Why cliffhangers keep you reading

What is the Zeigarnik effect?

In experiments published in 1927, the psychologist Bluma Zeigarnik gave people a series of small tasks, between 18 and 22 each, and interrupted half of them before they were finished. Asked afterwards which tasks they remembered, her first 32 participants recalled the interrupted ones about 1.9 times as often as the completed ones. The study as a whole involved 164 individual participants plus two group experiments.

Her own figures show how variable the effect was. Among those first 32 people, the advantage for unfinished tasks ranged from 500% for one participant to a 25% disadvantage for another. Even so, the finding became a staple of popular psychology, usually as the explanation for why an unresolved story nags at you.

Does the evidence hold up?

Not as a rule about memory. As early as 1964, Butterfield's review concluded that the effect was far from invariable and that people frequently recalled more completed tasks than interrupted ones.

In 2025 Ghibellini and Meier pooled 59 publications. Across 38 that tested recall, interrupted tasks were remembered about as well as finished ones: the ratio came out at 0.99. Where a standardised effect size could be calculated, in 8 publications, it was small (dz = 0.15). The authors also note that Zeigarnik's own way of averaging inflated her result.

In 1928 Maria Ovsiankina, using a similar approach, studied something different: whether people go back to an interrupted task when they are free to. That tendency held up. Across 21 publications, interrupted tasks were resumed about 67% of the time, well above the 50% expected by chance. The authors conclude that the urge to resume is a general tendency while the memory advantage lacks universal validity.

The curiosity studies are the most useful for learners. In Gruber, Gelman and Ranganath's experiment, people remembered 45.9% of trivia answers they had been very curious about a day later, against 28.1% for answers they cared little about. Faces shown while they waited for an answer were remembered slightly better too, though at a one-day delay only for the faces people recognised confidently. Kang and colleagues found something similar with surprising answers after one to two weeks, in a memory test of just 12 people.

Where did serialized fiction and the cliffhanger come from?

The oldest famous example is a frame story. In the Thousand and One Nights, Scheherazade begins a tale for the king each night but does not end it, and the unfinished story keeps her alive until the next.

Serial publishing made the technique commercial. Charles Dickens's Pickwick Papers came out in 20 monthly shilling parts from April 1836 to November 1837. The binder prepared 400 copies of the first part; sales of the later parts reached around 40,000. The Old Curiosity Shop followed in weekly parts in 1840–41, and with it the best-known story about serial readers: crowds on a New York wharf shouting to an arriving ship to ask whether Little Nell was dead. It is often repeated, but no contemporary evidence for it has been found, and it is better treated as legend.

Magazines did the same for detective fiction. The Strand Magazine sold over 300,000 copies of its first issue in 1891, and its first Sherlock Holmes story, A Scandal in Bohemia, appeared that July. When The Hound of the Baskervilles was serialised in 1901, it produced some of the magazine's highest circulation figures.

The word itself is younger than the technique. Thomas Hardy's A Pair of Blue Eyes, serialised in Tinsley's Magazine in 1872–73, leaves a character hanging from a cliff at the end of an instalment and is often named as the origin. The term, though, belongs to the film serials of the 1930s: the trade paper Variety used cliff-hanging in 1930 and cliff hangers in 1931.

What do cliffhangers do for language learners?

Their job is to bring you back, because coming back is what does the teaching. A reader who returns for the next chapter meets the same characters, places and words again, the repetition that our pages on learning with stories and narrow reading show makes text easier and words stick. Curiosity may help the memory side too, though the studies so far are small and none used language learners.

Every Contos chapter ends on a cliffhanger. Across the 1,534 free chapters in the Contos library (October 2026), in all 16 languages, 26% end on a question. Others stop mid-action. Three Spanish endings, at A1 and A2:

The research suggests two honest ways to use this. Stop at the end of a chapter rather than in the middle of a dull page, so the pull you feel is towards the story and not away from it. And don't rely on suspense alone: Wirz's viewers enjoyed cliffhangers no more than endings without them, so a fixed daily reading time does more to keep you going than any single ending.