Dupoux and Jacob's moral instincts: throwing out the baby, the bathwater and the bathtub.

We agree with Dupoux and Jacob (DJ) [1] that human moral capacities are, in ways yet to be understood, grounded in our biology. However, we disagree with DJ’s outright rejection of the linguistic analogy (LA). Fully cognizant that our present understanding of the moral faculty is no better, and perhaps is worse, than the state of affairs in linguistics circa 1950, we believe that it has opened up new and exciting research questions for scientists and moral philosophers alike, with empirical work emerging apace [2–4]. DJ claim to acknowledge that LA ‘usefully’ organizes inquiry into morality around the five central questions familiar from the study of language [5] and concede that recent studies of moral judgment [2,5–7] lend prima facie support to the value of LA. Yet they insist that the plausibility of LA depends on whether ‘the mechanisms underlying moral judgments make use of moral information encoded in a dedicated moral grammar (374)’. DJ’s attempted immolation of this straw person reveals not only their misunderstanding of the hypothesized moral faculty (MF), but also of the language faculty (LF). Here, we summarize some misconceptions: LF is not agreed to be, and perhaps is not, a module in the Fodorian sense [8]. It is plainly question begging to assert, ‘metacognitive processes of justification operating on explicit moral beliefs are fully part of the moral faculty (377)’. This claim is tantamount to saying that rhetoric is a proper object of scientific research. Hence, LA is not undermined, even if MF is not a module in the Fodorian sense. There is evidence that humans parse the world in ways strongly suggestive of a grammar of action [3,5,6]. In addition, just as the meanings of complex expressions can be fully determined by the meanings of their parts and the way in which they get put together without all those parts being fully recoverable through simple reversibility, failure of full reversibility from moral valences to structural descriptions of actions is not evidence for the absence of compositionality. The fact that part of our moral psychology depends on explicit beliefs is not counterevidence against an intuitive, unconscious component, viz. a moral grammar. Part of our language psychology also depends on