The Germans had empathy before we did. At least, the term “empathy” was coined in 1909 as a translation of the German “Einfühlung”. Einfühlung was a technical term, in late nineteenth-century German aesthetics, which referred to the mind’s supposed capacity to project itself into objects so as to make it appropriate for viewers to experience them as animate or beautiful. “Empathy” was not the first attempt at translation; “Einfühlung” was originally translated as “infeeling” (and the subsequent debate might have been rather different had we stuck with that). The term we know and seem to have come to love was coined by the Cornell psychologist, Edward Tichener, who, in an inspired linguistic move, drew on the Greek “em” (in) and “pathos” (feeling).
Since that time, the term has had a curious history. There were some in the Anglophone world who were interested in aesthetic theory and who knew the German debates. These included the remarkable figure of Vernon Lee (a pseudonym of Violet Paget), who, at least until the recent interest in the history of empathy, was better known as the author of short stories of the supernatural. Lee and her collaborator Clementina Anstruther-Thomson already had an interest in “experimental aesthetics”, carefully measuring the physical effects on each other that were caused by regarding the visual arts. Lee and the foremost German proponent of Einfühlung, Theodor Lipps, argued with each other in print, with, in retrospect, Lee’s scepticism of German theory-building coming out on top. The debate raged, drawing in people such as T.E. Hulme and possibly, although even then only for a short while, R.G. Collingwood. Lee’s neighbour, the great Renaissance scholar Bernard Berenson, appears to have borrowed from the debate for his influential notion of “tactile values” – although he (absurdly) accused Lee of plagiarising his ideas and didn’t speak to her for twenty years.
What emerged from this welter of activity is that Einfühlung, empathy, had at its heart a mystery: projection. What sense could be attached to the notion that we project our minds into objects? From Oxford, an earlier proselytiser for aesthetics E.F. Carritt wrote in his The Theory of Beauty, “we have here nothing but an attempt to explain in figurative language an unconscious process by which some beautiful objects may have become so”. It looked as if, after a brief flourishing, the notion of empathy was dead; replaced by theories such as the expression theory and formalism. Indeed, that seems to have been the case; the term then pretty much disappears from sight for around fifty years.
It is not clear why “empathy” stared to creep back into common parlance. My hunch is that it stemmed from the translation, in 1954, of Wilhelm Worringer’s Abstraktion und Einfühlung. The translator looked back to when the debate was live (the book had first been published in German in 1907), and employed the English term from then; hence, Abstraction and Empathy. Even then, take up outside of Art Historical circles was fairly slow. Nonetheless, an explosion of interest was about to happen.
What underpinned that explosion was not aesthetics at all. Early on in the debates, before Titchener had even coined the term, Lipps had suggested the work that had been done in aesthetics could be applied to solve the problem of other minds. In a slightly handwaving passage in 1907 he claimed that we could somehow mirror the mind of another in our own minds, and grasp both the fact that they had minds and also the content of their minds. In the later twentieth century this, our capacity to “read the minds of others”, was in the intellectual air. Psychologists were researching such matters as the basis for altruism, the acquisition of a “theory of mind” (our capacity to represent the mental states of others), and the underlying causes of autism. However, it was the philosophical debate about interpreting what was in the minds of others that, I suspect, took “empathy” to the centre of the stage. Since the mid-1960s, the dominant theory of mind had been “functionalism”; the claim that mental states were functional states – something that was caused by inputs (sensory information), had relations to other mental states, and caused outputs (appearance and behaviour). This came along with an account of how we could come to learn about another’s mental states: we employ some interpretative psychological theory we all carry around with us. This theory consisted of theorems such as “If a person desires X, and believes acting in such-and-such a way will bring about X, then, all things being equal, they will act in such-and-such a way”. In the mid-eighties this was challenged by Jane Heal in the UK, and, independently, Robert Gordon in the States. Rather than using theory to interpret each other, Heal and Gordon proposed a simpler alternative. We imagine ourselves being in the other person’s circumstances, work out what we would think or feel, and attribute those thinkings and feelings to them. That is, we discover what is going on in the heads of others by replicating the content of their minds in ours. Although this quickly got known under the name “simulation theory”, it originally went under the name “empathy theory”.
There were at least two other branches of philosophy that quickly got in on the act. The philosophy of the emotions began to study empathy as something interesting in its own right (we have Peter Goldie to thank for that). Moral philosophy, borrowing from the interest in psychology, stated to wonder about the relation between empathy and moral behaviour. Perhaps most astonishing of all, the concept escaped into the real world where it appeared, and still appears, unstoppable. Barack Obama warned of the dangers of “the empathy deficit”, and a plethora of books have appeared that push the view that empathy is the core of the solution to all of our problems.
The great difficulty that those who write about empathy face is that there is no agreement about what “empathy” actually means, and hence what, if anything it refers to. In a way, this danger was obvious. “Empathy”, as I claimed above, was coined in 1909. Does this mean that we did not have empathy before 1909? That seems unlikely. Or that we had empathy, but we did not have a word for it? That seems equally unlikely. This raises the interesting question as to when it is sensible to invent a word. One scenario in which it is sensible arises when we have a clear referent in the world, but no term with which to refer to it. I assume that this explains the coining of terms such as “television” or “helicopter”. Another might be when there is some clear logical or mathematical property, such as that named by “pi” or “binomial”. There will be more complicated scenarios, but a necessary condition for clarity, or so it seems to me, must be that the new term is coined to fit a gap in some reasonably stable existing discourse.
Did the term “empathy” emerge from a reasonably stable existing discourse? We have seen that the nineteenth-century German aesthetics discourse was no such thing – “Einfühlung” was a very unclear concept, used in several different ways with respect to several different problems. As we have seen, that discourse was the origin of the term. However, there is at least one other discourse that employed a term that appears akin to “empathy”: the discussion, by David Hume and Adam Smith, as to the foundations of morality. The term they used was “sympathy”. However, attempting to find a stable referent for “empathy” using that discourse faces something of a dilemma. Either the discourse was stable and the terms meaningful, or it was not. If it was, then “sympathy” is a perfectly adequate term and trying to co-opt it into “empathy” will cause unnecessary confusion (Louise Braddock takes something like this line). If it was not (and, at least where the empathy-like properties of sympathy lie it seems that it was not) then we are no better off as we have still not found a stable foundation for our term.
The current situation, then, is confusing. For those interested in the philosophy of mind, empathy is primarily an epistemological tool for working out what others think (sometimes called “cognitive empathy”). For those interested in the emotions and in moral philosophy, the focus is more on feeling what other people feel (sometimes called “affective empathy”). To attempt to find a way through the thicket, other distinctions are standardly drawn – although, to add to the confusion, not always drawn in the same place. There is “high level empathy” and “low level empathy” (sometimes called “basic empathy”). High level empathy describes the process if it is answerable to the will, or accessible to consciousness. Low level empathy is an automatic process, perhaps employing the mechanisms of mirroring expressions or behaviour.
From this we could perhaps derive some common core. Empathy at least involves this: imagining oneself (whether consciously or unconsciously) into another’s circumstances and replicating their mental states. For the epistemologists the relevant mental states are cognitive attitudes, and the result is grasping what other people are thinking. For the philosophers of emotion, the moral philosophers, and quite possibly the folk, the relevant mental states are feelings and the result is that one feels what the other person is feeling. We might want to add various accretions – for example, the folk usually take empathy to be a good thing and thus postulate a link between empathy and morality, or at least a desire to help. The more philosophically pure maintain it is morally neutral – knowing what is going on in someone’s head may inspire us to help them but we may also find it useful if we want to harm them.
In sum, empathy is a hodge-podge of ideas, not all of which seem to fit happily together. Thus, although I have claimed to have identified a “common core”, there is no guarantee that those are the ideas that will remain central. Indeed, since the current resurgence of interest in empathy it has become apparent that both elements of the common core are problematic. Consider the first: “imagining oneself into another’s circumstances”. Why should my imagining me in your circumstances enlighten me about what you believe or what you feel? That would only tell me what I would believe or feel in those circumstances – as you have different background beliefs, desires, and dispositions what you feel might be completely different. I could try to imagine being you in your circumstances, but it is not clear that that is even coherent. What is it to imagine of oneself that one is someone else? The problems with the second element include specifying the object involved in attempts to replicate someone else’s emotion. Consider someone (Albert) attempting to empathise with someone else (Bella) who is mourning a dead pet. Albert replicates Bella’s emotion. Does that mean the object of Albert’s emotion is the dead pet? No, as that leaves Bella out of the picture – after all, he is supposed to be empathising with Bella. So is the object of the emotion Bella? No, as that would not be a replication of Bella’s mental state and thus not a candidate for empathy at all.
Recently Anik Waldow and I have edited a collection on empathy. The papers have not exhibited any decisive turn in the debate; certainly not a convergence on an agreed definition. However, there is a drift – although not in all the papers – away from the common core in favour of accentuating two other elements of the hodge-podge of ideas. The first of these withdraws from the idea of imagining oneself into another’s circumstances in favour of an openness to the circumstances of another. To empathise with another is to be sensitive to their situation in a way that (in the words of Adam Leite and Katy Abramson) one becomes “attuned” to their needs. This is a move away from empathy as an epistemological tool, and closer to the link between empathy and care that is characteristic of empathy as conceived of by the folk. However, once the element of imagining being in another’s circumstances is abandoned, the second part of the core conception (replicating the mental states of the other) is left high and dry. Nobody thinks that mere replication is sufficient for empathy; the empathiser and empathisee could have the same mental state because of a common cause, or even by sheer happenstance. Hence, some different relation between the two mental states needs to be worked out. The obvious way forward is to stick with the claim that the empathiser feels what they feel because the empathisee feels what they feel. However, specifying the causal relation in a way that is immune to obvious counterexample has proven very difficult. Some years ago, in his “Empathy, Imagination, and Phenomenal Concepts”, Kendall Walton suggested a different approach – that the empathiser uses their mental state to refer to that of the empathisee. That is, Albert claims that Bella feels like this – where “this” refers to Bella’s mental state. This seems an interesting and enlightening idea that has links with other areas of the philosophy of mind.
Obviously, mentioning these changes only touches on the new work on empathy – I could go on to consider interesting work being done alongside theories of embodied cognition or novel ways of defending the status quo. My point, however, is only to raise the issue that the ground appears to be shifting in empathy studies. That, it seems to me, should provoke some reflection linked to the origins of the term. Given the term is so radically untethered, it is not obvious what these shifts signify. We are not moving towards a more accurate account of empathy – that presupposes that there is such a thing and that all we need to do is get a clearer view of it which is surely mistaken. A more accurate way to think about it is that we are doing and undoing the pieces of a jigsaw puzzle in the hope of getting a clearer picture. Given the history of the concept, however, it might be optimistic to think that any very clear picture will ever emerge.