There's an interesting discussion of empathy in a column by Paul Bloom at The New Yorker. Bloom begins by summarizing the generally good press that empathy gets these days, because it is seen as humanizing our responses to the sufferings of others. But the enthusiasm for it, he then says, is misplaced:
Empathy has some unfortunate features - it is parochial, narrow-minded, and innumerate. We're often at our best when we're smart enough not to rely on it.
What worries Bloom about empathy can be summed up in the phenomenon of 'the identifiable victim effect' - the fact that focusing on a single individual can move us to feel and act more effectively than some large statistic; the tale of one person's plight can take up more media time than instances involving huge numbers of deaths.
I think Bloom's point must be taken. But what, really, is this point? That empathy isn't self-sufficient? It certainly isn't. To operate effectively it needs the assistance of reason and the weighing of empirical evidence. I agree with him too - as I've argued on my own account in a recent paper - that 'it is impossible to empathize with seven billion strangers'. At the same time, how damning is it of empathy to point out that it is not all we require? Reason is also not all we require, but it is no less precious for all that. Empathy, in conjunction with other human faculties, is an invaluable way towards solidarity with others and humane action.
One other thing: the opposition between the 'identifiable victim' and the big statistic doesn't have to be rigidly maintained. How often have we all read a news story concerning genocide, famine or some other case of mass suffering, which picked out a particular detail, an individual's fate, by way of trying to bring home that the mass phenomenon was composed of myriad personal cases? It's the same point: advocacy for remedying the wrongs of the world has, and needs, more than one resource.