Triangulation psychology is not a trick somebody learned. Murray Bowen’s model holds that a two-person system is inherently unstable: it tolerates very little tension before pulling in a third person to carry the excess. The triangle is not the pathology. It is the smallest stable unit an emotional system has.
Which means the question was never whether there is a third person. There always is.
The short version
- Bowen’s triangle is automatic and universal, not a tactic. Everybody triangles.
- A pair cannot hold much tension. A third point absorbs the overflow.
- Standing in the middle has measured costs, over and above the conflict itself.
- The tell is not that a third person exists. It is that you never get to speak to them.
Bowen Did Not Describe a Weapon
The popular version of triangulation is a manipulator deliberately setting two people against each other. That is not what the concept was built to describe, and the mismatch matters.
The Bowen Center states the model plainly: a two-person system is unstable because it tolerates little tension before involving a third person. Add the third, and the tension can shift around three relationships instead of accumulating in one.
This is automatic. It is not planned, and it is not unusual. During calm stretches, two people hold comfortable insider positions while the third sits outside; when tension climbs, the positions rotate, and at high intensity the outside position becomes the one worth having. Nobody in a triangle is necessarily doing anything on purpose.
What It Costs to Be the Third Point
That the geometry is normal does not make it free.
Cheryl Buehler and Deborah Welsh followed 416 two-parent families across four annual waves, with adolescents aged 11 to 15 and reports collected from mothers, fathers and the young people themselves. Triangulation was measured with items asking, for instance, how often a parent would “try to get this child to side with you during family or marital disagreements”.
Triangulation at wave one predicted internalizing problems — anxiety, depressive symptoms, social withdrawal — three years later at wave four. It did so while controlling for marital hostility, and that control is the whole point. The damage is not merely that the parents were fighting. It is the job of standing in the middle of the fight.
Their model routes the effect through emotional reactivity. Being pulled into the middle raises how hard a young person reacts to the conflict, and that reactivity is what carries forward into anxiety and low mood. A direct path survives alongside it, which means reactivity explains part of the damage but not all of it.
So What Makes a Triangle Manipulative
If triangles are universal, then the popular question — is this triangulation? — is close to useless. Everything is.
The distinction that survives Bowen is not the geometry. It is who controls the third point. In an ordinary triangle, tension finds a third person; nobody appoints them. What the manipulation literature is reaching for is narrower: the third point is selected, its verdict is reported rather than heard, and it gets swapped whenever a different one would land harder.
“Even Sarah thinks you’re being crazy” is not manipulative because Sarah exists. It is manipulative because Sarah cannot be cross-examined. The tell is not the third person; it is that you are never permitted to talk to them directly. That unreachable witness is also where gaslighting gets its leverage, and it is a structure any account of what manipulation actually is has to make room for.
What This Does Not Establish
The limits are worth stating, because they are large. Buehler and Welsh studied two-parent families that were 91% European American, and their measure captured only the experience of feeling caught in the middle — not scapegoating, not detouring. The authors say directly that the associations were small in magnitude. This is a real effect, not a big one.
Bowen’s model, for its part, is a clinical theory rather than an experiment. Triangles are offered as a structural feature of every three-person system; the claim that they contribute to clinical problems is a claim about emotional intensity, not about anybody’s intentions.
So none of this identifies a manipulator, and none of it makes a third person suspicious. What it removes is the excuse that being put in the middle is nothing — which is, reliably, the thing said to whoever is standing there. It is also part of why people stay in toxic relationships: the third point is a position, and positions are hard to simply step out of.
Frequently Asked Questions
What is triangulation in psychology?
In Murray Bowen’s family systems model, a triangle forms when a two-person relationship pulls in a third to absorb tension it cannot hold alone. Bowen treated it as the smallest stable relationship system and as an automatic process, not a deliberate tactic used by one person against another.
Is triangulation always manipulation?
No, and treating it that way empties the word. Under Bowen’s model triangles are universal. What distinguishes a manipulative one is control of the third point: it is chosen rather than found, quoted rather than present, and replaced when a more useful third party is available.
Does being caught in the middle actually cause harm?
Buehler and Welsh found that triangulation predicted later internalizing problems in adolescents while controlling for marital hostility, so the effect is not just the conflict itself. The association was statistically reliable but small, and the sample was largely European American two-parent families.
Sources
- Triangles. The Bowen Center for the Study of the Family, Georgetown University Medical Center. Retrieved 16 July 2026.
- Buehler, C. & Welsh, D. P. (2009). A process model of adolescents’ triangulation into parents’ marital conflict: The role of emotional reactivity. Journal of Family Psychology, 23(2), 167–180. Retrieved 16 July 2026.
- Bowen, M. (1978). Family Therapy in Clinical Practice. Jason Aronson. Retrieved 16 July 2026.
