Agency and causal inference: minds default to intention

People spontaneously perceive intention and causal narrative even in simple moving shapes — shaping how coincidences get read as meaningful.

Established Supported by convergent, high-quality evidence.

Plain-language answer

Human perception does not stop at “shape moved toward shape” — it automatically supplies intention, personality, and a causal story: “the triangle is chasing the circle,” “the big shape is bullying the small one.” This happens even when the “shapes” are simple, silent, animated geometric figures with no biological features at all, and even when viewers are told in advance that the figures are meaningless. Perceiving agency and causal narrative in movement and co-occurrence appears to be an automatic, largely unavoidable feature of how minds process events, not a deliberate inference people can easily switch off.

Why it matters

This same default is doing a great deal of work whenever a coincidence feels meaningful rather than merely improbable. Two events occurring close together in time doesn’t just get logged as “these things co-occurred” — it gets automatically woven into a narrative in which one event relates to, anticipates, or was somehow “meant” alongside the other. That narrative-generating tendency is a feature of ordinary perception and cognition, present across cultures and ages, not a special mystical faculty or a sign of magical thinking.

Worked example: the chasing triangle

In the founding demonstration, observers watched a short silent film of two triangles and a circle moving around a rectangle with an opening in it. Asked to describe what they saw, almost every viewer spontaneously produced a social narrative — describing the shapes as “chasing,” “trapped,” “afraid,” or “fighting” — despite the film showing nothing but simple geometric shapes following programmed trajectories with no actual intentions, emotions, or social relationships whatsoever.(Heider & Simmel, 1944) The tendency to attribute agency was not a minority response or an artifact of suggestible participants; it was close to the default way people described what they had watched.

Applied to coincidences: when two events line up in a way that resembles a story (a warning followed by a near-miss, a thought followed by a matching event), the same narrative-generating apparatus that turned moving triangles into a drama of chasing and fear is available to turn a temporal sequence into a story of anticipation, warning, or connection — independent of whether any causal connection actually exists.

Common misconception

“I couldn’t help but feel it meant something, so it probably did” confuses the automaticity of a perceptual/cognitive process with evidence about the world. That a narrative interpretation arises easily and feels compelling is consistent with how human causal-narrative perception works in general — it does not, by itself, distinguish a coincidence that reflects an actual causal connection from one that doesn’t.

Limits and open questions

The Heider–Simmel demonstration shows that agency perception from motion and co-occurrence is close to automatic; it does not, by itself, explain the full range of individual differences in how strongly or readily different people build causal narratives out of ambiguous events, nor does it specify which real-world coincidences (if any) reflect a genuine causal link. Those remain separate, case-by-case questions.

Key takeaways

  • People automatically perceive intention, personality, and causal narrative in the mere movement and co-occurrence of simple shapes, with no biological cues present at all.
  • This narrative-generating tendency is a normal, largely automatic feature of perception and cognition — not evidence of a special faculty or a cognitive error confined to “superstitious” thinking.
  • Feeling that a coincidence “meant something” reflects how compellingly the mind builds causal stories, and is not, on its own, evidence that a real causal connection exists.

Sources

  1. Heider, Simmel (1944). An Experimental Study of Apparent Behavior. The American Journal of Psychology, 57(2), 243-259. https://doi.org/10.2307/1416950 ↩