Monday, 12 January 2009

Evil and Empathy


In my last posting I associated evil with moral disability rather than moral insanity as proposed by Grayling in the New Scientist of 17/05/08. I stated it is hard to be angry with someone who is disabled due to her disability. However it appears to follow if my definition of moral disability is accepted that it becomes hard to associate anger with evil. Intuitively anger seems to be a justified reaction to evil, perhaps even a necessary reaction, see Hugh Thompson's  anger in response to the My Lai massacre during the Vietnam War. Further accepting my definition might imply that if someone is morally disabled it is hard to hold her morally accountable for her actions. In this posting I will further examine the ideas of moral disability and understanding in an attempt to resolve these two problems.

In  moral insanity  I defined someone to be morally insane if she acted contrary to accepted moral dictates she understood due to her inability to feel empathic concern for others. In that posting I argued moral disability is a more useful definition than that of moral insanity. I defined someone to be morally disabled if she did not have the capacity to feel sympathy for others when making moral decisions. Critical to both these definitions is what is meant by moral understanding. I have previously argued moral understanding depends on an affective element. It might be objected of course moral understanding should not be dependant on our emotions. Moral understanding should be based solely on reason. This would certainly be the Kantian position. Let it be accepted that moral understanding is based on reasons. Reason and reasons need not be identical. Reasons may be purely logical or simply based on our feelings, emotions. It follows if moral understanding is based on reasons that moral understanding might at least include an affective element.

In order to get a better grip on the idea of moral understanding we must examine moral reasons. Internalism about moral judgement holds if someone makes a moral judgement that x is wrong then she has a motivation not to do x. I believe internalism concerning moral judgements is true. Moreover I doubt if someone who judges that doing x is wrong but has no motivation not to do x truly understands the meaning of moral judgement. In what follows I hold if I have moral reasons not to do x then I am also motivated not to do x. For instance if I believe stealing is wrong I am motivated not to steal. I have argued moral reasons might include an affective element. This affective element can be incorporated into moral theories in different ways. For instance Shaun Nichols believes we base our moral judgements on a set of moral norms and on an affective response to these norms, he calls this system of moral judgement a sentimental rules system (1). A different approach is taken by Jesse Prinz who argues that a moral judgement is right or wrong simply if the agent has dispositions to feel approbation or disapprobation towards it (2). It is important to note that in all moral theories based on our emotions the affective element plays an essential part in moral understanding, remove the affective element and the theory ceases to be a moral theory. Because our emotions play an essential part in our moral understanding according to these theories it follows these same emotions motivate us morally. It further follows my account of moral disability seems to be a perfectly adequate account. Unfortunately it also appears follow if someone is morally disabled using my definition then it is hard to hold her morally accountable for her actions even if she fully understands that society holds that these actions are morally wrong.

Let it be accepted that internalism concerning moral judgements is true. This means if someone makes a moral judgement that x is wrong she has a motivation not to do x. Let it be assumed someone makes a judgement that x is wrong and she has a non-affective motivation not to do x. The question I wish to explore is whether such a judgement could be regarded as a moral judgement. In order to answer this question we must ask why someone judges doing X is wrong. There would seem to be two possible reasons why she should be motivated not to act. Firstly she might judge doing x is wrong for him due to self interest. For instance he might believe stealing is wrong only because he believes she will be caught and punished. In this context her judgement would not be a moral judgement. I will not pursue this option. Secondly someone might judge doing x is wrong because society says doing x is immoral. In this second context she might also be motivated not to do x out of self interest. The question now becomes this, is she making a moral judgement in this second scenario? The answer to this question is important because if we believe she is making a moral judgement we can hold her morally accountable. If we accept Kohlberg’s account of moral development then she is indeed making a moral decision. Stage one of his account is based on an agent simply acting out of obedience or because she fears punishment. In stage two an agent is acting out of self interest. Accepting Kohlberg’s account of moral development means my account of moral disability would need re-examining. Someone whose moral decisions are affect free might be best described as someone whose moral development is stunted. Moral disability then might be a matter of degree much the same as physical or mental disability. It appears to follow that someone might be held morally accountable for some of his moral decisions and not for others. It further follows that not all moral decisions require an affective element. It would still further follow my account of moral disability is at best an inadequate account.

However accepting Kohlberg’s account of moral development creates a problem. This problem occurs because if decisions are made using the means available at stage 1 and 2 of his system there seems little, if indeed anything, to differentiate moral decisions from more conventional decisions. Consider a child who is taught not to hit his younger sister when she pulls his hair. His parents achieve this by withholding their approval and scolding him when he does so. He learns not to hit his sister because he wants his mum and dad’s approval, he acts out of self-interest. When he decides not to hit his sister if she pulls his hair his decision is a moral decision, albeit a primitive one, according to Kohlberg’s account of moral development. Now let us consider the same boy when he learns to use a knife and fork for eating as opposed to his hands. Once again his parents do so by withholding their approval and scolding him when he eats with his hands. Let it be assumed he decides to use a knife and fork for the same reasons he decides not to hit his sister. It follows the structure of his decision-making is identical in both cases. His decision not to hit his sister is a primitive moral decision according to Kohlberg’s account whilst his second decision appears to be one of simple etiquette. However it seems ridiculous to equate moral decisions with those of etiquette. Moreover children soon learn to distinguish between decisions prohibiting harm and more conventional decisions concerning etiquette at an early age, usually between the age of two and three. Further it seems highly improbable that a child learns to make this distinction due to an increase in her cognitive powers because sociopaths who have much greater cognitive capacities than her seem unable to make this distinction. This leads me to conclude that Kohlberg was mistaken in his belief that decisions motivated simply by obedience or self interest can be moral decisions. Moreover the fact that someone can make understand the difference between moral and more conventional decisions at an early age suggests that Kant was wrong to believe moral decisions must be rational decisions. It would seem that moral decisions must include an affective element. It follows my definition of moral disability might be an adequate definition.

Accepting that a moral decision must include an affective element has important implications. Firstly it would seem that some people are totally morally disabled if they feel no adverse emotions at all when harming others. Sociopaths would seem to be totally morally disabled. Secondly I have argued Kohlberg’s account of moral development at stages one and two does not concur with experimental evidence concerning children’s abilities to distinguish between decisions prohibiting harm and conventional decisions. But accepting my conclusion does not imply moral development doesn’t occur and that people cannot be partially morally disabled. Moral development might still occur in two ways. First we might develop new moral emotions. We might for instance come to have feelings of approbation towards those who act altruistically. Secondly we might learn how to understand the contexts, in which moral decisions are made as our cognitive powers increase, better. For instance we might come to see animals as capable of suffering and hence worthy of inclusion in the moral domain. It therefore seems that my initial account of moral disability is an incomplete account and that moral disability might be better described as follows.

  1. Someone is totally morally disabled if she does not have any capacity to feel sympathy when making moral decisions.
  2. Someone is partially morally disabled if she has only limited capacity to feel sympathy when making moral decisions or lacks the cognitive abilities to fully understand the contexts in which these decisions are made.

In the light of my revised definitions of moral disability I will now consider the two questions I posed at the start of this posting. Firstly if my revised definition is accepted is it appropriate to express anger when the totally or partially morally disabled commit evil acts? The discussion above suggests this question is inappropriate. I have argued if we judge an act as morally wrong that this judgement must include an affective response. It follows judging an act to be evil means we must of necessity feel some emotion. It seems natural to me that this necessary emotional response should be one of anger. It follows anger is an unavoidable response by us if we judge something as evil.

The second question I posed is this. If someone is morally disabled can I hold her responsible for her evil actions? I will answer this question in two parts. First let it be assumed that someone who is morally disabled acts in a mean or evil manner towards me or others and that her actions are not criminal actions. Perhaps for example they spread malicious gossip about me. Let it be further assumed this person is either totally or partially disabled as defined above, perhaps she is autistic. In such a case I believe I should make allowances for the person concerned and not apportion blame. I will of course be naturally angry as argued above. The second part of my answer to the above question is of much greater practical importance. Should we apportion blame to someone who is morally disabled and commits some evil crime? Our legal system is concerned with whether such a person is responsible for her actions. The cognitively impaired and young children are held to not responsible for their actions. The Courts may still take action even if they do not apportion blame. For example a sociopath may be sent to Broadmoor rather than punished. The criteria used in deciding if attaching blame is appropriate are based on determining whether the defendant has and is capable of using the requisite cognitive powers. Sometimes these cognitive powers may be overwhelmed and the legal system allows for this. For instance the case of B who refused to give informed consent to a caesarean section due to her fear of needles (3). In cases such as that of B the question the Court considered whether the defendant’s passions override her cognitive abilities. However in the light of the above I would suggest someone may have perfectly adequate cognitive powers and these are not overwhelmed yet still be morally responsible for her actions. She is morally disabled because she has limited affective capacities to guide her cognitive capacities. Someone who is cognitively disabled cannot be held responsible for her actions. I would suggest similarly someone who is affectively disabled should not be held fully responsible for her actions I would further suggest our legal system should not only be interested in whether a defendant has excessive passions overwhelming her cognitive capacities but also whether she has only limited affective capacities to guide her cognitive capacities. It might be argued that in practice this assessment is difficult, but is it any more difficult than assessing someone’s cognitive abilities? A defendant might well hide her cognitive abilities in order to be found not guilty due to diminished responsibility.


  1. Shaun Nichols, 2004, Sentimental Rules, Oxford University Press.
  2. Jesse Prinz, 2007, The Emotional Construction of Morals, Oxford University Press.
  3. Re B (Adult; Refusal of medical treatment) [2002] All ER 449

Engaging with Robots

  In an interesting paper Sven Nyholm considers some of the implications of controlling robots. I use the idea of control to ask a different...