Empathy and its Development: What is Missing?


Professor Alexandra Main

Alexandra Main, Psychological Sciences, University of California, Merced



July 2017 – Philosophers have discussed the phenomenon of empathy since the 1800s. Robert Vischer (1873) coined the term “einfühlung” (translated as “feeling into”), which was later translated into “empathy” by Theodor Lipps (Lipps, 1914). However, the empirical study of empathy in psychology did not gain much traction until the 1980s. Behaviorism had fallen out of fashion at this point, and psychology researchers became more interested in the “black box” of what was going on inside people’s heads. This spurred a burst of research on emotion. Though there has been debate over whether empathy is an “emotion” per se, one would be pressed to find an empathy researcher who did not believe that emotion is inherent to empathy.

How do psychologists study empathy?

In psychology, empathy is often defined as feeling what another feels. This seems simple enough, but upon further reflection, there are several caveats. Is a person empathizing if they feel what someone else is feeling because of emotional contagion (i.e., the tendency to “catch” others’ emotions), or is a cognitive understanding of why someone is feeling a certain way necessary? How are other emotional responses to others’ emotions, including sympathy and compassion, related to or distinct from empathy? How are vicarious emotional responses to others’ emotions that do not involve feeling the same emotion as another, such as feeling angry in response to someone being hurt, related to or distinct from empathy? Is empathy always a precursor to positive social behaviors, such as helping another person? What about when we feel too overwhelmed by someone else’s distress (i.e., personal distress) to help? Are we empathizing in this case?

These questions have led many researchers in psychology to argue that empathy is a multidimensional phenomenon. Davis (1983) argued that empathy has both affective (i.e., feeling what someone else is feeling) and cognitive (i.e., understanding someone else’s perspective) components. In other words, it is not enough to simply feel sad when we see someone who is sad, but we must also appreciate that the person is sad because their dog died. This view continues to pervade much of the thinking in the field and most empathy researchers distinguish between affective and cognitive components of empathy (see Brown et al., 2017; Cuff et al., 2016). Research with adolescents suggests that these different components of empathy have distinct developmental timelines (e.g., Van Lissa et al., 2014; Davidov et al., 2013) and there may be distinct neurological processing for more cognitively mediated vs. “pure” forms of empathy (Zaki & Ochsner, 2012).

An excellent example of research on the multidimensional nature of empathy comes from a study by Zaki and colleagues (Zaki et al., 2009). In this study, research participants watched videos of individuals telling emotional stories while in an fMRI scanner. Participants were instructed to rate using a dial how positively or negatively the storyteller was feeling at that moment while they were talking about their emotional event. These ratings were then mapped onto the ratings the storytellers had made of their own feelings to develop a measure of empathic accuracy – how accurate the participants were at judging the storytellers’ emotions. The researchers found that certain systems in the brain associated with understanding others’ mental states were particularly active during this task; specifically, the superior temporal sulcus and medial prefrontal cortex (Zaki et al., 2009).

Other neuroscience research examining individuals’ responses to the pain of others activate more “primitive” neural networks, including the anterior cingulate cortex, brainstem, and cerebellum (Singer et al., 2004). The researchers argue that the processes involved in “mentalizing,” or trying to understand other people’s point of view, is a neurologically distinct process from the more visceral reaction we experience when we witness someone experiencing an emotion, particularly distress or pain (Zaki & Ochsner, 2012). This suggests that not only are there multiple psychological ingredients that are necessary for empathy, but there is physical evidence in the brain of this multidimensionality as well.

What do we know about the development of empathy?

How we define and operationalize empathy has important implications for understanding its development. The work of Paul Bloom, Michael Tomasello, and others suggests that babies may not need to learn empathy; rather, they are capable of empathy from birth – entering the world as altruistic beings. In a clever early study, infants were either played a recording of another newborn’s cry, a synthetically-produced cry, or silence (Sagi & Hoffman, 1976). Infants were significantly more likely to cry in response to another newborn’s cry than in the other two conditions. Many have interpreted this to mean that infants have a rudimentary form of empathy from birth. Though there are competing explanations to this phenomenon, it is clear that even infants seek to connect with others, be it through the social smile emerging around 6-8 weeks (Anisfeld, 1982), intersubjectivity and joint attention emerging around 6 months (Butterworth & Cochran, 1980), or helping behavior emerging around 15 months of age (Zahn-Waxler et al., 1992).

Unlike research with adults that typically relies on self-report measures to study empathy, researchers needed to be a bit more creative in studying empathy in toddlers and children. Early studies with toddlers involved systematically observing children’s responses to another’s simulated distress (either the mother or an experimenter) by coding their facial expressions and behaviors (e.g., facial expressions of concern, personal distress, hypothesis testing, helping). This early work showed that between 12-24 months of age, children became more sophisticated in their responses to others’ distress (Zahn-Waxler et al., 1992). For example, in response to a mothers’ simulated distress (e.g., “hurting” herself by bumping her foot), infants engaged in more reparative behaviors as they got older, such as engaging in prosocial behavior (actively trying to relieve the parent’s distress). Infants also become less egocentric as they get older. Specifically, in the same study, infants engaged in fewer self-focused responses to the parent’s distress, including crying or rubbing the infant’s own foot, as they got older.

While very young children are sensitive to others’ distress, they are not necessarily very effective at using their empathic responses to motivate successful helping of others. For example, younger children may bring an experimenter or the caregiver their own toy in response to the adult’s distress. This reflects an understanding that the person is in distress, but a failure to demonstrate the capacity to respond to the adult in a way that will reduce the adult’s distress. Martin Hoffman’s seminal writings in the field of psychology argued that such a progression from “egocentric empathy” to veridical empathy was fundamentally a cognitive process involving self-other distinctions that become more differentiated and complex with development (Hoffman, 1982).

Other researchers posited that such development of other-oriented empathy fundamentally involves the developing ability to regulate one’s own distress in response to another’s distress. We might initially feel upset when we see someone hurt, but if we can turn our attention away from our own distress and focus on the other person, we will be better equipped to help them. Nancy Eisenberg demonstrated that children’s effortful control (a temperamentally-based individual differences in the ability to deploy attention and inhibit undesirable behaviors) is a precursor to sympathy. She argues that sympathy is a well-regulated empathic response that involves care and concern for others, as opposed to personal distress, which involves poorly-regulated distress in response to another’s distress that is focused on the self rather than on the other (Eisenberg et al., 2013). Though Eisenberg’s and Hoffman’s theories have their subtle distinctions, they overlap in their view of empathic development as fundamentally progressing from a “pure” emotional response to a more cognitively-controlled process.

What is missing in the study of empathy?

The research described thus far has provided great insights into the development and neurological substrates of empathy (see Uzefovsky & Knafo-Noam, 2017 and Zaki & Ochsner, 2012 for reviews). However, the focus of empathy research has traditionally emphasized the internal experience of the empathizer. This focus has often come at the detriment to understanding how empathy is an interpersonal, relational process. My colleagues and I have argued that current conceptualizations of empathy as feeling what another is feeling are not necessarily wrong, but they fail to capture the richness of the empathic process and do not provide many practical implications for how people can learn to be better empathizers (Main et al., 2017).

Consider the following example: A mother finds out that her teenage son has been sending text messages after he is supposed to be in bed. As punishment, she takes his phone away. The adolescent responds by withdrawing into his room (a common behavior among adolescents). The parent may interpret this withdrawal as anger at his phone being taken away. However, in reality, she failed to see that her son was in fact sad because he viewed his mother’s punishment as a lack of trust.

This example highlights several elements of empathy that are often missing in empirical research. First, empathy is inherently interpersonal in nature. If we think of empathy only as something happening within the mother (i.e., the empathizer), she was empathizing with her son by putting herself in his shoes. However, she made an incorrect assumption about the motivation for her son’s behavior. Second, empathy is contextually-bound. In this situation, the adolescent withdrew from his mother, which made it more difficult for her to empathize. However, in another situation, the adolescent may have been more willing to disclose the reasons behind his feelings, which would have facilitated empathy between mother and son, likely leading to a more positive long-term relationship. Third, empathy is a dynamic process. The mother may have initially assumed her son was angry about his phone being taken away, but if we were to stop our assessment there, we would miss whether she persisted in this belief, or if she attempted to interact with her son to determine if her initiate judgment was accurate. Without an appreciation of the corrective processes involved in real-time empathizing with others, we are missing the deeply dynamic nature of empathy. Fourth, this example highlights that the mother’s curiosity is central to effectively empathizing with her son. Below I provide examples from the real world as well as from research to highlight these underemphasized aspects of empathy.

Empathy is interpersonal. Although it may seem obvious that empathy is an interpersonal process – with whom would we empathize if not another person? – empathy is typically studied in solitary contexts. This is likely based on convenience (it is easier to study one person than two!), but theoretical perspectives that view that empathy is an internal experience. However, this neglects the fact that empathy is an interactive social process that depends not only on the empathic tendencies of the empathizer, but also on the openness or resistance of the social partner to being empathized with (Hollan, 2008; Ickes et al., 1997).

Consider a follow-up study by Zaki and colleagues in which the researchers examined the perceivers’ empathic accuracy and how empathic they generally reported themselves to be on a self-report questionnaire (Zaki, Bolger, & Ochsner, 2008). They found that people who scored highly on an empathy questionnaire, unsurprisingly, tended to be more accurate in their empathic judgments while observing the storytellers telling emotional stories. However, the perceivers were only accurate in the cases when the storyteller was highly expressive. This suggests that even highly empathic people may struggle to empathize if the person with whom they are trying to empathize is emotionally blunted or closed-off.

When the person we are trying to empathize with gives us feedback, such as telling us we are wrong and displaying an angry facial expression suggesting we misunderstood them, this helps us better empathize with those around us. In other words, when we consider the openness or resistance of others to being empathized with, we can appreciate the fact that empathy is a two-way street.

Empathy is contextually-bound. Another limiting factor to much of the research on empathy is that it is typically studied as a static trait (you expressed empathy or not in this situation, you are an empathic person or not). However, empathy is highly context-dependent. Sharing someone else’s emotional experience may be adaptive in some contexts, such as when we are motivated to help another by sharing their pain, but sharing someone else’s emotions may not always be the most effective route to empathizing. For instance, empathy is a crucial piece of effective conflict resolution (e.g., Levenson & Ruef, 1992), but sharing someone else’s negative emotions during a conflict is likely to lead to conflict escalation rather than resolution (Gottman et al., 2014).

John Gottman and his colleagues have conducted several decades of careful research in which they have observed couples discussing issues of conflict with one another and analyze the second-by-second emotion expression of each partner. The researchers consistently find that couples who allow their negativity to escalate, rather than engage in attempts to repair the relationship after one partner expresses negative emotion, report lower marital satisfaction and are far more likely to get divorced (see Driver et al., 2012). Thus, empathy in a conflict context may best be characterized as validating or showing interest in someone else’s emotions, not matching their emotion (Main et al., 2017).

Empathy is a dynamic process. Empathy does not occur at a finite point in time, but rather unfolds dynamically over time. Despite this, much of the empirical research on empathy uses self-report questionnaires about dispositional tendencies to empathize or requires participants to make one-time judgments about others’ emotions. However, empathy is often a corrective process. This is evident when we use others’ facial, verbal, and postural cues to determine whether we are accurate or not in our assessments of our partner’s emotions, such as when a wife displays a flash of anger toward her husband who immediately tries to make her feel better before validating her sadness.

Some individuals may not pick up on such cues, but others may use these cues to “correct” their assumptions. Without an appreciation of these individual variations and how contextual factors might influence these differences, we may be missing some very fundamental aspects of the empathy process.

The role of curiosity in empathy. Much of the research on empathy involves whether individuals successfully label a person’s emotions. However, empathy in the real world is more complex. Consider being at a party and you seeing a friend in the corner looking sad. It is not enough to identify your friend’s expression as communicating sadness. Rather, truly empathizing involves understanding why your friend is sad, which requires a curiosity about her with her environment.

This issue is particularly relevant in clinical settings. Recent research in behavioral medicine has addressed the potential for psychological burnout among physicians who share their patients’ distress (see Halpern, 2014). Thus, other empathic behaviors, such as curiosity about another’s emotional point of view, may be more effective in situations involving caring for others. Indeed, curiosity about a patient’s feelings and perspective facilitates patient disclosure, which predicts healthy outcomes in patients (Suchman et al., 1997). Simply labeling a patient as angry about a cancer diagnosis or displaying sympathy toward a patient who does not want others feeling sorry for her would not be considered empathic in real life, so it makes little sense that empathy is usually empirically studied in this way.

Empathy is culturally-situated. It is important to note that everything described thus far is culturally-situated. How we define empathy, express it to others, respond to others’ emotions, and how much different cultures value empathy can vary considerably.

Douglas Hollan’s ethnographic work highlights cultural differences in social expectations about how we express emotions and respond to the emotions of others – both of which have tremendous implications for how we think about empathy in different cultures (Hollan, 2012). For example, it is considered culturally inappropriate among the Yapese (a rural cultural group in Indonesia) to appeal for sympathy from others, while in another cultural group, the Toraja, expressions of vulnerability are key to positive social interactions. Furthermore, while we typically define empathy in Western society as an internal state (though I hope I have convinced the reader that we need to move beyond this), other cultural groups view empathy as an active, instrumental response (e.g., exchanging goods) rather than a passive sharing of another’s emotional experience (von Poser, 2011).

Taken together, understanding empathy from a cross-cultural necessitates that even the basic definition of empathy needs to be understood from a culturally-situated point of view.

Opportunities for future research

Psychologists have made tremendous progress toward better understanding of the internal experience of empathy. This emphasis on internal experience is likely because psychologists are inherently interested in what is going on in the human mind. But humans do not exist in a vacuum. In the real world, we are very rarely asked to label other people’s emotions explicitly without any feedback and we do not often encounter situations where we have only a one-time opportunity to empathize with others. Often psychology researchers emphasize tight control over experimental designs in the quest for internal validity. However, this may jeopardize the external validity of much of this research (Campos et al., 2011).

Research from other fields, including anthropology, behavioral medicine, and linguistics, can serve to complement the tightly-controlled research emphasized in psychology. For example, conversation analysis, a qualitative technique used in linguistics research, allows researchers to track behaviors over time, including empathy. One study found that empathic displays (e.g., validating statements) occurring early in a conversation may shut down a conversation by discouraging one’s partner from opening up, but such statements were highly effective at promoting empathic understanding later in the conversation (Kupetz, 2014). An important goal for future research would be to incorporate and develop quantitative methodologies that balance internal and external validity to better capture the phenomenon of empathy.

Another opportunity for depending our understanding of empathy would be to allow for more behavioral flexibility in our dependent measures. For example, while allowing subjects to interact with the target of empathy in a study adds a layer of complexity and messiness to a study, we may develop a deeper understanding of how people empathize with others in real time.

Empirical research on empathy typically starts with a priori assumptions of what kinds of behaviors are expected to be empathic (e.g., facial expressions of concern, overt helping behavior), but this reflects an emphasis on the form of empathy rather than its function. I have highlighted here that in some contexts affect matching might be appropriate, but in others, curiosity might be a more empathic response. Shifting our focus from assuming certain behaviors are inherently empathic to a greater appreciation of the contextual appropriateness of such behaviors would move the study of empathy toward greater clarity.

Despite the challenges associated with studying empathy from an interpersonal and ecologically valid perspective, it is time for empirical research to take a more dynamic and relational approach to shed light on how we can better empathize with others across the complex situations real life presents to us.



Anisfeld, E. (1982). The onset of social smiling in preterm and full-term infants from two ethnic backgrounds. Infant Behavior and Development, 5, 387-395.

Brown, M. M., Thibodeau, R. B., Pierucci, J. M., & Gilpin, A. T. (2017). Supporting the development of empathy: The role of theory of mind and fantasy orientation. Social Development.

Butterworth, G., & Cochran, E. (1980). Towards a mechanism of joint visual attention in human infancy. International Journal of Behavioral Development, 3, 253-272.

Campos, J. J., Walle, E. A., Dahl, A., & Main, A. (2011). Reconceptualizing emotion regulation. Emotion Review, 3, 26-35.

Cuff, B. M. P., Brown, S. J., Taylor, L., & Howat, D. J. (2016). Empathy: A review of the concept. Emotion Review, 8, 144-153.

Davidov, M., Zahn‐Waxler, C., Roth‐Hanania, R., & Knafo, A. (2013). Concern for others in the first year of life: Theory, evidence, and avenues for research. Child Development Perspectives, 7, 126-131.

Davis, M. H. (1983). Measuring individual differences in empathy: Evidence for a multidimensional approach. Journal of Personality and Social Psychology, 44, 113-126.

Driver, J., Tabares, A., Shapiro, A. F., & Gottman, J. M. (2012). Couple interaction in happy and unhappy marriages: Gottman Laboratory studies. In F. Walsh (Ed.), Normal family processes: Growing diversity and complexity (4th ed.), (pp. 57-77). New York, NY, US: Guilford Press.

Eisenberg, N., Betkowski, J., & Spinrad, T. L. (2013). Age-related changes in empathy-related responding. In D. Hermans (Ed.), Changing emotions, (pp. 17-23). New York, NY, US: Psychology Press.

Halpern, J. (2014). From idealized clinical empathy to empathic communication in medical care. Medicine, Health Care and Philosophy, 17, 301-311.

Hoffman, M. L. (1982) Affect and moral development. New Directions for Child Development,16, 83-103.

Hollan, D. W. (2008). Being there: On the imaginative aspects of understanding others and being understood. Ethos, 36, 475-489.

Hollan, D. W. (2012). Emerging issues in the cross-cultural study of empathy. Emotion Review, 4, 70-78.

Ickes, W. J., Marangoni, C., & Garcia, S. (1997). Studying empathic accuracy in a clinically relevant context. In W. J. Ickes (Ed.), Empathic accuracy (pp. 282-310). New York, NY, US: Guilford Press.

Kupetz, M. (2014). Empathy displays as interactional achievements – Multimodal and sequential aspects. Journal of Pragmatics, 61, 4-34.

Levenson, R. W., & Ruef, A. M. (1992). Empathy: A physiological substrate. Journal of Personality and Social Psychology, 63, 234-246.

Lipps, T. (1914): Aesthetik: Psychologie des Schoenen und der Kunst, Vol. 1, Grundlegung der Ästhetik, Leipzig/Hamburg: Voss.

Main, A., Walle, E. A., Kho, C., & Halpern, J. (2017). The interpersonal functions of empathy: A relational perspective. Emotion Review.

Sagi, A. & Hoffman, M. L. (1976). Empathic distress in the newborn. Developmental Psychology,12, 175.

Trevarthen, C. (1979). Communication and cooperation in early infancy: A description of primary intersubjectivity. In M. Bullowa (Ed.), Before speech: The beginnings of interpersonal communication. Cambridge: Cambridge University Press.

Uzefovsky, F., & Knafo-Noam, A. (2017). Empathy development throughout the life span. In J. A. Sommerville & J. Decety (Eds.), Social cognition: Development across the life span (pp.  71-97). New York, NY, US: Routledge/Taylor & Francis Group.

van Lissa, C. J.; Hawk, S. T., de Wied, M., Koot, H. M., van Lier, P., & Meeus, W. (2014). The longitudinal interplay of affective and cognitive empathy within and between adolescents and mothers. Developmental Psychology, 50, 1219-1225.

Vischer, R. (1873): Über das optische Formgefühl: Ein Beitrag zur Ästhetik, Leipzig: Credner.

Von Poser, A. (2011). Bosmun foodways: Emotional reasoning in a PNG life-world. In D. W. Hollan & C. J. Throop (Eds.), The anthropology of empathy: Experiencing the lives of others in Pacific societies (pp. 169–193). Oxford, UK: Berghahn.

Zahn-Waxler, C., Radke-Yarrow, M., Wagner, E., Chapman, & M. (1992). Developmental Psychology28, 126-136.

Zaki, J., Bolger, N., & Ochsner, K. (2008). It takes two: The interpersonal nature of empathic accuracy. Psychological Science, 19, 399-404.

Zaki, J., Weber, J., Bolger, N., & Ochsner, K. (2009). The neural bases of empathic accuracy. PNAS, 106, 11382-11387.

Zaki, J., & Ochsner, K. N. (2012). The neuroscience of empathy: Progress, pitfalls, and promise. Nature Neuroscience, 15, 675-680.

Print Friendly, PDF & Email

Traffic Count

  • 331520Total reads:
  • 77Reads today:
  • 265897Total visitors:
  • 65Visitors today: