Use of the Modular Method in Medieval Manuscript Illumination

Modular book of hours, opened at the beginning of the Vigil for the Dead. / Special Collections of the Universiteitsbibliotheek van Amsterdam

By Dr. Kathryn M. Rudy / 09.19.2016
Senior Lecturer in Art History and Medieval Studies
University of St. Andrews

From Piety in Pieces: How Medieval Readers Customized their Manuscripts


Toward the end of the fourteenth century, the manner in which manuscripts were made changed dramatically. In the late fourteenth and throughout the fifteenth centuries, book makers created their wares increasingly with what I call the modular method, an approach to construction that takes into account a division of labor and a need for efficiency, and that presupposes an owner who would expand the book later. This had serious implications for the ways in which book owners could make augmentations to their books.

Along with a shift in production methods came a shift in book type. Before ca. 1260, the psalter was the main text for private devotion. Psalters contain the 150 psalms, which a supplicant would read in fixed groups over the course of days or weeks, and thereby work through the entire text from cover to cover. A psalter’s text was therefore static. Although psalters continued to be made and used for the duration of the Middle Ages, beginning around 1260, the book of hours gradually replaced the psalter as the predominant book for private devotion.[1] Because of the canonical status of the psalms, and the longevity of the physical book, the process of change was slow. Psalters, of course, continued to be made, and combination books, such as the Liège Psalter-Hours, featured below, provided a transitional form.[2] Books of hours contained some standard texts: calendar, Little Office of the Virgin, Penitential Psalms and Litany, and Office of the Dead. They often appeared in this order, but their sequence was by no means fixed. Furthermore the book of hours usually contained other texts as well, such as the popular prayers O Intemerata and Obsecro te. Books of hours became increasingly widespread in France, England and the Netherlands in the mid- to late fourteenth century. In the early years of production, say, until ca. 1390, the book of hours was often made as a bespoke product, with a planner mapping out the entire book as one unit. Texts developed coevally with cycles of imagery—including infancy and passion cycles—which differed regionally and in which there was plenty of latitude.[3]

Just as the psalter might be read completely from beginning to end, it was also written in the same way, from beginning to end. In contradistinction to that, the book of hours contained many different texts, intended to be read on an as-needed basis. Users would dip into the texts that were appropriate for the moment, reading, for example, just the litany in one sitting, or just the vespers of the Hours of the Virgin. They might read the Hours of the Cross in the week before Good Friday, the Hours of the Holy Spirit to prepare for Pentecost, and the Penitential Psalms during Lent. Seasonal appropriateness and personal interest could determine which texts to read at any given hour and day. Whereas a psalter was canonical, there was never an entirely standard set of texts that made up the book of hours. Their composition always varied, both at the time of production and by force of an owner’s interests and desires.[4]

Furthermore, literacy increased considerably from the thirteenth to the fifteenth century, and new literate urban classes demanded more affordable books. These two forces—reading style and economics—meant that books of hours were written and designed differently from psalters. Books of hours (at least in the Netherlands) were made increasingly in smaller segments that could be brought together during binding. In other words, they were both read and produced in units—that is, modularly.[5]

Modular and Non-Modular Compared

To explain the modular method of manuscript construction with a concrete example, I compare two manuscripts, an earlier psalter-hours, and a later book of hours. The psalter-hours was made in the second half of the thirteenth century in Liège (HKB, Ms. 76 G 17). It is written in Latin and French and was made for a Beguine at the beguinage of St. Agnes in Maaseyck. This book’s structure is typical of many thirteenth-century productions. Historiated initials mark major text divisions with corresponding gold and painted decoration in the margin. Major psalm divisions, with their extra decoration, can occur on a recto or verso, at the top of the folio or in the middle. For example, the initial for Psalm 26, which depicts Christ healing a blind man, appears near the bottom of a verso folio (HKB, Ms. 76 G 17, fol. 20v; fig. 1).[6] When the scribe set out to write this text, he simply began at the beginning and continued to the end, filling the requisite number of quires. He left space for the rubrics and the decorated initials as he went along. The placement of historiated initials was by default dictated by the scribe.[7] The painter who made the bar borders was either the same person as, or else worked closely with, the person who painted the figures. These painters then sent their work back to the scribe, who at that point filled in the labels identifying St. John the Baptist and Moses, figures occupying the bas-de-page. The scribe also filled in Moses’s scroll. This manuscript reveals the degree to which the production required careful coordination between scribes and illuminators, who may have been working under the same roof. By contrast, in the late fourteenth and fifteenth centuries, scribes and planners had a different concept of book production, one marked by a sharper division of labor.

Fig. 2 Modular book of hours, opened at the beginning of the Vigil for the Dead. Special Collections of the Universiteitsbibliotheek van Amsterdam, Ms. I G 54, fol. 32v-33r. Image © Universiteitsbibliotheek van Amsterdam

A book of hours now in Amsterdam reveals that it was constructed according to a different set of principles (AUB, Ms. I G 54; fig. 2). I could have used any one of hundreds of books of hours to complete this comparison, but I chose this one because its binding is loose, which allows one to see the structure more easily. Each new text begins on the top recto of a fresh quire. Though the texts vary in length, the scribe simply used as many quires as were necessary for that text. He then began the next text on a blank recto at the beginning of a new quire. Some blank, ruled parchment invariably fell at the end. This book’s scribe has composed each section as a separate packet, and then sewn them together. A close look at the photograph reveals the division between two modules.

Changing fashion for types of devotional literature both necessitated and encouraged a new mode of production; that new mode of production in turn spurred further changes in devotional literature. Several forces—including literacy rates, new forms of private devotion, and economic forces—coalesced to fundamentally change how manuscripts were made in the decades just before the printing press. A closer look at the structure of the book of hours clarifies how this procedure developed. Let me explain.

Books of hours proved so popular that a high demand encouraged new, cheaper methods of production, which would also allow a wider audience to buy them. In Bruges around 1390, a new development occurred: a group of illuminators known as the Masters of the Pink Canopies began making full-page miniatures in ateliers separate from where the texts were written. These miniatures could then be inserted into the book-block before binding. Primarily they made full-page miniatures for books of hours for export to England. These were made in considerable quantities. Nicholas Rogers has identified 170 surviving books of hours made for this export market between 1390 and 1520.[8] The Masters of the Pink Canopies were some of the first artists to systematically exploit the new production and design concepts I have been outlining, which allowed production to swell. Pink Canopy manuscripts have for the most part been discussed in the art historical literature, because scholars have been interested principally in the miniatures and not in the bookish substrates those miniatures lived in.[9] Consequently there is no comparable name for the group of copyists who inscribed the books that contains their work. The fact that they form a group of “masters” with no identifiable individuals among them suggests that their labor was as interchangeable as the miniatures they made. The rather large number of surviving manuscripts containing their wares points to an efficient atelier (or group of ateliers) making miniatures for an export market of non-bespoke manuscripts.

For example, they produced a packet of images that was then bound into a book of hours for Sarum use in the last decade of the fourteenth century (Cambridge, UL, Ms. Ii.6.2).[10] This packet included images added to the Hours of the Virgin: an Annunciation, a Visitation, a Nativity, an Annunciation to the Shepherds, the Adoration of the Three Magi, the Massacre of the Innocents, the Flight into Egypt, and the Presentation in the Temple (fig. 3).[11] Included in this manuscripts are also single-leaf miniatures depicting single standing saints to preface short suffrages at the beginning of the manuscript, as well as a miniature depicting the Virgin in a radiant sunburst to preface the Salve Regina; a Crucifixion to preface a prayer to Christ’s limbs; an image of Christ as Man of Sorrows with the arma Christi which preface the Seven Penitential Psalms (fig. 4);[12] and a funeral service to preface the Office of the Dead (fig. 5).[13] These images were designed to enhance a book of hours, to make it more colorful and appealing, and to roughly gauge the interests of a recipient (piety to Mary and to Christ’s suffering, and to an assortment of popular saints).

Two features are striking about these images: first, their large number for such a relatively modest book of hours; and second, their sheer clumsiness. Although the full-page miniatures marking each canonical hour of the Virgin represent a lavish outlay of color, the individual paintings rely heavily on formulas and lack convincing spatial illusion. Patterned backgrounds—such as the swirling gold filigree on the black backdrop behind the Presentation in the Temple—add opulence, but in fact required little skill to apply. Likewise, voluminous drapery fills the pictorial space with color and patterns, but it obfuscates anatomy and structure. In the Presentation, an altar mostly covers Simeon’s lower body—his upper body appears as an indistinct swirl of drapery—while Mary’s body appears as a blue area of fabric. For the funeral service, the artist has reduced the number of figures to two and has avoided showing the figures’ hands, no doubt because they are difficult to draw. Instead the artist has filled most of the available space with a coffin, which—like the body of the Virgin earlier—is entirely covered with drapery. For the artist of limited skill, pattern (such as the red dots on the fabric) trumps volume, because showing three-dimensional forms in space is difficult. Flattened, patterned colorful shapes apparently fulfilled the buyers’ desires well enough, for the Masters of the Pink Canopies did a swift trade in miniatures. The opportunity to own colorful images, even incompetent ones, must have played a significant role in the rising popularity of the book of hours.

By allowing ateliers to specialize—to just make full-page miniatures, without having to copy all the texts—the division of labor streamlined production and also changed the imagery by simplifying and standardizing it. The Masters of the Pink Canopies began each sheet by drawing a standard-sized frame, topped with an eponymous pink canopy. These choices were givens. They then filled in popular subjects, often using patterns to further reduce the labor of having to come up with new compositions. For example, the composition with Christ as Man of Sorrows among the arma Christi (fig. 4) is one that appears in other manuscripts. A close free-hand copy appears in another book of hours made in Bruges for export to England (LBL, Sloane Ms. 2683; fig. 6).[14] This subject must have been a calculated choice for replication, because the image could be used in front of a variety of texts. In Cambridge, UL, Ms. Ii.6.2, it prefaces the Seven Penitential Psalms, whereas in Sloane 2683 it prefaces a prayer to the wood of the cross, to the crown of thorns, to the Five Wounds (taken one at a time), to the Virgin, and to St. John. A rubric prefacing the prayer indicates that it should be read before an image of the crucifix. The Pink Canopy miniature provides the required image of the crucifix alongside several other items mentioned in the prayer. Thus, this image was flexible and therefore interchangeable: it could find a home in nearly any book of hours. Second, its simplicity must have appealed to the half-trained Masters of the Pink Canopies. Depicting the naked figure of Jesus provided challenges for the artists, but they followed a formula, which made it easier. Otherwise, most of the surface is given to the arma Christi, which comprise simple geometric shapes and posed little challenge even to a maladroit painter. Part of the appeal and popularity of the arma Christi as a devotional aid must have been the ease with which it could be reproduced with little skill. By churning out ubiquitous infancy imagery, popular saints, and flexible motifs such as the arma Christi, these “masters” could contribute image-modules to accompany text-modules produced elsewhere.

Other examples confirm that the new high-volume miniaturists sought out simple imagery. Among the miniatures included in Cambridge, UL, Ms. Ii.6.2 is the Face of Christ (fig. 7).[15] Various elements of the standard iconography have been reduced, abstracted, or obfuscated. Veronica herself is missing altogether, and the face has swollen to nearly the full width of the page. Either the face was traced from a template or copied freehand based on transcribing simple geometric shapes. While visually “elevating” the subject, the cloth of honor obviates the need for drafting recessional space, which is difficult to depict convincingly. On the cloth and throughout the image, the artist used pattern (which is simple and repetitive) instead of modeling (which requires more skill to achieve nuanced gradations). These artists were not capable of achieving logical coherence: a blue blob below the central boss of the highly formulaic canopy makes no structural sense. A checkerboard floor indicates an attempt at depicting recession, although the artist has not quite understood the principle, and the floor just looks wonky. The features visible in this opening speak to a strong division of labor (painter, copyist, manager who assembles components), and to the deskilling of painters, whose work was reduced to copying boilerplate models. This yielded symmetrical designs, expressionless figures and unmodulated areas of color, but ultimately, richly colored books of hours. Using under-skilled labor must have driven prices down far enough to create a demand. The English ate it up.[16]

That the book of hours—and no other kind of book—initiated this production method makes intuitive sense, because its makeup is inherently predisposed to being assembled in modules. In fact, the conceptualization of the book as modular might have arisen from the ways in which the calendar was produced. Because calendars were typically made on two quires of three bifolia, to make twelve pages for the twelve months, and because they were ruled differently from the rest of the manuscript, they were made in a separate campaign of work. In the Southern Netherlands, special ateliers may have developed just to produce calendars (more research will have to be done to understand that situation for certain). Book makers must have realized that other texts could similarly be made of components that could be slotted into place. Thus, instead of producing entire books of hours, makers could produce components, and let customers choose which texts to bind together into a book of hours. This reduction to components became especially useful in the suffrages, where ateliers in Bruges produced single leaves with images of saints, and single leaves with prayers to those saints. Customers must have been able to select the saints they wanted according to desire and budget. For customers living further afield (such as the English clientele for Bruges books of hours), either the customer could send an agent, or the producer could roughly anticipate the consumer’s desires by including some of the most popular saints, including Sts George and Thomas, whom the English adored. Each customer could then further specify these by using local labor.

Book makers used modules as the basis of bare-bones books of hours, which owners could expand not only with further images, but also with further texts. An owner might want, for example, a “basic” book of hours, but with a copy of the Short Hours of the Cross. A scribe could produce this desired text in a single quire, and it, along with the relevant imagery, could then be incorporated in the final book. This modular manuscript construction had wide implications for bringing down the costs of books of hours and also for allowing owners to “personalize” them. It led to increased standardization and more variety at the same time. It also allowed owners to buy as many images as they cared to, or could afford. Medieval books were expandable, but this feature led to further design limitations, because book design still had to conform to long-established ideals around decoration.

The Hierarchy of Decoration

This new modular method had to take into account the hierarchy of decoration that already governed text-image decisions scribes and illuminators made when they produced manuscripts. Although this concept is widely understood by those who study manuscripts, the “hierarchy of decoration” often forms an unspoken set of assumptions in works of modern scholarship.[17] In a nutshell, the hierarchy of decoration means that the decoration of a book reiterates and reinforces the structure of the text. Each manuscript has its own internally consistent design logic. A manuscript can be highly decorated or barely decorated at all. Regardless of the degree of embellishment, the design of a medieval book’s page layout is always organized around an initial.[18] One finds the largest and most lavishly painted initials marking the beginnings of the most important texts. With the largest initials always placed at the top left corner of a page, their decoration can spread vertically upward and laterally into the margin. Smaller initials might mark either subdivisions within that text, or less important texts; these could come partway down a page. Historiated initials (those containing narrative scenes in their letter frames) are higher in the hierarchy than decorated initials (with no figures). Decorated initials of cascading sizes mark further divisions. The number of rulings it fills up indicates an initial’s size. For example, a manuscript might have 12-line historiated initials to mark the major texts, and three-line decorated initials to mark the internal divisions within the text, and one-line initials in alternating red and blue ink to mark new sentences or phrases.

Using initials of varying sizes to signpost the structure of the text could be likened to our modern system of making outlines, beginning with Roman numerals, then capital letters, then Arabic numerals, then lowercase letters, and so forth. Note also that in an outline each subsection begins on a fresh line and is indented appropriately. Increasing indentation fills the same role in a modern outline as the descending size of the initial in a medieval manuscript. In both systems the goal is to make the structure of the text manifest in the layout of the page.

Two rules governing the hierarchy of decoration are that each kind of initial should accompany a specific level of border decoration, and border decoration always emanates from an initial. A border decoration’s grandeur can be quantified by how many sides of the page it fills up, and by its material, which may be (in descending order of grandeur): gold and painted figures, gold and painted abstract designs, just paint, or penwork. In this system, gold trumps paint. Paint trumps penwork. Figurative imagery trumps abstraction. Within a single manuscript, initials of various sizes correspond to flourishing of respective intensities. When a book’s decorative program breaks these rules, it usually signals a later intervention, an unplanned component, an impromptu addition. An owner’s desire was often stronger than a decorative program’s consistency.

According to the hierarchy of decoration, miniatures of cascading sizes also correspond to the grandeur of the initials, with the full-page miniature at the top of the chain, followed by column-wide miniatures, and then small miniatures that don’t fill a column, and so on down. Each of these levels corresponds to a level of border decoration. For example, 12-line historiated initials painted in tempera and framed in gold might accompany painted and gilded border decoration on four sides, while in the same manuscript four-line decorated (painted) initials accompany painted border decoration on one side.

All this decoration helps orient the book’s user. More colorful pages signal the beginnings of more important texts. Smaller initials, with their accompanying decoration that spilled out into the margins and was therefore visible when one was flipping through the book, helped users find other text passages. Within a single manuscript the internal logic should be consistent; if it is not, then that is a sign that the manuscript was pieced together from disparate parts.

From the time when parchment codices were first made in the fourth and fifth centuries, the hierarchy of decoration was planned from the outset, because producing a manuscript followed these steps: the parchment was first ruled; then the scribe would write the text; the illuminator would then apply the paintings and decoration; and finally the book block would go to a binder. The system of the hierarchy of decoration required that scribes know from the beginning what size all of the initials should be and where the miniatures should go, because they would have to leave appropriate space for them. In other words, for the first 800-or-so years of codex production, the scribe knew what all the texts and decoration would be in finished work. It was all planned from the beginning.

With the new modular system, the scribe and the illuminator would work in separate ateliers. An illuminator would not make images for a particular book, but rather make “interchangeable parts” that could be added to any book. A book’s scribe would not know where, or even if, images would be added to his work. There was little contact between painters and scribes, because a manager (sometimes called a stationer) would direct their labor. With this system, the hierarchy of decoration could not be anticipated from the beginning. Separating the painters from the copyists suddenly made certain page layouts extremely inconvenient, namely, those that had both figurative paintings and inscribed words. Consequently, the new system suppressed certain design elements, including column-wide miniatures and other kinds of painting that the scribe would have to anticipate.[19] This is not to say that book makers abandoned illuminations. On the contrary, they used images even more fanatically. The old hierarchy of decoration and the new separation of the text copying from the image making meant that images would primarily be conceived as full-page miniatures, painted on single leaves, because such images could be slotted into the book at the beginning of the quire, and scribes would not have to plan for them. In order to face the text they accompanied, miniatures would almost always be inserted from now on as versos, that is, on the left side of the opening. This situation created a new standard.

Among the structural changes that followed from this shift was that now all new texts had to begin on a fresh quire so that the initial began on a fresh recto. Since the miniature had to precede the initial, the only place for it to go was on the preceding folio as a full-page miniature. Thus, with the new modular method, nearly all books that bore miniatures would have those miniatures on the left side of the opening, to face a text that began on the right side. In the pre-modular system, on the other hand, design elements would fall where they may. Page layout therefore ossified under the new modular method. Whereas in the past, each book was hand-designed and planned (at great mental cost and labor), now manuscripts could be made according to a general set of principles, so that the interchangeable units could be made remotely.[20]

All of the examples of the Pink Canopy Masters’ work, given earlier, follow this standard: full-page miniature on the left of the opening, incipit on the right. In the opening with the Face of Christ, for example, someone (the stationer?) has inserted the image across from the prayer Salve Sancte Facies (fig. 7). The image has clearly been made in a separate campaign of work from the text, as the two sides of the opening have different border decoration. Moreover, the decoration around the text block has been reduced to a minimum. Scribes and initial painters were undoubtedly under as much pressure as illuminators to apply their respective technes, repeatedly and efficiently, and to the minimum standard that the consuming public would permit. Initial painters would do only a cursory job. Most of the decoration in the manuscript would be borne by illuminators such as the Masters of the Pink Canopies. If customers wanted more decoration, they could pay for it by adding more miniatures.

Conventions for a hierarchy of decoration mandated that major texts be flagged with illumination and that major text breaks be flagged with large initials. It follows also that minor text breaks begin on a new line, and that items in a list each occupy a separate line. When major texts had to begin on a fresh quire, the remainder of the previous quire would remain blank. The backs of the miniatures were also blank, when they were added to an already-complete text. In the opening before the Face of Christ, for example, this modularization process generated one-and-a-half folios of blank parchment (Cambridge, UL, Ms. Ii.6.2; fig. 8).[21] The more modularized book making became, the more blank space it left in its wake. In short, the new production method multiplied the negative space of the book.

Modules and Blank Space

Another principle of medieval design was horror vacui, the fear of empty spaces. With its addition of newly blank spaces, the modular method came into direct conflict with this design principle. As components were added, blank space was created, and it demanded decoration. Scribes, illuminators, and book users often had the urge to fill those blank spaces up. A pervasive urge to fill empty space was already in play in the early Middle Ages, as on the text pages of the Book of Kells, where an illuminator has filled the line endings with colorful geometric shapes, and in one case a horse and rider, and has even gone so far as to fill the empty space of the round letters with pools of color (Dublin, Trinity College, Ms. 58; fig. 9).[22] Avoiding trapped and blank white space remains a design principle today.

Fig. 10 Calendar folio from the Hours of Catherine of Cleves, with painted line endings and beginnings. New York, Morgan Library and Museum, Ms. M. 917, p. 17. Image © The Morgan Library and Museum, all rights reserved.

Using line endings and following the basic principle of filling empty space continued throughout the manuscript era, and resulted in some flamboyant displays of color and gold, all in the interest of preventing blank parchment within the text block, for example, in the calendar copied into the Hours of Catherine of Cleves (New York, Morgan Library and Museum, Ms. M. 917; fig. 10). A scribe has begun each saint’s name on a fresh line because the alternating blue and gold initials in this manuscript demand a fresh line according to the hierarchy of decoration. Starting each on a fresh line creates unsightly blanks at the ends of the lines. The decorator has solved this problem by painting “line endings,” long and narrow geometric forms painted in bold colors and gold. Book makers had varying degrees of tolerance for blank space in other parts of the book, but the new modular method generated space at the ends of quires that sometimes proved too much to fill.

Modular manuscripts made in the Southern Netherlands for English export created excessive empty space. In England in the two centuries before the Henrician Reformation, book makers could not keep up with demand, which is why England imported so many books of hours from the Low Countries. Bruges book makers were put under pressure by the sheer scale of the demand to find new means of production, and began using more workers whose tasks became more divided. Now, each worker (or atelier) was delivering components, and the people making them never had oversight over the whole, finished product. In and around Bruges, this method of production created demand for stationers who assembled the components and must also have been able to guide production, so that suppliers made enough Jesuses and not too many Barbaras and just the right number of Vigils of the Dead. At the other end, post-consumption, a small industry must have grown up in England for scriveners to fill some of that blank space with English prayers, that is, to refine the roughly tailored products.

Cambridge, UL, Ms. Ii.6.2 had pages and pages of blank parchment when it was shipped from Bruges to England, and its English owners responded by filling it in. Much of this blank space appears on the backs of the 20 miniatures supplied by the Masters of the Pink Canopies. The first miniature in the book depicts the Trinity (fig. 11).[23] The English recipient responded to its blank back by writing on it, or more likely, by hiring several scriveners to write on it (fig. 12).[24] An English hand added one short prayer to the top of the page, and then someone else added a second. Emendations of this sort lent themselves to brief texts which would not exceed the available space. Scribes often chose a fresh blank page to begin a new prayer; therefore, additional prayers fill the top sections of many of the blank backs of other miniatures.

Sometimes the locations of the additions are significant. One the back of the miniature of the personal angel (Cambridge, UL, Ms. Ii.6.2; fig. 13)[25] the owner has added a prayer in which words are interspersed with crosses, probably signaling that the reader should cross him- or herself at those points (fig. 14).[26] Such prayers were understood to form a shield around their performer, who covered him- or herself with signs. The owner may have chosen this spot—in close proximity to the angel—who was also understood to protect the bearer, in order to concentrate the protective prayers in one area of the book. Deeper in the manuscript, sixteenth-century owners have added birth dates of family members to the back of the image of the Annunciation (fig. 15).[27] These notices have been added in English over a period of decades, the scribes layering them so that they form a chronology for the family. Of course by writing the names on the same piece of parchment as the Annunciation (fig. 16),[28] the family associated itself with the Ur-Christian birth.

Not only were the full-page miniatures made on singletons, but some of the prayer texts were, too, specifically the suffrages. Short texts that rarely fill a single folio recto and verso, suffrages provided a main vehicle for honoring the saints of whom worshippers had grown visually fond, in sculpted or painted form. They demanded images of them alongside the prayers. This posed a problem for manuscripts made in the modular method, for if the suffrage texts were copied continuously into a quire, it would not be possible to slot full-page miniatures into the packet. A solution was to treat both the images and the prayer texts as singletons, then fold and glue them into modules. In Cambridge, UL, Ms. Ii.6.2, an entire quire of suffrages, filling fols 10–17, comprises singletons: single-leaf images glued to single-leaf text pages for form four bifolia, which can then be nested and sewn to a thong. In other words, each bifolium is a confection, glued together from two halves, rather than a continuous sheet.

The result of this production method was empty space. For example, suffrage to the personal angel, like most suffrages, is quite short. It fills only the recto of fol. 15, which is a singleton, as is the full-page miniature facing it, depicting the personal angel doing battle (fig. 13). Both folios are therefore blank on the back. That blank faces another blank, the empty back of the next full-page miniature, forming two blanks in a row. An English scribe has been commissioned, however, to write out a Latin prayer in a bookhand to fill this otherwise blank parchment (fig. 17).[29]

Similar situations occur throughout the suffrages in Cambridge, UL, Ms. Ii.6.2. To take another example, the full-page miniature depicting St. George faces a short suffrage to that saint, both units produced as singletons (fig. 18).[30] Turning the page reveals two blanks facing each other, which have served as extra parchment for the English owners’ whims (fig. 19).[31] Three scribes have been hired to add prayers in Latin and in English to this otherwise blank opening. The final prayer added in this group is dedicated to St. Erasmus, who became popular in the late fifteenth century but was not originally included in this early fifteenth-century manuscript. These prayers in this opening have been written by different hands, and different again from the hands on the previous folios. It would appear, then, that adding prayers to the blank areas in the book was a cumulative process, that the owner looked for space as his or her needs changed.

Not only did the English scribes write on the blank backs of painted and inscribed singletons, but they also found space at the ends of quires, which usually offered ruled parchment. To the extra ruled lines at the end of the Hours of the Virgin, an English scribe has carefully added a prayer to St. Botolph (Botwulf) of Thorney, an English abbot (fig. 20).[32] St. Botolph was understood to protect travellers, which is why medieval churches dedicated to him were situated at the major gates of London: at Aldersgate, Aldgate, Billingsgate and Bishopsgate. This prayer was but one of many that the English owners added to the book to make the foreign product locally relevant.

Cambridge, UL, Ms. Ii.6.2 is but one of many manuscripts made with the modular method in Bruges for export to England, which produced not only an attractive product for English buyers, but one that buyers could adapt. These cumulative inscriptions suggest that a cottage industry developed in England for scribes to add texts to the blank areas of books of hours. The example I have discussed above, and many others, all have additions made by multiple hands, as if updating the manuscript were a continuous process. English patrons could add prayers as they came into fashion and thereby keep their books fresh and up to date. An impressive amount of empty space was opened up in these books by the modular method, and therefore they were capable of hosting significant additions, without having to add new parchment.

Modular manuscripts often contain blank folios, because scribes typically started new texts on a fresh recto. In doing so they were anticipating that a (future) owner could insert a full-page miniature to the book on the left side of the opening, so that it would face the initial on the right. The southern Netherlands produced the most extreme cases of wasted parchment. These manuscripts were extremely modularized.

Precursors Book Modules

Some of the elements of the modular method were already present in earlier systems of making manuscripts: account books made out of single leaves; thirteenth-century Parisian Bibles copied with the pecia system; image cycles prefacing psalters; and compilation volumes made in monasteries. I take these in turn and discuss how producers of books of hours drew on their innovations and streamlined them in the fifteenth century.

Accountants and merchants used a bundling technique to make booklets from single leaves. They would organize receipts generated by impaling them on a nail on the wall so that these sheets, once collected and bound, would become books. Jan Gossaert’s portrait of a merchant of around 1530 depicts such a record-keeping system (fig. 21).[33] A merchant has gathered single sheets one after the next on his nail. When the bundle is sufficiently thick, or a particular period had passed, say, a month, he would take the leaves off the wall and bind them. In this way, the resulting booklets would then serve as a chronological record of transactions. This departed from the regular way in which books were made on pre-ruled bifolia that were nested into stacks to form quires, which were then sewn into a binding.

When Southern Netherlandish stationers applied this idea to the illuminated manuscript, they were using an idea from work-a-day books and applying it to manuscripts that occupied a higher social stratum. With this new method of manuscript production, careful planning was unnecessary. In books of hours with suffrages to saints, the manuscript-makers compiled single sheets: a full-page image of a saint, followed by a single text sheet with a prayer to that saint, followed by the image and prayer to the next saint. Patrons could have as many saints as they wanted. As I will show later, book makers in the Southern Netherlands (which had a vigorous mercantile culture) took full advantage of this construction principle and made some packets out of single leaves arranged in a particular order. This also turned the manuscript-making procedure upside-down. In the past, the book would be planned, and the text inscribed with spaces left for the miniatures. But here, in this section of suffrages, the book maker begins with full-page singletons with images; it was the text quires that had to be cut up to accommodate them. They’re all singletons, just like the merchant’s receipts.

Another earlier book-making technique that the proponents of the modular method developed came from the way in which Bibles were copied. The university of Paris was founded around 1200 and flourished in the thirteenth century, attracting students from all over Europe. They created a high demand for the Bible, which was the main textbook used. Students needed a single-volume Bible that was handier and more economical than the enormous multi-volume Bibles that until then had been the norm. The new “Paris Bible” was small and made on thin parchment and written in highly abbreviated words so that it was physically manageable and as inexpensive as possible. Furthermore, it was copied according to a standard canonical exemplar so that all students would have the same texts in the same order.[34] Efficiency and standardization often go hand-in-hand. Although many used copies were in circulation by the end of the thirteenth century, the demand for small bibles far exceeded the supply in the first half of the century. This put pressure on the system to invent new ways to bring the costs down and to increase the speed of production, while maintaining a standard of quality. Taking manuscript production out of the hands of monks and putting it into the hands of secular urban scribes was the first step to increasing efficiency. In addition to urbanizing production and making the bible much smaller so that it would require less physical material, book makers began copying it in a new way, namely with the pecia system, which means “piece.” The university authorized copies of the bible and of other key works. These served as exemplars. Students could hire professional scribes who would each have access to one piece of the bible and could copy a volume as a team, each member with his own piece. In this way, the book was broken into modules and copied through a group effort.

This system, however, is different from the modular method of making books of hours, in that the quires of a bible did not correspond to discrete texts and were not therefore “stand-alone” units. For a bible to be a bible, it had to be complete and canonical. A book of hours, on the other hand, could comprise multiple units combined according to a buyer’s wishes. To be complete, it required very few units: owners often demanded more. Whereas the goal of the bible-copying pecia system was efficiency, regularity, and cost-cutting, the goals of the modular method in the fifteenth century were interchangeability, expandability, and of course cost-cutting as well.

Widespread in the fifteenth century, the practice of adding a quire made elsewhere may stem from an earlier practice of adding prefatory images to psalters. Certain luxury psalters were provided with a packet of images that formed a quire or two at the beginning of the text. They appeared in some quantity in the thirteenth century, resulting in semi-autonomous groups of folios containing pictures and no texts.[35] The exact composition of these images fluctuated but often included full-page images depicting saints, and perhaps a Passion cycle that served as elaborate frontispieces for the book. The images, which depicted post-biblical saints and a passion cycle based on the New Testament, did not “illustrate” the psalter, which is a text that comes from the Old Testament. Rather, they framed the psalter by imposing a Christian context onto it and provided a visual means by which to pray. Structurally these prefatory images were made separately from the text pages and the textual and visual components were only brought together later.

Modules that were made separately tended to come apart at the seams. Some image cycles have survived into the modern era loose, which has led some to believe that such image-only modules simply circulated as separate entities. For example, it is not clear whether a particular cycle of images made in Northern France, which now forms a separate book, was ever the prefatory cycle from a psalter (HKB, Ms. 76 F 5; fig. 22).[36] If it did, then the psalter has not survived or cannot be identified as such. What is clear is that illuminators designed image-oriented modules, and copyists did not need to plan for them from the outset. In the Northern French cycle, the illuminator has divided the page not according to the needs of a copyist, but for the needs of the pictorial unfolding of the story, where geometric separators hold the vignettes apart. Each page has been bisected horizontally and vertically to create four frames. However, for the story of Moses parting the Red Sea on fol. 7r, the illuminator has used the full width of the page rather than dividing it vertically by a bar, because he needed a wider landscape to tell the story. In short, no scribe planned the space for these images, and they have been made without respect to a particular text or word-oriented page layout. In that sense, they anticipate image-making procedures of the late fourteenth and fifteenth centuries.

Building a book of hours out of modules is closely related to making composite volumes out of separate parts or booklets. However, building manuscripts out of bookles was distinct from compiling texts. Whereas compilatio was a literary activity, assembling composite volumes was a physical activity. The former involved copying texts together, and the latter involved binding codicological units together.[37] Making a composite volume was a way to bring together small loose booklets, sewing them together into a set, and placing them in a single binding, which would protect them. A composite volume from the convent of Canonesses Regular of St. Agnes in Maaseik has several modules, each apparently copied by a different sister. These were brought together into a rigid binding and given as a gift to the sisters’ mother superior. In this way, the fact that the composite volume includes the labor of several sisters is integral to its function as a collective gift.[38]

In 1980 Pamela Robinson first called attention to volumes containing separate codicological units that have different scribes, and she applied the term “booklets” to them.[39] A booklet forms an autonomous unit and may comprise any number of quires. As such, the content, author, copyist, and date of one booklet may have nothing to do with that of its binding mates. According to Denis Muzerelle in his Vocabulaire codicologique of 1985, a manuscript made of booklets is a volume composite. He defined the volume composite as “a volume created by combining independent codicological units.” But as Peter Gumbert points out, this definition “leaves no room for the numerous cases where units are combined which are not quite independent and yet distinct.”[40] For example, the books of hours I discuss in this study comprise parts that were made as distinct units, but would not function independently. Gumbert suggests that the production of such books could be described as “articulated in blocks,” which could often mean that they were copied by different scribes. I agree and emphasize that the blocks of a book of hours might not be pre-determined at the outset of production; that the blocks could be interchangeable; and that they could be bound in many different configurations, most often with the calendar first, the Vigil last, and the texts in the middle varying widely. In criticizing the codicological terms defined by Muzerelle and by Maniaci,[41] Gumbert urges scholars to make “provisions for the many shades between ‘made in one piece’ and ‘built up out of independent items,’ and between ‘sensible’ and ‘random’ combinations.”

Rather than using the term “booklet,” “codicological unit,” or “production units” (a term Erik Kwakkel uses), I am using the term “modules,” because it’s more concise, has an adjectival form and can refer to single leaves and to parts of books of hours, which were made (semi-) independently. Building manuscripts out of modules has a long history. One of the earliest illustrated copies of the Gospels—the Rossano Gospels of the sixth century—has an “author portrait” of St. Mark that is probably not original, but added later.[42] St. Mark appears on a bifolium, which according to Kresten and Prato, has been added to the book. Retrofitting manuscripts to make them conform to new standards (i.e., having author portraits) by adding physical material to them is a practice nearly as old at the codex itself. As the Rossano Gospels demonstrates, book makers had been adding images made on separate leaves to existing books for a thousand years, but it was only in the final century of the codex on parchment that the production and insertion of such images became routine.

The modular method of producing and adding to manuscripts as practiced in the fifteenth century would take this idea—of separating labor and assembling books from components—to a higher degree of systemization. Responding to pressure to make more books (especially books of hours) cheaper, but to supply them with dazzling color and personality, book makers in the fifteenth century looked to the past for some ideas, and then updated those ideas and deployed them on a large scale.

Implications of the Modular Method

This new strategy for building books out of components has several implications. First, when manuscripts are made in modules, every major division falls on the recto of a new quire. This makes it very easy to slip a full-page miniature into the book so that it faces the incipit. Artists and groups of artists working in a similar style exploited this fact by constructing packets of single-leaf miniatures that owners could buy later and easily bind into their books.[43]

Fig. 23 Book of hours made in the modular method, unbound. Special Collections of the Universiteitsbibliotheek van Amsterdam, Ms. I G 50. Image © Universiteitsbibliotheek van Amsterdam

This appears, for example, in book of hours in the University Library at Amsterdam that is now unbound, which allows one easily to see the quires fanned out. Some of the quires contain a complete text, and other longer texts require multiple quires (AUB, Ms. I G 50; fig. 23). The first text, the Hours of the Virgin, fills five quires. The next text, flagged by a large decorated initial, fills two quires, and so on. It is clear the scribe did not produce this book of hours from beginning to end, but rather, as a series of booklets that could be arranged in any order and assembled just before binding. In fact, this is one of the reasons that Netherlandish books of hours come in so many varieties, with no standard organization. Often the Hours of the Virgin appears first, right after the calendar, but the Hours of Eternal Wisdom can also be first. Sometimes books of hours contain the Long Hours of the Cross, sometimes the Short Hours of the Cross, and sometimes both. Sometimes they contain one or two quires of prayers to the sacrament, or other non-essential prayers, such as indulgenced prayers. Clearly an array of possibilities was fostered by the new method of construction.

Fig. 24 Original production (left of the gutter) and an added quire, with incipit of an indulgenced prayer (right of the gutter). Special Collections of the Universiteitsbibliotheek van Amsterdam, Ms. I G 54, fol. 64v-65r. Image © Universiteitsbibliotheek van Amsterdam

Second, conceptualizing the book as a series of packets means that there is suddenly plenty of empty space—blank parchment—that could be filled later with more texts. Using packets was essentially wasteful, because the end of the text rarely fell at the very end of a packet. If it ended before the final folio, the rest of the quire would be left blank. In AUB, Ms. I G 54 (fig. 2), the end of the quire has six lines of ruled, empty parchment, while the new text begins on the fresh quire with an enormous gilt M. Many manuscripts built with the same procedure (using modules) have an entire folio of blank space at the ends of gatherings. Rarely did the scribe fill the quire perfectly, right up to the last inscribed line. Instead, the procedure usually left at least a few blank lines. That same scribe—or a different one—could fill these ruled lines with short texts; I have designated a category of brief texts used for this purpose as “quire fillers.” One convent (the St. Ursula convent in Delft, discussed later) made a virtue of these leftover spaces and kept short texts on hand to fill up these otherwise blank areas.

Third, it meant that books could be easily expanded. One could buy a book now with an eye to expanding it later. In fact, AUB, Ms. I G 54, introduced earlier, received an extra quire at the end (fig. 24). This packet was made in a different campaign of work, as the jarringly new borders reveal: the added modules have cheaper pen borders and lack gold, and they were clearly not made by the same decorator as the rest of the quires in the book. One can easily see that the style of decoration strongly contrasts with those of the original production. This quire also has a text that begins at the top of the first recto. The texts that this quire contains are heavily indulgenced and could be considered optional: they are not necessary components to a complete book of hours. They begin with the Adoro te, for which the reader of this booklet will receive 46,003 years’ and 59 days’ indulgence. The owner of the book, it would seem, received a basic book of hours, and later had it expanded with a quire full of indulgenced texts. This was easy to do, since such a packet was available as a module that could be slotted into the existing book. Doing so in no way disrupted the structure of the book’s contents, which were modular anyway.

Adopters of the Modular Method

Fig. 25 Modular method at the convent of St. Ursula. Diagram showing the modules comprising Cambridge, Trinity College, Ms. B.1.46. Diagram © Author (Click Image to Enlarge)

Although this is not a monograph about any particular atelier, a few production studios—secular workshops in Bruges and convents in Delft—will make multiple appearances. One of these is the St. Ursula convent in Delft, founded in 1454 or 1457, which probably produced manuscripts for its own use and for sale outside the conventual walls (or for donation to patrons, with the expectation of a gift in returm). These sisters built many books of hours from modules. Manuscripts connected to this convent share a certain set of “brand” features. They all have St. Ursula and the 11,000 Virgins listed in the litany as the first virgin; they all have painted or penwork borders typical of Delft; and they all have a particular kind of even, blocky script. A book of hours with the typical red and blue penwork of Delft and simple painted imagery was probably made by and for one of the inmates of this Franciscan convent in the 1470s (Cambridge, Trinity College, Ms. B.1.46).[44] At least two and possibly more closely related hands copied the manuscript in modules, so that the text for each office stands as a production unit.[45] A diagram of Cambridge, TC, Ms. B.1.46 demonstrates how the parts fit together (fig. 25). Modules in yellow are required in a book of hours. Those in green are common but not required. Those magenta and orange, the eighth and ninth modules, respectively, are idiosyncratic, and the book’s owner would have had to specify them or have them made separately. A single text fills the entire eighth module: the Hundred Articles, which apparently circulated only in female monastic houses. The ninth module contains prayers for the sacrament and indulgenced prayers.

In the cities of the western Netherlands (including Delft, Amsterdam, Haarlem, and Leiden), conventual scribes wrote in a corporate hand, which minimized the differences between the individual scripts.[46] As a result, this book of hours (Cambridge, TC, Ms. B.1.46) looks quite uniform, just as a Bible made in the pecia system looks quite uniform. Packets of quires comprise the normal texts (Hours of the Cross, Hours of the Virgin, etc.), as well as the texts that individuate this book. Based on these clues, I suspect that the sisters of St. Ursula probably assembled B.1.46 for their own use. It contains, for example, a prayer “to those who give us alms,” which, with its first-person plural form and the social relationship it implies, would make the prayer appropriate for a sister in a convent. Furthermore, a generalized convent sister is depicted in the margins of several folios. In manuscripts decorated for their own use, the sisters did not create lavish painted borders or gold, but rather limited themselves to penwork borders, albeit ones with some painted figures and historiated initials.

Fig. 26 Modular method at the convent of St. Ursula. Diagram showing the modules comprising Cambridge, Fitzwilliam Museum, Ms. 138. Diagram © Author (Click Image to Enlarge)

In those cases where an individual owner was purchasing the manuscript, the owner could apparently order a manuscript from the convent with just the text inscribed. He or she could then take the textual skeleton to a professional artist to have full-page miniatures added, as well as having the textual folios decorated in a matching style. This is what happened with a manuscript book of hours now in the Fitzwilliam Museum (Ms. 138; fig. 26). Bare modular quires, comprehending only text and no decoration, were all that the sisters of St. Ursula initially supplied. A professional artist easily slipped full-page miniatures in front of each module, because a new text always begins on a new quire. I suspect that the professional atelier then painted the incipit text folios so that they matched the borders around the illuminations. The resulting openings look uniform, as it they had been carefully planned from the beginning, when in fact, they are built out of components.

My discussion above has largely involved the sisters of St. Ursula and St. Margaret—for religious women produced many if not most of these modular quires. This is itself something new. Standard manuscript history, by contrast, goes like this: In the early Middle Ages, monks copied manuscripts in monasteries, largely for their own use. Around 1200, the University of Paris attracted students, who needed copies of the bible, which was the main textbook. Monasteries could not meet this demand, and a new, secular urban production of book making ensued. At the end of the fifteenth century, the market in manuscripts gave way to printed books. This is largely true, but there’s another, unexpected chapter: monasteries continued making manuscripts and existed alongside professional ateliers.

Around 1400, in the Northern Netherlands and in Germany, monasteries again became major players in manuscript production, this time making not bibles necessarily, but all kinds of manuscripts for use within the convent and especially outside it.[47] In fact, they overshadowed professional ateliers in the quantity of their production. For example, the Franciscan convent of St. Barbara in Delft produced manuscripts and acquired other manuscripts from outside the convent; their librarian produced a catalogue of their book holdings, which is nine folios in length and lists 109 items (HKB, Ms. 130 E 24).[48] (Later in the century they also operated some of the earliest printing presses, but that is the subject of a different book.[49]) Monastic inmates made books of hours and prayerbooks for themselves, but also for sale (or donation) outside the conventual walls. These manuscripts contain clues about their production in their script, decoration, calendars and litanies (in which the patron saints of the scribe’s monastic house are listed first among the confessors and virgins). Some of the convents had artists and illuminators, who could make miniatures. However, many convents did not, and they teamed up with professional artists, or pasted in prints purchased on the open market, to make a product complete with figurative images.[50]

In other cases book buyers purchased the written components from a convent, and then bought the miniatures separately from a professional, and then had all of the components bound together. Manuscript makers took advantage of this division of labor, and also took advantage of the collective labor in a convent in order to further streamline production. In order to make an internally consistent book of hours, which was actually written by five, six, or more hands, the sisters in a given convent had to learn to write in a similar, corporate style so that their efforts would be interchangeable and indistinguishable. Most convent sisters did not sign their work, but one group of sisters did: the sisters of St. Margaret in Gouda.[51] They collectively produced books that look as if a single scribe could have written them, but yet multiple people signed them. They were, in effect, anticipating by 500 years Henry Ford’s assembly line production, which relies on identical (or at least highly similar) fungible parts.

In the Northern Netherlands in the fifteenth century, urban convents made the bulk of surviving books of hours. Most of these convents housed women (although a few male monastics also produced manuscripts). Some professional ateliers also made books of hours, especially at the end of the century. Chief among them were the Masters of the Dark Eyes, who will reappear several times later. The large number of surviving books of hours indicates that they were affordable. In the north using the labor of women, which is nearly always more poorly remunerated than men’s, contained costs. In the Southern Netherlands, there were many more professional ateliers, and fewer monastic ones. In Bruges and its environs professionals made books of hours more affordable by dividing labor, reducing decoration of text pages as far as possible, and using simple templates to repeat motives. Manuscript makers in Bruges in the fifteenth century streamlined their production techniques so that most of the figurative imagery was executed on single-leaf miniatures, which were then brought together with text pages inscribed elsewhere. Both northern convents and southern ateliers used the modular method.

With the modular method book makers gained some efficiency by dividing labor, but at the same time they lost some material efficiency, as the new procedures wasted significant parchment. Manufacturing a book with single-leaf miniatures created empty parchment in the book, not only the blank backs of the miniatures, but the ends of the quires just before the inserted miniature. One suspects that demand for books was rising, labor costs were high, and the costs of raw materials, parchment in particular, were coming down, or else such wastefulness would not have justified the savings on labor. Above I showed how manuscripts made in Bruges and its environs often had enormous amounts of blank parchment in them. As a result, the exported books had plenty of room for the English owners to assert their personalities. Sisters at the convent of St. Ursula must have had lists of short religious texts on hand, because they left almost no ruled parchment blank. Instead, they filled it up with short sayings from church fathers, or brief prayers, or other texts I call “quire fillers.” In so doing, they trapped no white space on the page. They smoothed over the seams between modules. And they offered their customers even more value.

Blank parchment has not been discussed before because nearly all art historians concentrate on the fronts of miniatures. Photographs of blank pages make for boring plates. But I contend that the blank pages are key to understanding how the books were produced so efficiently, and how their new owners used them. Space was opened up by the modular process that might be used for two new kinds of products, both independently created to fill these spaces: quire-fillers and single-leaf miniatures.

Complicated Stratigraphy

In the fifteenth century English book producers could not supply enough books for lay votaries’ demands. As a result, the English imported books from the Southern Netherlands, where certain ateliers and artists specialized in making books of hours for Sarum use. Such manuscripts were made in modules, which resulted in many folios of blank or wasted parchment (the backs of full-page miniatures, and the blank ends of modules) bound into the book. English owners received their new, generalized books and almost always augmented them. Notating these layers of production is challenging; ideally a description should communicate the original and added physical material, as well as the original and added texts. When a manuscript is originally built out of components, designating “original” from “added” can be a slippery exercise.

Cambridge, UL, Ms. Ff.6.8 was made in the Southern Netherlands for export to England. It has prayer texts written in Latin, in brown ink by competent Netherlandish scribes, and numerous full-page miniatures, which have been bound in as singletons so that they face their relevant texts. The blank folios are concentrated at the beginning and end of the book and in areas with a high density of images. There are nearly as many hands among these additions as there are additions, suggesting that texts were added continually by several book owners.

The original modules are: the calendar; the suffrages; the Hours of the Virgin; Seven Penitential Psalms, Litany, and Vigil for the Dead; Commendatio animarum; and the Psalter of St. Jerome. Notable here is that the Penitential Psalms, Litany of Saints, and Vigil for the Dead form a single module. Whereas the litany invariably follows the Penitential Psalms and forms a unit with it, here the scribe has also tacked on the Vigil, which is a necessary and defining text of a book of hours. The scribe copied these texts together in order to make a somewhat larger module. Perhaps he did this for the sake of efficiency, as he knew that anyone ordering a book of hours would require these texts in this sequence.

On the other hand, the Commendatio animarum and the Psalter of St. Jerome were optional texts. These the scribe has copied in such a way that they each fill one quire. They could be slotted in easily, or omitted, as necessary. Further research may reveal that scribes manipulated these texts so that they would fill exactly one quire. In the two modules of suffrages, which each comprise a series of singletons, the solution was to alternate between image-text-image-text, so that the image of the saint always faces his or her prayer.

The scribe treated the Hours of the Virgin and various Marian prayers as one module. A full-page miniature appears on the left of the prayer’s incipit, and a text with a decorated initial stands on the facing folio (Cambridge, UL, Ms. Ff.6.8; fig. 27).[52]This is the standard design solution within the modular method. Clearly the manuscript met with approval. How much did the early readers have to use this book in order to rub off the bottom third of the miniature through handling? On the facing folio the text has also been heavily used, but has held up better, suggesting that the parchment and the paints were prepared in a different way. This further underscores the fact that the miniatures were made in one atelier, and the texts in another (and the decoration around the text folios in yet another), and that the various ateliers sourced their materials from different places and did not necessarily swap recipes. The paint of the miniature has not stuck well and has practically thrown itself from the page. Within the Hours of the Virgin, the scribe began some of the subtexts on a verso side (such as lauds, on 28v; fig. 28),[53]indicating that he or she did not anticipate a miniature to accompany the individual hours. On the other hand, the scribe left space at the end of the Hours of the Virgin in order to start the next prayer, the Salve Regina, on a fresh recto. This has allowed the book maker to insert a full-page miniature before the Salve Regina, specifically, an image depicting the Virgin of the Sun (fig. 29).[54]One can see at the opening with the Marian prayer beginning on fol. 51r that the text pages were painted at some distance from where the marginal decoration around the facing miniature was executed. Simply stated, the borders don’t match. Painters making single-leaf illuminations such as this one must have gotten significant mileage from this image type. Not only was it easy to paint—with the Virgin’s red and gold radiance represented as simple geometric shapes and patterning and most of the figure covered in drapery—but the image activated an indulgenced prayer, the Ave, sanctissima virgo Maria, which was much in demand. It could also be inserted before most other Marian prayers, such as the Salve Regina or the Magnificat.

The previous text in Ms. Ff.6.8—the Hours of the Virgin—finished on fol. 50r (fig. 30).[55]This means that the bottom of 50r and the entirety of 50v, consisted of blank ruled parchment, and that fol. 51r, which is the back of the full-page miniature, was also blank (fig. 31).[56]To transition from the end of one text to the jubilant, decorated beginning of the next cost 2 ½ pages of blank parchment. However, the English recipient has considered this white space an invitation to add further texts. On fol. 50r, he or she has inscribed the incipit of the Book of John, “In principio erat verbum….” This particular text was considered thaumaturgic, so that anyone carrying it would be protected from sudden death, without reading it. By no means a necessary text, it was nevertheless considered highly desirable to have in one’s book, and is therefore among those that owners most often squeeze in or find room for.

Wider and less confident, the English script of the “In principio” contrasts with the fine, even, decorous script of the Netherlandish scribe at the top of the folio. Notice, though, that the English scribe’s ink has flaked off severely, but the first campaign of work has stayed in place for more than half a millennium. It is easy to see from this page alone that the Netherlanders had a more developed book-making culture and could do a better job cheaper and faster than could the English at the time. That is why the English imported their books from their neighbors across the channel in the first place.

An English owner added (or commissioned from a scrivener) another prayer to the bottom of fol. 50v (fig. 31). This hand, also English, contrasts with that at the top of the folio and makes it clear that many additions were made to the empty space in this book, possibly over a period of several decades. Whereas the inscription of the incipit of the Book of John was made seemingly in that spot because there was sufficient space, the placement of this added text was not careless: the bottom of this folio had seven ruled blank lines, more than enough to inscribe a rhyming prayer to the Virgin, which then precedes her image.

Blank, ruled parchment also provided an opportunity to inscribe a prayer to St. Ethelburga, an English saint with a small cult (Cambridge, UL, Ms. Ff.6.8; fig. 32).[57] Stationers in Bruges must have been able to successfully make products for their English market, and this meant that the customers had ways of making their requests known through agents or middlemen. Netherlandish book makers apparently did not receive the message that Ethelburga was in demand, and her suffrage remained to be added locally.

An English owner found an entire blank opening before the image depicting Christ in Judgment, which prefaces the Seven Penitential Psalms (fig. 33[58] and 34).[59] That owner has, in fact, added a prayer to the Virgin in this blank area, one that has little to do with the Christological texts that preceded it or the Psalms that followed it (97v-98r). These additions, while sometimes placed carefully to give the new item a particular context are at other times added opportunistically, because the extra space was available.

This added Marian text appears in yet another rather poor-quality English hand. These additions suggests that either a single owner paid several scribes over a period of time to continually update the book, or else that a later owner took a stab at further personalizing the book. One of my observations, made after studying hundreds of such augmentations, is that a book that contains one is likely to contain many. Seeing one augmentation seems to encourage an owner to add more.

Overall, English scribes were worse than the Netherlandish scribes, even though the Bruges copyists were making hash out of some of the pages. I suspect that multiple people in Bruges wrote the various modules, which is one of the reasons to adopt a modular system in the first place, as several individuals could collaborate on a single product if they could work in a similar style. Such a system relies on standardization, although one can nevertheless see the seams between hands in the original part of the book, for example on fol. 11r, where there is a change of hand half-way down the page (Cambridge, UL, Ms. Ff.6.8; fig. 35).[60] (Compare the two instances of the word “Deus” above and below the seam in the middle of the page.) Saskia van Bergen has investigated the nature of ateliers in Bruges and found the workers picked up casual contracts, so that scribe A did not always work with artist B, but rather, they operated in an ever-changing and fluid network.[61] This is relevant to both the ways in which the books were produced and consumed.

Just as multiple hands can be seen at work in the script, the same is true of the images, which might be built up out of many pieces of different origin, and shoved together to reflect local interests. Text and image might be adapted to each other as part of this process, in a feedback loop of production (where new images need accompanying added text and vice versa). Cambridge, UL, Ms. Ff.6.8 speaks to a form of production in which it was possible to construct highly decorated manuscripts out of component parts. Studying the codicological structure of this manuscript reveals another aspect of this process: the suffrages (quires III and IV) have been assembled entirely out of singletons: an image singleton, followed by a relevant prayer text copied onto a singleton, followed by another image page, another text page and so on. Each of these folios has a stub that has been folded down so that all of the loose pages can be joined up and sewn together. In effect, the Southern Netherlandish book makers were assembling many codicological units, some comprising just a single folio, to build the suffrage modules. Single leaves depicting saints in full-page miniatures were quickly becoming the norm. In order to ensure that these images accompanied their relevant suffrages, the scribes had to adopt the same method: copying a single suffrage on a single leaf. This must have resulted in a situation by which customers could simply choose from stacks of saints those they wanted to include. They “expressed themselves” in this act of shopping. English customers had agents in Bruges who aided in these transactions. Alternatively stationers must have also assembled books of hours speculatively for the English market. In this case, they could stick to safe choices: the most popular female saints, plus the English saints George and Thomas of Canterbury, and St. Christopher, the most popular protective saint. Regardless of who was making the final selection, this form of book construction results in blank parchment that early owners could then fill.

Its English owner(s) added several prayers to Ms. Ff.6.8, including the apotropaic incipit of the Book of John; a suffrage to St. Sebastian (142v), who was thought to ward off the plague; an incantation to the names of Yahweh (143r); and a prayer headed “Tetragramaton” that lists the 72 names of Yahweh interspersed with red Maltese crosses (143v). It is clear, therefore, that the English owner(s) used the blank space to turn the book of hours into an apotropaic object, which would provide personal protection. It is also clear from the dirt ground into the pages and the wear incurred on the miniatures that they used the manuscript as prescribed, as a daily guide to prayer, reading it habitually and keeping it close at hand. They employed the blank parchment as a vehicle to carry new texts, especially apotropaic ones, thereby multiplying the functions of their beloved imported book.


Analyzing the book structurally, noting which material is original and which is added, and considering the blank areas and the additions made post-production help to illuminate the stratigraphy of the book. Noticing these features reveals workshop practices and shows how book makers in cities such as Bruges and Delft adopted a modular system. Not every atelier in the Netherlands built manuscripts out of modular units. Those that did seemed to be tapping into a manuscript production method that would allow multiple scribes to work on a single manuscript; to produce manuscripts at varying levels of decoration; to include several standard texts, but allow for the expansion with other, less popular texts; and possibly to streamline production, thereby reducing its attendant costs. It also allowed the manuscript to be expanded later, when more components could be added at the beginning, at the end, and at the fissures.

Whereas book makers in Bruges worked primarily in secular workshops, those in Delft, Amsterdam, Haarlem, and other cities in the Northern Netherlands were largely organized in convents. In the discussion above I have used Bruges manuscripts as examples, because their extreme form of modular construction is easily graspable. Later, I will treat manuscripts from other centers, which were made according to similar principles.

One might be inclined to think that the maker of the manuscript, locked in his atelier, and the owner who later glues a scrap of parchment into that manuscript or adds a quire to the end, stand at opposite ends of the manuscript’s life. But in fact, the processes of production and alteration of manuscripts could be stutter-stepped, incremental, and intertwined. A manuscript might go through several stages of “making,” involving many hands—including those of its owner, who might intervene in this creation on multiple occasions—from initially having the book constructed via a modular method that suited, as much as was possible at that initial moment, his prayer needs. He or she might intervene repeatedly to keep that book as it had been at its initial creation—up to date. Circumstances, devotion, and motivations would change as a book aged with its various owners. Book production methods were, from the 1390s onward, fully prepared to accommodate this fact.

People who owned books—especially books of hours—in the fifteenth century, made all kinds of modifications to them. Some of these modifications resulted from a new production method—what I’m calling the modular method. Building the book from additive and modular parts paved the way for owners to add even more modules (some of them quite chunky), to fill in the otherwise wasted space with text, and to slip in images at quire divisions. While the idea of modifying books had been around for centuries, here for the first time the methods of book making anticipated post-production alterations.

In the following two parts, I analyze these types of modifications structurally, from the simplest to the most complex, using examples drawn from a millennium of book making. People have modified books as long as books have been made. Owners of books of hours in the fifteenth century—the century of the greatest rise in literacy, the largest manuscript production and expanded markets for books—was also the era of the most book modification. They borrowed simple techniques of augmentation that did not require rebinding, as well as more complicated techniques that did. All of these interventions, from the full rebinding to the work of the quire-filler, will be considered within the larger frame of the contemporary use of books. In short, I want to reveal not just what was done to these books, but also what motivated owners to intervene.


1    For late medieval psalters, see A. Bennett, “The Transformation of the Gothic Psalter in Thirteenth-Century France,” in The Illuminated Psalter: Studies in the Content, Purpose and Placement of Its Images, ed. F. O. Büttner (Turnhout: Brepols, 2004), pp. 211–21, as well as the other essays in this volume. See also studies about individual manuscripts: Jane Geddes, The St. Albans Psalter: A Book for Christina of Markyate (London: British Library, 2005); Kathryn Gerry, “Cult and Codex: Alexis, Christina and the St. Albans Psalter,” in Der Albani-Psalter. Stand und Perspektiven der Forschung / the St. Albans Psalter. Current Research and Perspectives, ed. Jochen Bepler and Christian Heitzmann, Hildesheimer Forschungen, Band 4 (Hildesheim, Zürich and New York: Georg Olms, 2013), pp. 69–95; Stella Panayotova and Andrew Morris, The Macclesfield Psalter: “A Window into the World of Late Medieval England” (Cambridge: Fitzwilliam Museum, 2005); Lucy Freeman Sandler, The Psalter of Robert De Lisle in the British Library (London and New York: Harvey Miller & Oxford University Press, 1983); Lucy Freeman Sandler, Illuminators & Patrons in Fourteenth-Century England: The Psalter & Hours of Humphrey De Bohun and the Manuscripts of the Bohun Family (London: The British Library, 2014); Anne Rudloff Stanton, “The Psalter of Isabelle, Queen of England 1308–1330: Isabelle as the Audience,” Word & Image: A Journal of Verbal/Visual Enquiry 18, no. 4 (2002), pp. 1–27; James H. Marrow, “Text and Image in Two Fifteenth-Century Dutch Psalters from Delft,” in Spiritualia Neerlandica: Opstellen voor Dr. Albert Ampe hem door vakgenoten en vrienden aangeboden uit waardering voor zijn wetenschappelijk werk (Antwerp: UFSIA-Ruusbroecgenootschap, 1990), pp. 341–52. For early books of hours, see Claire Donovan, The De Brailes Hours: Shaping the Book of Hours in Thirteenth-Century Oxford, Toronto Medieval Texts and Translations (Toronto and Buffalo: University of Toronto Press, 1991).

2    Studies of individual examples include Judith Oliver, “Reconstruction of a Liège Psalter-Hours,” The British Library Journal 5, no. 2 (1979), pp. 107–28; Alexa Sand, “Vision, Devotion, and Difficulty in the Psalter Hours ‘of Yolande of Soissons,’” The Art Bulletin 87, no. 1 (2005), pp. 6–23; Alexa Sand, “A Small Door: Recognizing Ruth in the Psalter-Hours ‘of Yolande of Soissons,’” Gesta 46, no. 1 (2007), pp. 19–40; Alexa Sand, “Cele Houre Memes: An Eccentric English Psalter-Hours in the Huntington Library,” Huntington Library Quarterly 75, no. 2 (2012), pp. 171–211; Richard A. Leson, “Heraldry and Identity in the Psalter-Hours of Jeanne of Flanders (Manchester, John Rylands Library, Ms Lat. 117),” Studies in Iconography 32 (2011), pp. 155–98.

3    Dominique Vanwijnsberghe, “The Cyclical Illustrations of the Little Hours of the Virgin in Pre-Eyckian Manuscripts,” in Flanders in a European Perspective: Manuscript Illumination around 1400 in Flanders and Abroad: Proceedings of the International Colloquium, Leuven, 7–10 September 1993, ed. Maurits Smeyers and Bert Cardon (Leuven: Peeters, 1995), pp. 285–96; Dominique Vanwijnsberghe, “Le Cycle de l’Enfance des Petites Heures de la Vierge dans les Livres d’Heures des Pays-Bas Méridionaux,” in Manuscripten en miniaturen: studies aangeboden aan Anne S. Korteweg bij haar afscheid van de Koninklijke Bibliotheek, ed. J. A. A. M. Biemans, et al., Bijdragen tot de Geschiedenis van de Nederlandse Boekhandel (Zutphen: Walburg Pers, 2007), pp. 355–65.

4    See Paul Henry Saenger, “Books of Hours and the Reading Habits of the Later Middle Ages,” in The Culture of Print: Power and the Uses of Print in Early Modern Europe, ed. Alain Boureau and Roger Chartier (Cambridge: Polity, 1989), pp. 141–73; Paul Henry Saenger, Space between Words: The Origins of Silent Reading (Stanford: Stanford University Press, 1997); Kathryn M. Rudy, “Dirty Books: Quantifying Patterns of Use in Medieval Manuscripts Using a Densitometer,” Journal of Historians of Netherlandish Art 2, no. 1 (2010).

5    Farquhar, “The Manuscript as a Book,” pp. 40–41, describes the modular method in brief (although he does not use this term). Farquhar’s essay also draws on Delaissé, “The Importance of Books of Hours for the History of the Medieval Book.” Reinburg, French Books of Hours: Making an Archive of Prayer, c. 1400–1600, pp. 22–26, discussing a similar situation in France, calls books of hours assembled from parts “shop copy manuscripts.” However, in examples she gives such as Baltimore, Walters Art Museum, Ms. W. 269, fol. 76r, the miniatures are integral with the text pages and therefore do not reveal a physical separation of copyists from illuminators and therefore do not exemplify the production methods I am outlining here.

6    Folio from a psalter, with a historiated initial showing Christ healing a blind man, Psalm 26. The Hague, Koninklijke Bibliotheek, Ms. 76 G 17, fol. 20v.

7    J. P. Gumbert, “Times and Places for Initials,” Quaerendo 39 (2009), pp. 1–24.

8    Nicholas Rogers, Books of Hours Produced in the Low Countries for the English Market in the Fifteenth Century (M. Litt. thesis, Cambridge University, 1982), p. 1.

9    For example, see LBL, Sloane Ms. 2683, made in Bruges, ca. 1390–1400, and a description in Scot McKendrick, Flemish Illuminated Manuscripts, 1400–1550 (London: British Library, 2003), fig. 2. Surprisingly little has been written about the “Pink Canopies” group of manuscripts. Consult: Maurits Smeyers, Naer Natueren Ghelike: Vlaamse Miniaturen voor Van Eyck (ca. 1350-ca. 1420) (Leuven: Davidsfonds, 1993), pp. 90–91; Susie Vertongen, “Herman Scheerre, the Beaufort Master and the Flemish Miniature Painting: A Reopened Debate,” in Flanders in a European Perspective: Manuscript Illumination around 1400 in Flanders and Abroad: Proceedings of the International Colloquium, Leuven, 7–10 September 1993, ed. Maurits Smeyers and Bert Cardon, Corpus van Verluchte Handschriften = Corpus of Illuminated Manuscripts (Leuven: Peeters, 1995), pp. 251–65.

10    Paul Binski, P. N. R. Zutshi, and Stella Panayotova, Western Illuminated Manuscripts: A Catalogue of the Collection in Cambridge University Library (Cambridge and New York: Cambridge University Press, 2011), pp. 343–44.

11    Opening in a book of hours, with a full-page miniature by the Masters of the Pink Canopies depicting the Presentation in the Temple, facing complines of the Hours of the Virgin. Cambridge, University Library, Ms. Ii.6.2, fol. 55v-56r.–00002/106 and–00002/107

12    Opening in a book of hours, with a full-page miniature by the Masters of the Pink Canopies depicting Christ as Man of Sorrows among the arma Christi, facing the Seven Penitential Psalms. Cambridge, University Library, Ms. Ii.6.2, fol. 75v-76r.–00002/144 and–00002/145

13    Opening in a book of hours, with a full-page miniature by the Masters of the Pink Canopies depicting the Mass for the Dead, facing the Vigil for the Dead. Cambridge, University Library, Ms. Ii.6.2, fol. 94v-95r.–00002/182 and–00002/183

14    Folio in a book of hours, with a full-page miniature by the Masters of the Pink Canopies depicting Christ as Man of Sorrows among the arma Christi, facing a prayer to be read before an image of the crucifix. London, British Library, Sloane Ms. 2683, fol. 65v.

15    Opening in a book of hours, with a full-page miniature by the Masters of the Pink Canopies depicting the Face of Christ, facing the prayer “Salve sancta facies.” Cambridge, University Library, Ms. Ii.6.2, fol. 12v-13r.–00002/30 and–00002/31

16    Libraries all over the British Isles hold books of hours made in the Southern Netherlands for export. Consult Nicholas Rogers, Books of Hours Produced in the Low Countries for the English Market in the Fifteenth Century; Nicholas Rogers, “Patrons and Purchasers: Evidence for the Original Owners of Books of Hours Produced in the Low Countries for the English Market,” in Als Ich Can: Liber Amicorum in Memory of Professor Dr. Maurits Smeyers, ed. Bert Cardon, et al., Corpus of Illuminated Manuscripts = Corpus van Verluchte Handschriften (Leuven: Peeters, 2002), pp. 1165–81; Saskia van Bergen, “The Production of Flemish Books of Hours for the English Market: Standardization and Workshop Practices.” In Manuscripts in Transition: Recycling Manuscripts, Texts and Images: Proceedings of the International Congres [Sic] Held in Brussels (5–9 November 2002), edited by Brigitte Dekeyzer and Jan van der Stock. Corpus of Illuminated Manuscripts, pp. 271–83 (Leuven: Peeters, 2005); Saskia van Bergen, De Meesters van Otto van Moerdrecht. Een onderzoek naar de stijl en iconografie van een groep miniaturisten, in relatie tot de productie van getijdenboeken in Brugge rond 1430 (PhD thesis, University of Amsterdam, 2007).

17    Articles that have addressed this topic directly rather than implicitly include Claire Donovan, “The Mise-en-Page of Early Books of Hours in England,” in Medieval Book Production: Assessing the Evidence; Oxford, July 1988, ed. Linda L. Brownrigg, Proceedings of the Conference of the Seminar in the History of the Book to 1500 (Los Altos Hills: Anderson-Lovelace, 1990), pp. 147–61; Michael T. Orr, “Hierarchies of Decoration in Early Fifteenth-Century English Books of Hours,” in Tributes to Kathleen L. Scott: English Medieval Manuscripts: Readers, Makers and Illuminators, ed. Marlene Villalobos Hennessy (London: Harvey Miller, 2009), pp. 171–95. Although it discusses the development of penwork initials primarily and only the hierarchy of decoration secondarily, a fundamental work remains Patricia Stirnemann, “Fils de la Vierge. L’initiale à Filigranes Parisiennes: 1140–1314,” Revue de l’Art 90 (1990), pp. 58–73.

18    Gumbert, “Times and Places for Initials,” pp. 1–24.

19    Ibid.

20    This trajectory maps onto the history of the car industry in the early twentieth century. Henry Ford’s great breakthrough was to make interchangeable parts and put them together on an assembly line. This made the cars much cheaper than previous models that were made one at a time. Efficiency demanded more standardized components and created more regular results.

21    Opening in a book of hours with two blank folios in a row, immediately before the Face of Christ. Cambridge, University Library, Ms. Ii.6.2, fol. 11v-12r.–00002/28 and–00002/29

22    Text folio from the Book of Kells, gospels, written and illuminated ca. 800, possibly on Iona. Dublin, Trinity College, Ms. 58, fol. 89r.

23    Opening in a book of hours, with a full-page miniature by the Masters of the Pink Canopies depicting the Throne of Mercy, facing a prayer to the Trinity. Cambridge, University Library, Ms. Ii.6.2, fol. 10v.–00002/26

24    Folio in a book of hours, the reverse of the Throne of Mercy miniature, with prayer texts added in an English hand of the fifteenth century. Cambridge, University Library, Ms. Ii.6.2, fol. 10r.–00002/25

25    Opening in a book of hours, with a full-page miniature by the Masters of the Pink Canopies depicting the personal angel battling a dragon, facing a suffrage to the personal angel, inscribed in Bruges on a singleton. Cambridge, University Library, Ms. Ii.6.2, fol. 14v-15r.–00002/34 and–00002/35

26    Folio in a book of hours, with two prayer texts, written in an English hand of the fifteenth century, added to the blank back of a miniature. Cambridge, University Library, Ms. Ii.6.2, fol. 14r.–00002/33

27    Folio in a book of hours, with birthdays added to the blank back of a miniature in an English sixteenth-century hand. Cambridge, University Library, Ms. Ii.6.2, fol. 33r.–00002/61

28    Folio in a book of hours, with a full-page miniature by the Masters of the Pink Canopies depicting the Annunciation, painted on a singleton. Cambridge, University Library, Ms. Ii.6.2, fol. 33v.–00002/62

29    Opening in a book of hours, with blank space filled with prayers in an English hand. Cambridge, University Library, Ms. Ii.6.2, fol. 15v-16r.–00002/36 and–00002/37

30    Masters of the Pink Canopies, St. George, full-page miniature painted on a singleton. Cambridge, University Library, Ms. Ii.6.2, fol. 22v.–00002/46 and–00002/47

31    Blank space in the book of hours, filled with prayers in an English hand. Cambridge, University Library, Ms. Ii.6.2, fol. 23v-24r.–00002/48 and–00002/49

32    End of the Hours of the Virgin (inscribed in Bruges), and prayer to St. Botolph (added in England). Cambridge, University Library, Ms. Ii.6.2, fol. 58v.–00002/112

33    I owe this observation to Peter Stallybrass. He discussed it in a lecture at the 6th Annual Lawrence J. Schoenberg Symposium on Manuscript Studies in the Digital Age in Philadelphia, November 21, 2013. Jan Gossaert, Portrait of a Merchant, ca. 1530. Washington, National Gallery of Art.–50722.html

34    Richard H. Rouse and Mary A. Rouse, “The Book Trade at the University of Paris, c. 1250-c. 1350,” in La Production du livre universitaire au moyen age: Exemplar et Pecia: Actes du Symposium tenu au Collegio San Bonaventura de Grottaferrata en mai 1983, ed. Louis J. Bataillon, Bertrand G. Guyot, and Richard H. Rouse (Paris: Editions du Centre national de la recherche scientifique, 1983), pp. 41–123; Christopher De Hamel, Glossed Books of the Bible and the Origins of the Paris Booktrade (Woodbridge: D. S. Brewer, 1984); J. J. G. Alexander, Medieval Illuminators and Their Methods of Work (New Haven: Yale University Press, 1992), pp. 22–23.

35    Caroline S. Hull, “Rylands Ms French 5: The Form and Function of a Medieval Bible Picture Book,” Bulletin of the John Rylands University Library of Manchester 77, no. 2 (1995), pp. 3–24.

36    J. P. J. Brandhorst, “The Hague, Koninklijke Bibliotheek Ms 76 F 5: A Psalter Fragment?,” Visual Resources 19 (2003), pp. 15–25. Opening from a “picture bible” or a prefatory cycle from a psalter. The Hague, Koninklijke Bibliotheek, Ms. 76 F 5, fol. 6v-7r. and

37    Compiling, the act of selecting ideas from existing texts and arranging them into a new text, is a literary activity which compounds the auctoritates of the authors cited. Compilatio, which emerged in the thirteenth century, resulted in a new genre of literature, including the encyclopedia, whose goal was usefulness for the reader. On compilatio, see M. B. Parkes, “The Influence of the Concepts of Ordinatio and Compilatio on the Development of the Book,” in Medieval Learning and Literature: Essays Presented to Richard William Hunt, ed. J. J. G. Alexander and M. T. Gibson (Oxford: Clarendon Press, 1976), pp. 115–41; Alastair J. Minnis, “Late-Medieval Discussions of Compilatio and the Role of the Compilator,” Beiträge zur Geschichte der deutschen Sprache und Literatur 101 (1979), pp. 385–421; Diane Mockridge, “The Order of the Texts in the Bodley 34 Manuscript: The Function of Repetition and Recall in a Manuscript Addressed to Nuns,” Essays in Medieval Studies 3 (1986), pp. 207–16; Neil Hathaway, “Compilatio: From Plagarism to Compiling,” Viator 20 (1989), pp. 19–44.

38    Rudy, Postcards on Parchment, pp. 77–79, with further references.

39    For the definition of a booklet, see Pamela R. Robinson, “The ‘Booklet:’ A Self-Contained Unit in Composite Manuscripts,” Codicologica: Towards a Science of Handwritten Books 3 (1980), pp. 46–67; and Ralph Hanna III, “Booklets in Medieval Manuscripts: Further Considerations,” Studies in Bibliography, 39 (1986), pp. 101–12.

40    Gumbert, “Codicological Units: Towards a Terminology for the Stratigraphy of the Non-Homogenous Codex,” Segno e testo 2 (2004), pp. 17–42, at p. 19. Erik Kwakkel uses the term “production units,” in his article: “Towards a Terminology for the Analysis of Composite Manuscripts,” Gazette du livre médiéval 41 (2002), pp. 12–19.

41    Marilena Maniaci, Terminologia del Libro Manoscritto (Rome and Milan: Istituto centrale per la patologia del libro; Editrice Bibliografica, 1996).

42    See O. Kresten and G. Prato, “Die Miniatur des Evangelisten Markus im Codex Purpureus Rossanensis: eine spätere Einfügung,” Römische historische Mitteilungen 27 (1985), pp. 381–403. William Loerke, “Incipits and Author Portraits in Greek Gospel Books: Some Observations,” in Byzantine East and Latin West: Two Worlds of Christendom in Middle Ages and Renaissance. Studies in Ecclesiastical and Cultural History, ed. Deno John Geanakoplos (New York: Harper & Row, 1966), pp. 377–81. Loerke tries to refute Kresten and Prato, but according to John Lowden he has not succeeded. See John Lowden, “The Beginnings of Biblical Illustration,” in Imaging the Early Medieval Bible, ed. John Williams, Penn State Series in the History of the Book (University Park: Pennsylvania State University Press, 1999), pp. 9–60, here: pp. 20–21.

43    For an interpretation of what happens when votaries add images to manuscripts that were not made in the “modular method,” see my study of Liège, UL, Ms. Wittert 35 in Rudy, Postcards on Parchment.

44    The manuscript is not in M. R. James, The Western Manuscripts in the Library of Trinity College, Cambridge: A Descriptive Catalogue, 4 vols. (Cambridge: Cambridge University Press, 1900).

45    Kwakkel, “Towards a Terminology for the Analysis of Composite Manuscripts,” pp. 12–19.

46    There are far more recognizable conventual scribes from German convents, and Cynthia Cyrus has identified an impressive number of them. See Cynthia J. Cyrus, The Scribes for Women’s Convents in Late Medieval Germany (Toronto: University of Toronto Press, 2009).

47    Whereas Cynthia Cyrus has systematically studied the German material, conventual scribes in Netherlandish convents have not been the subject of an analogous study. Karl Stooker and Theo Verbeij, Collecties op Orde: Middelnederlandse handschriften uit kloosters en semi-religieuze gemeenschappen in de Nederlanden, 2 vols, Miscellanea Neerlandica (Leuven: Peeters, 1997), have catalogued all known manuscripts from Netherlandish convents, both male and female, and have identified scribes where possible.

48    The great nineteenth-century bibliophile Willem Moll made an edition of the catalogue in 1857, which was enlarged and corrected in Willem Moll, “De Boekerij van het St. Barbara-Klooster te Delft, in de tweede helft der vijftiende eeuw: eene bijdrage tot de geschiedenis der middeleeuwsche letterkunde in Nederland,” Kerkhistorisch archief IV (1866), pp. 209–86 (24–28).

49    Ina Kok, Woodcuts in Incunabula Printed in the Low Countries, transl. by Cis van Heertum, 4 vols. (Leiden: Brill, 2013).

50    On convent sisters making manuscripts illustrated with pasted-in prints, see Ursula Weekes, Early Engravers and Their Public: The Master of the Berlin Passion and Manuscripts from Convents in the Rhine-Maas Region, ca. 1450–1500 (London: Harvey Miller, 2004).

51    According to the book’s colophon, five sisters together (sister Aef Dircsdochter, sister Jacob Gherijtsdochter, sister Aechte Claesdochter, sister Maria Martijnsdochter, and sister Maria Gherijtsdochter) copied Jan van Ruusbroec, Vanden gheesteliken tabernakel (HKB, Ms. 129 G 4) in 1460. Fol. 207r: “Ende is ghescreven uut minnen des heilighen daghes ende buten die tijt des ghemenen arbeyts totter eeren gods ende salicheit der sielen alle der gheenre diet lesen ende horen lesen. Int jaer ons heren m cccc ende lx. van suster Aef Dircs dochter. suster Jacob Gherijts dochter. suster Aechte Claes dochter. suster Maria Martijns dochter ende suster Maria Gherijts dochter. nonnen. haren gheminden susteren tot een testament ende een ewighe ghedenckenisse na haerre doot. welker namen ghescreven moeten wesen inder herten gods. Amen.”

52    Matins of the Hours of the Virgin. Cambridge, University Library, Ms. Ff.6.8, fol. 23v-24r.–00008/47 and–00008/48

53    Lauds of the Hours of the Virgin. Cambridge, University Library, Ms. Ff.6.8, fol. 28v.–00008/58

54    Virgin of the Sun, full-page miniature inserted into a book of hours before the Salve Regina. Cambridge, University Library, Ms. Ff.6.8, fol. 51v-52r.–00008/104 and–00008/105

55    End of the Hours of the Virgin (original scribe) and beginning of a text added in England. Cambridge, University Library, Ms. Ff.6.8, fol. 50r.–00008/101

56    Prayer added to a formerly blank area. Cambridge, University Library, Ms. Ff.6.8, fol. 50v.–00008/102

57    Added prayer to St. Ethelburga (below a prayer to St. Thomas of Canterbury, later crossed out). Cambridge, University Library, Ms. Ff.6.8, fol. 16v.–00008/34

58    Prayer text added to the formerly blank parchment. Cambridge, University Library, Ms. Ff.6.8, fol. 97v-98r.–00008/196 and–00008/197

59    Christ in Judgment, full-page miniature, formerly blank on the back, inserted before the Seven Penitential Psalms. Cambridge, University Library, Ms. Ff.6.8, fol. 98v-99r.–00008/198 and–00008/199

60    Seam between hands. Cambridge, University Library, Ms. Ff.6.8, fol. 11r.–00008/23

61    van Bergen, De Meesters van Otto van Moerdrecht.