Multiple Hands in the Marginal Annotations of the Hebrew Bible Codex Madrid M1 (Biblioteca Historica Marques de Valdecilla, BH MSS1)

Abstract

This article presents a paleographical analysis of the marginal annotations that appear next to the biblical text in the Hebrew Bible codex known as Madrid M1. In order to do so, a new methodology has been developed. The ensuing analysis has identified a great number of annotations, largely located in the outer and intercolumnar margins, written by multiple hands. These later interventions have different aims and produce diverse results. The systematic additions associated with different Masoretic textual phenomena reveal an interest in organizing the information differently and offering the information on each phenomenon consistently by adding cases that are lacking. The non-systematic additions seem to be more the result or reflection of later revisions, readings, and uses of the manuscript. The annotations of both types show a clear intention to complete and expand the already abundant information found in the margins of Madrid M1, rather than an effort to correct it.

The annotations that appearnext to the text of the Hebrew Bible in the margins of most of the extant medieval Hebrew Bible manuscripts, referred to as Masora, have been almost exclusively studied for their textual content.¹ The assumption that one single Masora was transmitted orally until it was written down in lists and later added in the margins of the manuscripts in a single act of copying has deterred scholars from investigating and exploring the copious evidence of later additions made by various hands that can be seen with the naked eye in the annotations contained in most medieval Hebrew Bible manuscripts.² The few scholars who have recognized and worked with these annotations have focused on their content, on corroborating that they were correct from a textual point of view, and on checking the agreement between the annotations and the biblical text to which they are connected.³

This is the common way to approach any Masoretic annotation: whether added later or not, only the textual information is important and any other elements are ignored. In a previous study on the methodological approaches to the study of the Masora, I suggested placing the Masora itself at the center of the inquiry rather than studying it as a secondary element dependent on the biblical text.⁴ When this approach is adopted, the existence of different scribal hands becomes one of the important elements of the initial analysis of the Masora in a single manuscript.

This essay isolates and distinguishes the different scribal hands in the Masora of codex BH MSS1 of the Biblioteca Histórica Marqués de Valdecilla, University Complutense of Madrid (hereafter M1), in order to identify the marginal annotations written by those hands throughout the entire codex and to study not only their content but also the way they are formulated, their format, their placement, and their relationship with the biblical text. It also addresses the purpose behind the annotations added by different hands in the context of this codex. I have chosen this codex because I have partially edited its Masora and studied some aspects of it. Moreover, the existence of marginal annotations written by hands other than that of the principal scribe in the Pentateuch has been indicated in the ongoing edition of the Masora of this manuscript.⁵

Manuscript Description

M1 contains the Hebrew Bible in its entirety (apart from six missing folios ranging from Exodus 9:33 to 24:7).⁶ It consists of 340 unpaged, large-format folios (31.2 × 27 cm). It is written on high-quality parchment. The ruling—with horizontal lines and vertical boundary lines—and the double-margin pricking are visible; the space for the text and the Masora is marked.⁷

The biblical text is arranged in three columns of thirty-two lines each (fig. 1), except for the poetical portions of the Pentateuch (Exod. 15:1–19; Deut. 32:1–43), Judges (5:1–31; fig. 2), and Samuel, both of which are written in specially prescribed lines, as well as the poetical books (Psalms, Job, and Proverbs), which are distinguished by a hemistichal division.⁸

The order of the biblical books is that prescribed in the Babylonian Talmud.⁹ The text is written in Sephardic square script with Hebrew Tiberian punctuation. Most of the pericopes of the annual cycle (parashiyyot) into which the Pentateuch is divided, and the portions for the weekly synagogue readings according to the Palestinian triennial (sedarim) are respectively indicated in the margin by the abbreviation of the word parashah ( ) or the letter peh (פ), and the letter samech (ס). The “open” and “closed” portions are marked by leaving empty spaces. A summary with the total number of the verses, letters, and so on appears in a decorated rectangular frame at the end of each biblical book, except for the books of Isaiah, Jonah, Nahum, Habakkuk, Ruth, Job, Lamentations, and Nehemiah—which have been cut off and are missing—and the book of 1 Chronicles, as the two books of Chronicles are treated as a single book.¹⁰

Figure 1.

BH MSS 1, fol. 25v.

Figure 2.

BH MSS 1, fol. 97v.

The Masora is written in the upper, lower, and outer margins, and in the interstice between the columns of each folio. The annotations are mostly written in straight lines, with a large number written in figured patterns.¹¹ The annotations in figured patterns can be found in any of the margins of the folio, but they are especially common in the upper margin and the lower outer margin. The patterns of the figured Masora are mainly geometric and vegetal. There are thirty-seven exceptional patterns found in the outer margins. The most frequent motifs are of the so-called candelabra tree, but the patterns’ designs differ significantly from that of the most well-known candelabra in the Catalan Bibles and have no parallels in any other Toledan Bibles.¹² Their designs, elaboration, and complexity contrast with the simplicity of the rest of the designs (see the outer margins of figs. 1 and 2).

Besides the Masora given in the margins, a number of lengthy Masoretic rubrics are found at the end of the main divisions (Pentateuch, Prophets, and Writings).¹³ They are written in three columns of thirty-two lines each in a larger script, although not as large as that of the biblical text. Although this manuscript does not contain a colophon, it is assumed that the text, the vocalization, and the Masora were copied by a single person.¹⁴

The manuscript is dated to 1280, according to the note of purchase found on folio 334v, where it is stated that R. Yishaq and R. Abraham, both physicians, bought the manuscript in the year five thousand and forty of the creation of the world in Toledo.¹⁵ The manuscript also contains the Arabic numbering of chapters, and the Latin names of the books written in red ink. They were added by the convert Alfonso de Zamora (ca. 1474‒ca. 1545), to whose personal collection the manuscript belonged at a later stage. It has a sixteenth-century Complutense cloth binding and bears the old golden coat of arms of the Complutense University of Madrid. In the nineteenth century, the manuscript was transferred to the university’s Biblioteca Histórica Marqués de Valdecilla.

The importance and high quality of this manuscript are unquestionable. It is believed to have been used extensively for the composition of the Hebrew column of the Complutensian Polyglot Bible edited by Ximenez de Cisneros in 1514.¹⁶ A note written in Spanish in 1756 comments on the manuscript’s exceptional nature and the value of its marginal annotations.¹⁷ Finally, it is one of the manuscripts being consulted for the current critical edition of the Hebrew Bible, Biblia Hebraica Quinta.¹⁸

Analysis

Corpus

The Masora of M1 contains some sixty thousand marginal annotations. These annotations have features that differentiate them from the Hebrew biblical text, with which they share the page. While most of the characteristics are similar to those found in the Masora of other Hebrew biblical manuscripts, some are specific to this manuscript.

The marginal annotations in M1 are written in smaller-size letters than the biblical text. They are arranged around the biblical text in the upper, lower, inner, and outer margins, and in the inter-columnar spaces. The annotations located in the interspersed spaces within the columns and the outer and inner margins—denominated “Masora Parva” (MP)—are written in the middle of those spaces, usually starting on the same line as the lemma, with one word under another and largely horizontally.¹⁹ The annotations placed in the top and bottom margins—denominated “Masora Magna” (MM)—are written in three lines in the upper margin and four lines in the lower margin of each folio, very often combined with figured patterns placed under the fourth line and in the lower outer margins.

Each marginal annotation is linked to one or more words of the biblical text—the lemma—written on the same folio. The MP annotations are connected to their lemmas through their placement next to the line of the text, and the MM annotations through the repetition of the lemma in the annotation itself. Usually, a reference sign—a small circle that scholars today call the circellus—points the reader to the lemma to which the annotation should be attached. The circellus is often placed above the lemma when it consists of just one word, or between the words when the lemma consists of two or more words.

As a general rule, the information contained in the marginal annotations is expressed in a concise way, using abbreviations, and on many occasions, part of the information remains implicit. The MP annotations are the briefest, with the words often represented only by their initial letters or only by one letter (usually a Hebrew character with a dot above indicating the numerical value of letters). The MM annotations provide information in a more detailed and expanded way without giving up on conciseness and the use of abbreviations. They give the references to the biblical verses through catchwords (simanim), one or two words from the verse that is being referred to. However, the marginal annotations in M1 exhibit a tendency to be more expansive. This tendency is especially prominent in the MP annotations, which often give more than a numeral. Some MP annotations indicate that there is a corresponding MM annotation listing the verses involved, using the formula “their catchwords are written” in its abbreviated form (̇ ) Others give some more information and also the catchwords in the same way as the MM annotations. The consequences of this tendency in the MM annotations are catchwords formed by more than two words and a less frequent use of abbreviations. Finally, the expansive tendency is also seen in the high quantity of annotations included in M1, much higher than in the main Tiberian manuscripts attributed or related to the tenth-century grammarian Aron Ben Asher.²⁰ At times, there are so many annotations that there is not enough space for them on the corresponding page, so they are continued on the following page.

Methodology

Marginal texts are a problematic area of research in all manuscript traditions due to their idiosyncratic characteristics, and distinguishing scribal hands is an arduous task. The difficulty intrinsic to isolating hands in Hebrew manuscripts as well as the problems and limits of the use of script analysis have been acknowledged.²¹ Script analysis can be particularly difficult when the copyists employed the stereotypical scripts of their region without any visible idiosyncratic handwriting, as is the case with the thirteenth-century Sephardic Hebrew Bible manuscripts from the Iberian Peninsula. Moreover, the script analysis can be misleading, because the variations in the forms of writing do not necessarily imply the existence of different scribes. A scribe’s handwriting can vary for a number of reasons, including the circumstances, the speed of writing, the use of pens with differently cut nibs, lengthy breaks interrupting the copying process, and so forth.

To tackle these problems, Malachi Beit-Arié proposed taking into account other components that are not related to the script itself but that are considered to be the expression of the individual identity of a scribe, such as substitutes for the Tetragrammaton, graphic markers, the diverse stratagems for aligning the lines, the means for maintaining the sequence of the codex, their placement, and decorations.²² The analysis of these idiosyncratic auxiliary components of a single copyist in multi-handed Hebrew manuscripts demonstrates that although the writing style varies, the idiosyncrasy does not.

The difficulty in identifying and distinguishing hands increases when the focus is on the marginal annotations. In addition to the above-mentioned problems, the specific characteristics of the marginal annotations bring their own complications. The non-consecutive writing in the narrow margins and the shortness of most of the annotations, make it impossible to have the full range of distinctive features that can be identified in regular writing within the boundaries of the written area. All these difficulties, together with the focus on the content of the marginal annotations—making the identification of the hands involved in the act of copying irrelevant—explain the scarcity of attempts to isolate the hands in the Masora and, consequently, the lack of a clear methodology. The only extant work that distinguishes the hands in the Masora bases its analysis on the study of the following characteristics: ink color, size of the letters, forms for the sign of the Tetragrammaton, line shaping, and a comparison of the script of specific letters.²³ However, these elements are not sufficient to evaluate all the specific features of the marginal annotations that constitute the Masora.

Aware of the above-mentioned problems, and on the basis of Hebrew codicological studies and previous attempts to isolate the hands in the Masora, I propose a broader analysis that incorporates additional elements to characterize the handwriting and distinguish between the scribal hands in the Masora. In addition to the script analysis, I have incorporated the study of the specific components of the marginal annotations that may be the expression of an individual scribe and that are different from those used for the principal text: the graphic markers that indicate the abbreviation of words and technical terms, and the graphic markers and signs added to facilitate comprehension (separating annotations and catchwords, marking letters with numerical value, etc.) and used to highlight annotations. I have also taken into account elements related to the content of the annotations that can vary depending on each Masorete, such as how the same phenomenon is annotated, which technical term is employed, how the term is abbreviated, which formula is used to introduce information, and so forth. Finally, I have also incorporated the page layout of the manuscript into the analysis, given that the Masora has an assigned space at the moment the manuscript is produced, as indicated by the ruling for copying the MM annotations. The place occupied by each annotation may provide a clue to distinguish later hands.

Therefore, the approach taken to characterize and distinguish the hands in the marginal annotations of manuscript M1 entails examining the following elements: (1) the script, including ink color, size of the letters, and a comparison of the shape of specific letters (lamed, bet, mem, samek, heh, shin, ayin, qof, peh); (2) how information is recorded, including abbreviations, graphic markers, introductory formulae, technical terms, and so on; (3) the placement of the annotations within the folio; and (4) whether the lemma has a circellus over it and, if so, its appearance (size and ink color).²⁴ As a result of studying these characteristics, I have distinguished different scribal hands and characterized several hands in the corpus of the annotations of M1, which are described below. I have labeled each identified hand consecutively as hand1, hand2, and so on. Apart from the main scribe, which is labeled as hand1, the number assigned is by order of appearance, without any correlation to the role of the scribe. I mention—but do not label—the handwriting that appears only sporadically in the manuscript, without enough occurrences to be properly characterized.

Scribal Hands

Scribe 1 (Hand1)

As noted above, most of the marginal annotations were written by the principal scribe who was in charge of the biblical text. They are written in a uniform hand, in tiny calligraphic Sephardic square script, and in a dark brown ink (henceforth: hand1). The script is perfect and balanced. The MP annotations are written in a size smaller than the MM annotations. The shapes of the pairs of letters, which can be similar in some handwritings (beth/kaph, gimel/nun, daleth/resh, he/ḥeth, final mem/samek), are well differentiated.

Hand1’s specific features include the following (see table 1): the “middle” short stroke of the letter shin is attached to the upper part of the left arm; a very long ascender of the lamed with a dot on its upper part; the descenders of the final letters nun, peh, kaph, and tsadi are long; the left downstroke of the qof starts at the roof with a narrow neck and goes straight; the right stroke of the ayin goes obliquely beyond the baseline; and the right-hand downstroke of the gimel is straight and almost perpendicular to the baseline, going below the meeting point.

Table 1.

Unique features of scribal hand¹

Hand1 employs a two horizontal dots–circle–two horizontal dots symbol to separate the distinct annotations in the MM, and a dot to separate the catchwords given in each annotation. Both symbols are also used in the long MP annotations, those similar in content and form to the annotations placed in the MM. At times a colon-circle-colon symbol is used to delimit the ends of the paragraphs of the MM. The substitution for the Tetragrammaton consists of two horizontally adjacent yods, with a central supra-linear dot. The most common way to mark an abbreviation is a dot above the last letter. Less frequently, three dots are used, especially to abbreviate the word sefer. If the last letter is a shin or an alef, they are usually truncated and have a dot above them. The numerical value of the letters is marked with a dot above them. If the numeral is composed of two letters, each one has a dot above.

Two graphic signs are used to highlight some kinds of annotations or phenomena and draw the reader’s attention: a four-dot symbol and a drawing with three arms or any of its variations. The MP annotations are aligned one underneath the other in the middle of the columns’ interstice. The MM annotations are written downward from the upper guide line. The lines are regularly spaced.

Scribe 2 (Hand2)

A different hand is identified in the annotations giving the differences between Ben Asher and Ben Naftali (BA-BN).²⁵ These annotations are written in Sephardic square script using very light brown ink (henceforth: hand2; fig. 3).²⁶

Figure 3.

A comparison of hand1 and hand2 on fol. 39v.

The letters are slightly larger and wider than those of hand1. Hand2 has some specific features (see table 2): the middle stroke of the letter shin is attached to the meeting point of the left and right strokes in the lower left part. The right stroke of the letter tav reaches the base line, but the foot of the left stroke descends below it. The base of the letter peh descends to the left below the base line. The long ascender of the letter lamed has a dot at its end. The right stroke of the final kaph leans to the right.

The lemmas are not usually marked with a circellus unless they have another marginal annotation attached to them, usually from hand1.²⁷ Apart from some exceptions (fols. 3r–13r), the annotations give the two readings, usually abbreviating the names of the two Masoretes ( ) followed by the annotations in the outer and intercolumnar margins (in the middle of those spaces), and they are very often written vertically.

Table 2.

Unique features of scribal hand2

Figure 4.

Fol. 187r.

Scribe 3 (Hand3)

I have identified a third hand in the annotations given the differences between Eastern (Babylonian) and Western (Palestinian) Masoretes located mainly in the Prophets (see table 3).²⁸ The annotations are written in light ink and Sephardic square script with a tendency to slant the letters (henceforth: hand3). This tendency can be observed especially with the letter lamed when it is written between the lines (fig. 4). The letter ayin has a long right stroke going through the baseline obliquely with a curved end. The base of the letters nun and peh and the left foot of the letter tav also go through the baseline. Their long, curved bases are under the letter that follows them.

The lemma is not marked with a circellus, apart from some cases (fols. 122v, 126v) or unless it has another marginal annotation attached to it from hand1 (fols. 125v, 150v, 157r). Apart from some exceptions giving the Western reading or both readings, the general pattern is to give the Eastern reading, usually with the full technical term and less often with the abbreviated form ( ).²⁹ The annotations appear out of alignment with the rest of the MP annotations, and they are very often written vertically.

Scribe 4 (Hand4)

I have identified a fourth hand in most of the Eastern-Western annotations added in the Writings (see table 4). They are written in a regular Sephardic square script in light brown ink (henceforth: hand4).³⁰ The letter ayin has a short right stroke that goes through the baseline just a little bit. The long ascender of the letter lamed leans to the left and has a large dot almost at the end.

Table 3.

Unique features of scribal hand3

Figure 5.

An example of hand4 on fol. 266r.

The lemma is not marked with a circellus, apart from some cases (fols. 266r, 267r, 282v, 284r), or unless it has attached another marginal annotation from hand1. The general pattern is to give the two readings in the annotation using the abbreviations (Eastern) and (Western), respectively (fig. 5).

The annotations appear out of alignment with the rest of the MP annotations, and they are very often written vertically.

Table 4.

Unique features of scribal hand4

Scribe 5 (Hand5)

I have identified a different hand in most of the Ketib-Qere (K-Q) annotations located in the first books of the Prophets, mainly from folios 85v–152r (see table 5).³¹ They are written in a Sephardic square script with letters larger than that of hand1, thick strokes, using very dark ink (henceforth: hand5). The descender of the letter qof for the abbreviation of the term qere is leaning slightly to the right at its end with a thinner part at its top that often does not reach the top horizontal line. The circellus in the biblical text is also larger and done using very dark ink.

Figure 6.

A comparison of hand1 and hand5 on fol. 131v.

The lemma is always marked with a circellus. Very often the lemma has another marginal annotation attached to it written by hand1. The two annotations can appear side by side or with each one placed on opposite sides of the column. When each annotation is placed on one side of the column, the lemma in the text usually has two circelli (fig. 6), each one added by a different hand (as demonstrated by the differences in size and ink color).³²

Table 5.

Unique features of scribal hand5

The annotations usually follow the alignment of the rest of the MP annotations. Some of them are highlighted by a drawing with three arms.

Scribe 6 (Hand6)

Another different hand can be identified in most of the K-Q annotations located from folios 152v–167v and from folio 243r onwards (see table 6).³³ These annotations are written in Sephardic square script with letters even larger than hand5, thick strokes and in light ink (henceforth: hand6). The foot of the letter qof goes down straight and usually reaches the top horizontal line.

Table 6.

Unique features of scribal hand6

The lemma is marked with a circellus, which is also large, thick, and written in light ink. When it has another annotation attached to it written by hand1, only one circellus is found above the lemma in most cases.³⁴ Some of the annotations from this hand give not only the qere reading, but also information about how the word is written in the biblical text. They usually refer to the presence of a superfluous letter in one word, usually waw or yod (fig. 7).³⁵

The annotations are usually aligned. Some of them are highlighted by a drawing with three arms.

Figure 7.

A comparison of hand1 and hand6 on fol. 167r.

Scribe 7 (Hand7)

A new hand is identified in the K-Q annotations starting on folio 167v (see table 7). The annotations are written in Sephardic square script using light ink. The letters are regular and written parallel to the script line, in a smaller size than those of hand5 and hand6, but slightly larger than those of hand1 (henceforth: hand7). It is characterized by a pointed qof with a descender with a “neck” in a swan shape that does not reach the top horizontal line.

The lemma is marked with a circellus, which is small and done in light ink. The annotations are usually aligned. Some of them are highlighted by a drawing with three arms.

Table 7.

Unique features of scribal hand7

Scribe 8 (Hand8)

A different hand is identified in a set of the K-Q annotations located in the Prophets. They are written in irregular Sephardic square script and letters in a smaller size than those of the hand1 (henceforth: hand8; fig. 8).³⁶ The letter qof has a small body and a long descender that do not touch each other.

The lemma is not marked with a circellus, apart from two cases, or unless it has another marginal annotation attached to it from hand1. In the latter cases, the circellus is from hand1. The annotations are usually out of alignment.

Figure 8.

Fol. 198r.

Scribe 9 (Hand9)

One single hand can be identified in the seder markers located from folio 290v until the end of the manuscript (henceforth, hand9; see table 8). The letter samekh has a short right downstroke slanting downwards to the right that joins the base (which leans to the left), forming an oblique angle. A dot is placed above it.

Table 8.

Unique features of scribal hand9

The lemmas are not marked with a circellus. They are placed out of alignment. Apart from four cases, none of them has any kind of decoration.³⁷

Scribe 10 (Hand10)

One set of annotations in the Pentateuch (henceforth: hand10) written in informal square script has the following specific features (see table 9):³⁸ the right stroke of the letter heh is shorter than the left stroke and does not reach the baseline; the letter aleph is bigger than the other letters in the word, and its arms do not touch the body. The lemmas are mostly marked with a circellus. The annotations are placed out of alignment. They are highlighted by a waving stroke over the first word.

Table 9.

Unique features of scribal hand10

Scribe 11 (Hand11)

It is possible to identify another hand in another set of annotations, also in the Pentateuch, written in informal square script and a darker ink (henceforth: hand11).³⁹ It has a characteristic lamed with a small body, a long ascender undulating at the beginning and with a serif in the form of a flag at the end (see table 10). The lemmas are mostly marked with an irregular circellus. The annotations are placed out of alignment.

Table 10.

Unique features of scribal hand11

Scribe 12 (Hand12)

A different hand can be identified in a set of annotations in the Pentateuch that all have the letter lamed with a long right ascender that ends in a flag (henceforth: hand12; see table 11).⁴⁰ The lemmas are mostly marked with an irregular circellus. The annotations are placed out of alignment.

Table 11.

Unique features of scribal hand12

Sporadic Scribes

I have also identified more than one hundred annotations written by other different hands. The scarcity of occurrences of each of those hands, and the shortness of most of the annotations—some of which contain only one letter—complicate the analysis, making it impossible to properly characterize them and, consequently, attribute them to one particular scribe. However, it is possible to make some general observations.

The annotations are mainly written in informal and irregular Sephardic square scripts, with letters slightly larger than those of hand1, usually in light ink. They are placed without alignment in the margins, and the circellus over the lemma in the biblical text is bigger and in light ink. Moreover, there is one annotation written in Sephardic semicursive script (fol. 46r) and three annotations written in Sephardic cursive script (fols. 47v, 240r, and 265v). All of them are written by different hands, mainly in dark ink, out of alignment, and vertically.

Amendments/Corrections

A number of MP annotations have been corrected.⁴¹ Apart from the annotation on folio 165v—where a lamed is marked as deleted by two small lines crossing over it, and a new letter has been written at its side—the emendations consist of substituting the Hebrew characters indicating a number by rewriting over them (see table 12). The most common substitution is the letter bet to lamed and vice versa, but some other letters have also been substituted. When the substitution is from lamed to bet, the ascender of the lamed is canceled by drawing two or three small lines above it. Given the nature of these emendations, it is not possible to distinguish if they are the work of hand1 or of later hands.

Table 12.

Varieties of letterform substitutions and amendments

More numerous are the vestiges of proofreading in the MM annotations. Most of this editorial activity is the work of hand1. It primarily consists of inserting omitted words or letters by one of the following three techniques. The most common form is to write the omitted word or words in smaller script above the line, starting at the initial position where they should be inserted (fig. 9). A second, less common technique consists of writing the omitted words vertically in the insert location (fig. 10).

The third technique adds the omitted words in the margins (see table 13). Within the annotation, the location of the insert point is indicated by a reference sign. The most common is an overhead circle, but other reference signs are used sporadically, such as two dots over a circle, three vertical dots in a line, and two dots over a v. The omitted words are written in the margins, very often vertically, in the same size as the annotations, and they sometimes have a similar reference sign above (fig. 11). This technique is usually employed when there are more than two omitted words.

Hand1 also uses an overhead circle to indicate the cancellation of words. On folio 105v, the scribe has canceled one unfinished annotation by marking the first, middle, and last words with a circle over them. The fact that a new annotation is written after the cancellation shows that it was made in the process of writing the manuscript.

Figure 9.

Fol. 47r.

Figure 10.

Fol. 180v.

Figure 11.

Fol. 31v.

Table 13.

Insertion symbols, hand 1

A different hand is also identified inserting omitted words in the MM annotations. This hand marks the insert location in the annotation using a three-dot triangle sign and writes the omitted words in a smaller script in the margins, also marking most of them with a similar overhead reference sign. This hand also makes two corrections, by first canceling what is written and then providing a substitute in the margins. On folio 47r, the scribe cancels an entire siman by marking the first and last words with a three-dot triangle overhead. The alternative siman is written in the margin without any reference sign. On folio 25r, one word is canceled by marking it with a three-dot triangle, and the substitute word is given in the margins with a similar sign above it.

Some other reference signs are also used to mark the insertion of omitted words, but there are not enough elements to attribute them to one specific hand. Of particular interest is the use of a tailed circle.⁴² Several shapes, sizes, and inks can be distinguished (see table 14). The tailed circle located in the body of the text has its tail pointed toward the margins, and when one is placed above the corresponding marginal annotation, its tail points back toward the text. This reference sign is also used at times to mark the intended location in the biblical text of the omitted words that have been added in the margins.⁴³

Table 14.

Insertion symbols by several scribes

Figure 12.

Fol. 153v.

Figure 13.

Fol. 180v.

Finally, some other annotations have been corrected by crossing out the words and writing the substitute word over it (fig. 12) or by directly rewriting over them (fig. 13).⁴⁴

Content and Type of the Later Marginal Annotations

The content of the identified annotations in correlation with the different hands when it is possible can be described as follows. The annotations on the differences between Ben Asher and Ben Naftali are the work of hand2, who adds them systematically in the outer margins and between the columns of Pentateuch and Prophets.⁴⁵ The annotations giving the differences between Eastern and Western Masoretes start to be added systematically beginning with the Prophets (fol. 87v) and follow in the Writings mainly by two different hands. The annotations located in the Prophets are the work of hand3, and most of those added in the Writings are the work of hand4. There are also a few annotations written in different hands.⁴⁶

Annotations on the K-Q phenomenon are found throughout the manuscript. But while the annotations in the Pentateuch are mainly the work of hand1, from folio 85v, the beginning of the Prophets, and until the end of the Writings, the annotations from hand1 become less frequent, and several different hands are also identified as responsible for the K-Q annotations. Most of the annotations in the Prophets and the Writings are the work of hand5 and hand6. Very often, they add one annotation to lemmas that already have another annotation from hand1 also dealing with the K-Q phenomenon, but articulating the information around the group of words that share a common feature, as in the list headings.⁴⁷ In fact, in most of these cases, there is also one annotation in the upper or lower margin with the same heading that gives the references to the biblical verses through catchwords. As a result, three annotations on the same phenomenon are attached to a single lemma. Less often they give information about how the word is written. Finally, other hands are responsible for one or two isolated K-Q annotations.⁴⁸ Many of the K-Q annotations are highlighted by a similar graphic marker with three arms placed over them. But this marker does not seem to be specific to any of the different hands, although it is more common in some of them, for example in hand6, while hand8 does not use it at all.

Figure 14.

Fol. 213v.

Some of these later K-Q annotations seem to lead to a correction of the biblical text.⁴⁹ Especially notorious is the case of hand8, who corrected most of the lemmas he annotated. In all these cases, the word in the biblical text was written according to the qere form by hand1. The additions and erasures in the biblical text were made by the later hands to adapt the words to the ketib form (fig. 14). Most of these cases involve words containing waws or yods.⁵⁰

The annotations noting the presence of superfluous (yatir) letters—usually the letter yod—in the books of the Prophets and the Writings, forty-one in total, are mainly the work of hand5 and hand6.⁵¹ The rest are written by hand1 (e.g., fol. 150r) and other different hands.⁵² The same occurs with the annotations noting the middle of one biblical book, the Torah and the Bible, twenty-two in total. Apart from four annotations (fols. 42r, 73v, 90v, 283r) that are the work of hand1, the annotations in the Prophets and the Writings are mainly the work of hand5 (fols. 100v, 194r, 285v) and hand6 (fols. 163r, 231v, 243r, 278r). Several other hands can be identified as well.⁵³ Most of the annotations simply note “the midpoint of the book.” However, the information in the two annotations found on folios 39v and 40r is extensive, explicitly noting that they are the midpoints of the letters and words of the Torah and adding extra information. This information is similar to that provided in the Talmud to explain those midpoints.⁵⁴ Apart from one case, these annotations are primarily highlighted by a drawing with three arms over them (fig. 15) or any of its variations (fig. 16), and less often are surrounded by illuminated designs (fig. 17).

Some annotations record textual variants concerning the text, the vocalization, or the accentuation from model codices and other manuscripts. Those mentioned are: Sefer Mugah, Mugah; another Masora (Masora aheret) from another manuscript; Sefarim qadmonim, “ancient manuscripts”; Hilleli; and Sefer ha-Pastani. The annotations located in the Pentateuch are the work of hand10, which offers a reading according to the codex Mugah in seven instances. Another eight annotations added by different hands give alternative information, mainly from the codex Hilleli. All of them are located in the Prophets, five in Ezekiel.⁵⁵ One of those annotations is written in the first person: “In the Sefer ha-Pastani I saw [written] about it ‘twice with qamech’ ” (fol. 233r). In five cases, the biblical text has been corrected to make the word conform to the textual variant found in the source mentioned in the annotation.⁵⁶ There are also a few annotations by other hands that present differences of opinions (pelugta’) and variations in the accentuation.⁵⁷

Numerous annotations added by many different hands provide information about how a word is written, mostly regarding whether its spelling is plene (full) or defective.⁵⁸ Hand11 is noteworthy for annotating those spellings. The biblical text is corrected to conform the lemma to the plene or defective annotation in ten instances, adding the missing letter for the plene annotations or erasing it in the case of the defective ones.⁵⁹ Some other annotations provide information about an odd or unique spelling of a word, when one word is written as two but must be read as one, when one very long word that seems to be two words together is just one word, and when a letter is one of the large ones in the Bible.⁶⁰ Finally, other annotations give information about the vocalization or accentuation of its lemmas.⁶¹

Figure 15.

Fol. 296r.

Figure 16.

Fol. 255r.

Figure 17.

Fol. 163r.

The seder markers found in the books of the Pentateuch are the work of hand1 and are all enhanced by floral-like decorations in gold, red, and blue (fig. 18).⁶² The lemmas are not usually marked with a circellus unless they have another marginal annotation attached to them.⁶³ The seder markers in the Prophets are also mainly the work of hand1, but the decorations seem to be the work of two different hands. Floral-like decorations in gold, red, and blue, similar to those found surrounding seder markers from hand1, are predominant from folio 92v until folio 158v. However, from Jeremiah until the end of the Prophets, the predominant decorations are in gold and the motifs are more varied. The beetle-like design (fig. 19) and those with five-point, six-point, and eight-point stars are particularly notable.⁶⁴ In contrast to the decorations made by the other hand, the dot above the samek is not covered by the design and the color. Most of the lemmas have a circellus added by a later hand. However, the seder markers in the Writings were added by several different hands. From the beginning of the Writings until folio 290v, several different shapes of the samek can be distinguished. The dot over the samek is lacking in most of them. The lemmas are not marked. They are placed out of alignment. Apart from two cases (fols. 274r and 279v), all of them are decorated. The motifs are varied and most are colored in gold, while some others have no color at all.⁶⁵ The seder markers located from folio 290v until the end of the manuscript are the work of hand9. Finally, there are some cases where the seder marker is lacking, but there is a decoration in the corresponding place (fig. 20).⁶⁶

Although the division into paragraphs is indicated in the manuscript by spaces in the text (the “open” paragraph or petuha begins on a new line and the “closed” paragraph or setuma begins later on the same line where the previous paragraph ended), some annotations state that the paragraph is “closed.”⁶⁷ In all these cases, the preceptive space is lacking, and different hands try to emend this by providing information about what type of paragraph it is. The annotations are highlighted using different graphic signs. Another annotation (fol. 163r) notes that the paragraph is “open,” but this is written in the space left for the paragraph in the biblical text, not in the margins. The annotation contradicts what the space indicates, that the paragraph is “closed.”

Figure 18.

Fol. 8r.

Figure 19.

Fol. 200r.

Figure 20.

Fol. 281r.

Two main types can be established taking into account the content of these annotations and their number: systematic and non-systematic. The systematic additions are associated with specific Masoretic phenomena: the K-Q phenomenon, the differences between Ben Asher and Ben Naftali, the differences between the Eastern and Western Masoretes, superfluous letters, sedarim, the midpoint in the book, and the midpoint in the Torah. Those phenomena are annotated consistently throughout one or two parts of the Bible, mainly by two different later hands. The information given in these annotations is not completely new. Similar information is partially found in other parts of the manuscript written mainly by the principal scribe, hand1: (1) the number of the BA-BN differences in the Pentateuch is given in a list written in a figurative shape located in the third column of folio 81v, at the end of Deuteronomy, and in the long lists at the end of the manuscript, in the so-called Appendix I (fols. 82v–84r); (2) the list of variations between the Eastern and Western Masoretes is given in the third column of folio 159v at the end of the book of Kings, written in a figurative shape; (3) the K-Q phenomenon is already annotated by hand1 in the form of a list, grouping the cases that share one reading feature in the outer, intercolumnar, upper and lower margins, and in Appendix IV (fols. 338r–342r); (4) the midpoint of the books of the Pentateuch regarding words and letters is given in the list in a figurative shape located on folio 81v and in Appendix I; and (5) the list of the sedarim of the book of Kings is given on folio 159v.

The non-systematic annotations have been added here and there in the manuscript by many distinct sporadic hands. They cover a wide variety of topics: accents, vocalization, defective or plene spelling, odd spelling, other opinions/variant readings, yafeh, unique words, word division, large letters, closed paragraphs, and book summaries. The information gathered in the annotations of this type is not previously indicated in the manuscript, except for the annotation on folio 240r that contains redundant information already existing in a previous annotation.

Conclusions

The analysis of the script of the marginal annotations of M1, of how information is recorded on them, and of their placement within the folio has revealed a high number of annotations written by different hands other than the principal scribe, mainly in the outer, inner, and intercolumnar margins. Indeed, multiple hands have been distinguished, and twelve of them have been characterized. Most of the annotations are written in calligraphic Sephardic square script, although some annotations are written in an informal style, three in Sephardic cursive script and one in Sephardic semicursive script. It is quite common for the later annotations to be written vertically and very close to the biblical text, as the space for the MP annotations is already occupied. The content of these later annotations is varied and addresses the usual topics found in the Masora. While some hands only sporadically added annotations, others were responsible for systematic additions associated with different Masoretic phenomena.

The later annotations appear throughout the Bible, but while annotations without systematization written by many hands are more common in the Pentateuch, systematic additions are more frequently found in the Prophets and Writings. These later interventions have different aims and produce diverse results in the manuscript. The systematic additions that deal with the BA-BN differences, the Eastern and Western reading discrepancies, and the K-Q phenomenon reveal an interest in organizing the information in a different way. The people in charge of those additions deviate from the principal scribe’s initial plan for the Masora of this manuscript by setting down and highlighting in the outer and intercolumnar margins a series of phenomena often already found in other parts of the manuscript or expressed in another way. The most paradigmatic case is that of the K-Q phenomenon. The principal scribe tended to present the cases grouped into lists according to common reading characteristics in the upper and lower margins, and sometimes in the outer and intercolumnar margins. The later hands transformed this approach by replacing or adding annotations on individual cases. They called attention to the phenomenon by annotating each particular case separately in the Prophets and the Hagiographa and by adding a drawing over them. As a consequence of changing the approach to the phenomenon, several lemmas have up to three annotations attached. The fact that most of these additions were the work of two hands underscores the intentionality. As a result of these systematic additions, in its final form, this manuscript contains information in the outer and intercolumnar margins (such as BA-BN differences and Eastern and Western reading discrepancies) that—in most of the Masoretic Bibles—appears at their beginning or end. Moreover, these systematic annotations make the work of finding the important information easier by giving information about the lemma very close to where it appears in the body of the text, individualizing and highlighting it. The other systematic additions (midpoint annotations, superfluous letters, markings for the synagogue readings) aim to offer the information on each phenomenon consistently by adding cases that are lacking.

The non-systematic additions seem to be more the result or reflection of later revisions, readings, and uses of the manuscript. The annotations recording variants from the most accurate manuscripts—especially the only first-person annotation found on folio 233r—and those presenting other opinions attest that the biblical text and the marginal annotations of M1 were carefully checked and compared with other relevant manuscripts. The rest of the annotations are testimony to the continuous use of the manuscript over the ages by many different people, who added additional information they considered relevant for the transmission of the biblical text.

The numerous vestiges of proofreading and editorial activity in the marginal annotations made by hand1 and other later hands reveal a special interest in transmitting the Masoretic annotations as accurately as possible, a tendency not very common in the Masora of other biblical manuscripts. The fact that the proofreading activity consists mainly of adding omitted letters and words demonstrates that the copy of the Masora was already very accurate. However, the corrections to the biblical text brought about by marginal annotations of both types do not reflect an aim to correct the errors in the copy of the biblical text (unlike other cases in the manuscript, in which the corrections in the biblical text are the result of the proofreading of the principal scribe). Rather, they reflect the attitude toward the biblical text maintained by some individuals, who considered it erroneous when it disagreed with the information they had and had added in the margins and, consequently, corrected the text. This is confirmed by the annotations recording information from other sources or accurate manuscripts that, except for a few occasions, do not lead to a correction of the biblical text, even when they contain disagreements concerning the text, the vocalization, or the accentuation.

In summary, the identified later annotations show a clear intention to complete and expand the already abundant information found in the margins of M1 more than to correct it (although some alterations are found).⁶⁸ All of this work serves to enrich an already valuable, lengthy, and accurate Masora of this manuscript, making it one of the elements that endows it with great value. And once again, the exceptional nature of the Masora of M1, which has no parallel in any other thirteenth-century Toledan Bible, is confirmed.