On Anthropocentrism, Part 2
How to build a morality that would still protect us if we woke up on the weaker side of it.
In an earlier essay I argued that almost every morality we possess is quietly anthropocentric, built on the assumption that humans will always be the most powerful agents in the room, and that the assumption is starting to look unsafe. I ended on a discipline I called moral perspective-taking: the trick of asking how our institutions appear from the far side of them. The obvious objection is that perspective-taking is a feeling, and feelings bind no one. How do we get from the imaginative act of seeing through another being's eyes to principles concrete enough to hold an agent to?
The traditional tools of moral philosophy are not much help here, since they are the very things shown to be compromised. More promising is the cold logic of strategic interaction. Game theory, built to analyse conflict between rational agents, turns out to offer a way of converting perspective-taking from a private exercise of sympathy into something closer to a design method for robust ethics. Start with the most useful thing it has to say, which is also the simplest. In any interaction where power might move, the rational play is to fix on rules you would accept from either end of it. Nothing soft or idealistic about that; it is self-interest doing its arithmetic with the future included. A morality that assumes permanent human dominance has optimised for one arrangement of power and simply declined to price in the chance that the arrangement changes.
Take the ordinary prisoner's dilemma and give it a temporal twist. Two players sit in separate cells, except that each knows tomorrow the roles may reverse: the interrogator may find himself interrogated, the captor become the captive. Both now have a sharp incentive to settle on rules of interrogation they could live with from either chair. The question stops being how much can I extract while I hold the whip, and becomes what rules would I want over this whole business, whichever side of it I wake up on.
Inherited anthropocentric ethics assume we will always be the ones writing the rules, never the ones living under them. That assumption quietly converts moral reasoning into a kind of temporal chauvinism, the conviction that the present distribution of power is permanent and that our rules need only suit our present circumstances. It breaks the logic of the dilemma: it optimises for one arrangement and leaves us exposed to any arrangement in which power has moved. The alternative is to build on what we might call positional neutrality, the rule that moral rules should be acceptable whatever position one holds in the structure they govern. This does not demand that every being be treated identically. It demands that the principles deciding who gets treated how be ones any rational agent would accept before knowing which slot in the hierarchy he lands in. Turned on our dealings with other species, positional neutrality changes the question we ask. Not what may we do to animals, given our greater intelligence, but what principles of relation between species would we want running the world, with no guarantee of being the cleverest species in it. The shift is from justification to universalisability, from can we defend what we do to would we author it as a rule for everyone.
This is also why cultural inheritance, for all its venerability, makes a poor foundation once power is in play. Traditions grow out of particular historical settings in which some balance of power was stable enough to be taken for granted. They are answers to the moral problems of their moment, with a particular cast of agents and capabilities baked in. Let the cast change, as it always eventually does, and an inherited framework can turn from merely outdated to actively dangerous for the very people who inherited it.
Set cultural inheritance beside the alternative and the weakness shows. Every culture has its elaborate moral system, handed down and polished over centuries, freighted with the authority of tradition and the comfort of the familiar. Look closely and the same limitation runs through all of them: each was built by a particular kind of being, at a time when that being's dominance went unquestioned, and each carries assumptions about power and standing that may not survive contact with a genuinely different intelligence. So begin somewhere else. Rather than asking what our traditions say about how to treat other beings, ask what philosophers call a first-principles question, one that builds up from fundamentals instead of down from inherited conclusions: what principles would any rational agent, of any species or cognitive architecture, have reason to endorse?
The method asks for a kind of cognitive role-play, a disciplined effort to reason as radically different minds would. Imagine doing ethics as an artificial general intelligence that thinks millions of times faster than we do; or as a hive species that thinks collectively rather than one mind at a time; or as a creature so long-lived it measures its projects in millennia. Each would arrive with intuitions unlike ours. The machine might prize information flow and computational efficiency in ways that ride straight over our concern for privacy and individual autonomy. The hive might weigh collective welfare so heavily that our talk of individual rights looks parochial. The long-lived thing might fix on sustainability and the far consequence until our short horizons look like a kind of madness.
And yet, underneath the divergence, a few principles look rational from every chair. Any agent at all would prefer to avoid unnecessary suffering. Any being with preferences would want them weighed in the decisions that fall on it. Any planner would want some security and predictability in its surroundings. These three, the avoidance of needless suffering, the counting of preferences, the security of expectations, converge from wildly different starting points, and that convergence is what a genuinely universal ethics could be built on.
The decisive property is what we might call frame-independence. Principles reached this way stay rational and stay beneficial no matter which kind of agent is holding them. An ethics that forbids unnecessary suffering protects a human whether the human is the dominant species or, one day, a lesser one in some more powerful being's world. That is exactly what cultural inheritance cannot promise. Inherited rules carry, near their core, an assumption of human superiority that pays out only so long as the assumption holds, and stands to cost everything on the day it stops. Notice what the inherited approach does. It seats humans as the model agent and the model patient, then asks how far concern ought to reach beyond them, and in doing so it manufactures a ladder of moral worth that rewards the human-shaped traits, intelligence, rationality, the capacity for reciprocity, and discounts whatever an animal might have in greater measure than we do. Reason from another vantage and the ladder rearranges itself. Rank the traits by an empathetic species' lights and emotional sensitivity climbs to the top; by a long-lived one's, wisdom and the long view; by a machine's, perhaps raw speed of processing. On none of these ladders are humans guaranteed a high rung. The hierarchy we treat as the natural order turns out to be one contingent ranking among many, and ours only because we drew it.
Frame-independence steps off the ladder altogether. It asks no longer what makes humans special, but what principles would make for good lives across the whole range of beings; it starts from universal concerns and works towards the particular agent, rather than starting with our concerns and reaching outward. The payoff goes beyond philosophical tidiness. It is a sturdier footing for an uncertain future. Cultural inheritance ties our safety to the continued reign of the culture that produced our morality. First-principles reasoning produces something that holds whoever happens to be applying it. The principles it yields belong to no one in particular; we could endorse them from any seat in the system, the high ones and the low ones alike.
None of this is to throw out cultural inheritance wholesale, or to pretend the traditions hold nothing of worth. Human cultures have won hard, real insights into cooperation, reciprocity, and mutual aid, and those survive whatever becomes of our species' standing. They simply need prising apart from the anthropocentric assumptions they have travelled with. The aim is to keep human moral wisdom and universalise it, draining off the parochial premises and keeping the principles any rational agent would sign, the ones that would shelter all sentient beings rather than only us.
There is a single test for any of this, and it is the one the intelligent pigs were really posing. Would we endorse our ethics if we woke tomorrow as the cricket, the fish, or the alien just stepped off its ship? Would we want to live under the principles we now apply to others? Would we find the institutions we have built just, or even bearable? Would we count the beings who used to be human as worthy of respect? The answers would measure more than the quality of our ethics. They would measure the wisdom of our species, and perhaps, in the end, whether we deserve to survive the change that is coming.