Picture the maternity ward of a busy hospital. One infant begins to cry. Within moments another joins in, then another, until distress seems to travel through the room.

Psychologists have studied versions of this phenomenon for decades. Known as contagious crying, it has been observed even among very young infants. Recent cross-cultural research suggests that babies respond differently to another infant’s cry than to an equally unpleasant artificial sound. Something social appears to be happening. Yet caution is necessary. A crying newborn is not demonstrating mature empathy. The infant cannot understand another person’s predicament, imagine their perspective or decide to offer assistance. What we may be seeing is something more elemental: emotional contagion, an early sensitivity to the emotional states of others from which more sophisticated forms of empathy can later develop.¹

Now move forward thirty years.

A man stumbles in a crowded railway carriage. His shopping scatters across the floor. Apples roll beneath seats. People look away. Someone steps over a bag. Several passengers retreat into their phones.

The contrast poses one of the oldest questions about human nature. If we are biologically equipped to respond to another’s distress, why are we also capable of extraordinary indifference?

The answer is that the familiar question—is empathy innate or cultured?—sets up a contest that does not really exist.

Human beings appear to inherit capacities from which empathy can emerge. But those capacities develop through relationships, are shaped by culture, extended or restricted by social identity, influenced by circumstances and capable of deliberate cultivation. We are neither born universally compassionate nor manufactured as moral beings entirely by society.

Nature gives us capacities. Culture helps determine their boundaries. Practice influences how we use them.

That distinction matters because empathy itself is not one thing.

Psychologists generally distinguish at least two important components. Affective empathy involves sharing, or resonating with, aspects of another person’s emotional state. Cognitive empathy involves understanding another person’s perspective without necessarily sharing their emotion. Compassion overlaps with both but is better regarded as something distinct: concern for another’s suffering accompanied by some motivation that the suffering should be relieved.²

The differences are important. A skilled interrogator may understand another person’s fears with considerable cognitive empathy while feeling little concern for them. Someone overwhelmed by another person’s grief may share their suffering intensely but become incapable of helping. Compassion requires something further: not merely seeing or feeling another’s distress, but responding to it constructively.

The human story is therefore not about possessing or lacking empathy. It is about how a collection of ancient social capacities becomes a moral practice.

The First Layer: Capacity

Empathy did not begin with philosophy, religion or civilization. Its evolutionary roots are considerably older.

Among our closest primate relatives, researchers have repeatedly observed what is known as consolation behaviour. After a conflict, an uninvolved chimpanzee may approach the loser, embrace them, touch them or groom them. Long-term observations suggest that individual chimpanzees differ consistently in their tendency to console others, and that greater consolation is associated with social integration.³

Such behaviour cannot tell us exactly what a chimpanzee feels. We should resist turning animal behaviour into human psychology. But it does tell us something important: social responsiveness to another individual’s distress is not exclusively human and does not require modern civilization.

Charles Darwin understood this better than the caricature of “survival of the fittest” sometimes attached to his work. In The Descent of Man, Darwin treated sympathy as part of human social evolution. Groups containing individuals willing to warn, defend and assist one another could gain advantages precisely because cooperation matters to survival.⁴

Later evolutionary theory developed several mechanisms through which cooperation could emerge, including kin selection and reciprocal altruism. Robert Trivers demonstrated how helping behaviour can evolve between non-relatives when assistance is exchanged across repeated social relationships.⁵

Our ancestors were not solitary competitors wandering across the savannah. Humans evolved as intensely social animals. Children depended on prolonged care. Food was shared. Knowledge was transmitted. Coalitions formed. Injury and illness created dependence. Survival required an extraordinary capacity to monitor other people: Who is afraid? Who is trustworthy? Who is angry? Who needs help? Who intends harm?

The ability to register another person’s state became biologically consequential.

Modern neuroscience supports this broad picture, although not in quite the simple way popular accounts once suggested.

The discovery of mirror neurons in macaque monkeys during the 1990s generated enormous excitement. Certain neurons involved when a monkey performed an action were also active when the monkey observed an action performed by another. The discovery quickly escaped the laboratory and entered popular culture. Mirror neurons were invoked to explain imitation, language, art, morality and empathy itself.

The evidence is more restrained.

A 2021 systematic review and meta-analysis covering 52 studies found associations between activity in the putative human mirror-neuron system and some forms of cognitive and emotional empathy, but described the evidence as preliminary and methodologically heterogeneous.⁶ Empathy cannot be reduced to a single neural mechanism.

What neuroscience instead reveals is a distributed social brain.

In a landmark experiment, Tania Singer and colleagues used brain imaging to compare what happened when participants experienced pain themselves and when they knew that a loved one was receiving a painful stimulus. Regions including the anterior insula and rostral anterior cingulate cortex responded in both situations. Crucially, however, the overlap concerned affective aspects of pain rather than the complete sensory experience. The observers did not literally feel another person’s injury. Their brains represented something of its emotional significance.⁷

Empathy, then, is neither mystical telepathy nor a simple neural reflex. It is a suite of processes through which one nervous system models aspects of another.

Even oxytocin, perhaps the most famous chemical in popular accounts of human bonding, illustrates the danger of reducing compassion to biology.

Oxytocin unquestionably has important physiological roles in childbirth, lactation and social behaviour. Earlier experiments also suggested effects on trust and in-group cooperation. But the once-fashionable description of oxytocin as a “cuddle hormone” is now difficult to sustain. A large 2026 pooled replication found no meaningful general increase in trusting behaviour after intranasal oxytocin. Another 2026 preregistered study, however, found increased trust among a specifically selected population of men with low dispositional trust.

The apparently contradictory findings are instructive. Biology does not provide a kindness switch. Neurochemistry operates within persons, relationships and situations.

We possess machinery capable of social connection. What it produces depends partly on the world in which it operates.

The Second Layer: Boundaries

That qualification brings us to the more difficult half of the empathy problem.

Human beings are capable of extraordinary care. We are also capable of confining that care within remarkably narrow boundaries.

Evolution did not design a nervous system to contemplate the welfare of eight billion strangers. The social capacities from which empathy developed operated within relationships where identity, familiarity, kinship, reciprocity and reputation mattered.

Empathy therefore has a geography.

We tend to respond differently depending on who is suffering, whether we identify with them, whether we perceive them as responsible for their predicament and whether they belong to a group we regard as our own. Empathy can be selective. It can be biased. It can coexist with hostility towards outsiders.

This is where culture becomes decisive.

Biology gives us the ability to enter relationships with other minds. Culture helps teach us which minds count.

Family, religion, nation, race, class, ideology and moral tradition can all draw the circumference of concern differently. They can enlarge it or contract it. They can tell us that a stranger is a neighbour, or that a neighbour is an enemy.

The first lessons arrive long before a child can articulate any of this.

Attachment research suggests an association between the quality of early relationships and later empathic capacity. A 2022 meta-analysis drawing on 50 studies and more than 24,000 children and adolescents found that secure attachment was positively associated with empathy, while avoidant attachment was negatively associated with it.⁸ The relationships are not deterministic. Securely attached children do not inevitably become compassionate adults, and difficult childhoods do not condemn people to emotional incapacity. But the evidence supports something intuitively important: we learn how relationships work partly by being in them.

A responsive caregiver repeatedly performs a profound social lesson.

Your distress is visible.

Someone notices.

Someone responds.

Over time, children also learn to regulate their own emotions. This matters because mature empathy requires both openness and boundaries. To help another person, we must be able to approach suffering without becoming entirely consumed by it.

Culture then builds upon these early foundations.

Even the language used to describe empathic behaviour varies between societies. Japanese discussions of omoiyari, for example, commonly emphasize attentiveness to another person’s unspoken needs and sensitivity to interpersonal context. Anthropologist and psychologist Hidetada Shimizu has shown, however, that omoiyari should not be treated as evidence that Japanese individuals somehow possess a single national form of empathy. It is better understood as a cultural script: an ideal that people interpret, negotiate and sometimes resist.⁹

That distinction is crucial.

Culture does not program identical emotional responses into every member of a society. It provides narratives, expectations, rituals and categories through which human social capacities are interpreted.

And cultures possess another extraordinarily powerful technology for extending empathy: story.

A novel permits something remarkable. For several hours, a reader may inhabit the concerns of somebody who does not exist.

The historical claim that the rise of the novel somehow invented modern empathy is too strong. People were clearly capable of identifying with strangers long before eighteenth-century fiction. But experimental evidence does support a more modest proposition. A meta-analysis of 14 experimental studies found that fiction reading produced a small but statistically significant improvement in social-cognitive performance compared with nonfiction or no reading.¹⁰

The effect is small, but the principle is large.

Stories permit mental travel across boundaries.

Literature can take us inside a refugee camp without leaving home. Theatre can make an audience temporarily dependent upon the fate of a stranger. Documentary can attach a face to a statistic. Journalism can convert “casualties” back into names, families and unfinished lives.

None of these forms guarantees compassion. Propaganda uses story too. Narrative can dehumanize as efficiently as it humanizes.

Culture is not intrinsically an empathy machine.

It is a boundary-making machine.

The moral question is where those boundaries are drawn.

The Third Layer: Plasticity

If empathy is shaped by development and culture, how fixed is it?

For much of the twentieth century, debates about nature and nurture encouraged a misleading metaphor. Biology was imagined as the permanent blueprint; culture merely decorated what genes had already constructed.

Epigenetics has complicated that picture.

One of the best-known demonstrations came from Michael Meaney, Moshe Szyf, Ian Weaver and colleagues. Rat pups receiving different patterns of maternal licking and grooming developed different patterns of DNA methylation around a glucocorticoid-receptor gene in the hippocampus. These changes were associated with differences in regulation of the stress response, and cross-fostering experiments helped establish the importance of maternal behaviour.¹¹

The experiment is remarkable. But it does not show that affection switches “empathy genes” on while neglect switches them off.

Its real significance is subtler.

Experience can become biologically embedded.

The border between nature and nurture is therefore porous. Environments do not merely act upon a finished organism. Development is partly the process through which biology responds to environments.

The human brain remains plastic as well.

That does not mean anything can be changed by sufficient willpower, nor that adults can simply meditate themselves into different personalities. Neuroplasticity is not magic. But repeated mental and behavioural practices can alter patterns of neural activity, attention and behaviour.

Compassion training provides an intriguing example.

In a 2008 fMRI study, Antoine Lutz and colleagues compared expert and novice practitioners while they generated a loving-kindness and compassion meditation state and listened to emotional sounds. Expert meditators showed different patterns of activation in areas associated with emotional and social processing, particularly when confronted with sounds of distress.¹²

More importantly for the question of trainability, Helen Weng and colleagues subsequently tested adults receiving short-term compassion training. Participants who received the training subsequently behaved more altruistically in an economic redistribution task. Changes in altruistic behaviour were associated with altered activity and connectivity in brain systems involved in social cognition, regulation and reward.¹³

The study did not show that several weeks of meditation physically rebuilt the brain. It showed something both less sensational and more useful: compassion-related behaviour and the neural processes accompanying it can change with practice.

A later randomized trial likewise found changes in neural responses following four weeks of compassion meditation, although the mechanisms remain an active area of research.¹⁴

This suggests a distinction essential to any serious discussion of empathy.

Capacity is not destiny.

A predisposition can be exercised or neglected. Attention can be trained. Perspective-taking can be practised. Emotional reactions can be regulated. Habits can become easier through repetition.

The ethical traditions of the world intuited this long before neuroscientists entered the discussion.

Buddhist metta practice, Christian injunctions to love the neighbour and enemy, Jewish teachings concerning the stranger, Islamic traditions of rahma or mercy, Confucian cultivation of ren, and many Indigenous traditions of relational responsibility differ profoundly in theology and philosophy. Yet all reject the assumption that moral concern should simply be left to spontaneous feeling.

Compassion, in these traditions, is something practised.

That may be because spontaneous empathy is too unreliable to bear the entire weight of morality.

The Limits of Feeling

There is an uncomfortable truth here.

More empathy is not always better.

A person who absorbs every distress around them may become exhausted rather than helpful. Empathy can favour vivid individuals over statistical populations. We may feel intensely for one identifiable victim while remaining strangely unmoved by thousands.

Psychologists have explored this phenomenon under terms including psychic numbing and compassion collapse.

C. Daryl Cameron and B. Keith Payne found that people sometimes experience less compassion as the number of sufferers increases. Their experiments suggested something more psychologically interesting than simple emotional incapacity: people may regulate their compassion downward when they anticipate that caring will become costly or overwhelming.¹⁵

This changes the moral picture.

We do not simply possess a fixed reservoir of empathy that eventually runs dry. Sometimes we withdraw from feeling because feeling threatens to demand something from us.

Empathy therefore competes with self-protection.

It also competes with attention.

And that brings the problem directly into the twenty-first century.

Empathy in the Digital Environment

The digital age has produced one of the largest experiments in human social interaction ever conducted.

Billions of people now conduct substantial portions of their relationships through devices that did not exist during almost all of human evolution.

It would be simplistic to conclude that screens destroy empathy. Digital communication connects families separated by continents, permits marginalized people to find communities and allows suffering in one country to become visible almost instantly in another.

But digital environments change the conditions under which social cognition occurs.

Face-to-face interaction provides extraordinary quantities of information: timing, gaze, posture, vocal tone, hesitation, touch, facial movement and immediate feedback. Text-based communication removes many of those signals. Even video compresses them.

Psychologist John Suler described one consequence as the online disinhibition effect. Anonymity, invisibility, delayed interaction and reduced social authority can loosen behavioural restraints. Sometimes this produces unusual honesty and generosity. Sometimes it produces aggression that would be far less likely across a table.¹⁶

No “mirror-neuron brake” needs to be switched off to explain the phenomenon.

A simpler explanation is sufficient: psychological distance changes behaviour.

The architecture of platforms adds another layer.

Social-media systems compete for attention. Material capable of producing strong reactions can therefore acquire an advantage, particularly where engagement determines visibility. Outrage, humiliation, fear and tribal conflict can become socially contagious for reasons having little to do with whether individual users are naturally compassionate.

At the same time, scale presents another challenge.

For most of human history, a person might directly encounter the distress of dozens or hundreds of people. Today a phone can present war, famine, earthquake, abuse, extinction, political violence and personal tragedy before breakfast.

This does not prove that the human brain has a fixed neurological limit beyond which compassion shuts down. But research into psychic numbing and compassion collapse suggests why repeated exposure to large-scale suffering can encourage disengagement.

The problem is not that technology has removed empathy from human beings.

It is that technology has created environments capable of exploiting the weaknesses of empathy: its selectivity, its dependence upon attention, its susceptibility to group identity and its tendency to withdraw when emotional costs appear too great.

Those same technologies could be designed differently.

They can slow interaction rather than accelerate it. They can restore context. They can expose users to perspectives beyond their immediate group. They can prioritize meaningful communication rather than merely engagement.

Technology does not stand outside culture.

It has become one of culture’s principal architects.

Beyond Empathy

We can now return to the original question.

Is empathy innate or cultured?

The most defensible answer is: both, but not in the way the question implies.

What appears innate is not a finished moral faculty called empathy. Human beings arrive with capacities for emotional contagion, attachment, social attention and interpersonal learning. These capacities are part of an evolutionary inheritance shared, in simpler forms, with other social animals.

Development then elaborates them.

Relationships help children learn whether distress receives a response. Language supplies categories for emotions. Culture tells us whose suffering is visible. Stories allow us to cross social boundaries. Institutions reward some forms of concern and discourage others.

And throughout life, practice can influence what we attend to and how we respond.

The resulting picture can be summarized in three words:

Capacity. Boundary. Practice.

Biology contributes the capacity to resonate with and understand other minds.

Culture helps establish the boundary of whose experiences command our attention.

Individual and collective habits determine the practice through which concern becomes—or fails to become—action.

This also explains why empathy alone cannot serve as a complete moral philosophy.

Feeling another person’s suffering is valuable, but justice sometimes requires us to assist people whose suffering we cannot vividly imagine. Rights should not depend upon whether a victim tells a compelling story. Public policy cannot favour only those whose faces provoke an emotional response.

The task, therefore, is not simply to create a more empathetic society.

It is to build a society in which our capacity for empathy is disciplined by reason, widened by culture and translated into compassion and justice.

Schools can teach perspective-taking without pretending that prejudice disappears through a classroom exercise. Families can model emotional responsiveness. Literature and the arts can expose us to lives beyond our own. Journalism can restore individuality to statistics. Contemplative practices may help people regulate emotional reactions and cultivate concern. Digital platforms can be designed to encourage deliberation rather than reflexive antagonism.

None of this guarantees kindness.

That is precisely the point.

Empathy is not guaranteed.

Return once more to the railway carriage.

The man falls. Apples scatter across the floor.

For a fraction of a second, everyone nearby has a choice. Some may feel his embarrassment instinctively. Others may merely understand what has happened. Someone may be distracted. Someone may decide it is none of their concern.

And one person bends down.

That simple movement contains the entire argument.

The capacity to notice may be ancient.

The boundary that includes this stranger may have been culturally enlarged.

The habit of responding may have been practised for decades.

Biology made the moment possible. Culture helped give it meaning. Character turned possibility into action.

Perhaps the most important thing about empathy, then, is not whether we were born with it.

It is what we choose to do with the capacity we have.

This essay is part of a series reading kindness, empathy and compassion across religious, philosophical, literary and popular traditions.

The Backbone of Kindness: Rekindling Our Oldest Virtue in Troubled Times by Kevin J. Parker is published by Sane Earth, Sydney, in paperback, ebook and audiobook. Available on Amazon: https://shorturl.at/iO86g

Endnotes

1. C. Vreden et al., “Contagious Crying Revisited: A Cross-Cultural Investigation Into Infant Emotion Contagion Using Infrared Thermal Imaging,” Developmental Science 28, no. 2 (2025): e13608.

2. See, for example, the distinction between cognitive and affective empathy discussed in contemporary empathy research, and research treating compassion as a related but distinguishable motivational state. See also Jennifer L. Goetz, Dacher Keltner, and Emiliana Simon-Thomas, “Compassion: An Evolutionary Analysis and Empirical Review,” Psychological Bulletin 136, no. 3 (2010): 351–374.

3. Christine E. Webb et al., “Long-Term Consistency in Chimpanzee Consolation Behaviour Reflects Empathetic Personalities,” Nature Communications 8 (2017): 292.

4. Charles Darwin, The Descent of Man, and Selection in Relation to Sex (London: John Murray, 1871).

5. Robert L. Trivers, “The Evolution of Reciprocal Altruism,” Quarterly Review of Biology 46, no. 1 (1971): 35–57.

6. Soukayna Bekkali et al., “Is the Putative Mirror Neuron System Associated with Empathy? A Systematic Review and Meta-Analysis,” Neuropsychology Review 31, no. 1 (2021): 14–57.

7. Tania Singer et al., “Empathy for Pain Involves the Affective but Not Sensory Components of Pain,” Science 303, no. 5661 (2004): 1157–1162.

8. Xizheng Xu, Zhiqiang Liu, and Shaoying Gong, “The Relationship between Empathy and Attachment in Children and Adolescents: Three-Level Meta-Analyses,” International Journal of Environmental Research and Public Health 19, no. 3 (2022): 1391.

9. Hidetada Shimizu, “Japanese Cultural Psychology and Empathic Understanding: Implications for Academic and Cultural Psychology,” Ethos 28, no. 2 (2000): 224–247.

10. David Dodell-Feder and Diana I. Tamir, “Fiction Reading Has a Small Positive Impact on Social Cognition: A Meta-Analysis,” Journal of Experimental Psychology: General 147, no. 11 (2018): 1713–1727.

11. Ian C. G. Weaver et al., “Epigenetic Programming by Maternal Behavior,” Nature Neuroscience 7, no. 8 (2004): 847–854.

12. Antoine Lutz, Julie Brefczynski-Lewis, Tom Johnstone, and Richard J. Davidson, “Regulation of the Neural Circuitry of Emotion by Compassion Meditation: Effects of Meditative Expertise,” PLoS ONE 3, no. 3 (2008): e1897.

13. Helen Y. Weng et al., “Compassion Training Alters Altruism and Neural Responses to Suffering,” Psychological Science 24, no. 7 (2013): 1171–1180.

14. For a later randomized investigation, see “Effects of Compassion Training on Brain Responses to Suffering Others,” published in 2021, which reported altered medial orbitofrontal responses following a four-week compassion-meditation intervention.

15. C. Daryl Cameron and B. Keith Payne, “Escaping Affect: How Motivated Emotion Regulation Creates Insensitivity to Mass Suffering,” Journal of Personality and Social Psychology 100, no. 1 (2011): 1–15. See also Paul Slovic, “‘If I Look at the Mass I Will Never Act’: Psychic Numbing and Genocide,” Judgment and Decision Making 2, no. 2 (2007): 79–95.

16. John Suler, “The Online Disinhibition Effect,” CyberPsychology & Behavior 7, no. 3 (2004): 321–326.

17. On the changing evidence concerning oxytocin and trust, see Charlotte F. Kroll et al., “Absence of a Meaningful Effect of Intranasal Oxytocin on Trusting Behavior: A Registered Report with Pooled Equivalence Testing,” Cortex 198 (2026): 208–233. A separate preregistered 2026 study reported increased trusting behaviour among men selected for low dispositional trust, illustrating the continuing importance of population and context in interpreting oxytocin research.