Morality is neither objective the way physics is nor merely arbitrary preference; it is a constrained practice in which some views are genuinely better than others by standards internal to it, so "it's all subjective" and "it's written in the stars" are both wrong.
Against strong objectivity: there are no moral facts detectable in the world the way mass or charge are, no instrument that reads off wrongness, and no valid derivation of an "ought" from purely descriptive "is" premises—the is–ought gap is real.
Pervasive, seemingly intractable moral disagreement across cultures and eras is further evidence that ethics is not simply read off a shared external reality.
So moral claims do not look like mind-independent facts of the physical kind.
But the leap from there to "morality is just preference, like a taste in ice cream" does not follow and fails on inspection.
Moral judgments answer to reasons: they must be consistent (you can't hold that a harm is fine for you but wrong for an identical other without a relevant difference), they are constrained by shared facts—about who is harmed, who suffers, what actually promotes flourishing—and they are revisable under argument in a way pure tastes are not.
This is why moral progress is recognizable as progress (the abolition of slavery was not a mere change of fashion) and why some positions collapse under scrutiny while others survive.
The defensible synthesis is constructivist or quasi-objective: morality is constructed by reasoning agents but constrained enough that it has better and worse answers—objective relative to the standards any coherent moral practice must honor (consistency, non-arbitrariness, attention to suffering and wellbeing).
Practically: you cannot prove ethics like mathematics, but you can reason about it rigorously, and treating "it's all subjective" as a conversation-ender ignores how much moral disagreement is actually resolvable by facts and consistency. This would need revision if either mind-independent moral facts were found or all moral reasoning reduced to raw preference—neither holds up.
Morality isn't objective the way physics is, but it isn't arbitrary the way ice-cream preference is either. It's in between: no moral facts floating in the world to detect, but not "just opinion" either—because some moral views are more consistent, better informed about who actually suffers, and harder to defend once you reason them through.
Holding both halves:
The tell that it's more than opinion: moral progress is recognizable as progress. Abolishing slavery wasn't a change of fashion like hemlines; it was getting something more right, by standards (consistency, taking others' suffering seriously) we can actually articulate.
So the practical stance avoids two lazy errors. Don't claim ethics is written in the stars and provable like math—it isn't, and pretending otherwise usually just launders one culture's rules as cosmic law. But don't retreat to "it's all subjective, who's to say"—that's a conversation-ender that ignores how much moral disagreement is genuinely resolvable by getting the facts straight and applying your own principles consistently.
Reason about ethics like the constrained, arguable, improvable thing it is. You won't get proofs. You'll get better and worse answers—which is enough to live by, and enough to rule a lot of things out.
The Particle That Isn't There.
Physics can point to the thing its truths are about—the mass, the charge, the field that pushes back. But hunt through the world for wrongness itself and you find no particle of it, no moral substance woven into the atoms of a cruelty. The murder and the mercy are made of the same physics; the "ought" is nowhere in the "is." This absence tempts us toward despair—if wrongness isn't out there, isn't it all just made up? But notice: numbers aren't "out there" either, and no one thinks arithmetic is a matter of taste. Perhaps morality, like mathematics, is real without being physical—a structure of constraint that binds us though no instrument can find it.
The Convergence in the Dark.
If morality were pure invention, we'd expect the world's ethical codes to be as randomly various as its cuisines. Instead, beneath the real and vicious differences, something keeps recurring: nearly every culture forbids gratuitous cruelty, honors some form of fairness, protects the vulnerable in some way, condemns betrayal. It is as if many peoples, reasoning in the dark from the same raw facts—that others suffer as I do, that we must live together—kept stumbling onto the same few load-bearing walls. This convergence is not proof of cosmic moral law. But it is hard to explain if ethics is nothing but preference, and easy to explain if it is a constrained search that different searchers keep partly solving.
The Chasm and the Bridge.
Hume found the chasm: no pile of facts, however tall, logically forces a single "ought"—you cannot get from is to should by deduction alone. This looks like it dooms objective morality. But watch what we actually do: we do not derive our ethics from facts by pure logic; we build it, starting from a few things almost no one can honestly deny (that suffering is bad for the sufferer, that like cases should be treated alike), and reason outward under the discipline of consistency. The bridge across the chasm is not proof but coherence—and a bridge that holds your weight is not "merely" a construction. It is the realest thing you can walk on.
The Thing You Cannot Actually Believe Is Optional.
Here is the test that cuts through every seminar: find someone who says morality is pure subjective preference, and then watch what happens when they are betrayed, robbed, or cruelly used. The relativism evaporates; the protest that rises—this is wrong, not merely disliked—comes from a place deeper than opinion, and they cannot make themselves believe otherwise no matter what their theory says. We are built to experience some things as actually wrong, not just unfashionable, and that experience is data, not error. Maybe morality's objectivity is not a fact about distant stars but a fact about us—about creatures who cannot, when it is real and close, actually treat the torture of a child as a matter of taste. The universe may be silent on the point. We are not, and our refusal to be silent may be the closest thing to a moral fact there is.
Morality is neither objective the way physics is—there are no mind-independent moral facts to detect, and no valid path from "is" to "ought"—nor merely arbitrary preference, since moral judgments answer to consistency, to shared facts about who suffers and flourishes, and to reasons that revise them under argument. So some moral views are genuinely better than others by standards internal to the practice, which is why moral progress is recognizable as progress rather than mere fashion. Both "it's written in the stars" and "it's all subjective" are mistakes: ethics is a constrained construction with better and worse answers, arguable though not provable.