This post describes the status of science of morality investigations, plus my views on how those studies relate to each other, can be expanded on, and can and cannot tell us about right and wrong.

The science of morality approaches morality both from the “bottom-up” by looking for cross-cultural universals within descriptively moral norms and judgments, and from the “top-down” by applying game theory to understand the cooperation strategies that have enabled us to become such an incredibly successful social species. I’ll argue that both perspectives merge to a unified, robust understanding of cultural moral norms and our moral sense’s judgments as parts of cooperation strategies.

The “bottom-up” approach begins with gathering data about what people in different cultures say is moral and immoral. These surveys produce a superficially chaotic data set with diverse, contradictory, and strange elements. In philosophical terms, these surveys gather information about descriptively moral norms(ref6) and judgments (norms and judgments considered moral in at least one culture).

Oliver Curry and Jonathan Haidt are the two best-known bottom-up science of morality investigators. They have proposed separate theories about moral cross-cultural universals – the subset of moral norms and judgments that are empirically universal to all studied societies.

Oliver Curry’s “Morality as Cooperation” theory3 describes moral universals as solutions to seven cooperation problems regarding family values, group loyalty, reciprocity, heroism, deference, fairness, and property rights. Curry proposes that behaviors that are part of solving these problems are normative – what people ought to do.

Jonathan Haidt’s “Moral Foundation Theory”4 summarizes moral universals as five circumstances (others have been proposed since) that universally trigger moral judgments: (1) harm (reducing)/care, (2) fairness/reciprocity, (3) ingroup/loyalty, (4) authority/respect, and (5) purity/sanctity. Haidt proposes that behaviors that maintain these circumstances are normative5.

There is some overlap. Both agree with the consensus among science of morality investigators, that morality is about maintaining and increasing cooperation2,3,8,11,12.

Are Curry and Haidt proposing competing theories or just different perspectives on the same phenomena? And what is the origin and function of norms and judgments that are not cross-culturally moral? These questions will be easier to answer viewed from the “top-down” perspective.

Martin Nowak’s top-down approach uses game theory to identify five mechanisms7,8 for the evolution of cooperation: kin selection, direct reciprocity, indirect reciprocity, network reciprocity, and group selection. These mechanisms are as innate to our universe as the mathematics of game theory they are based on. They were available to our ancestors and were selected for and encoded in our biology and cultures by the benefits of cooperation they produced. Species that have not encoded these mechanisms into their biology and cultural norms would find it difficult or impossible to form highly cooperative societies.

The top-down approach predicts that some of Nowak’s mechanisms will necessarily be part of ‘morality’ for all highly cooperative species. The top-down approach thus enables us to make claims about what is universally moral for all intelligent species, not just what is cross-culturally moral for people (the limit of universality the bottom-up approach can claim).

What about descriptively moral norms that are not cross-culturally moral? Why do they exist? Are “universal” cultural moral norms and judgments best understood as moral absolutes or what? And why do people experience the strong intuition that what they believe is moral is what everyone ought to do regardless of their needs and preferences and violators deserve punishment?

Again, the top-down, game theory perspective on morality is informative and can explain that:

1) Exploitation of outgroups by ingroups (slaves must obey their masters, women must be submissive to men, and homosexuality is evil) have been considered descriptively moral10,11 because these norms can increase the benefits for an ingroup cooperatively exploiting outgroups. This ingroup cooperation (even if it is to exploit an outgroup) matches one of the templets encoded in our moral sense for recognizing descriptively moral behavior.

2) Marker norms (including sex and food taboos plus dress and grooming ‘moral’ norms) exist because they can improve the success of reciprocity strategies by enabling people to identify more reliably cooperative people who are committed members of their ingroups.

3) Versions of the Golden Rule are virtually universal in cultural moralities and are sometimes thought of as moral absolutes. Science can explain that versions of the Golden Rule are part of essentially all cultural moralities not because they are moral absolutes but because they are heuristics (usually reliable but fallible rules of thumb like virtually all moral norms) for initiating indirect reciprocity, perhaps the most powerful cooperation strategy known.

4) Our illusion of morality’s strange bindingness regardless of needs and preferences exists1,2,9 because our illusion of imperative bindingness is part of our motivation for punishing moral norm violators. Punishment is a necessary component of the cooperation strategies that make up human morality2. This illusion was encoded into the biology underlying our moral sense by the reproductive fitness benefits it produced.

As described above, the science of morality has made wonderful progress in the last several decades in understanding cultural moral norms and our moral sense.

However, I argue that both the existing top-down and the bottom-up approaches can be further expanded to reveal even more useful information.

The following claims may be implied in the literature, but I have not seen them explicitly stated.

What problem do Nowak’s five mechanisms solve?

I argue they all solve the cooperation/exploitation dilemma, a higher-level problem that is innate to our universe. This dilemma is how to sustainably obtain the benefits of cooperation without exploitation destroying future benefits of cooperation. This is a dilemma because exploitation is almost always the winning strategy in the short term and can be in the long term.

Identifying the cooperation/exploitation dilemma as the ultimate source of human morality suggests that the absence of exploitation is what separates what is universally moral from what is merely descriptively moral. Hence:

• Descriptive moral behaviors are parts of cooperation strategies.
• Universally moral behaviors are parts of cooperation strategies that do not exploit others.

Due to their origins in the mathematics of game theory, these principles are as species, environment, and time independent as the mathematics they are based on.

As a check, are these principles consistent with what we know about human morality?

Yes. Descriptively moral norms gathered as part of the bottom-up approach appear to be parts of cooperation strategies. And none of the cross-cultural universals – the seven moral problems identified by Curry and five circumstances identified by Haidt - include exploitation as a component. That the above two simple moral principles can explain such a huge, diverse, contradictory, and strange data set supports them as the basis of a highly robust theory of morality. (The data set they explain includes past and present cultural moral norms, Curry’s and Haidt’s moral universals, and Nowak’s five mechanisms.)

Competing explanations for cultural moralities such as “reducing harm”, “conflict resolution”, “maximizing happiness”, “increasing reproductive fitness”, and “cultural relativism” are not even remotely competitive in explaining this superficially chaotic data set.

Proposed counterexamples to the two principles are always welcome.

What can the science of morality tell us?

The science of morality can tell us the origin and function of descriptively moral behaviors and their universally moral subset. The science of morality can also inform us about instrumental oughts of the form “If you desire X and wish to obtain it using a universally moral means, you ought to do Y”. This knowledge can be useful for resolving moral disputes within groups that desire a similar X and wish to use this universally moral means (cooperation that does not exploit others). Like the rest of science, the science of morality cannot tell us what our goals somehow ought to be or what we ought to do regardless of our needs and preferences.

I’ll leave further discussion of cultural and philosophical implications for another time.

1. Blackford, Russell (2016). The Mystery of Moral Authority. Palgrave Macmillan.
2. Bowles, S., Gintis, H. (2011). A Cooperative Species: Human Reciprocity and Its Evolution. Princeton University Press.
3. Curry, O. S. (2016). Morality as Cooperation: A problem-centred approach. In T. K. Shackelford & R. D. Hansen (Eds.), The Evolution of Morality. Springer.
4. Graham, J.; Haidt, J.; Koleva, S.; Motyl, M.; Iyer, R.; Wojcik, S.; Ditto, P.H. (2013). "Moral Foundations Theory: The pragmatic validity of moral pluralism". Advances in Experimental Social Psychology. 47: 55–130
5. Haidt, J. (2012). The Righteous Mind: Why Good People Are Divided by Politics and Religion. Pantheon.
6. Harms, W., Skyrms, B. (2010) Evolution of Moral Norms. In Oxford Handbook on the Philosophy of Biology ed. Michael Ruse. Oxford University Press.
7. Nowak, MA. (2006). Five rules for the evolution of cooperation. Science. 2006 Dec 8.
8. Nowak, M., Highfield, R. (2011). SuperCooperators: Altruism, Evolution, and Why We Need Each Other to Succeed. Free Press.
9. Ruse, M. and Wilson, E. O. (1991) The Evolution of Ethics, in Religion and the Natural Sciences: The Range of Engagement.
10. Singer, Peter (1981) The Expanding Circle: Ethics, Evolution, and Moral Progress. Princeton University Press.
11. Tooby, J., and Cosmides, L. (2010). Groups in Mind: The Coalitional Roots of War and Morality, from Human Morality & Sociality: Evolutionary & Comparative Perspectives, Henrik Høgh-Olesen (Ed.), Palgrave MacMillan, New York, pp. 91-234.
12. Tomasello, M., & Vaish, A. (2013). Origins of Human Cooperation and Morality. Annual Review of Psychology, 64(1), 231-255.