The early Middle Ages and the beginning of the Hebrew codex

	Ready Reference
	Dead Sea Scrolls The Columbia Electronic Encyclopedia.

According to surviving books and excavated fragments of books in Greek, Latin and Coptic, preserved and found mainly in Egypt due to the dry climate, a new, more practical book form, the codex, was introduced into Mediterranean civilization during the second century of the Common Era. The revolutionary form, in which a number of papyrus or parchment sheets are folded into quires and stitched together at the center to form a series of opening pages, easily carried, stored and used, gradually replaced the old form of the roll. By 300 CE the codex and the roll were employed equally, to judge from surviving handwritten books, which also testify that the new form was first adopted and promoted by the Christians for copying the Bible and Christian literature. By the sixth century the roll was totally rejected by Christians for literary texts, and since then has been used only for documentary and liturgical purposes.

Literary evidence, as well as artifactual discoveries, seems to attest that the codex form was adopted by Jews much later, not before the Arabic period and Geonic literary activity, possibly as late as the eighth century. All references to books in post-Second Temple literature, namely the Mishnah and Tosephta, the Palestinian and the Babylonian Talmud, the halakhic and Palestinian Midrashim, relate to rolls, as noted by Rashi in the eleventh century: "All Sefarim [books] of the times of the sages were in roll form, like our Sefer Torah" (from his commentary to the Babylonian Talmud, Megillah 19a).

	Ready Reference
	Mishna The Columbia Electronic Encyclopedia.

Indeed, there are several references in this literature, from as early as the second half of the second century, to another form of recording information, called by the Greek term pinax, rendered in Hebrew by the word pinkes, multileaved wood or waxed wood tablets fastened together to form a sort of notebook. According to literary references, the Romans developed this device further by applying parchment. A passage in the Mishnah, Kelim 24:7, dating from not later than the middle of the second century, refers apparently to a papyrus pinkes. Although two later midrashic passages refer to folded notebooks, there is no clear evidence to support the assumption that the codex book was utilized by the Jews during talmudic times, nor are there archaeological findings to suggest it. A long period, from which hardly any Hebrew book has survived, separates the rich and diverse findings of books of the Judaean Desert of the late Second Temple period and the earliest surviving dated codex manuscripts, from circa 900. Even if one is justified in ascribing to a time earlier than 900 certain undated handwritten, mostly fragmentary, codices--such as Ms. Vatican Ebr. 66, Sifra with Babylonian vocalization, or the palimpsests found in the Cairo Geniza, in which Hebrew Palestinian texts are written on top of Christian texts, mainly in Christian Palestinian Aramaic and Greek--the long gap of about eight centuries is reduced by probably no more than a century or two.

From the period of that long gap, only a few dozen papyri and a few leather fragments have survived, most of them excavated in Egypt and dating from the Byzantine period. Of these, not a single fragment derives clearly from a codex, though some of the literary texts probably derive from single leaves. On the other hand, existing fragments attest that apart from the ritual Pentateuch scroll (Sefer Torah), which is used and written to this day, rolls were employed for Hebrew books until approximately the middle of the tenth century. A Latin palimpsest, preserved in Munich, Bayerische Staatsbibliothek Clm 6315 and 29022, was written in Northern Italy in the first half of the eighth century on fragments of a Hebrew roll which contained liturgy for Yom Kippur.

Quite a number of fragments of literary rolls were preserved in the Cairo Geniza. However, a major part of them derive not from the regular ancient form of roll--whose sheets are stitched or glued vertically and which is written and read from right to left and rolls from left to right--but from rolls whose sheets are stitched or glued horizontally and which is written and read from top to bottom in one single column and rolls from bottom to top. Such a papyrus roll, which had been used in the ancient world by the Greeks, Romans, and Egyptians for documents only, and was designated by the Latin term transversa charta, was indeed known to talmudic sources both for keeping records and for copying liturgy. To designate such a roll, the Tosephta uses the Greek/Latin term tomos/tomus, which is replaced by the Hebrew term takhrikh in the Mishnah and the Palestinian Talmud. It is interesting to note that the earliest Korans, as well as other Arabic literary texts, were also written on such rolls, and that the Arabic term for a codex, muaf, originally designated such a Koran roll.

NYPL, Dorot Jewish Division Ladino translation of Festival Prayer Book (Oriental cursive hand, Palestine?, 18th century).

NYPL, Dorot Jewish Division Prayer Book (Yemennite square hand, Yemen, 17th century).

NYPL, Dorot Jewish Division Prayer Book (Yemennite semi-cursive hand, Yemen, 17th century).

This Arabic term was borrowed and employed when the codex form was explicitly referred to, apparently for the first time, in Hebrew literature. It appears in the old Hebrew translation (Hilkhot Re’u) and in a later version (Halakhot Gedolot) of the Babylonian-Aramaic Halakhot Pesukot of the late eighth or beginning of the ninth century (the halakhah in which the term is employed is missing in the only existing, incomplete codex and Geniza fragments of Halakhot Pesukot). The Hebrew terms designating a codex, miaf as well as diftar, are found in colophons of the earliest dated Hebrew biblical codices of the tenth century. Like miaf, the term diftar was borrowed from early Islamic Arabic. The additional term derived from the Greek diftera (hide), a term used in talmudic literature to designate a certain sort of leather rejected for writing the Torah scroll.

To sum up: existing Hebrew manuscripts in the form of a codex which contain an explicit indication of their time of production date from circa 900 and later. Some codex manuscripts, mostly fragmentary, can be dated up to about a century or, at most, two centuries earlier. Indeed, literary evidence reflects the later adaptation of the codex, which had been introduced as a book form for Greek and Latin texts as early the second century, and became the usual book form in the fifth century. However, the virtual absence of surviving Hebrew books in any form from late antiquity to the High Middle Ages cannot be attributed only to their destruction by wear and tear or by conquerors and persecutors. One should also consider the possibility that the talmudic and midrashic literature, the so-called Oral Law, was indeed mainly transmitted orally until the Islamic period, as is indicated explicitly in a few talmudic sources, and attested by literary patterns and reciting devices contained in these texts.

NYPL, Dorot Jewish Division Joseph Gikatilla on the transmigration of souls from a compendium of early kabbalistic texts written in Sephardic (semi-cursive hand, North Africa, 16th-17th century).

NYPL, Dorot Jewish Division El mitnase le-khol rosh (Yotser for Shabbat Shekalim) from the Padua Ashkenazi Mahzor festival prayer book (Ashkenazic square hand, Germany, 14th century).

Handwritten books of the High and Late Middle Ages
Hebrew dated (and undated) manuscripts have survived from the tenth century on, and are found in hundreds of collections throughout the world. They represent a rich diversity of technological, scribal, and aesthetic traditions of book production by Jewish scribes and copyists continuing until the invention of moveable type and the spread of Hebrew printing at the end of the Middle Ages. Hebrew handwritten books were produced in many lands and over vast areas, due to the wide distribution of the Jews in the Orient and the Occident. They were made in Christian Europe; in Muslim Spain, North Africa, the Middle East, and West Central Asia; and in Byzantine Greece and Asia Minor.

	Thinking Point
	Hebrew manuscripts were produced all over the world during the Middle Ages. In what ways do you think these books can help us to track the movement of ideas and people throughout Jewish communities during this period? In what ways do you think these texts may fail to capture these travel patterns?

Our knowledge of the making of Hebrew books and the scripts employed in them derives from the study of the existing dated manuscripts, and is naturally confined to the High and Late Middle Ages: from the tenth century in the Orient; the late eleventh century in Italy; the late twelfth century in the Iberian peninsula, France and Germany; and the late thirteenth century in Byzantium and the Maghreb. Therefore, we know the various types of the Hebrew codex and its writings in their established form. Nevertheless, gradual changes in script and some radical shifts in bookmaking techniques can be noticed in most areas.

The variety of scribal practices that medieval Hebrew manuscripts demonstrate can be classified into several geocultural, rather than political, entities, each of which manifests its own combination of techniques and style and shape of script. The following geocultural terms are used to distinguish both specific types of scribal practices and script and the geographical distribution of these practices.

Orient: Egypt, Palestine, Syria, Eastern Asia Minor, Iraq, Iran and its surroundings. At the time of the earliest manuscripts, these territories were incorporated into a single political unit, the Abbasid Caliphate.

Yemen: A subtype within the Orient. Its distinctiveness manifests itself, particularly in the script, even in the earliest dated documents of the twelfth century.

Persia: A subtype within the Orient which includes Iran and Uzbekistan. Its distinctive characteristics emerged only at the beginning of the fourteenth century.

Sepharad: This entity encompasses vast territories, including not only the Iberian peninsula but also the Maghreb and even regions beyond the Pyrenees, Provence and Languedoc, which share the script and technical practices of Spain and North Africa from at least the beginning of the thirteenth century.

The affinity of these four Islamic territories and types is evident mostly in their scripts, which were influenced by Arabic script and calligraphy.

Ashkenaz: Central and northern France, England, and medieval Germany, and in the Late Middle Ages also Central and Eastern Europe. Some slight differences in script and codicological practices can be discerned between France and Germany.

Italy: Distinctive script and scribal techniques characterized Hebrew manuscripts produced in Italy as early as the time of the first dated ones, in the late eleventh century. However, some correlations can be noticed between early Franco-German manuscripts of the late twelfth and early thirteenth century and Italian manuscripts, which hint at the Italian origin of the Ashkenazic type.

Both these types share the background of Western Christian European civilization and are influenced by Latin script.

Byzantium: An independent type, clearly in its script, but also in some scribal practices, shared by manuscripts produced in Western Asia Minor, the Greek islands, Crete, Rhodes, and the Balkans, which constituted the late Byzantine Empire before its decline. This type may have been influenced by Greek script.

NYPL, Dorot Jewish Division Luah nimuse ha-mitkadshim veha-mitaharim copied by Samuel Archivolti (Italian semi-cursive hand, Bologna, Italy, 1546-7).

NYPL, Dorot Jewish Division Sefer ha-Razim compiled by Moses ben Jacob ben Mordecai (Byzantine semi-cursive hand, Greece, 1468).

Writing material
Medieval Hebrew codices were written on two kinds of material--parchment and paper. Only one large fragment of a papyrus codex was preserved in the Cairo Geniza, which can be dated to the ninth or eighth century. Papyrus was undoubtedly the writing material of biblical times in Palestine, but leather was introduced at the beginning of the post-exilic period at the time of the canonization of the Hebrew Bible. Most of the Dead Sea Scrolls were written on gevil, whose hair-side only was treated for writing. Early Hebrew codices in all regions were written on parchment, cattle hides treated so that both the hair-side and the flesh-side were suitable for writing. Rolls were written on only one side, whereas codices were written on both sides of their leaves.

Parchment preceded paper, as papermaking was introduced to the Arabs only in the middle of the eighth century, by Chinese prisoners of war in Samarkand, spread gradually through the Islamic countries, and reached Christian Europe, starting with Italy, only in the last quarter of the thirteenth century. Once introduced and manufactured, paper began to replace the expensive parchment and eventually became the chief writing material in all areas. In the Orient, paper replaced parchment more rapidly, as early as the beginning of the eleventh century. Deluxe copies were produced everywhere on parchment until the end of the Middle Ages and even thereafter.

Several types of parchment employed in medieval Hebrew manuscripts can be visually distinguished. The differences between them relate mostly to the treatment and the appearance of the hair-side. In Italy the parchment always preserves the natural differences between its two sides: the flesh-side is smooth, glossy, and much brighter, while the rough hair-side, although often scraped, retains its grain. In the Sephardic parchment the rough hair-side is generally not scraped, yet the grain pattern is barely visible. The flesh-side is bright.

In Ashkenaz we see a radical shift in the nature of the parchment. Early manuscripts were written on parchment which, like the Italian, preserved the differences between its sides. By the end of the twelfth century, a new technique was introduced, which minimized the differences between the sides. In Germany, this new technique was adopted gradually and evolved finally into a complete equalization of both sides in the mid-thirteenth century: the hair-side is rubbed and all its hair follicles removed; the flesh-side is also scraped; and both sides are very rough. In France also the minimizing technique was adopted, but there the sides remain distinctive, though sometimes not so clearly. Like the late Ashkenazic parchment, the Oriental shows a resemblance between the hair-side and the flesh-side, but while this effect was achieved in Ashkenaz by scraping both sides, it was accomplished in the Orient by smoothing and glossing. However, although the grain is barely visible in Oriental parchment, its sides differ in color, so it is usually possible to distinguish between them.

While parchment may have been produced by the Jews themselves, as some documents hint, and bears particular characteristics in Hebrew manuscripts, paper was manufactured in gentile papermills and was shared by all scribes. Consequently, Hebrew paper manuscripts do not differ from Arabic, Latin, or Greek paper manuscripts. In the Orient they were written on Arab paper and later, in Europe, on Occidental paper. The two kinds of paper, both produced from rags, differ greatly in their morphological patterns. Oriental paper has either no wire-lines at all, or only curly and not easily visible laid-lines; when it has chain-lines, they are grouped in twos, or threes, or twos and threes alternately. Occidental paper has clear and straight laid- and chain-lines, spaced evenly and, from the year 1282 on, watermarks, which were trademarks of the papermills. Just as watermarks in European paper provide a useful tool for dating, some the patterns in Arab paper are typical of certain regions and periods, as Oriental Hebrew paper manuscripts demonstrate.

Quiring
To construct a codex, the scribe or the stationer had to procure a number of sheets of either parchment or paper, of the same height but double the width of the page desired. The sheets were placed on top of each other and then folded down the center. When folded, each sheet would give a sequence of two leaves or four pages. Two sheets laid on top of each other before folding would give four leaves or eight pages, and so on. Later, each set of folded sheets was stitched at the center of its opening, and constituted a quire. A codex was formed by a number of sewn quires. A second set of threads passed horizontally through the first set united the quires. It was taken through them across the spine of the book and secured to the front and back binding covers.

The number of sheets folded to make a quire in Hebrew manuscripts varies from three to fourteen, but is usually uniform within a manuscript. Naturally, quires at the end of manuscripts, or at the end of parts of them, might be larger or smaller. Manuscripts produced in the same geocultural region share the same composition or compositions of quires, reflecting the degree of conformity and the power of tradition in Hebrew bookmaking. In some regions there was a difference between the composition of parchment quires and that of paper quires, the latter manuscripts showing less regularity and a larger number of sheets.

In the Orient the regular number of sheets in a paper or parchment quire was five (ten leaves, twenty pages), but in Persia and Uzbekistan, at least from the fourteenth century, Hebrew manuscripts were constructed of four-sheet quires (eight leaves). In Ashkenaz, regular quires, made of parchment or paper, were always constructed of four folded sheets (eight leaves). In Italy the regular composition of parchment manuscripts was five sheets (ten leaves). Paper manuscripts had no uniform quiring, and show a variety of compositions, but usually five, six, or eight sheets in a quire. In Sepharad the usual number of folded sheets in parchment codices was four (eight leaves), but around 1275, a secondary, much less common composition of six sheets (twelve leaves) was introduced. Paper manuscripts were frequently constructed of six-sheet quires, and less frequently of eight-sheet quires, but other compositions were also sometimes employed. All parchment codices produced in Byzantium have four-sheet quires, but paper quires have no uniform composition, though six sheets are most frequent.

The sheets of parchment quires in Hebrew manuscripts are arranged so that hair-side faces hair-side and flesh-side faces flesh-side. Consequently, each opening of a parchment codex has a uniform appearance, hair-side or flesh-side alternately. Quires usually start with the hair-side, but in Italy, and rarely in Sepharad, quires can start with the flesh-side.

In Sepharad, Italy, and Byzantium, quires were sometimes constructed by combining parchment and paper sheets. This practice was introduced after paper had become the cheaper material in these areas, as a compromise between the cheap but less durable paper and the much more expensive but stronger, more durable parchment. Parchment was used for the outer and innermost sheets of quires to protect the sheets of paper placed between the parchment sheets.

Ruling
The quires having been prepared, the next step in producing a codex was to rule the pages. The scribe could not start copying texts until horizontal writing and vertical boundary lines were drawn to guide the copying, to ensure the uniformity of the copy, and to help the scribe comply with traditional proportions and layout of the written space. Hebrew manuscripts show a variety of ruling designs and techniques, most of them typical of certain areas and periods. In Ashkenaz and Northern Italy one notices clear shifts of ruling practices in the mid-thirteenth and early fifteenth centuries respectively, while in the Orient and Byzantium a single technique was applied until the end of the Middle Ages.

NYPL, Dorot Jewish Division Haftarah for Second Day of Sukkot from the Padua Ashkenazi Mahzor (Germany, 14th century).

NYPL, Dorot Jewish Division Asaperah el hok pelaot (Kerovah for Purim) from the Padua Ashkenazi Mahzor (Germany, 14th century).

To guide the drawing of horizontal and vertical lines, and to ensure the uniformity of the ruling pattern within a codex, rows of pricks or small slots were made by knife, compasses, or other metal instruments down the outer margin of the folded quire (in the Orient, Italy, Byzantium, early Ashkenaz, and late Sepharad), or the outer and inner margins of the folded quire (in late Ashkenaz and early Sepharad). Horizontal lines were then drawn across the width of the unfolded sheet, or the leaf, from prick to prick, as were vertical boundary lines on each side of the written space or column. In paper manuscripts the drawing of the lines was probably guided by ruling-boards.

NYPL, Dorot Jewish Division Lel shimurim or Yisrael (Maaravah for Second Day of Passover) from the Padua Ashkenazi Mahzor (Germany, 14th century).

NYPL, Dorot Jewish Division Ha-melekh ha-yoshev (Shaharit for Rosh ha-Shanah) from the Padua Ashkenazi Mahzor (Germany, 14th century).

The lines in Hebrew manuscripts were ruled either by hard point (in all areas, but in Ashkenaz only in early manuscripts, and in the Orient only in parchment manuscripts), or by lead pencil (in late Ashkenaz), by ink (in late Northern Italy), or by ruling-board (in paper manuscripts in the Orient). While lead pencil and ink were applied to each page and ruling-board to each leaf, ruling by hard point was applied in various ways, depending on the ruling unit and the side of the sheet or leaf to be ruled, since hard point can rule successive leaves (or sheets) at one time. Thus, in early Sepharad, pairs of parchment leaves or sheets (sometimes even four) were ruled at one time on the hair-side. In early Ashkenaz, Italy, and Byzantium, each unfolded parchment sheet was ruled individually by hard point on the hair-side, while in the Orient each unfolded parchment sheet was ruled individually on its flesh-side.

Copying
Once the ruled parchment or paper quires were prepared, the scribe could start copying the texts requested. For writing he naturally needed a writing instrument and ink. He also employed certain specific techniques to ensure the uniformity of the margins of the copied text, and to indicate the proper order of sheets, leaves, and quires before binding.

1. Writing instruments and inks
Medieval Hebrew scribes employed two kinds of writing implements, the reed pen and the quill pen. The two pens differed in flexibility, which strongly affected the nature of the letter strokes and the style of the script. The reed pen, which was made from reed plants, was more rigid and produced more regular strokes. The quill pen, which was made from birds' feathers, was more flexible and produced varying strokes. The reed pen was used during the High Middle Ages by Hebrew scribes who lived within the Islamic territories, i.e., the Orient, the Maghreb, and the Iberian peninsula. The quill pen was employed by Hebrew scribes living in Christian territories, i.e., France, England, Germany, Italy, and probably Byzantium.

Inks were produced as dry sticks, which were mixed with water before use. In general, the main coloring component in the Orient was lampblack, while in the Occident it was always soluble iron salt. This difference resulted in the different colors of the written texts of Hebrew medieval manuscripts. While the Occidental manuscripts exhibit a variety of shades of brown, from dark to light, and sometimes reddish, yellowish or even greenish hues, Oriental manuscripts always exhibit dark written text, either black or dark brown.

2. Script
In each of the Jewish cultural areas, scribes employed a characteristic type of script: Ashkenazic, Italian, Sephardic, Byzantine and Oriental (which included two subtypes: Yemenite and, from the fourteenth century on, Persian). These types are known to us in the form established in the High Middle Ages, but the evolution of the Oriental type can be traced in part, due to the survival of papyri and datable epigraphic writings from the beginning of the Middle Ages and of dated codices and documents from the tenth century. The lack of early dated or datable codices or scripts from the other areas severely hinders our knowledge and understanding of the development of this diversity of types; only from historical information can we infer the factors which contributed to it. Thus, the Ashkenazic script seems to have evolved from the early Italian type, which had probably evolved from the Oriental. The Sephardic script, which was not confined to the Iberian peninsula, but was also employed in the Maghreb, Provence, and Sicily, most probably arose in North Africa.

Within the vast geographical distribution of the Ashkenazic, Sephardic and Oriental types of script, local variations can be seen, despite the strong tendency toward conformism and conservatism. Some stylistic differences can be observed between the Ashkenazic script employed in France and that in Germany. Local variation can be discerned within the Sephardic type between Iberia (Spain and Portugal), Provence, North Africa and, to some extent, Sicily and Southern Italy. Within the Orient, apart from the independent Yemenite script and the later Persian subtype, morphological differences exist between northeastern regions and southwestern ones.

NYPL, Dorot Jewish Division

Collected letters of early Hasidic leaders (Galilee, circa 1800).

These types of medieval Hebrew script should be multiplied by three, as almost each type was employed in three modes of writing: square, semi-cursive, and cursive. Thus, Hebrew medieval manuscripts exhibit a rich and very large variety of handwriting, amounting to some twenty completely different types of script. The difference between the three modes of each type is in the number of strokes needed to produce a letter: more strokes are required to produce square than semi-cursive characters, fewer strokes are needed to write cursive than semi-cursive characters.

The types of Hebrew scripts have not always been confined to their geographical boundaries. The special historical circumstances of medieval Jewry, the expulsion of whole communities by force and the frequent emigrations of individuals by choice, affected the script of many Hebrew manuscripts. These were written not in the local script of the geocultural area in which they were produced but in the native script of the immigrant copyists, who usually retained their native scripts for the rest of their productive lives. Thus, we find in Northern Italy, for instance, late medieval Hebrew manuscripts copied in not only the local Italian type, but also in Ashkenazic and Sephardic types, or, in Palestine, manuscripts copied in not only the local Oriental type, but also in Sephardic, Ashkenazic, and Byzantine types of Hebrew script. The codicological features, such as writing material, quiring, and ruling techniques, of codices produced by immigrant scribes reflect, however, the local tradition.

3. Scribal devices to maintain the left-hand margin
Although the ruled left-hand margin sets the limit of the lines, it does not guarantee that the lines will in fact end in a straight edge. Unlike Latin scribes, Hebrew copyists attempted to achieve a relatively justified end margin and preserve the uniform page layout of the book. Apart from aesthetic motivation and the influence of Arabic calligraphy, the roots of this marginal neatness are probably found in old halakhic rules and practices of writing the ritual Pentateuch scroll. Some devices can be traced back to the scrolls of the Judaean Desert.

NYPL, Dorot Jewish Division Mi-sod hakhamim (Petihah for Neilah) from the Padua Ashkenazi Mahzor (Germany, 14th century).

NYPL, Dorot Jewish Division Haftarah for Shemini Atseret from the Padua Ashkenazi Mahzor (Germany, 14th century).

NYPL, Dorot Jewish Division Af beri (Prayer for rain for Musaf of Shemini Atseret) from the Padua Ashkenazi Mahzor (Germany, 14th century).

To maintain the left-hand margin, Hebrew scribes developed and elaborated a broad range of scribal devices, some common to broad geocultural areas, some unique to one or another region. The Hebrew scribes used three methods: filling out short lines, preventing the margin from being exceeded, and writing protruding words or letters in such a way that the margin boundary was respected.

NYPL, Dorot Jewish Division

Eshkol ivui (Shevata for Shabbat Shekalim) from the Padua Ashkenazi Mahzor (Germany, 14th century).

The scribal devices for filling out short lines included (1) dilating one of the last letters of the last word in the line (the most common device in all areas, most effective with a square script, and particularly used on letters containing a long horizontal upper bar); (2) leaving space before the last word or, particularly in Sepharad, before the last letter of the last word; (3) inserting various graphic signs in the space at the end of the line when that space was too short to accommodate a complete word without protruding considerably into the margin (in the Orient and Sepharad such graphic fillers might consist of certain letters or parts of letters); and (4) filling in the space with as many letters of the next word as could be inserted and then writing the complete word, repeating those letters, at the beginning of the following line (Ashkenazic copyists tend to omit one or more of the strokes of the last letter of the anticipated word).

Devices for preventing the margin from being exceeded included (1) compressing letters of the last word (in the Orient and particularly in Sepharad, scribes might change their semi-cursive script into a cursive one) and (2) dividing final words so that the beginning was written at the end of the line and the rest at the beginning of the following line (this device was practiced only by Yemenite and Italian scribes). Devices for writing exceeding letters or words in such a way that the left margin was retained were (1) writing as much of the last word as possible within the available space, leaving a blank, and then completing the word to the left, in the border of the leaf; (2) writing up to the edge of the text block and then placing the excess letters above the word, without violating the margin; (3) writing words likely to exceed the margin diagonally, usually downward (the favorite practice of Oriental copyists, who made extensive use of it, following Arabic scribes); and (4) writing exceeding letters vertically upward (practiced in Ashkenaz).

4. Means of ensuring the correct order of the codex
Since scribes were copying texts on loose folded sheets arranged in quires, and only upon completion of the copy would they give the loose quires to the binder for stitching together, it was necessary to ensure the correct order of the quires at the time of binding. In later times, particularly when writing on paper, which was more vulnerable than parchment, scribes tended to employ additional means to ensure the order of the sheets or leaves within a quire.

These means included numerating the quires by Hebrew letters either at the beginning or at the end of the quire, or at both the beginning and the end. In Oriental manuscripts, additional numerating in Arabic words may occur at the beginning of quires. This numbering of quires ("signatures") was not practiced by Ashkenazic scribes. By the end of the Middle Ages copyists numerated all the leaves of the codex.

NYPL, Dorot Jewish Division

Elohim al domi lakh ke-kol (Kerovah for Shabbat Zakhor) from the Padua Ashkenazi Mahzor (Germany, 14th century).

Another way to ensure the correct order of quires, sheets, or leaves was by the use of catchwords. The most common method was to repeat the first word or words of the quire, or the leaf, at the foot of the preceding page. An alternative method was to repeat the last word of a quire, or a leaf, at the beginning of the following page. In the first method, catchwords are placed separately from the copied text (in the Orient they were frequently written diagonally, in Ashkenaz sometimes vertically), are frequently decorated simply, and are sometimes illustrated (in Ashkenaz and Byzantium). Sometimes, particularly in Sepharad, catchwords of both kinds were employed in the first half of a quire and at the end of a quire, as this was sufficient to ensure the order of the sheets within the quire.

After the text had been copied and the correct order of the loose quires assured, biblical and liturgical manuscripts, and sometimes other texts, had to be vocalized; this was usually done by a professional vocalizer, rather than by the copyist. To biblical texts the Masorah was usually added in the margins, generally by the vocalizer. Deluxe manuscripts were handed over to artists, who added decorations, illuminations, or illustrations. Only after the vocalizers and artists had done their work was the manuscript handed over to the binder, who completed the process of producing a codex.