by Ober, Josiah
The Greek city-states for the most part failed to establish command-and-control bureaucracies. Nor did the Greek world develop an integrated master social or theological narrative that would have provided ideological support for the necessity of individual obedience to central authority. It was to correct that ideological lacuna that Plato, in the Republic, proposed to introduce a “Myth of Metals” (each resident will be told that he or she was born with a gold, silver, bronze, or iron soul and is duty-bound to act accordingly) as a foundational Noble Lie that would sustain the rule of philosopher-kings in an ideal, highly hierarchical polis. As we have seen, there was no divine king capable of commanding obedience, either in Hellas or in the individual city-states. The laws of the Greek states were, for the most part, recognized by the Greeks themselves as products of human invention. While the Greeks did see their gods as a source of justice, and certain forms of criminal behavior were thought likely to incur divine wrath, there was nothing like the full-featured ideology of necessary obedience to a divine order that helped to sustain cooperation within societies predicated on a centralized royal authority.
Outside philosophical circles, the idea of setting up an absolute ruler in the interests of promoting order was limited to would-be tyrants. Greek oligarchs and democrats alike sought to prevent the emergence of tyrants, but the Greek world had enough experience with tyrants to allow a rough natural experiment that tests Hobbes’ theory of necessary absolutism. If Hobbes was right, poleis run by tyrants should consistently have outperformed citizen-centered oligarchies and democracies and, over time and across the ecology, tyranny should have driven out citizen-centered regimes. Although the history of Greek Sicily shows that tyranny was sometimes associated with local prosperity (ch. 8), tyranny certainly did not drive out oligarchy, much less democracy (chs. 9–11), and there is reason to believe that the highest performing democracies (notably Athens and Syracuse in its democratic era, 465–412 BCE) outperformed even the highest performing tyrannies. Again, we are faced with the puzzle of why and how that could have been. One way to solve the puzzle is to dig more deeply into the behavior of the “masterless” ants.21
ANTS AND INFORMATION EXCHANGE
I suggested above that Aristotle’s theory of humans as political animals can help to explain the classical efflorescence. But Aristotle’s natural and moral account of politics lacks an adequate “micro-foundation”: an explanation of individuals’ motivation for cooperation. Nor does it describe an underlying mechanism that would account for how it is that, once motivated, many individual social insects or humans could produce complex public goods in the absence of centralized direction. In the case of social insects, modern biological science fills the gap, providing explanations of both motivation and mechanism. Ants belonging to a single nest are motivated to cooperate by their close genetic kinship. In regard to the mechanism, ants are able to achieve their complex ends because they constantly and actively exchange information with one another. Albeit these are very simple bits of information, the aggregate effect of the many information exchanges is high-level coordinated behavior much more complex than the movements of schools of fish or flocks of birds—animals that are “gregarious but not political” in Aristotle’s scheme.
The discussion of ant behavior in this section (and elsewhere in this book) is based on the work of my Stanford colleague Deborah Gordon, a leading evolutionary biologist, who directs a long-term study of the behavior of nests of harvester ants at a site in the Chiricahua Mountains of southeastern Arizona. Gordon’s research explores how ant nests function as quasi-organisms, sustaining highly complex forms of collective activity without resort to anything remotely resembling centralized control. Gordon demonstrates how the collective behavior of the nest emerges and adapts over time in response to environmental change. It does so through countless exchanges of simple bits of information among thousands of individual ants.
In the following paragraphs, the world of harvester ants, as described by Gordon, is adapted to Plato’s simile, and thus to the physical world of the Greek city-states, by situating the nests in the immediate proximity of a pond, rather than in a desert. The imaginary “Platonic pond ants” discussed below have the behavioral traits of Gordon’s desert-dwelling harvester ants. Thinking about nests of ants as an extended thought experiment in collective action, imaginatively getting down on our hands and knees to peer more closely at the miniature ant world into which Plato’s simile and Aristotle’s taxonomy invite us, highlights for us what is most historically distinctive about the city-states of ancient Greece and offers us information exchange as a basic mechanism underpinning decentralized productive cooperation.22
Ants, as a taxonomic family of something like 14,000 species, have been extremely prevalent across most of the land surface of the Earth for tens of millions of years. Today ants comprise a large, if not accurately measurable, part of the total biomass of land animals. Ants have, in short, flourished. One key to their flourishing is their social behavior: Ant nests are hives of collective activity. The obvious point is that antlike social behavior is one route (of course not the only route) to collective flourishing for social animals. Comparing Greeks to ants risks confusing analogy with explanation. Yet if we employ the analogy carefully, as a heuristic device, it can help us to distinguish more clearly why the political and social organization of the Mediterranean/Black Sea world of the Greek city-states produced a historically remarkable efflorescence.
Entering the microrealm of ants dwelling around a pond, we notice that some of the anthills that interest us are located right on the shore of the pond, others lie back a ways, but all are quite close to the water. Once we move any distance from the pond’s edge, we find that all the ant nests belong to other species. These other ant species are in some ways quite different from the pond ants with which we are primarily concerned. The nests of ants of different species differ in appearance and behavior—the various ant species go about food gathering in distinctive ways, treat their dead differently, use different means to attack their enemies, and so on. By the same token, the nests of each species are alike in many salient ways. The species that we are focused upon has adapted to the immediate environs of the pond as its unique ecological niche.23
The nests of our pond ants vary considerably in size: Each nest is inhabited by several thousand to several hundred thousand ants. All ants belonging to a given nest recognize one another as nest-mates. They interact with their nest-mates in specific ways, behaving quite differently toward all other ants. Each nest has its own more or less well-defined territory. Within that territory, the ants belonging to a given nest work cooperatively; their activity prominently includes foraging to extract resources from their environment. Because there are many nests, because the nest territories are not perfectly well defined, and because resources are limited, there is periodic violent conflict among the ants from neighboring nests. The conflicts are both intra- and interspecies: The pond ants of a given nest protect their territory against pond ants from other nests—that is, against animals that are behaviorally like themselves—as well as against ants of other species, animals that differ substantially from them.
The nests constructed by our pond ants are all superficially alike in that each has a standard physical infrastructure. Yet each nest is a world of its own, with its own history, beginning from the day its founder-queen flew away from her home nest, mated, came to Earth in a new place, and began the new colony. If all goes well, the new nest will live for dozens of years, although no given ant, other than the queen, lives longer than a year. A successful nest grows in size over time. As the nest matures, the collective behavior of its ants changes in subtle ways—most notably, the instances of violent clashes with same-species ants of neighboring nests are likely to decline.
The ants of a given nest take on very different tasks: foraging for food, properly disposing of dead nest-mates, attending to the immature ants, and working on the tunnels and other infrastructural features
of the nest. If we observe closely and manage to distinguish one individual ant from another (perhaps, as Deborah Gordon’s research assistants do, by daubing them with spots of nontoxic paint) we notice something remarkable: An individual ant takes on different tasks at different times of day and on different days. Yet their physical appearance and genetic makeup is very similar: Most of the ants of the nest are morphologically almost identical. There is only one possible conclusion: The role assumed, day by day, by each ant in the work of the nest is specified by something supplementary to its genetic makeup. Yet, try as we may, we will not be able to find any form of top-down organization in any given nest, much less in the larger ecology of the ants living around the pond. No ant ever undertakes to organize the nests around the pond into a pond-ant empire.24 Nor does an individual ant ever determine what goes on within a given nest. In each nest, a single queen lays eggs. She is the common mother of all the ants of the nest, and the nest will die soon after she does. But she does not give orders, or advice, or direction of any kind.
As Gordon has documented, the extragenetic something that determines the behavior of individual ants, and organizes a mass of individual behavioral choices into productive collective activity that is responsive to changes in the external environment, is the information exchanged among thousands of individuals, through a plethora of binary interactions. If we pay close enough attention, we see that the ants of a given nest are constantly interacting with one another. When an ant encounters a nest-mate, she will typically touch the other’s antennae with her own, rather than ignoring her or trying to bite her head off, which are the two primary choices when encountering a same-species non-nest-mate. The result of each touch is the transfer of a discrete bit of information. It is through the multitude of these individual encounters, and the information that is exchanged in them, that the seemingly highly organized activity of the nest is brought about.25
Each bit of information exchanged by two ants meeting is very simple. Yet the sum of those many bits of simple information is profoundly powerful: It conditions what happens in the nest—which ant does what and when, and thus what collectively gets done by the nest acting as a quasi-organism. The aggregation of a great many very simple bits of data (“ant now leaving nest to forage,” “ant now arriving in nest with food,” “ant now arriving at nest without food”)—in a process that in some ways mimics certain forms of machine-computational intelligence—enables the nest to adjust its collective rate of foraging, for example. And if, as a result, the ants of a nest forage at a rate that is well suited to the environmental conditions, the nest does well as a collectivity: It brings in more in the way of essential resources (e.g., food, water) than is consumed by the energy-burning activity of the foraging individuals.
It is through many individual information exchanges that, collectively, the nest “knows” what needs to be done and thus is able to respond to environmental change. It knows to change its collective behavior based on a changing environment (e.g., more or less rainfall), and it “assigns” the necessary tasks to individuals in ways that conduce to the collective flourishing of the community. This emergent and decentralized collective intelligence, all the product of a mass of very simple information exchanges, is the secret to the ants’ success.
The “Platonic pond ant” thought experiment offers a way to think about natural collective self-organization, through information exchange, as a way for an extensive ecology of beings to flourish over time in a challenging and changing environment. It suggests that there is nothing preternatural about the efflorescence of the decentralized world of the Greek city-states. I would suggest, as a working hypothesis to be tested in the chapters to come, that the ancient Greeks reproduced the ants’ process of successful decentralized organization through constantly reiterated information exchange. The reproduction may be thought of as a sort of unconscious biomimesis, keeping in mind, of course, the very different scales of time and size that were involved and the great increase in complexity of outcomes that are made possible by human reason and communicative capacity. If this information-centered hypothesis is right, the key to effective decentralized human cooperation in the context of a state is enabling a wide variety of valuable (at a minimum: accurate and pertinent) information to be exchanged with great frequency by the residents of the state. The hypothesis would be falsified, of course, if, relative to central-authority systems, citizen-centered Greek poleis tended to discourage information exchanges. The evidence of Greek history does not, as we see, support the falsification condition.
LIMITS TO THE ANTS–GREEKS ANALOGY
Even at the highest level of imaginative generality, the ants–Greeks analogy can take us just so far. We must also attend to the striking disanalogies. Start with the pond itself: Although some species of ants are able to cross small streams by creating living ant bridges, ants do not intentionally venture out onto the water by choice. By contrast, the Greeks constantly went down to the sea in ships and did business in the great waters that defined their corner of the world. They exploited the bounty of the Mediterranean and Black Seas for fish and other marine products (e.g., shellfish for high-value purple die), but even more importantly they used the sea as a means of easy transport from one port of call to another—and thus as a means of facilitating exchanges. Travel by land in the mountainous geography that defined much of the Mediterranean–Black Sea zone was notoriously difficult, slow, and expensive. But the Greeks moved readily and rapidly across the surface of the two seas, in sailed ships powered by wind and oared ships powered by the strength and skill of men. Moving goods and people over the water vastly facilitated mercantile trade and information exchange. Overseas travel enabled the Greeks to take advantage of the diversity of their Mediterranean–Black Sea world, interweaving the crazy quilt of geographic, climatic, and social microzones into a complex network of cultural communication and economic interdependency.26
Next and at a more basic level, although, as we have seen, individual ants do exchange information and act accordingly, and although the behavior of the nest changes over time, ants do not learn in ways that would enable self-conscious innovation. The ant nest has no long-term collective memory, no knowledge of its own history, no access to an accumulated nongenetic store of useful knowledge. Although the physical infrastructure of the nest persists across many year-classes of ants, each year-class is on its own; there is no inherited wisdom to call upon; no narratives of past successes or failures. Each year-class makes do with its genetic inheritance and with the emergent properties of simple-information exchange. The experiences of each successive year-class die with it. This makes explaining behavioral changes in the collective behavior of a nest over time a major puzzle for students of ant behavior. But there is less mystery about how and why the behavior of a Greek polis changed, sometimes radically, over the generations or even within a single human generation, through the iterated processes of information-based innovation and learning.27
Innovation, in the world of the ants, is dependent upon the standard evolutionary mechanism of random mutation and adaptation. Unlike groups of humans, ants cannot innovate culturally. By contrast, human culture may change quickly, based on immediate experience, on a new idea, or on the interpretation of historical experience. The citizens of a Greek polis learned from the accumulated historical memory of their polis. They also learned from the historical experience of other city-states, and they borrowed and adapted institutions and cultural traits accordingly. Moreover, they learned from their non-Greek neighbors in ways that proved to be profound and persistent. Finally, the unique human capacity to use reason and to communicate complex ideas through language potentially enables very extensive forms of cooperation among as well as within communities. Although ants engage in complex forms of social cooperation with nest-mates, same-species ants of different nests are unable to cooperate on common projects that would be mutually beneficial to several nests, nor can they unite against common threats. In short, communities
of ants cannot know, nor can they make, nor can they make use of their own histories, as communities of humans can and do—for good and ill.
The human capacity to employ information of complex kinds, historical as well as current, to innovate and thus to drive big changes over short time horizons explains differences in ant and human timescales: The rise and fall of the polis ecology happened over a span of hundreds of years. Ant development must be measured across the span of millions of years. Nevertheless, if the analogy holds (albeit at a very high level of generality), the information-exchange-driven collective social behavior of ants will help us to understand how the decentralized ecology of Greek poleis might have produced a remarkable efflorescence in the absence of the centralized organization that, as we have seen, is often taken as the necessary condition of human flourishing.
Before that conclusion can be accepted, however, we still need to explain motivation: Why is quasi-antlike cooperation at scale chosen by the individuals who make up human communities? Hobbes supposed that, absent a third-party enforcer, humans could never manage to cooperate at scale.28 Even Aristotle, despite his teleological naturalism, declined to attribute human cooperation in complex communities simply to the fact of natural sociability—as we have seen, he supposed that law and education were essential to the successful maintenance of the kinds of social cooperation that conduced to human flourishing. The motivation (as opposed to the operational mechanism) of cooperation among ants is now regarded by natural scientists as unremarkable, insofar as all nest-mates are close relatives, and, as such, share a sociobiological “genetic interest” in collective flourishing. Moreover, lacking individual strategic rationality, ants have no way to distinguish individual from collective flourishing. If the ant analogy is to be of value, we need to explain how cooperation in a Greek polis, or even between poleis, might be well enough motivated to enable the mechanism of information exchange to gain traction among many individuals who were not closely related to one another genetically and who were quite capable of distinguishing individual from common interests.