A king can't sleep, so he demands extra stories from his exhausted storyteller. The storyteller obliges with a tale about a farmer who has just bought 2,000 sheep and needs to ferry them across a flood-swollen river in a tiny boat that fits two at a time — meaning 1,000 crossings. The storyteller describes the first crossing, then falls asleep. When the king shakes him awake, the storyteller says the farmer still has to transport the rest of the sheep. Let him finish before you ask for another story.
That story appears in the Disciplina Clericalis, a 12th-century Spanish collection drawn from Islamic sources, which means the sheep-counting-as-sleep-aid concept was already well established before anyone wrote it down, as Purple Motes documented. Cervantes reworked the same gag about 500 years later in Don Quixote — Sancho tells his master a version with 300 goats crossing a river one at a time and insists Don Quixote track each goat or the story ends.
The modern scientific take is less flattering to the sheep. Oxford researcher Allison Harvey tested 50 insomniacs in 2002 and found that people who imagined calming scenes — waterfalls, beaches — fell asleep more than 20 minutes faster than usual. The sheep-counters actually took longer than normal. Harvey's conclusion: "Counting sheep is just too mundane to effectively keep worries away." Journalists widely reported the study as proof that counting sheep doesn't work, but Harvey's paper never actually tested sheep-counting specifically. Her "counting" group had only 2 of 14 participants who counted at all, and they didn't necessarily count sheep.
Previously:
