Moral Minds: How Nature Designed Our Universal Sense of Right and Wrong

Marc D. Hauser Ecco: 2006. 512 pp. $27.95 0060780703

In Moral Minds, Marc Hauser makes an audacious claim about moral thought. He argues that morality is best understood in much the same way as Noam Chomsky described language: as the product of an innate and universal mental faculty. For Hauser, moral intuition is not the product of culture and education, nor is it the result of rational and deliberative thought, nor does it reduce to the workings of the emotions. Instead, it is human nature to unconsciously and automatically evaluate the moral status of human actions: to judge them as right or wrong, allowed or forbidden, optional or obligatory.

As Hauser is careful to point out, he is not the first to make the leap from a chomskyan theory of language to a chomskyan theory of morality: this analogy was proposed by the political philosopher John Rawls, the legal scholar John Mikhail of Georgetown University in Washington DC, and by Chomsky himself. But Moral Minds is the first detailed exploration of this idea. It is a trade book, highly accessible to a general audience and drawing on diverse examples from literature, popular culture and history. But it is also a deeply significant intellectual contribution: everything that's done in the new science of moral psychology in the coming years is going to be a response to this important and enjoyable work.

Certain deep parallels between language and morality make Hauser's proposal worth taking seriously. Chomsky has long observed that language is a system of knowledge, but what we know (competence) is different from how we use this knowledge in everyday life (performance). Linguistic competence is also unconscious: every English speaker knows that something is wrong with the sentence “John seems sleeping,” but only experts understand why. Similarly, moral intuitions are imperfectly linked to action — you can know the right thing to do but choose not to do it — and only experts can articulate adequate reasons for common-sense moral judgements. Finally, just as there are, arguably, innate principles of language, Hauser reviews an extensive array of cross-cultural, developmental, animal and neuroscientific studies that support the existence of innate principles of moral thought.


In other regards, however, language seems very different from morality. For one thing, linguistic knowledge is distinct from emotion. You might be disgusted or outraged by what somebody says, but the principles that make sense of sentences are themselves entirely cold-blooded. Your eyes do not well with tears as you unconsciously determine the structural geometry of a verb phrase. By contrast — and Hauser wrestles with this throughout Moral Minds — even those who accept that some moral capacity is innate often see it as inextricably linked to emotion. Perhaps the universal core of morality is a set of emotional responses — disgust, shame, sympathy, guilt and so on — that are triggered by certain situations. This hypothesis is supported by clear demonstrations that, at least in some circumstances, emotion precedes intuition. The psychologist Jonathan Haidt of the University of Virginia in Charlottesville has found, for instance, that we are sometimes first viscerally shocked by a scene (“Omigod, he's having sex with a chicken!”) and then converge upon a moral judgement (“There should be a law against that.”).

A different concern is that languages are combinatorial symbolic systems. An English speaker, for example, knows perhaps hundreds of thousands of words, and also knows principles of syntax that dictate how these words combine with one another to form sentences. There are other combinatorial systems in human cognition, such as number and music, but it's not clear that morality is one of them. Even if it is distinct from emotion, moral knowledge might be better characterized as a small list of evolved rules, perhaps simple (such as a default prohibition against intentional harm), perhaps complex (such as some version of the doctrine of double effect), but still very different in character from linguistic knowledge.

This combinatorial issue becomes relevant when it comes to differences in morality between people. The approach developed by Chomsky explains differences in human languages in terms of the parametric variation of universal principles. All languages have verb phrases, for instance, but in some of them, such as English, the object follows the verb, whereas in other languages the object goes first. Hauser makes an excellent case that the variation in moral systems is constrained in interesting ways, but he provides no evidence for parametric variation of the linguistic sort. Instead, as the cultural anthropologist Richard Shweder of the University of Chicago, Illinois, and others have argued, you get a difference in emphasis: all cultures value both purity and fairness, for instance, but some emphasize the former and others the latter. In addition, there can be dramatic variation in morality within an individual culture. As Hauser notes, for example, the culture of honour in countries such as Pakistan leads to a shockingly high number of 'honour killings' every year. However, that does not imply that all native Pakistanis agree on the moral permissibility of honour killings as they do on the linguistic grammaticality of well-formed Urdu sentences.

These differences call into question how much insight the research programme developed by linguists can provide into morality. For instance, some arguments for linguistic innateness are based on assumptions about the generative nature of language; these might not export well to the moral domain. In fairness, though, it might be that nobody has found these sorts of deep parallels because, before Hauser, nobody had really looked. Moreover, even if morality lacks certain interesting features of language, the very idea of an innate moral faculty is well worth investigating. Linguists have been exploring this idea for more than 50 years; in the study of morality, we are just getting started.