ancient origin of counting sheep to fall asleep

According to Disciplina Clericalis, a text written in early twelfth-century Spain from Islamic sources, a king every night heard stories from his storyteller. One night, the king, burdened with worries from the day’s business, did not feel like going to sleep. He demanded extra stories from his storyteller. But the storyteller himself wanted to go to sleep. The storyteller’s ingenious solution was to tell a story that required counting sheep.[1]

A farmer went to market and bought two thousand sheep. Returning home, he found his way blocked by a flood-swollen river. Along the shore was a small boat that could carry only two sheep across at a time. The farmer put two sheep into the boat and crossed over. The farmer needed to do that a thousand times in order to get all his sheep home.

According to Disciplina Clericalis, the storyteller fell asleep after stating that the farmer put the first two sheep into the boat. The king woke the storyteller and demanded that he continue. The storyteller responded that the story required the farmer to transport all the sheep across the river.[2] This meta-story clearly depends on common understanding of the practice of counting sheep to fall asleep. The practice of counting sheep to fall asleep thus must have been well-known prior to the early twelfth century.

The early seventeenth-century Spanish text Don Quixote reworked the frame story for counting sheep. Traveling at night, Don Quixote and his squire Sancho heard the roaring of water and loud, frightening, rhythmic banging of fetters and chains. Deluded with romantic fantasies of chivalry, Don Quixote was determined to risk death by approaching the noise. Sancho forestalled that action by hobbling Quixote’s horse. Quixote reluctantly resolved to wait vigilantly until dawn. Sancho helpfully told Quixote:

I’ll entertain your grace by telling stories the whole time, unless you want to dismount and stretch out for a little sleep, here on the green grass, the way knights errant do, so you can be better rested when day comes, and more fit for this unheard-of adventure that awaits you. [3]

Quixote angrily responded:

Am I, by any chance, one of those knights who look for rest when danger faces them? You sleep, since that’s what you were born for, or do whatever you want to, and I will do what best suits me.

Sancho began to tell a long-winded variant of the sheep story in Disciplina Clericalis. In Sancho’s version, a shepherd had three hundred goats that he had to transport across a river. The shepherd had to take the goats across one by one. Sancho explained:

you’d better keep track of how many goats the shepherd carries across, your grace, because if we forget a single one that will be the end of the story, and it won’t be possible to tell another word.

Quixote urged Sancho to assume all the goats were carried across, and get on with the story. Sancho then asked Quixote how many goats had already been carried across. Quixote didn’t know. Sancho then declared that the story had ended. Quixote responded:

you’ve told one of the most novel tales, or stories, or histories, anyone in the world has ever thought of, and the way you told it, and then ended it, is something never to be seen, and never ever seen, in the course of a lifetime, though I expected nothing less from your remarkable powers of reasoning. On the other hand, I’m not surprised, for conceivably this banging, which has never stopped, has troubled your brain.

That’s layers of nonsense built upon ridiculousness. Disciplina Clericalis was a highly popular work across Europe. Spanish readers most likely knew the tale of counting sheep. About five centuries after Disciplina Clericalis was written, the counting-sheep story had a novel ending in Don Quixote. That novel ending was in counting goat-sheep not being allowed to produce sleep.

when counting sheep, count this one

The animating spirit of Don Quixote’s evaluation of counting sheep continues in recent scientific work and associated story-telling. On January 24, 2002, news sources around the world reported news about counting sheep. In Britain, a story in The Guardian declared in its headline: “Trouble Sleeping? Don’t Count on Sheep.” The story reported:

New Scientist reports today that Allison Harvey, a cognitive psychologist at Oxford University, tested that classic recipe for numbing thought and quelling anxiety: counting sheep. She and a colleague divided 50 volunteer insomniacs into three groups, proposed a strategy for each and monitored the rates at which eyelids closed and breathing became regular.

One group was asked to concentrate on a distraction such as counting Southdown ewes in a field, or Merino lambs hopping over a stile. One group was left to its own devices. And one was asked to focus on a tranquil and relaxing suite of thoughts, such as a waterfall, or being on holiday.

Those who imagined torpid afternoons in the south of France, or lazy twilights in the Tyrol, went to sleep on average 20 minutes earlier than they would normally do on nights when they were not concentrating on faraway places. The sheep counters – and the ones who just lay there, wishing they could nod off – actually stayed awake for longer than their normal ration of restlessness.

“Counting sheep is just too mundane to effectively keep worries away,” Dr Harvey said.

The story in BBC News was titled “Sheep counting is tired technique.” That story reported:

The idea that you can nod off while imagining the woolly animals jumping through a hedge has been around for years, but scientists who have tested it on volunteers say other strategies are likely to be more effective.

Thinking about a calming waterfall or a tranquil beach was more likely to induce sleep, Allison Harvey, from Oxford University, UK, told New Scientist magazine. …

Harvey and a colleague took 50 insomniacs and asked them to use different techniques to get off to sleep.

Some were told to count those sheep; others to imagine a relaxing scene; and a third group was left to its own devices.

On average, those picturing a calming scene fell asleep more than 20 minutes earlier than on nights they did not try the technique. But both the sheep-counters and the controls took slightly longer than normal to fall asleep on the nights of the experiment. “Counting sheep is just too mundane to effectively keep worries away,” Harvey said.

The source article in the New Scientist, published the previous day, was entitled “Sleep scientists discount sheep.” The New Scientist article reported:

Harvey and her colleague Suzanna Payne asked 50 insomniacs to try different distraction techniques on certain nights, to see which helped them fall asleep more quickly. One group conjured up a tranquil and relaxing scene such as a waterfall or being on holiday, while a second were asked to think of a distraction such as counting sheep. A third group were left to their own devices.

On average, those picturing a relaxing scene fell asleep over 20 minutes earlier than on nights they didn’t try the technique. But both the sheep-counters and the controls took slightly longer than normal to fall asleep on the nights of the experiment. “Counting sheep is just too mundane to effectively keep worries away,” says Harvey.

The scholarly article that these three news articles referenced doesn’t actually mention sheep. The scholarly article states:

The “general distraction” group were told that during the pre-sleep period they should simply distract from thoughts, worries and concerns. No guidance was given as to a specific strategy that should be used to distract. …

Participants in the “general distraction” group thought through events that happened today (n=6), counted (n=2), meditated (n=2), subvocally hummed a favourite tune (n=1), blanked their mind whenever an unpleasant thought occurred (n=2), or focused on body relaxation (n=1). [4]

The “general distraction” group is the group that the news sources reported as “sheep-counters.” Yet only two of the fourteen tested insomniacs in that group counted, and those two did not necessarily count sheep. If you literally believed what was reported in the scientific news articles in The Guardian, BBC News, and the New Scientist, you might as well believe what is written in medieval chivalric romances.[5]

Medieval history is more meaningful than contemporary science journalism. Counting sheep for falling asleep has the authority of being, for at least a millennium, a widely recognized human practice. If you’re having trouble falling asleep, try counting sheep. That would be low-cost scientific work potentially offering significant results for you.

* * * * *

Read more:


[1] Petro Alfonsi, Disciplina Clericus, Sec. XII, from Latin trans. Hermes & Quarrie (1977) pp. 123-4. In Disciplina Clericus, the story of the king and the storyteller itself occurs within a frame of a pupil asking a master-teacher for stories about women’s guile.

[2] A similar story exists in the Cento Novelle Antiche, an Italian compilation of short stories from the end of the 13th century. In the Cento Novelle Antiche version, the storyteller urges the patron to imagine the sheep crossing “so that in the meantime you could sleep well at ease.” Novella XXXI. For the Italian text, see Novelle italiane dalle origini al cinquecento, a cura di Goffredo Bellonci, pp. 9-10. An English translation is available at Elfinspell, mislabeled correctly labeled as “Novella XXX” for its source text. Different source texts have different numberings.

[3] Miguel de Cervantes, The Ingenious Gentleman Don Quixote of La Mancha, vol. 1, ch. 20, from Spanish trans. Raffel (1999) p. 112. Subsequent quotes from Don Quixote are from id., pp. 112, 114, and 114. A version of the story exists in Avellaneda’s continuation of Don Quixote. There the connection of the story to falling asleep is more distanced. See Yardley (1794) Bk. 3, Ch. V, p. 84.

[4] Harvey & Payne (2002) pp. 270-1.

[5] The common text across all three news articles suggests a lightly rewritten press release. The medieval stories of counting sheep show more creative story-telling. All three modern scientific-journalism articles refer to 50 insomniacs. The final sample studied actually consisted of 41 insomniacs. Harvey & Payne (2002) p. 271. That’s undoubtedly a trivial number relative to the total number of persons who have tried counting sheep to fall asleep throughout history.

[image] Lundy Sheep. Photograph thanks to Michael Maggs, Wikimedia Commons.


Harvey, Allison G., and Suzanna Payne. 2002. “The management of unwanted pre-sleep thoughts in insomnia: distraction with imagery versus general distraction.” Behaviour Research and Therapy. 40 (3): 267-277.

Hermes, Eberhard and P. R. Quarrie, ed. and trans. 1977. Petrus Alfonsi. The Disciplina clericalis of Petrus Alfonsi. Berkeley: University of California Press.

Raffel, Burton, trans. and Diana de Armas Wilson, ed.. 1999. Miguel de Cervantes Saavedra. Don Quijote. New York: W.W. Norton & Company.

Yardley, William Augustus, trans. 1784. Alonso Fernández de Avellaneda. A continuation of the history and adventures of the renowned Don Quixote de la Mancha. London: Printed for Harrison and Co.

9 thoughts on “ancient origin of counting sheep to fall asleep”

  1. What an excellent article!

    Just to correct the footnote regarding the mislabeled Story XXX of the translation of the “Cento Novelle Antiche”, that is how the story is numbered by Roscoe in his text and may well be right, depending on the edition used. The first two printed editions of this collection in the 16th century, one by Gualteruzzi (the original edition, “Le Ciento Novelle antike”) and one by Borghini, “Libro di Novelle et di bel Parlar Gentile”, which is the one used by Roscoe, was printed 47 years later. The stories were not the same in each, although both had certain stories in common. I suspect the numbering was different. There are only 8 manuscripts of the text, and all differ in various respects. See the excellent Introduction by Edward Storer in his translation:

    So the numbering depends on the edition or manuscript used and unless you have access to Borghini’s text then who is to say that that is not the number used there? If you have checked that version and found it is mislabeled, let me know and I will add a note on the page stating that Roscoe misnumbered that tale.

    There are many typos on my site, without a doubt, some through the hand transcription and coding process by me, some due to errors in the text being transcribed (although I try to emend these), but this is not one of them.

    Here is Storer’s translation of this tale, where it is numbered XXXI:

    I look forward to reading more of your blog,


  2. Count sheep! (to help one go to sleep) is an idiom formed by phono-semantic matching (PSM). In this case, a Latin source phrase was transliterated to a Hebrew pun which was translated to other languages. It is indeed ancient and may have occurred as early as the Roman occupation of the Middle East.

    Idioms formed via PSM form an interesting class. Most of them are foreign words or expressions transliterated to common words of the target language which retain the semantics of the source. As such, they represent disguised code-switching, usually so well disguised that the speaker / writer and hearer / reader (and, it seems, some linguists 🙂 do not realize that a switch has occurred. English examples: raining cats and dogs, let the cat out of the bag, left holding the bag, kick the bucket, and many others.

    Biblical idioms have been translated to many languages. (Escape) by/with the skin of my teeth (Job 19:20) means “barely, hardly, with difficulty” because B’3oR SHiNai בְּעוֹר שִׁנָּי was a near homonym of B’QoSHi בקושי when the aiyin had a ​​velar G/K-sound​ as in 3aZa עַזָה = Gaza.

    You can download these files to see my thoughts about idiom formation via PSM.
    ​ ​
    This method of formation explains why PSM idioms tend to be relatively inflexible. Both the word divisions and parts of speech in the source and the idiom may not match.

    In some languages, some idioms are formed by reversal.
    This occurs in both Hebrew and English.

    In the story about Lot’s wife (Genesis 19:26), the words “looked back from behind him” are a not-so-subtle hint that we must read N’TZiV MeLakH נְצִיב מֶלַח (pillar of salt) backwards to understand that Lot’s wife suffered a stroke or thrombosis (modern Hebrew shin-bet-tzadi שָׁבָץ , literally, caused by mud). She became mud-struck.

    In the English idiom “conniption fit” (a panic attack), CoNNiP[tion] is a reversal of PaNiC and FiT is a reversal of TiFF which is a reversal of FighT.

    You may use this material in any way you think appropriate and may forward it to anyone you think would enjoy it.

Leave a Reply

Your email address will not be published. Required fields are marked *

Current month ye@r day *