Siblings, alliances and coalitions: the invisible geometry of a family

Two adult sisters come to a session, forty-five and forty-eight years old. The conversation starts with a practical matter — who will look after their mother, who can no longer manage alone after a fall. Within fifteen minutes they are no longer discussing a visiting rota but a sentence from 1991: “She always let you do whatever you wanted.”

Anyone with a sibling knows how quickly this happens. The sibling relationship has an extraordinarily long memory and the peculiar quality of reverting, around certain topics, to the shape it had when you were ten. This is not a sign of immaturity. It is a sign that this is the only relationship that lasts longer than any other in our lives — and that it took shape at a time when none of those involved had any say in what it would look like.

In this piece I explain how the geometry of alliances and coalitions forms in a family, what triangles are and why a child so often falls into them, why two siblings from the same family often grow up as people from two different worlds, what happens to comparison in adulthood, and where the old configurations reawaken most reliably — in caring for ageing parents.

The longest relationship you will have

If you have a brother or sister, that is statistically the longest relationship of your life. You usually outlive your parents, you meet a partner later, you have children later still. Your sibling was there before them and will most likely be there after them.

The psychologist Victor Cicirelli, who researched sibling relationships across the whole lifespan, pointed to their particular mixture: siblings are at once equals and unequals, at once allies and rivals, at once the most familiar and sometimes the most foreign. Nobody else witnessed your childhood from the inside. And that is precisely why a sibling has a power others don’t — they can confirm your memory or erase it with a single sentence: “That’s not how it happened at all.”

Salvador Minuchin described the sibling subsystem as the first social laboratory in life. In it a child learns to negotiate, to share, to lose, to form alliances, to defend themselves and to forgive — all without adult supervision and with very fast feedback. When this subsystem works well, it is one of the strongest protective factors in childhood. But when adult tensions break into it, the laboratory turns into a battlefield.

Triangles: why tension between two always seeks a third

Murray Bowen noticed something that has since become one of the central insights of family therapy: a relationship of two is unstable, a relationship of three is stable. When tension between two people exceeds what they can bear, they relieve it almost automatically by drawing in a third. This is called triangulation.

A triangle is not in itself pathology — every family has countless ones and most are harmless. The trouble starts when a triangle becomes permanent and when the third party is a child. Then the tension between the parents isn’t resolved, only relocated. The parents keep their balance, and the child is given a job they never asked for and are not equipped to do.

You will probably recognise all the forms this takes:

• The child as go-between. “Tell your father that…” Messages travel through the child, who becomes the family switchboard — and gains a view of things that overwhelm them.
• The child as confidant. One parent tells the child about their difficulties with the other parent. The child feels important and chosen — and at the same time betrays the other parent, because they are hearing things that put them on one side.
• The child as the reason. The parents stay together “for the children” and redirect their tension into worry about them. The child feels this, even if it is never spoken about.
• The child as arbiter. When the parents argue, both turn to the child for confirmation. Whoever has the child on their side wins.

The crucial distinction is between an alliance and a coalition. An alliance is closeness between two that harms nobody — mother and daughter enjoy cooking together, father and son go to matches. A coalition is closeness between two against a third. And a coalition across the generational line — parent and child against the other parent — is one of the most burdening configurations a child can be given.

When a child becomes a parent to their parents

The extreme form of this geometry has its own name: parentification. The concept was brought into family therapy by Iván Böszörményi-Nagy and worked out in detail by Gregory Jurkovic. It is a reversed arrangement in which a child takes on tasks that an adult should be doing.

There are two forms. Instrumental parentification means practical tasks — the child cooks, looks after younger siblings, deals with bills, translates for the parents, accompanies a sick parent. Emotional parentification is quieter and often more burdening: the child becomes the one who comforts their mother, regulates their father’s mood, makes sure no argument breaks out at home, and takes the family’s temperature all the time.

A degree of responsibility does a child good and strengthens them. Harm arises when the tasks are constant, out of proportion to the child’s age and — most importantly — unacknowledged. A child who carries a lot and receives gratitude and recognition for it also takes competence away from the experience. A child who carries a lot while it is taken for granted takes away above all the feeling that they are worth only as much as they are useful.

A meta-analysis by Lisa Hooper and colleagues showed that reported childhood parentification is associated with a higher incidence of mental health difficulties in adulthood — with the effect being moderate and in no way a matter of fate. Many a person who was a parentified child is extremely reliable, empathic and capable as an adult. The cost often shows elsewhere: in not knowing what it is like for someone to look after them for once.

In a family with several children, parentification almost always falls on one — usually the eldest or the most sensitive. And this is exactly where the resentment siblings carry for decades is born: one carried, the other was protected, and neither of them chose it. I have written more about the roles children are given in families in the piece on roles in the family.

Why the same family produced two different people

This is one of the most surprising findings of behavioural genetics and it explains a great deal in the therapy room. In the mid-eighties Robert Plomin and Denise Daniels posed a question that sounded naive: why are children from the same family so different? The answer that twin and adoption studies led to was radical. The environment really does shape personality — but it is mostly the non-shared environment, that is, what is particular to each child, not what they share.

A shared house, shared parents, shared income and a shared village explain surprisingly little. Far more is explained by what was different for each child: what your mother was like when you were born and what she was like three years later; whether your father was unemployed at the time or being promoted; what place you were given among the siblings; how a teacher responded to you; what temperament you had and what response that temperament provoked in your parents. Erik Turkheimer and Mary Waldron later showed that these individual factors are very hard to measure — but that the effect of the non-shared environment is nonetheless undeniable.

From this follows a sentence I often use in therapy and which often takes a large part of the burden off siblings: every child has their own parents. Not metaphorically — literally. The first-born got inexperienced, frightened, perhaps stricter parents. The youngest got tired, more permissive, perhaps also more disappointed ones. The same person, a different version.

That is why an argument between two adult siblings about “what our father was really like” is an argument that cannot be won. Both are right. Each is describing the father they had.

The same picture, two geometries

The difference between the two arrangements is easiest to see drawn out. On the left is a family in which the adults’ tension stays between the adults. On the right is the same trio, in which the tension has spilled across the generational line — and in which the child has a job that isn’t theirs.

What we really know about birth order

Since we are on the subject of differences between siblings: more claims circulate about birth order than the research will bear. The popular account says that first-borns are responsible and conscientious, middle children are negotiators, and the youngest are rebellious and charming.

The largest study on the topic to date, published in PNAS in 2015 by Julia Rohrer, Boris Egloff and Stefan Schmukle on the basis of data from more than twenty thousand people in three countries, showed something more sober. There is practically no systematic effect of birth order on personality traits — emotional stability, extraversion, conscientiousness, agreeableness. The only difference that could be reliably confirmed was a small advantage for first-borns on intelligence tests and in self-rated intellectual ability.

This is useful to know for two reasons. First, because it relieves: if you don’t recognise yourself in your “prescribed” role, there is nothing wrong with you. And second, because it redirects attention to where things are genuinely interesting. Whether you were born first or third is unimportant. What matters a great deal is what place you were given in this particular family — who was the sensible one, who the difficult one, who was mother’s and who father’s, who carried and who was carried. Birth order is only one of the factors that shaped that place, and often not the strongest.

Comparison that doesn’t end at eighteen

Many people expect jealousy to stay in childhood. It doesn’t. It merely changes form: from toys it moves into grades, from grades into career choice, from career into the success of the children, the house, the marriage.

The work of Karl Pillemer and J. Jill Suitor, who studied differences in parents’ relationships with individual children in large samples of American families, is particularly interesting here. Their findings are consistent on two points. First: differentiation is common — most mothers can say which child they are emotionally closest to and which one they would turn to first in trouble. Second, and more importantly: what is decisive for siblings’ psychological wellbeing is not who was actually favoured, but the perception of differential treatment. Even those supposedly privileged feel worse — the favoured child often carries guilt and doubt about whether they are loved for themselves or for the role.

In the therapy room this comes out in sentences that seem trivial and that people preface with an apology: “I know it’s childish, but…” It isn’t childish. It is the most fundamental question of all: was I enough? And because that question is addressed to a parent, siblings cannot settle it between themselves — but they can stop sending each other a bill that isn’t theirs.

Caring for ageing parents: where everything reawakens

There is a period in which the old family geometry activates almost without fail: when parents fall ill or become frail. Within a few weeks everything that has been under the lid for three decades comes to the surface.

Why then? Because all the old questions are suddenly back on the table at once: who is responsible, who was closer, who “deserved” more, who left and who stayed. Practical decisions — a care home, nursing, inheritance, who takes them to the doctor — become carriers of old meanings. An argument about a visiting rota is rarely an argument about a rota.

The old patterns of distribution repeat too. The one who was a parentified child takes on the most — and feels alone again in doing it. The one who was protected holds back — and feels guilty. The one who was in a coalition with one parent still has a special bond with them that the others don’t understand. And a parent who needs help sometimes — often unconsciously — goes on setting the children against one another, because that was always how they stayed at the centre.

The good news is that this is also one of the most rewarding periods for family therapy. Adult siblings have something children don’t: the possibility of coming to an agreement. And it is often enough for someone in the room to say out loud what each of them has been thinking privately for thirty years.

What can be done with this geometry

The geometry cannot be erased, but it can be seen — and with that it loses much of its power. In practice the work runs in several directions.

• Naming the triangle. When an adult says for the first time “I was my mother’s confidant and my father’s rival”, a great deal in their story falls into place.
• Returning burdens to their owners. A child who carried a parent’s distress symbolically sets it down in therapy — not by accusing, but by acknowledging that it was never theirs.
• Separating the relationship with a sibling from the relationship with the parents. Very often siblings are not settling accounts with each other but, through each other, with a parent.
• Working on differentiation. The ability to stay yourself and in contact at the same time is exactly what gets you out of coalitions — I write about this in the piece on differentiation.
• Drawing a genogram. Triangles, coalitions and cut-off contacts show up on paper within minutes, even across three generations. What is passed down the generations along this route I have described in the piece on emotional inheritance.

Work with siblings, or with adult children and parents, takes place within family therapy; I am a therapist in training in marriage and family therapy and I work under regular supervision. Not everyone has to come — it often begins with a single person who wants to understand why they always feel ten years old around their brother. If you recognise yourself in this, you can write to me.

And if you would like the broader frame — how it is that one member’s shift moves everyone else — you will find it in the piece on the family as a system.