Jonathan Haidt’s six foundations of morality

Jonathan Haidt and colleagues propose that human beings have six different areas of moral concern: care, fairness, liberty, loyalty, authority, and purity. They argue that individuals and cultures differ in how they define and value these six areas, which they call “Foundations.” For instance, secular Westerners are unusual in giving very heavy weight to the first three. Within the USA, conservatives care about all six and realize that liberals are only concerned about the first three; but liberals fail to grasp conservatives’ concerns for loyalty, authority, and purity. Therefore, conservatives can predict liberals’ answers to specific questions quite accurately, but liberals do not understand conservatives–with consequences for elections and public debates.

I find these results illuminating. For example, it rings true that liberals are pretty immune to traditional concerns about purity; and even when we oppose something because it’s impure (such as completely consensual incest), we refuse to invoke concepts of purity in our reasons. At the same time, we liberals may actually have our own purity concerns. For example, I would never leave a glass bottle in the woods, and I would be at least mildly upset to see one there. I suspect the damage caused by an inert glass bottle is minimal; producing my breakfast probably hurt nature much more than a little littering. But I regard the woods as pure, and a human deposit there as a pollutant–in the moral sense.

We could call my reaction “irrational,” but that would presume that only care, fairness, and liberty are rational concerns. Haidt, who is explicitly Humean, would say that reason is and ought to be the slave of the passions. We human beings happen to have six (not just three) bases for our moral passions. Liberals are just odd; we are not distinctively rational.

The main method is factor analysis influenced by conceptual categories. Factor analysis, which is used very widely, takes people’s actual responses to surveys and looks for clusters that suggest an unobserved “factor.” For example, people may not know that they are driven by a concern for care, but their answers to a range of more concrete questions may cluster together empirically, and “care” may be a good label for this cluster. Since “Individuals are often unable to access the causes of their moral judgments,” the real causes must be unconscious and instinctive. We may not know that care is one of our Moral Foundations; that is an empirical finding.

For Haidt et al., moral factors must have two characteristics in addition to statistical clustering: 1) they must be found empirically in many cultures across the world–although it is to be expected that some cultures will miss some factors–and 2) there must be a Darwinian evolutionary explanation for why human beings would have developed these concerns.  Darwinian explanations are supposed to be not mere stories that illustrate evolutionary theory but findings of biology or archaeology.

Finally, Haidt’s method is cumulative and falsifiable. He and his colleagues propose six “Foundations” right now, but they expect to add more.

A lot of assumptions are built into these methods. Arguably, the assumptions are vindicating themselves pragmatically by yielding valuable insights; but it may also be the case that the assumptions drive the methods, which determine the results. Either way, I see these premises:

1. Haidt et al. are interested in very broad categories (“Foundations”) and will even sacrifice tighter correlations if necessary to bring more survey response under broad headings. One could instead investigate the pervasive but subtle differences that arise between two people of the same cultural and ideological background, or within one person’s thinking about one moral concern. For example, I may think about care in five different ways in respect to one complex interpersonal situation.

2. Haidt et al. broaden the definition of morality to include matters–such as authority and purity–that are viewed as extra-moral by, for example, Kant. But they still name six domains as moral and exclude other psychological constructs that may explain our judgments (e.g., competition, lust, and hatred) from the list. This seems to imply a substantive theory of what counts as “moral,” despite Haidt’s repeated claims that he is a naturalist who seeks to describe, not prescribe. What if I insist that authority is not moral or that Will to Power is?

3. Haidt acknowledges that moral norms and beliefs change, but the Foundations Theory does not seem useful for explaining such changes. Haidt is a liberal who argues that liberals are rhetorically hobbled by our inability to invoke loyalty, authority, and purity. Nevertheless, the country has moved dramatically in a liberal direction on social issues that are entwined with purity and authority (e.g., gender roles). How could that happen so quickly?

4. This stream of research is concerned with the often unconscious causes of our moral judgments. It is not concerned with the way that we articulate reasons and arguments to justify our conclusions. This is because Haidt et al.–like most empirical psychologists–find that our explicit reasons are poor predictors of our concrete judgments. Our reasons are not causal.

And yet I would maintain that our reasons can be good or bad, and it matters that they are good. We are morally accountable for the quality of our reasons. It is not only important to judge and do right, but to think right. Besides, we sometimes form judgments contrary to affect because of our reasons, or we decide what to do when our instincts conflict by considering the applicable reasons, or we persuade other people by citing reasons, or we construct institutions on the basis of reasons that then constrain our behavior. Thus the ways we connect our ideas have at least some significance.

Ultimately, I think that the Foundations Theory is a form of science. It is a cumulative and falsifiable body of research, generating useful findings. Yet one can start in other places and also make progress.

I would start with the assumption that human beings are evolved animals whose brains are the product of Darwinian design. But I would also note that we human beings live and think in immensely complicated social settings that have emerged from the uncoordinated choices and actions of billions of people over thousands of years. Emergent phenomena, such as actual languages, religions, governments, and communities, are at the heart of our moral concerns. Although human beings with evolved characteristics have made these institutions, they take on lives of their own and influence us profoundly while also structuring the decisions we make. For instance, our institutions officially embody norms that we may find onerous and undesirable once applied. Although we may not be psychologically influenced by the abstract ideas that we endorse, we are politically influenced by the abstract ideas that we have written into our constitutions and catechisms. I think most of the action is at this level–in how we navigate complex and emergent systems rather than how we apply very basic biological drives.

Although people from different backgrounds hold contrasting values, there are some commonalities across time and culture. Haidt et al. and many other empirical psychologists assume that these commonalities came first and are morally fundamental, and the differences are relatively superficial. Michael Walzer summarizes this view: “Men and women everywhere begin with some common idea or principle or set of ideas and principles, which they then work up in many different ways. They start thin, as it were, and thicken with age.” “But our intuition is wrong here,” Walzer writes. “Morality is thick from the beginning, culturally integrated, fully resonant.” When we find commonalities in distant places, these  are simply “reiterated features of particular thick and maximal moralities.”

[Sources: Jonathan Haidt, The Righteous Mind: Why Good People are Divided by Politics and Religion (New York: Pantheon, 2012); Jesse Graham, Brian Nosek, Jonathan Haidt, Ravi Iyer, Spassena Koleva, and Peter H.Ditto, “Mapping the Moral Domain,” Journal of Personality and Social Psychology, Vol 101(2), Aug 2011, 366-385; and Michael Walzer, Thick and Thin: Moral Argument at Home and Abroad (Notre Dame: University of Notre Dame Press, 1994), p. 4, 10.]

About Peter

Associate Dean for Research and the Lincoln Filene Professor of Citizenship and Public Affairs at Tufts University's Tisch College of Civic Life. Concerned about civic education, civic engagement, and democratic reform in the United States and elsewhere.
This entry was posted in philosophy, Uncategorized. Bookmark the permalink.
  • Wow. Let me first say that I used to consider myself a traditional conservative – and I didn’t think I had changed so much. But I admit I cringed when I read your intro to Height’s work, to wit: “Westerners are unusual in giving very heavy weight to the first three. Within the USA, conservatives care about all six and realize that liberals are only concerned about the first three; but liberals fail to grasp conservatives’ concerns for loyalty, authority, and purity. Therefore, conservatives can predict liberals’ answers to specific questions quite accurately, but liberals do not understand conservatives–with consequences for elections and public debates.”

    Conclusions such as those in Height, et al. only feed the US far-right’s broad, damning
    generalizations of “liberals – a habit that may well be the single most crippling blow to the public discourse today. I mean, after demonizing someone, where can a conversation go? Your response was measured and pretty generous, considering. You certainly fit our motto: Be Cool. Be PolitiCat.

    https://www.facebook.com/ThePolitiCat

  • The Independent Whig

    What liberal issues are entwined with purity and authority, and how so?

    • PeterLevine

      I was thinking of things like premarital sex. I think a liberal would typically hold that a 17-year-old has a right to consensual sex with a peer. But that violates traditional norms of authority (in this case, parental control) and purity (in this case, chastity or virginity).

      • The Independent Whig

        I think, in your example, it is liberalism’s LACK, or ABSENCE, of “entwinement” with the foundations of sanctity/degradation and authority/subversion that explains why a liberal would typically hold that a 17 year old has a right to consensual sex.

        It’s important to note, I think, that none of this is cut and dried, black and white. We’re talking about aggregates and averages, and trends and tendencies. Liberals tend to do, feel, or think X more than do conservatives, and vice versa. If the moral foundations really are products of evolution, then we all have all of them to some degree. I think it’s fair to say that in none of Haidt’s data does anyone score a
        complete zero on any foundation.

        I think Haidt offers better examples of the use of the sanctity foundation by liberals. One is environmentalism – your bottle in the woods example; protecting and preserving the purity of the planet against its degradation by human actions (e.g. Global Warming). Another example is food, in two ways. The first is the anti-GMO movement, and the second is the pro-organic movement. Both of which tend to be more prevalent among, but not exclusive to, liberals. He observes that neighborhoods near Whole Foods stores tend to be more liberal. Or, probably more correctly, the executives at Whole Foods know their market and are shrewd in their selection of store
        locations.

        Examples of liberal respect for authority are harder to come by, except, of course, for the liberal sacred value of “reason” and by extension liberal deference to the “authority” of “reason” and of “experts.” See, for example, “Rationalism in Politics” by Michael Oakeshott, “Kindly Inquisitors” by Jonathan Rauch, “A Conflict of Visions,” “The Vision of the Anointed,” and “Intellectuals and Society” by Thomas Sowell, and Haidt’s discussion of The Rationalist Delusion in The Righteous Mind and his discussion of the cult of reason during the French Revolution in a talk he gave at the Stanford Center for Compassion and Altruism Research and Education (CCARE) entitled “When Compassion Leads To Sacrilege.” (
        http://vimeo.com/m/24606251 )

        I think your question about how the country has moved to the left so quickly regarding gender roles can be responded to in a couple ways. First, it hasn’t been all that quick. People have been pushing the envelope of gender roles since the Women’s Suffrage movement in the late 19th century. Second, the move to the left can also be explained by the fact that conservatives understand where liberals are coming from. After all, all the liberal foundations are also conservative foundations. Conservatives are not lacking in feelings and intuitions of care, fairness, or Liberty. Their scores in those categories in Haidt’s data are very similar to the scores of liberals. Conservatives can and do tend to empathize with liberals on related issues and can be persuaded by arguments based on those foundations.

        I disagree with your statement that “Foundations Theory does not seem useful for explaining such changes.” I think it explains them quite well, for example, as in the paragraphs above. And also in the sense that Haidt says we are “predisposed but not predestined” by moral foundations; that moral foundations are part of the “first draft” of the human psyche that we are born with, but that the experiences we have in life revise and update that draft. The intuitive elephant is a pattern recognition module, shaped by hundreds of millions of years of evolution to be sensitive to the particular patterns that helped or hindered our genetic ancestors to survive and thrive in the social world we humans created for ourselves. AND, our ability to sense and understand those patterns is refined, enhanced, even changed, or suppressed, but the experiences we have throughout
        life. Those changes can come by way of experiences that are repeated over long periods of time, or through single searing experiences, as in the Keith Richards example in his book. In a way, those pattern recognition skills become the psychological/social equivalent of the muscle memory that athletes develop from, for example, repeating free throws hundreds of times a day, or from playing in so many basketball games that they get a sense, or a “feel” for the game that allows them to “see” and react to things that us mere mortals would miss.

        As to your question, “What if I insist that authority is not moral or that Will to Power is?” The answer is just a few paragraphs above it in your own note. Insisting is not enough. You’d have to show that authority does not satisfy the “two characteristics in addition to statistical clustering” that moral factors must have, and that will to power does satisfy them AND statistical clustering.

        Regarding your statement “This seems to imply a substantive theory of what counts as “moral.” Well, yes, the explanation of that theory is the subject of “The Righteous Mind,” and it is only after he has completed that explanation that he defines what counts as moral on page 314 of that book. One of the criticisms of his work is that he doesn’t give a crisp definition of the criteria for “foundationhood.” He acknowledges that criticism and addresses it in his paper Moral Foundations Theory: The Pragmatic Validity of Moral Pluralism.

        I think Haidt agrees with your sentiment that reason matters. I think it’s fair to say that the belief held by many that he thinks reason doesn’t matter is a source of considerable frustration for him. He has pointed out, for example, that his Social Intuitionist Model, on page 55 of the book, involves reason in several ways. Also, see this post by Haidt on the New York Times “Opinionator” blog: http://opinionator.blogs.nytimes.com/2012/10/07/reasons-matter-when-intuitions-dont-object/?_r=0

        I think he’d also agree with your sentiment that we are influenced by the abstract ideas that we have written into our constitutions and catechisms. He has said, “Culture and the psyche make each other up.” (quoting somebody else, I think) And the culture that we swim in provides the experiences that refine, enhance, and change the pattern recognition modules that are moral foundations, and morality.

        • PeterLevine

          Thanks–this is useful and insightful.

          As a point of clarification: I was trying to make the same point that you do in your first paragraph. It’s not that liberals’ views of an issue like teen sexuality show a concern for authority or purity. It’s rather that liberals pushed for a change in norms that had been entangled with conservatives’ views of authority and purity.

          You make a good point that the change in those norms didn’t happen overnight; it had a history. I continue to think that the change came surprisingly quickly if our moral beliefs have evolutionary bases and are largely unconscious. Then you would expect change to come on the scale of centuries, not years or decades.

          Let me try one more example about the scope of morality. People have mostly believed in male sexual prerogatives over women. The idea that men want to, and are entitled to, have sex and that women should not want it much but should submit to men is not universal, but it is very common across cultures and times. It also has an evolutionary basis. So why not say that it is “moral”? We could add “male sexual dominance” to the list of Moral Foundations and merely note that modern Americans happen to be uncomfortable with it.

          My answer would be based on philosophical argumentation rather than empirical observation. I would argue that male sexual dominance is a human sin/frailty, not an aspect of our morality. But this would seem to be a point of disagreement with Haidt et al.

          • The Independent Whig

            Re: your first paragraph. I suspected that but I wanted to “put it out there” so to speak to confirm or deny my suspicion. We’re on the same page.

            Re: your second paragraph. I guess it depends on where you start to measure change. One could argue, for example, that the change started with the events leading up to Magna Carta, in which case it has indeed been centuries. I think the discussion then gets sorta sidetracked into definitions of what we mean by fast and slow, or what constitutes “change. ” No matter how you look at it I think we’d agree, as the old TV commercial said, “You’ve come a long way baby.” (I think it was a commercial marketing cigarettes to women, how’s that for an issue about gender roles?)

            I think you’re right about the scope of morality. Morality is made up of much more than just six moral foundations. Even Haidt calls them the low hanging fruit, and solicits ideas for additional ones, or challenges to the current ones. There’s also more to it, probably, than just foundations; maybe other psychological factors or personality traits are part of the equation too. And if we go down that road, then a question that woul naturally arise is which is the chicken and which is the egg?

            Still, I think Moral Foundations Theory offers a brand new lens through wich we can take a fresh look at human social behavior, both now and looking backwards through history. I think it adds lots of new insights into why things happen, and happened, the way they do.

            It also gives us clues as to steps we could take to shrink the size of the political divide and reduce the amount of demonization that flows back and forth across it.

          • The Independent Whig

            Regarding your final paragraph about philosophy:

            I’m not a big fan of philosophy per se. In general layman’s terms I would subscribe to the “philosophy” that all philosophies are wrong because they rest on a limited set of assumptions that can’t possibly explain everything, or upon a particular moral matrix (set of moral foundations).

            Your statement that male sexual dominance (all your examples seem to be about sex) is a sin/frailty is normative, not descriptive. It’s a personal judgement derived from your own moral matrix. That’s not a philosophy, that’s an opinion.

            I think Haidt is on the right track to remain as descriptive as possible and to avoid normative opinions like yours as much as possible.

            If the goal is civic engagement and civic renewal then isn’t it most important, even imperative, that it be based on the empirical reality of how the human mind actually works, and why people think and behave the way they do? If not, then what’s the point, other than to advance your own personal political opinion/agenda, an agenda that could very well be a Procrustean imposition of a square peg “philosophy” into the round hole reality of fundamental human nature?

          • PeterLevine

            A moral statement such as “women should not be treated as inferior to men” is not merely an opinion. It is an opinion of mine that I can defend with reasons. By making such a point, I make myself accountable for offering supportive reasons. If we think that it’s merely an opinion, then there is no moral difference between, say, a Nazi and an American political moderate. They both have opinions, which just happen to be different. It would also just be an opinion that killing people for fun is fine. Normative claims are not the same as empirical claims, but they are not merely subjective, either. Some are right and some are wrong–many are uncertain–and all require normative reasons and reasoning.

          • The Independent Whig

            I agree that there is such a thing as normative claims that are right and wrong. And I agree that normative claims should be supported with reasons and with reasoning. But I think you might have misconstrued the point I was trying to make when I used the word “opinion.” So to avoid getting into a semantic debate over the definition of “opinion,” I’ll try again to explain myself without using that word.

            The Nazis thought they were morally right and offered supportive reasons and reasoning.

            Al Qaeda thought it was morally right to fly planes into buildings and posted many videos on the internet offering supportive reasons and reasoning.

            ISIS thinks it’s morally right to decapitate people and similarly offer supportive reasons and reasoning.

            The reasons and the reasoning – the philosophies – of the Nazis, Al Qaeda, and ISIS are based on the moralities – the moral matrices – within which those groups operate(d). Similarly, the moral statement that women should not be treated as inferior to men, or any moral statement an American political moderate might make, and the reasons and reasoning (the philosophy) offered in support of those statements, are based on the moral matrix of the person making the statement.

            A moral matrix, or a morality, is a “closed epistemic world” that “has within it everything it needs to prove itself. And it has within it defenses against any possible argument that could be thrown at it.” “So it becomes impossible to think beyond.”

            (Quotes are from an audio interview of Haidt, here:
            http://www.onbeing.org/program/jonathan-haidt-the-psychology-behind-morality/6341 , transcript here: http://www.onbeing.org/program/jonathan-haidt-the-psychology-behind-morality/transcript/6347#main_content

            The limits and the extent of the epistemic world within which each of us lives, and beyond which it becomes impossible for each of us to think, are defined largely by the degree to which each of us employs each of the moral foundations.

            Moral foundations are cognitive tools of social perception (pattern recognition modules, as in my previous note), subconscious social intuition (pattern interpretation modules), and conscious social reasoning (argumentation and persuasion that the patterns perceived and understood are actually there, and meaningful).

            Since all philosophies and all moral statements are products of the moral matrices from within which they originate the question becomes, therefore: Does the matrix, and therefore the reasons and reasoning offered in support of the moral statements that arise from it, reflect an accurate grasp of human nature that is supported by empirical evidence, or does it not?

            Haidt’s empirical findings show that the more foundations one employs the better one understands human nature (see Haidt on Colbert and Moyers, easily found online), and the better able one is to understand and empathize with one’s political opponents (as you summarized in your original piece, as Haidt describes on pages 334 and 335 of The Righteous Mind). He also finds that the philosophies of philosophers who employ all the foundations tend to be more congruent with human nature than those of philosophers who employ fewer foundations, or who employ them to a significantly lesser degree (see pages 337-338 of The Righteous Mind specifically, but all of Haidt’s work, really).

            So, I ask again, if the goal is civic engagement and civic renewal then don’t you think it is important, even imperative, that it be based on the empirical reality of how people actually perceive, understand, and reason about, and act in the social world? [i.e., that a) intuitions come first, strategic reasoning second, b) morality is about more than care and fairness, it’s also about liberty, loyalty, authority, and sanctity, c) morality binds and blinds (The three sections of Haidt’s book), d) that the more foundations one employs the better grasp one has of basic human nature, and e) the philosophies of philosophers who employed all the foundations have been borne out by the empirical evidence of disciplined scientific study?]

            In other words, don’t you think civic engagement and civic renewal should be based on the empirical evidence about why people think and act the way they do? Don’t you think that civic engagement and civic renewal would be easier, more productive, and maybe even more fun, if everybody actually understood where people who think differently are coming from? And shouldn’t civic engagement and civic renewal START with fostering mutual respect and common understanding through increasing everyone’s knowledge of those things? How can we possibly expect people to better understand one another and through that get along better if they never learn why those things can be so hard to do?

            Or, do you think civic engagement and civic renewal is somehow possible if half the population stays firmly ensconced in their “comforting delusions” (from Haidt’s 2008 TED Talk) that the other half just doesn’t get it, when in fact it is they who don’t get it?

            Or, are you interested only in advancing your own particular sense of right and wrong, and/or that of the particular moral tribe(s) you belong to?

          • PeterLevine

            Hi again,

            What I want is a conversation with diverse people who exchange ideas and work together, reflect on the experience, and change our views so that they are more just and good. I see that conversation as the very heart of civic engagement.

            A crucial precondition is that I, like others in the discussion, are modest and fallibalist about my own views. I am more than likely wrong. That is because there is such a thing as being morally (more) right, but it’s a hard pursuit. In the effort to reason about what is more right, it is very important to understand where other people are coming from and to do so as charitably as possible–really grasping their views. Yet we are not merely trying to understand; we are seeking to improve.

            If everyone belongs to a “closed epistemic world,” and there is more than one such worldview in the world, then deliberation with people from different “worlds” is impossible. That’s what “closed” implies: “So it becomes impossible to think beyond.”

            Since my very first book in the early 90s, I have been arguing that people differ in their moral views, but it is a high-modernist prejudice to presume that the difference takes the form of isolated islands of moral thinking: matrices, in Haidt’s metaphor. I argue that individuals hold many moral views that tend to be connected, forming conceptual networks. The number of nodes in each person’s network is more like 6,000 than six. Further, our networks overlap with each others’, even though each one is unique. Deliberation is a matter of explaining nodes and connections to other people, and it is frequently effective, leading to change in their respective networks.

            In that sense, it is both possible and obligatory to “think beyond,” and no one has to have a “closed epistemic world,” although some people have moral networks highly centralized around a few nodes. They look like Haidt’s model, but we call them ideologues.

            These links might be helpful:

            http://peterlevine.ws/?p=14034
            http://peterlevine.ws/?p=13812
            http://peterlevine.ws/?p=13806

            Thanks a lot,
            Peter

        • The Independent Whig

          I’m not a big fan of philosophy per se. In general layman’s terms I would subscribe to the “philosophy” that all philosophies are wrong because they rest on a limited set of assumptions that can’t possibly explain everything, or upon a particular moral matrix (set of moral foundations).

          Your statement that male sexual dominance (all your examples seem to be about sex) is a sin/frailty is normative, not descriptive. It’s a personal judgement derived from your own moral matrix. That’s not a philosophy, that’s an opinion.

          I think Haidt is on the right track to remain as descriptive as possible and to avoid normative opinions like yours as much as possible.

          If the goal is civic engagement and civic renewal then isn’t it most important, even imperative, that it be based on the empirical reality of how the human mind actually works, and why people think and behave the way they do? If not, then what’s the point, other than to advance your own personal political opinion/agenda, an agenda that could very well be a Procrustean imposition of a square peg “philosophy” into the round hole reality of fundamental human nature?

  • There’s no mention in those “foundations” of honesty … which is consistent with my experiences with Haidt.