Science Break: Morality

A sense of morality is a universal human trait. Every single human who has taken a breath, no matter where on this planet, or when in time, has had a strong understanding of what is right or wrong. People may not be able to articulate those beliefs, and in fact most people treat morals as obvious and axiomatic to the point where they don’t consciously consider them or where they came from. That in itself is interesting, but what is really fascinating is that these moral “truths” in fact vary from person to person, and especially culture to culture. Is this just one of life’s mysteries, or can science shed some light on this topic? Well of course it can, or else this would be a very short article!

This month we’ll veer away from the hard sciences into social psychology, as work in this field sheds quite a bit of light on the topic of morality, providing some models that can be deployed to better understand the moral architecture of a particular group or person. I stumbled on this while reading Steven Pinker’s latest book, The Better Angels of our Nature: Why Violence has Declined (Pinker, 2011). His section on morality, just part of the larger topic of the book, touches on a more systematic way of looking at the human condition. Morality, the largely subconscious compass that guides our intentions, actions, opinions and decisions is really at the heart of who we are, so I believe it’s a topic of universal interest.

Now right off the bat, readers are going to question the statement that morals vary greatly from person to person and culture to culture. Of course there are certain moral issues where the vast majority of people are on the same side – it’s wrong to steal, it’s wrong to murder, and so on. But beyond those obvious ones is a minefield of contentious taboos related to some of life’s fundamentals – birth, death, violence, eating, sex, ownership, defecation, etc. Recently I read that a big health issue in rural India is a deep-rooted cultural preference to defecate outside. To a North American this sounds simply disgusting, whereas I would imagine to an Indian villager, the thought of doing their business in a little room shared with others is equally disgusting. So that’s the kind of thing I’m talking about, and we can all think of examples that we find shocking – historical attitudes in eastern Mediterranean regions towards homosexual acts between heterosexual males; attitudes towards infanticide in many cultures; the role of women in others. Within our own multicultural society at any given time there will be a number of hot issues – fracking, abortion, Israel versus Hamas, vaccination, etc. – where values clash. Can we make sense of all this?

Recently, two books have changed my perspective on what is “normal” for humans. I’ve always believed, besides relatively minor local variations, that humans are largely the same everywhere, especially in terms of their fundamental beliefs. But then I read Jared Diamond’s The World Until Yesterday: What Can We Learn From Traditional Societies (Diamond, 2012) in which he contrasts our (i.e. the developed world’s) attitudes and practices with those in societies he terms “traditional”, those retaining aspects of hunter gatherer societies that presumably modern societies evolved from. This is truly eye opening. And then I just finished Savage Harvest (Hoffman, 2014) which, in piecing together Michael Rockefeller’s 1961 disappearance in Papua New Guinea, explains the belief systems underlying cannibalism on PNG’s SW coast. Contrary to common notions about cannibalism, one of our sacrosanct taboos, it was fundamentally woven into their world view, and considered right, moral, necessary. This new appreciation for the broad range of human beliefs has me re-examining my own basic beliefs about what it actually means to be human.

But despite big differences in fundamental beliefs, it is not a random landscape – beliefs cluster around areas that are central to human existence, and different attitudes tend to repeat themselves. Every culture will have strong beliefs around things like birth or death, but few or none around things that don’t matter much in life, and completely separate cultures will tend to develop similar beliefs, for example that there is an afterlife, or that a crime should be punished. For some time psychologists have been trying to describe and categorize this segmentation of beliefs, and distill it all into a universal framework with which all people’s beliefs can be measured and modelled. The objective they are seeking is to define a small and simple set of components that are independent, can be measured and together completely encompass the full range of human beliefs. An example of a similar system (modelling something different, but also psychological) that many in the corporate world will be familiar with is the Myers-Briggs Type Indicator (MBTI), which involves a test that can measure and define people’s preferences when it comes to forming opinions and making decisions.

A convenient paper I found (Bruce, 2013) summarizes three currently popular models of human morality – Mary Douglas’ Grid-Group Cultural Theory (CT), Richard Shweder’s tripartite theory, often referred to as Big Three Ethics, and Jonathan Haidt’s Moral Foundations Theory (MFT). Pinker (2011) covers a fourth, Alan Fiske’s Relational Models Theory (RMT).

I am not favouring these four models in particular, I’m just using them as examples of the tools being developed to understand morals, and will leave it to readers to pursue the topic further and develop their own opinions. One thing that is immediately evident is that different from most psychological theories we’re familiar with that focus on the individual, when it comes to moral beliefs, these models acknowledge the heavy influence of sociocultural factors on belief systems. People are heavily influenced by the societies they live in, and as social animals inevitably absorb and embody the morals of their own cultures.

Grid-Group Cultural Theory

This theory perfectly captures the two main dimensions of interpersonal relations – the degree of freedom a society “gives” a person to be themselves, especially in person to person interactions, termed Grid, and the degree to which the group a person belongs to defines his or her self-identity and esteem, termed Group. Figure 1 illustrates this system, and the quadrants it defines. A high + Grid society restricts and dictates what members may do based on individual identification, whereas a low – Grid society doesn’t restrict that sort of thing much at all. In a low – Group social environment people have a low connection with the social group they fall into, and act more or less independently from the group; in a high + Group society, people’s self-image and behaviour is largely dictated by group rules, often not formalized.

When I ponder this model a number of thoughts spring to mind. First, it almost perfectly captures the difference between East and West. Western societies would generally fall into the low Grid / low Group quadrant, whereas Asian cultures would be mainly found in the high Grid / high Group quadrant. I believe that the central theme of many films and novels popular in the West is the tension between a person’s desire for personal freedom and expression, and the restrictions that society (often foreign) places on them. We probably gobble up these stories because they affirm the cultural quadrant our society aspires to: they are the cultural propaganda we consume to bolster our case against values in other societies that are in conflict with our own.

But then even within any given society, one could map out subsets. For example, I think you could argue a person’s income level would help dictate where they would fall along the Group dimension, with higher income people valuing the hierarchy that protects their favoured status, and low income people feeling relatively helpless about their lot in life. Somewhat related, one could draw an age-dependent vector from idealistic youth culture, diagonally up to the conservative hierarchical values held by the older segments of society. The archetype at the centre of the axes, the “hermit” who avoids all social interaction, gives me a chuckle – we all know a few of those, and I sometimes feel like retreating into that little bubble.

But the Grid-Group theory leaves me with a sense that while it has captured one key factor driving morality, it hasn’t actually modelled morality at all. In the words of Bruce (2013), CT lacks the moral lexicon we’re looking for. This is where Big Three Ethics comes in.

Big Three Ethics

Richard Shweder’s work has had an enormous impact on the study of differences between societies. His Big Three Ethics – Autonomy, Community and Divinity – represent areas where fundamental moral belief differences between cultures tend to cluster. There are clear similarities between CT’s Grid and Shweder’s Autonomy, and CT’s Group and Shweder’s Community, but with a difference. In the CT model a person’s actual level of Grid and Group influence is measured, whereas in Shweder’s model he is looking at what levels of Community and Autonomy (and Divinity) are idealized in a society.

Hypothetically, a society that values purely Autonomy sees itself as being made up of individuals, and morality as a set of beliefs and rules that allows these individuals to live their lives as they choose. A society which values only Community sees itself as a single whole, with morals being rules and beliefs protecting that whole and allowing it to operate according to its rules. Lastly a community that values only Divinity views morality as the rules and beliefs which protect the divine essence and keep it pure. In practice the theory accepts that a culture will be defined by the weighted mix of the three Ethics. Observation has shown that typically one or two Ethics are very dominant, with the third often not being much of a factor. For example, a North American religious community would weigh Divinity very high, and then the second most important factor, either Autonomy or Community, would very specifically define them: for the Amish, Community would be very important and Autonomy not at all; whereas a church-going Episcopalian community could weigh Autonomy fairly high similar to general US culture, and Community not very high at all.

Each community will express their important Ethic(s) in their own unique way. For example, many religious communities value purity under Divinity in the form of food taboos or things like circumcision, baptism and so on, while Western democracies place a high value on rationalism and justice under the Community Ethic.

Big Three Ethics, “offers a start to developing a moral vocabulary to fill in Grid and Group,” (Bruce, 2013) but it doesn’t fully inform the complete moral landscape that is observable among humans; being limited to just three very broad Ethics means the many unique and specific types of moral beliefs are not defined and therefore are not measurable. For that, we’ll take a look at Moral Foundations Theory.

Moral Foundations Theory

Moral Foundations Theory (MFT) has been built on the base formed by the first two theories. In fact, Haidt was directly inspired by Shweder’s work, and started adding categories as he tried to better understand the broad range of moral beliefs he observed in various diverse cultures through the lens of Big Three Ethics. An early form of MFT contained four categories, Pinker (2011) lists five, and in the excellent Wikipedia overview on the theory (Wikimedia Foundation, Inc., 2014) there is a list of six:

1. Care/harm for others, protecting them from harm.
2. Fairness/cheating, justice, treating others in proportion to their actions (Haidt has also referred to this dimension as Proportionality.)
3. Liberty/oppression, characterizes judgments in terms of whether subjects are tyrannized.
4. Loyalty/betrayal to your group, family, nation. (Haidt has also referred to this dimension as Ingroup.)
5. Authority/subversion for tradition and legitimate authority. (Haidt has also connected this foundation to a notion of Respect.)
6. Sanctity/degradation, avoiding disgusting things, foods, actions. (Haidt has also referred to this as Purity.)

Haidt essentially took Autonomy and split it into Care/harm and Fairness/cheating, and similarly split Community into Liberty/oppression and Loyalty/betrayal. So again here we see the bloodlines of the theory extending back to Douglas’ CT model. Sanctity/degradation is essentially the same as Divinity, and Liberty/oppression is a new dimension created by Haidt.

MFT was initially developed to analyze differences between cultures, much as Big Three Ethics was, but social psychologists are finding that it is a tool that is broadly applicable, and are using it to model other social differences, such as those between different political ideologies. For example, it has been shown that people we call liberals will tend to value Care, Fairness and Liberty and not so much the other three, while conservatives value all six fairly evenly. The fact that it can be applied in this way to analyze almost any aspect of human belief suggests to me that MFT is getting pretty close to nailing the objective I mentioned in the introduction – it is a tool that can be used to measure and uniquely identify any person’s or group’s moral attitudes and beliefs. What researchers can do now is measure the levels of the six variables using standard statistical methods to ensure reliability, and the results will give a unique moral “fingerprint” for the test subject.

An aspect that I find particularly interesting is that Haidt basically views morals as a form of heuristic, although I am not sure he ever used that term. Haidt’s approach, “suggested that “moral judgment is caused by quick moral intuitions” while moral reasoning simply serves as a post-hoc rationalization of already formed judgments. (Wikimedia Foundation, Inc., 2014)” This perspective makes perfect sense, as heuristics are a set of evolved mental short cuts which allow us to make quick decisions in situations where we lack sufficient information to make logical ones, as I discussed in my December 2012 Science Break article on heuristics, and morals deal almost entirely with areas lacking complete information. Research has shown that the brain circuitry for heuristics gets shaped and modified through experience and learning, especially during childhood, so this would explain how moral beliefs can be so different from one community to the next.

Relational Models Theory

In his Relational Models Theory (RMT), Alan Fiske defines four basic categories of morals (Bolender, 2011): Communal Sharing (having something in common and sharing resources), Authority Ranking (arrangement into a hierarchy), Equality Matching (striving to maintain egalitarian relationships), and Market Pricing (use of ratios). One of the key assumptions underlying the theory is that resource allocation within a group or society is an important driver of the development of morals. Pinker (2011) provides a table showing his opinions on how Big Three Ethics, MFT and RMT line up relative to each other, which I’ve modified to include Liberty/oppression (Table 1). I’m not sure if my decision to combine MFT’s Authority/respect and Liberty/oppression into RMT’s Authority Ranking is valid, but I can argue a case for it. At any rate, the fact the three models can line up so closely shows how similar they are, but obviously there are differences.

Pinker argues that Fiske’s Communal Sharing is a combination of MFT’s Purity/sanctity and Loyalty/betrayal. I believe this flips things around, changing Divinity from an external factor which drives or creates morals around purity and other taboos, and shifts that collection of rules into a result of the moral imperative of Communal Sharing. In other words a community which values Communal Sharing will develop many of the aspects of what we call religion, especially the taboos and rituals which bond a community together, including a common belief in a deity or deities, as a way to support and protect the sanctity of the group and its sharing of resources among members.

I’m not going to go into great detail on the four RMT components other than this brief summary. Communal Sharing is just that – it captures moral beliefs associated with a community’s sharing of resources. Authority Ranking contains all moral beliefs associated with the power structure of the group. Equality Matching captures all morals associated with reciprocity and the you-scratch-my-backand-I’ll-scratch-yours social etiquette that is so important to humans. And lastly, Market Pricing / Rational-Legal contains all the rules that a modern society contains around how resources are controlled and allocated in a modern economy; these latter could be argued to be beliefs and taboos, I suppose.

Fiske has noted that his scale can be lined up, “along a scale that more or less reflects their order of emergence in evolution, child development, and history: Communal Sharing > Authority Ranking > Equality Matching > Market Pricing (Pinker, 2011).” To me, the category of Market Pricing doesn’t quite fit or ring true, and is perhaps a result of cultural arrogance in that we feel our own modern western society is different and better than others, and as the ultimate expression along a trajectory of continual improvement, deserves its own category to contain rules and other aspects which we feel make our culture superior – our laws, our complex markets, our supposedly more rational way of going about things. Perhaps if we took a step back we could see that all these trappings are variations of aspects contained in the other three categories.

The models I’ve described were designed to analyze belief systems and compare different cultures’ moral frameworks. But they’ve provided so much more – we now have a language or grammar with which to discuss many aspects of morals and the social norms that are built around them. Pinker (2011) states that each moral norm of any importance applies to a situation involving the following: one of Haidt’s relational models, a social role (husband, friend, employee, neighbor, etc.), a context (work, gym, store, highway, etc.), and a resource (labour, money, food, land, etc.). In any functioning society its members are expected to understand and apply the appropriate rule or norm to each situation. When a social norm is broken, this evokes some degree of moralistic outrage and punishment. When the resource in question is considered sacred in some way, then a taboo has been violated, and the response to the violation is much stronger. So if a person violates RMT’s Equality Matching by not thanking a friend for a gift, then the punishment is the social equivalent to a slap on the wrist, such as not being invited to parties for a while. But if the violation involves something sacred, perhaps MFT’s Purity/sanctity norm in a Middle Eastern country, then the response could be as severe as death by stoning. As an aside, sociopaths appear to exploit an ability to operate outside of society’s moral norms, something the rest of us find extremely difficult to do, and they are so brazen about it that the rest of us are caught unawares and are quite easily taken advantage of.

Philip E. Tetlock has done a lot of very interesting work using Fiske’s RMT as a starting point. For example he shows how the relational model within which an economic exchange of resources takes place affects or complicates that exchange (McGraw & Tetlock, 2005). Examples of this would be pricing distortions involving the exchange of resources imbued with some special property (like a polyester jersey with a famous athlete’s signature on it), or that certain awkwardness when selling something to a friend that clearly is too valuable to give for free. In another line of research he defines three types of tradeoff – Routine, Taboo, and Tragic (Tetlock, 2003). Routine tradeoffs involve choices within one relational model such as donating one’s estate to the church rather than one’s relatives; Taboo tradeoffs have a person choosing between a secular value and a sacred value, such as selling your children into slavery; a Tragic tradeoff is between two sacred values, such as a society that valued Communal Sharing and Equality Matching wrestling with whether to exile lepers to a remote island to protect the rest of society.

To wrap up, I want to highlight an extremely interesting aspect of RMT that Pinker (2011) points out, which will allow me to end with a mention of the hard sciences. Pinker is a neurolinguist by training, and his most interesting books involve explorations of the neurological underpinnings of human skills, senses and behaviours – he’s an expert in this area. What he notes is that the first three RMT categories (that is excluding Market Pricing / Rational-legal) can be connected to neurological systems in our brains. Following the morals held within the Communal Sharing category is likely rewarded neurologically by the release of neurotransmitters such as dopamine, endorphins and oxytocin. Authority Ranking morals can be seen as components of the dominance systems we’ve inherited from our primate ancestors, involving testosterone circuitry. Regarding Equality Matching, much research has shown how finely tuned the human brain is to keeping track of fairness in the give and take relationships we have with others in our communities. And lastly in a general sense there is the heuristics aspect to morals, which is also neurological. So I believe that these psychological taxonomies which aim to model, define and categorize human morals, will eventually lead away from the social sciences and into neurological explanations.

In the course of writing this article I find myself viewing moral conflicts or stances more analytically, particularly using Haidt’s six moral foundations. As I get in the habit of doing this, it actually seems quite simple and easy to do so, and morality no longer seems to be an impossibly murky area, and I no longer feel compelled to try to view it in black and white terms. Instead I now feel that anything involving morality – a person’s or group’s moral position, or a particular situation – can be described and understood using the moral foundations and whether they are respected or violated:

• Is someone or some people being harmed or cared for?
• Is some party gaining a benefit without paying the cost for it?
• Where on the scale between liberty and oppression do things lie?
• Is an agreement or coalition being honoured or ignored?
• Is an authority being respected or ignored?
• Is some sacred personal or community value being sanctified or degraded?

Once broken down this way, I feel the core issue(s) can be quickly identified. It doesn’t necessarily change my opinion, but it does help me understand other points of view better.

Note: September’s article The Math of Mosaics contained an embarrassing but funny error. I referred to the Tomb of Hafez star in Figure 4 as having 17 points. David Monk pointed out to me that it clearly has 16, and flatteringly wondered if it was a cunning joke or test, given that I’d gently ridiculed mathematicians for identifying the Gonbad-e Qabud Tower in Figure 11 as having 8 sides when it obviously had 10. I was tempted to say yes, I’d planned it, but no, I didn’t – it was a mistake!