skip to content

New Pages of Sefer Tagin: JTS ENA 2753.16 and National Library of Russia (NLR) Box N.91

Joseph Ginsberg

I’ve recently spent a lot of time working my way through the ENA collection at the Jewish Theological Seminary in New York with some colleagues from the Princeton Geniza Lab. One benefit of browsing through hundreds of Genizah fragments, many only lightly cataloged, is the unplanned discoveries that happen along the way. It was while surfing the Genizah’s cool waves of dust and paper that we stumbled upon a peculiar fragment filled with strange looking letters and markings, which I excitedly identified as a new fragment of a rare medieval text, Sefer Tagin.1

Sefer Tagin (or Tagei2) has been featured in this space more than once in the past.3 It is an ancient treatise which lists out certain letters, called Otiyyot Meshunnot, that should be written in Torah scrolls in a strange or unusual manner, often by adding little crownlets (Tagin) or whorls to them.4 In many versions of the text, the list is introduced by a rather legendary account that traces the history of the text all the way back to the stones that Joshua erected in Gilgal, which were inscribed with the text of the Pentateuch and plastered for posterity.5

Unfortunately, Sefer Tagin was preserved very poorly, and almost no copies of the complete text survive. The editio princeps, based on a manuscript in Paris (and only published in 1866!) seems to be riddled with errors.6 In fact, so little is known about the text and its history that scholars in the modern day have been forced to try and cobble together some sort of urtext, piecemeal, from bits of manuscript and quotation – an endeavor that still has much ground to cover.7 Moreover, comparing the text between the different textual witnesses that have survived, most of which are relatively late, reveals extraordinary variation. This is what makes Genizah fragments of Sefer Tagin so important, and Marc Michaels has done an excellent job describing and analyzing nearly all of the known Sefer Tagin fragments from the Genizah in his recent monograph on the subject.8 Still, how many editions, revisions, and reworkings of the text circulated in antiquity is unclear, though it typically appears in one of two formats: the list can either be organized in order of each embellished letter (lexically),9 or in order of the biblical text.10 Both of these new fragments, JTS ENA 2753.16 and NLR Box N.91, belong to the latter category.

As an aside, it is generally believed that the former recension is the original – it makes more sense that someone would have reorganized the list from alphabetical into lexical order for easier scribal use than the reverse, though that does not help date any particular version of the text, since it is impossible to know when the material may have been reorganized. 

Sefer Tagin, especially the versions ordered according to the biblical text, served as a general Tiqqun Soferim – a model guide for scribes – and so often contain interpolative notes that relate to characteristics of Torah scroll production in general. The most common of these notes11 indicate which sections (Parshiot, or pericopes) should be left open (פתוחה) and which closed (סתומה). Yedidya Shelomo Norzi, in his Minḥat Shai, makes use of these notes several times when discussing variant traditions surrounding the pericopes.12 These notes are almost certainly not original to Sefer Tagin itself, and were presumably inserted later once the text was reworked into a broader scribal manual. With the addition of these notes, Sefer Tagin might be considered a pseudo-Masoretic work, and, depending on when the text was formalized, a prototypical one as well.13 

The first fragment, JTS ENA 2753.16, is a rather neat page of unlined paper (measuring around 18.2 × 13.4 cm) containing 15 lines of rabbinic semi-cursive text, likely from the 12-13th centuries.14 The verso is blank. There are several small holes in the paper, and one somewhat larger tear in the center, but these lacunae are small enough that whatever text has been lost is almost completely reconstructable. There are a few illustrations in the margins of some forms of the Otiyyot Meshunnot included in Sefer Tagin, but they do not seem to be directly related to the text itself. Interestingly, the page begins with a standard introductory formula and a single line of Sefer Tagin to Genesis before abruptly skipping to the beginning of Exodus. The text may have been compiled directly from a Torah scroll or itself copied from an earlier text of Sefer Tagin.


JTS ENA 2753.16

Detail from JTS ENA 2753.16, recto, published with courtesy of the Jewish Theological Seminary Library.


This fragment is clearly connected to T-S Misc.24.182, another two-page fragment of Sefer Tagin identified by Dr Kim Phillips in 2019. The sheets of T-S Misc.24.182 are the same size and material as ENA 2753.16. The handwriting is extremely similar (almost certainly from the same scribe) and the number of lines are around the same. The style of Otiyyot Meshunnot, which can greatly vary between different sources, matches. But are they part of the same copy of Sefer Tagin? This is – surprisingly – a difficult question to answer, because ENA 2753.16 and T-S Misc.24.182 both contain Sefer Tagin to the beginning of Exodus.


T-S Misc.24.182 P2 recto

 T-S Misc.24.182 2r


There are two ways to resolve this problematic occurrence, neither completely satisfactory, and it may not be possible to determine which one is correct:

●    ENA 2753.16 may have been used by the scribe of T-S Misc.24.182 as scrap or practice. The former's blank verso and marginal doodles lend themselves to such an explanation, and this might also help explain the sudden discontinuity – from Genesis to Exodus – in the text.

●    These two shelfmarks might be from two separate copies of Sefer Tagin, either both copied from the same source or one copied from the other.15 In truth, the scripts are not perfectly identical: for example, the final Nun in ENA 2753.16 is the mirror image of the one in T-S Misc.24.182, which would make sense if these are two separate but near-contemporary copies. 

As mentioned earlier, the text of Sefer Tagin in ENA 2753.16 begins with a standard introductory formula (an abbreviation of a Psalmic verse) followed by a single line of letter embellishments to the beginning of Genesis. On the following line, the list skips to the beginning of Sefer Tagin to Exodus, which occupies the rest of the page. This jump – whether deliberate or erroneous – is very strange, but it may be due to the following error. The first entry in every edition of Sefer Tagin is the word אלהים from Genesis 1:1. However, the scribe of ENA 2753.16 seems to have accidentally skipped it and began the list with the second and third entries, ורוח and מרחפת from Genesis 1:2. Upon writing out the fourth entry, which is אלהים again, this time from Genesis 1:3, the scribe likely realized that they had skipped the same word earlier on, and set aside the page (without adding any embellishment to אלהים) – only returning to it later, having decided to use the rest of the space for Sefer Tagin to Exodus. 

Speculation about the fragment’s history aside, its real value lies in the textual variants it gives for Sefer Tagin, especially for the part – most of the fragment – that is not duplicated in T-S Misc.24.182. The following is a transcription of ENA 2753.16 with comparison to T-S Misc.24.182, as well as the variants included in a recent edition of Sefer Tagin.16


[Unusual or embellished letters are given in bold. For their forms, see the manuscript image above and the chart below. Lacunae and reconstructions are marked with square brackets.]

1. ע׳מ׳י׳ ע׳ש׳ו׳

This is an abbreviation of עזרי מעם יהוה עשה שמים וארץ (Psalms 121:2), used as introduction.

2. בראשית ורוח מרחפת (Genesis 1:2) ויאמר אלהים (1:3)

The list opens with the beginning of Genesis, as noted by the first word. The scribe, as mentioned earlier, has skipped over the first entry to Sefer Tagin of Genesis, אלהים from 1:1, and has not embellished the second אלהים, from 1:3, at all. 

3. ויקם פ׳ (Exodus 1:8) הנה עם (1:9) נתחכמה פן ירבה (1:10)

The list for Exodus begins by noting that Exodus 1:8 heads an open Parshah. Other sources include ירך יעקב of Exodus 1:5 and ותמלא of 1:6. It appears that ירבה from the end of this line is an error, because it is repeated on the following line without any embellishment for the Resh, which is the way this word is included in T-S Misc.24.182. However, most sources only note an embellishment to the Yod, just as this list has for the ירבה of 1:12.

4. ירבה (1:10) וישימו (1:11) יענו ירבה יפרץ (1:12) קשה (1:14) העבריות (1:16)17

Most sources do not include the ירבה of 1:12 to begin with, and only embellish the earlier word from 1:10.

5. פתם (1:11) ויצו (1:22) טוב ותצפנהו (2:2) ותחמרה (2:3) ותתצב (2:4)

It appears that the scribe accidentally left out פתם from 1:11 earlier, and so added it here. T-S Misc.24.182 contains this same transposition, and adds in וילך ס׳ after ויצו, indicating that Exodus 2:1 heads a closed Parshah. This is corroborated by JTS MS 2968, which contains a standard list of the pericopes, and, in the margins, comparative annotations that reference Sefer Tagin. Other sources also add an embellishment to the final Kaph of that וילך as well as ירחים from 2:2.

6. ונערתיה (2:5) ותפתח העברים זה (2:6) היליכי (2:9) משיתהו (2:10)

Other sources add לדעה from 2:4 and נער בוכה from 2:6, but do not include היליכי at all.

7. ויהי פ׳ (2:11) ויך (2:12) ויבקש (2:15) עזבתן (2:20) ויהי פ׳ ויאנחו (2:23)

Exodus 2:11 and 2:23 are each noted to be an open Parshah. In T-S Misc.24.182, עזבתן is erroneously added above the line before the first ויהי פ׳ of 2:11, probably due to confusion with the identical ויהי פ׳ of 2:23 which it should actually precede (as it does here). Sefer Tagin for this part of Exodus in T-S Misc.24.182 ends here.

8. ויזעקו (2:23) נאקתם ויזכר (2:24) וידע א[לה]ים (2:25) ומשה ס׳ חרבה (3:1)

Exodus 3:1 is noted to be a closed Parshah.

9. אסרה (3:3) מתוך הסנה (3:4) [תקרב18] הלם של נעליך אדמת קדש (3:5)

Other sources include the Samech of אסרה as well, and ויקרא of 3:4. The inclusion of מתוך here is very interesting: all the unusually written Vavs in Sefer Tagin are conjunctive prefixes, never a Vav from the middle of a word. No other source includes this entry.

10. ורחבה הכנעני (3:8) צע[קת] (3:9) [עו]ד אלהים (3:15) לך (3:16) זכרי (3:15) זקני (3:16)

Other sources include כה תאמר of 3:14. The last part of this line is misordered – presumably, זכרי of 3:15 was accidentally left out and then added later.

11. ואני ידעתי (3:19) נפלא[תי] (3:20) [ו]שאלה אשה (3:22) השליכהו (4:3) ויוצאה

The Nun of ואני ידעתי has the rare distinction of being one of the few Otiyyot Meshunnot mentioned in the Masorah,19 which notes that this letter is a נון עקומה, or twisted Nun.20 Other sources include the Nun of נפלאתי and והכתי of 3:20, but not אשה or השליכהו at all. The line’s last entry, ויוצאה, is an interesting case – no other source includes an embellishment for this word, but it is also not entirely clear what the scribe’s intention is here. The Tsadi is written large and rigid, and is likely referring to a rare form of uncrowned Tsadi which in most versions of Sefer Tagin only appears twice in the Pentateuch.21

12. [מצ]רעת (4:6) ישמעון22 ושפכת (4:9) והוריתיך (4:12) ויחר אף (4:14)

Other sources include the final Nun of ישמעון, but not the Ayin. Nor do they include the final Kaph of והוריתיך.

13. והוריתי (4:15) המטה הזה תעשה האתת (4:17) וילך פ׳ לך

Other sources do not include the Mem of המטה or תעשה. Exodus 4:18 is noted to be an open Parshah.

14. לשלום (4:18) מתו המבקשים (4:19) אחזק לבו (4:21) ו(ה)[ת]גע לי (4:25) לך

Other sources include the Qoph of המבקשים but not the Mem, though some include both. The word תגע has no embellishment,23 and no other source includes it.

15. ויפגשהו וישק (4:27) וילך ויאספו (4:29) ראה (4:31) וישפט (5:21)

Other sources include וימאן of 4:31 and צעקים of 5:8.

Overall, ENA 2753.16 largely matches other versions of Sefer Tagin, but is a bit sparser. It also includes some very peculiar variants. 

It was while preparing this piece that I stumbled upon a second new fragment of Sefer Tagin in an entirely different Genizah collection. The Russian National Library, Box N is an unorganized jumble of mostly liturgical fragments, assumedly from the Genizah. Hidden among them – under the shelfmark N.9124 – is a single, well-worn page which contains Sefer Tagin to the beginning of Genesis. Unfortunately, this may be the lone leaf that has survived: it seems there are no more pages of this compilation under any of the “box” shelfmarks.

This fragment, unlike ENA 2753.16, has writing on both sides of the page, each containing around 15 lines of text. The recto is in poor condition: the paper has large stains and the ink has been rubbed off in several areas, making many words difficult or impossible to read, though it is possible to reconstruct nearly everything by closely reading the faded text, paying careful attention to spacing, and comparing to other versions of Sefer Tagin. The verso, which continues the list (ending with Genesis 5:1) is in better condition and entirely readable. It is difficult to determine the age of this fragment without a color photograph, which would likely reveal much about the type of paper, ink, and hand employed. With a necessarily wide range of confidence, it was probably written between the 12th and 14th centuries. 


detail from NLR Box N.91

Detail from NLR Box N.91 (from the collections of The National Library of Russia and The National Library of Israel's “Ktiv” Project)


This fragment was clearly part of a larger Masoretic compendium,25 as the top line of the recto ends a standard Masoretic list of the letters in the Bible that are written miniscule:

26 …ש׳ ואת פרמשתא     [ת׳ ו]את פרשנדתא 

What follows is a list of the Tagin, titled “אילו תגין דתורה” – “These are the Tagin of the Torah.” Though the list is ordered according to the biblical text, it does not contain any interpolated Masoretic notes (like those in ENA 2753.16). As such, it might be considered a somewhat more “pure” version of the text. The scribe was careful to distinguish between words belonging to the biblical text, which are written in a formal hand, and words that clarify which verse the list is referring to, which are written in a less formal rabbinic semi-cursive. This, too, might point to the accuracy of the text. The Tagin are thick and noticeable, and there are dots included in between each entry in order to keep everything properly spaced. As with ENA 2753.16, the text may have been compiled directly from a Torah scroll or itself copied from an earlier text of Sefer Tagin. The following is a transcription of NLR Box N.91 with comparison to the text of the same recent edition of Sefer Tagin.


[Unusual or embellished letters are given in bold. For their forms, see the manuscript image above and the chart below. Lacunae and reconstructions are marked with square brackets.] 



1. איל[ו] תגין דתורה

2. אלהים27 דבראשית (Genesis 1:1) ◦ ורוח אלהים [מרחפ]ת (1:2) [◦ אלהים]

Other sources add the Ḥet of ורוח and מרחפת (though they are not necessarily in agreement about each letter). This fragment may have included other embellishments to מרחפת as well.

3. דיהי אור ויהי אור (1:3) ◦ אלהים דאת האור ◦ כ[י טוב ◦] אלהים

Other sources include יהי אור of 1:3, but do not include either אלהים of 1:4.

4. דבין האור ובין (1:4) ◦ ויקרא [אל]הים לאור [◦ ולח]שך קרא לילה (1:5) ◦

Other sources include ויקרא of 1:5, but do not include אלהים.

5. אלהים דיהי רקיע (1:6) ◦ אלהים דאת הרקיע (1:7) ◦ ו[יק]רא

Other sources do not include אלהים of 1:7.

6. אלהים לרקי[ע (1:8) ◦] אלהים דיקוו המים (1:9) ◦ ויקרא

Other sources do not include אלהים of 1:8 or 1:9

7. [א]להים ל[יבשה ◦] וירא אלהים כי ט[וב (1:10) ◦] אלהים

Other sources do not include any embellishments for 1:10.

8. דתד[שא הארץ (1:11) ◦] וירא אלהים דיום שליש[י (1:12) ◦ א]להים

Other sources do not include אלהים of 1:12.

9. ד[יהי מארת] (1:14) ◦ אלהים דשני המאורות28 (1:16) ◦ אלהים 

Other sources do not include אלהים of 1:16.

10. דברקיע השמים (1:17) ◦ אלהים דיוםרי29 די[ו]ם רביעי (1:18) ◦

Other sources do not include אלהים of 1:17 or 1:18, but include ולמשול of 1:18.

11. אלהים דישרצו המים (1:20) ◦ אלהים דאת התנינם (1:21) ◦

Other sources do not include אלהים of 1:21.

12. [א]להים דפרו ורבו (1:22) [◦] אלהים דתוצא הארץ (1:24) ◦

Other sources include ויברך of 1:22 (which this fragment includes later, in 2:2), but do not include אלהים.

13. [אל]הים דתוצא הארץ (1:24) ◦ אלהים דאת חית [ה]ארץ ◦

The first entry of this line has been accidentally duplicated from the previous line. Other sources do not include any embellishments to 1:25.

14. אלהים דכי טוב (1:25) [◦] אלהים דנע[ש]ה אד[ם] ב[צ]למנו

Other sources do not include אלהים of 1:26, but include נעשה.


1. כדמותנו (1:26) ◦ ויברא אלהים דבצלמו ◦ אלהים דברא

Other sources do not include any embellishments to 1:27.

2. זכר ונקב׳ (1:27) ◦ ויברך אלהים [◦] ויאמר אלהים (1:28) [◦] אלהים

Other sources include ויברך of 1:28 (which this fragment includes later, in 2:2). A few sources do include the first אלהים of 1:28, but not the second.

3. דהנה נתתי (1:29) [◦] ירק עשב (1:30) ◦ אלהים דאת כל אשר עשה

Other sources do not include אלהים of 1:29, or any embellishments to 1:31.

4. והנה טוב מאד (1:31) ◦ ויכל אלהים (2:2) [◦] ויברך אלהים30

Other sources include ויכל of 2:2, as well as ויקדש and מלאכתו of 2:3. A few sources also include ברא of the same verse.

5. אלהים דלעשות (2:3) [◦] בהבראם [◦] אלהים דארץ ושמים (2:4) [◦]

Other sources do not include the He of בהבראם, but include עשות of 2:4. It is possible that this manuscript also embellishes the Resh of בהבראם, though this is unlikely – on the following line, where this word is accidentally rewritten, it does not include any Tagin. The single Tag included here looks unlike all the others, and it may just be a slip of the pen or a tear in the paper. Unfortunately, there are no other instances of an embellished Resh in this fragment to compare to.

6. שיח השדה (2:5) ◦ אלהים אלהים דלעשות בהבראם31

Other sources do not include אלהים of 2:7, but include עפר and נשמת, as well as לנפש.

7. דאת האדם (2:7) ◦ ויטע אלהים דגן עדן (2:8) [◦] ויצמח אלהים

Other sources do not include אלהים of 2:8 or 2:9.

8. דמן האדמה [◦] נחמד למראה [◦] למאכל ועץ (2:9) [◦] פישון דהוא

Other sources include למראה and וטוב of 2:9.

9. הסובב32 (2:11) [◦] גיחון הוא הסובב (2:13) [◦] ויקח אלהים דאת האדם (2:15) [◦]

Other sources include any embellishments for 2:15.

10. אלהים דעל האדם (2:16) ◦ אלהים דלא טוב היות האדם (2:18) ◦

Other sources do not include any embellishments for 2:16 or 2:18.

11. ודבק באשתו (2:24) [◦] ערומים דהאדם ואשתו (2:25) [◦] אלהים

Other sources include ויישן and בשר of 2:21, as well as באשתו of 2:24.

12. דלא תאכלו (3:1) ◦ כי טוב העץ (3:6) ◦ ותפקחנה דעיני [◦] עירמם

Some sources include ערום of 3:1. Most sources include אף of the same verse, and a few sources include אלהים of 3:1, though most do not. Other sources include ונחמד of 3:6, though not עירמם of 3:7.

13. הם ויתפרו (3:7) [◦] כי עירם אתה (3:11) ◦ הנחש השיאני (3:13) [◦] גחנך תלך (3:14) [◦]

Other sources include מתהלך of 3:8, האדם of 3:9, both קלך and ואחבא of 3:10, and ועפר of 3:14.

14. לקול אשתך (3:17) ◦ תצמיח לך (3:18) [◦] כתנות עור וילבשם (3:21) [◦]

Other sources include בעבורך of 3:17 and ודרדר of 3:18.

15. לדעת טוב ורע (3:22) [◦] וישלחהו דמגן עדן (3:23) ◦ להט החרב המתהפכת (3:24) [◦]

Other sources are in agreement here.

16. ותאמר דקניתי איש (4:1) [◦] למה חרה (4:6) [◦] לקחת את דמי אחיך (4:11)

Other sources do not include למה of 4:6.

17. ומפניך אסתר (4:14) ◦ לקרא בשם (4:26) [◦] אדם דביום ברא אלהים (5:1)

Other sources include פני of 4:14, מלפני of 4:16, and both לפצעי and לחברתי of 4:23. They also include ביום of 5:1.


This fragment seems to contain a somewhat alternative recension of Sefer Tagin. With regards to the word אלהים, on the one hand, the use of embellishments is greatly expanded relative to all known versions – this fragment notes an embellishment to nearly every single instance of אלהים in the first chapter of Genesis! (The lone exception is the second instance of אלהים in 1:21.) On the other hand, when it comes to other words, this fragment is abstemious, marking fewer embellishments than what is typical. Interestingly, this contraction seems to be non-random – some of the differences between this fragment and other versions of Sefer Tagin appear to be systematic. This manuscript, at least for the section that is extant, never marks any embellishment for Shin33  (typically adorned with seven Tagin), and, likely, nor for Resh34 (typically adorned with one Tag), which other sources include around seven times each in the same portion of the Pentateuch. There also never appears the famous "curled Pe",35  though even in standard versions of Sefer Tagin this embellishment is rare in the first several chapters of Genesis. One might have used the extra instances of אלהים to argue that this fragment is relatively young, as textual variants, especially those related to similar words, tend to build up over time. However, taking into account the relative reticence of this manuscript regarding other entries, the case becomes significantly more murky. This list also includes some rather unusual embellishments not found in other sources, such as Ayin with two Tagin. Whether this irregular version of Sefer Tagin is a result of pure textual variance, regional scribal differences, or blatant error is difficult to determine. In all likelihood, the text was influenced by a mix of these factors. 

Where these fragments fit in the overall framework of the different Sefer Tagin recensions and Otiyyot Meshunnot traditions is difficult to ascertain, especially considering the complex connection between ENA 2753.16 and T-S Misc.24.182, as well as the lack of available background information about NLR Box N.91. I will necessarily leave the task of fitting all the pieces together for a much larger and more general study of Sefer Tagin, to which I have no doubt that these fragments will contribute significantly. 


letter embellishment chart



Bargès, J.J.L. ספר תגין [Liber Coronularum]. Paris, 1866. 

Basser, Y. ספר תגי [Sefer Tagei]. Toronto, 2010. 

Betser, Z. (ed.) מנחת שי על התורה [Minḥat Shai on the Torah: Critical Edition, Introduction and Notes]. World Union of Jewish Studies, 2005. 

Buber, S. יריעות שלמה [Yeriot Shelomo]. Warsaw, 1896. 

Dotan, A. “Masorah.” In Encyclopaedia Judaica, edited by C. Roth & G. Wigoder (ed.). Jerusalem, 1972.

Michaels, M. Sefer Tagin Fragments from the Cairo Genizah. Cambridge Genizah Studies, Volume 12. Brill, 2021.

Razhabi, Y. האותיות המשונות בתורה [Irregular letters in the Torah]. In תורה שלמה [Torah Sheleimah], edited by M.M. Kasher, Volume 29. Jerusalem, 1992. 

Van Voolen, E. et al. (ed.) The Image of the Word: Jewish Tradition in Manuscripts and Printed Books (Catalogue of an Exhibition held at the Jewish Historical Museum, Amsterdam, 14 September – 25 November 1990). Amsterdam, 1990.


1 Much appreciation is due to Dr Eve Krakowski, who pointed this manuscript – then just a peculiar-looking Genizah fragment – in my direction. This discovery speaks just as much to the value of collaboration as it does to the value of browsing. Many thanks to Marc Michaels and Dr Ben Outhwaite for their helpful comments and edits as well.

2 In Hebrew and Aramaic the name can appear four different ways: תגי, תאגי, תגין or תאגין.

3 See the fragment of the month for August of 2020, by Jen Taylor Friedman, ( and a conversation with Marc Michaels on the Genizah Fragments blog, from 1st September, 2021 ( 

4 For an excellent (though a little outdated) collection of different traditions regarding these letters, see Y. Razhabi, האותיות המשונות בתורה, which comprises the second half of volume 29 (titled כתב התורה ואותיותיה) of M.M. Kasher’s Torah Sheleimah.

5 Joshua 4:20, cf. Deuteronomy 27:3 and Nahmanidies there, who mentions Sefer Tagin. For another legendary account about the origin of Sefer Tagin, see JQR Vol. 3, No. 4 (April, 1913), pg. 531. 

6 J.J.L. Bargès, Liber Coronularum (Paris, 1866). Based on Ms. Paris, Bibliothèque Nationale, héb. 837 (likely from the 15th century). Cf. University Library Johann Christian Senckenberg, Frankfurt am Main, Germany Ms. Oct. 143, which is copied from the Paris manuscript.

7 It may be possible to reverse-engineer the text of Sefer Tagin by analyzing ancient Torah scrolls, codices, and fragments which contain these usual letters themselves.

8 Sefer Tagin Fragments from the Cairo Genizah, Cambridge Genizah Studies, Volume 12 (Brill, 2021).

9 As found in the Paris manuscript and the Ibn Gaon Bible Codex (about which see footnote 16).

10 As found in certain manuscripts of the Maḥzor Vitry. See Ms. London, British Library, Add. 27200 – 27201 (Margoliouth 655, 13th cent.), which was the basis for the Hurwitz edition (Berlin, 1893), as well as Ms. Parma, Biblioteca Palatina Cod. 2574 (De Rossi No. 159).

11 For an example of other types of Masoretic information included in Sefer Tagin, see the verso of Kreuzenstein, Sammlung Graf Wilczek, Inv.-Nr. 5667, HDS (from the European Genizah), which illustrates the proper layout for the poetic sections of the Pentateuch: . Another example is T-S Misc.24.182, which includes the words which should be dotted.

12 A complete list is included in the index (pg. 412) to Zvi Betser’s edition of the Minḥat Shai (World Union of Jewish Studies, 2005). 

13 Interestingly, the Masorah itself may rarely note curved or oddly shaped letters (beyond the more typical large and small ones), as in Exodus 3:19, so there is definitely some overlap between the two literatures. 

14 Based on Dr Ben Outhwaite’s evaluation of T-S Misc.24.182, found on page 33 of Marc Michaels’s aforementioned book. 

15 Because both copies include many of the same errors and peculiarities, it is very unlikely that they are completely independent. 

16 Y. Basser, Sefer Tagei (Toronto, 2010). This edition is mainly based on the version of Sefer Tagin found at the end of the Bible codex written by Shem Tov ben Abraham Ibn Gaon (A low-resolution color photo can be found in Marc Michaels’ book, and high-quality images can be found at This manuscript, now owned by Mr. Jacqui E. Safra (JUD. 022), formerly Ms. Sassoon 82, was in the Florsheim Trust collection in 1990. See The Image of the Word: Jewish Tradition in Manuscripts and Printed Books (Catalogue of an Exhibition held at the Jewish Historical Museum, Amsterdam, 14 September – 25 November 1990). While this manuscript was still in Tripoli, a copy of the Sefer Tagin found therein was made for Solomon Buber by a certain Alexandrian Rabbi, Elijah Ḥazzan. Buber then lent this copy to Moses Gaster in London (see Buber’s Yeriot Shelomo, pg. 26). Basser includes comparisons between the Ibn Gaon Bible and the Paris manuscript, as well as several other sources. 

17 Other sources note specifically the second העבריות of 1:16, and not the same word earlier in 1:15. I have assumed our manuscript means the same.

18 This word is missing in the manuscript due to the central tear. However, it is almost certainly תקרב, which most sources include directly before הלם.

19 Besides from large and small letters, which is a different phenomena, and unrelated to Sefer Tagin.

20 Mentioned in Aron Dotan’s comprehensive article “Masorah,” pg. 1410 in Encyclopaedia Judaica, edited by Cecil Roth (Jerusalem, 1972). While this Masoretic note appears in several late editions of the Pentateuch with Masorah Parva, I have not seen any manuscripts containing this note. It is also somewhat mysterious why this specific embellishment, as opposed to any other, is mentioned in the Masorah, though it is likely related to the fact that the Tur (and Rokeaḥ) both mention this unusual letter in their allegorical commentaries to this verse.

21 Deuteronomy 28:68 and 32:37. It is also somewhat possible that this is one of the interpolative Masoretic notes discussed earlier, being a directive to write the Tsadi large, though I don’t know of any such tradition, and it is not included in any of the standard Masoretic lists of majuscule letters.    

22 There is a single character after this word that is likely just a filler, mistake, or test of the pen. 

23 Unless the Tav looking very much like a He is somehow an unusual type of embellishment which I have not previously encountered.  

24 The fragment appears on pages 169 and 170 of the National Library of Israel’s microfilm of NLR Box N.

25 This is not unusual: very often Sefer Tagin appears as part of a Masoretic compilation. An excellent example of this is another Genizah fragment, T-S NS 287.11, which contains the end of Sefer Tagin and the beginning of several other Masoretic lists. 

26 There may be more writing on the edge of the page, but it is impossible to read. 

27 The He is missing an embellishment, but this is likely accidental. I have added the embellishment so that it matches all the other instances of this word. 

28 Unlike the text here, in the Masoretic text this word is spelled doubly defective.

29 A test of the pen or mistake.  

30 Unusually, the He here has four Tagin, unlike all the others which have three, though this might just be an error.

31 The text here is dotted, making it clear that it was accidentally duplicated from the previous line. 

32 Unlike the text here, in the Masoretic text of the Bible this word is spelled defective.

33 This is not very rare: I have seen several manuscripts that do not include this (or any) embellishment for Shin. There is likely an entire family of traditions here. 

34 See the note to verso, line five.

35 The Pe with three Tagin, on the other hand, does make an appearance. 

If you enjoyed this Fragment of the Month, you can find others  here

Contact us: 

The manuscripts in this article are part of the Cairo Genizah Collection. To see more items from this collection visit: