The London borough of Westminster, home to the Houses of Parliament and long associated with power, order and continuity, was an unlikely setting for a deeply disturbing death in March.

An 18-day-old baby girl, Mariam, died after falling from a third-floor window of a family flat. Her mother, Zahira Byjaouane, has been charged with her murder. The court heard that the child’s father had been in the kitchen preparing formula. Emergency services arrived within minutes, but the infant died shortly afterward from a severe head injury.

There is nothing outwardly extraordinary about the setting: a central London apartment, a baby, the routines of early parenthood. Yet the allegation that a mother may have caused the death of her own newborn jars sharply with one of the most deeply held assumptions in any society.

Why do mothers kill their children? The question is unsettling. It feels almost forbidden, as if even asking it risks normalizing what should remain inconceivable. The prevailing assumption is that the bond between mother and child is not merely strong, but elemental: a natural attachment that precedes social influence and resists external pressure.

And yet cases such as this prompt a difficult question. If maternal love is as instinctive and unbreakable as commonly assumed, how can it fracture? The most familiar answer is also the most reassuring. Faced with behavior of this kind, explanation usually turns to mental disturbance: breakdown, disorder or some form of psychological impairment sufficient to account for the rupture. In some instances, such explanations are relevant. But they also serve a wider function by allowing the assumption of an enduring maternal instinct to remain largely intact, disturbed only in exceptional cases.

A mother’s love

The question becomes more — not less — difficult when we consider how often it has had to be asked. Modern cases suggest not a single anomaly but a recurring, if rare, phenomenon. In England, women such as Rachel Tunstill concealed pregnancies and killed their newborns shortly after birth. Others, like teenager Paris Mayo, gave birth alone and in secrecy before acting in panic and isolation. In the United States, Megan Huntsman killed multiple infants over a period of years, each birth hidden, each death initially undiscovered.

And then there are cases that resist comprehension, such as that of Susan Smith, who, in 1994, drowned her two young sons, not in panic, but in what appeared to be a thought-out attempt to remove an obstacle to a desired relationship. In 2020, in England’s West Bromwich, the world got a glimpse into what happens when a mother, unconstrained by conscience and unmitigated by compassion, vents her emotions on her two-month-old baby and beats the girl to death, fracturing her skull, ribs, legs and other body parts.

All these killings differ profoundly in motive and circumstance: Panic, concealment, calculation, prolonged abuse. Yet they share a disquieting feature — the apparent breakdown, absence or displacement of the attachment that’s assumed to bind parent to child.

Infanticide is not new. It appears in legend as well as in modern courtrooms. In the Greek myth of Medea, a mother kills her children in an act of revenge against their father. The story has endured not simply because of its extremity, but because it violates what is widely taken to be a basic human instinct.

One response is to treat such cases as rare breakdowns of an otherwise universal maternal bond. Mothers love their children; exceptions prove the rule. But the persistence and variety of these cases suggest a more unnerving possibility. Perhaps maternal attachment is not as fixed or automatic as we assume, but more contingent and variable. It may develop through proximity, recognition, repetition and reinforcement, as reflected in the routine practice of placing newborns immediately in their mother’s arms to encourage early bonding. Under certain conditions, it may fail to take hold at all.

This doesn’t deny that the bond between parent and child is usually deep and enduring. It suggests that its strength may derive less from biology than from the social conditions under which it is formed. If so, the question changes. It is no longer simply why some mothers kill their children.

The natural and the social

We often speak of actions or emotions as coming “naturally,” as if they require no instruction or social context. The phrase implies ease and inevitability. But closer inspection tells a different story.

What appears natural is often the result of a long and largely invisible process of learning. Language is the obvious example. It feels instinctive, almost automatic, yet it’s acquired through years of immersion, imitation and repetition. The same is true of manners, moral judgments, emotional responses and, crucially, attachments.

Human beings are not born knowing whom to love or how to love, or even how to respond to love. These capacities develop through socialization: an ongoing engagement with others such as parents, family members, partners, and with the physical environments in which life proceeds. Through this process, individuals absorb patterns of behavior, internalize expectations and develop bonds that come to feel natural and indispensable.

Over time, habits harden into dispositions. Emotional responses become almost reflexive, relationships deepen, strain, break and reform. A sense of self emerges not in isolation, but through constant interaction with others. Because this process is continuous, its outcomes can appear innate.

Maternal attachment is often understood in precisely these terms: as a biological instinct that emerges automatically. In many cases, something like this does seem to occur. Bonds can forge quickly and, once established, are often resilient and so capable of enduring repeated strain, as when parents persist with a substance-dependent child whose cycles of relapse test but rarely sever attachment.

Yet this resilience raises a difficult question. Is there any stronger attachment than that between a mother and her child? The question seems almost rhetorical. That is, until we consider situations in which that bond is displaced by more immediate pressures. Accounts of severe dependency, for example, describe parents whose need for a substance reshapes judgment and priority to such an extent that obligations which would ordinarily be characteristic are neglected. These cases are extreme, but they make a relevant point. Attachment is not simply a matter of feeling, but of what, in a given moment, takes precedence.

Social bonds

If attachments are shaped rather than simply biologically triggered, a different possibility surfaces. The maternal bond may depend, at least in part, on conditions that can be weakened, disrupted or never fully established. Proximity, recognition, stability and support are not incidental to attachment; they help constitute it.

In a different context, in 1969, the criminologist Travis Hirschi argued in his Social Bond Theory that strong social bonds restrain harmful behavior. Individuals refrain from certain actions not only because they are irrational or pathological, but because they are tied emotionally and socially to others. When those ties weaken, the restraints they provide can loosen.

Hirschi developed this framework to explain more ordinary forms of deviance, and it can’t be directly mapped onto cases of infanticide. But the underlying insight is suggestive: If attachment is contingent rather than automatic, then its failure — however rare — becomes more intelligible.

The infanticide taboo

The question returns, now in altered form: Why do so many mothers kill their children? Not simply because they lack an instinct that others possess, but because the processes through which attachment is formed and sustained have, in certain cases, broken down. Sometimes it was gradual, sometimes it was abrupt and sometimes it happened under pressures that remain largely invisible.

Across societies, the overwhelming majority of parents form enduring attachments to their children. But the strength of these ties lies not only in their apparent naturalness, but in the social processes that sustain them. From early life, individuals are embedded in networks of expectation and obligation. Families, communities and institutions reinforce the idea that children are to be protected and cared for. Over time, these expectations are internalized, becoming part of what individuals experience as their own commitments.

Powerful prohibitions also exist. The killing of a child is not simply illegal; it is one of the most deeply entrenched taboos in any culture. Its force lies not only in punishment, but in the moral revulsion it provokes. It operates as a brake on behavior that most people never countenance.

Yet adherence to norms is not absolute. As the sociologist David Matza observed, individuals can “drift” in and out of alignment with moral frameworks. Under conditions of pressure, isolation or desperation, the constraints that ordinarily guide behavior may weaken. Actions that once seemed unthinkable can become, if only momentarily, possible and even probable.

Seen in this light, the rarity of infanticide does not confirm that maternal attachment is unbreakable. It suggests that the conditions under which it is formed and sustained usually hold. When they do not, through secrecy, strain, dislocation or the absence of meaningful attachment, the result, though mercifully rare, moves from the incomprehensible to the explicable.

[Ellis Cashmore is the author of The Destruction and Creation of Michael Jackson.]

[Lee Thompson-Kolar edited this piece.]

The views expressed in this article are the author’s own and do not necessarily reflect Fair Observer’s editorial policy.