Luke Roelofs is a philosopher of mind at NYU's Center for Mind Brain and Consciousness. They live in New York with their spouse Katlyn and their cat Sebastian, and they like insects, board games, and coffee.
This week at The Junkyard we’re hosting a symposium on Heidi Maibom’s recent book The Space Between: How Empathy Really Works (OUP 2022). See here for an introduction from Heidi. Commentaries and replies are appearing Tuesday through Friday.
* * *
Recentering Perspectives: Commentary from Luke Roelofs
The Space Between offers two much-needed things: a defence of empathy’s value and a reorientation of how we should analyse it. I won’t say too much about the defence: Maibom responds to critics of empathy both in philosophy and in the wider culture, making the case that empathy is a multi-faceted skill that often requires effort and care, but which provides a lot of valuable things when done well, some of which we can’t get elsewhere. In particular, empathy is not opposed to either rationality or impartiality, but is a key contributor to, even component of, both. On this score I’m in full agreement.
What I want to dwell on is the picture of empathy the book develops and deploys, which I think somewhat blurs or cross-cuts some common distinctions that philosophers use to analyse empathy. That’s not to say it’s idiosyncratic; the book’s picture is woven out of both everyday examples and empirical results, and left me suspecting that the picture of empathy given here matches pretty well how a lot of non-philosophers think of it.
Maibom’s account of empathy revolves around perspectives and perspective-taking, and the book is organised around explaining these terms. It starts with visual perspectives as a sort of familiar microcosm of a perspective, then moves to focus on the well-established empirical difference between what are termed agent-perspectives, observer-perspectives, victim and perpetrator perspectives, and interactor-perspectives. Crucially, perspectives here are holistic structures: they don’t consist in any specific mental state, but in what stands out, what is emphasised and what is neglected, how things are organised, etc. For example, the difference between a bird’s-eye view of a house and the view from in front of it isn’t well captured as a difference in the specific colours, shapes, and textures that are visible, but in how those are organised. Likewise, the different perspectives of an agent engrossed in action and of a detached observer of their action can’t be captured by any catalogue of the particular desires, thoughts, or feelings one has and the other lacks; it’s a large-scale difference in how their mental lives hang together.
This is important because analyses of empathy (especially those that link it to notions like simulation or imagination) often start with particular states, and the imaginative recreation thereof. We distinguish affective from cognitive empathy, or give separate analyses of propositional imagining and sensory imagining, or argue about whether there are imaginative analogues of desire or emotion (e.g. Currie and Ravenscroft 2002, Doggett and Egan 2007, Roelofs 2022). But shifting, for instance, from the perspective of an observer to that of a victim, involves changes that are both cognitive and affective, that involve desires and perceptions and all sorts of different states fitting together into a characteristic structure. An analysis that is focused on piecemeal simulation of individual states is liable to make this central phenomenon harder to see clearly.
That’s not to say that analysing specific simulated states and analysing holistic structures are sharply opposed. The central example of emotion illustrates this: Maibom centres empathising with emotions in part because emotions are in themselves holistic states, which bring with them a structure of cognition, attention, sensation, motivation, etc. Part of the epistemic value-added of empathy exploits this holism: we see a certain sort of smile or frown on someone’s face, which tells us to imagine feeling a certain sort of emotion, and finding that the simulated emotion discloses a certain sort of wider world, directing our attention to possibilities or analogies or causal connections that we wouldn’t otherwise have thought of. More broadly, it’s obvious that we can’t expect to recreate another’s perspective in its entirety (we have neither the imaginative capacity nor sufficient evidence), so we will have to make do with partial recreations. But there is still a crucial difference between simply recreating one isolated part of the other’s perspective, and attempting to partially recreate its overall structure. It’s the second, I take it, that Maibom wants to refer to as ‘empathy’, and to defend the special epistemic value of. Imaginatively recreating particular states of the other piecemeal, and slotting them into a perspective that remains organised in an egocentric way, is something that may have its own value but falls short of empathy.
A second notable feature of the book’s treatment of perspectives is that it de-emphasises their tie to individual identities. We talk a lot about ‘so-and-so’s perspective’, and about how ‘everyone has their own perspective’, and so it might seem natural to think of perspective-taking as fundamentally about taking ‘the perspective of’ some specific individual, so that we can always ask whose perspective is being taken (e.g. Goldie 2011, Coplan 2011, Roelofs 2021). But this is not the case. Although Maibom does treat perspective-taking as a key way of understanding particular individuals, what’s analytically central is not the sort of perspective that there are 8 billion+ of (mine, yours, Joe Biden’s, etc.), but the sort of perspective that there are around 3-5 of (agent, observer, interactor, etc.). That is, what really matters are a relatively small number of relatively invariant ways that our experience of a situation can be structured, and we can shift among these independently of their belonging to any particular person. This comes out especially clearly in the chapter on self-knowledge, where Maibom discusses Sartre’s example of the sudden shift in perspective brought about by ‘the look’ of an observer, even if the observer isn’t really there. When I’m unreflectively engrossed in some activity (like peeping through a keyhole) and hear a noise behind me, I experience a precipitous onrushing of self-consciousness, becoming aware of myself and my shameful activities from an observer-perspective. This shift from an agent-perspective to an observer-perspective isn’t undermined if it turns out that the noise was actually just the wind. Consequently, critiques of empathy that ask obsessively about who is being imagined - the other person, you in their situation, you-as-them, or someone else - are liable to miss what matters.
Let me end with one area where I thought more discussion could have helped. Critics of empathy’s role in morality often point to the apparent possibility of taking someone else’s perspective just in order to better predict, manipulate, or torment them. Maibom considers this worry, and responds that empathy can have pro-social motivational effects even if it doesn’t have those effects always and everywhere, and even if they are sometimes outweighed by other motivations. But I worry that more needs to be said about what exactly the relationship is between these phenomena. If I take another’s perspective and feel no warmth or altruistic inclination whatsoever, have I empathised with them or done something else? And either way, what makes the difference between this and the more common case where perspective-taking breeds rapport and a motivation to help (even if a faint or easily outweighed one)? To put it another way: to the extent that perspective-taking helps me better understand how reality in fact is, its impact on how I subsequently act seems to fit within a familiar schema (roughly, now I know more about what to do to satisfy my pre-existing goals). But The Space Between brings out a lot of ways that perspective-taking can be transformative beyond just teaching me new facts about others. It can, very obviously, change what I want, how I feel, what seems good or important. But it’s not clear what familiar schema this fits into (is it a brute change of preference? Is it learning new moral facts? Is it ‘infecting’ me with other people’s preferences?). Why does it sometimes happen and sometimes not? Is it justified, unjustified, or just a thing that happens? I didn’t come away from The Space Between with a clear sense of how to answer these questions, and to that extent I felt like I was still missing a piece of the story of ‘how empathy really works’.
References
Currie, G., and Ravenscroft, I. (2002). Recreative Minds: Imagination in Philosophy and Psychology. Oxford: Clarendon Press.
Coplan, A. (2011). “Understanding Empathy: Its Features and Effects.” In Goldie, P., and Coplan, A., Empathy: Philosophical and Psychological Perspectives, Oxford University Press: 3-18.
Doggett, T., and Egan, A. (2007). “Wanting Things You Don’t Want: The Case for an Imaginative Analogue of Desire.” Philosophers’ Imprint 7(9): 1-17.
Goldie, P. (2011). “Anti-empathy.” In Amy Coplan & Peter Goldie (eds.), Empathy: Philosophical and Psychological Perspectives. Oxford University Press: 302-317.
Maibom, H. (2022). The Space Between: How Empathy Really Works. New York: Oxford University Press.
Roelofs, L. (2021). “Imagine if They Did That to You! The Complexity of Empathy.” In Kind, A., and Badura, C., Epistemic Uses of Imagination, Routledge: 279-297.
Roelofs, L. (2022). “Longings in Limbo: A New Defence of I-Desires.” Erkenntnis. https://doi.org/10.1007/s10670-021-00505-7
Response from Heidi L. Maibom
Luke Roelofs raises a series of questions, not all of which I have the space to elaborate on here, regarding the connection between perspective taking and prosocial motivation. An increasingly common criticism of the kind of empathy I am concerned with in the book is that it is possible to empathize with others this way without being moved by what is simulated at all. In the book, I object to this line of reasoning by pointing out that to assume that perspective taking always has prosocial effects is unreasonable since agents have many concurrent motivations at any one time, many of which may trump any motivation coming from the empathic engagement itself. Roelofs finds this response unsatisfactory, or at least, not fully satisfactory.
Roelofs asks—very reasonably to my mind—whether someone who takes somebody’s point of view in a purely exploitative-instrumentalistic way has really empathized with that person at all. A first line of response is the following. Personally, I have always been suspicious of claims that torturers are excellent empathizers, which is why they are so good at what they are doing. In my framework, one does not have to take an agent perspective on someone to know what will hurt them. There may be things that hurt that one will not understand from an observer perspective, but one can figure out how to hurt others perfectly well from such a perspective. Psychopaths, who appear to do little perspective taking, are very good at manipulating others on the basis of knowing what hurts or scares them. I would therefore be inclined to deny that a torturer empathizes with his victim, for instance. Having said that, I don’t think we can rule out that conflicting motivation trumps empathically induced caring. This often happens when we make somewhat paternalistic decisions on behalf of children or pets. I shall refrain from pointing out the callousness of most of humanity regarding the plight of other people and nonhuman animals.
There are more questions, of course, some of which I look forward to exploring in future work. But I want to touch a little on what makes the difference between one instance of someone experiencing empathic affect and another doing the same. There is a tendency to focus on the emotion or the mental state of perspective taking, as if the wider cognitive economy was irrelevant. Of course, we all agree that it is. Nonetheless, many of the criticisms raised against empathy proceed in apparent ignorance of this fact. But we have evidence that it does matter who the person is who empathizes with someone (e.g. Sautter et al. 2014). We also know that emotion regulation plays an important role in empathy (Sprinrad & Eisenberg 2014, Maibom 2019). This suggests that we must look to the greater cognitive economy, or to the whole person (dare one suggest), for an answer to some of Roelofs’ questions and that demands a much more extensive inquiry than I have engaged in in The Space Between. It is, however, an inquiry well worth engaging in, as Roelofs rightly suspects, although it does require us to rethink how to investigate the mind. I have made some suggestions as how to do this elsewhere (Maibom 2022). We may need to talk about the empathic person in addition to the cognitive-emotional phenomenon.
Lastly, I do intend the argument in The Space Between to be that an agent perspective on others and an observer perspective on ourselves help us see morally relevant facts that we would otherwise not see, because they each bring out different ways of mattering. More should be said about this, of course, but I have run out of space.
Reference
Maibom, H.L. 2019. Empathy and Emotion Regulation. Philosophical Topics, 47:2, 149-163.
Maibom, H.L. 2022. Don’t Worry, Be Happy? Synthese, 200:67, https://doi.org/10.1007/s11229-022-03597-y
Sautter, J.A., Brown, T.A., Littvay, L., Sautter, A.C., and Bearnes, B. 2008. Attitude and divergence in business students: An examination of personality differences in business and non-business students. Electronic Journal of Business Ethics and Organization Studies, 13 (2). http://ejbo.jyu.fi/pdf/ejbo_vol13_no2_pages_70-78.pdf
Spinrad, T. & Eisenberg, N. 2014. Empathy and morality: A developmental psychological perspective. In H.L. Maibom (Ed.) Empathy and Morality. New York: Oxford University Press, 59-70.