Unresolved Tension
Unresolved tension is the fifth of the six ingredients in Adam Alter's model of behavioural addiction, set out in Irresistible in 2017: unfinished business occupies the mind in a way finished business does not, so a deliberate non-ending is a retention mechanic rather than a storytelling accident. It is a practitioner framework rather than a research finding. The research usually cited under it has moved, and the effect that survives is the urge to resume an interrupted task rather than a memory advantage for it.
What it is
The idea is conventionally built on the Zeigarnik effect, the claim that an open loop is stickier in memory than a closed one. Bluma Zeigarnik's 1927 experiments used interrupted laboratory tasks. The familiar waiter story, in which a waiter recalls an unpaid order in precise detail and forgets it once the bill is settled, is an anecdote attributed to Kurt Lewin's observation rather than Zeigarnik's data, and it should be told that way.
The evidence has since been gathered up. A 2025 meta-analysis by Roberto Ghibellini and Beat Meier found no general memory advantage for unfinished tasks, and a robust tendency to resume them, at about two thirds of opportunities. So the mechanic is better grounded in the pull to go back than in better recall, and describing this ingredient as built on a memory effect needs rewording.
In effect
The commercial versions are easy to see once named. Serial built a podcast format on refusing to close the loop from episode to episode. The Sopranos ended mid-scene, cutting to black rather than resolving anything. Netflix's post-play autoplay is the mechanised form: the next episode begins before the viewer has decided to watch it, so the loop is never available to be closed.
The reward-system explanation that often accompanies the ingredient should be handled carefully. The claim that anticipating an uncertain reward activates dopamine pathways more than a certain one does is a real finding, but the attribution given in Irresistible was not traced, and the canonical work is Christopher Fiorillo, Philippe Tobler and Wolfram Schultz in monkeys in 2003.
What it does not say
It does not say that cliffhangers improve memory. The meta-analytic record does not support a general memory advantage for unfinished tasks.
It does not describe a tested effect of its own. It is ingredient five of a framework, and the framework has not been tested as a framework.
It does not follow that every open loop retains an audience. Resumption is a tendency measured on laboratory tasks, and the leap from that to a viewing or listening habit is an inference rather than a finding.
Sources
- Irresistible, Adam Alter, 2017. Named as ingredient 5 of six, with the full treatment in ch. 8, "Cliffhangers", p. 191, which supplies the Serial, Sopranos and Netflix post-play examples.
- Zeigarnik, B. (1927). Experiments on interrupted laboratory tasks. The waiter story is an anecdote attributed to Kurt Lewin's observation, not Zeigarnik's data.
- Ghibellini, R., & Meier, B. (2025). "Interruption, recall and resumption: a meta-analysis of the Zeigarnik and Ovsiankina effects." Humanities and Social Sciences Communications, 12. Abstract only. No general memory advantage for unfinished tasks; a robust resumption tendency at about 67 per cent.
- The uncertain-reward dopamine claim attributed in the book to Gregory Berns was not traced. The canonical finding is Fiorillo, C. D., Tobler, P. N., & Schultz, W. (2003), in monkeys.
- Evidence status not-a-research-claim, following the verification verdict of 2026-09-25.