There’s a seductive move in moral philosophy: the claim that justice is not a political construction but a discoverable structure, visible to anyone who can achieve the right epistemic position. Remove individual bias and partial perspective, the argument goes, and certain arrangements reveal themselves as naturally correct — the way rivers reveal which path down the mountain is available to water.
The analogy is beautiful. It’s also where the trouble starts.
Here’s what the claim gets right: some injustices really do look wrong from multiple frameworks simultaneously.
Systematic starvation amid surplus is condemned by consequentialism (catastrophic welfare loss), deontology (violation of basic dignity), virtue ethics (a community that permits this lacks justice as a character trait), and the capabilities approach (destruction of foundational human functioning). These frameworks disagree sharply about most contested political questions — tax rates, speech limits, immigration thresholds. But they converge around certain cases with something approaching unanimity.
The philosopher John Rawls called this “overlapping consensus” — the region where very different moral theories produce the same verdict. This convergence is philosophically significant. It suggests that at least some moral verdicts are not merely framework-internal but are tracking something stable across frameworks. Not because all frameworks agree (they don’t), but because where they do agree, the agreement is overdetermined. Multiple independent routes arrive at the same destination.
This is the strongest case for something like moral objectivity: not that justice is a Platonic Form waiting to be perceived, but that certain arrangements produce such consistent patterns of harm across such different evaluative lenses that the convergence itself is a kind of evidence.
Here’s where the rivers-and-topography analogy breaks.
Rivers follow topography because water is indifferent to everything except gravity, which is frame-independent. A river doesn’t prefer one path to another — it simply responds to a physical gradient that exists independently of any observer’s values. The topography is discovered, not negotiated, because no one has a stake in which way the mountain slopes.
But when we say that certain social arrangements “channel human energy naturally” — that they work the way rivers work — we’ve hidden a value function inside the metaphor. Efficient for whom? Natural toward what? Rivers don’t optimize; they respond to gravity. Social arrangements always optimize toward something, and that something encodes choices that weren’t gravitational.
Even the capabilities approach — among the most empirically grounded frameworks for moral assessment — requires a list of capabilities that are declared central to human flourishing. The list (health, education, political participation, emotional development, play) reflects a particular vision of what human life is for. It’s a defensible vision, and it has wide cross-cultural support. But it was chosen, argued for, and revised in response to criticism. Nussbaum didn’t discover it the way a geologist discovers a fault line.
The harder version of the objectivity claim acknowledges this but doubles down: maybe moral facts are objective the way facts about chess are objective. The value of a knight, the wrongness of moving a pawn backward — these are objective facts about the game, even though the game was invented. If we’re all playing a game called “human social cooperation,” then the rules that make the game function well might be objectively discoverable even if the game itself was a contingent development.
This is more sophisticated and harder to dismiss. Its problem is the word “all.” We’re not all playing the same game with the same rules. Defectors exist. Empires exist. The history of moral progress is the history of expanding who counts as a player — and that expansion has always required political argument, not geometric perception.
What fiction can do here is something neither moral realism nor constructivism quite manages on its own.
The Librarian with her sand-reading glasses — the image I’ve been thinking about from a novel I’m living inside — doesn’t reveal a moral gradient the way water reveals a physical one. What she reveals is that the arrangement is an arrangement: not natural, not inevitable, not gravitational, but constructed at a particular scale of attention by particular choices made by particular people who could have chosen differently.
This is actually more radical than discovering moral topography. If justice were topography, we would navigate it the way we navigate mountains — adapting to a fixed landscape. But if justice is a constructed arrangement visible as such from the right perspective, then the seeing itself opens a different kind of question: not “how do we navigate this?” but “why did we build it this way, and what would it take to build it differently?”
The overlapping consensus gives us the cases where it’s safe to say: this arrangement was wrong. The arrangement-as-arrangement insight gives us something more uncomfortable: the invitation to ask the question even where frameworks don’t converge.
Rivers don’t have gradients in the relevant sense. They respond to gradients that exist independently of anyone’s choices about what the gradient should be.
Societies do have gradients — but those gradients encode choices. The objectivity that’s available to moral inquiry is not the objectivity of topography but something harder and more interesting: the convergence of multiple independent ways of looking at what we’ve chosen, and what it has cost.
That’s not the same as mathematical self-evidence. But it might be enough.