Is there Value in Teaching Moral Values?

In Japan, being good will soon be a formal subject in school education. The Japanese education ministry must develop textbooks and curricula to teach morality, and tests to grade it, which occasions a host of interesting practical questions (for a thoughtful list). In addition to practical questions about how to implement such a program, there are theoretical questions about whether trying to do so is wise. Moral education has a troubled past in Japan; it was scratched in the 1960s by the Americans because they suspected it taught racism and blind obedience to the emperor. Nevertheless, the idea of teaching moral values seems to be gaining steam internationally. In Britain, for instance, the Jubilee Centre for Characters and Values at the University of Birmingham hopes to form the moral character of British pupils (another attempt). And it just announced the launch of a free online course designed to teach people how to build a morally good character.

Such initiatives, well intended or not, are grounded in the belief that we can effectively teach morality. The Jubilee Centre, comprising educators, philosophers, psychologists, and theologians,  states that moral character is teachable. According to a poll that they conducted, 95% of British parents belief that teachers can form the moral character of their pupils. Most philosophers, in contrast, appear to be sceptical about the possibility of teaching moral values. Jesse Prinz stated that moral interventions or teachings produce weak positive effects at best, but more often have no effect or even backfire. In their call for moral enhancement, Ingmar Persson and Julian Savulescu seem to have little regard for the power of teaching moral values, and recommend developing biological interventions instead.

Even those who believe that it is possible to shape the moral character agree that doing so is not easy. James Arthur, the Jubilee Centre’s director, commented that: “Many schools do not know how to teach character.” Most psychologists would probably say: Neither do I! And a good portion of them would assume that it cannot be done. Take racism, for instance. Persson and Savluescu point out that knowing that racism is wrong – something that could presumably be taught – often does not produce the desired changes in behaviour. Most leading experts who study implicit, automatic racial biases would agree. In spite of a great deal of research, we do not yet know how to effectively change the automatic biases that underlie racial behaviour. Hence, many experts suggest that we should focus instead on establishing effective guidelines and rules that prevent the biases from influencing everyday behaviour.

The definition of moral character underlying the efforts of the Jubilee Centre underlines the difficultly inherent in teaching moral values: “We understand character to encompass the morally evaluable, reason-responsive and educable part of human personality.” This definition, a bit loaded, rests on the idea of a reason-responsive morality. While admirably optimistic about the moral nature of humans, this claim stands in stark contrast to decades of research on morality, which shows everything but a morality responsive to reason.

The most effective moral interventions, to my knowledge, try to avoid reasoning, and focus on the training of compassion via meditation. Richard Davidson’s group, for instance, let 20 people complete a 2-week long compassion training for 30min each day and 21 people complete a 2-week long emotion-regulation training for 30min each day. When they studied the effects of these interventions, they found that participants in the compassion training group showed more altruistic behaviour (albeit in an economic game) than the emotion regulation group, elicited by different neural responses to picture of human suffering. These studies often come with their own set of problems such as a murky picture about how the meditation affects altruism, small samples sizes, and potent demand effects. And they might face additional challenges. As outlined by Persson and Savluescu, such interventions might often fail to reach the people who would “need” it. And British pupils might perceive it challenging to wish themselves and others freedom from suffering, as compassion meditation wants you to do.

Science, at this point in time, remains sceptical about the chances of teaching and changing morality. But morality is nothing special in this regard. Humans aren’t very good at changing their behaviour, whether it is stopping smoking, exercising regularly, saving money, eating healthier, or loving their neighbours. Good intention or not. But the low success rate doesn’t stop us from trying – witness the perennial New Year’s resolutions – and maybe there is some value in it. Teaching morality alongside mathematics might signal to pupils that morality is as important as other subjects. And parents, trying to raise their children to be good people, might not succeed very often, but their attempts have a strong social signalling function. These social signals, from individuals and societies, have better chances of fostering innate social instincts than rational appeals to be good.

  • Facebook
  • Twitter
  • Reddit

Comments are closed.

Authors

Subscribe Via Email

Affiliations