ONE The Castes

Movement One · The Colony

Chapter 5

The Castes

8 min read · 1,862 words


A young ant, freshly emerged from its pupal case, is pale and soft. Its cuticle has not yet hardened. For the first days of its life it does not leave the deep chambers of the nest. It stays near the brood — the eggs, the larvae, the pupae from which it itself recently emerged — and it tends them. It cleans them. It moves them between chambers as temperature and humidity shift. It feeds the larvae from its own body. It does almost nothing else.

It was not assigned this work.

No ant inspected the new arrival and directed it to the nursery. The new ant found itself where new ants find themselves — deep in the nest, among the brood, surrounded by other ants doing brood care. And it did what the ants around it were doing.

The work is not trivial. A nest is a column of chambers at different depths, each at a different temperature, and tending brood means knowing — without being told, without a thermometer — which chamber the eggs belong in this hour. The young ant does this. It does it well. And it does it for no reason it could give, because it has never done anything else and there is nothing else around it to do.

A caste is not a posting. No one hands it out.


Watch the same ant over the weeks that follow.

It does not stay in the nursery. As it ages, it moves outward. It spends less time with the brood and more time in the chambers around them — maintaining the nest, moving soil, processing food brought in from outside, attending to the queen. Later still it patrols the interior corridors and the regions near the entrance. And at the end of its life, in many species, it becomes a forager. It leaves the nest entirely, into the heat and the danger, to find food and carry it home.

The work an ant does tends to track its age. The youngest ants do the work deepest in the nest. The oldest do the work furthest from it. The ant migrates outward across its own lifetime, from the brood chamber to the desert floor, and the work it does migrates with it.

There is a logic to the direction. The work furthest from the nest is the most dangerous. A forager's life expectancy, measured from the day it begins foraging, is short. Often a matter of days. So the colony sends out the ants it can most afford to lose. An ant that is going to die soon anyway is sent to do the dying work. An ant with weeks of brood care still in it is kept safe in the dark.

No ant decided this. There is no actuary in the nest computing the expected return on each worker's remaining days. The pattern emerges from a single fact: the work an ant does follows its age, and age moves in one direction.

This gives the colony something a fixed roster could never give it: every worker is a forager-in-waiting. The colony does not maintain a reserve of trained foragers held back for emergencies. The reserve is the whole population, each ant somewhere along a single track that every ant is travelling.


The word caste can mislead.

In some species, there are castes in the strict physical sense. A soldier is built differently from a worker — larger, with a heavier head and stronger mandibles. Its body was committed to its role before it hatched, fixed by how it was fed as a larva. The soldier cannot become a forager any more than it can become a queen.

But the ant tending brood and the ant foraging outside may be physically identical. They may have hatched from indistinguishable eggs. Nothing in their bodies commits them to their work. The difference between them is where they are, how old they are, and what they have recently been doing.

This is the difference between a body and a state.

A physical caste is a body. It is decided once and does not change. The ordinary worker's caste is a state, a condition the ant is currently in, which can change, and does. The ant is a nurse the way a person is asleep: by circumstance, not by construction.

And like any state, it can be read off from a few variables.

The ant's age is one. The hormones circulating in its body are another — chemical levels that shift as it matures and that bias it toward one kind of work or another. Its position in the nest is a third; an ant deep among the brood encounters brood, and an ant near the entrance encounters the outside. And the rate at which it meets other ants is a fourth.

These variables push on each other. Age shifts the hormones. The hormones shift where the ant tends to go. Where it goes determines what it encounters. What it encounters feeds back on its state. The ant does not weigh these inputs. It is carried by them, the way a leaf is carried by a current it does not perceive.

The encounter rate is the variable that lets the colony talk to the ant without anyone speaking. An ant moving through the nest meets other ants constantly, brushing antennae, reading in those brief contacts something about who the other is and where it has been. A forager that has been out in the sun smells different from a nurse that has been deep in the brood. So when an ant counts its contacts, it is sampling the composition of the traffic around it — how many of the ants it is meeting are doing what.

No ant takes a census. Each ant feels only the small stream of contacts that passes through its own patch of nest. The encounter rate tells the ant where it is without the ant ever knowing where it is.


Suppose a predator takes the foragers. A bird, a lizard, a horned toad parked at the entrance picking off ants as they leave — and over an afternoon, a large fraction of the colony's foragers are gone.

A planner would now face a problem. The roster has a hole in it, and the ants qualified to fill it have just been eaten.

The colony does not face this problem, because there is no roster.

What happens instead is that the younger ants accelerate. The middle-aged ants, the ones patrolling the interior, begin foraging sooner than they otherwise would. The ants behind them shift outward to fill the space the patrollers left. Within a day or two, the colony is foraging again, with younger ants doing the work that older ants were doing before the predator came.

How did they know to do this?

They did not know anything. The ants that used to encounter returning foragers now encounter their absence — fewer contacts, a thinner stream of the chemical traces foragers leave on each other and on the nest. That change in encounter rate pushes on the state of every ant that experiences it. And the ants near the threshold tip over it sooner.

The hole in the roster is filled not because anyone noticed the hole, but because the hole changed what every nearby ant was sensing, and the sensing changed the state, and the state changed the work.

The same thing runs in the other direction. Remove a chunk of the ants tending brood, and foragers can revert. Ants that had moved all the way out to foraging will move back in and return to brood care.

An ant that was a forager becomes a nurse again.

If the role were a body, this could not happen. A soldier cannot un-grow its head. But states can be re-entered. The forager and the nurse were one ant in two conditions, and the condition followed the colony's need — not because the ant tracked the need, but because the need changed what the ant encountered.


There is no manager in the nest. No ant holds a list of which jobs are filled and which are open. The closest thing to a central figure, the queen, lays eggs and does nothing else; she has no idea how many foragers are out, and she would not act on the number if she had it.

And yet the colony maintains, at every moment, a distribution of workers across tasks that is roughly appropriate to its situation. More foragers when food is abundant. More nurses when there is brood to raise.

The allocation is real. The allocator is not.

Change the surroundings and you change the states. Change enough states and you change the colony's distribution of labor. The information about what the colony needs is held as a pattern of encounters, distributed across thousands of bodies, and each body responds only to its own small slice of that pattern.

A human organization divides its labour with a personnel department, job descriptions and a manager, and it takes weeks to fill a vacancy. The colony fills its vacancies in a day, because it never had positions to fill. In the organization, the knowledge of what is required is held by a manager, and the work is done by others; the gap between knowing and doing is where the delay lives. In the colony, the thing that tells an ant more foragers are needed is the same thing that makes it forage. To sense the need is already to begin meeting it.

A caste, then, is a thing the substrate does to an ant.


Watch a small team that works well. No one wrote down who does what, and yet the work is divided. One person has drifted into handling the things that go wrong. Another has become the one who talks to outsiders. A third holds the institutional memory — the one you ask when no one remembers why a thing is the way it is. None of them applied for these roles. They settled into them, over months, out of who was near what and who happened to pick up the thing that was dropped.

And when one of them leaves, someone else slides into the role, often someone who did not do that work before, because the work is still there to be done and the gap changes what everyone around it encounters.

Watch any organization that has lasted. Its roles are realer than the people in them. People arrive, occupy a role, shape it a little, and leave; the role persists, waiting for the next occupant. The org chart is usually written after the fact — a description of an allocation that already happened, dressed up as the cause of it.

The ant tending brood did not choose the nursery. It found itself among the brood, and it did what the conditions made it do, and when the conditions changed it did something else.

So does almost everyone reading this.

6 of 25

100 Million Years Ahead

Prologue

  1. One Ant, August 1993

Movement One · The Colony

  1. 1Brain or Colony?
  2. 2What the Ants Are Doing
  3. 3How the Ant Decides
  4. 4The Pheromone Trail
  5. 5The Castes
  6. 6How a Colony Survives a Decade
  7. 7The Queen Is Not in Charge

Movement Two · The Architecture

  1. 8The Six Things Every Colony Has
  2. 9The City
  3. 10The Market
  4. 11The Scientific Community
  5. 12The Body
  6. 13The Brain
  7. 14The Language
  8. 15The Ledger
  9. 16Why the Pattern Holds

The Hinge

  1. 17The Two Materials

Movement Three · The Implications

  1. 18What Superintelligence Actually Is
  2. 19The Ceiling of the Single Model
  3. 20Alignment Is a Substrate Property
  4. 21What Civilization Already Is
  5. 22The Next Hundred Million Years

Epilogue

  1. A Note on Reading

Apparatus

  1. Notes on Sources