In Normativity and Water: The Analogy and its Limits, Sharon Street claims that if synthetic ethical naturalists want to secure moral objectivity it will be by becoming Kantian constructivists or non-naturalists, thus giving up the naturalist view.
Synthetic ethical naturalists (SEN) want to synthetically identify moral properties with natural properties, in the analogous way ‘water’ has been identified with H2O. In the case of ‘water’, we collect samples of it (e.g. the stuff in lakes, rain etc.), we study what is relevantly similar in the samples we collect, use reflective equilibrium, and then we determine that water is synthetically identical to H2O. In the case of ‘good’, we collect samples of it (e.g. donating to charity, helping an injured person etc.), we study what is relevantly similar in the samples we collect, use reflective equilibrium, and then we determine that good is synthetically identical to some natural property N (e.g. maximizing happiness.)
Street offers her own Moral Twin Earth scenario where ideally coherent Twin Earthlings discover a synthetic identity with ‘good’ and some natural property O, which entails torturing a cat. Depending on how the SEN responds to this unintuitive result, they will be forced onto Kantian constructivism or non-naturalism. Kantian constructivists think there are objective moral truths, and these truths are grounded in what it means to be a rational agent. (This is arguably not a robust moral realist view, because morality is based on facts about the agent’s own evaluative standpoint.) On a non-naturalist view there are objective moral truths, and these truths hold independently of our evaluative attitudes.
Street anticipates various SEN responses and replies accordingly.
SEN Response 1: Earthings can say that Twin Earthlings are wrong.
Reply: On this response, Earthlings and Twin Earthlings are talking past each other—what Earthlings call wrong is different from what Twin Earthlings call wrong—each have a different referent. If objectivity is to be secured, Earthlings and Twin Earthlings are supposed to be in genuine disagreement about how to live.
SEN Reponse 2: Twin Earthlings have made a rational mistake in their synthetic identity with ‘good’ and some natural property O (which entails torturing a cat.)
Reply: Your position is really a Kantian constructivist, where you make a link between rationality and morality.
SEN Response 3: Twin Earthlings have made a mistake—based on some fact independent of the empirical results and reflective equilibrium—in their synthetic identity with ‘good’ and some natural property O.
Reply: Your position is really a non-naturalist position.
My concluding thoughts:
I see other options the SEN can take to block the move to the O theory:
- If they want to draw some identities between the meaning of ‘good’ and the meaning of their natural property N, then it will push them to analytic naturalism.
- If they want to say it is not a meaning identity but a substantive conceptual truth that ‘good’ is identical to natural property N, then it will push them to a type of non-naturalism defended by Cuneo and Shafer-Landau. (In this case, I have the intuition that there wouldn’t be genuine disagreement between Earthings and Twin Earthlings.)
Recall, there is a necessary empirical aspect to SEN, but I question if it really is empirical rather than an a priori conceptual house-cleaning. SENs want to say their synthetic identity is analogous to the paradigmatically empirical water=H2O example, but I think it is more analogous to milk = ‘the liquid mothers feed their young’. The water case is an identity with a structural property. The milk case is an identity with a functional property (though it’s arguably a priori rather than empirical.) Since moral terms are taken to be identical to a functional property (some normative theory) rather than a structural property, the analogy is closer to milk. The challenge to a SEN is to find a partner-in-crime, where there is a synthetic identity between some X and a functional property that is empirical. Perhaps the best option for SENs is a type of functionalism in philosophy of mind. So the next questions is: Are moral terms more like milk (a priori?) or more like functionalism?
On the face of it, I think synthetic ethical naturalism should be implausible for those with metaethical realist intuitions. On the SEN view, it is an epistemic possibility that raping babies for fun could be good; we have to wait for the empirical data to come in. (Side note: a similar problem holds for Sam Harris’s analytic naturalism, where well-being could consist of raping.) Moral truths are to be discovered by empirically polling people, finding the referents of those terms, and finding the correct underlying normative theory—using reflective equilibrium throughout. In my opinion, this is nothing more than taking a poll of people’s preferences—a majority-makes-right view—a I point I won’t argue for here. I think this quote from Gilbert Harmon is apropos:
I now think that it was a mistake for me years ago (in Harman 1975) to treat moral relativism as a claim about the meaning of moral judgments, even when made by moral relativists. But I continue to think of myself as a moral relativist. What I think I have learned from Sturgeon and Boyd is that Moral Relativism is best treated as a version of Cornell Moral Realism! And vice versa! (Harman, 2012, p. 2)