The sound is the same.

The message is changed by the context.

And social insects, of course, take this to the extreme.

Oh, absolutely.

That queen bee substance, the inhibitor, inside the hive, it controls caste.

But outside the hive, during the nuptial flight, it's the primary sex attractant.

And during swarming, it's the assembly scent that keeps the swarm together.

One molecule, three totally different meanings, all depending on context.

So after all that complexity, we still have to come back to the one thing that really separates us from animals,

true syntax.

Right.

True syntax means the order of the signals changes the meaning.

AB means something different from BA.

And as we said, that has never been convincingly demonstrated in any animal system.

The closest things, like the head bobbing sequences of lizards, are just fixed signals for species recognition.

You can't rearrange the bobs to mean something new.

What you do see are Markov processes.

This just means that what an animal does next is influenced by what it just did.

The sequence isn't random.

But that doesn't create new meaning, does it?

No, that's the crucial point.

These predictable sequences don't create some new emergent meaning.

They just reflect the animal's internal motivational state, not some kind of grammatical rule.

Let's move to a really interesting type of signaling,

met communication.

This is communication about communication.

Right.

This is a signal that changes the meaning of other signals that are happening at the same time.

The simplest form is status signaling.

So a dominant rhesus monkey.

Exactly.

It's not just that he acts aggressively.

His whole posture, the confident walk, the calm gaze, the erect tail, it signals his high rank, and it communicates his own knowledge of that status.

It tells other monkeys how likely he is to attack versus retreat.

But the most familiar type is the play invitation.

For sure.

This is a signal that basically says, don't take what I do next seriously.

The dog's play bow, when it crouches down with its front legs out, is actually a ritualized part of an aggressive sequence.

But in that context, it filters the meaning of the pounce or bite that follows.

It says, this is mock fighting, not real fighting.

And primates have the play face, which looks kind of like a human smile.

And that's its function.

It's a filter that says, this is all for fun.

The normal rules are suspended.

Without that metacommunicative signal, the actions would escalate into real conflict.

And finally, we have to talk about mass communication.

This is information that can only be passed from group to group, not individual to individual.

The fire ant recruitment system is a beautiful example.

The number of workers that leave the nest to go to a food source is a linear function of the total amount of trail pheromone laid down.

It's a collective quantitative response.

So how does the colony get the number right?

The individual ants act like an electorate.

They check out the food.

And if it's high quality, they lay down more pheromone.

If it's low quality, they lay down less.

So the collective signal presented to the colony is a sum of all those individual votes.

And that controls how many workers get recruited.

It matches the response to the size of the reward.

And the Honey Bee's hive cooling system works in a similar way, but based on receiver demand.

Yeah.

The hive's air conditioning is tuned by how willing the workers inside are to take water from the foragers coming in.

If the hive cools down, they're less thirsty.

So they're less eager to take the water.

That signals to the foragers to stop bringing it in.

The collective demand controls the whole group's activity.

The complexity here really demands a more quantitative approach, which brings us to the densest part of this chapter, information theory.

This is how we try to move from just describing signals to actually measuring them.

Right.

The goal is to formalize what communication is.

We say it happens when the probability of animal B doing something, given that animal A sent a signal, is different from the probability of B just doing that thing anyway.

If the signal changes the odds, information was sent.

And the basic unit of measure here is the bit, a binary digit.

What does that actually mean for, you know, a bee or a crab?

Think of it like this.

One bit is the amount of information you need to decide between two equally likely options.

Yes or no?

Left or right?

If you have eight equally likely messages, you need three bits of information to pick the right one, because two to the power of three is eight.

But in biology, messages are almost never equally likely.

One signal might be used all the time, and another is really rare.

So for that, you use the Shannon -Weiner formula.

It accounts for those unequal probabilities.

And conceptually, why does that give you a lower number of bits?

Because if a system is really predictable,

if one signal is used 80 % of the time, there's less uncertainty.

And information in this theory is all about the reduction of uncertainty.

A highly predictable system just isn't carrying that much information per signal.

No, in a perfect world, the information sent would be the same as the information received.

But biological systems are noisy?

Very noisy.

And the noise comes in two forms.

First, there's ambiguity.

One signal can trigger multiple different responses.

It's vague.

And second, there's equivocation.

Different signals can all trigger the same response.

The receiver can't tell them apart.

So to find the actual amount of transmitted information, you have to measure all the potential information and then subtract all the information that was lost to that noise.

Exactly.

And when you do the calculations, the results can be pretty sobering.

A system might have, say, almost two bits of potential information in the signal.

But after you account for all the ambiguity and equivocation, you find that only 0 .22 bits actually got through.

So most of it is just lost to noise?

It is.

The system is inefficient, but maybe that redundancy is what makes it work in a messy real world.

But even with all that noise, when you look at the raw rate of information transfer bits per second, you find something pretty amazing.

The empirical results are startling.

Mantis shrimps, with their complex visual displays, can transmit information at a rate of up to 8 .58 bits per second.

And here's where it gets really interesting.

That upper rate for a mantis shrimp is actually inside the estimated lower range for human speech, which is about 6 to 12 bits per second.

It means that while their vocabulary is tiny, some animals can communicate essential information at a speed that's surprisingly close to our own.

When they need to be fast, they really approach our level of efficiency.

Information theory was even adapted for continuous signals, like the bee dance, by looking at the variance.

Basically how scattered the new bees are when they arrive at the target.

And the analysis confirms our earlier estimate.

Bees transmit about 4 bits for direction and 3 bits for distance.

It's like a 16 -point compass and an 8 -step ruler.

And if you compare the bee dance to a fire ant trail, you find something counterintuitive.

Wilson says the directional information of the ant trail actually increases the longer the trail is.

That seems backwards.

It does, but it makes sense when you think about the geometry.

The width of the pheromone trail, the absolute error,

stays about the same.

But as the trail gets longer, that same amount of absolute error becomes a much smaller angular error relative to the nest, which is now far away.

So the longer you follow the trail, the more precisely you are aimed at the target.

So information theory is a powerful tool, but Wilson is really clear that applying this kind of math to messy biology is really hard.

The biggest problem is what he calls primer effects.

These are signals that don't cause an immediate change in behavior, but instead change the receiver's physiology over time through the neuroendocrine system.

What are some examples of these slow burn effects?

A male ring dove's courtship bowing doesn't trigger immediate mating.

It actually stimulates the female's hormone production over a couple of days, getting her ready for nesting.

Or chemicals in mouse urine that can alter the reproductive cycle of other mice.

And how do you measure a two -day hormonal shift in bits?

The framework kind of falls apart.

It does.

It's really hard to quantify.

Another huge problem is just the sheer complexity of behavior, especially in primates.

A lot of their signals are so graded and subtle that you can't just divide them into neat little categories to plug into the formula.

So with all this noise and complexity, if there's one word that characterizes animal communication, Wilson says it's.

Redundancy.

Yeah.

Just saying the same thing over and over and over again.

Animal displays can be incredibly repetitive.

It seems so inefficient, though.

Why not just say it once and be done with it?

Well, he gives a few key adaptive reasons.

First, reassessment.

If you're in a constantly shifting social relationship, like two rivals sizing each other up, you need to constantly repeat the signals to reassess the situation at every moment.

And second,

sustained arousal.

Right.

Those slow burn hormonal changes need constant stimulation.

So you need multiple, maybe slightly different signals with the same redundant meaning to keep the physiological system primed.

Third is just insurance.

If there's a good chance your signal might get missed, like a lizard in a dense leafy forest, you have to repeat it to make sure it gets through the noise.

And finally, the redundancy might only be apparent to us.

What looks like repetition to a human observer might actually be packed with subtle contextual information.

Like the warbler who changes the ratio of his two songs depending on where he is in his territory.

To us, it sounds like he's just singing the same two songs.

To another warbler, that ratio is communicating a lot about his arousal or his confidence.

Exactly.

The signals aren't truly redundant.

They're highly specific to the context.

Wow.

That was a tour de force.

We started by defining communication as an adaptive relationship, all about changing probabilities.

We saw the profound limits of even the best animal systems, like the honeybee dance, which only carries a handful of bits.

Then we built up the framework of signals, the clear on or off discrete signals, and the nuanced scalable graded signals.

We saw how Darwin's principle of antithesis shapes opposing intentions, and how specificity can range from the microscopic precision of a moth pheromone to the generalized nature of an alarm call.

We looked at the strict economy of the animal repertoire, probably limited to about 40 displays by cognitive constraints, and how animals get around that with enrichment devices.

Manipulating physics with Q &K, building signals into the environment, and especially creating massive complexity through composite signals and context.

We spent serious time with information theory, seeing how you can actually try to measure information transfer in bits.

And we found that some animals can operate at speeds surprisingly close to the low end of human speech, even if most of that potential information gets lost to noise.

I think what really stands out is that contrast.

Human communication is characterized by its infinite flexibility.

Animal communication is characterized by maximum adaptive efficiency.

They take that small fixed vocabulary and they squeeze every last drop of meaning out of it through context and composites and physical structure.

Which leaves us with a final thought for you to chew on.

Considering all the complexity added by context and ambiguity in our own modern lives, if we were to apply Wilson's rigorous bit counting definition to, say, an average 60 -second social media exchange, two friends firing memes and short context -heavy messages back and forth, how many bits of true non -redundant information do you think are actually being transmitted?

And how much of it is just noise or redundancy that's only there to sustain the social bond?

Thank you for joining us for this deep dive into the basic principles of communication.

We hope this has given you a clear, accessible, and detailed understanding of this really crucial foundational chapter in sociobiology.