Persian fortune-telling
Four arts under one name. We compute one of them.
What the name covers
“Persian fortune-telling” is not one thing. It covers at least four: opening the Dīvān of Ḥāfeẓ at random and reading the couplet you land on (fāl-e Ḥāfeẓ); reading the shapes left behind in a coffee cup; the planetary period systems Persian astrologers handed on to the Arabic and then the Latin world; and ʿilm al-raml, the art of striking dots and turning them into figures.
Bokki computes exactly one of those four. The rest of this page is about which one, how it is calculated, and why not the others.
The one we compute: ʿilm al-raml
ʿilm al-raml — “the science of the sand” — took shape in the Arabic-speaking world around the ninth century: dots struck into sand, counted odd or even, gathered into figures. Bokki’s world atlas pins it to Arabia, because that is where it began, and we would rather the map and the article say the same thing.
But Persia is where the art was worked out at length. The practitioner’s Persian name, rammāl, is still an everyday word for a diviner, and it was through the Persianate world that raml travelled into India as ramal. In the twelfth century Latin Europe took the same art and called it geomantia. One set of sixteen figures; several centuries of commentary in several languages.
How Bokki casts it
A figure is four rows, read top to bottom as Fire, Air, Water and Earth, and each row is either a single dot (odd, active) or a double dot (even, passive). Four rows with two states each is sixteen figures — no more, no fewer — which is why the whole art fits into four bits.
Bokki casts four Mothers, then derives the rest the classical way. The four Daughters are the Mothers transposed, columns becoming rows. The four Nieces, the two Witnesses and finally the Judge are each the row-by-row sum of the two figures above them, where odd plus odd makes even. That addition is exactly an exclusive-or, which is why the classical checksum still holds in our code: the Judge always has an even number of active rows. The Reconciler is the Judge added back to the first Mother.
And here is the part to be honest about. The Mothers are drawn by a deterministic generator keyed to today’s date and to whatever you typed in the focus box — not by your hand in sand. So the same question on the same day gives the same figures, and tomorrow gives different ones. It is a reproducible draw, not a cast, and we would rather you knew that than believed something warmer and untrue.
The sixteen names we did not write
On Bokki’s geomancy page the figures carry Latin names — Via, Populus, Carcer, Coniunctio. That is not because Latin is the true language of this art. It is the language twelfth-century Europe translated it into.
Authentic Arabic and Persian names exist. We have not put them on the page, and the Arabic and Persian screens say why in their own words: our codex does not yet hold a table of all sixteen checked against sources, and writing sixteen names we have not verified would be inventing rather than computing.
That choice costs something real — an Arabic or Persian reader meets their own tradition under someone else’s labels — and we would still rather pay it than print a guess. When the sourced table lands, the names change.
What Bokki does not compute
Fāl-e Ḥāfeẓ is the one people ask about most, and Bokki does not do it. Two reasons, and the second matters more than it looks. The ritual is a book in your hands: you open the Dīvān and read what is in front of you. And while the Persian text of Ḥāfeẓ is centuries old and long out of copyright, every English or Korean Ḥāfeẓ you can find is a translation — a separate work with its own author. We are not going to build a feature out of somebody else’s translation.
The Persian solar calendar is not computed either. Its year turns at the exact instant of the March equinox, which would be a lovely thing to calculate, and we simply have not built it. Bokki’s calendars today are the Gregorian, the Korean 만세력 with its 24 solar terms, and the Maya count.
Firdāria — the planetary time-lord periods Persian astrology carried into the Arabic and Latin traditions — are not computed. They do resemble the Vimśottarī daśā that Bokki does compute: both hand fixed spans of years to planets in a fixed order. Whether that resemblance is descent or convergence has been argued over for a long time, and a footnote on a fortune site will not settle it. What we can tell you is which one is in our code, and it is the Indian one.
Coffee-cup reading sits in Bokki’s atlas as a tradition we honour and deliberately do not imitate — it lives in a cup, a hand and a room, not in a form field. And the Zoroastrian calendar, where each of the thirty days carries the name of a yazata, is not computed.
One thing we did do
Persian is one of the thirteen languages Bokki speaks, and the geomancy page is written in it — right to left, in Persian words, including the paragraph that admits we have not sourced the Arabic and Persian figure names yet.
It is a small thing. It is also the difference between inviting someone in and handing them a page that was written for somebody else.
Closing note — one of four arts, sixteen figures, and a checksum that always holds
In short: “Persian fortune-telling” covers at least four separate arts, and Bokki computes exactly one — ʿilm al-raml, the geomancy that Persian scholars carried and Latin Europe later translated. Four rows of Fire, Air, Water and Earth, each a single dot or a double dot, give sixteen figures and no more; deriving the Nieces, Witnesses and Judge as row-by-row addition is exactly an exclusive-or, which is why the classical checksum still holds in our code — the Judge's active rows are always an even number. The four Mothers are drawn from a seed keyed to today's date and your typed focus, not from a hand in sand, so the same question on the same day always returns the same sixteen figures.
The other three named arts — fāl-e Ḥāfeẓ, firdāria, the coffee cup — stay uncomputed on purpose: a book you hold, a translated poem with its own separate author, and a lineage that lives in a cup and a room rather than a form field. And the sixteen figures still carry their Latin names, not Arabic or Persian ones, because we have not yet sourced a checked table — a gap we would rather show than fill with a guess.