The Israel Bible: this ain't no Koren edition

The Israel Bible (2018) comes from Menorah Books, which its copyright page calls an imprint of Koren Publishers Jerusalem. Its acknowledgments praise Koren as Israel's foremost Bible publisher and thank Koren's staff by name. One might assume from all this that the Hebrew text of the Israel Bible is the same as that in other Koren editions of the Tanakh, but it is not. Not even close.

The Hebrew text of the Israel Bible is WLC, version 4.18, as tanach.us distributed it in November 2014 and as Sefaria imported it in early 2015, when Sefaria's import script stripped off its accents along with other content such as its bracket-notes.[1] The Israel Bible's website shows, by default, the same Hebrew text as the printed book. The Israel Bible's copyright page credits the JPS Tanakh for the English; it credits nothing for the Hebrew. The website's "About" page does name a source, MAM, but MAM is only the text of a "Cantillation" layer that the website hides by default.[2]

So the Israel Bible, sold under the name of a prominent Israeli Bible publisher, reproduces, without any editing,[3] an outdated diplomatic edition of a single manuscript, the Leningrad Codex. A diplomatic edition tries to show one manuscript as it stands, quirks and all. That is the right thing for a student of that manuscript and the wrong thing for almost everyone else. A general-purpose Bible should give its readers the Masoretic consensus, with the reading aids that Jewish editions have long provided. Notably for an explicitly Zionist edition, the Israel Bible ignores the work of the great Israeli Bible scholars like Breuer, Dotan, and Yeivin.

I refer to WLC as "outdated" above, but WLC's quality as a diplomatic edition is beside the point here.[4] The point is the Israel Bible's many departures from the Masoretic consensus, whether they come from a good transcription of the word in question or a bad one. A departure that comes from a bad transcription is slightly more frustrating, but that doesn't matter much.

The Israel Bible leaves out the accents, a reasonable choice for readers who do not chant, though this forfeits the help with stress that most accents provide. Some prayer books provide stress help by marking nonfinal stress with meteg/silluq. The Israel Bible provides no such help, despite the interest in pronunciation that it demonstrates through its many transliterations, which put the stressed syllable in capitals.

And beyond stress, the Israel Bible gives no help with the following distinctions:

Such distinctions are made in Koren editions, so here the misleading association with Koren hits particularly hard.

A sampling of quirks

In each of the 48 rows below, the Israel Bible's form is shown above a consensus form of the same word. (I take the consensus form to be MAM's form.) MAM's form is shown without accents or meteg, to provide an "apples-to-apples" comparison with the Israel Bible's form.

Another word

FormsWhereNotes
אֶל־Exod 10:28
page 176
IB has a segol instead of a pataḥ under the alef: the preposition "to" instead of the negative "do not". Judg 19:23 and Jer 47:6, below, have the reverse. The Israel Bible transliterates every verse that has a study note, 1,084 verses in all, Deut 2:9 among them, and there its transliteration has al (al ta-TZAR, do not harass): it follows the consensus, against its own Hebrew.
אַל־
אֶל־תָּצַרDeut 2:9
page 452
אַל־תָּצַר
אֶל־תִּמְרֹדוּJosh 22:19
page 595
אַל־תִּמְרֹדוּ
וַיֶּחֱזַק־Judg 19:4
page 655
IB has, on yod–ḥet–zayin, segol–ḥsegol–pataḥ instead of pataḥ–ḥpataḥ–segol: the vowels of the qal instead of those of the hiphil.
וַיַּחֲזֶק־
אַל־בֵּיתִיJudg 19:23
page 657
IB has a pataḥ instead of a segol under the alef: the negative "do not" instead of the preposition "to", the reverse of Exod 10:28, Deut 2:9 and Josh 22:19.
אֶל־בֵּיתִי
אַל־Jer 47:6
page 1197
אֶל־
הַמַּלְאכִים2 Sam 11:1
page 767
IB has, on mem–lamed, pataḥ–sheva instead of sheva–qamats: the vowels of "the messengers" instead of those of "the kings".
הַמְּלָאכִים
הַחֳדָשִׁיםIsa 66:22
page 1083
IB has a ḥataf qamats instead of a ḥataf pataḥ under the ḥet, the vowel of "the months" instead of "the new".
הַחֲדָשִׁים

Two vowels on one letter

FormsWhereNotes
תְִּירָאוּםDeut 3:22
page 457
IB partners the consensus ḥiriq with a sheva that is nonsensical, non-robustly encoded, and typographically mishandled.
DetailsThe Israel Bible fails to use a CGJ (as at Exod 20:4, below), and thus leaves itself open to variable presentation depending on whether the environment does Unicode normalization or not. Normalization puts the sheva first, which in some fonts gives a false "ḥataf ḥiriq" look.[5]

As it turns out, the order of the two marks is immaterial due to IB's font's mishandling of this (admittedly outlandish) combination of marks. In IB's PDF (I have not inspected the paper edition), the font draws the ḥiriq and the sheva superimposed: the ḥiriq's dot lands almost on the sheva's upper dot, so the page shows something like a sheva with a misshapen top dot.

The word as page 457 of the Israel Bible's PDF draws it
תִּירָאוּם
קוָֹיIsa 49:23
page 1054
IB has a nonsensical ḥolam male dot on vav rather than a ḥolam ḥaser dot on qof.
קֹוָי
לְקוָֹוLam 3:25
page 1813
לְקֹוָו
וִּכְיוֹםHos 2:17
page 1347
IB has a nonsensical ḥiriq below a shuruq.
וּכְיוֹם
וִּקְרָאJonah 3:2
page 1394
וּקְרָא

Missing vowels and dots

FormsWhereNotes
אָבִיהGen 34:11
page 102
IB has a final he without qamats.
אָבִיהָ
סֹחֲרֶיהIsa 23:8
page 1003
סֹחֲרֶיהָ
מִפָּנֶיהHos 2:4
page 1346
מִפָּנֶיהָ
חַיֶּיהProv 31:12
page 1700
חַיֶּיהָ
מימֵי רַגְלֵיהֶם2 Kgs 18:27
page 940
IB has no tsere on the first mem of this qere.
מֵימֵי רַגְלֵיהֶם
שֵיבָהIsa 46:4
page 1048
IB lacks the sin dot that appears in the consensus.
שֵׂיבָה
וִירוּשָׁלָםZech 2:2
page 1436
IB has the Jerusalem-related word without its orphan ḥiriq.
וִירוּשָׁלִָם
הַיַּיןProv 20:1
page 1679
IB has no ḥiriq under the second yod.
הַיַּיִן
אֶל־חכּוֹLam 4:4
page 1816
IB has no vowel under the ḥet.
אֶל־חִכּוֹ
לִבּוEccl 2:23
page 1827
IB has a bare vav instead of a ḥolam male.
לִבּוֹ

A sheva instead of a full vowel, or the reverse

FormsWhereNotes
וּבְנְיָמִןExod 1:3
page 151
IB has a sheva instead of a ḥiriq under the bet.
וּבִנְיָמִן
וְמַטֵּהNum 2:14
page 343
IB has a sheva instead of a shuruq dot.
וּמַטֵּה
וְפֻםDan 7:20
page 1891
וּפֻם
וּפְלִשִׁתִּיםJudg 15:14
page 647
IB has a ḥiriq under the shin instead of a sheva.
וּפְלִשְׁתִּים
וַהֲרְגְנֻהוּJudg 16:2
page 647
IB has a sheva instead of a pataḥ under the resh.
וַהֲרַגְנֻהוּ
הַיַּרְחְמְאֵלִי1 Sam 27:10
page 737
IB has, on yod–resh, pataḥ–sheva instead of the reverse.
הַיְּרַחְמְאֵלִי
וּבְבְכִיJoel 2:12
page 1367
IB has a sheva instead of a ḥiriq under the first bet.
וּבִבְכִי
בְּשְׂפְתוֹתֶיךָPs 45:3
page 1516
IB has a sheva instead of a ḥiriq under the sin.
בְּשִׂפְתוֹתֶיךָ
מִמְּתַכַּבֵּדProv 12:9
page 1665
IB has, on the mem–tav, sheva–pataḥ instead of ḥiriq–sheva. (Unlike IB's vowels, the ḥiriq–sheva vowels of the consensus make sense, being the vowels of the hitpael participle's prefix.)
מִמִּתְכַּבֵּד
הוֹדַעַתָNeh 9:14
page 1956
IB has a pataḥ instead of a sheva under the ayin. IB has no dagesh in the tav.
הוֹדַעְתָּ

Qubuts-alef-shuruq

FormsWhereNotes
הָרֻאוּבֵנִיNum 26:7
page 412
IB not only includes the LC's quirky qubuts, it does so without the rafeh on alef that makes this quirk less confusing (though still confusing). Its transliteration of this word as la-ru-vay-NEE (in Josh 12:6) is hard to justify against its Hebrew pointing. IB's 2 Kgs 10:33 form of this word lacks a dot in the vav; whether the word is more or less confusing this way is unclear. It certainly adds to the general confusion surrounding this word, which also appears in near-consensus form in IB 12 times. (By "near-consensus form" I mean with no qubuts, albeit with no rafeh either.) See also מאום at Job 31:7, below. See MAM's relevant FOI.
הָראֿוּבֵנִי
לָרֻאוּבֵנִיDeut 3:12
page 455
לָראֿוּבֵנִי
וְלָרֻאוּבֵנִיDeut 3:16
page 456
וְלָראֿוּבֵנִי
לָרֻאוּבֵנִיDeut 4:43
page 462
לָראֿוּבֵנִי
לָרֻאוּבֵנִיJosh 12:6
page 571
לָראֿוּבֵנִי
וְהָרֻאובֵנִי2 Kgs 10:33
page 918
וְהָראֿוּבֵנִי
מֻאוּםJob 31:7
page 1752
IB not only includes the LC's quirky qubuts, it does so without the rafeh on alef that makes this quirk less confusing (though still confusing). See also ראובני at Num 26:7, above.
מאֿוּם

Miscellaneous

FormsWhereNotes
וִשְׁאֵלְךָGen 32:18
page 98
IB has a ḥiriq instead of a shuruq dot.
וּשְׁאֵלְךָ
אִם־אַמְנוֹן2 Sam 13:33
page 776
IB has a ḥiriq in a word that is כתיב ולא קרי, so a reader would be encouraged to read it aloud. Elsewhere the Israel Bible leaves such a כתיב ולא קרי word unpointed (six places), with nothing else to mark it; it brackets only the qere words that it adds. At 2 Sam 15:21 the Israel Bible also has a maqaf after it. These two words are the subject of my "Saying the quiet part out loud".
אם־אַמְנוֹן
כִּי אִם־2 Sam 15:21
page 781
כִּי אם
וְנִקְוּוּJer 3:17
page 1094
IB has a dot in the first of the two final vav letters.[6]
וְנִקְווּ
קְוּצּוֹתָיוSong 5:11
page 1785
IB has a shuruq dot in the first vav instead of a qubuts under it. That vav is then a shuruq, a vowel directly after the qof's sheva, which makes no sense.
קְוֻצּוֹתָיו
וְהַדֶּקֶתDan 2:45
page 1875
IB has a segol under the dalet instead of a tsere. IB's transliteration, v’-ha-DE-ket, spells the syllable with e, its spelling of segol. So its transliteration follows its Hebrew here, not the consensus.
וְהַדֵּקֶת
שְׁלִיַחַDan 5:24
page 1886
IB has pataḥ under yod.
שְׁלִיחַ
דָויִד1 Chr 3:1
page 1982
IB has ḥiriq on yod rather than vav.
דָוִיד

The Decalogues

The Decalogues are among the most challenging passages of the Bible to present. So, it is no surprise that the Israel Bible did not rise to the challenge. Each Decalogue has two cantillations, the טעם תחתון and the טעם עליון, which differ in a few of their non-accent marks as well as in their accents. (It is a little-known fact that the accent system interacts with the non-accent marks, i.e. the accent system is not, as an edition like IB pretends, fully separable from the system of vowel marks and dagesh.) In each of the 8 rows below, the Israel Bible's form is shown above MAM's עליון form. (We choose the עליון because it is the closer of the two cantillations to IB's text.)

FormsWhereNotes
עַל־פָּנָיַExod 20:3
page 198
IB has the תחתון's qamats under the nun, and the עליון's pataḥ under the yod rather than the nun.
עַל־פָּנַי
עַל־פָּנָיַDeut 5:7
page 463
עַל־פָּנַי
מִתַָּחַתExod 20:4
page 198
IB has both the תחתון's qamats and the עליון's pataḥ under the tav. At Deut 5:8, below, it has the qamats alone.
DetailsIB does not use a CGJ between the two vowels (as at Deut 3:22, above), so its encoding is not robust to normalization, though that doesn't matter since for display purposes, IB got kind of lucky and its PDF (I have not checked the printed page) draws them at the same point: in its Hebrew typeface a qamats is a pataḥ with a stem added, so the pataḥ lands exactly on the qamats's bar and vanishes. Anyone who copies or searches the PDF's text, or the website's, still gets the pataḥ.
מִתַּחַת
כָּל־מְלַאכְתֶּךָExod 20:9
page 198
IB has the תחתון's dagesh in the kaf. Correctly reflecting an error in the LC, WLC 4.18 has no rafeh here. Contrast with Deut 5:13, below.
כָל־מְלַאכְתֶּךָ
לֹא־תַעֲשֶׂהExod 20:10
page 198
IB has the תחתון's maqaf.
לֹא תַעֲשֶׂה
לֹא־תַעֲשֶׂה־לְךָDeut 5:8
page 463
לֹא תַעֲשֶׂה־לְךָ
מִתָּחַתDeut 5:8
page 463
IB has the תחתון's qamats under the tav instead of the עליון's pataḥ. At Exod 20:4, above, it has both. See also my note on the LC at Deut 5:8.
מִתַּחַת
כָּל־מְלַאכְתֶּךָDeut 5:13
page 463
IB has the תחתון's dagesh in the kaf. Correctly reflecting the LC, WLC 4.18 has an important rafeh here, but Sefaria stripped it.[7]
כָל־מְלַאכְתֶּךָ

In a long line

The Israel Bible is the latest in a long line of Hebrew Bibles slapped together from a Unicode conversion of WLC. One expects that of free websites. One expects better of printed books, which cost real money to make and to buy, but the line includes those too:

  1. Zondervan's A Reader's Hebrew Bible (2008) names its source, WLC version 4.4, and marks the places where WLC differs from BHS. Various follow-ons to this publication exist, including a second edition of this Reader (2026), as well as A Reader's Hebrew and Greek Bible, whose third edition came out in April 2026. But I assume that these follow-ons continue to use WLC, probably not even a more recent version of WLC.
  2. Hendrickson's A Hebrew Reader for the Pentateuch (2024) says in its preface that its Hebrew is WLC. This is the book I reviewed in "WLC Considered Harmful". The same publisher's A Hebrew Reader for the Psalms (2021), mostly a fresh transcription from the Aleppo Codex, shows what it could have done instead.
  3. Crossway's The Hebrew Old Testament, Reader's Edition (2021) cites its use of WLC: in its publisher's words, it "combines the text of the Westminster Leningrad Codex (WLC)" with glosses. With no qualifier such as "based on", that implies WLC's text verbatim.
  4. Timothy A. Lee's Hebrew Bible Readers say on their copyright page that their base text is based on WLC, and I can attest that "based on" means quite closely based on, i.e. with little or no editing.
  5. GlossaHouse's illustrated series does not say where its Hebrew comes from, but a review of its Genesis volume says that it is WLC with vowels and without accents, which is just what the Israel Bible gives its readers.

Use of ḥataf vowels under non-gutturals

One expects to find an occasional "ḥataf-NG": a ḥataf vowel under a non-guttural, i.e. under a letter other than the אהח״ע letters (alef, he, ḥet, and ayin). The Masoretes held a ḥataf necessary to mark a vocal shewa under a guttural but optional under other letters, and the manuscripts differ widely in how often they have one there (Yeivin, ITM §391). Cases of ḥataf-NG appear more frequently in IB than in a consensus edition such as MAM. These "excess" appearances, beyond the consensus, are not wrong, but they may surprise an intermediate student of Biblical Hebrew.

The counts below include resh but leave out the divine-name-related cases that might be included if their implied qere is ignored.[8] IB has 431 cases of ḥataf-NG, 225 of which are in agreement with MAM.[9] None of the cases in agreement with MAM is a ḥataf pataḥ: under a non-guttural, MAM has only a ḥataf qamats or a ḥataf segol. That leaves 206 cases of excess ḥataf-NG in IB, 97 of them under resh alone. (MAM has only one ḥataf under resh, in רֳאִי at Gen 16:13, and IB has this one too.)

Unlike IB, MAM has a special notation for a ḥataf under a non-guttural in authoritative manuscripts. That notation is the varika mark, which MAM has 674 times, as on the lamed of MAM's וּלְﬞהַבְדִּיל at Gen 1:18, where IB has a ḥataf pataḥ. So, though MAM has only a simple sheva rather than a ḥataf vowel on these letters, the varika mark records the ḥataf vowel that might surprise the intermediate student. MAM marks 194 of IB's 206 cases of excess ḥataf-NG with the varika, which makes something close to agreement between the editions for those cases.

Conclusion

Through a series of examples, we have demonstrated the perils of just slapping together a Hebrew Bible using WLC. The result is nothing like a Koren Bible, which is a consensus edition biased towards the printed tradition and providing pronunciation aids that have become common. Nor is the result like MAM, a consensus edition biased towards the manuscript tradition but not slavishly diplomatic to any one exemplar of that tradition.[10]

Related work

I have written about WLC before, in my series "Undoing and redoing the work of the Masoretes" and in "WLC Considered Harmful", my review of A Hebrew Reader for the Pentateuch. That review lists the quirks the Reader inherited from WLC, and every one of them that survives the stripping of the accents is in the Israel Bible too.

Other works by me about the Masoretic Text can be found in this index of my documents.

Notes

  1. Sefaria's import script, parseLeningradCodex.py, read tanach.us's XML (build 24.2), leaving out WLC's bracket-notes, and, for its "Tanach with Nikkud", applied a regular expression that deletes all accents, meteg, rafeh, the paseq code point, the extraordinary upper and lower dots, and every parenthesized one-letter marker: the setumah and petuḥah markers and the inverted nuns.
  2. We suspect that MAM, rather than IB, is the source of the website's transliterations! Our unpublished research of 2026-11-10 on the Israel Bible's transliterations found that the website began to transliterate every verse in August 2023, when it began to show MAM's text, and that where MAM's text and WLC's differ, its transliterations mostly follow MAM.
  3. The Israel Bible added a ḥolam to YHVH-Adonai, wrote the ḥolam ḥaser of a consonantal vav (as in מִצְוֺת) as a sin dot (!), and, in Jerusalem-related words, inconsistently separated the orphan ḥiriq from the lamed's own vowel with a hair space. Of the 633 words with the orphan ḥiriq, 144 have no hair space, almost all of them in Joshua through 2 Kings, and of the four ירושלמה words, whose lamed has an orphan sheva instead, only 1 Kgs 10:2 has a hair space.
  4. WLC is not even a particularly good diplomatic edition. It began as a digital copy of the body text of BHS, a diplomatic edition of the LC first published over the decade 1967–1977. WLC has not evolved much beyond BHS. BHS, though a great achievement when it was first published, is, by today's standards, no longer a particularly good diplomatic edition. (BHS was revised in 1984, 1990, and 1997. What changed in those revisions is unknown, and this lack of transparency is one of the many criticisms that can be levelled against BHS.)

    Worse, the Israel Bible presents this diplomatic text without ever expressing any surprise or doubt, and a diplomatic presentation that never does so is somewhere between nonsensical and irresponsible. Of the 48 words sampled in "A sampling of quirks", 45 are marked with surprise or doubt in at least one of three sources:

    • WLC 4.20's bracket-notes call 28 of them anomalous or of questionable validity
    • MAM notes surprise or doubt in 14 of the 21 that are in sections where MAM relies on the LC
    • Dotan's BHL marks its reading of 3 as doubtful, and effectively marks 43 of them with surprise by banishing them to Appendix A.
  5. A ḥataf ḥiriq appears five times in the Aleppo Codex: see near-Aleppo.
  6. This is a known mark of the LC and of the Aleppo Codex, though they write it only sometimes: a dot in a consonantal vav before a shuruq. Yeivin, as MAM's introduction (chapter 2) quotes him, takes it to show that the vav is pronounced almost as a vowel. Dotan calls it "the mappiq point in the letter waw followed by the vowel u", counts twenty-one words with it in the LC, and leaves it out of BHL's text (Foreword, p. xiv); MAM leaves it out too, as Breuer did. The Israel Bible keeps it and explains nothing, so its readers can only take it for a dagesh or a shuruq.
  7. WLC itself has since dropped this rafeh too. WLC 4.22, released in August 2020, removed all twelve rafeh marks that WLC 4.20 had, the seven in the Decalogues among them. See my notes on this change in WLC 4.22.
  8. The cases left out are the ḥataf segol under the yod of YHVH-Elohim, and, at Ps 144:15, IB's ḥataf pataḥ under the yod of YHVH-Adonai and MAM's ḥataf segol under the shin prefixed to it.
  9. MAM has 6 cases of ḥataf-NG where IB has none. At 3 of them IB has a simple sheva (Ezek 44:13, Ezra 4:9, and 2 Chr 5:7), and at the other 3 a full vowel: a qamats (Num 35:20 and 1 Sam 26:10) or a segol (2 Sam 6:5).
  10. MAM even occasionally diverges from the Aleppo Codex, the greatest exemplar of the manuscript tradition.