a blog about things that I've been thinking hard about

The Biology of Morality

29 July, 2005
morality is not just about "being moral"

Morality is not just "there exist rules, which everyone must follow"

Morality is also about determining the rules, and deciding when to follow the rules, and passing judgement on others.

Your brain is programmed to process moral truths the same way it processes other "truths".

So morality feels absolute, even though it isn't actually absolute.


What is Morality?

I am going to start with a hypothesis:

Morality is a system for defining rules for judging the behaviour of individuals in a society.

As a system of judgement, morality has three major defining characteristics:

Although morality is socially defined, a society does not exist independently of the people who form that society. So morality can only exist because the individuals within a society cause it to exist. If morality is a fairly constant aspect of human society, then there must be some moral instinct which causes morality to exist in the way that it does. We can presume that this moral instinct creates the tendencies underlying the three major characteristics, i.e. that people tend to create and use systems of moral judgement such that:

In the above I have not given a precise definition of "society", and we might wonder if society is defined as a tribe, or a country, or perhaps the whole world. It actually makes sense to leave this term somewhat undefined, as it is a function of the circumstances of each individual, who may belong to various different societies, to varying degrees. Each person will share and use the systems of morality relevant to each society that they belong to, and they may or may not consolidate the different systems. For example, a typical person living in a large city may belong to their family, their co-workers, some mates they regularly go to the pub with, the neighbours in their street, and the country they live in. They will interact with moral systems defined within each of these societies, which may or may not have similar rules in each case.

The above characteristics define morality, but they are not the only factors that determine the nature of specific rules that exist within a moral system. In fact, given only the characteristics specified, the rules of morality could be quite arbitrary, such as "don't wear white shoes in the autumn".

The other factors that determine morality could be called "secondary" factors. It is not that these factors are not important in the determination of moral systems, as indeed they play a major role in the apparent consistency and similarity between different moral systems. But they are separate from the moral instinct per se, and they would exist and have an effect on our behaviour even if we were not moral beings. These factors include:

The Morality of Not Being Stabbed

The role of secondary factors can be illustrated by considering a possible moral rule about what should happen when one person wishes to stab another with a knife. There is a conflict here: person A wishes to stab person B, but person B does not wish to be stabbed. If there was a moral rule, it might state that it is OK to stab someone, or it might state that it is not OK to stab someone.

Now most people, most of the time, definitely do not want to be stabbed. On the other hand, any desire that people have to stab other people is much less frequent, and in most cases less intense than the desire not to be stabbed. So, if there was going to be a moral rule, and this rule was to be applied universally, then most likely it would state that is was not OK to stab someone.

The characteristic of generality would also apply to such a moral rule, and a rule that was specifically about stabbing with a knife would tend to evolve into a rule about not physically hurting other people by any means.

Empathy also plays a role in our views about people hurting each other. If we see someone being stabbed, we feel upset to see their suffering. Empathy is something that can exist independently of any socially constructed system of morality; for example a mother will always empathise with the suffering of her children. Empathy does have similar goals to morality, in that it encourages reciprocal respect for the well-being of others, but I think it is useful to consider it as a factor that contributes to the development of moral systems, rather than as part of morality itself.

Moral Activities

I have defined morality as being a system of judgement. But really there are a number of specific activities relating to morality, which include:

There is no guarantee of consistency between these various activities. For example, a given moral rule may be advocated and accepted by society, but little effort may be made to observe the relevant behaviours and pass judgement on them. Or people may judge certain behaviours to be immoral, even though everyone does them anyway.

Morality and Language

There is a fourth important characteristic of morality, which is to some extent implicit in the other three, which is that morality often depends on language, in that moral principles are or can be expressed using natural language (although there is not necessarily any one definitive expression of any particular moral principle). Language enables the statement of very abstract principles, and it makes it straightforward for the principles to be communicated from one person to another. Language is used to discuss and advocate moral rules, and to consider specific hypothetical examples that these principles might apply to. Finally, we use language to express the moral judgements that we make about the behaviour of others, and also to defend our own actions in relation to the moral principles that they might be judged against.

Language is a uniquely human characteristic, and the dependence of morality on language would explain why morality is also only found in humans, and not even in those animals most closely related to us.

The Purpose of Morality

Given that morality follows a tendency to define a universal shared system of maximally abstract rules applicable to the behaviour of all people, you might suppose that the purpose of morality is to impose a uniform standard of judgement on the behaviour of all the people of the world.

But you would be wrong. Because the purpose of morality must be the same as the purpose of any other aspect of human behaviour, which must in fact be the same as the purpose of any purposeful aspect of the behaviour of any living organism. This purpose is long-term reproductive success, or, to put it in more common-place terminology, having lots of grandchildren. The moral instinct must be an adaptation, something that has been selected for because it increased the long-term reproductive success of the people who had it. All the activities associated with morality (thinking, discussion, advocacy, judgement, enforcement) must occur because they benefit the expected long-term reproductive success of those who engage in them.

A common response to this proposition is to consider extreme examples of moral behaviour, such as the person who follows social codes which require self-sacrifice in the name of the greater good (and such behaviour does occur). But these examples are extreme. Most examples of moral behaviour are not so extreme, and on average they increase the expected reproductive success of the people performing them. Consider the opposite: what would happen if you woke up tomorrow with no concept of what was right or wrong. Would you even last through the day?

The moral instinct is just that – an instinct, and like other instincts it is sometimes a blunt instrument – it does not produce the desired result (long-term reproductive success) in all circumstances.

A classic example of an emotional instinct in the "blunt instrument" category is anger. We can think of anger as an instinct which causes us to behave violently when we are frustrated. Anger has evolved because this is a useful response in many circumstances. But we can all think of examples where a person's angry response has been counter-productive. These examples show that anger is not a perfect strategy, but they do not prove that it is not adaptive on average. A more refined strategy might be possible, but the question then arises as to how such a strategy could be genetically specified, and how much information it would take to specify it, and how long it would take for such a specification to evolve.

Moral Truth and Scientific Truth

The process of determing moral systems of judgement bears some resemblance to the process of scientific investigation:

Now science is a search for a single scientific truth (at least it is if you believe in the Theory of Everything), and the similarities of moral thought and discussion to scientific thought and discussion suggest that the goal of morality is a search for some single "moral truth".

But this is not the case, and confusion between these two types of "truth" underlies two major fallacies about morality.

The first of these is the naturalistic fallacy. If scientific truth is like moral truth, then perhaps moral truth can be determined scientifically. This inevitably leads to arguments where "is" statements have to be converted at some point to "should" statements. But there can never be any a priori accepted method for performing such conversions.

The opposite fallacy does not have a standard name, but it underlies much of what passes for moral philosophy. It consists of the assumption that there does exist a single moral truth and that existing imperfect systems of morality are progressing (or they should be progressing) towards this single truth. A good name for this fallacy is the fallacy of moral truth.

I have already stated that morality is based on tendencies towards universal applicability and maximum generality. So if there is some moral truth to be discovered, philosophers usually assume that it can be reached by increasing the applicability of whatever rules they already believe in, and by increasing the generality of those rules.

But, morality is strongly influenced by secondary pragmatic factors, including, most importantly, what it is that people actually want. And what people want can depend a lot on the particular circumstances in which they live. Over-generalisation leads to moral precepts that go beyond what is pragmatic, and beyond what has any relevance to long-term reproductive success.

For example, if it is wrong to kill people, maybe it is also wrong to kill other animals. Specific examples help to justify this view, since we naturally empathise with other animals that are similar to us, and to those animals we keep as pets, and to the young of other animals that look cute (who would argue that you should be allowed to club baby fur seals to death for their pelts?). But where does such generalisation stop? Should we be allowed to kill houseflies? What about bacteria?

The conclusion is that generalisation just doesn't work for morality the way it does, for instance, in fundamental physics. Physics has rules like conservation of energy, and the irreversibility of thermodynamic processes, and these rules have no exceptions. Moral rules always seem to need exceptions, which are often caused by application of other moral rules, and no one has ever managed to formulate meta-rules that resolve these conflicts in a well-determined manner that everyone considers to be morally satisfactory.

The reason that generalisation taken to an extreme doesn't work for morality is that the tendency towards generalisation is just part of a stragegy for members of a society to settle on a system of morality that is appropriate to their circumstances. Generalisation helps to prevent manipulation of a moral system for the benefit of particular individuals. For that reason it also makes people more willing to accept moral propositions advocated by others, if those propositions are seen as not being designed to favour one person's interests.

But generalisation is not a final goal in itself, and exceptions can be found to any moral principle that causes too much trouble and imposes too high a cost. In such cases, people will adjust the moral rules to incorporate relevant exceptions (and eventually they will generalise the justifications for the exceptions). Also, sometimes people will publicly agree with a moral principle, but in practice they will ignore it, and not make any significant effort to enforce it. For example, we all agree that it is wrong for some people to suffer poverty while others enjoy being rich, but hardly anyone gives most of their money to the poor.

Scientifially Determining Morality versus Scientifically Understanding Morality

It is wrong to suppose that science can determine what is moral and what is not moral, because moral truth is not the same thing as scientific truth. But science can certainly be used to try and understand how morality works, and to understand the processes by which "moral truths" are discovered and by which they change over time.

If someone is advocating a particular moral principle, and they scientifically understand the mechanisms by which morality works, then that might help them to express their arguments in a way which makes them more likely to be accepted by other people. So it is not impossible for science to play some role in the determination of morality as it exists in a society.

So is Morality Absolute or is it Relative?

A lot of philosophical, religious and political argument about the nature of morality is centred around the dichotomy of "absolute" versus "relative". Absolute morality corresponds to the notion of "moral truth", which I have already shown is not as absolute as it seems. Relative morality is a more slippery concept. It is supposedly the opposite of "absolute", and for fans of absolute morality it usually translates into "make it up as you go along". In other words, "relative" is regarded as a synonym for "arbitrary", and arbitrary morality is no morality at all.

The theory of the moral instinct does imply that morality is "relative" to the circumstances of a person and to the existing morality of the society that the person lives in. But this does not mean that morality is arbitrary. The moral instinct includes tendencies towards commonality (between different members of a society), universal applicability and generality. The tendency towards generality has the effect of smoothing out differences between the different circumstances that occur within different societies. For example, one society might have knives and say "don't stab people with knives", and another society might have guns and say "don't shoot people with guns", but these moral rules will tend to generalise to "don't hurt or kill other people with weapons", which is the same in each case.

As I have pointed out, the nature of moral thought is similar to that of scientific thought, and this similarity is responsible for the illusion of absolute "moral truth". In particular the moral instinct creates a strong tendency for different people to reconcile their moral differences, in a manner that minimises any possibility of manipulation to benefit particular individuals (although of course such manipulation will occur when people can get away with it).

Comparing Different Moral Systems

This process of reconciliation strongly affects our perception of the relationships between the moral systems of different societies and also between moral systems that have existed at different times in history.

If two societies A and B exist without any contact between them, then each will develop separate systems of morality. Because of the way that morality develops, there will be some principles common to both systems. But there will probably also be some differences.

Eventually the two societies might come into contact with each other, and they will be exposed to each other's moral systems. The initial reaction of each society will be that the other society has a moral system that is "wrong" in some respects (i.e. anywhere that it disagrees with that of the first society). But this might be followed by an awareness within each society that some aspects of the other society's moral system are not necessarily so bad, or may even be "better" than their own.

As the two societies become more and more exposed to each other, their moral systems will reconcile with each other, eventually turning into just one system. This of course perpetuates the illusion that there is just "one true morality" which applies to everyone.

When we consider societies separated in time, such reconciliation is not possible. The society of the present can judge the society of the past, but those living in the past can only guess what their future might be like. And those living in the present have already been exposed to the circumstances and the reasoning that informed the development of past moral systems. Morality will always seem to "progress", as people become aware of new ways of thinking about morality, and new generalisations that can be applied to existing moral principles (although morality can also seem to be constantly "breaking down", as circumstances change, and moral principles that seem irrelevant or costly are ignored and then discarded).

Often we will look at societies of the distant past, and be appalled by their seeming immorality – for example, their tolerance of feudalism and slavery. However, if we lived in such societies ourselves, we would not have been exposed to the reasoning that causes members of present-day societies to consider such things as "wrong". In past times, a slave might have objected to being a slave, but that is not the same thing as everyone deciding that slavery is immoral. Despite the absence of many aspects of modern morality in earlier societies, those societies still had their own morality, and it was just as important for members of those societies to be aware of that morality and to participate in the processes of moral judgement and moral reasoning as it is for people living in modern societies.

The very processes by which moral "truth" is determined create the illusion that there is such a "truth", and that our own personal version of that truth is the correct absolute version of it (or the best version so far), and that the past history of moral thought consists of "progress" towards our current notion of right and wrong.

But it is just an illusion. To break the illusion, we need to see the processes of moral reasoning for what they are – part of an evolved system for increasing the long-term reproductive success of individuals belonging to a species whose members live together in complex societies.

Vote for or comment on this article on Reddit or Hacker News ...