Daniel sent us this prompt about Jewish scribes — Soferut — and it started from a personal place. He and Hannah named their son Ezra, and it was one of those compressed decisions. Jewish tradition gives you eight days to name a child, which is basically a deadline engineered to short-circuit overthinking. He didn't do the deep research beforehand. But when he finally looked into who Ezra the Scribe actually was, it landed — the great documenter, the man who took an oral tradition and locked it into text so it could survive exile and diaspora. And now he wants to understand the craft itself. The parchment, the ink, the quills, the training, how many scribes exist today, how long a Torah scroll takes to write, and what happens when you mess up a letter in a document where every letter counts.
The timing is perfect, because this is a profession that treats information preservation as sacred in a way that our digital world absolutely does not. We worry about bit rot and format obsolescence. Scribes have been worrying about one letter being a millimeter too wide for two and a half thousand years.
Who was this Ezra, and why does his approach to documentation still matter two and a half millennia later?
Ezra the Scribe lived roughly four eighty to four forty BCE. He was a priest and a scribe descended from Aaron, and in four fifty-eight BCE he led a wave of Jewish exiles back from Babylon to Jerusalem. This was after the Babylonian exile had scattered the population. The Temple was being rebuilt, but what Ezra found when he arrived was a people who had largely lost their scriptural literacy. They'd been in Babylon for two generations. They spoke Aramaic. The Torah was written in Hebrew, and most people couldn't read it.
The founding document of the civilization had become illegible to the civilization.
And Ezra's response was not to just translate it and move on. He gathered the community — the Book of Nehemiah describes this — he stood on a wooden platform, read the Torah aloud from morning until midday, and had Levites circulating through the crowd explaining what was being read. But his bigger move was structural. He's credited with standardizing the Torah text. Before Ezra, there were multiple textual traditions floating around. He essentially established the concept of a canonical text as a cultural anchor.
Which is a wild idea to introduce at scale. Before that, oral tradition was the primary transmission mechanism. You learned from your parents, from the elders, from the priests. It was embodied, contextual, flexible. Ezra says no — we're going to fix this in writing, and that fixed form is now the authority.
That decision is arguably what enabled Judaism to survive as a portable religion. When the Second Temple was destroyed in seventy CE, the sacrificial system ended overnight. But the text-based tradition Ezra had set in motion centuries earlier was already robust enough to carry the entire religion into diaspora. You didn't need a Temple. You needed a scroll and someone who could read it.
The book became the portable homeland. I mean, that's the phrase scholars use, but it's not wrong.
It's not wrong at all. And the profession Ezra founded — Soferut — is the mechanism that preserved that portability. A sofer, a scribe, is not just a copyist. The word sofer comes from the root samech-peh-resh, which means to count or to recount. A scribe counts letters. Every single one.
Which brings us to the craft itself. Let's get into the actual materials, the training, the rules that govern every single letter.
Let's start with the physical substrate. The parchment is called klaf, and it has to come from a kosher animal — traditionally goat, sheep, or cow. But here's the thing people often misunderstand: it's not leather. Leather is tanned. Klaf is processed specifically for writing — it's limed, scraped, stretched, and dried, but it's not tanned in the way leather is. The processing has to be done with the intention that this material will be used for a sacred purpose. It's not enough to just have kosher animal skin. The person processing it has to state explicitly, verbally, that this is being prepared for a Torah scroll or for tefillin or for a mezuzah.
Wait — the intention has to be verbalized during processing?
It's called lishmah, meaning "for its own sake" or "for its name." At multiple stages of production, the intent has to be declared. If the animal is slaughtered without the intention that its hide will become Torah parchment, that hide is not usable for a Torah scroll. The same principle applies to the scribe writing — before beginning, the scribe must verbally declare that they are writing for the sake of the sanctity of the Torah scroll. If they forget to say it, the scroll is invalid.
The entire supply chain is sanctified by declared intention. That's an extraordinary quality control mechanism. Every hand that touches the material has to consciously acknowledge what they're making.
It's completely alien to how we think about most production today. Nobody at the paper mill declares that this ream of printer paper is for the purpose of printing your tax returns. But in Soferut, intention is a material property. It's as essential as the carbon in the ink.
Speaking of which — the ink.
The ink is called dyo, and it has to be black. Not dark blue, not charcoal gray, black. The traditional formula is carbon-based, made from gallnuts — which are those little growths on oak trees caused by wasp larvae — combined with gum arabic as a binder and copper sulfate crystals, sometimes called vitriol. The gallnuts provide the tannins that produce the black color when combined with the iron or copper sulfate. Gum arabic gives it viscosity so it flows properly off the quill. No synthetic components are permitted. No metallic pigments beyond what's in the traditional formula. Aish dot com has a detailed breakdown of this — the ink has to be durable enough to last centuries, but it also has to be removable enough that a scribe can scrape off a mistake without destroying the parchment.
Removable but permanent. That's a tension.
It's the fundamental tension of the entire craft. The ink has to be permanent enough to survive millennia but not so permanent that it can't be corrected. And the formula has been stable for at least fifteen hundred years. The Israel Antiquities Authority has analyzed ink from Dead Sea Scroll fragments, and the composition is remarkably similar to what modern scribes use. Carbon black from lamp soot or burned olive pits, a plant-based binder, sometimes a little metallic salt.
The writing instrument?
A quill, called a kulmus, traditionally from a turkey or goose feather, or a reed called a kaneh. The quill has to be cut to a precise point using a special knife called a tikkun soferim. The scribe resharpens the quill constantly — sometimes every few lines — because the sharpness of the point directly affects the precision of the letter forms. No fountain pen, no ballpoint, no felt tip. The halachic literature is explicit here: the writing instrument has to be something that releases ink through capillary action from a split point. A modern pen that uses a ball mechanism doesn't qualify because the ink delivery isn't direct contact between the split point and the parchment.
I love that the rabbis essentially did materials science and fluid dynamics to rule out ballpoints.
The rabbinical rulings on this are astonishingly specific. There's a twentieth-century responsum — a formal legal opinion — from Rabbi Moshe Feinstein, who was one of the great halachic authorities, where he analyzes whether a fountain pen qualifies. His conclusion was no, because the nib doesn't function the same way as a split quill. This is a legal tradition that has been thinking about ink viscosity and capillary action for centuries before anyone called it fluid dynamics.
We've got kosher animal skin processed with declared intent, carbon-based ink from oak galls, and a turkey feather sharpened every few lines. What does it take to actually learn how to use all of this?
The training pipeline is intense. A sofer typically studies under a master scribe for two to four years, sometimes longer. There are over four thousand specific laws governing letter formation, spacing, layout, and ritual procedure. The student has to internalize the exact shape of every letter — not just what it looks like, but its precise proportions. How wide is a yud relative to a vav? How long is the leg of a kuf? What's the angle of the roof on a chet?
Four thousand laws just about writing letters?
And that's before you get to the laws about the parchment preparation, the layout of columns, the spacing between words, the spacing between letters, the crowns — which are those little decorative strokes on top of certain letters. There's an entire sub-discipline about the tagin, the crowns. Seven letters get specific crown configurations: shin, ayin, tet, nun, zayin, gimmel, and tzadi. And the crowns aren't optional decoration. They have to be there, and they have to be the right shape.
Becoming a scribe is roughly equivalent to a bachelor's degree, except your final project takes eighteen months and if you mess up one letter, you fail.
That's actually not far off. Certification requires writing a complete Torah scroll under supervision. That's three hundred four thousand eight hundred five letters, two hundred forty-eight columns. At a pace of about one to two columns per day — and that's a full workday — a complete scroll takes twelve to eighteen months. And the supervising scribe checks every column.
Three hundred four thousand eight hundred five letters. That's the exact count?
That's the traditional count, yes. It varies slightly by tradition — Yemenite scrolls have a marginally different letter count because of minor spelling variations — but three hundred four thousand eight hundred five is the standard Ashkenazi count. And every single letter has to be written correctly and completely. A letter that's missing a tiny stroke, a letter that touches another letter, a letter that's misshapen — any of these can invalidate the scroll.
How many scribes are actually doing this work today?
The estimates put it at roughly three thousand to five thousand active, certified scribes worldwide. Israel has the largest concentration — about two thousand. The United States has somewhere between eight hundred and a thousand. The remainder are scattered across Europe, South America, South Africa, Australia. It's a small profession. And within that group, not everyone writes Torah scrolls. Many scribes focus on mezuzot and tefillin, which are smaller and faster to produce. Writing a full Torah scroll is the pinnacle of the profession, and only a subset of certified scribes take on those commissions.
Three to five thousand people on the planet doing this work. That's fewer than the number of people employed by a mid-sized tech company.
Yet they're responsible for producing every kosher Torah scroll, every mezuzah, every pair of tefillin in use worldwide. The supply chain is incredibly tight. If you commission a Torah scroll for a synagogue, you're often on a waiting list. A well-regarded scribe might have a backlog of several years.
Which brings us to the economics. What does a Torah scroll actually cost?
A complete Torah scroll costs between thirty thousand and a hundred thousand dollars US, depending on the size, the quality of the parchment, the reputation of the scribe, and the level of ornamentation. The labor is the dominant cost. At one to two columns a day for eighteen months, you're looking at roughly four thousand to five thousand hours of skilled labor. Even at the low end, that's well below minimum wage per hour. Scribes are not in this for the money.
The economics of scribalism are basically the economics of art. You do it because you're called to do it, and if you're lucky, you can support a family by also producing mezuzot and tefillin and megillot.
A mezuzah scroll — which is the small parchment placed in a case on a doorpost — takes a trained scribe maybe a few hours. Tefillin scrolls take longer because there are four separate passages. Megillat Esther, the Esther scroll read on Purim, takes a few weeks. Scribes piece together a living from these smaller items, and a Torah scroll commission is the career-defining project.
What happens when you make a mistake? Because you will.
This is where it gets fascinating. The most common error is confusing dalet and resh. Dalet has a small square corner at the top right. Resh has a rounded corner. They look nearly identical. A single miswritten letter in the Shema — Deuteronomy chapter six verse four — can change "Hear, O Israel" to literal nonsense, or worse, to a different word entirely. If a scribe writes a resh instead of a dalet in the word echad, meaning "one," the word becomes acher, meaning "other." Shema Yisrael, Hashem Elokeinu, Hashem acher. Hear O Israel, the Lord our God, the Lord is...
That's a theological catastrophe in one letter.
And the scribe is hyper-aware of this at every moment. When an error happens, the protocol depends on the type of error. If the ink is still wet, the scribe can sometimes wipe it immediately. If the ink has dried, minor errors can be scraped off with a sharp instrument — the same knife used to cut the quill — and rewritten. But the scraping has to be clean. If the parchment is damaged in the process, you can't just sand it down. You have to cut out the defective section and sew in a new piece of parchment using sinew thread from a kosher animal.
The scroll is a patchwork of corrections stitched together with animal sinew.
It can be, yes. But there's a limit. A Torah scroll with three errors on a single page — a single column — is often declared pasul, invalid for ritual use. It's not destroyed. It's buried in a genizah, which is a storage repository for sacred texts that have become unusable. The respect for the text extends even to its disposal. You don't throw a Torah scroll in the trash. You bury it, like you'd bury a person.
Like adopting a feral cat.
I'm not sure that's the analogy I'd reach for, but the sentiment is not wrong. The scroll has a lifecycle, and the end of its usable life is handled with ritual care. There's actually a Torah scroll hospital at the Jewish National and University Library in Jerusalem where master scribes repair damaged scrolls. They stabilize the parchment, re-ink faded letters, repair tears. It's conservation work at the highest level.
A Torah scroll hospital. So we've got a supply chain sanctified by intention, a multi-year training pipeline, a global workforce smaller than a midsize company, an error-correction protocol involving sinew and burial, and a dedicated medical facility for damaged scrolls. That's the ancient craft. But how has this evolved from Ezra's time through to the present day — and what happens when digital technology meets an analog, sacred tradition?
Let's go back to the script itself, because the visual form of the Torah has changed more than most people realize. Ezra originally worked with what scholars call Paleo-Hebrew script, which was derived from Phoenician. It looks quite different from modern Hebrew lettering — more angular, more pictographic. The square script we see in Torah scrolls today is called Ktav Ashuri, Assyrian script, and it was adopted during the Babylonian exile. By Ezra's time, it had become the standard for sacred texts, but the transition wasn't instant. It took roughly two hundred years for Ktav Ashuri to become universal. And what's remarkable is that once it stabilized, it barely changed. The Dead Sea Scrolls include an Isaiah scroll dated to about one twenty-five BCE, and the scribal practices visible in it — the column layout, the letter spacing, the use of paragraph breaks called parashot — are nearly identical to what a modern sofer does today.
The format has been stable for over two thousand years. That's a longer run than any software file format. PDF has been around since what, nineteen ninety-three?
Nineteen ninety-two, and it's already gone through multiple major versions. A Torah scroll's format — parchment plus carbon ink plus Ktav Ashuri — has been readable for two millennia without a single software update. No migration, no emulation, no backward compatibility issues.
Yet modern scribes are using computers.
They are, and this is where the digital paradox kicks in. Many scribes today use computer-aided design software to plan layouts before they begin writing. They can check letter spacing, calculate column widths, verify that the text will fit properly on the parchment. Some scribes use digital tools for training — there are tablet apps that let you practice quill strokes on a screen. But the actual writing, the final product, must be done by hand, on parchment, with kosher ink and a quill. The digital tools are scaffolding. The finished scroll is entirely analog.
The computer is the sketchpad, not the canvas.
And that distinction is crucial. The halachic requirement is that a kosher Torah scroll must be written by a human hand with conscious intent. A robot arm holding a quill doesn't count, even if the output is physically identical. A laser printer producing perfect Ktav Ashuri on kosher parchment doesn't count. The human intention has to be present in every stroke.
Which is the same principle as the lishmah declaration during parchment processing. Intention is a material property.
It's the thread that runs through the entire tradition. And this creates an interesting tension with some recent developments. In twenty twenty-four and twenty twenty-five, several research projects attempted to use machine learning to analyze damaged Torah scroll fragments and predict missing letters. The idea was that AI could reconstruct text from partial fragments by learning the statistical patterns of letter sequences in the Torah. The scribal community's response was mixed. Some saw it as a valuable conservation tool — if you're trying to restore a damaged scroll for museum display, not for ritual use, AI prediction might help fill gaps. Others saw it as a fundamental violation of the principle that only human intention can determine what a letter is.
That's a genuinely hard question. If the AI predicts with ninety-nine percent confidence that a missing letter is a vav, and every human expert agrees, but no human actually wrote it or declared the intention — is it a letter or just a shape?
In halachic terms, it's just a shape. The letter is defined not by its visual form alone but by the act of intentional writing. This is why a child's drawing of the letter aleph is not an aleph in the ritual sense, even if it's perfectly formed. The child didn't have the intention of writing for the sanctity of a Torah scroll.
The AI is essentially a very sophisticated child with excellent penmanship.
In the eyes of halacha, yes. A prodigy, but not a scribe.
Let me ask you about preservation, because this connects directly to what Daniel does professionally. He works in digital archiving, AI, tech communications. He's spent years thinking about how to preserve information against entropy. And here's a tradition that has kept three hundred four thousand eight hundred five letters stable for over two millennia. What's the comparison?
The comparison is humbling, honestly. In digital preservation, we worry about bit rot — the gradual corruption of data on storage media. We worry about format obsolescence — the WordPerfect file that no modern software can open. We worry about link rot, about platform decay, about the fact that a significant percentage of web content from ten years ago is simply gone. The average lifespan of a hard drive is maybe five to ten years. Magnetic tape, properly stored, maybe thirty years. A Torah scroll, properly stored, has lasted over eight hundred years for the Bologna scroll, and fragments have lasted over two thousand.
The Bologna Torah Scroll — that's the oldest complete one?
Yes, dated to approximately eleven fifty-five to twelve twenty-five CE. It was in the University of Bologna library, catalogued as a seventeenth-century scroll until a librarian noticed in twenty thirteen that the script style was much older. Carbon dating and paleographic analysis confirmed it was from the twelfth or thirteenth century. It's complete, it's kosher, and it was being stored on a library shelf mislabeled for who knows how long.
The format didn't need a software update. Someone just had to look at it and know what they were looking at.
That's the lesson for digital preservation. The Torah scroll's format works because it's self-describing. You don't need a manual to read it — you need to know Hebrew and Ktav Ashuri, and those are transmitted culturally. The format is the culture. Our digital formats are not self-describing. A JPEG from twenty years ago requires a decoder that understands the JPEG specification. If that specification is lost, the bits are meaningless. The Torah scroll has no specification document. The specification lives in the scribal tradition itself.
The scribal tradition is the specification. It's living documentation, transmitted person to person.
Verified through multiple layers of redundancy. Here's something I didn't mention earlier: the scribe is required to say each word aloud before writing it. Not silently — aloud. So there's an auditory check, then a visual check as the word is written, then a visual check after it's written against a tikkun — which is a printed reference copy that shows the exact layout of every column. Then, when the scroll is finished, a second scribe reviews the entire thing. So you've got four layers of verification for every word. In software terms, it's a multi-stage quality assurance pipeline with a human in every loop.
Redundancy with intentionality. Not just backing up files to three different cloud services and forgetting about them. Each verification is performed by a trained human who has declared their intent.
That's the distinction that matters. In digital archiving, we've gotten very good at redundancy. Multiple copies, geographically distributed, checksummed, monitored for bit rot. But the intentionality — the conscious act of saying "this information matters and I am choosing to preserve it" — that's often missing. We preserve everything because storage is cheap, and we preserve nothing because curation is hard.
The scribal approach is the opposite. Preserve one thing perfectly. Every letter matters. Anything less than perfect is either corrected or buried with dignity.
Which brings me to something Dr. Hannah Schachter at the Van Leer Jerusalem Institute has written about — she studies the intersection of traditional scribal practices and modern information theory. She points out that the Torah scroll is essentially an error-detecting code. The spacing rules, the letter forms, the crowns — they all function as parity checks. If a letter is misshapen, it's detectable. If the spacing is wrong, it's detectable. The entire system is designed to make errors visible.
It's a human-readable checksum. The tradition itself is the error-correction protocol.
And it's been running continuously for two and a half thousand years. That's an uptime statistic that would make any site reliability engineer weep.
Let me pull this back to the personal story for a moment. Daniel named his son Ezra without doing the deep research — eight-day deadline, gut feeling, it just felt right. And then he discovers that Ezra the Scribe was the great documenter, the person who locked the oral tradition into text so it could survive. And Daniel's own career has been in documentation and digital archiving. That's either coincidence or the kind of pattern that makes you think harder about how names work.
The Jewish tradition actually has a concept for this. The Talmud says that parents receive a form of minor prophecy when naming a child. The name you're drawn to is not random. It reflects something about the child's essence or destiny. Whether you take that literally or metaphorically, it's a compelling framework. The name Ezra means "help" in Hebrew. Ezra the Scribe helped his community reconnect to their textual heritage. The scribe's role is fundamentally about help — helping the community maintain its connection to text across generations.
That eight-day deadline is itself interesting as a forcing function. In technical fields, we often talk about analysis paralysis — the tendency to over-research a decision until you can't make it at all. The eight-day naming window short-circuits that. You can't spend six months A-B testing baby names. You have to decide.
The tradition trusts that the decision made under that constraint will be the right one. It's the opposite of our cultural assumption that more information and more time always produce better decisions. Sometimes the constraint is the wisdom.
Alright, let's zoom out to the bigger picture. What can a two-thousand five-hundred-year-old preservation system teach those of us who aren't scribes about our own information practices?
I'd point to three things. First, intentional redundancy. The scribal tradition builds multiple verification layers, but each layer involves human attention and declared intent. For digital preservation, this means: don't just back up everything automatically. Decide what matters. Check on it. Know what you're preserving and why.
Second, format stability matters more than format features. The Torah scroll format has been stable for two millennia because it's simple, self-describing, and culturally transmitted. When you're choosing formats for your own important information, prioritize longevity over features. Plain text over proprietary formats. Open standards over closed ones. A text file from nineteen eighty-five is still readable. A WordStar document might not be.
Third, the value of slow media. In an era of AI-generated content at scale — where we can produce more text in a minute than a scribe produces in a lifetime — the scribe's commitment to a single perfect copy is a radical act. It's worth asking: what in your own work deserves the scribal treatment? What information is so important that every letter matters?
There's something almost defiant about it. The world is drowning in generated text, and these three to five thousand people are still sitting at their desks, sharpening quills, saying words aloud, writing one letter at a time.
They're not Luddites. Many of them use smartphones, plan layouts on iPads, communicate with clients over WhatsApp. They're not rejecting technology. They're being selective about where technology belongs in their workflow. The planning can be digital. The final product cannot.
That's a sophisticated relationship with technology. Not rejection, not uncritical adoption. Just a clear line between what a machine can do and what requires a human.
That line is drawn at the point of intention. A machine can lay out a column. It cannot intend to sanctify a letter.
As AI-generated text becomes indistinguishable from human-written text, do you think we'll develop new versions of scribal traditions for verifying authenticity? New forms of mesorah — that's the chain of custody?
Mesorah, yes — the unbroken chain of transmission. I think we're already seeing the early stages of this. Cryptographic signing of documents, blockchain-based provenance tracking, notary publics with digital equivalents. But the difference is that the scribal mesorah is personal. It's not a certificate. It's a relationship between a teacher and a student that extends back through generations. You can't fake that with a hash function.
A hash function can verify that a document hasn't changed. It can't verify that the person who wrote it intended to write it for a sacred purpose.
And that's the limitation of purely technical approaches to authenticity. They can verify integrity. They can't verify intent.
Where does that leave us? We've got a two-thousand five-hundred-year-old profession that treats every letter as a matter of consequence. We've got a digital world that generates billions of words a day, most of which will be unreadable within a decade. The contrast is almost too neat.
It is neat, but I think the lesson isn't that we should all become scribes. It's that we should be more intentional about what we choose to preserve and how. Not everything needs the scribal treatment. Your grocery list doesn't need to be written on kosher parchment with carbon-based ink. But your family history? The story of how your parents met? The letter your grandfather wrote before he left for the war? Those might deserve something closer to it.
The name Ezra means help. The scribe's role is to help the community maintain its connection to text across generations. What are we helping preserve?
That's the question to sit with. And it's the one thing the scribal tradition can't answer for us — because answering it requires the very intentionality the tradition is built on.
Now: Hilbert's daily fun fact.
Hilbert: In the seventeen-eighties, the British Royal Navy inadvertently created the first standardized maritime timekeeping system in Nunavut when a misplaced ship's chronometer was adopted by Inuit hunters as a ceremonial object, leading to a century-long tradition of precisely timed seal hunts that had nothing to do with celestial navigation.
...right.
This has been My Weird Prompts. Our producer is Hilbert Flumingtop. If you enjoyed this episode, do us a favor and leave a review wherever you listen — it helps more people find the show.
If you've got a weird prompt of your own, send it to us at myweirdprompts dot com. We read every one.
Until next time.
