Prev Next

This is part 1 of 2 of a change from k1q1×2 to k2q2. UXLC rejected this change.

I think I understand the impulse behind this WLC change, but nonetheless I find the change inadvisable. So I not only support UXLC’s rejection of this change but also advise WLC to revert this change in some future version.

My guess is that the impulse behind this change is that we are primarily dealing with a word boundary issue here, so a single k2q2 seems more appropriate than two adjacent k1q1 constructs, a configuration we sometimes abbreviate as k1q1×2. (There is a ḥaser/malei issue later in the second word, but the primary issue is the word boundary issue.) The word boundary issue is about a difference between qere and ketiv as to where the word boundary falls within the letters ושוריאשכל[י]לו.) The table below shows the details.

Because we are primarily dealing with a word boundary issue, it is an understandable impulse to group these 4 words into a single k2q2 construct (which is what WLC 4.22 now has) rather than group them into two adjacent k1q1 constructs (which is what WLC 4.20 had). The table below shows what these two grouping strategies look like in Michigan-Claremont terms.

Although the impulse is understandable, it is not consistent with the diplomatic spirit of WLC to allow such an impulse to override what we see in the manuscript. What we see in the manuscript is the choice to group these words into two adjacent k1q1 constructs. BHS and BHQ agree with the manuscript, though of course we’d be willing to reject their reading if it contradicted the manuscript.

Other k2q2 in WLC

The table below shows two other k2q2 in WLC that are instructive to compare with this Ezra 4:12 one, since they, too, are about a word boundary.

In the case of Ezekiel 42:9, we find that the manuscript supports WLC’s k2q2 grouping (page 299B col 3 line 22):

But in the case of 2 Samuel 21:12, we find that the manuscript does not support WLC’s k2q2 grouping. Rather, it supports a different k2q2 grouping than that found in WLC! The manuscript’s k2q2 reaches back one word earlier than WLC’s k2q2 (page 181B col 2 line 12):

Using square brackets to set off the k2q2 grouping, here’s how WLC and the manuscript divide up the qere letters of these three words:

Let’s look at the six other k2q2 cases in WLC. None of these cases are about a word boundary, but it will still be interesting to see whether, in each case, the manuscript does or does not support WLC’s k2q2 grouping.

In the case of 1 Kings 17:15, we find that the manuscript does not support WLC’s k2q2 grouping (page 197B col 3 line 24):

In the case of 1 Sam 20:2, we find that the manuscript supports WLC’s k2q2 grouping (page 161B col 2 line 17):

In the case of Isaiah 52:5, we find that the manuscript supports WLC’s k2q2 grouping. It is slightly surprising that this k/q is framed as a k2q2 at all, since the second word of both qere and ketiv is לי, i.e. there is no variation in the second word. Here is the manuscript image (page 240B col 3 line 3):

In the case of Job 38:12, we find that the manuscript supports WLC’s k2q2 grouping (page 408A col 2 line 17):

In the cases of Job 38:1 and 40:6, we find that while the manuscript supports WLC’s grouping of the two qere words together, the manuscript does not support WLC’s treatment of the ketiv letters as a single word. In other words, the manuscript supports a k1q2 grouping rather than a k2q2 grouping.

In both 38:1 and 40:6, the paseq after the nun should not be interpreted as relevant to the ketiv, i.e. should not be be interpreted as splitting the ketiv into two words. I find it surprising and unfortunate that WLC went to the trouble of devoting a special bracket note, ]M, to these cases. A little more scrutiny might have revealed that there is only a mistake in transcription, not an anomaly in the manuscript. Here the WLC transcription has a little more work to do than usual because, as with all k/q, WLC is not just transcribing BHS (or later, ל) but also converting (in this case incorrectly) from pointed-ketiv, unpointed-qere representation to the opposite representation.

Note that, extraordinarily, in Job 38:1 and 40:6, the scribe has reiterated some of the pointing of the ketiv on the qere letters, pointing the מן of the qere with ḥiriq, gaʿya (meteg), and paseq. (I’m not sure why this pointing was done, and I don’t think it is relevant to the issue at hand, but nonetheless it seemed worth mentioning.) Here is the manuscript image for Job 38:1 (page 408A col 2 line 5):

Here is the manuscript image for Job 40:6 (page 408B col 2 line 13):

Having completed our review of k2q2 in WLC, we can say that some are supported by the manuscript, and some are not. I would urge WLC, in future versions, to use k2q2 only when supported by the manuscript.

k1q1×2, k1q1×3, etc. in WLC

It would be nice to also review all cases of multiple adjacent k1q1 in WLC to see if all of them are supported by the manuscript. I wonder whether, for example, some instances of k1q1×2 in WLC should in fact be represented as k2q2. I do not propose to do this (possibly rather large) survey at the moment, but I will discuss two cases here.

Here is the image for Proverbs 21:29 (page 417A col 1 line 14):

In WLC’s 2 Samuel 5:2, the k/q differences of the first two words are about a word boundary, but the manuscript does not support that. Indeed the first k1q1 in WLC, הייתה/היית, corresponds to a normal word in the manuscript, as the following image shows (page 171A col 1 line 20):

In other words the manuscript considers the ה of the qere המוציא to come from nowhere, not from the end of הייתה. Note that in ל, the pataḥ intended for this ה is an orphan, floating out in a lonely fashion before the ketiv letters מוציא.

There is an odd mark after והמבי. It is a bit like a maqaf, but the thickness of the stroke makes it look more like part of a letter than a maqaf. Is it the top right part of an א that was abandoned, and for some reason not erased? If it is part of an abandoned א, was this א the א of the qere or the א of the next word, את? (I don’t think this mark is relevant to the k/q issues at hand, but nonetheless this mark seemed worth mentioning.)

The Aleppo Codex (Ms א) is rarely relevant to WLC, but here it is perhaps useful to look at it. Ms א treats these words quite differently than ל does. Unlike ל, Ms א does consider the ה of the qere המוציא to come from the end of הייתה, as indicated by the following:

• היית rather than הייתה are the qere letters
• the final ה of הייתה is pointed with pataḥ, i.e. הייתהַ
• the qere words היית המוציא are grouped together

Also unlike ל, Ms א considers והמבי to be a normal word, not a ketiv needing a final א to form a qere. Ms א does note the lack of a final א, with “חס א” with a splotch above that may have been a numeral, perhaps a vav meaning “6 times lacking [a final] א”. All this can be seen in the following image:

The table below shows WLC’s pointed qere words and the pointed qere words implied by Mss ל and א.

We can summarize this table as follows:

• WLC agrees with א on word 1, היית.
• All agree on word 2, המוציא.
• WLC agrees with ל word 3, והמביא.

So we could say that WLC’s k1q1×3 in Samuel 5:2 is a mix of the traditions we see in ל and א. But according to its charter, WLC should hew to ל. As to why WLC diverges from ל here, I very much doubt that א actually influenced BHS. Rather, we can see this mix of traditions in the Second Rabbinic Bible (RB2) (Second Venice Miqraot Gedolot), so that’s almost certainly where WLC/BHS got this mix from (full page):

This case is typical of most cases where WLC/BHS blatantly diverges from ל. In most cases of blatant divergence from ל, WLC/BHS is conforming to its base text, which, like almost all printed editions before BHS, hewed close to RB2. In other words, this is a typical example of a pattern that, as a whole, betrays the methodology of BHS. Although one might imagine or hope that BHS was created by transcribing ל from scratch, in fact BHS was created by starting with an RB2-type text and editing it in most but not all places to conform to ל.

In a similar way, most modern, general-purpose Hebrew Bibles were created by starting with some version of WLC and then editing it in most but not all places to conform to general-purpose expectations. This methodoloy is often not publicly disclosed, but certain telltale WLC artifacts leak through, betraying this methodology.

Conclusion

Finally, we should admit that qere grouping may have been left up to individual scribal discretion. In other words, qere grouping may not have been viewed as meaningful by the Masoretes. Just as we do not distinguish a normal lamed letter from the occasional elongated one, perhaps we should not be concerned to preserve qere grouping, particularly when we don’t like the manuscript’s grouping.

That having been said, it seems more consistent with the diplomatic spirit of WLC to err on the safe side by preserving qere grouping. For example this is what BHS did, and WLC continues to do, with respect to gaʿya (meteg) placement. In the 50-or-so years since the publication of BHS, scholarship has reached a consensus that gaʿya placement was not viewed as meaningful by the Masoretes. I.e. the consensus is that gaʿya placment was left up to scribal discretion. Still, at the time, perhaps it was the right decision for BHS to err on the safe side by preserving gaʿya placement. I urge WLC to follow this example and err on the safe side by preserving qere grouping.