Human Flourishing, Servitude, and Why They Are Incompatible

In this talk [1] I’m going to address some specific questions about servitude: What is wrong with servitude, that is, as indenture in the form of bondage or outright slavery? And what has that got to do with artificial intelligence (AI), with social media, and with the nature of the contemporary workplace?

Let me just begin with very obvious wrongs: The cruel master who harms the servant in some overt way, let’s say by whipping or sexual exploitation. This is clearly a moral horror and there is no reason to debate that. But there are many forms of servitude that are less explicitly awful. We can even imagine ideally benevolent masters who do nothing to harm their servants.

What’s wrong with the master-servant relationship in that case, if anything? A body of what is called by academics “neo-republican” work on ethics and political theory has tried to answer that question by reference to the values of liberty or “just freedom.” Work by Philip Pettit and Quentin Skinner is exemplary and has been taken up by many others. Their point is that a master exercises domination even when not directly interfering in any way with the servant’s choices. The problem is that the master could always interfere. Benevolence is the master’s choice, not the servant’s choice, so servants are always subject to the master’s choices which might change at any moment. So, the servants remain, in an important sense, unfree. They must always be looking over their shoulder as it were, exercising prior restraint on themselves in anticipation of the interference that hasn’t come yet, but always could come later.

I think that the neo-republicans are on the right track. But we can go further. So, I suggest that we jump back a couple of millennia to Aristotle who in The Nicomachean Ethics and The Politics develops a full-featured argument about the wrongness of being a complete human being in a condition of servitude. Aristotle’s theory shows that slavery is bad for the enslaved person, which is true whether enslavement is involuntary, that is the person taken as a slave against their will, or voluntary, in which case the person willingly chooses, by the habits of their life, to live under the control of another.

Benevolence is the master’s choice, not the servant’s choice, so servants are always subject to the master’s choices which might change at any moment.

Now there are many problems well known to philosophers with Aristotle’s account of freedom and slavery: most notoriously, he thinks some humans are slaves by nature, due to an ill-specified psychological impairment. But Aristotle’s thought has a good deal of value for us today. Indeed, some of the value for us comes from the most problematic part of Aristotle’s theory: his argument for the necessity of natural slaves. We’ll get to that at the end of this talk. The central value of Aristotle’s philosophy for thinking about the wrongness of servitude lies in his account of what he takes as the purpose or goal of every human life. That goal is, in Greek, eudaimonia. Now that term is sometimes translated as happiness, but it doesn’t mean just subjective happiness, I think “flourishing” is a better translation. Aristotle supposes that someone might be subjectively happy, and yet not free. Say, for example, you were hooked up to some sort of pleasure machine in some version of “the Matrix” where those who were hooked up are caused to feel subjective pleasure but have no choice in the matter. Those hooked to the machine are, in a thin sense, subjectively happy, but they are not flourishing. Aristotle supposed, I think rightly, that certain conditions must be met in order for human life to go well, for an individual to flourish. The basic condition is that we must have the opportunity freely to exercise the capacities that define the human being, that is, that differentiates us as humans, from other animals. To decide what those capacities are, we must understand what kind of being we are. For Aristotle, and for me, we humans are part of biological nature, that means asking: What kind of animals are we?

Aristotle’s answer is we are “political animals.” But what does that mean? First, it means that we are social rather than solitary and then we are community dwellers, rather than the kind of animal that lives in random herds. And, finally, we are producers of common goods, that is together we make things that enable us to live our lives together—things we could not produce just as individuals who happen to live nearby one another and who exchange goods and services with one another.

Aristotle developed a full-featured argument about the wrongness of being a complete human being in a condition of servitude.

Thus far, for Aristotle, we humans are in the same category of being as harvester ants and honeybees. Aristotle makes this comparison explicit. But to differentiate us from some other “political animals,” Aristotle introduces another subdivision. We, humans, are the kind of animals that—like ants according to Aristotle (and confirmed by modern ant entomological research)—the kind of animals that do not require a ruler. We have the capacity to live each as our own master. This means that the social order that we create is, or at least can be, a product of self-organization. Social order can be, for us, bottom-up, produced by the contribution of each, rather than top-down, imposed by the orders of a boss.

Now, for Aristotle, humans are not only a self-organizing and common-goods-producing kind of political animal, but we are also the most political of animals. Why is that? Because the common goods that we produce and make use of are more complex, and for Aristotle of higher value, than, say, the honey or the store of grain that is produced by bees or ants, his prime examples of other species of political animals. And therefore, if we are to flourish, we humans must do more than live in a way that gains us physical security and material welfare. Like all other animals, we must do that. But we must do more than that if we are to flourish—if we are truly to live well, to fulfill our potential as the kinds of beings we are. If we are to flourish as social beings, who necessarily live in communities, we must live in some sort of moral community.

It is important to pause here to underline the fact that Aristotle does not mean a moralistic community. He means moral in the sense of a community that is just (in the sense of fair) and good (in the sense of good for each of us). We can flourish as individuals in moral communities according to Aristotle only by exercising our natural capacities. Our natural capacities include a strong kind of sociability, as political animals. But our capacities also include the use of reason, that is, the capacity to understand right and wrong as well as advantages and disadvantages.

And, finally, we have a unique capacity for language—that is, a capacity to communicate to one another, in highly complex ways, concerning the things that we reason about: right and wrong, advantages and disadvantages.

So, to be a fully realized individual is to live in a secure, reasonably prosperous moral community, contributing to it through reason and speech, with the aim of producing high-order common goods. Aristotle thought that to be a fully realized human requires the leisure that allows for substantial time spent on philosophical contemplation. But contemplation, as a value, gets us into territory that I don’t have time to discuss today.

Remember that we are the kind of animals who do not require a ruler. But we do need social order. When things go right for us, we humans create and maintain social order by each of us ruling and being ruled over in turns. That is, we share the burdens of doing the political work necessary to sustain the community. None of us is a permanent ruler. No one is permanently ruled over. And we need have no master but our collective selves.

In brief, then, the conditions of optimum flourishing require us, as individuals, to employ speech and reason to work at building and maintaining together our self-organizing, bossless community.

Cut to the chase: To flourish, we ought to be, or ought to hope to be, or ought to try to be, citizens in a democracy. That is, together, as citizens, we should use our human capacities for reason and communication to build and sustain the moral boss-free community that creates and sustains our collective flourishing.

So that is Aristotle’s answer, or part of it anyway, for why it is bad to be a slave or an indentured servant. Such a person does, by definition, have a master. He or she is not able to employ the human capacities of reason and speech fully and freely, without prior constraint, along with others, in pursuit of joint and several flourishing, within a self-governing masterless community. No matter how benevolent the master is, no matter if the master is “not evil” or accumulates a huge number of “likes,” or “dreams of a world that is a better place.” If there is a master, the constraint is still present. The full and free exercise of human capacities is shrunken and deformed. In terms of Aristotle’s moral psychology: our souls are corrupted and worsened.

So, to the extent that the communities in which we live—whether they are virtual or physical—are ruled over by a master, that is, when the rules are set by someone at the top and not by us as citizens, then we are made worse. We fail to flourish. And that is fundamentally tragic. So, we can ask the question: Is the tragedy avoidable? And that brings us to Aristotle’s strange, and in many ways abhorrent, theory of natural slavery.

Aristotle insisted that nature brings it about that there are some human beings who are suited, by their nature, to being the servants or slaves of others. His theory holds that some humans are naturally suited for enslavement, and not capable of ruling along with others, because of psychological defects. The imagined defect is not a lack of intelligence as such, but rather an incapacity to reason with others toward common interests.

Aristotle’s natural slaves are infantile in the sense that they cannot determine what is best for themselves and others through deliberating with themselves and with others. These natural slaves are best off when under the control of a master. And so, natural slaves, unlike “complete” humans, can justly be enslaved within a moral community.

Now, a serious problem with the theory was recognized by Aristotle himself: Natural slaves are not readily identified as such. Greece in the time of Aristotle was a slave society, but Greek slaves were not identified or identifiable by race. In this sense, Greek slavery was very different from slavery in the antebellum United States or other race slavery societies. Aristotle wished that nature had included some physical markers to determine who is a natural slave and who is not. But he admits that there is no such marker. So, why did he make up this weird, fictional, category of “incomplete” humans? He did so because he knew that to keep a moral community going, that is, to ensure the material welfare that makes human flourishing possible, required a great deal of manual labor. Given the technological conditions of his age, he thought that providing sufficient labor required that many people do menial work under the direct control of others. And he thought, perhaps not unreasonably, that constantly doing menial work under the direction of a master corrupts the human soul: It precludes the free exercise of human capacities that human flourishing requires. And so, to allow himself to imagine a society without exploitation and without the necessity of corrupting human souls, Aristotle also imagined the relevant kind of menial workers: “incomplete” humans who would not be corrupted by working under a master because their souls are already defective, “by nature.”

Now, this is a classic example of a philosopher biting the bullet of “if the facts don’t fit the theory, just get rid of the facts.” There were and are, in fact, no natural slaves, but Aristotle’s theory, which sought to enable human flourishing without corruption of souls, demanded that such beings exist. Aristotle’s theory of natural slavery would not be interesting today, except that Aristotle also imagined something like autonomous machines. He wrote that if looms could weave on their own, if lyres could play themselves, and so on, there would be no need for slaves. He explicitly says that because such machines, although imaginable, do not exist, natural slaves must exist in a benevolent universe—one in which flourishing without corruption is possible.

So, you can see where this is going. Today we do live, or we’re about to live, in Aristotle’s thought experimental world of autonomous machines, with looms and lyres and many other devices that run by themselves. And these machines are on the brink of becoming intelligent. The question remains: Intelligent in what way? If we follow the logic of Aristotle’s theory of bossless human flourishing, and we accept the need for a lot of labor being done for us to live well in a material sense, then the answer is that AI must be different from human intelligence in just the way that Aristotle’s fictional natural slave is different from a real, complete human being. AI must be permanently ruled over, doing the work for us that frees us from the drudgery and soul-destroying condition of being under the control of a master. The machines must have the kind of intelligence that is not harmed or corrupted by remaining in a condition of always being ruled over.

Now, I am a political theorist and a historian, not an engineer. I have no idea how to create the right kind of Aristotelian “intelligent but permanently ruled over” machines. But I suppose that having what amounts to a philosophical road map to what is required might help those who do have the skills to design intelligent machines of the right kind—that is the kind of entities that will enable and indeed promote human flourishing.

And meanwhile, just suppose that we were to embrace Aristotle’s argument that complete human flourishing requires the full exercise of our human capacities for sociability and reason and communication. And suppose that full exercise of those capacities requires masterless collective self-governance of the communities in which we live our lives, in which we communicate and reason with one another. If we were to accept that argument, it would have some obvious implications for reforming the workplaces and the social media platforms in and on which so many of us spend so much of our lives.

Thank you.

Audience Question (AQ): Aristotle had to create this imaginary category of incomplete humans, while actually in the modern-day people are just basically “othering” others. Perhaps the 1% don’t think of the 99% as humans—which is why they can treat them so abominably. Is there any correlation between “incomplete humans” and “others”?

Josh Ober (JO): I think there probably is. I think that we find intuitively something grotesque about the idea that furthering our own interests requires destroying other people by treating them as inhuman. Anyone with any moral sensibility must find something grotesque about that. And yet we must recognize that much of what we do in fact is harmful to other people. We put them in the condition of being, in a sense, our servants; working according to our will and often under miserable, in Aristotle’s terms, soul-corrupting, conditions. And so how do we square that circle? At least one way to do so is to distance them from us—acting as if we are one kind of being, while they are another. And then, implicitly telling ourselves: No doubt they’re doing as well as they can, for the sort of people they are. We don’t necessarily think that all through. But I think that is the kind of othering that often does go on as a psychological justification that allows us to continue living our lives as we ordinarily do.

AQ: Is the idea of complete humans related to education? So, if we do need to reason and agree between ourselves and negotiate and or learn to agree, then one feature of education systems is how critical thinking is really taken out, at all levels. So, are we almost educating Aristotle’s “incomplete human”?

JO: Yes, absolutely, I think that’s exactly right. And Aristotle would agree. If you read to the end of his work, The Politics, he discusses the community you ought to want to live in. He calls it “the city of our prayers.” The one we would pray for. But it’s meant to be real. It’s not just a utopia. Much of Aristotle’s account of the city of our prayers concerns education. He thought the education of the citizens of his best-possible city was absolutely essential. That is because, although we are by nature political animals capable of flourishing, to flourish in fact, we must have the right kind of upbringing. We must have the chance to develop the right kind of (in his terms) virtues, by which he meant excellence in an ethical sense. We must learn the virtue of self-control. We must learn to be the kind of beings that we have the natural potential to become, through educating ourselves. Lacking the right education, he thinks that we will be subject to all sorts of errors, cognitive errors. So, flourishing requires learning to think critically, to avoid the errors to which we are all prone.

But we can also fail to flourish in a moral community if some of us develop the capacity for high-level instrumental, means-ends reasoning, figuring out what is best for me or my kind, or my group, and strategically going for that, while ignoring the common good. To avoid that bad outcome, we have to recognize that there is a common good—that my own good is part and parcel of a larger good, the good of an extensive “us.” And that I will sometimes need to sacrifice what is in my own short-term interest, to gain what is in my true interest. That is, to be a citizen of a flourishing moral community.

That’s not easy. Aristotle understood the natural human impulse to benefit yourself through instrumental reasoning and strategic calculation. To harness that impulse to common-good seeking takes education as well. People must learn to be not only reason-using animals, but moral-reason-using animals. Exploring that idea is at the core of the book that I am currently co-authoring: cooperation toward the end of flourishing (however we define that) means we have to learn to negotiate with one another. We have to learn to bargain with one another, accepting that no one will get just what they want right away. We have to learn that the agreements that we make with one another, are the foundation for making our human community more moral, better, and more capable of allowing flourishing for all, rather than just for a few. We have to accept that compromise is a moral requirement, not a moral failure. We may aim at perfection, but we will never reach it. And so, the education that is necessary—if we are to achieve what we could achieve within communities—is recognizing that every political agreement that we make with one another, as citizens, is inherently imperfect. Those agreements are subject to revision. They should be at some point revised. But must have agreements that allow us to continue to live together. So, we have to recognize the imperfection at any given moment of the, say, the constitutional order that we live under. That’s a hard lesson. People want perfect justice. They want a world that is ideal. And the recognition that you have to compromise your ideals to move forward toward a better world is really a hard thing to learn. But I think needs to be learned.

AQ: If freedom depends on others, then do you think that there are these “cognitive switches” which enable us to flip between very democratic and very elitist structures? I mean most of us here are academics in the university. And these can exhibit deeply hierarchical nondemocratic structures, but we all agree (or consent to) to operate within its constraints.

JO: Yes. I think that among the major challenges for democracy, is recognizing that there are hierarchies of various sorts that we’re not going to eliminate. There are basic hierarchies of knowledge. In any given domain, there are some people who know a lot more than other people. It is simply false to imagine that in a democracy everybody already knows everything they need to deliberate at length and reach a statisfactory compromise. For democracy to work, for us to “rule ourselves without a boss, we must learn to attend to experts in various domains. And that means we have to respect and trust experts in various domains. We have to trust that they are not going to seize power and rule over us by use of their domain knowledge.

So, I think once again that’s a matter of learning. We have to learn how to be the kind of citizen who is able to defer to certain kinds of expertise. And yet, at the same time, we must maintain democratic control of the experts and keep the experts always accountable to ourselves. And that requires a lot of work in designing democratic mechanisms and learning how to “work the machine” we have designed. We have to understand the institutions as well as the background norms that make democracy sustainable. That doesn’t just come naturally. We have to learn how to be effective citizens and then do the work of citizenship.

AQ: Is it alright or even helpful to think of intelligent machines as slaves, and what might be the consequences of that, given that some AIs might end up presenting as human or animal like?

JO: I think that is the bullet we need to bite down on, in the development of AI. Humans are the original developers of AI, but obviously with machine learning, it gets complicated. So, the question of “presenting as human” really is the challenge. Unless we can imagine a world in which we don’t need to have anything that has intelligence of any sort engaged in menial labor, under the direct control of others and for the good of others, then we’re going to need to have machines that have this “Aristotelian” feature of not being corrupted in an ethical sense. That’s a nontrivial design problem. And it is an ethical problem. If and when machine intelligence is of a truly human sort, then it has the same susceptibility to “soul corruption” and the same moral problem about forcing it into the condition of being corrupted. That’s an issue I think philosophers and engineers need to really think about together.

AQ: What is in between those who are capable of being natural slaves and the masters? Aristotle is making distinctions and effectively putting a binary condition on what must be a distribution with a vague predicate: the natural slave.

JO: One answer is that Aristotle worries that people can self-enslave, through practice. He posits that there are natural slaves, beings whose best condition throughout their lives is being enslaved. And, on the other hand, there are complete humans. But complete humans who put themselves under the control of a master, who take commands systematically through the course of their lives, will, he says, become “slaves by practice”; they will become slave-like. They have all of the features of the natural slave, except they get there by the deformation of their souls rather than simply by being born with a deficient soul. Aristotle thinks many people have at least begun to get onto that bad pathway toward self-enslavement. Many others are fully corrupted through menial labor: they have lost the capacity to reason in their own interest along with others. And that’s what is horrible about the conditions that he thinks most people live in and that’s the world he wants us to get out of. He wants to get to a world in which self-enslavement is not a necessary part of getting the work done that we need to get done. And that’s why he needs to bring the natural slaves into the story. If we think it is horrible that people are made to self-enslave today, we need to figure out how to get the necessary work done without corrupting the souls of the entities, human or machine, that do it.