The Hebrew Bible is the sacred, foundational text for peoples across the globe. For thousands of years, the faithful have closely studied, analyzed, and meditated on Scripture to acquire wisdom and attain spiritual enlightenment. The Hebrew Bible is composed of twenty-four books divided into three sections: the Torah (Pentateuch), the Prophets, and the Writings. Christians call these texts the Old Testament, and Catholic, Orthodox, and Protestant denominations all incorporate them into their biblical canons. Muslims, too, regard the Torah and the Psalms as divinely revealed books, and many biblical stories made their way into the Qur’an and later Islamic works.

The narratives and teachings of the Hebrew Bible permeate every aspect of Western humanistic endeavor: from painting and sculpture to literature, music, and philosophy to the biblical epics of early cinema. The personages of the Old Testament—Adam and Eve, Noah, Abraham and Sarah, Isaac and Rebecca, Jacob and Leah, Joseph and his brothers, Moses and Aaron, Pharaoh, Job, Joshua, Samson and Delilah, Ruth, Saul, David and Goliath, Solomon, Elijah and Elisha, Jonah, Daniel, Nebuchadnezzar, and Mordechai and Esther—are as familiar to us today as they were to generations of our ancestors. Copied, printed, and translated into scores of languages the world over, the Hebrew Bible is arguably the most influential book in history and constitutes the bedrock of Western culture.

Codex Sassoon, dating from circa 900 CE, is comprised of the twenty-four books of the Hebrew Bible supplied with punctuation, vowels, cantillation marks, and an extensive collection of text-critical notes known as the Masorah. Only one other full Hebrew Bible dating from this early period survived into the modern era: the Aleppo Codex. The latter was produced in Tiberias circa 930 CE and has long been recognized as preserving an exceptionally accurate version of the biblical text. However, almost two-fifths of that book, including the vast majority of the Torah and parts of the Prophets and Writings, were lost under mysterious circumstances sometime between the late 1940s and the late 1950s. By contrast, to this day Codex Sassoon contains almost the entirety of the Hebrew Bible. It is thus the earliest most complete copy of the Hebrew Bible extant.

In the preparation of this catalogue, the research of two outstanding scholars has proven indispensable. Professor Yosef Ofer of Bar-Ilan University in Israel, one of the preeminent authorities on the history of the biblical text, has been intensively studying Codex Sassoon for the past several years (together with his students Dr. Neriah Klein and Tal Peretz) while preparing an edition of its Masorah for publication. Professor Judith Olszowy-Schlanger of the University of Oxford and the École Pratique des Hautes Études, a leading expert on Hebrew paleography and codicology, has published widely on medieval Jewish book culture, with a special focus on documents produced in the Middle East. Both of them submitted written reports that, together, form the basis for the present catalogue. Sotheby’s is grateful to them for their trailblazing scholarship, and to Professor Paul Fenton, Professor Miriam Goldstein, and the Sassoon Family Archives for providing information that aided in the cataloguing of this historic manuscript.

Throughout the early modern period, the oldest Hebrew biblical texts of which scholars were aware were produced during the Middle Ages. This changed dramatically in 1947 with the discovery of the Dead Sea Scrolls, possibly the most important archeological find of the twentieth century. Preserved in the arid caves of the Judean Desert, these documents date as far back as the third century BCE through the first century CE, during the period when the Second Temple still stood in Jerusalem. Among this corpus of somewhere between eight hundred and one thousand manuscripts are about two hundred thirty scrolls containing parts of all the books of the Hebrew Bible (except the book of Esther), including a complete copy of the book of Isaiah. Needless to say, the discovery of the Dead Sea Scrolls revolutionized the study of the Hebrew Bible and significantly revised historians’ understanding of Second Temple Judaism and early Christianity.

About a half-century earlier, Europeans visiting Fustat (Old Cairo) had made another major discovery when they uncovered a huge trove of approximately 400,000 early medieval book fragments and records collectively known as the Cairo Genizah (a genizah being a traditional repository for the respectful disposal of worn Hebrew texts). The Genizah’s contents range from the sacred to the mundane and originate in a geographically diverse array of Jewish communities in Europe, North Africa, and especially the Middle East, with the earliest among them deriving from the late ninth century. Because only a minute number of fragmentary Hebrew documents produced between the mid-second and late ninth centuries survive, scholars refer to this time interval as a “silent period” in the history of Jewish literature. Astonishingly, not a single codex comprising the Hebrew Bible is extant from these seven centuries of silence.

What contributed to the end of this silence and the onset of a new age of intensified manuscript production? Some historians have proposed that at least part of the answer may lie in the Jews’ transition from copying books in scroll form to using the (relatively) new technology of the codex. This highly durable book format—in which sheets of papyrus or parchment inscribed with text were folded and sewn together as quires that could later be assembled and bound in a single volume—had been invented in the Graeco-Roman world in the first centuries CE and quickly diffused throughout pagan and Christian society. The codex was found to be much more user-friendly and versatile than the scroll because it allowed for easier navigation of a given work, as well as the inscription of text on both the recto (front) and verso (back) of a writing surface, thereby greatly increasing a book’s economy of space. With the rise of Islam in the seventh century, the codex became the primary, if not exclusive, medium for the transmission of the text of the Qur’an, especially as the religion spread during the Muslim conquests of the seventh and eighth centuries. In Jewish literature, the codex is first mentioned in an eighth-century Middle Eastern source, and the word used, mitshaf, derives from Arabic mus'haf, suggesting that the new book form reached Jewish communities via contact with Muslims. Thereafter, Jews enthusiastically adopted the codex and Hebrew book production exploded, while use of the scroll was gradually relegated to ritual contexts, such as the public reading of the Torah scroll in the synagogue.

In ancient times, including the period of the Dead Sea Scrolls, the books of the Hebrew Bible were copied without punctuation, vocalization, or accentuation. Words were separated by spaces, paragraph divisions broke up the flow of the text, and four letters (alef, heh, vav, yod) were frequently used to mark certain vowels, but even with these devices, the correct pronunciation and parsing of the biblical text into intelligible units was not readily apparent from its appearance in a scroll. Instead, Jews in antiquity relied on inherited reading traditions, passed down orally from one generation to the next, in order to understand the Hebrew Bible.

In about the seventh or eighth century, after hundreds of years during which Hebrew was no longer a commonly spoken language, Jewish Bible scholars commenced a project to develop systems of committing the reading traditions to writing and thus preserving them for posterity. The most influential of these systems—named “Tiberian” for the city of Tiberias on the Sea of Galilee where it originated—would, with time, become the standard one used throughout the Jewish world to record the vowels and accents of the Hebrew Bible.

In addition, extensive lists counting the number of occurrences of a given spelling, vocalization, or accentuation in the Bible—collectively known as the Masorah (lit., Tradition)—were drawn up in order to stabilize the biblical text and ensure that scribes would copy it correctly. The masoretic tradition of Aaron ben Moses ben Asher (first half of the tenth century), scion of a famous family of Tiberian Masorah scholars (also known as Masoretes), came to be considered particularly accurate and authoritative.

It was no coincidence that this activity took place concurrently with the introduction into Jewish book culture of the codex, which allowed for the addition of vowels, accents, and the text-critical notes of the Masorah to a page of biblical verse. Whereas rabbinic law prohibits the insertion of any extra-biblical text or diacritical marks into Torah scrolls used for ritual purposes, the new format of the codex enabled and encouraged the addition of all the crucial masoretic systems and notes. Codices that feature all of these elements are known as Masoretic Bibles, and those that closely follow Ben Asher’s tradition are prized above all others.

Most Masoretic Bibles include two types of Masorah notes inscribed in minuscule script on each page: the Masorah parva, added between columns of biblical text and in the outer and inner margins, and the Masorah magna, generally written in three or four lines in the upper and/or lower margins. Among other functions, the former usually transmits in highly abbreviated form the number of occurrences of a given textual feature in the Hebrew Bible, while the latter often expands on that number by citing those occurrences, sometimes formulating a rule or mnemonic to aid the scribe’s memory. Masoretic Bibles were not meant to be used liturgically in the synagogue; instead, they served scholars and copyists intent on maintaining and transmitting the masoretically accepted and standardized version of the biblical text.

Two examples taken from Codex Sassoon will help demonstrate the function of the Masorah:

The text of the Song of Deborah from the book of Judges appears on p. 222. The first line of this poem includes the word u-barak (and Barak; Judg. 5:1) with a small circle over it, referencing the outer margin where the Masorah parva indicates the number of occurrences of this word-form using the letter gimel (three). If we now look in the upper margin, the last line of the Masorah magna tells us that u-barak indeed occurs three times in the Hebrew Bible, though in two different senses: twice as a proper noun (listing Judg. 4:16, 5:1) and once as a common noun (listing Job 20:25). Note that a slightly different form of this word, u-berak, occurs once, in the book of Nahum (3:3), but is left off of this masoretic list precisely because it is vocalized differently. Thus, these comments of the Masorah parva and magna help the scribe avoid error when copying the relevant verses, making sure to vocalize the text correctly.

Moving down two lines, a small circle suspended above the phrase barekhu Hashem (bless God; Judg. 5:2) points us to the Masorah parva in the outer margin, which notes that this form, lacking the direct object marker et, occurs five times in the Hebrew Bible. Again, the first and second lines of the Masorah magna in the upper margin list these five occurrences by quoting snippets from the relevant verses (Judg. 5:2, 5:9, Ps. 103:20–22), concluding with the observation that all other instances of the words for “bless God” include the direct object marker et. Here, too, note that the form berekho Hashem, spelled identically but vocalized differently, occurs twice (Gen. 27:27, Isa. 19:25) but is again omitted from the current listing due to its variance.

Students of the Masorah know that no two manuscripts contain exactly the same collection of notes. The selection, placement, and formulation of these comments vary based on the predilections and tastes of a given volume’s Masorete(s). Each manuscript is therefore a unique creation, but taken together the extant Masoretic Bibles give us a window onto the Herculean labors of the Masoretes in attempting to record every detail of the Hebrew Bible’s nearly 307,000 words.

The simplest explanation for why so few complete Masoretic Bibles produced before the end of the first millennium CE survive is that so few were created. Copying the Hebrew Bible’s text, vowel points, and accent marks, together with the Masorah, onto large pieces of parchment was an expensive proposition that only wealthy patrons could afford to sponsor.

Codex Sassoon is distinguished both by its antiquity and by its completeness. As one of only two of the earliest Masoretic Bibles from the tenth century containing all twenty-four biblical books to have survived over a millennium, Codex Sassoon is approximately contemporary with the aforementioned Aleppo Codex but is significantly more complete. Whereas the Aleppo Codex is currently lacking almost all of the Pentateuch (approximately 200 folios) Codex Sassoon is almost complete, probably lacking no more than eight folios containing biblical text (Gen. 1:1–11:2a and 31:29b–37:2) toward the start of the volume with additional partial losses at the front and rear. Codex Sassoon is thus the earliest most complete Hebrew Bible; only the Leningrad Codex, produced in 1008–1010, is more complete. It is definitively the most significant early biblical manuscript in private hands.

The importance of Codex Sassoon as a witness to the Masoretic Text of the Ben Asher tradition has been recognized by researchers since at least the late 1960s. In the Torah section of the volume, for example, its consonantal skeleton deviates from Ben Asher’s tradition in only about twenty places, fewer than most other early biblical manuscripts. By comparison, the Leningrad Codex, for example, deviates from Ben Asher’s tradition in about one hundred twenty places in the Torah. Israel Yeivin, one of the foremost Masorah scholars of the previous generation, writes that Codex Sassoon “reflects a faithful Masorah that is very close to the Masorah of [the Aleppo Codex], is of approximately the same time period, and follows a similar approach.” Likewise, Mordechai Breuer, another leading Masorah expert, characterizes these two codices as “among the finest of the ancient manuscripts.”

More recent study has pointed up the significance of Codex Sassoon’s masoretic notes, which are rich both in quantity and quality and which draw upon ancient sources, including Babylonian Masorah and even the Masorah of the Aleppo Codex, as will be elaborated below.

The combination of these characteristics in a single volume endows Codex Sassoon with enormous value as the earliest most complete copy of a text fundamental to Western history and society.

The late ninth and tenth centuries were times of great upheaval in the Levant. The two caliphates of the period, the Abbasid and the Fatimid, had to contend with each other, the Byzantine Empire, and a wide range of tribes and clans for control of Egypt, the Holy Land, and Syria. Between the years 878 and 1000, these areas (or parts thereof) passed between the Tulunids, Qarmatians, Abbasids, Ikhshidids, and Fatimids. Despite the political chaos, the contemporary Jewish population was mostly protected by these Muslim authorities, and important Jewish centers developed (or continued growing) in Fustat, Jerusalem, Ramla, Tiberias, and Damascus (Aleppo was under Byzantine control for part of this era). It is in this context—the region of “Greater Syria” around the year 900 CE—that Codex Sassoon came into being.

While the identity of Codex Sassoon’s scribe is unknown, several inscriptions scattered throughout the volume allow us to chart its travels. The earliest of these, on p. 373, is a deed of sale drawn up circa 1000 CE by the skilled scribe Tsedakah ben Daniel of Jerusalem. The document records the sale of the codex by Khalaf ben Abraham, the business partner of the elder Abu Sa‘id Eli ben Hananiah al-Sirafi, to Isaac ben Ezekiel al-Attar and is witnessed by the scribe himself and a certain Mubarak ben Wahib Kohen. The price is not given, but we can assume that it was expensive, as the manuscript contains nearly four hundred folios (792 pages), which would have required about two hundred average-sized sheepskins to produce. The vast resources needed to create this manuscript speak to its status as an object of extreme luxury and rarity.

Of the five men mentioned in this deed of sale, one may have left traces in the contents of the Cairo Genizah. The seller might be identified with a certain Khalaf ben Abraham ben Khalil of Tripoli (Lebanon), who was involved in a dispute with his business partner Khalluf ben Aaron sometime before the year 1011. According to a Genizah manuscript (Oxford, Bodleian Library Ms. Heb. b. 11.8), Khalaf and Khalluf traded between Ramla and Damascus. If the two Khalaf ben Abrahams are one and the same person, we can suggest that the sale of Codex Sassoon, a luxury item that may have been jointly owned by Khalaf and Abu Sa‘id Eli, took place somewhere in Syria, rather than Egypt (which would also explain why most of the other people involved in the sale do not show up in the documents of the Cairo Genizah).

The aforementioned Isaac ben Ezekiel al-Attar later transferred ownership of the codex to two of his sons, Ezekiel and Maimon, and designated it as an inalienable endowment, as memorialized in inscriptions on pp. 190–191, 373, 792, some of which explicitly state that the codex should not be sold, redeemed, or pledged.

Despite these exhortations, the volume did eventually leave the family’s possession, as we learn from a thirteenth-century (or slightly later) dedicatory inscription written across the upper margins of pp. 567 and 587 (the first pages of quires 31 and 32, respectively), wherein the codex is consecrated to the synagogue of Makisin. (The writing on p. 567 covers an older, partially erased ownership inscription that started on the facing page—the beginning of the Writings—but is difficult to decipher.) It was likely bound anew at this stage, since the same hand that added the inscription also numbered the book’s quires. The practice of donating important biblical manuscripts to synagogues is well attested among Middle Eastern Jewish communities. The Aleppo Codex, for example, was kept in synagogues in Jerusalem and Fustat before being transferred to the Central Synagogue of Aleppo (whence it takes its name). It is therefore not unexpected that Codex Sassoon would have made its way into a Jewish house of worship.

Makisin (present-day Markada, in northeastern Syria) was a small medieval town on the road to Mosul. Much cotton was grown there, and a pontoon bridge over the Khabur River facilitated travel and trade. Surprisingly little is known about the Jewish history of Makisin except that Obadiah the Proselyte, the famed early twelfth-century Norman-Italian convert to Judaism, passed through the town on his way from Aleppo to Baghdad. Still, it can be inferred that the local Jewish community must have been a prominent one to have held such a prized manuscript as Codex Sassoon.

An inscription on the book’s last page tells an intriguing tale: the town of Makisin was attacked (perhaps by Tamerlane’s troops in 1400), and the codex was removed from the synagogue. It was entrusted to the care of a private individual, Salama ibn Abi al-Fakhr, who was required to return it as soon as Makisin would be rebuilt. The handwriting of this note suggests the end of the fourteenth century or later. It is possible that the book itself suffered damage during the destruction of the town and was subsequently repaired. If so, its survival adds yet another layer of wonder to this rare Hebrew Bible.

Following the deposit of Codex Sassoon in private hands, its trail goes cold until the twentieth century. It resurfaces in correspondence between Professor Aron Freimann (1871–1948), a famous librarian, bibliophile, and bookseller based in Frankfurt am Main, and David Solomon Sassoon (1880–1942), who assembled the most significant private collection of Judaica and Hebraica manuscripts in the twentieth century. On March 19, 1929—a bit over seven months before the Wall Street Crash of 1929—Freimann penned a letter to Sassoon, who was living in London, offering to send him a “Bible manuscript … that is certainly among the oldest known,” with a £600 price tag. (In 1925, the average annual earnings of a man working in Great Britain stood at about £260.) Sassoon expressed interest in inspecting the codex and on April 8, having completed an initial examination, countered with a £350 offer. After some additional back-and-forth, Freimann wrote to Sassoon on April 15 that the volume’s owner, who lived in the vicinity of Angora (present-day Ankara, Turkey), had accepted Sassoon’s counteroffer on the condition that he would pay straightaway. This he did, sending Freimann a check in the amount of £350 on April 18, thanking him “very heartily” for all his trouble, and wishing him a happy (and kosher) Passover. Sassoon assigned the codex the shelf mark 1053 and had it rebound and restored by His Majesty’s Stationery Office Press British Museum Bindery later that year. (The destruction and loss of much of Freimann’s personal archive as he escaped Germany for America in April 1939 make discovery of the identity of the Angora owner unlikely at this point.)

Sassoon recognized the manuscript’s antiquity early on. This is documented in both an unpublished article he composed in 1930 about his book acquisition practices and in his 1932 collection catalogue, Ohel Dawid, where he dated this Bible to the tenth century. Sassoon had a keen interest in the Hebrew Bible and masoretic studies, and Ohel Dawid begins and ends with manuscripts of the Hebrew Bible. Codex Sassoon is its very last entry, giving pride of place to this monument of medieval civilization, which now carries the name of its illustrious owner.

In 1970, the Sassoon family began selling parts of its magnificent Judaica collection through Sotheby’s. One of its blockbuster manuscript auctions took place in Zurich in November 1978, when Codex Sassoon sold at a hammer price of 550,000 Swiss francs (about $320,000/£160,350). It later emerged that the manuscript had come into the possession of the British Rail Pension Fund, which had acquired three other manuscripts from the 1975 and 1978 Sassoon sales as part of its art investment program, begun in 1974. These were loaned to the British and Bodleian Libraries in London and Oxford, respectively, with Codex Sassoon housed at the British Library and assigned the shelf mark Or. 62f (it was also referred to as “Loan 16”). From August to December 1982, Codex Sassoon was featured as the first item in an exhibition of “Hebrew Manuscripts from the Sassoon Collection” held at the British Library.

In December 1989, the British Rail Pension Fund auctioned Codex Sassoon at Sotheby’s London and it now sold for £2,035,000 (about $3.19 million), setting a world record for a Hebrew manuscript and achieving the second-highest price for any manuscript (after the twelfth-century Gospels of Henry the Lion). The codex then entered the collection of Jaqui Safra, a prominent and passionate collector of art and historical objects from Geneva, where it remains to this day.

The main script and the masoretic scripts of Codex Sassoon belong to the Northeastern subgroup of Hebrew square handwriting. This type of script originated in Iran and Mesopotamia but was also used in communities that followed the Babylonian Jewish rite in the eastern Mediterranean, including in the Land of Israel and Egypt. From a paleographical point of view, the script of Codex Sassoon has some affinities with that found in fragments that have survived from a codex of the Writings dated 903 CE and written in Gunbad-i-Mallgàn (probably present-day Dogonbadan, Iran), according to its colophon, as well as with that of undated biblical fragments with Babylonian vowels (e.g., Oxford, Bodleian Library Ms. Heb. d. 26.5–9) and of a Judeo-Arabic Bible commentary by Yusuf ibn Nuh, copied circa 1000 in Jerusalem (St. Petersburg, National Library of Russia Ms. Evr.-Arab. I 1756). The following features are typical of the Northeastern script and point to Codex Sassoon’s Eastern origins, probably in “Greater Syria” (rather than in Egypt): the alef whose lefthand downstroke departs from the body of the letter at some distance from the headline, is curved, and ends either with a foot facing left or curling up to the right; the very short final nun; the pear-shaped pe with a pointed roof; and the general lack of shading (creating a difference in width between horizontal and vertical lines).

The volume’s anonymous scribe copied the biblical text using a pointed reed, which allowed him to make curvilinear strokes and thus write more quickly. His is a careful hand that demonstrates a certain aesthetic sense with regard to the elegantly extended ascenders of the letter lamed. The copyist made sure to procure fine sheets of parchment for this project, which he generally arranged in quires of ten leaves (quinions), as was customary for large Eastern Hebrew manuscripts. Each quire starts with the hair side of the parchment, and the text itself was usually laid out in three columns per page (as was also common for medieval biblical codices), except where dictated by scribal custom (e.g., the biblical songs) or other considerations. The vocalizer, who added the vowel points and cantillation marks in a lighter brown ink, may have been identical with the primary scribe, and two separate Masoretes were responsible for the codex’s Masorah. It is possible that the first set of masoretic notes was written by the scribe of the biblical text.

One of the more interesting features of Codex Sassoon is the collection of Masorah notes inscribed in its margins—and this for two reasons. First, these notes were added by two separate Masoretes working sequentially, and the process by which the second Masorete worked can be partially reconstructed. Second, many of the second Masorete’s notes contain valuable information unknown from other surviving sources. What follows is an overview of both the form and content of the Masorah of Codex Sassoon.

Perhaps the most immediately visible difference between the notes of the two Masoretes is the ink color: Masorete 1 wrote in light brown ink, whereas Masorete 2 used dark brown, almost black, ink. While their handwriting is typologically similar, there are some variations in how they formed certain letters (e.g., alef, final nun, shin), how they abbreviated (using one or three dots), and how they concluded their Masorah magna notes (a mandorla shape in the case of the first Masorete, one or two colons in the case of the second Masorete). Finally, Masorete 1 did not use the upper margin to copy Masorah magna, instead concentrating his comments generally in two lines in the lower margin of each page, arranged as follows: two-thirds of the comments were placed under the two outer columns of biblical text, and the last third was placed under the inner column of biblical text. By contrast, Masorete 2 added his Masorah magna notes in both the upper and lower margins of the page and wrote in justified and uninterrupted long lines (except where the ascender of the letter lamed extended into the upper margin).

It appears that the first Masorete had originally appended an estimated total of 1,800 Masorah magna notes to the codex’s pages. Of these, 224 survive to one degree or another, the rest having been erased, presumably by the second Masorete. While the reasons for the erasure are unclear, it has been suggested that Masorete 2 may have noted the flaking of the text written by Masorete 1 (see below) and therefore decided to remove it completely and add his own notes instead. Masorete 2 probably intended to inscribe his own Masorah in place of his predecessor’s throughout the book. His success in doing so, however, was partial; it seems the work was cut short, so that the Masorah is now distributed as follows:

• On pp. 1–102, Masorete 2 generally added Masorah magna notes in three lines in the upper margins and in two lines below those of Masorete 1 in the lower margins. On pp. 95–102, his Masorah magna in the upper margins has been mostly preserved, but the lower margins have been repaired with blank slips of parchment. 
• On pp. 103–444, the second Masorete appears to have systematically erased all of the first Masorete’s Masorah magna and Masorah parva, adding his own set of both types of notes in their place.
• On pp. 445–726, Masorete 2 seems to have deleted much (though not all) of the work of Masorete 1 but only added his own Masorah parva notes, perhaps because he was interrupted before he could return to inscribe the Masorah magna (cf. p. 452). On pp. 723–726, he apparently did not even have the chance to add his own Masorah parva, with the result that the intercolumnar spaces are now almost completely blank.
• On pp. 727–746, the first Masorete’s notes remain intact, and there is no evidence of additions by the second Masorete.
• Finally, on pp. 747–792, Masorete 2 again appears to have removed the comments of Masorete 1 and replaced them with his own.

From the foregoing analysis we gain insight into the stages of the second Masorete’s work: first he erased his predecessor’s notes, then he added Masorah parva, and finally he appended Masorah magna. This order of operations may reflect a tradition among Masoretes about how to furnish a manuscript with the Masorah.

Notes written by the first Masorete generally record the number of occurrences of a given word or phrase throughout the Hebrew Bible; only two out of 224 surviving comments of his treat the subject of how a word is spelled, a favorite topic of the Masorah. Some of this Masorete’s work also focuses on the distinction between when a given word is pronounced with vs. without dagesh hazak (gemination) or on the frequency of use of a particular cantillation sign. The vast majority of these notes have their parallels in other sources, suggesting that they were copied either from the margins of other biblical manuscripts or from independent Masorah lists.

Most of the Masorah comments found in the manuscript today are the work of the second Masorete. Two facts help to date the period of his activity. First, in an important Masorah magna note inscribed at the beginning of the book of Numbers (p. 112), Masorete 2 mentions having consulted “the work of the great teacher Aaron ben Moses ben Asher in his work on the [biblical] cycle known as al-taj.” Scholars have argued that the reference here is to the Aleppo Codex, which, owing to its high degree of accuracy stemming from Ben Asher himself having edited the text repeatedly over the course of many years, was crowned with the honorific al-taj in Arabic or ha-keter in Hebrew, both meaning “the crown.” The Aleppo Codex was created circa 930 CE in Tiberias and remained there for about a century before being purchased by members of the Jewish community of Jerusalem. When the Crusaders captured Jerusalem and plundered the synagogue in which the codex was kept in 1099, Egyptian Jews sent ransom money to redeem the book, which was then transferred to Fustat. Because the second Masorete did not append the traditional blessing for the dead (“of blessed memory”) to Ben Asher’s name in the note under discussion, some have suggested that Ben Asher was still alive at the time, meaning that the second Masorete must have been working in Tiberias in the first half of the tenth century. Others have argued—based in part on paleographic grounds, in part on evidence of later scribal practice—that this omission is inconclusive and that Masorete 2 probably accessed al-taj either later in the tenth century or while it was in Jerusalem in the eleventh century.

In any case, it seems likely that this Masorete did not have to travel to Egypt to see the codex. This can be deduced from the second fact mentioned above, namely that some of this Masorete’s comments were damaged over time, including in the upper margins of pp. 5–6. Carbon-14 testing of the parchment sewn into the codex to repair these pages yielded a date in the latter half of the twelfth century. It is reasonable to assume that the work of Masorete 2 was completed several decades before the manuscript sustained the wear that necessitated its restoration. The most logical conclusion, then, is that the second Masorete was active sometime between the creation of the Aleppo Codex and the end of the eleventh century.

The notes of Masorete 2 tend to focus on orthographic issues: is a word plene, spelled with vowel letters, or defective, spelled without them—and where throughout the Hebrew Bible does it occur in that form? Many of these comments seem to be his own creations, without any known direct parallels in other collections of Masorah. In other cases, he copied preexisting Masorah notes but added the number of times a particular word occurs within the biblical corpus, or disambiguated a masoretic reference to a specific verse by including more of its words than had been provided by the source on which he was relying. His intention to be as accurate as possible in writing these comments is evidenced by his having occasionally (e.g., pp. 113, 256, 452) left space in them so that he could later return to add verse tallies or citations once he had had the opportunity to review the biblical text.

The second Masorete also functioned as a corrector, apparently using the Masorah to edit the primary scribe’s work and ensure its accuracy. This is especially true in the Torah and the book of Joshua, where the corrected biblical text demonstrates a strong affinity to Ben Asher’s tradition. We gain insight into this Masorete’s editorial methodology by noting the presence of small hook symbols suspended above about sixty-six words scattered throughout the volume, presumably meant to indicate that a correction was required.

While working on Codex Sassoon, the second Masorete made use of an impressive body of ancient masoretic material. As mentioned above, the most significant of these sources was the Aleppo Codex, cited as al-taj at the beginning of Numbers. This comment is, in fact, likely the first known reference to the Aleppo Codex and probably constitutes its earliest recorded attribution to Ben Asher.

Aside from this note, there are three additional pieces of evidence for the use of the Aleppo Codex by Masorete 2. First, in a Masorah magna comment in the lower margin of p. 452, the Masorete cites Ben Asher and the “Western” Masoretes (from the Land of Israel) as spelling the word u-refatim (Jer. 33:6) with the vowel letter yod, while the “Eastern” Masoretes (from Babylonia) spell it without that letter. It is uncommon for Ben Asher’s tradition to be contrasted with that of the Easterners; his usual opponent in masoretic matters is Moses ben David ben Naphtali. This likely indicates that the second Masorete was not simply copying from a previous scholar’s masoretic treatise but instead had access to the Aleppo Codex and that he checked its text when writing this comment.

Second, in a Masorah parva note near the lower margin on p. 538 Masorete 2 records that the word va-taggishun (Amos 6:3) appears nowhere else in the Hebrew Bible and is spelled defective (without the yod) according to Ben Asher. (In fact, it seems that he corrected the primary scribe’s work here, erasing the yod that had originally mistakenly been included.) Ben Asher’s masoretic debates generally revolve around issues of vocalization, cantillation, and stress rather than orthography. This again suggests that the second Masorete was able to consult the Aleppo Codex when composing his Masorah.

Third, Masorete 2 compiled a long list of the Aramaic words in the Hebrew Bible that end with the letters he or yod (as opposed to alef) and the vowels tsereh or kamats, which he divided into five parts that he then copied onto folios containing Aramaic text from the books of Daniel and Ezra (pp. 752–753, 770–771). The contents of this list closely mirror those of a page of Masorah material that had been appended to the front of the Aleppo Codex but is no longer extant, and so it is highly probable that the second Masorete used al-taj when he worked on these notes.

In addition to the Aleppo Codex, Masorete 2 cites several other ancient Tiberian authorities, including the Masoretes Moses ben Mohah and Phinehas Rosh ha-Yeshivah (p. 330), as well as the now-lost model codex known as the Mahzora Rabbah (pp. 184, 330, 751). Moreover, in twenty-five instances (including the one mentioned above) he explicitly quotes the Masorah of the Easterners from Babylonia to contrast it with that of the Westerners from the Land of Israel. He even incorporates some of the terminology unique to the Babylonian Masorah into his own comments, including language like de-hakha, de-gabbei, be-kullam, and shalma de-shalma. All of this indicates that the range of his masoretic scholarship and of the sources to which he had access was wide.

Like many other medieval volumes of a similar age, Codex Sassoon sustained wear and tear over the centuries. Some of this wear was surely due to normal use, and some probably derived from the act of erasure of the first Masorete’s notes on both sides of a given page. The book may even have been damaged during the aforementioned destruction of the town of Makisin.

The manuscript is remarkable, however, for the lengths to which its owners went to preserve it; multiple conservation efforts are evident in its pages. These include: the careful stitching of tears that opened up along the parchment’s deeply scored ruling lines; the re-inking (using carbon-based black ink) of writing that had faded or flaked, especially on the flesh side of the parchment; and the attachment of slips of parchment (some of them with text written on them) to shore up margins that had worn thin or eroded over time. On several folios where repairs resulted in the loss of biblical text, later hands—at least one of them dating from the twelfth or thirteenth century (pp. 5–6)—filled in what was missing. Carbon-14 testing of three pieces of restoration parchment yielded likely dates in the latter half of the twelfth century (pp. 5–6, 173–174) and the latter half of the fourteenth century (pp. 253–254), suggesting repeated investment in the codex’s preservation due to a recognition of its value and importance.

In addition to restoration, Codex Sassoon underwent rebinding, probably on at least two occasions in premodern times. As mentioned above, the first likely took place in the thirteenth century or later, when the volume was consecrated to the synagogue of Makisin. The same hand that wrote the dedication inscription on pp. 567 and 587 also added Hebrew-character signatures, or quire marks, in the upper-inner corner of the first page of each quire (some of which are no longer visible due to wear or repair), presumably to aid the binder in reassembling the quires in their proper order.

A second hand inserted pairs of identical symbols in the lower margins of a number of the book’s openings, usually at (or around) the end of one quire and the beginning of the next one. These symbols are reminiscent of the ones used in medieval Christian Arabic manuscripts bound in major monasteries in Syria and the Sinai, suggesting that Codex Sassoon was entrusted to Christian craftsmen at some stage. Perhaps it was these (or later) binders who trimmed the margins of the manuscript’s leaves, inadvertently removing some of the masoretic and biblical text in the process. Still, their efforts helped protect the volume from further damage, preserving it for the benefit of future generations of students and scribes.

Codex Sassoon, created circa 900 CE, is the earliest most complete example of a single volume containing all twenty-four books of the Hebrew Bible. As such, it helps illuminate the process by which the Masoretic Text was standardized and transmitted among medieval Middle Eastern Jewish communities. It also contains invaluable masoretic information unavailable from other sources and gives us insight into the methods used by the Masoretes to record that information. The care and dedication with which Codex Sassoon was repeatedly repaired over the centuries, as well as its consecration to a synagogue, give eloquent testimony to the esteem in which it was held by its owners and users. It is only fitting that such a significant and valuable witness to the masoretic tradition should have entered the Sassoon collection, one of the largest and most important private Judaica libraries ever built.

In the Middle Ages, in the absence of a functioning temple in their ancestral homeland, Jewish communities began to conceive of the Hebrew Bible itself as a portable temple. One of the earliest sources for this idea is Sefer dikdukei ha-te‘amim, authored by none other than Aaron ben Moses ben Asher. Toward the beginning of this work, Ben Asher explains that the “Torah, Prophets, and Writings … they all correspond to this order in the temple: the Holy of Holies, the Holy, and the Courtyard of the Tent of Meeting.” The parallelism thereby established between the historical, structural temple and the contemporary, textual temple appears in turn to have influenced the coining of a term current among later medieval Spanish Jewry: mikdashyah, the sanctuary of God. The Hebrew Bible-as-mikdashyah became a site for contemplation, communion, and connection, both with the Divine and between generations.

It seems appropriate, then, to close this catalogue note with the blessing for Isaac ben Ezekiel al-Attar and his descendants included in the aforementioned deed of sale (p. 373):

The order of the biblical books in Codex Sassoon mirrors that of other Masoretic Bibles such as the Aleppo Codex and the Leningrad Codex:

Torah (pp. 1–190)
pp. 1–39: Genesis
pp. 39–82: Exodus
pp. 82–111: Leviticus
pp. 111–153: Numbers
pp. 153–190: Deuteronomy

Former Prophets (pp. 191–373)
pp. 191–217: Joshua
pp. 217–243: Judges
pp. 243–306: I–II Samuel
pp. 306–373: I–II Kings

Latter Prophets (pp. 374–565)
pp. 374–419: Isaiah
pp. 419–475: Jeremiah
pp. 475–526: Ezekiel
pp. 526–565: Twelve Minor Prophets

Writings (pp. 566–792)
pp. 566–632: I–II Chronicles
pp. 632–686: Psalms
pp. 687–708: Job
pp. 708–725: Proverbs
pp. 725–728: Ruth
pp. 728–732: Song of Songs
pp. 732–739: Ecclesiastes
pp. 739–743: Lamentations
pp. 743–751: Esther
pp. 751–766: Daniel
pp. 766–792: Ezra-Nehemiah

Format: 792 pages (approx. 13 3/4 x 11 1/2 x 6 1/8 in.; approx. 350 x 290 x 155 mm) (collation: i1 [i1–9 lacking], ii10, iii7 [iii1–3 lacking], iv–x10, xi8, xii–xxviii10, xxix8, xxx9 [xxx8 canceled], xxxi–lx10, lxi8, lxii5 [lxii6 canceled]) on Eastern raqq-type parchment; premodern quire signatures in Hebrew characters in upper-inner corner of first page of each quire (often damaged or obscured toward front of volume); premodern binder’s marks in lower margins at and around start of each quire; modern pagination in pencil in Arabic numerals in lower margins at center (cited), p. 291 left unnumbered; modern (binder’s?) foliation in pencil in Arabic numerals in upper-outer corner of recto at somewhat regular intervals on pp. 201–687; written in Eastern square script in iron-gall and mixed inks; generally triple-column text of 28 lines per column (often 29 in the Writings) (though cf., e.g., pp. 571–572 [26 lines], 333–334 [27 lines], 715–718 [30 lines]); the biblical songs, the books of Psalms, Job, and Proverbs (pp. 632–725), and the beginning of the book of Ruth (p. 726) laid out differently; ruled in blind on the flesh side; justification of lines via dilation or contraction of final letters and use of anticipatory letters and space fillers (average justification: approx. 10 7/8 x 10 3/8 in.; approx. 275 x 265 mm); biblical text usually fully vocalized and accentuated in light brown ink (except where it repeats, e.g., pp. 120–121, 682); Masorah magna (on some pages) and Masorah parva written in micrography in margins (Masorah parva erased from pp. 717, 719, 723–726 and not replaced); where Masorah magna is present, it generally occupies three lines in upper margins and four lines in lower margins and sometimes includes vocalization and/or accentuation; small semicircles placed above words to which Masorah notes refer; three letters pe placed at beginning (e.g., p. 463) or end (e.g., pp. 372, 447, 499, 535, 549, 671) of column to indicate the space was intentionally left blank; space fillers added to lines (e.g., pp. 346, 371) to indicate the space was unintentionally left blank; no catchwords; the letter pe generally added in the margins to indicate the start of a new parashah in the Torah, often accompanied by a tally of verses placed at the end of the previous parashah; the letter samekh added sporadically in the margins of the Torah (e.g., pp. 12, 74–77, 163–169, 177–178, 185) but fairly regularly in those of the Prophets and Writings to indicate the start of a new seder; later headers added to the Prophets and Writings; later numeration of the psalms; corrections, marginal insertions, and strikethroughs in hands of primary and subsequent scribes (e.g., pp. 12, 67, 78, 84, 204, 266–267, 309, 314, 338, 347, 410, 466, 482, 507, 543, 551, 601, 619, 681, 716).

Special Features: Scribe ends each biblical book through Psalms by spacing out its last lines; within Psalms, its latter four sections are separated from the preceding section using two blank lines (pp. 646, 657, 665, 671); pp. 54–55, 187–188, 222–223, 302–304 feature biblical songs with ornamental layouts; pp. 204, 248, 277, 472, 583, 589–591, 768, 772 feature other texts with special layouts; Psalms, Job, and Proverbs are laid out stichographically, as is traditional; Masorah parva marks certain textual halfway points (e.g., pp. 18, 88, 92, 172, 500, 545, 681, 747).

Condition: Probably lacking about twelve whole folios, eight containing biblical text and four others at the front which may have contained frontmatter (nine folios before p. 1 and three between pp. 22–23); parts of outermost column and/or upper/lower margins of pp. 1–22, 27–36, 635–642, 647–686, 689–726 lacking (mounted on modern paper); two outer columns of pp. 23–26 and outermost column of pp. 687–688, 727–792 lacking (mounted on modern paper); premodern repairs of parchment (sometimes including restoration of text) frequently throughout, though at times starting to come undone; pages closely cropped, sometimes resulting in loss of masoretic or biblical text; text often partially abraded (sometimes re-inked) on flesh side, in part due to twentieth-century application and removal of Japanese conservation paper; scattered staining; episodic small holes in text; occasional tape repairs (loose along lower edges of pp. 215, [291]); lower margins of pp. 79–80, 453–454, 547–548, 571–574 excised but not replaced with newer parchment; the bifolium comprising pp. 299–300, 305–306 does not observe Gregory’s rule; modern repairs in upper edges of pp. 429–430, lower edges of pp. 159–160, 619–626, 645–646, and intermittently in outer edges; small puncture on pp. 599–600; pp. 669–672 loose.

Binding: Bound in whole dark goat over boards with leather joints, paneled in blind, slightly worn in places; turn-ins blind-tooled; spine in six compartments with raised bands; British Library shelf mark (“Or. 62f”) pasted in upper and lower compartments, Sassoon Library shelf mark (“1053”) numbered in gilt in lower compartment; modern paper flyleaves and pastedowns; Sassoon bookplate added to pastedown of upper board.

Aaron ben Moses ben Asher, Sefer dikdukei ha-te‘amim, ed. Seligman Baer and H. L. Strack (Leipzig: L. Fernau, 1879), 2.

Anon., “Manuscript Sale Brings in £900,000,” The Daily Telegraph (November 22, 1978): 15.

Anon., “Hebrew Manuscripts from the Sassoon Collection: An Exhibition in the British Library 10 August–31 December 1982,” 3 (no. 1).

Anon., “Hebrew Bible Sold for $3.19 Million,” The Washington Post (December 6, 1989).

Mordechai Breuer, The Aleppo Codex and the Accepted Text of the Bible (Jerusalem: Mossad Harav Kook, 1976), viii, xi–xii, xxxii, 90, plate 7.

Mordechai Breuer, Nussah ha-mikra be-“keter yerushalayim” u-mekorotav ba-masorah u-be-kitvei ha-yad (Jerusalem: Keren ha-Masorah, 2003), 9–10.

Brian L. Frye, “New Art for the People: Art Funds & Financial Technology,” The Chicago-Kent Law Review 93,1 (March 2018): 113–139.

Dov Genachowski, “Ha-shiddukh ha-muzar bein mishpahat sassoon le-po‘alei ha-rakevet ha-britim,” Yedi‘ot aharonot (November 24, 1989): 18.

Mordechai Glatzer, “Melekhet ha-sefer shel keter aram tsovah ve-hashlakhoteha,” Sefunot 4 (1989): 167–276, at pp. 256 n. 15, 260–263.

Norman Golb, “The Autograph Memoirs of Obadiah the Proselyte of Oppido Lucano and the Epistle of Barukh b. Isaac of Aleppo,” v–vi, 5, available at: https://oi.uchicago.edu/sites/oi.uchicago.edu/files/uploads/shared/docs/autograph_memoirs_obadiah.pdf (accessed February 19, 2023).

Rachel Heuberger, Aron Freimann und die Wissenschaft des Judentums (Tübingen: Max Niemeyer Verlag, 2004), 15–16.

Neriah Klein, “Ha-keter be-fustat: keter aram tsovah bi-te‘udot min ha-genizah ha-kahirit,” Ginzei kedem 17 (2021): 179–209, at pp. 182–183.

Guy Le Strange, The Lands of the Eastern Caliphate: Mesopotamia, Persia, and Central Asia from the Moslem Conquest to the Time of Timur (Cambridge: Cambridge University Press, 1905), 97.

Jacob Mann, The Jews in Egypt and in Palestine under the Fāṭimid Caliphs, vol. 2 (Oxford: Oxford University Press; London: Humphrey Milford, 1922), 29–30.

Yosef Ofer, “Keter aram tsovah—le-or reshimotav shel m. d. cassuto,” Sefunot 4 (19) (1989): 277–344, at pp. 302–304.

Yosef Ofer, “A Masoretic Note in the Aleppo Codex Concerning the Composite Words,” Textus 21 (2002): 209–233.

Yosef Ofer, “Kunteresei ha-masorah shel keter aram tsovah al pi he‘etek be-arkhiyyono shel yitshak seligman baer,” in Shmuel Vargon et al. (eds.), Iyyunei mikra u-parshanut: minhot yedidut ve-hokarah li-menahem kohen (Ramat Gan: Bar-Ilan University Press, 2005), 137–162, at pp. 147–148, 157–160.

Yosef Ofer, The Masora on Scripture and Its Methods, trans. Michael Glatzer (Berlin: De Gruyter, 2019), 8, 43, 46, 134, 199.

Judith Olszowy-Schlanger, “The Hebrew Bible,” in Richard Marsden and E. Ann Matter (eds.), The New Cambridge History of the Bible, vol. 2 (Cambridge: Cambridge University Press, 2012), 19–40.

Meir Persoff, “Hebrew Bible Fetches £2m,” The Jewish Chronicle (December 8, 1989): 48.

David Solomon Sassoon, Ohel Dawid: Descriptive Catalogue of the Hebrew and Samaritan Manuscripts in the Sassoon Library, London (Oxford: Oxford University Press; London: Humphrey Milford, 1932), xiii, xxii, xliv, xlviii, 528 n. 1, 1111–1112 (no. 1053), plate 6.

Sotheby’s, Ohel Dawid: The Library of Hebrew and Samaritan Manuscripts Collected by Rabbi David Solomon Sassoon (London: Sotheby & Co., 1975?), plate 2.

Sotheby’s, A Further Thirty-Three Highly Important Hebrew and Samaritan Manuscripts from the Collection Formed by the Late David Solomon Sassoon … Tuesday 21st November, 1978 (Zurich: Sotheby Parke Bernet A.G., 1978), 14–16 (lot 6).

Sotheby’s, Western Manuscripts and Miniatures, London Tuesday 5th December 1989 (London: Sotheby’s, 1989), 47–51 (lot 69) = Sotheby’s, Four Extremely Important Hebrew Manuscripts from the Tenth to the Fifteenth Century: The Property of the British Rail Pension Fund, London Tuesday 5th December 1989 (London: Sotheby’s, 1989), lot 69.

David Stern, The Jewish Bible: A Material History (Seattle: University of Washington Press, 2017).

David Stern, “On the Term Keter as a Title for Bibles: A Speculation about Its Origins,” in Shmuel Glick, Evelyn M. Cohen, and Angelo M. Piattelli (eds.), Meḥevah le-Menaḥem: Studies in Honor of Menahem Hayyim Schmelzer (Jerusalem: JTS-Schocken Institute for Jewish Research, 2019), *259–*273.

Israel Yeivin, Keter aram tsovah: nikkudo u-te‘amav (Jerusalem: Magnes Press, 1968), 3, 79, 362–364.

Israel Yeivin, Mivhar kitvei-yad shel ha-mikra (Jerusalem, 1972), plates 31–36.

Israel Yeivin, Introduction to the Tiberian Masorah, trans. and ed. E. J. Revell (Missoula, MT: Scholars Press for the Society of Biblical Literature and the International Organization for Masoretic Studies, 1980), 21–22.

Israel Yeivin, Ha-masorah la-mikra (Jerusalem: The Academy of the Hebrew Language, 2003), 9, 19.