A friend recently pinged me about a claim about Turkish coffee reading that has been making the rounds (most recently as a TIL post). Supposedly, kahve falı, the practice of flipping your drained cup onto the saucer and reading the dried grounds, began in the Ottoman harem about five hundred years ago. The idea is that it offers women a licensed way to gossip, vent, and raise taboo subjects without consequences.
I’ve written before about how harem women were political actors, and that rituals which look like superstition often do concrete social work, so he was right that it would interest me. Unfortunately, I think a lot of details in the claim are at least unsubstantiated. But since I took the time to dig into why it’s not something I’m willing to totally buy, I thought y’all might be interested in what I dug up. Basically, I think the timeline is roughly right, but the harem origin rests entirely on modern speculation, based on the real social-loophole function of these sorts of fortune-telling rituals as they appear in the modern day.
On the timeline of divination from cups
Coffee seems to have first been brewed as a hot drink in Yemen in the fourteenth or fifteenth century, largely in Sufi devotional circles. It spread into the Ottoman eastern Mediterranean during the sixteenth century. The Ottomans absorbed coffee-drinking Egypt in 1517, and Yemen itself under Süleyman in 1538. Istanbul’s first coffeehouses opened in the mid-1550s; they were founded by Syrian merchants in Tahtakale. So the “500-year-old tradition” from the Reddit thread rounds things up a bit, but, fine, whatever.
Divination more generally, tho, is much older than coffee in the Islamic world. Fal covers a whole family of techniques: bibliomancy from the Quran or from the poetry of Hafez, dream interpretation, geomancy, astrology, and such. Adding divination tables to the end of Qurans became widespread in the sixteenth century, so an Ottoman coffee drinker of the 1550s may already have been used to asking a text or an object about the future. The Ottoman and Safavid courts commissioned monumental illustrated falnama, aka “books of omens”, and the palace employed a chief astrologer. Letter magic and talismans pervaded sixteenth-century Ottoman society at every level, from the palace elite down to the Sufi lodges. Jurists condemned the divinatory arts, although the elites at court patronized them, because humans cannot get away from conservative/libertine cultural conflicts I guess.
Of course in a lot of places, it was the elites doing the divination without condemnation from stick in the mud guys. Almost four thousand years ago, Babylonian seers read omens in the livers of sacrificed sheep. They trained on clay teaching models that mapped the liver’s surface into zones, each inscribed with what a mark found there would mean (seriously check it out it’s a pretty neat artifact). Excavators in Israel found a cache of knucklebones shaved, drilled, and weighted for divination dating to about 2,300 years ago. Given the trends of human history it’s hard to imagine it’s the case that coffee came to town and invented a whole new social practice.
Particularly when you consider tea-leaf reading, which in the West is well-known mostly because of the Harry Potter books. It’s really more of a Chinese thing, though: tea was a national drink by the Tang dynasty, with Lu Yu’s eighth-century treatise codifying it. Tea sellers trace tea-leaf reading back to ancient Chinese drinkers spotting symbols in their cups. There’s a “Highland Seer” volume describing leaf-reading as an established peasant practice in Scotland. Victorian hostesses later brought Roma readers into their parlours. Scottish tea-leaf reading developed the same women-centered, domestic, licensed-talk structure; the 18th-century Tea Cup-Reading book attributed to a “Highland Seer” describes generations of spae wives reading cups, and by the early 1800s a lady of the manor was reading fortunes for her servants after breakfast.
Whether the British teacup reading borrowed from the Ottoman coffee cup, from Europe’s older wax and lead pouring, from China directly, or reinvented the practice independently is a mystery as far as I know. Either way, drink dreg fortune-telling is hardly unique to the Ottomans. You can even do it with wine. The Old Testament mentions Joseph’s skill in divination by use of a cup!
How harem origin story originates in modern journalism
Wikipedia’s tasseography article says coffee fortune-telling “first appeared in the Ottoman Palaces in the 1500s” and that bored harem concubines told each other fortunes in order to chat and gossip, but Wikipedia itself flags this as better source needed; Food52’s version quotes a Turkish coffee company founder saying readings began “first appearing in the Ottoman palaces as a communication tool” for palace wives, but we know almost nothing about Ottoman palace interiors because Western diplomats weren’t allowed in there; BBC Travel says the art “can be traced back at least 500 years” and was “a way to trade news and gossip,” but the source is “Gizem Salcigil White, who has been working for more than a decade to introduce Turkey’s unique customs and reverence for coffee to the United States through her business, the Turkish Coffee Lady.”
I don’t want to impugn her motives or anything, but she’s not exactly somebody with lacking in motivation to play up the harem intrigue angle. And for all we know, the badly-sourced Wikipedia claim was her source…
Meanwhile, the scholarship I’ve looked at that reconstructs the harem from archival documents, like Leslie Peirce’s Empress of the East (which I reviewed around a year ago; it’s a good book), builds a picture of daily life from stipend registers, training hierarchies, patronage networks, and building campaigns. There just aren’t that many sources that can speak to underhanded gossip networks inside the harem walls. So the evidence for the origin story is really thin. Like, harems were deliberately opaque, Ottomans observed a cultural taboo against writing publicly about the sultan’s household, and most European “eyewitness” harem accounts were almost certainly fantasies composed by men who never got past the gate… the same way that modern-day tabloids claim “eyewitness” accounts of Britney Spears’ alien baby or whatever.
But even if we wanted to take it at face value, at least some of the comments on that Reddit thread do not align with what we know of harems. As you may recall from When Royal Marriages Keep Power Inside the Family, the harem was a training institution and a power center, the women’s counterpart of the palace school that produced viziers. The main risks for harem women were factional, like backing the wrong prince or crossing the sultan’s favorite, and hiding behind fortune-telling doesn’t really offer much protection against that kind of thing because plausible deniability only gets you so far in an active power struggle.
Really, I just worry that we’re at risk of anachronistically pushing modern practices onto the Ottoman past. Or, on the flip side, attributing a beginning to something much more recently than it actually began, due to a lack of sources. Seriously, divination from dregs in cups goes back a really long way.
But that doesn’t mean it’s not interesting! Or that it’s not still widespread, which is in some ways more interesting.
Laura Shannon, who researches women’s ritual practices in Greece and the Balkans, finds grounds-reading across Greece, the Balkans, Asia Minor, the Caucasus, and the Middle East. It’s practiced mainly by older women in domestic settings; Armenian communities keep dedicated readers called kooshags working in diaspora cafés. And while it’s definitely not always a “do gossip safely” thing, a 2025 discourse-analysis study by Betil Eröz and colleagues, “’Neyse Halim Çıksın Falim’: Turkish women’s intimate discourse in fortune-telling sessions through coffee cup readings”, documents how the readings help structure intimate discussions about marriages, money, and in-laws that would be harder to raise cold. I’ve experienced something similar with tarot card reading in high school, and the same mechanism turns up among Turkish Gen Z-ers.
The cup offers a discussion outline, so the reader can say vague but difficult things (”I see an enemy close to you”; “there is trouble on a road you are considering”) without owning them, and the party being “read” can disclose useful information — as much as they’re comfortable with — as if merely confirming what the grounds already revealed. The ritual was a way of trading news and gossip, including which wives were favored by the Sultan, according to Salcigil White… but I’m right back to: how exactly can she possibly know that? We have almost no records from that time!
Now, sure. This is the same kind of mechanism I like to discuss in the context of practical rituals. Kahneman makes a related point in Thinking, Fast and Slow, that “the expectation of intelligent gossip is a powerful motive for serious self-criticism.” A ritual that licenses gossip is doing important social work along the lines of what I outlined in my article about how gossip and murder were critical to the domestication of Homo sapiens sapiens. There are, separately, lots of cases of institutions doing concrete social work under the cover of the “sacred,” like the famine-preventing welfare functions of religious organizations, or ritual plague prevention. The “ritual” labels in the archaeological record often mark perfectly practical behavior — particularly social lubricant events like Thanksgiving or sporting events. And divination certainly does add deniability, since the message ostensibly comes from fate or the gods rather than the speaker.
I’m just not convinced that this tradition started with coffee from Yemen making its way into Ottoman harems.
What do y’all think? Am I being needlessly skeptical?
I'm thinking this is an excellent example of what peer review should look like.
Great article! Thanks for taking the time to write it