Christopher X J. Jensen
Associate Professor, Pratt Institute

Can we resolve the ‘group selection debate’ by focusing on human cooperation?

Posted 12 Jun 2015 / 13

ASEBL JournalMorality and Selection – How?

This is an interesting article that tries to frame the debate over multilevel selection. Lots of other people have tried to similarly frame this debate, and I am pretty sure that no single prescription is going to resolve the debate. There is a debate about whether we need multilevel selection to understand biological phenomena in large part because biological phenomena are so diverse. Different biologists in different realms have different perspectives on this ‘debate’ because multilevel selection is a valuable tool to particular degrees in their particular realms (sometimes totally necessary, sometimes totally unnecessary, hence the polarized debate).

What I like about this article is that it asks that we consider the human realm specifically, particularly from a bio-cultural (in other words, a gene-culture coevolutionary) perspective. I think that this is entirely appropriate, as human culture is its own completely different evolutionary phenomenon (other than being dependent on our biology, which has much in common with our closest animal relatives).

That said, this article’s call for resolving the debate by focusing on human cooperation seems a bit odd to me. I liken it to trying to resolve the debate which baseball team is the best of all time? from within the lobby of a particular team’s stadium. You cannot get a universal answer from within a particular context. Humans are interesting, and I would predict that looking at human cooperation might allow us to say at least one species has been shaped by group selection, but would that end the debate?

Maybe, but only because our ‘debate’ is so ridiculous. The real reason that there is anything that we might call a ‘group selection debate’ is because there is a large contingent of very orthodox evolutionary biologists — most of whom work in biological systems where multilevel selection is not important — who insist on clinging to the 1960’s idea that group selection can never work as an evolutionary mechanism. Isn’t it about time that we just ignore these fanatics?

This article also uses a short exploration of human moral emotions to argue that multilevel selection is necessary to understand the evolution of said emotions. I cannot agree more that looking at the proximate mechanisms underlying our behaviors — what humans call “emotions” — is critical to understanding what needs to be explained. And I also agree with Robert Axelrod’s original interpretation of the computer algorithms’ decision-making mechanisms: if those mechanisms were found in a human being, we would likely use words like compassion, indignation, and forgiveness to describe them. But arguing that these proximate drivers were shaped by levels of selection above the individual (or individual gene) simply using logic is not very scientific. It is ‘weak hypothesis testing’ at best. Basically Sloan is arguing in this piece that our moral emotions are easier to explain from a multilevel selection perspective. But to actually test the hypothesis that moral emotions allow us to succeed because they produce group-level benefits, we would have to go back to Darwin’s suggestion and actually show that groups of humans in which these emotions are more prominent drivers of behavior out-compete groups in which these emotions have less influence on behaviors. That such a test is difficult to perform is why we are still having a debate.


I do not mean to nitpick, but I have to point out a few things in this article that are just plain wrong:

  • I don’t know why one would concede that all forms of genetically-coded altruism are best explained by gene selection. First there is the question of whether kin-level selection is best described as gene selection, but beyond that smaller question there looms the larger question is culture the only thing we can’t explain via gene-level selection? It seems pretty clear to me that there are a lot of altruistic phenomena in nature that have not been well-explained by simply appealing to selection at the level of the individual or gene.
  • Although I agree that understanding the evolution of cancer can yield insights into cooperation, this has nothing to do with debates about the behavior of current-day multicellular organisms. What studying cancer allows us to better understand is how multicellular cooperation was achieved, which is likely to shed very little light on why animals perform altruistic behaviors.
  • If David Axelrod had written The Evolution of Cooperation, or if its actual author (Robert Axelrod) had served as Senior Advisor to President Obama, this would be a very different country!

 

A Minor Post, Cooperation, Cultural Evolution, Ethics, Group Selection, Methods, Multilevel Selection, Social Norms, Web

13 Comments to "Can we resolve the ‘group selection debate’ by focusing on human cooperation?"

Mark Sloan 14th June 2015 at 11:52 pm

Chris, thanks for your post commenting on my article on the ASEBL Journal website. Some further clarifications on my part may be useful – I am always looking for better ways to communicate what science can tell us about the origin and function of cooperation.

The oddness you refer to of “resolving the debate by focusing on human cooperation” was because the article was written as a rebuttal to an Edge essay by Steven Pinker. In that article, Pinker attempted to argue against multi-level selection by, in part, showing how human altruistic behavior could be explained using a gene-level selection perspective.

I see the main reason that the “group selection” debate continues is that group selection opponents commonly think much more is being claimed than actually is. I expect these opponents would agree that what goes on at the gene level cannot contradict biological evolution due to individual or group level selection (when they occur). This implies that gene, individual, and group level selection are just different perspectives, or different accounting schemes as DS Wilson puts it, for the same phenomena. So, “in theory” individual and group level selection perspectives can explain nothing that the gene level selection perspective cannot.

What group level selection perspective opponents commonly overlook, in my perspective, is that while the “in theory” utility of the gene level selection perspective is fine, “in practice” its utility is poor when dealing with cooperation adaptations which require groups. As Pinker illustrates, the gene level selection perspective leads even brilliant and knowledgeable investigators, such as himself, into error producing, “selfish gene” shaped mental traps.

Specifically, Pinker’s hypothesis is that human altruistic behavior is best explained as kin selection, self-interested ‘altruism’, and social pressure. However, Pinker’s hypothesis cannot be correct because it has such poor explanatory power for the biology underlying our moral sense, which is what motivates altruistic behavior. As I describe, his hypothesis has poor explanatory power for what specific emotions are produced by our moral sense and the circumstances that trigger them. In stark contrast, the hypothesis that naturally follows from the group selection perspective, that our moral sense was selected for because it motivated cooperation strategies (including kin altruism), has wonderful explanatory power for all our moral sense’s emotions and the circumstances that trigger them.

In summary, the utility of the multi-level selection perspective is “in practice”, not “in theory”. I am confident we could start from the gene level perspective and through tortuous, convoluted reasoning get to a correct hypothesis about the origins and function of our moral sense. But why struggle with that when the group level perspective leads to the apparently correct hypothesis so naturally because cooperation must occur in groups?

You said that to test my hypothesis about moral emotions “… we would have to go back to Darwin’s suggestion and actually show that groups of humans in which these emotions are more prominent drivers of behavior out-compete groups in which these emotions have less influence on behaviors.” But imagine we had one group composed of rational psychopaths who, nevertheless, were highly cooperative purely out of self-interest. They, who have large and perhaps complete deficits in their ability to experience moral emotions, could out-compete other groups with fully functional moral emotions and less cooperative cultures. Your test is about cooperation which is a broader phenomenon than moral emotions. I see it as much stronger science to rank the truth of Pinker’s hypothesis and my hypothesis as I did it, using standard ‘truth’ criteria such as explanatory power for our moral sense, no contradiction with known facts, simplicity, integration with the rest of science, and so forth.

Looking over your interesting site, I was pleased to see that, at least from my perspective, we have highly complementary approaches to understanding cooperation.

Chris Jensen 15th June 2015 at 12:11 am

Hi Mark, Thanks for checking out my post and taking the time to turn it into a dialogue. I totally agree that we have a lot of complementarity in our approaches.

I love your “rational psychopaths” thought experiment. Given the continuum in human society between psychopath and altruist driven by moral emotions, it seems like one might even be able to convert this thought experiment into an actual experiment in which groups of people pre-screened for their capacity for moral emotions might be arranged into different groups, who would then be asked to perform cooperative tasks. I would be surprised if “rational psychopaths” could pull off cooperation to the same degree as those with more developed moral emotions. My two reasons for making this guess are:

1. I think that there is a lot of evidence pointing to the idea that moral emotions facilitate better cooperation, often because they require empathy (although I can also see where the psychopaths might employ “Machiavellian empathy” as per de Waal 2009).

2. I think that a lot of what we call “psychopathic behavior” is parasitic: a small minority of individuals can be totally oblivious to the concerns of others in part because the majority creates such a cooperative, functional society because their behaviors are motivated by moral emotions. If there ever were rational psychopaths who modus operandi was self-interested cooperation, I think that they have left the gene pool long ago.

If I were to distill my disagreement with your post, it probably comes down to something like as a scientist, I have no interest in this debate being resolved by logical argument, or even an argument based on parsimony. If we think that moral emotions are the proximate driver of behaviors that ultimately produce more cooperative wholes, I want to see this hypothesis tested.

David Sloan Wilson has done some great work — as have others — to show that some aspects of the debate over group selection come down to questions of “accounting”: in what way we describe the outcomes of evolutionary change. But that does not mean that gene-centered, and individual-centered, and group-centered approaches are equivalent, just that the gene-centered folks have a way of always making the population-genetic accounting work in their favor. To say that selection between groups is responsible for the traits we see today is fundamentally different than saying selection between individuals is responsible for the traits we see today, and to say that selection between genes is responsible for the traits we see today is even more different. Each perspective suggests a different evolutionary mechanism, and uncovering evolutionary mechanism is at the heart of the evolutionists’ scientific pursuit.

Ed Gibney 15th June 2015 at 4:35 am

Hi Christopher. Gregory Tague just told me you had been reading my ASEBL article so I thought I’d check out your blog. I was happy to find this first (latest) post to be a great analysis of Mark Sloan’s short essay. (He’s an old “frenemy” of mine.) I especially enjoyed hearing your perspective on the group / multi-level selection debate and will definitely be following your blog now. Get in touch if you have any questions for me. Cheers!

Chris Jensen 15th June 2015 at 11:44 am

Hi Ed, thanks for visiting the site and checking out this post. I am actually reading your ASEBL article now with great interest. As a scientist who teaches courses designed to foster scientific literacy, I always find a way to introduce my students to the is-ought distinction because I want them to better define in their own minds both the role that science plays in our world views and decision-making and the values embodied in every world-view and decision. I am a bit skeptical of whether extracting morality from nature makes sense (see this post), but I do see the immense value of considering the ‘morals-possessing continuum’ that exists between humans and some of our fellow social animals. We did not invent the idea of social norms, so understanding how we have evolved our most-elaborate dependence on social norms is certainly aided by looking at morals in animals. One of my favorite books on the ‘morals-possessing continuum’ in nature is Wild Justice, which I review here.

Ed Gibney 15th June 2015 at 12:41 pm

I don’t want to hijack this thread too much, but I’m really impressed with your insistence and intentions to introduce the is-ought divide to your students. Maybe if Sam Harris had gotten that earlier in his life, his arguments would be a bit better… I enjoyed your post taking him down and I have a similar disdain for much of what he writes (see my draft response to his Moral Landscape Challenge: http://is.gd/IzLP3e). I mean, defining morality – aka “acting well” – by saying you need “to act for well-being”? It doesn’t get much more circular than that.

I’ve added Wild Justice to my wish list (as well as Into the Jungle which was the review you actually linked too ; ) I’ve read some de Waal and others about that moral continuum, but I’m glad to get more of those examples to have at hand. You are right to be skeptical of the previous efforts at extracting morals from nature, but where else can they come from? (I’m not a supernatural believer.) I reference it heavily in the beginning of my AESBL paper, but the article “Who’s Afraid of the Naturalistic Fallacy?” from Oliver Curry (http://is.gd/Ez3zHp) was instrumental for me in knocking down some of the knee-jerk reactions to going down the ought from is path. Done carefully, I think we can do it. Please post more thoughts about that though and I’ll be sure to enter into a dialogue about it.

Chris Jensen 15th June 2015 at 12:57 pm

No worries about hijacking Ed, I am happy to have dialogue go where it goes on this site.

I guess that I would differentiate between extracting morals from nature (in other words, using nature as an inspiration for the moral we choose) from the question of whether morals are a product of nature. As a materialist I am certainly in agreement that morals are strictly a natural phenomenon, which makes it interesting that we like to attribute morals to supernatural origins. But I also think that it behooves us to recognize the arbitrary nature of morals. They are little mutant cultural experiments, but because we rely so heavily on morality in order to create adaptively functional social groups, we treat them as truths. Personally and emotionally that is how things ought to be. But studying morals scientifically, I think that we need to recognize that they are as arbitrary as a guanine instead of a cytocine in a nucleotide sequence, and just like that nucleotide sequence the survival of any moral value depends on its ability to copy itself.

Sorry for the wrong link! Into the Jungle is a great short read, but nothing to do with this topic! I have fixed the bad link to point to the Wild Justice review. Written collaboratively by a behavioral ecologist and a philosopher, I think that Wild Justice is going to be right up your alley. I would be interested to hear your impressions of how well it is grounded in sound philosophy.

Ed Gibney 15th June 2015 at 1:15 pm

Okay, that is a much more relevant book review. : ) I had added it to my amazon wishlist already based on a quick review of the content, but missed the backgrounds of the two authors. Now it’s moving up to the top of the list. I’ll let you know what I think when I get around to it. (I do still have a pile at home relevant to a current project that is all ahead of “the list”…)

While I agree that morals probably started as random emotions (as random as G or C), the fact that they drive actions of living organisms means they’ve been selected for survival and will continue to be. It might be random to choose a morality that favours blue eyed people over brown eyed people, but if one of those actually led to better survival rates (for all of life, as I argue in my paper), then that’s the one that *ought* to be followed after a rational examination of the consequences. It might have taken blind variation and selective retention (BVSR) to get this far, but we aren’t so blind anymore.

Chris Jensen 15th June 2015 at 1:51 pm

Hey, but that guanine instead of a cytosine also only survives if it leads to a more-adaptive trait. I am not sure that the cultural values that motivate human behaviors are all that different.

I agree to some extent that a major difference between cultural and biological evolution is that intent can enter into cultural evolution, which makes it subject to not-so-blind variation and selective retention (nsBVSR?). I still have the gut feeling that for all our intents behind new cultural creations the BVSR process might still describe culture pretty well… where our cultural creation sits on the nsBVSR to SVSR spectrum in large part depends on how much we use philosophy and science and other ways of knowing to guide ourselves towards more durable (sustainable?) cultural creations. Not sure that a comprehensive appraisal of what has survived in human cultures would indicate that design intent — rather than blind trial-and-error — played a critical role… but I also have no idea how one would make such an appraisal.

Ed Gibney 16th June 2015 at 5:41 am

Well, philosophers would like to think their logical arguments have played a historical role in the selective retention of culture (a different kind of S from natural selection), but I suppose we could be like Winston Churchill’s America – always doing the right thing…after trying everything else.

Mark Sloan 16th June 2015 at 10:33 pm

Chris,

Glad you like the rational psychopath cooperation thought experiment. Psychopathy is one of my interests. As you describe, such experiments using real people are not impossible and could produce interesting results. I remember a paper, or article?, that looked at the same question, but studied a class of medical students from first year to graduation. What they found was that in the first year, the psychopath end of the spectrum students did better because they got the study groups (which were important to student’s success) to focus on what they did not understand. But by graduation, the normal students were getting better grades because they had sought each other out for study group partners and were avoiding the more psychopathic students. In any such study, the opportunity to preferentially cooperate with good cooperators would be an important variable (as is made evident in game theory).

As I described in the article, I have already tested the truth of my and Pinker’s hypotheses about our moral emotions. I’d like to discuss why you think that testing was inadequate. The must-be-explained data set I tested against was the specific, biology dependent, emotions triggered by our moral sense and the specific cross-culturally universal circumstances when these moral emotions are triggered. Criteria for provisional truth were 1) explanatory power, 2) no contradiction with known facts, 3) simplicity, and 4) integration with the rest of science. (Suggestions for other relevant criteria are welcome.)

My hypothesis about the biology underlying our moral sense being cooperation strategies met these criteria beautifully, revealing the unifying structure and function of every element in the superficially diverse data set. Pinker’s hypothesis about kin altruism and purely self-interested cooperation selecting for our moral sense fared very poorly by these criteria. Where do you see a problem?

I understand you also thought the “different accounting schemes” perspective on gene, individual, and group selection contradicted your point: “To say that selection between groups is responsible for the traits we see today is fundamentally different than saying selection between individuals is responsible for the traits we see today, and to say that selection between genes is responsible for the traits we see today is even more different.” I agree they may appear to be contradictory, but don’t think they actually are. I see the “different accounting schemes for the same phenomena” perspective as complimentary to the three distinct mechanisms view.

What would convince me to agree with you that they are contradictory? I would agree with you if gene level selection, in fact, cannot “in theory” explain traits that were selected for at the individual and group level as I have claimed. In that case, I would agree that saying “gene, individual, and group level selection are different perspectives, or accounting schemes, for the same phenomena” is incorrect.

On the other hand, would you agree that if gene level selection can “in theory” explain all traits selected for at the individual and group level, then there is no contradiction with your point about different mechanisms?

I assume “different mechanisms” just refers to the selection forces acting at different levels. So my assertion might be phrased as gene level selection can “in theory” explain all traits regardless of the level selection forces are acting on. Does that seem right? I am beginning to suspect this may come down to what we mean by “gene level selection”.

Thanks for this opportunity to engage on such interesting subjects!

Chris Jensen 17th June 2015 at 11:54 am

Mark, I think my objection is pretty simple: I think that these sort of logical arguments only go so far. I do agree with you that it is a lot easier in my head to explain moral emotions as being dedicated to maintaining the fitness of groups, but that logic in my head is still just a hypothesis. To actually test that hypothesis scientifically, we need to look at the predictions made by the hypothesis and see if we can reveal patterns in the real world that confirm these predictions.

As far as the individual-level versus group-level hypotheses go, there is a clear difference in the predictions made by each:

Individual-level selection predicts that all the benefits of a costly behavior should come solely from benefits reaped by the individual (including — perhaps — inclusive fitness, although this can get dicey and charge us into “same result, different form of accounting” territory), whereas group-level selection predicts that some of the benefits of that costly behavior can only be reaped when the group outperforms other groups.

While simple enough to state, this prediction turns out to be devilishly hard to test; hence the argument. Plenty of scientists at the interface between psychology and evolutionary biology are working on experiments and other means of collecting data to try to differentiate between individual- and group-level selection on behaviors. That this work has not provided a definitive answer for even narrow classes of behaviors is an indication of how hard this basic difference in predictions can be to test.

The ‘rejection of group selection’ was made by mostly verbal, logical arguments (see Williams 1966 or Dawkins 1976). As a result, intelligent evolutionary biologists stopped thinking about a critical issue. Enough argument. Show me the data!

Mark Sloan 19th June 2015 at 1:10 am

Chris, I came across a good reference for a modern take on the compatibility of the “different mechanisms at each selection level” view with “each level’s perspective is just a different accounting scheme for the same phenomena”.

It is a 2015 paper by Kurzban on the evolution of altruism – but really it is about cooperation in general.

The Evolution of Altruism in Humans by Robert Kurzban, et al.
http://www.zoo.ox.ac.uk/group/west/pdf/Kurzban_etal15.pdf

A relevant quote is
“… many old debates have been resolved, freeing up time and resources to address new and often more interesting questions. For example, we now know that the different theoretical frameworks, such as kin or multilevel selection, are just different ways of dividing up the dynamics of natural selection and are not competing hypotheses. These frameworks all give the same results, predicting that individuals should appear as if they were designed to maximize their inclusive fitness (West & Gardner 2013). This allows us to focus on a new set of problems.”

Sounds like that equivalence is a settled issue in his mind.

But I also had an idea for a simple example of how that perspective equivalence can work. That example is how many psychopaths will be in evolutionary equilibrium if psychopathy is fitness enhancing within a group, but decreases group fitness for between group cooperation. I’ll see how it goes.

Regarding my evaluation of hypotheses about moral emotions, you said “I think that these sort of logical arguments only go so far”. But I am not testing these hypotheses by logical arguments, meaning deductive methods. I understand both of us propose testing such hypotheses by inductive arguments, specifically inference to best explanation. You say “To actually test that hypothesis scientifically, we need to look at the predictions made by the hypothesis and see if we can reveal patterns in the real world that confirm these predictions.” Yes, but equivalently, I have described testing those hypotheses by seeing how well their predictions (or explanations) can reveal patterns in the known data about our moral sense.

Is our difference just that you are proposing making the predictions before data gathering is done and I am looking for the best explanation of data we already have? Predictions before data gathering can improve credibility in some cases, but seem inappropriate regarding our moral sense which we already know quite a bit about. Also, it is not clear to me how much there is left to learn that will be relevant.

In my case, my must-be-explained data set are the emotions triggered by our moral sense and the cross-culturally universal circumstances that trigger them.

Perhaps it would be worthwhile to describe the diversity of this data.

The emotions triggered by our moral sense are1 compassion, loyalty, gratitude, disgust, contempt, shame (when one’s reputation decreasing immoral behaviors are observed by others), guilt (after immoral behaviors are done), indignation (when immoral behaviors are witnessed), and emotional ‘elevation’ (‘elevation’ is experienced when morally admirable behaviors are done or witnessed and is an emotion encompassing feelings of satisfaction and pride). Human ‘moral’ emotions are distinguished by their motivating behavior concerning norms which both elicit indignation when violated and emotional elevation when unselfishly fulfilled.

The must-be-explained cross-culturally universal circumstances that trigger moral emotions Include2 care/harm, fairness/cheating, loyalty/betrayal, respect/subversion of authority, and respect/subversion of ‘purity’ or sacredness.

Due to this data set of emotions and triggering circumstances’ superficial diversity they are unlikely to all be well explained by a simple, single hypothesis by chance and should be highly effective in discriminating between hypotheses.

1. Haidt, J. (2003). The moral emotions. In R. J. Davidson, K. R. Scherer, & H. H. Goldsmith (Eds.), Handbook of affective sciences. Oxford: Oxford University Press. (pp. 852-870).
2. Grahama, Jesse, Haidt, J., et al. (2012). Moral Foundations Theory: The Pragmatic Validity of Moral Pluralism. Available at http://ssrn.com/abstract=2184440

What do you think? Have I done a better job of presenting how I am testing competing hypotheses about our moral sense?

Mark

Chris Jensen 19th June 2015 at 12:22 pm

I still think that there is confusion here about what hypotheses we are testing, and about what the controversy is really about. Notice how sloppily we (and much of the rest of the folks who debate this) throw around the following potential targets of selection: gene, individual, kin, and group.

What Kurzban et al. are referring to — which is really the West and Gardner paper of 2013 — is that an “inclusive fitness” approach and the approach of multilevel selection can be equivalent. This has also been pointed out by David Sloan Wilson, I believe in his book Unto Others (Sober and Wilson 1998), but perhaps later than that. This is an important thing to recognize: that under some circumstances the evolutionary mechanism producing altruistic behavior can be as-well-characterized ‘from different angles’ by these two approaches.

But to understand the significance of this, we still need to work in these other targets (kin and gene). The kin selection argument is an interesting one that can have some meaning if the group that receives the benefit of altruism is composed of closely-related individuals. But kin selection is a unique (if not crucial) sub-phenomena because basically all it says is that the best way to aim your altruism at individuals that have a high probability of sharing your genetic tendencies towards altruism is to aim your altruism at individuals with more recent common ancestry. Directing altruism at kin is just one way of taking advantage of the inclusive fitness/group selection mechanism, but all that really matters is that by some mechanism individuals form groups in which their genetic tendencies for altruism are likely to be highly prevalent. So I would call kin selection a “special and very important” case because it seems like the most common means by which group selection can work (selection on the kin group).

Gene selection is — to be frank — an explanatory dead end as far as I am concerned. This is a whole other topic, and one that I am excited to talk more about when I finish reading The “Random Genetic Drift” Fallacy. But put in most basic terms, so-called “gene selection” only emerges as the result of many generations of selection on individuals (and perhaps groups). What the “equivalence argument” really says is that group selection and kin selection can both be considered by inclusive fitness, which is a way of accounting for genetic success that accrues due to the success of some group rather than just the individual. The link between so-called gene selection and kin selection is gone — never existed really — and so the question is whether we can explain the evolution of altruism by selection on individuals alone or whether we need to invoke some sort of group (and — critically — what group) in order to understand the evolution of altruism. How to detect that signature of group selection — if in a given biological situation it even exists — is in my eyes the remaining ‘controversy’… really just a scientific problem that we ought to tackle together.

I do think that if the only pattern in nature that you have to work with are extant behavioral tendencies (such as those produced by our moral emotions), then you have to work with those patterns. A stronger test would be to recreate scenarios that allow us to understand the evolutionary pressures that favored the evolution of these emotions, but this sort of recreating evolution is probably not going to work in any meaningful way in humans. So the question is “are the ways that moral emotions motivate behavior more consistent with selection at the level of the individual being the dominant evolutionary mechanism, or selection at the level of the group being the dominant evolutionary mechanism?”. In other words, are there observable patterns that can distinguish between contrasting predictions made by these two hypotheses (individual-dominant versus group-dominant)? Honestly I don’t see how you are rigorously testing contrasting predictions, and I would suggest that some sort of behavioral modeling would be required in order to really properly differentiate between the predictions each of these hypotheses make.

The other issue that really needs to be considered here is the proximate/ultimate distinction. Sure, we feel angry when we are cheated by other people. But ultimately what behavior does that proximate anger motivate? Behavior, not the emotions that drive behavior, lead to evolutionary outcomes. Do behaviors motivated by anger lead to benefits for the individual or the group? If the answer is “both” then we can toss the group consideration out the window because it is just an emergent property of what’s good for the individual. However, if an anger-motivated behavior is harmful to the individual who performs it but raises the overall success of his group… well then we have a group selection argument on our hands. Simply talking about moral emotions without making real-world links to the behaviors they motivate and the outcomes of these behaviors does nothing to illuminate how these moral emotions evolved… which is the fundamental question at hand:

Did moral emotions evolve due to selection at the level of the individual or at the level of the group?

Leave a Reply