A Three-Person Model of Empathy

Empathy and Self-Loss (Step 1)

The first step of the proposed architecture of empathy consists in acknowledging the impressive degree to which human beings are capable of what one could call the “multiple facets of empathy” (Leiberg & Anders, 2006), so that one might go as far as calling man the empathic animal (Tomasello, 1999, comes close to this). The facets of empathy include involuntary copying mechanisms; emotional mimicry; feeling distress when observing the suffering of others; mind reading; deciphering others’ thoughts, states, or emotions; and “imagining how it feels to experience something” (Halpern, 2001, p. 85), as well as caring for others. In previous literature, these forms have sometimes been distinguished as either “cold” or “hot” forms of empathy, that are either distanced-analytical or simulative-emotional. Additionally, Batson (2009) provides an overview of eight uses of the word “empathy,” and Thompson (2001) and Singer and Lamm (2009) provide an overview of concepts related to empathy.

It is relevant to some of the views I develop here that several of the forms of empathy discussed by these authors relate to both human–human interaction and to engagement with fiction. In fact, there would probably be no fiction if we did not have the ability to imagine how it feels to be another or to be in another’s situation. Furthermore, what needs to be stressed is that empathy is by no means only a matter of well wishing and the positive acceptance of the other. Some facets of empathy allow competitors to better understand and hence undermine each other. Schadenfreude is not a marginal phenomenon of empathy. This distinguishes empathy from sympathy, which does not require a simulation of the other or insight into his state of mind, but does require a positive attitude of being “for” the other (Chismar, 1988; Wispé, 1986). However, as will later be suggested, certain forms of empathy involve aspects of sympathy or care.

A key aspect in several forms of empathy is the focus on temporal processes, actions, or changes of state. This seems true for mirror neuron activity in action observation (Rizzolatti & Sinigaglia, 2008). Pain perception is also more likely to trigger empathy when the pain is acute and not chronic (Saarela et al., 2007). Stagnation, inaction, and emotions without development or trajectory seem less likely to induce empathy. This aspect of empathy is also important for narrative empathy which involves specific temporal developments that first draw recipients in, but also “promise” an end to the empathic engagement when the development has come to an end (Breithaupt, 2009; Keen, 2007).

Research in the past decades has brought to light a slew of potential mechanisms for how we slip into the shoes of others, ranging from the so-called mirror neurons (for an overview see Rizzolatti & Sinigaglia, 2008), the basic mimicry abilities of infants, the discussions about “theory of mind” or cognitive mentalizing (Singer 2009; Thompson, 2001), and the evolutionary biologists’ meditations on the social intelligence of people (Tomasello, 1999). These discoveries give us not only fascinating insights into the mechanisms of empathy, but also suggest that people cannot help but empathize with others. It also seems that empathic attention can jump or even oscillate quickly between different people. This at least seems to be a lesson that filmmakers utilize when they employ sequences of shots from different perspectives and thereby guide viewers “through” different points of view (Coplan, 2006; Curtis, 2007). Another aspect of the omnipresence of empathy is the intensity with which people seem to share feelings. It sometimes may seem as if the inner perspective of what one imagines others are feeling, or what they are thinking, is part of one’s immediate perception. We can assume that we are biologically prepared for many forms of empathy, and our social environment seems well suited to develop these capabilities. Social beings like humans live in a world full of empathic noise, so that they almost constantly and involuntarily adopt others’ perspectives.

Being hyper-empathic sounds like an excellent adaption for a hyper-social being. Still, we should ask about the costs of this ability. The cost of empathy is not only the impressive brain power of human beings that require a large amount of energy and an extended period of care during childhood (Dunbar, 1992). The cost of empathy may also include danger of self-loss (Breithaupt, 2009; Preston & de Waal, 2002). Self-loss can be described as a possible effect of simulating, adapting, or otherwise engaging with the perceived perspective, state, or identity of another and thereby losing, ignoring, or forgetting one’s own perspective, interests, or state. Mild forms of self-loss are a common social and aesthetic experience. A reader of fiction may “identify” with characters in specific situations; an onlooker can “feel the pain” of someone else, even though the pain is not his; or a discussant may adapt to the viewpoint of the other and thereby lose sight of his own interests. Perhaps this is a factor in why face-to-face negotiations often yield surprising solutions. However, it is possible to imagine failures of empathy, in which individuals lose their identity for extended periods of time and may even act against their individual interests. Political and religious cults seem to have this force, as well as violent extremes such as hostage situations and intense training settings such as graduate school or boot camp (Carver, 2007) that may result in Stockholm syndrome or Survival Identification syndrome (see Fuselier, 1999; Regner, 2000).

Blocking Empathy (Step 2)

Here we get to the second step of the proposed model. If human beings are hyper-empathic (Step 1), we have to ask how they focus their empathic attention. Considering what has been said about self-loss, we need to ask what protects humans from excess of empathy. Put simply: How is empathy diverted, channeled, pulled away, filtered, in a word: blocked? These questions aim at whether and how individual control plays a role in the process of empathy. More pointedly, one could say: empathy, the understanding or simulating of others, only happens because our emotional attention towards others is mostly jammed, blocked, and filtered. Without such a (partial) blockage, we would live in a world of constant loss of perspective, in which we would involuntarily adopt the perspectives of not only all the people with whom we get in contact, but also those of animals, mythical creatures, and perhaps even objects.

A range of control mechanisms and learned forms of callousness (Shirtcliff et al., 2009) that extend to the multiple facets of empathy are to be expected. On the level of neuro-circuit activity, we can assume that there are inhibitory tendencies. For example, Marco Iacoboni proposed that a separate group of “super-mirror-neurons” might have the task of controlling other mirror-neurons (Iacoboni, 2008). Even without such “super-mirror-neurons,” the relation of neuron activity and full empathy is not well understood. Whereas it seems that certain neuro-circuits operate quasiautomatically, prereflexively, and prerationally, including the mirror neurons (Fogassi et al., 2005; Gallese, 2001, 2008), this does not mean that these activities trigger awareness or clearly measurable emotional response. (Mirror neurons are used here only as an example. Since their discovery, mirror neurons have received a great deal of attention; however, this does not mean that there are not many more neuro-circuits to consider. Many forms of empathy may not require mirror neurons.)

At the other end of the spectrum of control mechanisms are the more conscious efforts to selectively understand or distance oneself emotionally from others. Blocking empathy can be a trained form of behavior. A surgeon should probably not relate too much with his patients. One may reason that most health professionals (Halpern, 2001), as well as counselors and legal professionals (Madeira, 2006), develop skills to limit and allow empathy selectively.

We can expect a range of different control mechanisms in between inhibitory mechanisms operating at the level of individual neurons to the large-scale processes involved in conscious control of empathy. Still, how these other techniques work and how they can best be classified is an open question. What constitutes such blockage mechanisms of empathy and by whom or what are they steered? Which roles do higher level systems (such as “consciousness”) play (Libet, 2004; Shriver & Allen, 2005)? Considering the range of different cultural attitudes about empathy, we need to ask: Which roles do cultural techniques and learning play? And under what circumstances is empathy nevertheless allowed? How are the emotions that are recognized in others and simulated interpreted, filtered, and focused? For example, it seems reasonable to speculate that someone who observes the mishap of another may have less empathy if he also attributes the “fault” of the mishap to the other (excluding children, handicapped people, etc.). It has been confirmed in a related case that the brain reacts with less empathy when the observed person in pain has previously been perceived as being unfair (Singer et al., 2006). Hence, it seems at least plausible that attributions like “it was his own fault” also play into the intensity of empathy.

Other considerations in regard to the blocking mechanisms are the excesses, abuses, and dark sides of empathy. Some forms of sadistic behavior do not stem from an absence of empathy but from the morally wrong use or excess of it when abusers want the pain, humiliation, or degradation of others (for the case of torture, see Sussman, 2005). In these cases, either some empathic control or suppression of joy in another’s pain seems to be a morally desirable ability.

This article will not focus on the blocking mechanisms—for further suggestions see Tan (2009).

Unblocking the Blocking of Empathy (Step 3)

The third step of the model consists of those techniques that circumvent the blocking mechanisms and allow us, nevertheless, to experience empathy.

These learned control mechanisms of empathy (Step 2) can help to explain individual differences that are more difficult to explain from a general disposition toward empathy alone. Hence, the advantage of the three-step model of empathy is to link empathy in its mechanism with secondary mental processes that can occur without empathy. As long as we simply assume a one-step model of empathy, we can only account for individual differences by saying that some individuals have stronger core mechanisms than others (such as mirror neurons, better imagination, more theory-of-mind ability, better copying mechanisms, etc.). However, the three-step model could account for a wide range of variation within the same person who may have learned to use their empathy selectively, blocking it in some cases and not in others. Instead of simply focusing on empathic abilities or the lack thereof, as in autism, the focus on blocking and unblocking mechanisms allows us to examine the secondary mechanisms that are not part of the empathy facets per se, but that serve as triggers or blockers. Hence, a three-step model of empathy can give a better account of the complexity and individual or cultural differences of empathy. As a result, the model predicts that human empathy is the strongest (be it in terms of intensity or duration) when it is paired with secondary mental activities that help to bypass the blocking mechanisms of empathy. As mentioned, these secondary mechanisms may include, but are not limited to, narrative thinking (Breithaupt, 2011; Keen, 2007; Voss, 2004), fairness perception (Singer et al., 2006), causal fault attribution, past experiences, and temporal development (Saarela et al., 2007). This can also include conscious decisions to empathize.

In the following, this article will discuss one of these secondary mental activities that can trigger empathy and thus bypass or unblock the empathy inhibition, namely a process of side-taking in situations of conflict.

Three-Person Settings of Empathy

Most theories of empathy take a scene involving two people as their starting point: an observer and an observed, the former of whom registers the behavior and emotions of the latter “as though he himself were the agent.” Most theories thus portray empathy as a particularly close observation in a two-person scene that also takes into account the intentions and emotions of the other person. Depending on the respective theories, the observed experience is either copied in one’s own neural or conscious experience, and mentally performed in sync with the action and emotion (as in mechanisms of simulation, mimicking, etc.), or pieced together into a functional, calculable whole of the other person (as in mind reading and theory of mind). However, it seems that this focus on observation within a two-person model may be too narrow for the most social animal.

Certainly, two-person scenes are, evolutionarily speaking, ubiquitous, and the infant–mother relationship is decisive for all mammals (for a model of empathy derived from the mother–child bond, see Blaffer Hrdy, 2009). However, what sets apart primates and other social mammals living in groups is their increased possibility for complex social scenes with three or more individuals with different roles, and it is precisely these mammals (primates, dogs, dolphins, certain birds, etc.) that seem to be most capable of empathy. It is possible, then, that the origin of a more fully developed empathy lies in social scenes, or perhaps vice versa—social structures are enabled by empathy.

In what follows I would like to propose a mechanism that enables empathy as a result of taking sides in a three-person scene. The proposal will be that side-taking is one way to trump empathy’s blocking mechanisms. The connection between choice, side-taking, and decisions (to empathize), and empathy is not well understood or adequately researched. The following provides a proposal of how a specific form of empathy might be triggered by side-taking decisions.

It has often been noted that, in the realm of animals, human beings are the side takers or judgment makers. Side-taking and coalition-building do occur in other animals, notably in chimpanzees (de Waal 1998; Melis, Warneken, & Hare, 2010), but apparently much less so than in humans. Side-taking presupposes that a choice can actually be made and is not predetermined (or does not feel predetermined to the decision maker). And it presumes that a choice between different positions is registered by the decision maker. This means that there must be some form of discord between two parties. “Discord” here simply means that there is some tension between the observed individuals or that they have conflicting interests. This can occur in direct conflicts like fights or arguments, or in ritualized conflicts like sport competitions or judicial processes. This can also be the case in indirect conflicts such as disagreements of opinion about how a group should behave, or in popularity contests, gossip, political advertisements, or erotic behavior. In general, all groups and constellations of persons in which the individuals play different roles can be considered. In the family, for instance, the individuals inhabit different relationship niches as father; mother; first, second, or third child, etc.; and these unavoidably produce conflicts of interest.

Furthermore, side-taking assumes that the fight or conflict is registered as such, that is, that the different tendencies of the combatants are recognized by the observer. The word “tendency” is used here instead of “intention,” since “intention” is often related to internal states and circumstances of belief that are not perceptible from an external perspective, while a “tendency” is built into the action itself. This emphasis on the external description, that is, tendency, builds on Anscombe’s insight that human action can always be described in different ways and descriptions are based on external recognition of an action (Anscombe, 1957). In order to take sides in any given conflict, a position must be adopted from which one tendency and one side takes priority over the other.

It is impressive how quickly human beings arrive at side-taking decisions. If indeed human beings are fast decision makers, we should ask what happens with a quick decision about another human being. Is it quickly and constantly revised? Apparently not that often. There is a large body of both anecdotal and also more systematic accounts that suggests that most decisions about other human beings (in scenarios with or without side-taking) tend to last for a long period of time. This is documented for first impressions of facial judgment (Todorov, Pakrashi, & Oosterhof, 2009) as well as for legal judgment (Porter & ten Brinke, 2009). If the data about the speed of side-taking decisions are indeed robust, one may wonder how and why the inclination toward quick judgment is halted once a first judgment has been reached. How and when can a quick decision about someone be sustained?

One answer may be: When one humanizes him or her. If it appears that the one whose side we have taken acts in a way that we find especially agreeable, understandable, or consequential given his or her situation, it seems that we have chosen the correct side. In short, after the quick side-taking, empathy may come in and maintain the quick decision. Empathy would then be a secondary act, a consequence of taking sides. I share feelings with the other because I have decided for him.

At first (and maybe even second) glance, this suggestion seems counterintuitive: I have “shared feelings with” somebody, because I have decided for him and taken his side. One would expect: I decide for somebody, because I have empathy with him. A familiar example may help make the thesis more plausible: sports competitions, in this case soccer. Whoever observes a game without specific self-interest in the outcome can certainly appreciate the performances. However, it often only becomes exciting when he or she has decided for a team and from then on lives or dies with its players. Only then can the events of the game become experienceable, only then can the attempted shot become a second of horror or a moment of hope. Every foul also becomes an emotional event, because one either feels pain with the player and curses the fouling opponent, or dismisses the player rolling around on the ground as a great melodramatic actor but no soccer player.

When side-taking is not based on predetermined decisions, the selection remains vacillating, uncertain, and in need of legitimization. Given this uncertainty, the question becomes how one can feel good about the decision. My suggestion is that empathy allows emotions to be released, and these emotions confirm the initial decision. In short, empathy can be regarded as a mechanism for strengthening a decision. The more clearly I feel the pain of the one for whom I decided, the stronger my rage will be against his adversaries, and I will side even stronger with the one for whom I already decided. Vice versa, the suffering of the opponent, against whom I decided, can release negative empathy such as schadenfreude. Schadenfreude and negative empathy are also means to the legitimization of my own side-taking and thus prevent desertion (of the side I have chosen). According to this model, the function of empathy is to strengthen the taking of sides and to confirm created alliances. Empathy is the medium through which the quick act of taking sides becomes more durable. And the prior act of side-taking blends empathy with sympathy.

Based on these introductory remarks, we are now in the position to describe which kind of empathy might be involved in sustaining side-taking decisions. Here, the combination of the following processes should be considered:

To understand how side-taking and empathy interact, it is important first of all to stress again that empathy does not in itself involve side-taking (as in sympathy). As noted before, some of the facets of empathy can be employed to compete with the other (Bräuer, Call, & Tomasello, 2007), to directly harm him or her (Sussman, 2005), or can be part of some form of enjoyment as in sadistic impulses or schadenfreude. Perhaps the common association of empathy with sympathy or positive inclination stems partly from the fact that empathy and side-taking often form an alliance as proposed in this article, but this is by no means an automatic process. Psychopaths, for example, do not seem to lack the ability to have theory of mind, but they do not “care” for the other; do not combine theory of mind with sympathy (Harenski, Harenski, Shane, & Kiehl, 2010).

What is empathy without side-taking or prior to side-taking? Or put differently: What does empathy have to add to side-taking? Depending on which form of empathy is invoked, there may be two key forms of enhancement of the side-taking.

If a form of “hot” empathy is used—which is a form of empathy that involves the simulation of the experience or emotion of the other within one’s neuronal network—then the resulting empathy can be described as giving emotional weight to the chosen side. If we ourselves “feel the pain” of someone else or copy their stress or joy, we have to first give mental space to the other. The expected result of this activity then should be to prolong the side-taking and deepen it, since it receives additional experiential qualities.

If a form of “cold” empathy is used—which is a form of empathy that results in the understanding of the other’s mind, reasoning or emotions, without simulating the experience in one’s own neuronal networks—then the resulting empathy likely involves the constructing or reconstructing of some sequence of events. The “understanding” involved in cold empathy typically allows us to imagine knowing how the other is feeling, because we know or guess that x just happened to him (or because we understand that his situation gives him few options and we see what is coming). For example, my daughter comes home from school in a bad mood, so I may reason that she failed an exam. Hence, I have empathy with her as long as I believe that it is her exam which causes her moodiness. This sequencing or narrating has the additional effect that it “makes sense” of the other’s situation. Obviously, this may be a short circuit of the observer. The other behaves in a certain way. The observer explains this behavior because he knows or imagines that it was caused or triggered by another prior event. Hence, since the observer just added this prior event to explain the behavior, it consequently also makes sense to him. Imagination plays a necessary role here since the observer needs to intuit which prior event might have triggered the behavioral response (Sherman, 1998).

Sense-making is not the same as justifying behavior, but it comes close since it assumes some pattern of reasonable expectations. The behavior of the other makes sense to me since I could reasonably expect that his or her behavior is triggered by his or her situation. This process of sense-making could be described as producing a narrative order (Sternberg, 1990, 1992). Even if I do not agree with the behavior morally, it does not surprise me (or only minimally so; Norenzayan, Atran, Faulkner, & Schaller, 2006). Hence, the effect of cold empathy for quick side-taking is to bestow “sense” to the behavior of the other, making it appear as an adequate or at least common behavior in a situation. And this often comes close to normalizing and justifying the behavior, thus helping to sustain the side-taking: since the other acts in a normal, human way, he deserves my side-taking (or, conversely, when we have a prior negative image of the other we also predict and expect further negative actions).

To be sure, the terms “hot” and “cold” empathy are not used as empirical categories, but as merely heuristic idealization. It is quite possible that “cold” empathy for most people also causes some simulation and thus becomes lukewarm empathy. For a more complex discussion, see the debate concerning simulation theory and theory theory (Kögler, 2000; Stueber, 2006).

The point here is that the overall effect of (cold or hot) empathy for side-taking is to prolong, solidify, and quasi-justify the quick decision of taking someone’s side. The prediction is that empathy in a scene of three people leads to a higher degree of commitment to and loyalty with the chosen side than would be the case in a scenario without side-taking.

Situations that induce side-taking, meaning observations of conflict, seem to be especially well suited to trigger empathy, that is, to bypass empathy’s blocking mechanism. As indicated before, a combination of two reasons could account for this special status of such three-person scenarios: (a) the human tendency to quickly make judgments or side with someone, and (b) the relative simplicity of theory-of-mind observations since the impact of the adversary is likely to be clearly marked (that is, aesthetical).

Not every scene that involves three people is actually a three-person scenario in the described sense. Likewise, some scenes that only seem to involve two people may nevertheless involve three positions. It is easy to imagine that an observer may have empathy with a beggar because the observer can imagine countless scenes where the beggar was rejected by others (the third person), even if they are absent. The observer can also have empathy with another because the observer realizes that he himself wronged the observed. Hence, the observer occupies two positions, one as a party in conflict with the other, and one as an observer of that very conflict.

The basic situation of fiction corresponds to the already mentioned three-person situation, insofar as the reader/viewer takes on the reserved role of the third (the observer) and sides with one of the characters. One can also come down to the side of the “villain,” and this decision too can be legitimized through narrative empathy (for issues of fiction and judgment, see Phelan, 2007). In such a case, one may empathize with the villain’s attempt to escape punishment. Even villains become people in the eyes of the observers when they reach the scaffold, Adam Smith observed (Smith, 1759/1976). Legitimization of choosing the villain’s side can also take place, against moral feeling, because the villain is the more interesting character, who does or says what no one else in the fictional world dares to.

In general, there has been a remarkable absence of research on empathy in three-person scenarios. In particular, few or no experimental designs seem to distinguish clearly between two- or three-person scenarios. Fairly simple fMRI experimental designs may give some indication of the validity of the proposed three-person model by exposing subjects to videos with different scenarios of a conflict between two individuals. A case can be made that Stockholm syndrome is a case of a three-person scenario with hostage, hostage taker, and police (Breithaupt, 2009). One could describe the hostage as an observer of the conflict between police and hostage taker, with the hostage siding with one or the other. If the hostage sides with the hostage taker, which is the exception (Fuselier, 1999), Stockholm syndrome may be the result. The strong presence of a third party (police) in conflict with the second (hostage taker) distinguishes torture from hostage scenarios, and there is little to no evidence for identification or Stockholm syndrome in torture (Regner, 2000).

Patterns of Side-Taking

How does one choose a side (assuming one is not already or quasiautomatically connected with one party)?

For this decision, a number of forms are available. In the following, only the most interesting for our purposes will be discussed.

Conclusion

Whereas this article can only speculate about the precise mechanisms behind blocking and enabling empathy, it suggests a process by which one can consider the influence of individual learning and the cultural impact of empathy. This also gives some hope for the learning and cultivation of empathy as a virtue. However, it also should caution putting too much hope on empathy as a bond in society. After all, if the model is correct, empathy is especially strong in situations of side-taking. This means there are not only winners but also losers in the empathy race. These described mechanisms also show how empathy can be manipulated for various purposes, including in legal contexts. The winner of sympathy and empathy in a courtroom may not be the innocent but the one with the better story. Should judges have empathy? Only if they have learned to block and control it. But how one does this, we hardly know.