“One Writing Says”: What Japan’s Oldest History Book Taught Me About the Bible’s Contradictions

I am not a biblical scholar. I work on water data, pipelines and APIs, that kind of thing. But about a year ago the YouTube algorithm got its hooks in me, and I haven’t come up for air.

It started with a Dan McClellan clip, I think, one of his TikTok responses where he’s calmly taking apart someone’s claim that “the Bible doesn’t contradict itself.” That led me to Religion for Breakfast (Andrew Henry’s channel, which you should not start on a weeknight). Then Bart Ehrman’s lectures and blog. Then Joel Baden’s work on the Pentateuch. Before I knew it I was reading about the Documentary Hypothesis at midnight and annoying my wife by narrating the differences between the Priestly and Jahwist flood accounts over dinner.

What hooked me wasn’t the contradictions themselves so much as the argument about the contradictions. Critical scholars like Baden and Ehrman point out that the Hebrew Bible contains irreconcilable accounts of the same events: two creation stories with different sequences, two interwoven flood narratives with different numbers of animals, three incompatible stories of how Saul became king. Their explanation is that these are fingerprints of a compilation process. Different source traditions, written at different times by different communities, stitched together by later editors who kept the contradictions rather than smoothing them out.

On the other side, apologists, people like Gleason Archer, Craig Blomberg, or more careful scholars like Iain Provan and V. Philips Long, argue the contradictions are only apparent. Read more carefully, they say. Attend to genre. The ancient authors were at least as smart as you. There’s always a way to harmonize, and if you can’t find one, that’s your problem, not the text’s.

For a while I was stuck between these. The critical arguments felt more historically grounded, but the harmonizers had a point: shouldn’t we extend charity to ancient texts? Shouldn’t we assume competent authorship before slicing the text into hypothetical sources?

Then I thought about my grandfather, and a book his people wrote twelve hundred years ago.

---

The Book My Grandfather’s People Wrote

My grandfather was Japanese. I grew up Jewish on one side, Japanese on the other, and I won’t pretend to have deep expertise in either tradition’s textual history. But I knew enough about Japanese mythology to feel a jolt of recognition reading Baden and Ehrman. The dynamics they described for the Hebrew Bible, multiple sources, ideological editing, contradictions preserved rather than resolved, I’d seen that before.

The Nihon Shoki (日本書紀), the “Chronicles of Japan,” was finished in 720 CE. Thirty volumes, creation of the world through the reign of Empress Jitō in the late seventh century. It was compiled by court scholars under Prince Toneri, continuing a project Emperor Tenmu (r. 673–686) had started decades earlier. They wrote it in classical Chinese, not Japanese, because the audience was continental: this was Japan’s calling card to the Chinese and Korean courts, proof that it was a civilized nation with a history worth respecting.

The Nihon Shoki had a job to do. Its mythological opening established that the imperial family descended from Amaterasu, the sun goddess. Its historical chapters constructed a chronological framework stretching back to a founding date of 660 BCE, a date modern scholars recognize as padded to rival Chinese antiquity. Legitimate the dynasty, glorify the state, assert Japan’s place in the East Asian order. It was a political document dressed up as a comprehensive history.

The compilers were working from a whole library of sources that didn’t agree with each other. Court chronicles, clan genealogies, regional traditions, Chinese dynastic histories, Korean records. Different, often contradictory versions of the same events. And what they did with those contradictions is the reason I’m writing this.

They kept them all.

---

“One Writing Says”

In the first two chapters of the Nihon Shoki, the mythological books covering the “Age of the Gods,” the compilers deployed a remarkable editorial device. After presenting a “main text” account of a mythological event, they appended variant versions, each introduced with the formula issho ni iwaku (一書に曰く): “one writing says.”

One writing says the world began when heaven and earth separated from chaos. Another says a thing existed in the midst of the void. Another says the land was young and floated like oil. Another says there was first a deity.

One writing says Amaterasu, Tsukuyomi (the moon), and Susanoo (the storm god) were born when Izanagi washed his face after returning from the underworld. Sun goddess from his left eye, moon god from his right, storm god from his nose. Another writing says the three were born from the union of Izanagi and Izanami together, before Izanami died. That’s not a minor difference. If Amaterasu was born from a unilateral act of purification, her supremacy has a different theological basis than if she was born from the primal couple’s union. The variants imply different models of divine authority.

The compilers didn’t pick one. They didn’t harmonize. They wrote “one writing says,” set the alternatives next to each other, sometimes dozens for a single episode, and moved on.

I remember the moment this clicked for me. This is what critical scholars describe for the Hebrew Bible, except in the Nihon Shoki, the editorial process is visible. The compilers showed their work.

---

Why They Didn’t “Fix It”

If you watch enough apologetics content, you’ll hear this objection to the Documentary Hypothesis: “If a redactor combined these sources, why didn’t he smooth out the contradictions? Are you saying the editor was too stupid to notice?”

It’s a fair question. The Nihon Shoki answers it.

Its compilers were not stupid. They were elite scholars working in classical Chinese, a literary language that took years to master, under the direct supervision of an imperial prince. They were producing a text that would represent Japan to the most sophisticated literate civilization in the world. They had every opportunity and every incentive to produce a clean, consistent narrative. They chose not to.

Comprehensiveness signaled authority. Every source tradition came from somebody, a clan, a region, a priestly lineage. Excluding a tradition meant excluding its transmitters. The issho apparatus let the compilers include everyone without adjudicating between them. “We didn’t leave anyone out” was a political statement as much as an editorial one. David Carr has argued that something similar was at work in the Pentateuch’s final redaction: keep everyone’s tradition in the text and everyone has a stake in the text.

The message also operated above the level of the contradictions. The Nihon Shoki variants disagree about how Amaterasu was born, how the heavenly descent happened, which route the gods took. But every single variant tells the same story at a higher level: the imperial family descends from the sun goddess, and the gods established the order of the world. The contradictions are in the details. The ideology is in the structure. Critical scholars observe the same thing in Genesis. The two creation accounts disagree about sequence and method, but both say YHWH alone created everything. The two flood narratives disagree about numbers, but both deliver divine judgment and a covenant with survivors.

And there’s a rhetorical dimension too. If five different sources, transmitted by five different institutions, all tell the story of the heavenly descent, each differently, the reader encounters a chorus. The details conflict, but the central claim is multiply attested.

Meanwhile, harmonization has costs. To make two contradictory accounts agree you have to add things neither says, or subtract things one of them does say. Every harmonization introduces a reading the text doesn’t contain in order to explain away something it does contain. The Nihon Shoki compilers sidestepped all of that. They let each source speak in its own voice. Inconsistency was the cost. Fidelity to the sources and the authority that fidelity conferred was the payoff.

---

The Control Case

This is where the Nihon Shoki does something for the biblical debate that arguing about Genesis alone can’t do. It gives us a control case.

When Ehrman or Baden point to contradictions in the Hebrew Bible, apologists can always fall back to: “Those are only apparent contradictions. You’re not reading carefully enough. The original author/editor intended coherence.” That’s circular, it assumes what it’s trying to prove, but it’s hard to break the circle because we don’t have the biblical redactor available for comment. We can’t ask him whether he thought Genesis 1 and Genesis 2 were in tension.

With the Nihon Shoki, we don’t have to ask. The issho apparatus is the compilers’ comment. They literally wrote “one writing says this, another writing says that.” They knew the accounts contradicted each other. They preserved them that way. They did not claim harmony.

Try applying the harmonization hermeneutic here. A committed harmonizer could say: “The variant about Amaterasu being born from Izanagi’s purification and the variant about her being born from the primal couple’s union aren’t really contradictory. Maybe the purification was a form of union, metaphysically speaking. Maybe one is the proximate cause and the other the ultimate cause.” With enough ingenuity you could reconcile them.

But we know that reconciliation would be wrong, because the people who compiled the text didn’t reconcile them. They flagged them as alternatives from different sources. A harmonizer would be imposing a reading the text itself rejects.

The biblical case is structurally the same. When someone reconciles the two creation accounts, the question is whether the reconciliation reflects what the ancient redactor intended or what a modern reader needs to be true. The Nihon Shoki pushes hard toward the latter. Ancient editors of compiled texts were comfortable with real contradiction. The assumption that they wouldn’t have been, which is the load-bearing assumption of the whole harmonization enterprise, is falsified by the Japanese case.

---

What This Doesn’t Prove (and What It Does)

The Nihon Shoki doesn’t prove the Hebrew Bible was compiled from contradictory sources. That’s a question for Hebrew Bible scholars, and they’ve been working on it rigorously for two hundred years.

What it does is close off a particular escape hatch: the claim that no ancient editor of an authoritative text would have tolerated real contradiction. One did. We can watch him do it. The text he produced became the foundational document of Japanese imperial ideology for over a millennium. Its contradictions were noted and debated across centuries of medieval commentary, and nobody treated them as grounds for rejecting the text.

I expect someone will object that the Nihon Shoki isn’t “scripture” the way the Hebrew Bible is, and so the comparison doesn’t hold. But the comparison isn’t about theological status, it’s about how the texts were put together. The Hebrew Bible didn’t spring into existence as “scripture.” For most of its compositional history, the texts that became the Bible were court records, legal codes, political histories, propaganda. Exactly what the Nihon Shoki compilers were working with. The sacralization came later.

What the Nihon Shoki proves is that the type of compositional process critical scholars describe, compilation from multiple sources, ideological motivation, preservation of contradiction, is historically real. It happened. It produced a text that worked for centuries. The harmonizer who insists it couldn’t have happened in the biblical case owes us an explanation for why one text should be exempt from a dynamic that demonstrably occurs in another.

---

The View from Two Traditions

I’m not an expert. I’m a data scientist with a YouTube habit and a family tree that touches two of the traditions I’m talking about. I don’t read biblical Hebrew or classical Japanese. Scholars of both texts would push back on things here, and they should.

But sometimes being outside both traditions is useful. When you study the Hebrew Bible from inside the Jewish or Christian tradition, the theological stakes can make it hard to see the editorial process for what it is. When you study the Nihon Shoki from inside the Japanese tradition, the nationalistic stakes can do the same. Standing with a foot in each, which is closer to my actual situation, you can see the family resemblance. Both texts were products of state power trying to tell a story about itself. Both drew on older traditions that disagreed with each other. Both preserved the disagreements because the disagreements served the project.

The difference is that one tradition spent a thousand years building an industry around explaining why the contradictions aren’t really there.

I don’t think the contradictions in the Hebrew Bible are a scandal. They don’t make the text less interesting. If anything, seeing the Bible as a compiled text, a conversation between traditions preserved by editors who thought the conversation mattered more than a single consistent voice, makes it more interesting. The Nihon Shoki compilers understood something the harmonizers don’t: a text that contains every voice is more authoritative than one that contains only one.

---