Whilst I can’t exactly fathom how I missed such a Dead Sea Scrolls Manuscript (I mean, it’s not like I haven’t done transcriptions for the other Wadi Murabba’at scrolls… oh wait, yes I did!), it does however give a good opportunity to discuss the Dead Sea Scrolls and the transcription process for TWTY, using this very manuscript as an example.
Prior to 2011, the only way to see manuscripts of the Dead Sea Scrolls (which include those from Wadi Murabba’at, Wadi Sdeir, Nahal Hever etc.) would either a) be a scholar and have access to them for a scholarly article/Masters/Ph.D.; b) have a copy of one of the numerous volumes from the Discoveries in the Judaean Desert series (DJD), which range from the ‘oh that’s not that expensive (£96)’, to the ‘OH MY WORD I’LL NEED TO REMORTGAGE MY BLASTED HOUSE! (£282.50)’ – you’re looking at £6,000+ for the entire collection (40 Volumes at the moment); or c) find a book about a Dead Sea Scrolls manuscript produced by a scholar, that also includes the image(s) (or facsimile as a reproduction of a manuscript is more commonly known).
Obviously, getting all those volumes and producing a free to use transcription of the facsimiles would not have been an easy job, but there was some earlier assistance, and that was the production of the Dead Sea Scrolls Bible (DSSB), published in 1999. This at least gave the mere ‘public’ the opportunity to see a translation of the Scriptural Dead Sea Scrolls manuscripts, which also provided lots of supplementary information, such as a list of the names or designations given to Dead Sea Scrolls manuscripts (such as 4QGenn, 3QLam etc.), and a glimpse into the Hebrew letters or words extant in the manuscripts. This gave people the ability to use the list of manuscripts, and shown extant or non-extant letters, to create preliminary transcriptions of the manuscripts in question, as the Masoretic Tanakh was already digitised by this time. This unfortunately didn’t give anyone the capacity to produce a proper transcription of the manuscripts, as how many letters per line etc., etc., was not given in the DSSB.
It was, however, a fantastic start.
Nevertheless, role on 2011, and in conjunction with Google, the Israel Museum was able to provide a high-resolution ultra-version of several of the larger Dead Sea Scrolls, one of which was the Great Isaiah Scroll. Being able to see quite probably the most famous Dead Sea Scroll in all of its glory was a fantastic achievement for both.
It wouldn’t stop there, however. Again in conjunction with Google, the Israel Antiquities Authority released The Leon Levy Dead Sea Scrolls Digital Library (LLDSSDL) online. This gave photographs (both original and newer ones, and they have continued to provide more recent ones too), of nearly all the Dead Sea Scrolls discovered. To call this a ‘treasure trove’ would be a huge understatement.
Clearly the Internet has done a lot of good to assist with Dead Sea Scrolls research (especially the behemoth that is Google), and before 2011 you could quite easily find images of several of the Dead Sea Scrolls dotted around online. But not nearly as much as the collection now on the LLDSSDL.
In the recent release of the Dead Sea Scrolls Image Sources page on TWTY, I spent quite a few days going through all the images again, and came across several things I had missed, one of which was an entire manuscript designated MurIsa (short for Wadi Murabba’at Isaiah)! How it was missed I cannot explain. Notwithstanding, checking through my copy of the DSSB and some much earlier notes on the Dead Sea Scrolls, I did find a few peculiar things that might solve the mystery.
In its original discussion of the Isaiah manuscripts among the Dead Sea Scrolls, the DSSB states this (image taken from the book itself), which I have underlined the reference to Wadi Murabba’at Isaiah:
So the DSSB mentions it in the initial discussion. But what about when it comes to the actual content of the manuscript?
At the start of each chapter from a Scriptural Book, the DSSB gives a list of the manuscripts that contain words from the chapter, and how many verses it has extant.
One should expect therefore to find Wadi Murabba’at Isaiah listed somewhere at the start of one of the chapters of Isaiah. I say one should expect, but in this case, we don’t!
Wadi Murabba’at contains text from Isaiah 1:4-14. Here’s what the DSSB has for this section:
As can be seen, this is somewhat bizarre. MurIsa isn’t given in the list of manuscripts containing text from Isaiah Chapter 1, yet is given in the notes on variant readings from the main one used in the DSSB! This may explain why I found an old note in one of my notebooks that just says “MurIsa?” Needless to say that even now, information on MurIsa is somewhat scarce (most places I’ve found just give the name, but no indication of verses or content), so whilst not a great excuse, it goes someway to explaining how MurIsa was missed for so long a time.
Let’s have a look at the manuscript in question, and describe the transcription process used for TWTY.
Using the infrared image on the LLDSSDL, we can see that the manuscript has quite a few letters and words that are very easy to read. But how can we determine that this manuscript contains text from Isaiah, and not some other place in the Tanakh that has similar words? Perhaps it isn’t even a Scriptural book at all?!
The first full word that can be seen is on line 2 (I’ll explain why I don’t refer to it as line 1 later on), which reads as אחור. A search in the Tanakh shows that there are only 24 places that have the word אחור. This therefore narrows our choices for this manuscript to Gen 49:17; 1 Sam 1:22; Isa 1:4, 28:13, 42:17, 44:25, 50:5, 59:14; Jer 15:6, 38:22, 46:5; Psalms 9:4, 35:4, 40:15, 44:11, 19, 56:10; 70:3, 78:66, 129:5, 139:5; Lam 1:8, 13, 2:3. The final choice is that this isn’t a manuscript containing text from any Canonised Scriptural book, but is a different non-canonised book or previously unknown composition instead.
Our next full word is on the next line, and is כל. This is a very common word in the Tanakh, so doing a search on this would be somewhat fruitless. What we can glean from this though is that the manuscript most likely contains the defective (short) rather than the plene (full) spellings of words.
Right next to כל is ראש meaning head. Now on its own, ראש is seen 99 times in the Tanakh, reducing our choices quite a lot, but how many times is it preceded by the common noun כל? The answer is four times: Isa 1:5; Jer 48:37; Ezek 29:18; and Amos 8:10. This reduces our pool significantly, for if we compare to our previous extant word, there is only one viable conclusion: we’re likely looking at a manuscript of the book of Isaiah.
Could it be another composition? Possibly, but directly following כל ראש is one extant letter, and part of another. The extant letter is ל, and in the text of Isaiah, what directly follows כל ראש is לחלי, granting a big clue that we’re looking at a direct copy of the text of Isaiah. The rest of the extant words and letters also fit in exactly with how the text of Isaiah has descended through the centuries.
From here we can start to build the transcription.
Based on the easiest to see letters, the following can be given as to what the text of the manuscript looked like:
Clearly this isn’t a ‘transcription’ of the manuscript just yet; there’s no indication of missing letters etc., but this is all added later on. Regardless, as we can fit the text of Isaiah into what is seen on the manuscript, we can start to deduce several things about it:
1) From the extant text, there are no variants from what came to be the standard Masoretic Hebrew Text of Isaiah. Therefore it is more than likely there’s no variants in the non-extant portions either.
2) The line length ranges from around 22 letters to 30 letters, with an average amount of 6 words per line.
3) Like other Hebrew manuscripts, words are not split up from line to line. This is in contrast to Greek and Latin works, which happily split words up from one line to the next.
All the above helps to determine the non-extant text. We can then produce the following:
Now I’ve left some lines (3, 4, 5, 7, 8, and 9) without indication of where the final extant letters appear. This is due to the fact that there’s at least a bit of several letters that are visible, but not in their entirety.
Looking at line 3 in the image, we can clearly see the letters כל ראש ל, but can also see the right-hand side of another letter:
As we learned before, the rest of the word following ל is most likely חלי, with the right hand side of the letter extant above being from ח. If we compare this to how the fully extant ח are within the manuscript (line 2, line 15), then this becomes even clearer.
Applying this to the rest of the just-about-extant letters seen, we can produce the transcription of lines 2-10 thusly:
On lines 4 and 5, there’s a bit of scarring on the manuscript that has rubbed away two full letters, and quite a good chunk of a few more, but these at least have left some evidence. On lines 7 and 8, we can just about see the bottom or tops of the last couple of letters; in this case there is enough left of them to give more than just a ‘?’ in their place.
The really tricky part comes in determining exactly which letters we can see in those lines that have barely a dot left of a letter.
The reason I’ve not designated the line beginning with אחור as line 1 is due to the fact that there’s a bottom right corner dot of one letter just above the initial א of line 2:
But what letter is this? Is it even possible to determine what letter it can be from just a small dot as this? Well, yes is the answer; although there is always a very certain degree of doubt as to the identification. This however is where knowing the line-length comes in handy.
We know from the earlier discussion that the line length is around 22-30 letters, averaging at around 6 full words (though there is a minimum of 5, max of 9 based on the word-length). Also, the extant portion, as mentioned, demonstrates that this manuscript contains no variants. Therefore we can deduce that this dot is from a letter, 22-30 letters preceding אחור in Isaiah 1:4, at an average of around 6 full words.
The 6 full words prior to אחור in Isaiah 1:4 are the following:
יהוה נאצו את קדוש ישראל נזרו
Not only is this 6 full words, but also 23 letters which is within our range of 22-30. But is this right? Can that dot be the bottom of the letter י at the start of יהוה? Checking for the fully extant י’s in our manuscript (see lines 18 & 19), this dot is too low and too far to the right-hand side for it to be the bottom of a י. We’re looking really for a letter that has a sort of hook to the bottom right.
If we take our maximum amount of letters (30), and apply it to the words before אחור, we end up with עזבו את יהוה נאצו את קדוש ישראל נזרו. Here we have 29 letters, 8 words, right within our range once more. Unfortunately we have no extant ע’s in our manuscript to see how the scribe would’ve written the letter. Therefore is this an ע? Here we need to look at similarly dated, and similarly styled manuscripts to see how they wrote the letter ע to give a guestimate on how the letter may’ve looked in MurIsa.
MurIsa has the closest similarity with MurDeut, both of which are dated to have been produced 20-84 CE in a post-Herodion style of writing. Thankfully, MurDeut has some extant ע’s that can be looked at.
Taken from Fragment 1 Column 1:
Here we can see the style of ע that was prevalent in post-Herodian style handwriting: a long, sloping right line, with an almost vertical and smaller left line going up from the middle of the right line. This therefore removes the letter ע and the word עזבו as being what we can see extant in the first line of MurIsa; the letter ע wouldn’t have left a bottom-right mark.
What would be our next choice? As noted above, we could easily remove the word עזבו from our sentence, giving us את יהוה נאצו את קדוש ישראל נזרו. Not only is this still within our range (25 letters, 7 words), but the letter א, as seen numerous times in MurIsa, always leaves a bottom-right hook.
Therefore the dot visible in line 1 is from the bottom-right corner of the left downward stroke of the letter א.
We can then give the following:
This fits what is detectable in MurIsa, but also all the available ranges and evidence from other manuscripts that are at our disposal.
One of the last things to decide upon is what letter we can just see the top-right hook of in line 11:
Again, if we take the average line length, we can leave lines 11-13 for the time being and look at what can be decided for from line 14 onwards:
We begin in line 14 with some text from near the start of v11. This indicates that the start of v11 must come in one of the lines above, more than likely line 13 directly above. Therefore as נצורה at the start of line 10 is the last word in v8, then what we must have in lines 10-13 is text from Isaiah 1:9-10.
Taking our average line lengths, and omitting the amount of letters that can be surmised from the transcription, line 10 has 17-25 letters left to fit in, and line 13 has 15-23 left to fit in. Adding in the total possible letters for lines 11 and 12 into the equation, this leaves us with around 76-108 letters from verses 9-10 to place within the brackets.
Accepting the evidence from the extant letters, and the conclusion that this manuscript has no variants from the standard text of Isaiah, the total number of letters in Isaiah 1:9-10 is 89, meaning we’re looking at line lengths of around 22-26 letters for lines 10-13.
One has to presume therefore that the letter we see at the start of line 11 is from somewhere in v9. As the manuscript shows no evidence of gaps between verses, v9 will have started straight after the final word of v8 in line 10. We are also looking at a letter that has a sort-of right-slopping top-right hook.
Accepting the text of v9 as follows:
לולי יהוה צבאות הותיר לנו שריד כמעט כסדם היינו לעמרה דמינו
Then comparing these letters to the extant ones we see in MurIsa, there are very few letter candidates that could produce something as seen in line 11, with the only one I would argue being the letter ש. There’s more than a few extant ש’s in MurIsa, which look like this:
As is distinguishable, though never written exactly alike, the start of the right-side hook is nigh-on perfect for what we can see at the start of line 11. Whilst not 100%, the likelihood that it’s another letter is quite low, as no other letter gives such a right-curving mark.
The transcription of MurIsa is now able to be completed, and is as below:
The above detailed process actually doesn’t take as long as I’ve taken to explain it; most of the decisions are made very quickly once a base-text has been established for the manuscript in question. If this was an unknown composition, then what I’ve explained above would be extremely complicated, and it’s very difficult to conclude what the missing letters are, or what letter one can just about see scrapes and dots from. Thankfully we do actually have a base text to work from, making transcriptions of the Scriptural manuscripts a lot easier.
Thus wraps up the transcription process used for the manuscripts seen on TWTY, and the discussion of MurIsa. Hope it’s been informative, and not quite as mundane or boring as such explanations can be 🙂