RD. Alexander (1987) has proposed that the ultimate grounds of ethics are evolutionary. Humans have the capacity for moral reasoning and negotiation, together with a system of appropriate emotions, such as shame and guilt when we do wrong, and a sense ofmoral wellbeing when we do right, because these abilities tended to increase our ancestors' inclusive fitness in the past. Fairly simple mathematical models can illustrate how this could work-one example is the Prisoner's Dilemma problem in which two individuals must act cooperatively if they are to maximise their 'survival points' over a series of interactions (Axelrod, 1984). In the Prisoner's Dilemma, mutual cooperation leads to an intermediate score for both parties; mutual defection gives both a low score, but if one cooperates and the other defects, then the cooperator scores no points, while the defector scores maximum points. Over a series of games the program that achieved the best scores was one called TIT FOR TAT, which cooperated unless its partner defected, but after a defection subsequently did whatever the partner did on the previous round, thus 'punishing' defection and 'rewarding' cooperation. It might appear tllat an evolutionary tlleory of the origin of ethics can never be compatible with the inclusion of nonhumans in our moral systems, because any tendency to give weight to their interests would always be disadvantageous; would tend to reduce inclusive fitness, and hence would involve a basic conflict with the ultimate grounds of human ethics. I hope to show that this strong view of the impossibility of animal rights is untrue, although there are grounds for holding a weaker view that proposes that, when interests conflict, humans will always have some tendency to favour relatives over non-relatives, friends over strangers, and other humans over nonhumans. Alexander himself believes (pp. 156ff.) that tlle selective value of indirect reciprocity may be enough to account for a human tendency to give sentient animals some status in their moral systems. He suggests that a person who (for example) was observed by other humans to be kind and reliable when handling pet animals would appear to them to be more trustworthy than someone who behaved cruelly. Trustworthy people would tend to be given preferential treatment, and, if it is true that behaviour towards animals tends to be a good predictor of future behaviour towards other humans, then tllere would be selection both for kindness towards animals and for a tendency to approve of such kindness in other people. The approvers benefit by being able to make successful predictions about the behaviour of other people because they are enabled to 'invest' help in partners who are likely to reciprocate.
Read full abstract