Recent attempts to revive the Platonic thesis that moral knowledge is innate have attempted to piggy-back on the perceived success of the Chomskian arguments for the thesis that linguistic knowledge is innate. The most important of these arguments has been the Poverty of the Stimulus argument. Chomskian poverty-of-the-stimulus arguments primarily deny that empiricist learning would allow children to be as competent with the language as they are. Childhood first language acquisition is for the most part invariant with respect to environmental input. If that’s right, then empiricist accounts of language acquisition are inadequate. This is thought to lead to the positive result that the only way to explain fully children’s linguistic competence is by reference to innate knowledge. According to Fodor, “The bottom line of Poverty of Stimulus Arguments, as Chomsky uses them, is that innate domain specific information is normally recruited in first language acquisition” (Fodor 2001). The ideas or propositional knowledge we acquire about some domain, the domain of language in this case, is explained by an innate mechanism specific to that domain.

How natural it seems to run a parallel argument in favor of a kind of moral nativism.

Just as empiricist accounts of learning are inadequate to explain children’s linguistic competence, so also are they inadequate to explain children’s moral competence (Dwyer 1999; Harman 2000). For example, children exhibit the capacity to make genuinely moral judgments since they are sensitive to the moral/conventional distinction in their use of terms like “wrong”. It has been shown that children as young as 3 years old will regard moral transgressions as more serious than conventional transgressions and regard the former as more universal, independent of any authority figure, and closely correlated with considerations of harm (Smetana and Braeges 1990; Smetana 1993; Nichols 2002a). So if children are told that some households have no rule against, say, running in the house or eating dessert before dinner, then even if their parents do have rules against these practices, most children will judge that it is not wrong for children who live in the households where the rules do not exist to do these things. However, children will judge that hurting another is wrong, even if they are told that some parents have no rule against hurting others. They make similar claims across different cultures. And this recognition emerges early in the developmental psychology of children – earlier than we should expect if children learn it only from explicit moral instruction from caregivers.

Susan Dwyer has presented a poverty-of-the-stimulus argument for moral nativism based on these data. According to Dwyer:

Absent a detailed account of how children extrapolate distinctly moral rules from the barrage of parental imperatives and evaluations, the appeal to explicit moral instruction will not provide anything like a satisfactory explanation of the emergence of mature moral competence. What we have here is a set of complex, articulated abilities that (i) emerge over time in an environment that is impoverished with respect to the content and scope of their mature manifestations, and (ii) appear to develop across the species (Dwyer 1999, 173).

This is the same anti-empiricist conclusion as reached in the case of language acquisition and Dwyer takes this to suggest the same positive result: there is in human beings an innate mechanism specific to the moral domain. Dwyer writes, “We all come into the world equipped with a store of innate moral knowledge which, together with our experience, determines our mature moral competence” (176-177).

The thesis that there is innate moral knowledge would be undermined if there were an alternative hypothesis capable of explaining the relevant empirical data. An alternative proposal for explaining our capacity for moral judgment implicates human affective response mechanisms. Some studies suggest that the capacity for drawing the moral/conventional distinction is linked to emotional development (Blair 1995; Nichols 2002a). For example, individuals with normal affective responses to suffering and distress cues more reliably draw the appropriate distinction between moral transgressions and conventional transgressions than individuals whose affective response mechanisms aren’t fully developed or functioning properly. For example, psychopaths are more likely to treat moral transgressions as most others would treat conventional ones with regard to the transgression’s seriousness, authority contingence, universalizability, and justification (Blair 1997). The more typical emotional responses to such transgressions (the responses of those in the normal population) explains our tendency to attribute to harm norms their distinctive status as nonconventional.

This affect-based proposal seems to explain the capacity for moral judgment without reference to innate moral propositional knowledge. However, the poverty-of-the-moral-stimulus argument remains intact without the characteristic positive conclusion. Nothing in the environment is playing the critical role in the development of moral judgment; rather, it’s native emotion contributing to moral competence. Writes Shaun Nichols, “of course, there is still a crucial innate contribution to distinctively moral judgment, but the contribution is affective, not propositional…” (Nichols forthcoming).

This proposal apparently fits nicely with an expressivist semantics for moral terms since it explains how individuals will come to call something “morally wrong” typically only if they disapprove of it. According to moral expressivism, moral judgments are expressions of pro- or con-attitudes. Affect is necessarily implicated in moral judgment. According to the proposed view of moral psychology, noncognitive mechanisms of the mind provide innate biases that shape moral judgment. Arguably, the best account of how our moral psychology works provides empirical evidence for moral expressivism.

But maybe if Dan is right, and moral attitudes are logically complex, the dichotomy between the innate contribution being either affective or propositional is false. If the characteristic moral emotions, and the more general moral attitudes, are as fully propositional as they are affective, then moral nativism has not been undermined. Moral propositional knowledge has its basis in innate affective mechanisms.

One possible way to cash out this idea is to say that pro- and con-attitudes project value and disvalue, but the attitudes can still be true perceptions of value and disvalue insofar as they are outputs of a properly functioning moral sense or something like that. If human affective mechanisms aren’t functioning properly, they won’t reliably lead to the appropriate attitude and the correct moral judgment.

Whether this model of moral judgment is plausible seems to depend on there being the right sort of objective appropriateness conditions for the affective responses in question. For example, certain situations call for fear, certain situations call for contempt, and certain situations call for the relevant pro- or con-attitudes. The underdeveloped suggestion here has been that the appropriateness conditions are determined by the innate emotional response mechanism in the properly functioning person, where these emotional responses are now considered propositional and affective. The data about psychopathy at least seems to fit this suggestion. So it’s possible that the best account of how our moral psychology works provides empirical evidence for a kind of cognitivist expressivism.


Blair, RJR. 1995. A Cognitive Developmental Approach to Morality: Investigating the Psychopath. Cognition, 57, 1-29.

Blair, RJR. 1997. Moral Reasoning and the Child with Psychopathic Tendencies. Personality and Individual Differences, 26, 731-739.

Dwyer, Susan. 1999. Moral Competence. In Murasugi and Stainton (eds.), Philosophy and Linguistics. Boulder, CO: Westview Press, 169-190.

Fodor, Jerry. 2001. Doing without What’s Within. Mind, 110.

Harman, Gilbert. 2000. Explaining Value. New York: Oxford University Press.

Nichols, Shaun. 2002. Norms with Feeling: Towards a Psychological Account of Moral Judgment. Cognition, 84, 221-236.

Nichols, Shaun. forthcoming. Innateness and Moral Psychology. In Carruthers, Laurence, and Stich (eds.), The Innate Mind: Structure and Content. New York: Oxford University Press.

Smetana, J. 1993. Understanding of Social Rules, in M. Bennett (ed.), The Development of Social Cognition: The Child as Psychologist. Guilford Press.

Smetana, J. and Braeges, J. 1990. The Development of Toddlers’ Moral and Conventional Judgments. Merrill-Palmer Quarterly 36, 329-346.

6 Replies to “Poverty of the Moral Stimulus

  1. Interesting. I think this article is quite persuasive. It makes sense that there would be a “moral ability” innate in humans. If this is indeed the case, though, wouldn’t it also make sense that such an ability would be subject to evolutionary forces? The reason I bring this up is because I could easily see people extrapolating normative claims from the fact (if it is a fact) that humans have an innate moral capacity. I think this would be a mistake, though. Say the propositional content of our innate moral ability is discovered – does that mean that this content is normative? This is the classical is/ought distinction. Now, if we were to discover that it was not possible for humans to deviate from the content of their innate moral ability, then it wouldn’t be possible to distinguish is from ought.
    However, Chomsky strongly believes that humans have a creative capactiy when it comes to their innate linguistic abilities, which he sees as leaving room for freedom of the will. Logically, then, one should posit the same freedom morally. In this case, is can be distinguished from ought.
    So what are your thoughts about how an innate moral ability interfaces with normative ethics?

  2. Kyle-
    What a fantastically interesting post. Thanks! I’ve got about 36 things I’d like to say, but I’ll restrain myself to just a few.
    As someone who is deeply skeptical that there is anything more than a conceptual gap between the propositional and the affective, I very much hope that your Aristotelian proposal that moral judgments rest on true perceptions of value is correct. My main observations is that it doesn’t seem to me obvious that
    (a) because the capacity for issuing moral judgments (which we can presume is tracked by the capacity to draw the moral/conventional distinction) is linked to emotional development that
    (b)moral judgment is a kind of native emotional competence rather than being (or involving) a form of propositional knowledge.
    Might it not be true that the appropriate moral development is in evidence because of some form of propositional knowledge? I gather that the poverty-of-stimulus argument is meant to deny this, since this alleged propositional knowledge is underdetermined by children’s exposure to moral claims and propositions. So I’d like to know more about what the stimulus-knowledge gap consists in: What is that children ‘know’ about morality that we would not expect them to know given the amount of ‘moral stimulus’ they are subject to? After al I think we could identify a good many things about morality that children do not know. To pursue the Chomskyian analogy again, Chomsky posits his deep grammar as a way to explain the gap between children’s exposure to the rules of their language and their apparent facility with it. What’s the moral knowledge gap exactly? And is there something in the moral domain that would play the role here of deep grammar?

  3. Among psychologists, there is a broad consensus that there is something built in that makes people have the sorts of moral views that we have. That still leaves open questions like what exactly that “something” is, whether it’s one thing or many things, and what sorts of learning people undergo. All of those questions have been subjects of research, though (see, for instance, Piaget & Kohlberg on moral development). Shweder has argued that there are 3 kinds of morality: ethics of autonomy (based on rights, duties, liberties, etc.), ethics of community (based on cooperation, social goods, punishing slackers, etc.), and ethics of divinity (based on purity, sin, etc.).
    There is strong evidence for the claim that our moral intuitions are based on emotions in addition to cognitions. See, for instance, Jon Haidt on Moral Emotions.
    For a not-quite-up-to-date look at morality from the point of view of evolutionary psychology, I recommend Robert Wright’s The Moral Animal.

  4. Kyle:
    I like this proposal. It makes a lot of sense to think that there is a or several deep moral grammar(s)analogous to the Chomskian linguistic deep grammar. There may be a parallel between the norms governing linguistic behavior (entering and exiting conversation) and the norms governing moral responses that can be exploited, in the way J. Heath does in his “Transcendental Argument for Morality” in PPR (2003).
    A question I have about this thesis. Does moral innateness make the amoralist empirically or conceptually impossible? If moral nativism implies that moral responsiveness is not learned then there cannot be an amoralist, in the sense of a character who has learned the moral language but does not have the parallel moral responses. A character who lacked the approporiate moral responses could be a psychopath, lacking the appropriate affective equipment, but it appears could not be an amoralist.
    Great post

  5. Thanks for these comments. Mathew, the data I’ve cited suggests the pervasiveness in normally functioning humans of normative claims invoking harm norms. I’m trying to work this into my preferred metaethical view. What this means for normative ethical theory I’m not in a position to say right now, but I know that Joshua Greene looks at this in some papers, including an especially neat one on Trolley cases (
    Michael, those are good questions. One proposal offered to undermine the negative conclusion of the POMS argument (the anti-empiricist conclusion) is that caregivers have more severe reactions to their children’s violations of moral rules than they do to violations of conventional rules, or that they take greater pains to reinforce moral rules. Children would only have to distinguish between these different cues provided and respond accordingly. But it seems unlikely that this could do the trick. There just aren’t enough of such cues. The speech patterns that call for children to obey moral rules are certainly no different than the ones that call for them to obey conventional rules. “Stop hitting your sister!” isn’t distinguishable in its form or, usually, any other way from “Stop playing with your food!” A child will hear “Stop doing that!” after having done either (at least, mine will), and likely without any discernable difference in tone or severity or whatever. There doesn’t seem to be enough available in typical behavioral correction or instruction that would cue a child to whether or not there’s been a moral or a conventional rule violated. But children, from a young age, have the ability to make this distinction.
    About the amoralist: there’s a difference between a moral standard (“Thou shalt not hit thy sister”) and a moral judgment or belief (“It is wrong to hit one’s sister”). First, the judgment does, while the standard doesn’t, have truth conditions. Second, endorsement of the standard does, while having the belief doesn’t, entail the motivation to act in accordance with it. This is, I think, David Copp’s position, which I like. It’s compatible with my thesis and it permits the possibility of the amoralist.

Comments are closed.