skip to content

The Uncommon Bible: T-S AS 44.35

Ben Outhwaite

I spent the first couple of years of my time as a researcher in the Genizah Research Unit transcribing, revising and preparing for publication the third and fourth volumes of Malcolm Davis’s Hebrew Bible Manuscripts in the Cambridge Genizah Collections, which covered the Additional Series, the largest and most fragmentary part of the Cambridge Genizah. The Bibles that Malcolm had so studiously described there are scrappier than those he found in the T-S Old Series, the subject of his first catalogue, and even the New Series, the subject of his second. Most of them do not derive from lovely, artfully produced model codices, but from more modest productions. They were probably often the work of the ordinary people who owned and used them (‘owner-produced’), rather than commissioned from the finest scribes of Fusṭāṭ.

The palaeographer Colette Sirat’s Hebrew Manuscripts of the Middle Ages (2002, as translated by Nicholas de Lange) divides the biblical codices of Muslim lands into four main categories: (a) Great Bibles, such as Aleppo, Leningrad and the like, encompassing what are termed ‘model codices’; (b) Common Bibles, which were ‘more modest’, usually without masora, but with vowels and cantillation; (c) Bibles with translations; and (d) Bibles with Arabic translation and commentary (such as Saʿadya’s Tafsīr or the commentary of Yefet b. ʿEli). Her classification is fully informed by a familiarity with the biblical manuscripts found in the Cairo Genizah, though, as always with the immense Genizah Collection, it is not difficult to find exceptions that seem not to slot into any of these categories. David Stern’s recent The Jewish Bible: a material history (2017) talks about three ‘distinct generic’ types of Bible in the Middle Ages, ‘the Masoretic Bible, the liturgical Pentateuch, and the study Bible’. There is clear overlap between some of Stern’s and Sirat’s categories, but no great contradiction since one classification is more concerned with form and the other function. Both these works echo, to a certain extent, Moshe Goshen-Gottstein who produced some of the earliest attempts to classify the different kinds of Bible manuscript one finds preserved in the Genizah. In his ‘Biblical Manuscripts in the United States’ (1962, which tackles the Genizah fragments of the Jewish Theological Seminary Collection, as well as the nicer codices preserved in the great American libraries), Goshen-Gottstein distinguishes ‘study codices’ from what he terms ‘listener’s codices’. While the former are differentiated from Great (Masoretic) Codices by an absence of masoretic notes, the latter were intended not for study but for ‘everyday use’ and, he remarks pointedly, ‘not written in order to please future hunters of variant readings’ (Goshen-Gottstein 1962, 38–40). This latter point is telling and explains why not a lot of work has been done on the manuscripts classified as ‘Common Bibles’ or ‘Listener’s Bibles’: they are perceived as holding no useful text critical data.

It is unlikely that these Common Bibles preserve variant readings of great value to the recovery of pre-masoretic biblical traditions. Israel Yeivin, in his Introduction to the Tiberian Masorah (1980, as translated by the E. J. Revell) suggests ‘most are fragments of “vulgar” texts, some with Masorah, without accents, with many extra vowel letters, and so on…’ (Yeivin 1980, 30–31). But this does not diminish their value as cultural artefacts: they form a large body of evidence for ordinary Jewish engagement with the text of the Hebrew Bible in the Middle Ages. It is a Jewish custom (codified by Moses Maimonides in the Mišne Tora, Hilḵot Tǝfillin, Mǝzuza vǝ-Sefer Tora 7:1) that everyone (including kings – on the basis of Deuteronomy 17:18) should, at some point in their lives, copy a Bible, and the congregation of Fusṭāṭ seem to have lived up to it, no matter their competence or resources. Those who could afford it discharged their responsibility through commissioning a Masoretic Bible, as Ibn Yazdād did with Codex Leningrad B19a. Others had to fall back on their own abilities, even when these were apparently quite meagre. The evidence of these ‘handmade’ Bibles is scattered in the Genizah Collection, particularly richly in the T-S Additional Series.

The term Common Bibles is a useful one since it avoids the negative connotations of ‘vulgar’ or even ‘popular’ (both of which I used in the early days!), and it is suitably descriptive of these manuscripts’ most recognisable feature: in the Genizah, such texts are very common. There are loads of them, of all shapes, sizes and qualities. They are overwhelmingly Tiberian Bibles, since Babylonian and Palestinian-vocalised texts are far fewer, and the Tiberian text was the principal form of the Bible used by the congregations of Fusṭāṭ.

Normally the quality of a Tiberian biblical text is assessed by the text’s closeness to the Standard Masoretic Text, as found in Codex Leningrad B19a or the Aleppo Codex. This is not a new idea, and dates back to the Middle Ages (Maimonides, for instance, compared unfavourably the copies of the Bible he saw in circulation in Fusṭāṭ against the exemplary Ben Asher text of the Tāj, the Aleppo Codex). It’s not, however, a useful yardstick to use when examining the Common Bible on its own terms, since inevitably they will fall short, often in quite egregious ways. While some were written by practised hands using reliable texts to copy from, others have no such signs of quality, and indeed probably did not have such pretensions. They will have been used for different purposes: for practising a reading, for learning the Hebrew language, for personal study or devotion, or just to serve as a lap Bible in the synagogue. To this end, some will have only vowel signs, some both vocalisation and cantillation, or others just a partial use of reading signs; gaʿya – definitely a sign for advanced users – rarely occurs to any great extent. Further textual paraphernalia might be present, but are often lacking: paraša or seder markers, kǝtiv and qǝre, correct layout of the parašiyyot etc.

T-S AS 44.35 is a good example of a Common Bible at the less stringent end of the Tiberian spectrum. It’s a small bifolium, made of paper, and contains Lamentations 2:13–18 and 3:51–4:2. It probably came from a pamphlet-type book (termed a daftar in contemporary book lists). Perhaps it contained all the Megillot or Ketuvim, or just Lamentations. It may be connected with the liturgical reading of this book on Tišʿa be-ʾAv. It is unpretentious and, if compared to the Ben Asher ideal, lacking in all kinds of ways: no distinguishing dot for sin and šin, no cantillation, no dageš, no rafe, and certainly no masoretic paratext of any kind. In all likelihood, however, the writer of this book had no intention of producing a model text. It’s not an ignorant attempt at a Bible, and may potentially be ingenious in its approach to common problems.

T-S AS 44.35

The front of T-S AS 44.35

Its particular interest lies in where it diverges from Standard Tiberian, where it can reveal aspects of the writer’s pronunciation of the text. There is much divergence: the use of Tiberian vowel signs is idiosyncratic even by Common Bible standards. A quick glance reveals the qibbuṣ sign written back-to-front and the u vowel as a digraph with qibbuṣ and waw, or even qibbuṣ and šureq. A closer look shows that although it diverges greatly from the standard system of vocalisation, it seems to be following an internal logic of its own.

Detail of T-S AS 44.35, digraphs at Lamentations 2:15

It is likely that this manuscript was not copied from a standard text of the Bible, and may perhaps be the result of writing-down by dictation or memory alone. Evidence of this can be found in the two uses of the Tetragrammaton, at Lamentations 2:18 and 3:58, where the Masoretic Text has אדני on both occasions. The two different written terms for God are of course for piety pronounced identically. Further evidence lies in the ellipsis of quiescent alef, e.g., רושם for ראשם at Lam. 2:15, ולו for ולא at Lam. 2:17, and רושי for ראשי at Lam. 3:54 – to mention just a few. Another hint to the primacy of the oral component is the spelling שב at Lam. 2:14 for שוא, though it has been corrected above the line – identical pronunciation, different spelling. Similarly, מחש{ב}ותם at Lam. 3:60, where the ב has been added above the line suggests the same.

It is where it comes to vocalisation and reading signs that the fragment reveals the greatest differences from the Masoretic Text. As is frequently seen in Common Bibles, the dageš sign is not used. A lack of dageš forte suggests that the nuances of consonantal length did not bother people too much. The avoidance of dageš lene is usually obviated by the habitual presence of rafe, but here rafe is not used. Silent šǝwa is also usually not marked in this manuscript – after all, why trouble to mark a ø vowel when you can just do without a sign at all (even the Aleppo Codex occasionally misses out a silent šǝwa, בְּקִרבּֽוֹ at Job 20:14, for example)? When it falls under one of the bgdkft consonants, however, it is usually marked, and in these cases it must serve primarily not to mark the (absence of a) vowel, but the fricative pronunciation of the consonant, which explains why rafe is not used: it is superfluous when šǝwa plays that role.

Vocalic šǝwa is similarly rarely used. Instead pataḥ is written in its place, e.g., וַאַנַחַמֵיך  at Lam. 2:13, נַבִיאַיִך  at Lam. 2:14, בַנֵי at Lam. 4:2, and in the special case of šǝwa under the first of two identical consonants after a long vowel (such cases exercised various grammatical commentators), עוֹלַלָה at Lam. 3:51 – among dozens of examples. There is nothing intrinsically wrong with this (though a Masorete might hold his head in his hands), since in all these cases the correct Tiberian pronunciation of the vocalic šǝwa is as a short pataḥ. Rather than sticking to the opaque use of šǝwa, the writer of this fragment prefers a more phonetic transcription. This also explains their use of ḥireq when šǝwa occurs before yod, e.g., בִיוֹם  at Lam. 3:57, and the qameṣ in place of šǝwa under lamed in לָהָשִיב at Lam. 2:14. Standard Tiberian pronunciation gives šǝwa before yod the quality of a ḥireq, and šǝwa before a guttural the quality of the vowel under the guttural (unless the šǝwa itself is also under a guttural, in which case it has the quality of the unmarked šǝwa, i.e., pataḥ). All of this can be read in the Masoretic treatises, and can be seen in the Karaite transliterations of the Bible, but it can also be revealed in plain sight in such informal reworkings of the Tiberian system as we find in Common Bibles.

Detail of T-S AS 44.35, pronunciation of šǝwa revealed [lɔhɔːʃiːv]

Ḥaṭef isn’t used either, since the use of full vowel signs, such as pataḥ, for šǝwa, renders it redundant too. The nuances of vocalic length and syllable structure are not reflected in this manuscript, and indeed were unlikely of much consequence to the average user of the Hebrew Bible in the Middle Ages.

There are some pronunciation differences from the usual Tiberian revealed in this text. The vowels sǝgol and ṣere occasionally interchange, as do (rarely) the o and u vowels, suggesting the potential influence of Palestinian pronunciation on the quality of vowels, or the intrusion perhaps of the vernacular (Arabic) with its different vocalic inventory. However, vocalically, the text is far more in accord with the Tiberian standard pronunciation than it deviates from it, even if graphically (or graphemically) it commits all manner of vocalisation sins. The preservation of distinct Tiberian features, lost in later transmission, such as the quality of unmarked šǝwa as a pataḥ, or the various pronunciations of šǝwa before yod or a guttural (which have been levelled in Sefardi pronunciation, for instance), point to a deep familiarity with the Tiberian pronunciation tradition and speak to the value of manuscripts such as this, which rework the Tiberian system in revealing ways.



Goshen-Gottstein, M. (1962), ‘Biblical Manuscripts in the United States’, Textus 2: 28–59.

Outhwaite, B. (forthcoming), ‘The Tiberian Tradition in Common Bibles from the Cairo Genizah’, in G. Khan (ed.), TBA.

Sirat, C. (2002), Hebrew Manuscripts of the Middle Ages.

Stern, D. (2017), The Jewish Bible: a material history.

Yeivin, I. & Revell, E. J. (1980), Introduction to the Tiberian Masorah.


Cite this article

Outhwaite, B. (2019). The Uncommon Bible: T-S AS 44.35. [Genizah Research Unit, Fragment of the Month, January 2019]

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

Contact us:

The zoomable images are produced using Cloud Zoom, a jQueryimage zoom plugin:

Cloud Zoom, Copyright (c) 2010, R Cecco,

Licensed under the MIT License