For Anscombe, psychology provides a foothold against the Nietzschean slide. Moral theory begins within. Lying is wrong, not because it violates some moral law, as Kant had it, or because it tends not to maximise utility, but because it’s deceitful, and being deceitful doesn’t tend to fulfil a person, or in any way help her to flourish. To see what we should do, we must look at who we are.

Whether or not they’ve bought her particular moral theory, philosophers since have tended to accept that psychology is the foundation on which all such theories should be built – that before we can understand morality, we must understand people. A happy product of this has been to make their output broader and more tangible – less about abstract concepts and more about real-life facts. Sydney philosopher Craig Taylor’s Sympathy is, in this sense, typical of the new wave. Drawing on Dickens and Austen, Swift and Wordsworth, the Holocaust and the Russian Revolution, it seems not just more readable, but more relevant, as a result.

But Taylor wants to swim against the current: Anscombe’s dictum is impracticable, he declares, because it rests on a false divide. The idea that moral theory must build on psychological theory assumes that the latter can be constructed separately – to assume that ‘moral actions themselves will not feature in our most fundamental account of human nature. However that is just what I will deny.’ Before we can conceive, or even begin to conceive of, psychological theories (and, for that matter, moral theories), Taylor declares, we must perform certain moral acts. It is doing what we ought to do that lets us think about who we are, and about what else we ought to do. Taylor starts by pointing out that not everything people characteristically do seems attributable to some underlying motive: a suicide case may dodge a car, and a self-consciously tough guy can wince in pain. Some of the ways we respond to the world, and therefore ‘act’ in the world are automatic, Taylor concludes: human nature is comprised not just of what we characteristically think and feel, but also some of what we characteristically do.

Step two is epistemological. Automatic responses are not just a part of human nature, they are a part of human theories, particularly the theories we have about one another. The fact that other people think and feel in the first place, Taylor explains, is not itself a matter of theory, but a ‘pre-reflective given’, something we just know from the response we automatically have to them. As Wittgenstein put it: ‘I am not of the opinion that he has a soul.’ So the theories we go on to have about those thoughts and feelings – our psychological theories – in that sense begin with our acts.

And moral theory? Moral theory begins, Taylor states, with the idea that another person’s suffering may make a claim on us, and this idea begins with the automatic response we may have when we witness that suffering: that of sympathy. Sympathy tells us that we are ‘one among many’: it ‘establishes a connection according to which we recognise other people’s experiences, hopes, joys and sufferings and so on, as like our own’. A slave owner may be perfectly aware that his slaves are suffering, but without sympathy he will not see that they are suffering as human beings – as he would – and therefore not see their moral claim.

However, Anscombe would probably agree that moral theory is bound up with considering human suffering (though I don’t know where this leaves foetuses or calves), and that this logically seems to entail an appreciation that such suffering exists. And she may well accept that this appreciation isn’t wholly abstract and intellectual – that somewhere, at its beginning, lies something like the automatic response of sympathy. It’s not clear how Taylor opens up a false divide by simply calling this automatic response an act. But he’s not simply doing that. Sympathy is a moral act. ‘The claim that sympathy answers is not merely that one’s suffering be relieved but … that one’s suffering be acknowledged. What is important is not merely that you help me when I am suffering and in need, but that you know how it is with me.’

Let’s think about this. To recognise a moral theory, as members of a society do routinely, is to recognise a reason to do what that theory prescribes. A hungry Christian can’t say: ‘I believe that “Thou shalt not steal” but I see no reason not to steal that apple.’ Now imagine a situation where she leaves the grocery hungry on the basis of that reason alone: that she felt no sympathy whatsoever for the struggling grocer, but wanted to be moral and therefore did ‘the right thing’.

According to Taylor, though, she hasn’t: the moral worth of an action depends partly on why you do it. But is this really how we understand morality? As Claudius (or, at least, Robert Graves) had it: ‘There are scoundrels with stony hearts, virtuous men with equally stony hearts, virtuous men with golden hearts … and scoundrels with golden hearts’ — the humane aren’t always so human. Is Taylor sure that his false divide isn’t actually just between people who are moral, and people who are moral and nice?