There Is No Single Object Called “the Septuagint”

“The Septuagint says” is one of the most common sentences in biblical scholarship.

It is also frequently one of the least exact.

What says it? A papyrus fragment of Deuteronomy? The earliest Greek form that an editor has reconstructed for Isaiah? Codex Vaticanus? Codex Sinaiticus? A medieval manuscript? Alfred Rahlfs’s twentieth-century manual edition? A Göttingen editor’s eclectic text? A quotation in a church father? One of the Jewish revisions commonly associated with Aquila, Symmachus, or Theodotion?

Those are not interchangeable objects. They do not all have the same date, contents, history, or evidential value. Some are physical artifacts. Some are textual forms inferred from artifacts. Some are ancient revisions. Some are Christian codices containing many books. Some are modern printed judgments about what an earlier Greek translator probably wrote.

The singular noun quietly turns them into one witness.

The proposition of this article is:

The term Septuagint is a necessary convenience but a dangerous ambiguity unless it is attached to a particular biblical book, textual stage, manuscript, or critical edition.

This is not a denial that Greek translations of Hebrew Scripture existed. It is a demand that the witness be identified before it is permitted to testify.

Six Objects Under One Name

The word Septuagint is used for at least six distinguishable things.

The first is a story about an origin. The second is a scholarly reconstruction of an early translation. The third belongs to a history of revision. The fourth is a material manuscript. The fifth and sixth are modern editions.

They can overlap. A Göttingen editor may judge that a reading in Vaticanus preserves the Old Greek. A church father may quote a Jewish revision. A Christian codex may contain an early translation in one book, a revised Greek form in another, and later corrections in both.

Overlap does not create identity.

The International Organization for Septuagint and Cognate Studies says that the collection is better described as an “anthology” than a homogeneous whole.¹ The introduction to the New English Translation of the Septuagint likewise explains that Old Greek is the more exact term for the earliest recoverable Greek rendering of an individual book or translation unit. It retains Septuagint because usage has made the name convenient, not because the corpus has been proved to descend in every part from one committee and one day.²

That admission should govern ordinary speech.

A Manuscript Is Not a Reconstructed Text

Suppose Codex Vaticanus contains one wording, Codex Sinaiticus another, and several later manuscripts a third. A critical editor studies those readings, quotations, daughter versions, translation patterns, and scribal tendencies. He then prints the wording he judges most likely to precede the surviving variants.

The printed result is not another ancient manuscript.

It is an editorial conclusion. Its letters may never appear in that exact combination on any extant ancient page. The conclusion may be learned, responsible, and highly probable. It remains a reconstruction.

This distinction is especially important when someone says, “The Septuagint is older than the Masoretic Text.” The sentence may conceal several different claims:

• A particular Greek artifact is assigned an earlier production date than a particular Hebrew codex.
• The Greek wording in a later artifact is judged to descend from an earlier translation.
• An editor’s reconstructed Old Greek reading is judged to reflect an early Hebrew exemplar.
• A modern printed Greek text is being compared to a medieval Hebrew manuscript as though the printed edition itself were ancient.

Only the first compares two physical objects. The second and third compare inferred textual histories. The fourth commits a category error.

The age of a manuscript and the age of a reading are not identical. A later copy can preserve an ancient reading. An earlier copy can preserve a translator’s mistake or an already revised form. That is why the actual object must be named.

A Printed Page Can Hide a Committee of Witnesses

Rahlfs’s Septuaginta, revised by Robert Hanhart, is a manual edition. It gives readers a compact Greek text across a large corpus. Its usefulness depends upon editorial selection. It is not a photographic copy of a pre-Christian Bible.

The Göttingen Septuaginta: Vetus Testamentum Graecum has a different purpose and scale. It treats individual books with a much fuller apparatus and attempts to reconstruct their earliest attainable Greek forms. The current Göttingen Psalter project openly describes its goal as a critically reconstructed text drawn from a large and layered manuscript tradition.³

The distinction is not an accusation. It is printed on the title page of the work: this is a critical edition.

NETS therefore uses Göttingen editions where available and Rahlfs elsewhere. Its editors expressly warn that no edition can claim to be the definitive text of the original Greek translator.⁴ Even a translation called A New English Translation of the Septuagint must first decide which edited Greek text it is translating.

Consequently, “NETS reads,” “Rahlfs prints,” and “Göttingen reconstructs” are not pedantic substitutions for “the Septuagint says.” They identify three different acts.

The Old Greek Is Also a Reconstruction

Old Greek is the standard scholarly term for the earliest recoverable Greek translation of an individual biblical book or unit. It is more exact than using Septuagint for everything, but it must not be made magical.

The Old Greek is not necessarily the wording of an extant scroll labeled “original translation.” It is normally reconstructed from later manuscripts, early fragments, daughter versions, patristic citations, and the translator’s habits.

Consider what the phrase “the Old Greek of this passage” can involve:

1. Establishing which Greek witnesses preserve the book.
2. Distinguishing copying errors from meaningful variants.
3. Identifying later revisions toward a Hebrew text.
4. Learning the translator’s normal equivalents and syntax.
5. Deciding which reading best explains the others.
6. Printing the result as the earliest recoverable form.

The work may be excellent. The conclusion may be stronger in one book than another. It remains an E5 textual reconstruction under this series’s evidence rules, not an E1 physical artifact.

When the reconstruction is then used to infer a lost Hebrew parent text, another inferential step has been added. When that retroverted Hebrew is proposed as a correction of the received Hebrew, a third step has been added.

Compact language must not conceal that ladder.

Williams Sees the Ambiguity and Draws the Wrong Conclusion

H. D. Williams quotes Karen Jobes and Moisés Silva to the effect that, strictly speaking, there is no single thing called the Septuagint. He then turns their warning about the ambiguity of a collective noun into the conclusion that the Septuagint is a “non-entity” and that its very name should be rejected.⁵

That does not follow.

There is no single object called “English literature.” There are poems, plays, sermons, novels, letters, editions, and manuscripts. It would be absurd to conclude that English literature is imaginary.

There is no single object called “the Masoretic tradition” if by single object one means one physical manuscript containing every stage, scribal note, reading tradition, and printed edition. That does not make the Hebrew tradition nonexistent.

Corpus diversity refutes careless singularity. It does not erase the corpus.

Williams is right to object when “the Septuagint” is treated as a complete pre-Christian codex or one uniform textual authority. He is wrong to make the ambiguity of the label disprove the physical fragments, the individual Greek translations, or the later manuscript traditions to which the label is conventionally applied.

The correction is not to ban the noun. It is to control it.

The Strongest Objection

Someone may answer that specialists already understand these distinctions. Requiring “Codex Vaticanus,” “Old Greek,” or “Rahlfs-Hanhart” every time will make ordinary writing unreadable.

Technical language does require reasonable abbreviation. Once an article has defined LXX as the Old Greek of Genesis according to the Göttingen edition, it need not repeat the entire formula in every sentence.

But the initial definition cannot be skipped when the argument depends upon age, authority, or textual difference. Those are precisely the places where ambiguity does argumentative work.

Compare these statements:

The Septuagint adds a clause to the Hebrew.

---

1. International Organization for Septuagint and Cognate Studies, “A Prospectus for a Commentary on the Septuagint”, preamble and principles. The prospectus calls the collection an anthology, treats the parent text as necessary for understanding translated Greek, and requires book-specific treatment. ↩
2. Albert Pietersma and Benjamin G. Wright, eds., A New English Translation of the Septuagint, front matter (Oxford: Oxford University Press, 2007), xii–xiii, xviii–xix. The introduction reserves Old Greek for the earliest recoverable rendering of an individual book or unit and explains the collection’s diversity. ↩
3. Göttingen Academy of Sciences and Humanities, “The Editio critica maior of the Greek Psalter”. The project describes a layered tradition, approximately 1,300 catalogued Psalter manuscripts, Christian revisions, and the preparation of a critically reconstructed text. ↩
4. Pietersma and Wright, NETS, xix. NETS uses Göttingen editions where available and Rahlfs’s manual edition elsewhere, while expressly recognizing the provisional character of an editor’s reconstruction. ↩
5. H. D. Williams, The Septuagint: The Character of God’s Words Is Not Found in the Septuagint, supplied PDF, PDF pp. 13, 16–17. Williams accurately reproduces the warning about the term’s multiple referents but then treats diversity as proof of nonexistence. ↩

Codex Vaticanus and Rahlfs-Hanhart contain a clause absent from the received Hebrew; the Göttingen editor judges it part of the Old Greek.

---

1. International Organization for Septuagint and Cognate Studies, “A Prospectus for a Commentary on the Septuagint”, preamble and principles. The prospectus calls the collection an anthology, treats the parent text as necessary for understanding translated Greek, and requires book-specific treatment. ↩
2. Albert Pietersma and Benjamin G. Wright, eds., A New English Translation of the Septuagint, front matter (Oxford: Oxford University Press, 2007), xii–xiii, xviii–xix. The introduction reserves Old Greek for the earliest recoverable rendering of an individual book or unit and explains the collection’s diversity. ↩
3. Göttingen Academy of Sciences and Humanities, “The Editio critica maior of the Greek Psalter”. The project describes a layered tradition, approximately 1,300 catalogued Psalter manuscripts, Christian revisions, and the preparation of a critically reconstructed text. ↩
4. Pietersma and Wright, NETS, xix. NETS uses Göttingen editions where available and Rahlfs’s manual edition elsewhere, while expressly recognizing the provisional character of an editor’s reconstruction. ↩
5. H. D. Williams, The Septuagint: The Character of God’s Words Is Not Found in the Septuagint, supplied PDF, PDF pp. 13, 16–17. Williams accurately reproduces the warning about the term’s multiple referents but then treats diversity as proof of nonexistence. ↩

The second statement lets the reader ask the right questions. Does Sinaiticus agree? Do early papyri survive at the place? Is the clause present in all Greek families? Could it be an expansion? How confident is the reconstruction? Is the inferred Hebrew wording extant anywhere?

The first statement closes those questions before they are asked.

Precision is not clutter when it reveals the evidence.

The Rule for the Rest of the Series

From this point forward, every use of Septuagint will be governed by four questions:

1. Which biblical book or translation unit?
2. Which physical manuscript, textual family, or revision?
3. Which printed or critical edition?
4. Is the statement directly attested or reconstructed?

If a sentence cannot answer those questions, it probably asks the word Septuagint to carry more weight than the evidence.

The doctrine established in Packet One remains unchanged. The Greek version is a translation and therefore subordinate to the authentical Hebrew. Packet Two adds a historical discipline: before the stream can enter the witness box, the court must know which channel of the stream has arrived.

The next article will apply that discipline to the oldest and most influential explanation of the name—the story of Aristeas and the seventy-two translators.

Notes and Further Reading

See also Alfred Rahlfs and Robert Hanhart, eds., Septuaginta: Id est Vetus Testamentum graece iuxta LXX interpretes, rev. ed. (Stuttgart: Deutsche Bibelgesellschaft, 2006), and the book-specific volumes of Septuaginta: Vetus Testamentum Graecum.

Previous in the series: “Three Translation Supremacies, One Inversion.”

Next in the series: “Aristeas and the Seventy: What the Story Does—and Does Not—Prove.”