The London Codex and the Scribe Who Hid His Name

Somewhere in the Near East, early in the tenth century or perhaps before, a scribe finishing a large Torah codex decided to sign it without writing a colophon. He hid his name in the margins instead. Among the Masoretic notes that collect rare words, each list could be written in any order, so he arranged the entries so that their initial letters spelled out Nissi ben Daniel ha-Kohen, followed by a blessing, “may God preserve him.” Then he added, written out in full, the phrase “according to the name of the scribe, vocaliser and masoretic annotator.” Kim Phillips, who has been “excavating” the codex layer by layer, notes that the scribe did this three times.¹ The British Library’s description lists the places where the name is inscribed in the Masoretic notes: folios 40r, 113v and 139r.²

Centuries later the beginning and end of his codex were lost. In 1540 someone supplied the missing Genesis and Deuteronomy on paper, in a Yemenite square script, and an owner wrote a note in Persian, in Hebrew letters, on the last leaf. That note is now illegible.³ On 9 May 1891 the British Museum bought the composite volume, and it received the shelfmark Oriental 4445.⁴ Within a decade Robert K. Douglas, the Museum’s Keeper of Oriental books, introduced its Hebrew catalogue by calling it “the earliest known copy of the Hebrew Pentateuch.”⁵ It is often called the London Codex.⁶

That claim no longer stands in that form, and the codex’s date and origin are still debated. What has not changed is the reason scholars keep returning to it. Or. 4445 preserves nearly two thirds of the Torah as it was first written, pointed and annotated in the age of the Tiberian Masoretes. Its margins carry a Masorah older in form than that of the great complete Bibles, and a note that seems to speak of Ben Asher as a living man.

A Pentateuch in Two Ages

The book is a composite. The British Library’s description, based on George Margoliouth’s catalogue, records 186 folios. Folios 1–28, 125, 128 and 159–186 are paper and were added in 1540. A considerable number of leaves are mutilated, and some, such as folios 122–24 and 126–27, are badly damaged.⁷ In Margoliouth’s own description of 1899, the vellum leaves measure about 16½ by 13 inches. The original arrangement of the quires was already uncertain then, because leaves had been cancelled in places and many folios had been joined together irregularly.⁸

What survives of the ancient manuscript begins in the middle of the story of Joseph, at Genesis 39:20, on folio 29, and runs to Deuteronomy 1:33.⁹ The books begin at folio 1v (Genesis), 42r (Exodus), 85r (Leviticus), 115r (Numbers) and 158v (Deuteronomy).¹⁰ Christian David Ginsburg, who made the codex the first manuscript in his list of witnesses, described it as running “on continuously from Gen. XXXIX 20 to Deut. I 33,” except for the two interior leaves that had been replaced.¹¹

The later restoration has its own history. Phillips dates the replacement leaves to 1539 and describes their scribe as being of Yemenite or Persian origin. “It’s OK, you can say it,” he writes of a page from the sixteenth-century leaves: “no, they are nowhere near as beautiful as the original pages.” He distinguishes four, or four and a half, chronological stages in the book: the original codex, an interim phase, the replacement phase, which came in two parts, and the latest phase.¹² Margoliouth printed the mutilated colophon of the newer part from folio 186v. He also noted, on folio 1r, a list of books in a Persian hand of the sixteenth or seventeenth century.¹³

Three Columns Without Stretched Letters

The ancient pages are laid out in three columns of twenty-one lines each.¹⁴ Ginsburg admired the script: “large, bold and beautiful characters,” furnished with vowel points and accents.¹⁵ Margoliouth found the hand identical with the square script of the Codex Babylonicus Petropolitanus, the dated manuscript of the Latter Prophets of 916. Both codices were written with a reed, “and the ink used is thick and shiny.”¹⁶

Several details of the writing caught both cataloguers’ attention. The upper stroke of the letter lamed is greatly lengthened in the first line of a page. The left edges of the columns are ragged, because, as Margoliouth put it, the scribe did not know “the use of (or not caring to employ) elongated letters.”¹⁷ Ginsburg made the same observation more emphatically: the dilated letters that later scribes stretched to fill out a line “did not then exist, and are indeed a modern device.”¹⁸ When a word would not fit, the scribe filled the gap with dots or unfinished letters rather than abbreviating.¹⁹ Margoliouth noticed that parts of the letter alef were usually used as fillers. He wondered whether this hinted that the scribe’s name began with that letter, but found “no sufficient certainty on the point.”²⁰ The name, when it was found, began with nun.

Phillips, comparing the scribal habits of the great codices, describes the same filling technique. The original scribe of Or. 4445, which he calls B, “predominantly used raised dots at the end of the line, raised dots between the last words of a line, the first stroke(s) of the alef,” and less often part of the first letter of the next line.²¹ The same scribe, he adds, “was very diligent in marking virtually all spirantized begadkfat letters with rafe,” and “assiduous in marking non-consonantal final he with rafe.”²² Ginsburg recorded other peculiarities of letter form. The left leg of he is closed at the top like that of ḥet, yod has a longer shaft than usual, and final nun is hardly longer than zayin. The vowel qameṣ still has its primitive form, a pataḥ with a dot beneath it.²³

Before the Verse Divider

One feature in particular persuaded the early cataloguers that the codex was very old. Every later manuscript Ginsburg collated marked the end of a verse with a sign like a colon, the sof pasuq. In Or. 4445 that sign “was originally entirely absent.” The end of a verse was marked only by the silluq accent under its last word, followed closely by the first word of the next verse. When a later Masoretic annotator added the two dots, Ginsburg wrote, “they are frequently forced in between the verses for want of space.”²⁴

Margoliouth drew the same conclusion from the same evidence. He noted that verse divisions were regularly used in the St Petersburg manuscript of 916 but were originally absent from Or. 4445. That fact, “together with the stiffer and somewhat less formed type of writing used in the MS., appears unmistakably to point to its greater antiquity.”²⁵ He added a third consideration in a footnote: the three-column layout. William Wright had observed that Syriac manuscripts in three columns “are scarcely to be met with after the seventh century.”²⁶ Ginsburg, describing the Petersburg codex, made the comparison directly. Its letters were “almost identical” with those of Or. 4445, but it had “already the verse-divider or SophPasuk (:) which is still absent in Orient. 4445.”²⁷ The journal’s essay on Codex Babylonicus Petropolitanus describes that dated partner.

A Pointing Without Fixed Rules

The vowels and accents of Or. 4445 follow the Tiberian system, not the superlinear Babylonian pointing of the 916 Prophets. Margoliouth thought the pointing was “in all probability contemporary with the consonantal text.”²⁸ Its most noticeable feature is the scarce and irregular use of the meteg and gaʿya, the small vertical strokes that mark secondary stress. Margoliouth gave examples. On folio 120a the same word appears with a meteg in one column, without it in the next, and with a gaʿya lower down. The manuscript, he concluded, “appears to have been punctuated at a time when no fixed rules for the use of either ‘metheg’ or ‘ga’ya’ had yet been established.”²⁹

Ginsburg reached the same verdict, and stated it in a sentence that sums up the codex’s value for the history of pointing.

As this is a most accurately written MS. and as the accuracy extends both to the vowel-points and accents, it is evident that it belongs to a period when the superfine speculations about the Metheg and the Gaya had not as yet asserted themselves.

Christian D. Ginsburg, Introduction to the Massoretico-Critical Edition of the Hebrew Bible, 1897³⁰

The point mattered to him because, as he wrote elsewhere, the presence or absence of the meteg or gaʿya “constitutes fully nine-tenths of the differences” between Ben Asher and Ben Naphtali, the two redactors whose disagreements fill the medieval lists. Yet “the oldest and most beautifully written Codex of the Pentateuch” very rarely used either sign.³¹ The journal’s essay on Mishael ben Uzziel and the Book of Differences describes the lists in which those disagreements were recorded.

The only full-length study of the codex, a 1983 doctoral thesis at the School of Oriental and African Studies by David Lloyd Lyons, refined the picture. Lyons found the vocalization and accents “accurately marked, reflecting in the main the norms and tendencies of Aaron Ben-Aser in the model Aleppo Codex,” with only sporadic errors. Its particular character lay in “certain features of the metheg-marking” and in its economy with one of the accent signs. In one class of meteg, Or. 4445 was actually more regular than the Aleppo Codex. Overall, he concluded, the codex reflected “a slightly less-developed stage than that reflected in the Aleppo and Leningrad Codices.”³²

Ben Asher Without a Blessing

In the outer margin of folio 40b, beside Genesis 49:20, the Masoretic annotator wrote a note that both nineteenth-century scholars found remarkable. It records how “the great teacher Ben-Asher,” as Ginsburg translated the title, accented a pair of words in four verses. Ben Asher used the conjunctive accent merkha in all four places, while others joined the words with a maqqef.³³ Margoliouth noted that the note “is not at present known to exist anywhere else,” and that the received Masoretic text does not follow Ben Asher’s reading in this instance.³⁴

What struck them most was what the note left out. Medieval writers normally followed the name of a dead scholar with a blessing, such as “his rest is in Paradise.” Here there is none. Ginsburg took this as “additional evidence that the Massorah in question was written in the life-time of Ben-Asher.”³⁵ He found a second mention of Ben Asher without the formula, in the annotator’s note on Leviticus 20:17, on folio 106a. Margoliouth drew the same conclusion: “no euphemism for the dead is used after Ben Asher’s name, thus apparently showing that he was alive at the time.”³⁶

The note does not identify which Ben Asher is meant. Ginsburg dated the Masorah of the codex to “the life-time of the Ben-Ashers circa A. D. 900–940,” using the plural for the family of Masoretes to which Aaron ben Asher belonged.³⁷ If the annotator was writing while a Ben Asher still lived, then the margins of Or. 4445 come from the lifetime of the Ben Asher family itself. The journal’s essay on the schools of Ben Asher and Ben Naphtali follows that family and its rivals.

The Masorah in the Margins

The Masoretic apparatus is laid out with great regularity. Two lines of the Masorah Magna generally run along the upper margin of each page and four along the lower. The outer margins and the spaces between the columns hold the Masorah Parva.³⁸ Ginsburg believed both had been added about a century after the text was written, by the annotator or naqdan who revised it. “The Massorah which is here exhibited in its oldest form,” he wrote, “frequently uses a terminology different from that employed in MSS. of the eleventh and twelfth centuries.”³⁹ Margoliouth agreed, though more cautiously, that the Masorah was written “about a century (?) later than the text.”⁴⁰

Lyons described the Masorah as “profuse.” Masorah Parva, enumerative Masorah Magna and cumulative Masorah are recorded throughout. Some notes resemble the Babylonian Masorah in style and form, “but no direct influence can be assumed,” while the cumulative lists are specifically Tiberian. The apparatus is relatively accurate. Its counts and descriptions “apparently surpass (somewhat) those of the Leningrad Codex,” but fall short of the Aleppo Codex in precision.⁴¹ Lyons compared the cumulative lists with the Masoretic compendium Okhlah ve-Okhlah and concluded that the two collections “cannot be treated as variant recensions of a single compendium.”⁴² That compendium is the subject of the journal’s essay on the lists of Okhlah ve-Okhlah.

Ginsburg chose a page of Or. 4445 to show students what the Masorah looked like at its beginnings. In his introduction he printed the Masorah of folio 94b, covering Leviticus 11:4–21, in parallel with the notes on the same verses from nine other manuscripts and the first edition of the rabbinic Bible.⁴³ He also reproduced the page as a collotype facsimile, which, the caption records, was presented to the Trinitarian Bible Society by the Rev. William Bramley-Moore.⁴⁴ One of its notes, on Leviticus 11:21, gives what Ginsburg called “the oldest form of the List” of places where the written text has lo, “not,” and the reading is lo, “to him.” The note counts fifteen instances, and then adds that “there are two other passages outside this Massorah,” Isaiah 49:5 and 1 Chronicles 11:20.⁴⁵ For Ginsburg this showed that the Masoretes of different schools supplemented one another’s lists.

Another note shows the Masoretes guarding the smallest particles of the text. In Exodus 39:33–43, where the object marker et appears many times, sometimes with the conjunction “and” and sometimes without, the Masorah of Or. 4445 records the exact pattern verse by verse. It ends: “This is according to the Codex Mugah.” Ginsburg called this “the earliest Codex quoted by the Massorites, as far as I can trace it.” He read the note as “manifestly a protest against these variants,” since manuscripts and ancient versions disagree in almost every verse of the passage.⁴⁶ Margoliouth noticed another authority quoted on folios 30b and 31b, which he took to be “an ancient lost codex” like the others the Masoretes named.⁴⁷ The journal’s essay on the lost model codices of the Masorah gathers what is known of those books.

Margoliouth observed two more habits of the annotator. Some rubrics of the Masorah Magna carry vowel points. The Masorah Parva, as a rule, does not mark a qere in the margin where there is no audible difference between the written form and the form read aloud.⁴⁸ Compared with Ginsburg’s great edition of the Masorah, the Masorah Magna of Or. 4445 contained only one or two rubrics not found there. It frequently expressed them differently, however, and in many cases less fully.⁴⁹

Open and Closed Sections

The text of Or. 4445, both cataloguers agreed, belongs to the Western or Palestinian recension on which the received text is based.⁵⁰ In one respect, however, it differs constantly from later Masoretic Bibles: the division of the text into open and closed sections, the paragraph breaks of the Torah. Ginsburg counted the differences book by book. In the surviving chapters of Genesis and in Exodus, Leviticus and Numbers, the codex shows “no fewer than 116 variations” from the received text.⁵¹

The remarkable thing, in Ginsburg’s view, was how few of these the later annotator corrected. Out of 116 variations he altered only six, and in one of those six he changed a break that already agreed with the received text so that it no longer did. At Exodus 9:13, where the codex has a closed section as later Bibles do, the annotator wrote that it ought to be open.⁵² In places where he wanted no break at all, Margoliouth noted, the annotator wrote a term meaning “straight on.” Margoliouth found one form of the expression on folio 48b “thus testifying to the Persian affinities of the MS.”⁵³

The codex also records the older divisions of the Torah. The sedarim, the units of the triennial reading cycle, are marked only twice, at Genesis 43:14 and 46:8. Ginsburg remarked that the second does not appear in the official lists or in any other manuscript he collated.⁵⁴ The annual weekly portions match the received text. The number of verses in each portion is written in the space at its end, and the word parasha stands in the margin.⁵⁵ The journal’s essay on the sedarim explains the Palestinian cycle those two marks belong to.

The verse counts sometimes depart from the later tradition. For the last portion of Genesis, Or. 4445 gives 84 verses where the received text has 85, which Ginsburg thought probably reflected a different recension. At the end of Numbers it states that the book contains 1,285 verses rather than 1,288. Since its figures for the separate portions agree with the other manuscripts in every case but one, Ginsburg concluded that the scribe “committed an error in the summing up.”⁵⁶

A Slip at Leviticus 21

The codex also preserves an ordinary copying error. On folio 107a the whole of Leviticus 21:24 was at first left out. Both that verse and the first verse of the next chapter begin with the same Hebrew word, “and he spoke,” and the scribe’s eye skipped from one to the other. Ginsburg cited it as the first example in his chronological survey of this kind of error, a mistake “in operation in all ages and in all countries.” The missing verse was added by a later hand.⁵⁷ The annotator then had to deal with the paragraph break that the inserted verse now occupied. Ginsburg notes that he rightly marked that an open section belonged there.⁵⁸

Part of a Whole Bible

A note on folio 40a suggests that Or. 4445 was once more than a Pentateuch. Among the Masorah’s lists of unique words is the statement, in Margoliouth’s translation, that “on the authority of the scribe and the punctuator,” a particular word occurs only once in the Bible. Margoliouth pointed out that the word occurs in Isaiah 10:15, and a similar form in Daniel 11:37. A list that counted it as unique across the whole Bible, on the authority of this scribe and punctuator, implied that they had produced a whole Bible. He concluded that “one is justified in assuming that Or. 4445 is but a small portion of a complete Bible written and punctuated in exactly the same ancient manner.”⁵⁹ He added that the scribe and the punctuator were possibly the same person.⁶⁰

The name in the margins later answered that question. In the acrostics Phillips describes, Nissi ben Daniel names himself as scribe, vocaliser and Masoretic annotator.⁶¹ The British Library’s description takes the same view: Nissi ben Daniel, “probably also the punctuator of the manuscript.”⁶² Lyons summarized the codex as a Tiberian Pentateuch “vocalized and annotated by a certain Nissi B. Daniel at the beginning of the 10th century C.E.”⁶³ That sits uneasily with Ginsburg and Margoliouth, who both thought the Masorah had been added about a century after the text.⁶⁴ The acrostic formula claims all three tasks for one man, and the question of who wrote which layer of the margins is one of the things Phillips’s layer-by-layer study sets out to answer.⁶⁵

Nissi ben Daniel and the Road to Persia

Where Nissi worked is less certain. The British Library dates the codex between 920 and 950 and gives its place of origin as Egypt or Palestine.⁶⁶ Phillips has gathered evidence that points east instead. One scholar, he writes, suggested a Persian origin because the name Nissi or Nissim was very common in and around modern Iran at the time. More recently Yosef Ofer proposed that a place name is hidden in one of the acrostics. On folio 139r, after the name and the formula about the scribe, eight further list entries follow. Read as an acrostic, their letters give a word that Ofer interprets as “from Shaqabad,” a place Phillips identifies with modern Shaqlawa in north-eastern Iraq.⁶⁷ Margoliouth, as noted above, had already seen Persian affinities in the annotator’s vocabulary.⁶⁸

In the second part of his study Phillips adds a piece of evidence of his own. Hardly any early Hebrew Bible manuscripts can be placed securely in the Persian region, he notes: only four, of which two are inaccessible and one too fragmentary to help. The remaining one is National Library of Israel Heb. 8° 2238, a manuscript containing a single Torah portion, Numbers 13–15, dated by its colophon to 1107.⁶⁹ At Numbers 13:22, where a singular verb appears among plurals in the story of the spies, both manuscripts carry the same Masoretic note in the same very unusual terminology. Where the standard Masorah uses the technical words sevirin and rabbim, these notes use Aramaic terms that Phillips calls “very, very unusual indeed.” In Ginsburg’s Masorah, a work of over 2,500 pages, one of these terms appears only three times. Phillips found it nowhere in the Masorah of the Aleppo or Leningrad codices, or in about ten other early manuscripts he checked.⁷⁰

Phillips is careful about what this proves. “With just one note we can’t say anything with certainty,” he writes, but “it’s another intriguing connection between Or. 4445 and Persia.”⁷¹ He also has a firm view on one older claim. Scholarship on the codex, he notes, has focused on its unusually plentiful Masoretic notes and on “the text itself as a potentially pre-Ben Asher witness to the Tiberian biblical text (spoiler: it isn’t).”⁷² The later history of the codex, in any case, is Persian. Its ownership colophon is written in Judaeo-Persian, and Phillips thinks a continuing Persian location best explains the interim periods.⁷³

How Old Is It

The date has moved more than once. Ginsburg was not consistent. In his chapter on the Masorah he called the codex “not later than the middle of the eighth century,” with a Masorah about a century later, “about the middle of the ninth century.”⁷⁴ In his chapter on Ben Asher he called it “undoubtedly of the early part of the ninth century.”⁷⁵ In his formal description he wrote that “though not dated, the original MS. was probably written about A. D. 820–850,” with the Masorah added around 900–940.⁷⁶ Describing the dated Petersburg codex of 916, he called Or. 4445 “at least half a century earlier.”⁷⁷ Margoliouth judged that it “may safely be looked upon as older” than the Petersburg codex and was “probably written about the middle of the ninth century.”⁷⁸

The arguments behind those dates were palaeographical and typological: the missing verse dividers, the irregular meteg, the three columns, the older terminology of the Masorah. They placed Or. 4445 at the very beginning of the age of pointed codices. Later scholarship moved the date forward. The British Library now gives 920–950.⁷⁹ Lyons placed Nissi’s work “at the beginning of the 10th century.”⁸⁰ Phillips writes only that the original codex “appears to have been produced in the 10th century (maybe).”⁸¹ On any of these dates the codex is among the oldest surviving pointed Torah manuscripts, but no longer the earliest of all.

From the British Museum to the Screen

The British Library records the acquisition simply: “Mrs C. D. Sassoon & Co.: purchased by the British Museum on 09 May 1891.”⁸² Margoliouth described some features of the manuscript in The Academy in April 1892, as Ginsburg noted.⁸³ Ginsburg made it the first manuscript in his account of the witnesses to his edition. He departed from his usual order by country and library so that “the oldest two Codices, viz. Orient. 4445 in the British Museum and the St. Petersburg Codex dated A. D. 916” could stand at the head of the list.⁸⁴ The journal’s essay on Christian David Ginsburg follows the collator who described it. In 1899 Margoliouth’s catalogue included a photographic plate of the codex, chosen, Douglas explained, as one of the “specially famous MSS.”⁸⁵

In the twentieth century Or. 4445 became one of a small group of codices that every student of the Masorah cites. Phillips writes that “for the past century or so, Or. 4445 has served as one of the six major biblical codices for the discussion of the Masoretic Text.”⁸⁶ In the conventional sigla, A is the Aleppo Codex and B is Or. 4445, beside C for the Cairo Codex of the Prophets, L for the Leningrad Codex, and S and S1 for the two Sassoon manuscripts. For details of all of them Phillips refers readers to Israel Yeivin’s Introduction to the Tiberian Masorah.⁸⁷ The journal’s essay on Israel Yeivin describes that handbook.

The codex is also now easy to see. It was digitised by the British Library’s Hebrew Manuscripts Digitisation Project, funded by the Polonsky Foundation. Its images and description are in the public domain and are published on OPenn by the University of Pennsylvania Libraries.⁸⁸ The Library’s own catalogue currently notes that its digital images are “currently unavailable” in its viewer.⁸⁹ Phillips’s essay on Hebrew codicology, written while he was working on the codex, reproduces the codex’s quire numbers, many of which were worn or cropped away in the margins when the manuscript was bound.⁹⁰

Layers of a Single Book

Or. 4445 is a book made in layers. A scribe wrote, pointed and annotated a Torah, and perhaps a whole Bible, in the generation of the Ben Ashers, and hid his name in his own Masorah. A later hand forced verse dividers into the narrow spaces between his verses, and an annotator corrected six of his paragraph breaks and left more than a hundred alone. Five centuries after that, a copyist on paper replaced the lost beginning and end so that the book could be read again from Genesis to Deuteronomy, and a Persian owner signed the last leaf. In 1891 it entered the British Museum, and scholars spent the next century arguing about its age.

Each layer is still visible on the page. The hub page on Codex Orientales 4445 places it among the great codices of the Masoretic tradition.

Notes

1. Kim L. Phillips, “Manuscript Archaeology. Oriental 4445, Layer by Layer (Part I),” Kim’s Substack, 19 May 2026, at kimlphillips.substack.com. *
2. British Library, “Or 4445 תורה (Torah),” manuscript description and images, Hebrew Manuscripts Digitisation Project, published on OPenn, University of Pennsylvania Libraries, at openn.library.upenn.edu. *
3. British Library, “Or 4445,” OPenn. *
4. British Library, “Or 4445,” OPenn. *
5. G. Margoliouth, Catalogue of the Hebrew and Samaritan Manuscripts in the British Museum, part 1 (London: British Museum, 1899), no. 64, at archive.org. *
6. British Library, “Or 4445,” OPenn. *
7. British Library, “Or 4445,” OPenn. *
8. Margoliouth, Catalogue, part 1, 36. *
9. Margoliouth, Catalogue, part 1, 37. *
10. British Library, “Or 4445,” OPenn. *
11. Christian D. Ginsburg, Introduction to the Massoretico-Critical Edition of the Hebrew Bible (London: Trinitarian Bible Society, 1897), at archive.org, 469. *
12. Phillips, “Oriental 4445, Layer by Layer (Part I)”. *
13. Margoliouth, Catalogue, part 1, 39. *
14. British Library, “Or 4445,” OPenn; Ginsburg, Introduction, 470. *
15. Ginsburg, Introduction, 469–470. *
16. Margoliouth, Catalogue, part 1, 36–37. *
17. Margoliouth, Catalogue, part 1, 37. *
18. Ginsburg, Introduction, 470. *
19. Ginsburg, Introduction, 165. *
20. Margoliouth, Catalogue, part 1, 37. *
21. Kim Phillips, “A New Codex from the Scribe behind the Leningrad Codex. L17,” Tyndale Bulletin 68.1 (2017), 1–29, at tyndalebulletin.org, 14. *
22. Phillips, “L17”, 17. *
23. Ginsburg, Introduction, 473–474. *
24. Ginsburg, Introduction, 473. *
25. Margoliouth, Catalogue, part 1, 37. *
26. Margoliouth, Catalogue, part 1, 37. *
27. Ginsburg, Introduction, 475. *
28. Margoliouth, Catalogue, part 1, 37. *
29. Margoliouth, Catalogue, part 1, 37. *
30. Ginsburg, Introduction, 474. *
31. Ginsburg, Introduction, 278. *
32. David Lloyd Lyons, “The Vocalization, Accentuation and Masora of Codex Or. 4445 (Brit. Mus.) and Their Place in the Development of the Tiberian Masora” (PhD thesis, School of Oriental and African Studies, University of London, 1983), abstract at openalex.org. *
33. Ginsburg, Introduction, 249. *
34. Margoliouth, Catalogue, part 1, 38. *
35. Ginsburg, Introduction, 249–250. *
36. Ginsburg, Introduction, 470; Margoliouth, Catalogue, part 1, 38. *
37. Ginsburg, Introduction, 470. *
38. Ginsburg, Introduction, 470. *
39. Ginsburg, Introduction, 470. *
40. Margoliouth, Catalogue, part 1, 38. *
41. Lyons, “Codex Or. 4445,” abstract. *
42. Lyons, “Codex Or. 4445,” abstract. *
43. Ginsburg, Introduction, 423–424. *
44. Ginsburg, Introduction, plate before 469. *
45. Ginsburg, Introduction, 428. *
46. Ginsburg, Introduction, 429–430. *
47. Margoliouth, Catalogue, part 1, 38. *
48. Margoliouth, Catalogue, part 1, 38. *
49. Margoliouth, Catalogue, part 1, 38. *
50. Ginsburg, Introduction, 470; Margoliouth, Catalogue, part 1, 37. *
51. Ginsburg, Introduction, 470–472. *
52. Ginsburg, Introduction, 472. *
53. Margoliouth, Catalogue, part 1, 38. *
54. Ginsburg, Introduction, 472; Margoliouth, Catalogue, part 1, 38. *
55. Ginsburg, Introduction, 473; Margoliouth, Catalogue, part 1, 38. *
56. Ginsburg, Introduction, 74, 82. *
57. Ginsburg, Introduction, 169. *
58. Ginsburg, Introduction, 472. *
59. Margoliouth, Catalogue, part 1, 38–39. *
60. Margoliouth, Catalogue, part 1, 38. *
61. Phillips, “Oriental 4445, Layer by Layer (Part I)”. *
62. British Library, “Or 4445,” OPenn. *
63. Lyons, “Codex Or. 4445,” abstract. *
64. Ginsburg, Introduction, 470; Margoliouth, Catalogue, part 1, 38. *
65. Phillips, “Oriental 4445, Layer by Layer (Part I)”. *
66. British Library, “Or 4445,” OPenn; British Library Archives and Manuscripts Catalogue, “Or 4445. Pentateuch,” record 032-003376314, at searcharchives.bl.uk. *
67. Phillips, “Oriental 4445, Layer by Layer (Part I)”. *
68. Margoliouth, Catalogue, part 1, 38. *
69. Kim L. Phillips, “Manuscript Archaeology. Oriental 4445, Layer by Layer (Part II),” Kim’s Substack, 26 May 2026, at kimlphillips.substack.com. *
70. Phillips, “Oriental 4445, Layer by Layer (Part II)”. *
71. Phillips, “Oriental 4445, Layer by Layer (Part II)”. *
72. Phillips, “Oriental 4445, Layer by Layer (Part I)”. *
73. Phillips, “Oriental 4445, Layer by Layer (Part I)”. *
74. Ginsburg, Introduction, 423. *
75. Ginsburg, Introduction, 249. *
76. Ginsburg, Introduction, 469–470. *
77. Ginsburg, Introduction, 475. *
78. Margoliouth, Catalogue, part 1, 36. *
79. British Library, “Or 4445,” OPenn; British Library catalogue, “Or 4445”. *
80. Lyons, “Codex Or. 4445,” abstract. *
81. Phillips, “Oriental 4445, Layer by Layer (Part I)”. *
82. British Library, “Or 4445,” OPenn. *
83. Ginsburg, Introduction, 474. *
84. Ginsburg, Introduction, 469. *
85. Margoliouth, Catalogue, part 1. *
86. Phillips, “Oriental 4445, Layer by Layer (Part I)”. *
87. Phillips, “L17”, 2. *
88. British Library, “Or 4445,” OPenn. *
89. British Library catalogue, “Or 4445”. *
90. Kim L. Phillips, “Preaching to the Quire,” Kim’s Substack, 4 September 2026, at kimlphillips.substack.com. *