The Ein Gedi Torah Scroll?

A Sensational New Technology

Lying in storage for over forty years, a lump of carbonized parchment found in the 1970s excavation of the synagogue in Ein Gedi, has been partially deciphered using remarkable technology.^[1]

How sensational is this scroll?

It is quite remarkable, at least to me, that scientists have used advanced imaging technology paired with advanced computing to read a small, badly charred mush (see photo above). I have seen the technology developed by, and used at the Israel Antiquities Authority for reading Dead Sea Scrolls that have already been unrolled, which makes use of different lighting and filters.^[2] The new combination of imaging-and-computing technology developed by a team at the University of Kentucky, working in conjunction with the Israel Antiquities Authority and the scholars at Hebrew University, however, takes it to a whole new level.^[3]

But what is the significance of this discovery to our understanding of the Bible?

The Stabilizing of the Biblical Text

We have a plethora of biblical scrolls from the second century BCE – 68 CE from Qumran, near the Dead Sea, and from other nearby sites, from a slightly later period. The Aleppo and Leningrad codices, which contained the entire Hebrew Bible, date from approximately 930 and 1008 CE, but for the period between these scrolls and these early manuscripts, we have fewer than ten extant biblical scrolls. Moreover, the difference between what we find in Qumran and what we find in the tenth and early eleventh century codices is vast.

The Dead Sea Scrolls have yielded many texts that suggest that a pluriformity of biblical texts, rather than a single, uniform text of the Bible, existed for much of the Second Temple period. The Qumran texts differ one from another in small details (e.g., spelling of words), but also in larger matters (e.g., absence or presence of particular words or units, order of the textual units).  

“Our” biblical text, the Masoretic text (MT), is attested there—or more precisely, the proto-Masoretic text is attested, namely the MT without vowel points and cantillation marks, which were developed just more than half a millennium later.  But also found at Qumran are Hebrew texts that are similar to the Samaritan Pentateuch (SP), and those similar to what was likely the text that underlies the Septuagint (LXX), the early Greek translation of the Bible.  Proto-MT texts are found at Qumran, and according to some statistical analyses even dominated, but it was not the only text in town.

Scrolls reflecting biblical books were also found in various places near the Dead Sea (e.g. Wadi Murabba’at), though significantly fewer than those found at Qumran.  Most of those other scrolls are related to sites occupied through the Bar Kochba revolt (132-135 CE).  These are extremely similar to proto-MT; they largely agree with the wording and order of MT, but contain small variants, e.g., in spelling (especially full, מלא versus plene, חסר spelling).

The comparison of the Qumran scrolls to the second century Dead Sea Scrolls suggests the consonantal text of the Bible was standardized among Jews, and Hebrew texts now reflected in the LXX or the Samaritan were rejected, at the end of the Second Temple period (by 70 CE), or perhaps soon thereafter, perhaps partly in reaction to that catastrophe.

From the Bar Kochba Rebellion to the Cairo Geniza

My late teacher Nahum Sarna pointed out in a class, long ago, that the rabbinic system of midrash, especially as used by R. Akiva (who according to rabbinic tradition was martyred during the Bar-Kochba revolt), assumes a stable text whose slight idiosyncrasies (e.g., unusual spellings, an unexpected missing or present particle [i.e.[את]) may be exposited. This only would have worked if the community had an agreed-upon, precise text.^[4]   

By the time we get to the Cairo Geniza, the manuscript scraps found there, as well as the great early biblical manuscripts such as Aleppo and Leningrad B19a (the earliest preserved complete Bible) agree with one another, and preserve, almost without exception, the consonantal text of the MT; in fact, these texts define the MT.

The MT Style of the Seventh to Eighth Century MS Ashkar 2

Some scholars believed that the MT was the invention of the Masoretes in the second half of the first millennium CE. Nevertheless, certain pieces of evidence indicate that is incorrect, and that the Masoretic text of the Torah had stabilized earlier.^[5] Duke University owns a remarkable manuscript—MS Ashkar 2, which, based on paleographical analysis and carbon dating, is from approximately the seventh to eighth centuries.  A matching part of this fragment is now found in the Shrine of the Book in Israel. 

The Duke fragment contains the Song of the Sea from Exodus 15, written in the form of alternating brickwork mandated by rabbinic literature.  The text of both fragments is identical to MT and confirms that MT was not the invention of the Masoretes in the late first millennium, but reflects proto-MT, which was, as described above, one of several text-types that existed in the Second Temple period, and became the standard text of the Jewish Bible soon thereafter.^[6]

The Ein Gedi Scroll: Summarizing the Findings

Where Was the Scroll Found? 

Many early Jewish synagogues had a special Torah niche and a nearby cabinet that contained scrolls of biblical books. (These cabinets were relatively small, unlike the large, often ornate cabinets that now adorn many synagogues.) These stone niches have been excavated in many synagogues;^[7] the accompanying wooden structures have not survived, though they are depicted in some ancient Jewish art.  The Ein Gedi scroll was found near one of these niches.^[8]

When Was It Written? 

Carbon dating suggests that the Ein Gedi scroll was copied between the third and fourth century CE—a logical date given the synagogue’s destruction in the sixth century.^[9] Paleographical analysis, namely analysis of the letter shapes, however, suggests an earlier date, overlapping, at the latest, with the Dead Sea Scrolls found outside of Qumran. (Perhaps a later scribe preferred an older, archaic style of hand-writing.)

A Proto-Masoretic Text

The question everyone wanted to know about the Ein Gedi scroll before it was deciphered was:  What text did it contain?  If it was some sort of Torah text—a likelihood given where in the synagogue it was discovered—would it conform to proto-MT, or would it be, like some of the Qumran Scrolls, close to the Samaritan, or the Hebrew text underlying the Greek, or perhaps, like some Qumran scrolls, a mixture of these traditions? Most scholars expected it to be proto-MT, and this is what was found.

In cases thus far deciphered in which other versions (LXX and SP) differ from proto-MT, the Ein Gedi scroll is always identical to the proto-MT.  As the authors of the recent Textus article  observe (p.11):^[10] “Based on the evidence of these two columns, the Ein Gedi Scroll can be characterized as strictly proto-Masoretic.” 

A Leviticus Scroll (Not a Torah Scroll)

So far, only two columns of the scroll have been virtually unrolled.  Neither column is complete—each preserves only part of 18 of the likely 35 original lines: column 1 covers Leviticus 1:1-9, and column 2 covers Leviticus 2:1-11. 

Column one is preceded by a blank column of significant width, suggesting that it began the scroll.  In other words, contrary to what has been widely reported, this document is not a complete Torah scroll, but a Leviticus scroll.  This should not be surprising—no complete Torah scroll has been recovered from the Dead Sea Scrolls. We do have scrolls that incorporated two books such as Genesis and Exodus, but it is unlikely the leather or parchment technology existed in this period to write out the entire Torah on a single scroll—it would have been too heavy and unwieldy.

What We Learn from the Ein Gedi Scroll

In sum, the Ein Gedi scroll is not a Torah scroll in our sense.  Its find-site suggests that it was likely a Leviticus scroll used for the Torah reading and that, like the later MS Ashkar 2 scroll, its consonantal text conformed to the (later) MT. This supports the probability proto-MT had become standard soon after the destruction of the Second Temple—something we had already thought before this new Leviticus scroll was published based on the evidence of the second century CE Dead Sea Scrolls and the way that the rabbis expounded the biblical text.  But it is always good to get manuscript confirmation of this idea.

The authors of the Textus article write (p. 14) that the scroll’s “clear affinity with MT essentially confirms the model(s) of textual development that had already been developed by scholars before its discovery.” I would urge caution, however, before stating unequivocally that the proto-MT was the only text used by the Jewish community in this period. In fact, I would not have been totally shocked if the Ein Gedi manuscript did not fully represent proto-MT.

Questions Not Yet Unanswered 

We do not know exactly how, when, and why a single text was created for the Torah.^[11]  The question of how alternative texts—we know from Qumran and the evidence of the LXX that they certainly existed—were suppressed is even more difficult ponder: How in a pre-print age could alternative texts be suppressed (or recalled), and a single, standard text be published?  Perhaps archaeological finds, combined with new technology, will one day answer this question as well.^[12]