Starting, as I usually do these days, with wictionary, “empathy” is defined as “the intellectual identification of the thoughts, feelings, or state of another person”, and alternatively as “[the ] capacity to understand another person's point of view or the result of such understanding”. Wictionary also gives a third definition, namely “paranormal ability to psychically read another person's emotions”, but this is not how I shall be using the word “empathy” in this article.



Moving now on to Wikipedia, it appears that “observing another person's emotional state activates parts of the neuronal network involved in processing that same state in oneself, whether it is disgust, touch, or pain”. In plain English: the correct perception of someone else’s feelings tends to involve experiencing those same feelings oneself.

Our capacity for empathy is limited, of course. As David Pearce has put it (see comment thread on Giulio Prisco’s recent article on The Promises and Perils of Mind Uploading), “Evolution hasn't endowed us all with the hyper-empathising condition of mirror-touch synaesthesia. This is because it's genetically adaptive for us to be quasi-psychopaths.” And of course some people really are psychopaths, in the sense that they have an abnormally low capacity for empathy, and/or have (for one reason or another) a reduced willingness to exercise that capacity.

One of the most vivid illustrations of our limited capacity for empathy is the now-famous Stanford prison experiment, in which twenty-four male students were selected to take on randomly assigned roles of prisoners and guards in a mock prison situated in the basement of the Stanford psychology building. As noted in the relevant Wikipedia entry, the participants adapted to their roles well beyond expectations of the psychology professor conducting the experiment, Philip Zimbardo, to the extent that the guards enforced authoritarian measures and ultimately subjected some of the prisoners to psychological torture. The entry also notes that many of the prisoners passively accepted psychological abuse and, at the request of the guards, readily harassed other prisoners who attempted to prevent it. Even Zimbardo, adapting to his role as the superintendent, permitted the abuse to continue for several days until his future wife drew his attention to what was happening, and was sufficiently traumatized by the experience that it took him more than 30 years to write his book (The Lucifer Effect) on the subject. Zimbardo states in that book that one if his motivations for writing it was the relevance of the experiment in explaining what took place at Abu Ghraib. Zimbardo had set up the experiment in order to study the effect of a person’s environment on their personality and behaviour, the experiment provided a spectacular confirmation of the effect that such contextual factors have.



The ethical question to be addressed, here, however, is why any of this matters, and as I have repeatedly argued here, such ethical questions are always, in the end, matters of choice rather than of truth. The question is what kind of choice we want to make, and what this means for our understanding and practice of empathy.
In his recent article “After Happiness, Cyborg Virtue”, James Hughes points out that, having espoused a version of utilitarianism that seems to closely mirror my own, the gathering pace of research in hedonic psychology and emerging neurotechnologies make utilitarianism an “unattractive moral logic”, while a version of the “capabilities approach” of Amartya Sen and Martha Nussbaum better supports the technoprogressive endeavor. While I personally do not see the technoprogressive endeavour as being an end in itself, I have remarked in response to this article that it provides one of the strongest arguments so far that there is an ethical system that I might actually find more attractive than utilitarianism. In the capabilities approach, the goal is not happiness but rather accomplishment, an idea that seems to be closely related to the concept of “flourishing” that is now popular in positive psychology. What these two approaches have in common, in any case, is that they are both determinedly consequentialist in their outlook: that is to say, they both take the position that our ethical choices in the here and now should be determined on the basis of their likely consequences in the future. This is also the position on which the following analysis is based.

Nature has selected us to have a capacity, albeit limited, for empathy because such a limited capacity helps us to steer our way through our social environment in the way that best serves our interests (or more precisely that of our genes to the extent our environment still bears some similarity to the one in which our stone age ancestors evolved). Jonathan Haidt has called this ability “Machiavellian reciprocity” to illustrate the fact that, on the one hand, we need to be willing to serve the needs of others in order to achieve social status and get resources, while on the other hand that willingness needs to be selective and based on a positively biased perception of our self-worth. To serve the needs of others requires us to understand what they are, and empathy clearly plays a role in this. By extension, a capacity to be selective in our empathy allows us to be selective in our willingness to do so, and also to conveniently ignore evidence that might suggest that we are not quite such wonderful human beings as we like to think we are.

What is good for the individual (and his or her genes), however, is not necessarily the same as what is good for society at large, and it is fairly obvious that increased levels of empathy are likely to be helpful as a means of adapting individual behaviour in a way that improves the prospects for society as a whole (whether we are talking about happiness, capabilities, flourishing, or something similar).

Empathy still needs to be selective, however. For example, excessive empathy with someone suffering an extreme form of sexual jealousy could lead us to perpetrate or assist an act (such as murder) that would certainly be considered unethical. More generally, an extreme form of “mirror-touch synaesthesia” would effectively drive us insane, particularly in this modern world where we are constantly inundated with empathy-inducing messages encouraging us to buy this product, join this cause, spare some change, work overtime to assuage a panicking boss, come home early to spend time with the kids, etc, etc.

In summary: empathy is the ability to correctly perceive and to some extent mirror the feelings of others. Nature has selected a limited and selective capacity for empathy in individuals, and this limited and selective capacity is essential to allow us to steer ourselves effectively through life. At the societal level an overall increase in empathy seems likely to be beneficial, but still it is important to understand that too much or misplaced empathy can be a bad thing. Finally, given the extent (as shown in the Stanford prison experiment) to which our personality and behaviour are context-depend it is essential for mindfulness techniques to be taught at the earliest possible age as a means to promote consistent, values-based behaviour. Ultimately, this is probably more important than empathy per se.