Ants and Anima

Life is something built

This is rather conjectural and a bit off the wall. That isn’t to say it is wrong, but it might have a little bit of a fever dream flavor to it. Sometimes good things happen like that; a few weeks ago I woke up in the morning and understood why the imperial fractional measuring system was better than metric. Sometimes brains just work, and you don’t know why.

In the spirit of fever dreams, I am also not editing this. Fresh from the brainpan, baby!

---

The minimum viable unit for ants is the colony.

Not the individual ant, nor a handful of random ants, but a whole colony. The number of heads that make up a colony is variable across species: generally larger ants like carpenter ants have a few hundred colony members at, while some smaller species have thousands or possibly millions. The number of bodies isn’t the important part. What matters is the distribution of types.

Ants come in four basic types, or castes:

• Workers: The vast majority, all sterile females¹, sisters and aunts of each other, that do all the basic “ant” jobs. Some species such as leaf cutters have subtypes of more specialized workers.

• Soldiers: Specialized workers, usually larger and with oversized heads to hold powerful jaw muscles. Specialized for combat, but also contribute to foraging and other worker roles². Also all sterile females.

• Drones: Males, generally winged, drones serve to provide male gametes, generally once. Short lived, drones usually will swarm out of the colony during mating season to mate with new queens to start new colonies.

• Queens: Despite the appellation, queens do not rule but serve the egg laying reproductive role. Queens typically are unique to each colony, laying thousands of eggs, and are replaced by a new queen when their fertility wanes. Female, obviously.

At least three types, workers, queen and drone, are needed for a functioning colony, and typically all four will be extant. A colony without a queen (and the larvae to replace her) cannot reproduce members and dies, no matter how many resources abound. Likewise, any number of queens can’t start a colony without a drone to fertilize them, and after laying and hatching the first brood of workers the bloated queen cannot fend for herself if they die.

Just as the ant castes can no more exist separately from each other as my fingers can exist separately from my stomach, or my stomach from my mouth. Each caste is as specialized as different organs in the body, albeit a rather simple body³.

The workers are the hands that do, digging the nest, caring for the young, finding the food, cleaning the home, and generally caring for the other parts. All things that ants do, that make a colony a colony, are done by workers, save one.

The soldiers are the fists that defend and strike the enemy. Less efficient than hands for most jobs, but particularly useful for one, they join workers in their work, and are joined in turn by workers in combat.

Queens are the reproductive organs. Creating new both new members of the body and new members of the species, queens are protected closely. As the longest lived members of the colony, often living decades compared to the handful of years of the workers, and creating the other colony members as partial copies of herself she is in many ways the colony. Yet she can and will be replaced as needed, and by herself can perform none of the functions that make the colony function.

Drones are the mobile gametes of the colony, moving out to meet other queens and start a new colony, or sometimes moving into another established colony to help provide seed for new workers. Like most mobile gametes, they are not long lived in any case.

Individual ants are thus analogous to the cells of the various organs (the queen being large enough to be quite a few cells in once place.) No individual worker is particularly critical, so long as there are enough to continue their functions en masse. Even the queen can be replaced if necessary.

Further, just like cells of the body’s organs, individual ants are extremely decentralized.

The queen doesn’t govern or provide coordination; she is no more aware of what is going on outside her chamber than any given ant. How would she even spread messages of her intentions? Tiny ant cellphones? Miniature myrmidon memos?

But then, how do the other ants know what to do?

For a species that focuses on hierarchical coordination, this is kind of a mind boggling question. Ants have tiny little brains, yet they perform a multitude of tasks, within the nest and without. As it turns out, there are not digger ants, brood nurse ants, foraging ants, janitor ants, etc.; individual ants switch between jobs during the day, and indeed at different times of life, but how do they know what to do when?

Are there middle management ants telling them what to do, coordinating activities based on some grasp of the overall needs of the colony?

No, as it turns out⁴. Each ant decides what to do all by itself. Yet, somehow, you don't wind up with every ant deciding it wants to spend a wet Thursday morning inside in the warm digging new tunnels. Apparently the little devils have a suitably simple method of deciding what to do: they count how many ants they have seen doing certain jobs, and if they haven't seen a task being done much they swap over to that⁵. If it seems like lots of other ants are taking care of larvae and you haven't seen much of anyone carrying food around lately, you go do that⁶.

Ants also use pheromones (stinky chemicals) to communicate. A forager wandering out leaves a trail behind her so she can make her way back to the nest. (This is why ant trails tend to look so winding: they don’t think to cut off corners or take short cuts but just follow it back as is.) If she finds food she will head back while leaving another trail announcing food lies at the far end of it. Other ants pick up that signal and follow it out to the food, leaving their own reinforcing trail on the way back. As more and more ants do this, more and more ants pick up the signal and go collect the food. As the amount of food dwindles, however, ants start coming back empty mandibled, and don’t reinforce the trail, which dissipates over time. Before long ants stop following it in favor of searching for new sources or following other food trails that are more powerful.

In both cases, the minimum decision making unit, the level of agency, is the individual ant. Each individual does what it thinks is best based on signals sent by the other individuals doing their thing. Cooperation occurs without the cooperating agents knowing what the overall state of the system is, or really what is the outcome of their actions at scale. A few rules about when to stop doing one job and start doing another, along with some about what to do in the presence of what pheromones, and when to produce them, and the result is incredibly complex behavior from a meta organism composed of nearly mindless moving parts⁷.

The individual ant has no grasp of the whole; so long as other ants it meets smell like it, they are part of the colony, but that’s as far as it goes. The individual decision making ant does not make decisions with the colony in mind, and cannot grasp the needs or goals of the colony as a whole. Its decision making is entirely myopic.

The individual colony, the minimum viable unit of antdom, has no grasp of individual ants, anymore than I have a grasp of what individual cells in my spleen are up to. There is no centralized decision making process giving orders to all the individual cells, or even the individual caste organs; all communication is two way, with every part affecting the behavior of every other part.

Yet the colony persists, long past the lifespan of any individual member, even the long lived queens, and has identifiable needs outside the needs of its individual cells. No worker needs an egg chamber⁸, but the colony needs them and so they are produced. The colony's health ebbs and flows, controlling more or less territory, spawning more or fewer child colonies, in ways separate from the health of individual ants, and often in ways inimical to it.

The colony is a being with existence above that of the beings that make it up, a being whose motivations, needs and goals are perhaps incomprehensible to the decision making beings within it.

Asking a random worker foraging for food what is best for the colony is probably a lot like asking a random stomach cell what is best for me: it will probably say that getting some food in it would be a good idea. No single organ makes the decisions alone, but through the constant feedback and needs of the parts, often without the parts themselves being conscious of their feedback. The ant serves an entity that it cannot fully comprehend, made up of multitudes of beings just like itself.

Humans and ants have a lot in common like that.

We also inhabit huge decentralized systems, what we economists like to call markets, where individual decision makers use limited local knowledge to make guesses about what is best to do, typically based on what other people are doing. We don’t use pheromones⁹, but instead use prices to let us know if some course of action is more worthwhile or not. We have developed limited hierarchies¹⁰, within which behavior is coordinated, but the vast majority is dictated by individuals who can't comprehend the whole.

But the close readers among you will notice I have shifted the scale of the simile. Earlier I likened the ant colony to the single human, the ants her cells, the castes her organs, the colony the whole body. But the individual human is not the minimum viable unit of humanity, only the decision making agent. The minimum viable unit is the family, the male and female pairing that allows for the creation of new humans, just as the colony can create new colonies¹¹. Yet even that is not enough. Humans, being rather more complicated beings than ants, require not just a family but a tribe, a society to really flourish¹². The optimal unit of humanity is a society of families and individuals.

Yet the relationship of the society to the individual is not so different for humans and ants.

There is no central decision making entity for the society, only a complex web of interdependencies. The society knows little of the individuals, and indeed has needs different from and often inimical to the individual. The larger the society, the more the well being of the whole deviates from the well being of individual members¹³.

The individual cannot understand the needs of society, or grasp its goals and ends. Decisions are made at the myopic, local level by individuals based on their individuals goals, motives and desires, only incidentally coinciding with the needs of society.

Society emerges from the behaviors of the individuals who serve its ends while serving their own.

That doesn’t make society any less real a being than a human made of cells, or an ant colony made of ants.

Humans in fact have words for beings we can’t quite grasp, whose motives are beyond our comprehension, yet whose goals we serve, knowingly or not. As Matt Ridley put it (roughly based on my memory): I don’t believe in a god, but if I were looking for one, the invisible hand of the market would be a good candidate.

Societies come in all sizes, large and small, some nestled within others, some standing apart. Perhaps it is no wonder then that we have words for gods and sprits, pantheons and monotheistic deities. Beings that are multifaceted and others that reflect merely an aspect of our experience.

For while the ant cannot grasp the whole of the colony, the way that all of the behaviors and experiences of a thousand other ants shapes the world it lives in and emergent goals of the whole, it no doubt can sense that there is something more than what is directly before it. Humans, no doubt perceive much more clearly the existence of society, of some extended self we are apart of, of beings emerging from our collective behaviors. Agents of whom we are a part, but which we can no more fully comprehend than my thumb can grasp the process that created the book whose page gave it a paper cut.

The parts cannot grasp all of but the most simple of wholes, and we are surrounded but a great many wholes. Call them what you like, but we are of many such beings.

1. Ok, in most species… ants, like their close cousins wasps, vary quite a bit on their theme. Some species have only queens reproducing, usually one but sometimes multiple at once. Some other species have workers that can reproduce, which are amusingly called “gamergate” types, and don’t have a queen. There is probably a type that has both queens and gamergates. Ants are a varied bunch.

2. Sometimes the distinction between workers and soldiers is very clear, and sometimes it is almost absent, either due to many gradients of workers across a spectrum of sizes, or just all workers being roughly the same size and shape. It might be more correct to group all the sterile females into a worker type, with worker specializations between combat and non-combat roles, since all will perform both types of work, just with different efficiency. It doesn’t really affect the discussion either way, so I am just going to role with the four caste division I am used to.

3. Like nematode simple, but humans are easier to imagine.

4. Having written that, I am almost certain that, someday, a species of ant with a “management” caste will be discovered. Presumably they will have evolved mandibles capable of handling tiny clip boards.

5. This is on a daily basis. Over the course of their lives most ants have a broader pattern, with young ants staying home and taking care of interior jobs, and older ants with less life to lose if they get snatched by a bird tending to do the more dangerous foraging type jobs. Again, not always exactly true for all species, and there are probably free loading ants in any colony, but there you are.

6. I find ants very interesting, but I also find the idea of being one of the researchers who watches them closely enough to figure out this is what they do horrifying. It takes all types to make a world.

7. This is why ants are such a popular topic for studying emergence by agent based modelers.

8. Except those gamergates.

9. For this at least.

10. But our managerial morphologies generally do not have clip boards for mouths.

11. Of course there would be a foot note. Some ant species have colonies that send out drones and queens that start a new colony, while others require another colony to provide the pair. So depending on the species the minimum viable unit of ants might be two colonies, an ant colony family.

Freaking ants, man…

12. And, you know, not wind up with kids with wildly varying numbers of thumbs per hand.

13. An empire can throw away the lives of hundreds or thousands of men in a failed war and survive. A city state cannot (ask Thespiae.)

Sub-footnote: Thespiae was a Greek city state that was doing pretty well, until a series of setbacks lost it much of its male population, notably losing ~700 in the battle of Thermopylae, after which Xerxes burned Thespiae down, and then the Battle of Delium against the Athenians, after which Thespiae was to become essentially a ghost client of Thebes and never was a big player again until siding with Alexander. Even then its history gets pretty spotty after that.