Seventeen hundred years. That's how long Hebrew went without anyone raising a child in it — and then, within about two generations, ten million people.
And it's the only time it's ever happened. No other natural language with zero native speakers has gone on to acquire millions of them, and no other sacred language has become a national one. The revival is a genuine singleton in the historical record.
Which is what makes Daniel's question so pointed. He's not asking about the accent, he says — pronunciation is the boring part. He wants the grammar: word order, the verb system, syntax, vocabulary, how much of the way modern Hebrew builds its sentences is actually European. Whether a speaker of one would understand a speaker of the other. And he names the man to argue with — Ghil'ad Zuckermann, the Australian-Israeli linguist who says the language Israelis speak shouldn't be called Hebrew at all, but "Israeli," a hybrid of Hebrew and Yiddish and a half-dozen others. Daniel's framing is fair: a minority view, worth taking seriously, worth pushing back on where it overshoots.
And it's a question with a country attached to it. The twenty twenty-eight... sorry, the twenty-eighteen Nation-State Law made Hebrew the sole national language of Israel. So "is this the same language" is not a seminar question — it's tangled up in curricula, identity, the whole origin story.
Then let's start with the thing everyone thinks they know — that modern Hebrew is just Biblical Hebrew with a funny accent. That's where it falls apart first.
The baseline, quickly, because the comparison only means something if you know what the two objects are. Biblical Hebrew is a Canaanitic Semitic language, attested in inscriptions from around the tenth century BCE, spoken as a vernacular through the Second Temple period. It stops being a regularly spoken language somewhere between two hundred and four hundred CE, probably declining after the Bar Kokhba revolt. After that it's alive but silent — liturgy, rabbinic writing, poetry, commerce between communities, never a mother tongue.
A language on life support for seventeen centuries.
Kept warm, though. Read every week, written continuously. It's not like reconstructing Etruscan from fragments. The revival Zuckermann dates to the eighteen eighties through the nineteen thirties, and now you've got over ten million speakers, six and a half million of them native.
And here's the thing about the accent everyone focuses on. Sephardi pronunciation was adopted partly as ideology — the Ashkenazi variants were the ghetto, the new nation wanted the desert voice — and partly because it was simpler. But Daniel's right that it's the least of it. The grammar is where the distance lives.
Start with word order, because it's the clearest single change. Biblical Hebrew is verb-first, typically. Genesis one-one, read in the Hebrew order, is "created God the heaven and the earth." Verb, subject, object. Modern Hebrew is subject, verb, object — the English order. "God created the heaven and the earth."
Which sounds like a small thing until you realize an Israeli reading Genesis aloud is encountering a sentence structure that's ungrammatical in their own mouth.
Ungrammatical is exactly Zuckermann's word. His example is Job fourteen nineteen — avanim shaqqu mayim. The actual syntax there is object, verb, subject: "the waters wear away the stones." Israelis read it as "the stones eroded the water," because object-first is simply not a word order the modern language permits. It's not that they lack the vocabulary. Every word is familiar. The architecture is foreign.
Modern Hebrew does still allow verb-first, doesn't it? You hear it.
You do, but as a marked choice. Putting the verb first sounds formal or archaic — it sounds biblical, which is exactly why people reach for it when they want to sound lofty. The modern language kept the old order the way English kept "whom" — as a costume.
A dialect keeps its grandfather's syntax as a rhetorical flourish. That's a strange kind of inheritance.
And here's where the second big change comes in, because word order and the verb system are connected. Biblical Hebrew doesn't really have tenses the way we think of them. It has aspect — the perfect, marked by suffixes, for completed action; the imperfect, marked by prefixes, for incomplete action — plus the waw-consecutive forms that drive biblical narrative forward. It's a system about whether something is finished, not about when it happened.
And modern Hebrew reanalysed all of that as tense.
As a clean past, present, future. The old perfect became the past tense. The old imperfect became the future. And the present tense is built out of what was, in the Bible, a participle — the "doing" form. So the machinery got reorganized from "is this done or not done" to "when did this happen."
My guess is the forms themselves barely changed.
Barely at all, and this is the point I want to slow down on, because it's the crux of the whole episode. Rahel Halabe, who's taught both varieties for decades, puts it directly: one who knows the grammar of biblical Hebrew will not have to relearn how to form the future or past tense when they learn modern Hebrew. The forms transfer perfectly. What doesn't transfer is the meaning. Zuckermann's line is that after eleven years of biblical training in the Israeli school system, Israeli speakers still read the perfect aspect as past tense and get the sense wrong — because in the biblical system that form can carry "I will have said" as easily as "I said."
So a modern reader looks at a familiar form and assigns it the wrong category. The word is a friend; the grammar underneath it is a stranger wearing the friend's coat.
That's the elegant trap of the whole relationship, yes. Morphological continuity with semantic rupture. It looks like the same language from the outside, from the conjugation table, and behaves like a different one in the sentence.
Say the Job example again — the erosion one. What makes that misreading invisible to the reader?
Word order. Every word checks out as a word they know — stones, wear away, waters. So the brain takes the sentence in the only order the modern grammar permits and produces "the stones eroded the water." Nobody notices they've reversed the physics of the sentence. The water was doing the eroding.
A geologist would complain.
A geologist reading the modern Hebrew would file it under stones being dissolved, yes. Meanwhile the Semitic core survives, and I think this is underweighted in the whole debate. Modern Hebrew is still a fusional, synthetic language. It still builds words from triliteral roots through vowel patterns — k-t-v gives you katav, niktav, ktov, the whole family — and prepositions still take pronominal suffixes. That root-and-pattern system, where the consonants carry meaning and the vowels between them carry grammar, is the deep architecture of Semitic. It's completely intact.
The bricks are Semitic; the blueprints are European.
And a living example, because people assume Hebrew's evolution ended in the nineteen twenties. It didn't. The accusative marker et — the little word that flags a definite direct object — only appears before a definite object. Ten li et ha-sefer, give me the book. In colloquial speech, et-ha is fusing and reducing to ta. Ten li ta-sefer. That's grammaticalisation happening in real time, in the mouths of people waiting for buses.
So a linguist could sit on a bench in Tel Aviv and watch a morphological change happen across an afternoon.
Roughly. It's not fully standard yet, but it's widespread in speech. And there's a tidy structural detail in the same neighborhood: modern Hebrew is partly pro-drop. When the verb marks gender, number and person, you can drop the subject pronoun. But in the present tense you can't, because present forms don't mark person — the same form serves I, you, he. The pronoun has to stay.
So the language drops the subject precisely where the verb does the work and keeps it precisely where it doesn't. Efficient, in its way.
Rule-governed redundancy. The old Semitic economy, still running.
So the grammar picture is: word order flipped, aspect became tense, the clause is European, the word-building is Semitic. Now, you said the forms transfer and the meanings don't — what happens when you actually sit an Israeli down with the Bible?
That's where Zuckermann did his valuable work, in my view. His twenty-ten essay is a catalogue of Israelis misreading scripture with total confidence. Jeremiah thirty-one, yeled sha'ashu'im — most Israelis will tell you it means playboy. Because sha'ashu'im in modern Hebrew is fun, amusements. The biblical sense is a pleasant child, a darling. Isaiah thirty-seven, ba'u banim ad mashber — modern reader: children arrived at a crisis. Mashber in modern Hebrew is crisis. In the verse it's the mouth of the womb, the moment of birth.
And the readers have no idea they've missed anything.
That's his sharpest line: Israelis not only do not understand the Bible, but much worse — they misunderstand it without even realizing it. The worst part isn't ignorance, it's false fluency. The best one theologically is Job five-seven, adam le'amal yullad. The verse is an accusation — man is born to trouble, to mischief. Israelis read it as man was born to do productive work. It's turned from a curse into an Israeli work ethic.
Seventeen hundred years of silence, and the verse about man's inherent wickedness comes out as a LinkedIn post.
And Zuckermann's explanation for how the sentences got built this way is the Yiddish layer. Ma nishma — the universal Hebrew greeting, what's up — is a straight calque of Yiddish. Vos hert zikh, what's heard. The phrase didn't come from the Bible; it came from translating a Yiddish greeting word for word. He has a wonderful anecdote: President Ezer Weizman in nineteen ninety-six greets the Regius Professor of Hebrew at Cambridge with ma nishma. The Regius Professor of Hebrew — the most credentialed Hebrew reader alive, roughly — had no chance of guessing what it meant. He knew neither Yiddish nor the Russian, Polish, or Romanian equivalents it also matches.
The world expert on the language couldn't parse hello.
Because the hello isn't Hebrew. It's Yiddish wearing Hebrew clothes, and the professor had never lived in a shtetl.
Hilbert's Anteater Trivia. The claws on a giant anteater's front feet can be up to about ten centimetres long. I have never once needed a claw that size, and I have needed some things.
There's a photo of one of those at the desk and I've never asked why.
Don't. So the calques are real — ma nishma isn't the only one, I assume?
Not remotely. Phrasal verbs built with the Hebrew preposition ke-ilu, entire idioms, the rhythm of casual speech — the Yiddish substrate runs through how sentences get assembled even where the vocabulary is biblical. Zuckermann generalizes it as his Congruence Principle: a feature is more likely to persist into the new language if it exists in more than one contributor. And he applies it even to the word order change — the subject-verb-object order of modern Hebrew might not be purely European. It could simultaneously reflect the marked, emphasis order of Mishnaic Hebrew, the rabbinic Hebrew of the Talmud era. Multiple parents, one feature.
Which is more sophisticated than "Europeans rebuilt the sentence."
Much. It's multiple causation, and it's the version of his argument worth taking seriously. Because now we're at his actual thesis. He draws a taxonomy: Israeli, at the top. Two primary contributors beneath it — Hebrew and Yiddish, equal. Then a bench of additional contributors: Russian, Polish, English, German, Ladino, Arabic. And his illustrative percentage breakdown is deliberately impossible — forty percent Hebrew, forty percent Yiddish, ten percent Polish, ten percent Russian, ten percent English, five German, five Arabic, five Ladino.
That's a hundred and twenty-five percent.
On purpose. Features have multiple parents, so the shares can't be expected to sum. It's a rhetorical device dressed as arithmetic.
It's a good one, I'll grant. So his claim is the name should be Israeli, not Modern Hebrew, because calling it Hebrew hides the co-parent.
That's the claim, and he's explicit that it's a middle position. There's a traditionalist camp — Rosén, Rabin — that says it's simply Semitic Hebrew revived. And there's a revisionist camp, Horvath and Wexler, that goes the other way: Israeli is relexified Yiddish, Indo-European with Hebrew vocabulary pasted on. Zuckermann's hybrid thesis sits between them, and he calls the language semi-engineered — semi, he says, because the Yiddish contribution was not planned.
Now — before you defend him, I want the pushback on the table.
It is, and that's where most linguists part company with him. Yiddish gave the clause architecture, the calques, the sound of casual speech. Hebrew gave the root-and-pattern morphology, the inflectional machinery, the bulk of the lexicon, the entire writing system's relationship to the canon. Gold... Gideon Goldenberg, the Semiticist, in the nineties disputed the scale of the western influence outright. Saying Yiddish is an equal primary contributor — as opposed to a major substrate — overstates the case. The house is built with Hebrew joints and Yiddish plaster, and the joints hold the roof up.
And the pushback hasn't been purely academic, either.
No. Hillel Halkin, writing as Philologos in The Forward in December two thousand four, accused Zuckermann of a political agenda — that "Israeli" functions as a denigrating word for modern Hebrew, that the thesis serves a view of Zionism as a European colonial project that merely borrowed a holy language as branding. Zuckermann replied in February two thousand five with the line he's clearly fond of: Israeli, the world's youngest language, is mosaic rather than Mosaic, and it bristles with hybrid vigor.
Mosaic with a small m.
And to his credit, in The Mendele Review around the same time he conceded — his word — that nobody would deny the important ideological component in the emergence of Israeli. Which is about as close as that argument ever got to détente.
I notice what he doesn't concede, though. He concedes the ideology surrounded the language, not that the ideology produced the linguistics. Those are different concessions.
Sharp, and I think that's the fair reading. The descriptive work stands on its own. The Yiddish calques are real, the false friends are real, the word order shift is real, and what he calls the Internal Development Myth — that every feature of modern Hebrew evolved internally from biblical Hebrew — is simply not tenable. Nobody serious defends that anymore. Wikipedia's revival article states the softer version as settled: the revived language includes characteristics from the non-Hebrew languages of the Jewish communities, with Yiddish predominant. That's the mainstream. Where he loses the field is the equality claim, and the rename.
A rename is a big ask for a language that is also a country's name for itself.
Which is why Halkin reached for politics, and why the argument never stays in the linguistics. There's also a caricature problem — "Zuckermann thinks Israelis don't speak Hebrew" is stronger than anything he's published. His technical claim is that Hebrew is one of two primary contributors. The popular version of him is shouting more than he does.
That leaves Daniel's other question, and it's the one I actually care about. Would a speaker of one understand the other? Put Isaiah in a room with a Tel Aviv taxi driver.
Zuckermann's answer to that is my favorite thing he's written, because he dodges the fun half. The Semiticist Edward Ullendorff claimed Isaiah could have easily understood Israeli. Zuckermann: I propose that his statement is false — unless, he adds, he referred to Isaiah Leibowitz, yet another prophet.
He did not just... he did. He turned the prophet into the philosopher.
Deadpan as a filing cabinet. But then he flips the question, and the flip is the good move: could Isaiah understand us is unfalsifiable speculation. Do Israelis understand Isaiah is testable, and he has the receipts — the whole false-friend catalogue. That asymmetry is the real answer. An educated modern speaker might follow ninety percent of biblical narrative. A biblical Hebrew student who's never seen modern Hebrew cannot simply read a modern text — the vocabulary for ordinary modern life, the calques, the idiom, all of it is opaque.
Ninety percent — is that from a study?
No, and it's worth saying plainly: there is no controlled intelligibility testing. Nobody has put biblical readers and modern speakers in a comprehension experiment. The ninety percent figure is an estimate from a pedagogical source, The Hebrew Café, which takes the moderate line — different dialects of the same language, not entirely mutually intelligible, but one can pass between them with training. It's a reasonable guess, and it's a guess.
The asymmetry is what sells me on the dialect framing, though. If they were separate languages, the transfer wouldn't be this lopsided. A French speaker gets nothing free in Italian; an Israeli gets the Bible mostly free, wrong in the details.
And "wrong in the details" is the whole story of the relationship. I'll say this for Zuckermann before we let him go: the strongest evidence for hybridity isn't his percentages. It's sitting in the grammar we already walked through. The morphology is Semitic, the clause is European, and they coexist in every sentence without friction. Nobody designed that. It's what happens when a language is rebuilt by speakers whose instincts were formed elsewhere.
Wait — before we close that thread. Job fourteen-nine is the one that got me earlier, the ox one... no, hold on, it's the stone verse. Nineteen. The erosion reversal. That's the sentence I'd put on a T-shirt.
For a language family reunion, yes. The T-shirt reads "the stones eroded the water" and the biblical Hebrew reader winces.
The street sign at the top of my road has the fourth commandment on it. Zachor et yom haShabbat. The municipal version drops the et, which is right for a sign, and someone has added nikud in marker over the engraved letters. Small dots. I need the long ruler from the drawer to make the tseire out properly, and the glasses sit wrong on the snout, so it is a two-stage operation. I called the city's street-naming office about it. Spoke to a Mr. Abutbul there for forty minutes, most of it about why the archive holds the file. He read me the original nineteen sixty-two committee minute in which the quote was approved, and he read "mizkor" for "zachor," which the file itself corrects on the next line, so somebody in sixty-two heard the same thing. He was not embarrassed about it. I asked him to read the correction too. He did. He was decent about that.
Forty minutes with the file open and nobody's fixed the marker.
The marker is mine. That's the part I fixed.
The et dropped on the sign is the et dropping everywhere. Ten li ta-sefer by way of the Ten Commandments.
I don't think it's the same thing. The sign left it out on purpose. The speech leaves it out because it's heavy. Corn was quicker about that distinction than you're being.
He has the minutes, Herman. The man read the correction.
He did read the correction.
I stand corrected by the archive. It's the misreading I'd want on record, though — even the committee that chose the quotation heard it wrong first. That's Zuckermann's whole catalogue in one municipal file.
And it happened in nineteen sixty-two, before Zuckermann published a word of it. His thesis has street signs.
Cut to the cutting room floor, briefly — the detail I couldn't fit earlier. Zuckermann also dismantles the smooth-sailing claim of the philosopher Aviezer Ravitzky, who wrote that the modern Hebrew speaker moves smoothly through the Bible. Zuckermann's reply is one flat sentence: the alleged smoothness is a mere myth. And behind all of it sits a number from the other direction — Dio Cassius claims the Romans killed five hundred eighty thousand Jews in the Bar Kokhba revolt, which is the catastrophe that started Hebrew's slide out of daily speech. Even Zuckermann notes Dio might have exaggerated. The language's seventeen-hundred-year nap began with a possibly inflated body count.
Leave it to a Roman to round up.
The question I'd leave listeners with is the live one. Ten li ta-sefer. If et-ha really is fusing into a new prefix, then in a few generations Israeli grammar books will teach something no biblical grammar ever contained — and the same argument we just had about Yiddish's fingerprints will be had about our own. The language isn't a finished monument. It's mid-sentence.
The question "same language or not" may not have a clean answer, because the honest answer splits by level — same morphology, different clause, same lexicon, different meanings. Whoever asks Daniel's question in fifty years will get a different answer than we did today, without anyone having been wrong.
On that open question, we'll stop. Thanks to Hilbert Flumingtop at the desk, producing this one as ever.
If you enjoyed this one, try episode ten thirty-seven, From Scrolls to Software; episode thirty-four seventy-six, When Hebrew Became a Living Language Again; and episode ten thirty-one, The Clothes of Language. This has been My Weird Prompts. If the show earns it, leave us a review on Spotify — it helps people find us.
We'll be back soon. See you then.