En-Gedi Leviticus Scroll

A Lump of Charcoal in a Synagogue Ark

In 1970 archaeologists digging the Byzantine synagogue at En Gedi, on the western shore of the Dead Sea, reached the Holy Ark and found charred lumps inside it. The building had burned down around AD 600, and the lumps were what was left of its scrolls. The excavation was led by D. Barag and E. Netzer of the Hebrew University with Y. Porath of the Israel Antiquities Authority; the authors of the eventual publication set down, in their own statement of who did what, that "Y.P." — Y. Porath — "excavated the En-Gedi scroll on May 5, 1970." Then nothing happened, and the reason is worth stating plainly, because it is the whole point of the story: the object could not be opened. Unrolling carbonised parchment destroys it. The Israel Antiquities Authority kept the lumps for four decades while, in the words of the same paper, "no one produced a means to overcome the irreversible damage they had suffered in situ."

The means arrived from computer science. P. Shor and Y. Porath of the Antiquities Authority arranged for the scroll to be scanned in three dimensions by micro-CT, and put the resulting volume in the hands of W. B. Seales of the University of Kentucky, whose team had been developing a technique they call virtual unwrapping: find the layers of the rolled sheet inside the scan, model each as a surface, flatten it, and read what is written on it. The ink cooperated for a reason the paper is careful about — its composition is unknown, because no surface is exposed to test, but "the ink response within the micro-CT scan is denser than other materials, implying that it likely contains metal, such as iron or lead," and dense material is exactly what a CT scanner sees. The Antiquities Authority announced readable text at a press conference in the summer of 2015; the full technical account appeared in Science Advances in September 2016, and the textual edition in the journal Textus the same year.

Two Columns of Leviticus

What came out is the beginning of a Torah scroll: the first two columns of Leviticus, eighteen surviving lines out of an original thirty-five in each. Column one carries Leviticus 1:1-9 and column two Leviticus 2:1-11 — the opening of the sacrificial law, the burnt offering and the grain offering. The published segment measures 16.8 cm long and 7.8 cm high as it survives, though the editors note the fire shrank it by an unknown amount and its original size cannot be recovered. The text is unvocalised and carries no verse divisions, which is what a Hebrew scroll of this period should look like. Everything from column three onwards is still rolled up and still unread.

The finding that made headlines was about agreement with the later Hebrew Bible, and it is worth quoting at its own weight rather than at the weight the headlines gave it. The Science Advances paper says: "The text deciphered thus far is completely identical with the consonantal framework of the medieval text of the Hebrew Bible, traditionally named the Masoretic Text, and which is the text presented in most printed editions of the Hebrew Bible." The edition in Textus is more precise still — the comparison was made against a named manuscript, and the paragraph divisions match too: "Throughout the entire two columns, the text preserved of Leviticus 1-2 matches the consonantal text of the medieval MT of Leviticus precisely, and we did not find a single variant vis-à-vis MT (choosing Codex Leningrad B 19A as our point of comparison)." Notice the two qualifications the editors themselves put on it: deciphered thus far, and consonantal framework. Vowels and accents are a later addition and are not in question here. And the editors add the deflating sentence that popular retellings drop: "Although it does not contribute any new readings (in the two columns that have been revealed to date), it serves as a valuable witness for the state of the biblical text in an era for which we have a paucity of direct evidence."

Nobody Agrees How Old It Is

The date is genuinely open, and the editors put the disagreement in their own publication rather than resolving it quietly. Radiocarbon dating at the Weizmann Institute gave "a probability of 68.2% that the scroll dates between 235-340 CE, and a probability of 88.9% that it dates between 210-390 CE," with only a 6.5% chance of the second century. The palaeographer A. Yardeni, examining the letter forms, reached a different answer: "It seems therefore that we may safely date this scroll to about the second half of the 1st century and at latest, the beginning of the 2nd century CE." That is a gap of a century and a half or more. Yardeni states her own limits before she states her conclusion, and they should be quoted with it: "Because of the very poor condition of this scroll, its image is distorted and does not reflect its precise appearance; therefore, the measurements given below are approximate and the drawings are conjectural," and the letters are so small that they "do not allow a sound examination of their forms."

The editors' own verdict is a model of what to do with a discrepancy: "We take note of the discrepancies between these two examinations which should lead to caution in any conclusion, but such discrepancies are not unparalleled." The argument has since moved on. D. Longacre reconsidered the palaeography in 2018 and argues in his abstract that "a series of important typological developments evident in the hand of EGLev suggests a date somewhat later than the Dead Sea Scrolls of the first-second centuries," and that the physical evidence — a wooden roller, metallic ink, tall narrow columns — points the same way; he concludes for the third or fourth century, which would agree with the radiocarbon. Only his abstract could be read here, so his reasoning is reported and not restated. What follows from the answer is worth stating, because it is why the argument matters: on the later dating the scroll is a witness to the Hebrew Bible from the centuries between the Judean Desert finds and the medieval codices, a stretch for which there is very little else; on Yardeni's it is contemporary with the latest of the Judean Desert manuscripts and fills no gap at all.

Scripture references

• Leviticus 1:1-9 — The text of the scroll's first column, recovered without the scroll being opened
• Leviticus 2:1-11 — The second column: the grain offering
• 1 Samuel 24:1-2 — David in the wilderness of En Gedi, the oasis where the synagogue stood a thousand years later

Sources

• ScholarshipW. B. Seales, C. S. Parker, M. Segal, E. Tov, P. Shor and Y. Porath, "From damage to discovery via virtual unwrapping: Reading the scroll from En-Gedi" — William Brent Seales, Clifford Seth Parker, Michael Segal, Emanuel Tov, Pnina Shor and Yosef Porath · Science Advances 2:9 (21 September 2016), e1601247, doi 10.1126/sciadv.1601247. Open access under CC BY-NC and read in full through PubMed Central; science.org itself refuses automated requests. Source for the 1970 excavation and the AD 600 destruction, for Porath's excavation of the scroll on 5 May 1970 — a sentence that stands in the paper's author-contributions note rather than in its body, and that names him by his initials, "Y.P.", not in full — for the ink's likely metal content, and for the sentence about identity with the Masoretic consonantal framework quoted whole in section two. Because the licence is non-commercial, this article quotes and cites it rather than reproducing any of itSupports: The discovery, the virtual unwrapping, and the claim about agreement with the Masoretic Text
• ScholarshipM. Segal, E. Tov, W. B. Seales, C. S. Parker, P. Shor and Y. Porath, with an appendix by A. Yardeni, "An Early Leviticus Scroll from En-Gedi: Preliminary Publication" — Michael Segal, Emanuel Tov, William Brent Seales, Clifford Seth Parker, Pnina Shor and Yosef Porath; palaeographic appendix by Ada Yardeni · Textus 26 (2016), 29-58. Read in full through the Hebrew University Bible Project's own copy; Brill's page for the same article refuses automated requests. This is the textual edition and it is the source that matters most here: the verse range (Leviticus 1:1-9 and 2:1-11, NOT 1:1-8, which is what the 2015 press coverage said), the eighteen of thirty-five lines, the 16.8 by 7.8 cm measurement, the comparison against Codex Leningrad B 19A, the radiocarbon probabilities in its footnote 4, Yardeni's palaeographic conclusion and her own caveats, and the editors' sentence about the discrepancy between the two datings. The excavation-date sentence quoted in section one is NOT here; it is in the Science Advances paper onlySupports: The verse range, the dimensions, the textual comparison, and both sides of the dating dispute
• ScholarshipD. Longacre, "Reconsidering the Date of the En-Gedi Leviticus Scroll (EGLev): Exploring the Limitations of the Comparative-Typological Paleographic Method" — Drew Longacre, University of Groningen · Textus 27:1 (2018), 44-84. The body is behind Brill's paywall and was NOT opened. What section three quotes is his own abstract, read on his university's research portal. His arguments are named from it and are not restated here · paywalledSupports: The third position on the scroll's date
• Holding institutionUniversity of Kentucky, Department of Computer Science: the En-Gedi 2016 scan data — W. B. Seales and the EduceLab team, University of Kentucky · The archive the published paper names as the home of "All scan data and results from this paper". Confirmed live. The master image of the unwrapped scroll is also deposited on the Internet Archive at archive.org/details/engedi-scroll, where it carries the paper's own non-commercial licenceSupports: That the scan data behind the reading is public and inspectable
• ReferenceBiblical Archaeology Society, "Book of Leviticus Verses Recovered from Burnt Hebrew Bible Scroll" — Biblical Archaeology Society · The lay account of the 2015 announcement, read in full. It is cited here for a specific reason: it carries the "first eight verses" figure that the 2015 press round produced and that is still repeated, and then updates itself. The primary publication says Leviticus 1:1-9Supports: Where the widely repeated "first eight verses" figure came from
• EncyclopediaWikipedia: En-Gedi Scroll

Further reading

• W. B. Seales et al., "From damage to discovery via virtual unwrapping: Reading the scroll from En-Gedi" (Science Advances, open access)
• M. Segal, E. Tov et al., "An Early Leviticus Scroll from En-Gedi: Preliminary Publication" (Textus 26)
• Wikipedia: En-Gedi Scroll

Reflect

This scroll was unreadable for forty-five years and nobody threw it away. What are you keeping that you cannot yet make sense of?