2
Some Conceptual Preliminaries

In the last chapter I gave a rough characterization of empathy as using our imaginations as a tool so as to adopt a different perspective in order to grasp how things appear (or feel) from there. One manifestation of this, which plays a prominent role in current debates, is simulation:1 we use our own minds as a model of other minds. I shall introduce the idea by borrowing an analogy from Jane Heal:

We can get at the key idea by considering the familiar example of the model aircraft in the wind tunnel. (The case has some misleading features, which we shall need to remark later, but it will serve to get us started.) Suppose that we know, in general terms, that aerofoils provide lift, that aircraft are liable to become unstable in some circumstances and the like, but lack any detailed quantitative theory of aerodynamics. We do not have a set of usable equations relating all the significant variables, such as body shape or wind speed, to the upshots, such as lift and stability, in which we are interested. How can it be that we may nevertheless arrive at detailed quantitative predictions on these matters? Here is a possible method. If we are convinced (for example, by inductive generalisation, or as a consequence of theoretical assumptions) that a model aircraft will behave similarly to a real aircraft of the same shape, at least in a usefully wide range of circumstances, then we may test models with varying shape in varying wind speeds and so on, measuring the quantitative outcome in various respects and using those figures as a basis for the needed detailed predictions of the actual aircraft. We use the model aircraft to simulate the real aircraft.

(Heal 1996a: 64-5)

2

As Heal points out, we could learn something only if the 'model aircraft will behave similarly to a real aircraft of the same shape'. Analogously, simulation will only work if the model (our mind) behaves similarly to that of which it is a model (I shall refer to the person to whom our attention is directed as 'the target'). If we assume our mental machinery and that of the target are roughly the same then all we need to do is find some way of replicating their inputs into our machinery, and this should yield an output from our machinery which is the same as the output from their machinery. Whether or not differences between us render empathy epistemically unreliable is, as we shall see, a matter for considerable dispute. There is also an issue of what can be simulated. Wind tunnels can test for certain properties of aircraft, but they will not answer all of an engineer's questions. Analogously, the scope of empathy - that is, which kinds of mental states can be simulated - is also open to dispute.

In addition to worries about the reliability and scope of simulation, two further issues arise. First, there are differences in the means by which we marshal the resources needed to take on the perspective of the target. The simplest case might be that in which we are in direct perceptual contact with the target. For example, you might be watching someone out of the window about to get into his car. As he does so, without realizing it, he drops his keys. You could simulate being in his situation; you 'input' that you put your hand in your pocket and find nothing there. Your simulation yields the thought that you would check all your other pockets, before looking on the ground around you. Having noted these 'outputs' in your own case, you attribute them to the target. This is, in fact, what the target does; your simulation has been successful. In cases in which you are in direct perceptual contact with the target, you do not need to exercise your imagination; all you need to do is to think what they are thinking. As we shall see, some count this as an exercise of the imagination (after all, the input is not the belief that you have lost your keys). In addition to direct perceptual contact, there are various other means by which you could marshal the resources you need. Someone could describe a set of events from the perspective of the target. You would not need to imagine (in the sense of 'make up') the circumstances; you are being told them. However, it might require an exercise of the imagination to adopt the target's point of view; to come up with the right inputs from what you have been told. Finally, we might simply imagine a set of circumstances from the perspective of someone in those circumstances; it might be some actual person, some hypothetical person, or some type of person (such as a Roman soldier who trudged along this road (Goldie 2000: 204)). This might require some effort; you might need to imaginatively engage your senses and engage in visualizing or imagining sounds. Indeed, as we shall see, the effort required arguably renders the process epistemically unreliable.

The second issue concerns the nature of the inputs. As it is a simulation, the inputs are not true of the simulator. That is, when I simulate the person who has dropped his keys, the input is not 'He has dropped his keys' but 'I have dropped my keys'. What is important is that these simulacra of beliefs ('make-beliefs', as they have come to be called) turn the simulator's cognitive machinery in the same way that beliefs turn the target's cognitive machinery. The output (check the other pockets, look on the ground) stop short of action by the simulator; you do not check your pockets, or look on the ground around where you are sitting. Rather, the output is some kind of representation of the motivation to so act; which motivation you then attribute to the target. Empathy, construed as simulation, needs it to be the case that make-beliefs have a similar enough effect on the simulator's mental machinery as do beliefs on the target's mental machinery. If not, the output from the simulator's mental machinery will not be a reliable guide as to the target's mental state. As we shall see, this too is open to dispute. We should note, as it has caused some confusion, that generally we regard the make-belief states as tracking the truth about the world, even if they are not true of us but true of someone else. We make-believe that there is nothing in our pockets because we believe that the target believes there is nothing in their pockets. That is, the distinction here between make-beliefs and beliefs is not the distinction between stuff we make up and stuff we do not make up, or fiction and non-fiction.

The characterization of empathy as simulation, which has the broad purpose of finding out what is going on in someone else's head, whatever that might be, is more characteristic of academic philosophy than it is of discussions outside that discipline. Outside academic philosophy there is more of a focus on affective or emotional states (which I shall call 'narrow empathy'). That is, to empathize with someone is not to imagine the world from their perspective so as to discover what they are thinking or what they will do next, but to imagine the world from their perspective so as to feel what they feel. For those for whom 'empathy' means 'narrow empathy', there is an essential link with the emotions. Those who do think thus might wonder why three entire chapters of this book are taken up with something that seems to lack this essential link. My defence of this is twofold. First, we should not take it for granted that academic philosophy is the outlier. In the most in-depth monograph on empathy to date, Karsten Stueber claims that it is the philosophers who have grasped the interesting core of the concept (a claim supported by the brief history of the concept given in the last chapter):

I would object to the claim that empathy as a vicarious sharing of an emotional state should be understood as the only right way of defining and explicating the concept of empathy, as it is sometimes asserted in this context. Empathy as understood within the original philosophical context is best seen as a form of inner or mental imitation for the purpose of gaining knowledge of other minds.

(Stueber 2006: 28; emphasis in original)

The second consideration is that empathy as a broad epistemic concept, and empathy as a narrow affective concept, share a common core: they both centrally involve taking on the perspective of the other. Although the philosophical account has come to be known as 'simulation theory', a name that seems to take it away from our concerns here, it is altogether possible that it might instead have been known as 'the empathy view' (see Davies and Stone 1995: 1; and Goldman 2006: 17). Furthermore, there are several wrinkles in sorting out a defensible version of the broad notion of empathy, which carry over to sorting out a defensible version of narrow empathy.

Narrow empathy comes in two forms. First, the link with the emotions might simply be that we imagine being in someone's circumstances so as to find out what they are feeling. We can consider the kind of example that Barack Obama had in mind in his talk of 'the empathy deficit'. We make the effort to take on the perspective of someone who has been made redundant. We simulate all kinds of input: worry about the effect on his or her family, worry about paying the mortgage, belief that people with his or her set of skills face limited opportunities and so on. We find ourselves experiencing something like a feeling of panic....