The Making of Early Copies of the French Translation
of Boccaccio's De mulieribus claris
An illuminated manuscript is a nexus of various branches of medieval studies. This insight underlies many recent studies of late medieval manuscripts. Derek Pearsall has written: "...palaeography, codicology and manuscript studies generally are now more than ever before to be seen as an integral part of the study of literature and literary history, and vice versa. The fact of their interdependence becomes clearer and clearer as research advances, and the nature of that interdependence needs to be similarly clarified." Scholars have emphasized the attention paid by authors like Guillaume de Machaut and Christine de Pizan to the production of copies of their texts. The traditional disciplinary boundaries separating studies of the textual and pictorial contents have been broken down: literary and art historians alike have explored text-image relationships. With the advent of codicology, the interests of the art historian and palaeographer have begun to merge. Art historians have sought to examine the paintings in manuscripts not in isolation but as part of an integrated study of a book as a totality. The appreciation of a manuscript as a product of a carefully conceived plan and of the coordinated activities of different specialists has led historians of the medieval book to explore the social and economic contexts involved in the book industry.
These perspectives inform this present study which will examine the production of three early copies of a French translation of Boccaccio's De mulieribus claris. On September 12, 1401, an anonymous translator completed work on the French version , entitled Des femmes nobles et renommées. On New Year's day of 1403, Jacques Rapondi, a Lucchese merchant established in Paris, presented a copy of the translation to Philippe le Hardi, the duke of Burgundy. In February of 1404, Jean de la Barre, receveur general des finances in Languedoc and Guyenne, presented a copy of the same text to Jean de Berry. Both of these copies are currently preserved in the Bibliothèque nationale in Paris. On the basis of the Burgundian inventory of 1420, MS fr. 12420 can be identified as the copy given to Philippe, while the ex libris of the Duke of Berry written by his secretary, Jean Flamel, appears at the beginning of MS fr. 598 and the signature of Jean de Berry appears on folio 161. A third early copy of the text is in the Bibliothèque Royale in Brussels (MS 9509). Although this manuscript appears in inventories of the Burgundian collection, its date and original ownership are unclear. Meiss has dated the Brussels' manuscript to between 1410-1415 on the basis of his assessment of the style of the miniatures. Although the more limited miniature cycle of this manuscript distinguishes it from MS fr. 12420 and MS fr. 598, examination of the manuscript reveals its close association to the other copies.
These manuscripts are significant for both the intellectual and art historian. They are early testament to the French interest in the writings of "modern" authors, especially the works of Boccaccio. The heavy emphasis Boccaccio places on the mythology and history of the ancient world apparently appealed to the growing "humanist" interests within the intellectual circles associated with the French court at the beginning of the fifteenth-century. Bella Martens pointed to the art historical importance of the miniatures in these books in the developing naturalism of early fifteenth-century French manuscript illumination. She attributed the miniatures in MS fr. 12420 and MS fr. 598 along with a group of other manuscripts to an artist she dubbed the Master of 1402. More recent assessments by Millard Meiss and Patrick de Winter, while agreeing with Martens on the stylistic importance of the miniatures, have seen that each manuscript is dominated by a distinct artistic personality. With the exception of a few miniatures, Meiss has attributed the miniatures of MS fr. 12420 to an artist he has called the Coronation Master after a representation of the Coronation of the Virgin this artist added to a copy of the Golden Legend (Paris, Bibliothèque nationale, MS fr. 242). In turn, Meiss used the miniatures of MS fr. 598 as the eponymous manuscript on which to base the corpus of an artist he named the Master of Berry's Cleres Femmes. After earlier attributing both manuscripts to the same group of miniaturists, de Winter has subsequently come to agree with Meiss's attribution, although he identifies the principal artist of MS fr. 12420 after this manuscript as the Maître du livre Des Femmes nobles et renommées de Philippe le Hardi. De Winter has distinguished three distinct hands in the miniatures of MS fr. 598. Meiss has attributed the miniatures of MS 9509 to an artist related to his Master of Berry's Cleres Femmes.
The studies, to date, have focused primarily on examinations of the style and iconography of the miniatures. Little consideration has been given to other aspects of the making of these books, most notably to the copying of the text and the secondary decoration. These issues will be the focus of this paper, which will examine the discrete stages of the production of the books from the laying out of the page and the scribal work, through the secondary decoration, to the creation of the miniature cycles. What this examination will reveal is that the manuscripts are closely linked in each of these stages. The connections go beyond general similarities based on standardization of book making practices. We will see that some of the same scribes and decorators were active in each manuscript. Although I agree with Meiss and de Winter that the miniatures were created by distinct artisans, an assessment of the relationship of the miniatures to the text suggests that the miniaturists were not working independently but worked under the supervision of a planner. The implications of these connections between the manuscripts in their various stages of production suggest that they represent a veritable first edition of this new text.
THE SCRIBAL ACTIVITY
Before the scribes could have begun their work, the lay-out or mise-en-page needed to be defined. All three copies have nearly identical lay-outs, with a two column format and similar ruling patterns. In both MS fr. 12420 and MS fr. 598, there are 35 rows of text per column, while in MS 9509 there are only 33 rows. The measurements of the text blocks in the respective manuscripts although not identical are similar. In laying out the manuscript, it was decided to have allowed between 11 and 14 lines for a miniature and between 4 and 6 lines for an initial to introduce each new text. The type of script to be used also needed to be decided. All three manuscripts are written in the same script, gothica textualis.
The striking similarities in the lay-outs of the three manuscripts could reflect the standardization of manuscript making practices, and not necessarily a direct relationship in their production. Many of the similarities noted above can be found in a large number of other books produced in Paris at the same time. Similar lay-outs are found in copies of texts like the Bible historiale (e.g. Paris, Bibliothèque nationale, MS fr. 159), the Livre des propriétés des Choses (e.g. Brussels, Bibliothèque royale, MS 9094), and the Légende dorée (e.g. Paris, Bibliothèque nationale, MS fr. 242). The choice of this mise-en-page for the Boccaccio text was probably intended to place it in the context of the clerkly tradition of authoritative texts to which these other texts belong. The miniatures introducing the Dedication and Prologue show Boccaccio dressed as a cleric thereby attesting to the author's link to the tradition of Classical and Christian auctores.
Closer examination of the scripts of the manuscripts reveals the work of three distinct scribes. This identification is based on three different criteria: general aspect of the scripts, particular letter formations, and general scribal habits. The general aspect of the script of the scribe here designated as A is very regular and controlled with the letters being comparatively widely spaced. Among the distinguishing letter formations, the descender of this copyist's "g" generally forms a long, fluid curve going off to the left. The letter "y" generally has a faint, hair-line descender trailing down to the left, and this letter is characteristically dotted with a short stroke that ascends from the left to the right. As a scribal habit, Scribe A frequently fills the empty space at the end of a line of text with a faint vertical stroke crossed by horizontal dashes. The framing device for catchwords appears regularly and is virtually a signature for Scribe A's work.
In contrast to the work of Scribe A, Scribe B's script is smaller, more compact, and angular. The letter "y" used by this scribe can easily be distinguished from that of Scribe A by the hairline descender which loops upwards to the right and the stroke above which descends from the left to the right. This scribe's "g" has a descender which is frequently joined with the upper part of the letter and is placed off-center to the right. Scribe B fills out empty spaces at the end of lines with a pronounced minim stroke crossed by a dash and a dot beneath it.
The work of Scribe C has the appearance of being more fluid and more hastily written than that of the other scribes. The "y" has a hairline dash which ascends from the left to the right. This copyist's letter "s" is typically closed with the lower part being more pronounced. The "a" created by this scribe is characteristically closed. Period marks, especially if they appear at the end of a line, have pronounced hairline tails. Like Scribe A, Scribe C decorates his catchwords with distinctive decoration which is again almost a signature for this scribe's work.
The work of these three scribes can be found in more than one copy of the Boccaccio text. Scribe A and Scribe C worked on all three copies, while Scribe B limited activity to MS fr. 598 and MS 9509. Usually the scribes were responsible for the rubrics and running-titles on pages which they transcribed, but exceptions suggest the distinction in stages between the transcription of the text and the addition of rubrics and running-titles. In MS fr. 598, the rubrics and running-titles in Scribe B's portions were written by Scribe A, while in MS 9509, Scribe C transcribed the rubrics and running-titles for Scribe A's stint.
Closer examination of the division of labor between these scribes in relationship to the physical structure of the manuscripts gives important insights into the production of these books. Two types of scribal break are evident. The first type is represented by instances where there is a change in scribes in the middle of a gathering or even between a recto and verso of a leaf. Such breaks suggest consecutive copying with one scribe picking up where another left off. The second type is represented by instances where there is a break in scribes after a short gathering not containing the standard eight leaves, or where there is a break at the beginning of a gathering which introduces a new text. This type of break suggests the division of the text into separate scribal units which could have been distributed to the various copyists who could have worked concurrently on the different scribal units.
Diagramming these two types of scribal breaks gives a better picture of the production of these manuscripts. In the case of MS fr. 12420, Scribe A was responsible for the whole manuscript up to the recto of folio 127. Scribe C was responsible for the verso of this leaf and the remainder of the gathering. Significantly this break occurs in a gathering before a short quire of six leaves (fols. 130-135) completing a scribal unit. Scribe C was responsible for the entirety of the next scribal unit which extends from folio 136 to folio 167, the end of the manuscript. This suggests that Scribe C actually began activity with folio 136 and worked concurrently with Scribe A who was completing a stint, and then for some reason Scribe A suspended work with Scribe C completing the unit, in all probability after the last scribal unit had been completed .
In MS fr. 598, three distinct scribal units can be defined. Scribe A was again responsible for the beginning of the manuscript. Here this copyist's first stint was only the first 38 leaves. While Scribe A was working on the first scribal unit, Scribe B probably began the next unit extending from folio 39 to folio 135. For some reason Scribe B suspended activity on this unit on folio 86, and then likely began work on the last scribal unit extending from folios 136-162. Scribe A, picking up for Scribe B in the second scribal unit, was responsible for folios 87-112r, while Scribe C completed this scribal unit (fols. 112v-135). In MS 9509, there are only two scribal units. Scribe A and Scribe B collaborated on the first unit extending from folios 1-59, with Scribe A responsible for only the first 18 folios. While Scribes A and B were working on this first unit, Scribe C probably began the second unit, transcribing folios 60-131, while Scribe B reappears on folio 132 and completes the manuscript (fols. 132-165).
Dividing a text up into discrete scribal units with different scribes working concurrently expedited the production of a book. Comparing the length of the scribal units gives a rough estimate of the relative amount of time it took to transcribe the different copies of the text. For example, the amount of time it took to produce MS fr. 12420 corresponds to the length of the first textual unit, assuming, as argued above, that the first and second units were transcribed concurrently. The amount of time it took to produce MS fr. 598 is determined by the second unit. This again assumes that this unit was begun at the same time as the first unit, and that the third unit was written concurrently with the second. For MS 9509, the estimate is based on the second unit. Since the first unit in MS fr. 12420 is 135 leaves and the second unit in MS fr. 598 is composed of 97 leaves, it took 72% of the time to transcribe MS fr. 598 as opposed to MS fr. 12420. Comparably, MS 9509, which has a longest unit of 106 leaves, took roughly 79% as long to write as MS fr. 12420. Recent studies allow us to conjecture an approximate number of days it took to complete the scribal work on the respective manuscripts. Donal Byrne, in a study of a Sallust manuscript in Geneva, interpreted a set of roman numerals at the bottom of leaves to be records of the days it took to write the manuscript. He concluded that the scribe completed between 3 and 3.5 leaves per day. Since this manuscript is similar to the Boccaccio manuscripts in format and script, the scribal work of MS fr. 12420 could have taken between 38 and 45 days, that of MS fr. 598 between 28 and 32 days, while that of MS 9509 between 30 and 35 days.
THE SECONDARY DECORATION
The conclusions reached in the study of the scribal activity are closely paralleled in an examination of the secondary decoration. The decorative plan of the manuscripts, like the lay-out of the page, reflects the standardization of the Parisian book industry. Each new text is introduced by a large gold ground foliate initial. Along the margin adjacent to the large initial runs a bar staff with painted and gold rinceaux sprouting along its length and at the top and bottom of the page. Dentelle paragraph marks appear in the body of the text. This decorative plan can be found in a number of books produced in Paris during this period.
Closer examination reveals the significant relationships between the three copies of the text. I have demonstrated elsewhere the ability to isolate the contributions of individual decorators. Although unity and coherence of decorative plans were essential concerns for decorators, a division of labor can be articulated on the basis of identifying distinct repertoires of motifs and characteristic treatments of decorative components. Such an examination of the decoration of the Boccaccio manuscripts reveals the work of six distinct decorators. Of the decorators, the one here labelled as the A Decorator is clearly the finest. His work is distinguished by the delicacy of the treatment of the various decorative components, most apparent in the white highlighting of bodies of the gold ground foliate initials and of the dentelle paragraph marks and line endings. This delicacy is also apparent in the thin pen line used in the stems and tendrils of the penned line rinceaux and the outlines of the initials and of the painted rinceaux leaves.
The high quality of the A Decorator becomes immediately apparent by comparing the work by this craftsman to that of the B Decorator, whose white highlighting of the initials and of the dentelle paragraph marks and line endings is more simplified in the repertoire of decorative motifs. The painted lines lack the control and delicacy of those of the A Decorator. The pen line of the B Decorator is thicker and lacks the subtlety of that of the A Decorator. This can be seen in the more angular cusps created by the ink outlining of the initials, and the less fluid stems and tendrils of the penned line rinceaux.
The general appearance of the work of the C Decorator is finickier and more cramped (Fig. 15), most apparent in the pen line employed by this decorator. The stems of the penned line rinceaux are frequently broken up by curls and dashes, while the outlining of the painted rinceaux creates more pronounced cusps.
The tendrils projecting from the stems of the penned line rinceaux are signatures for the work of the D Decorator (Fig. 16). These are formed by a pronounced curl usually followed by a loose "S" shaped curve. This decorator frequently includes a dragon projecting from the top of the bar staff.
The general aspect of the E Decorator suggests a rapidity of execution, perhaps most apparent in the tendrils of the penned line rinceaux generally formed by a comma stroke followed by a quick "2" shape. Clusters of gold balls with the distinctive tendril decoration frequently appear in the borders associated with this hand. The E Decorator also creates denser clumps of painted rinceaux from which frequently projects a gold tongue shape.
The F Decorator typically adds trilobed painted leaves in a concave space created by the bar staff adjacent to initials. The decorator also increases the number of cusps in the ink outlining of the staffs and initials. The gold leaves of this decorator are also typically thinner than those of the other decorators. The concave space left on the right side of paragraph marks are distinctive of this decorator's contributions.
The following tables identify the contributions of these distinct decorators:
The decoration of MS fr. 12420 was the responsibility of the A, B, and E Decorators, while five decorators worked on MS fr. 598 ( A, B, C, D, and E Decorators) and two decorated MS 9509 (B and F Decorators). Only the B Decorator worked on all three copies of the text, while the A and E Decorators worked on both MS fr. 12420 and MS fr. 598. The three other decorators (C, D, and F Decorators) contributed to only one of the copies. The collaboration of some of the decorators in multiple copies of the text reaffirms the conclusion reached in the discussion of the scribal activities that the different manuscripts are linked in their process of production. Since there is no correlation between the work of specific scribes and decorators, there is no evidence to conclude that any of the scribes also acted as a decorator.
The division of labor in the decoration gives a better picture of the production of the books. In the majority of cases a single decorator was given responsibility for the decoration of a gathering. It seems, especially in the early stages of production, groups of gatherings were turned over to a decorator at a single time. For example, in MS fr. 598, the A Decorator was given the first three gatherings, while the B Decorator was given the next nine gatherings. The decorators could have worked concurrently on their stints. In the latter portions of MS fr. 598, the division of labor is more complex. With the gathering beginning on folio 95, decorators appear to have been given individual gatherings with the C, B, and D Decorators each being given a single gathering. The pattern in MS fr. 598 becomes more complicated with the gathering beginning on folio 119 where as many as three different decorators participated in the decoration of an individual gathering. Here the division of labor is by bifolio. It is noteworthy that in these latter portions of MS fr. 598 appear the only contributions of our C and D Decorators in the Boccaccio manuscripts. This suggests that the makers of MS fr. 598 were under pressure to complete the manuscript as soon as possible, necessitating the more complicated pattern of the division of labor and the appearance of these other decorators who were apparently called in to help speed up the process of production. These conclusions parallel those we have already reached in our discussion of the division of labor in the scribal activity.
In contrast to MS fr. 598, the division of labor of MS 9509 is much simpler. Here, only two decorators participated with each apparently being given groups of gatherings at a single time. The way the B and F Decorators leap frog through the manuscript suggests that they were working concurrently with each other and with the scribes. Groups of gatherings could be parcelled out to the decorators as the scribes completed their work on them. The decoration of MS fr. 12420 presents an intermediate case between the apparently hectic pace of MS fr. 598 and the leisurely one of MS 9509. The three decorators divided their labor by groups of gatherings in the early stages of production. It is only with the gathering beginning on folio 114 that a more complex pattern becomes apparent with decorators dividing the labor by bifolios.
THE MINIATURE CYCLES
As noted earlier, MS fr. 12420, MS fr. 598, and MS 9509 were dominated by distinct artistic personalities. In contrast to the other stages of production, there is no evidence of individual miniaturists working on more than a single copy of the Boccaccio text. Acknowledging this, the striking similarities in the selection and composition of scenes in corresponding miniatures in the various copies of the text needs to be explained. Instead of concentrating on the issue of miniature style, the focus here will be on a comparison of the different miniature cycles and their relationship to the texts they illustrate.
The relationship between text and image can be demonstrated by considering the illustrations associated with the story of the infamous Olympias, wife of Philip of Macedon and the mother of Alexander the Great (Figs. 1, 2, & 3). The miniatures show the crowned figure of Olympias, wearing a wimple and being supported by two women, while armed figures approach Olympias. This illustrates Boccaccio's account of how the jailed Olympias, aware of the approaching executioners, prepared herself and with the assistance of two servants stood fearlessly to face her death. The miniatures include a pile of bodies among which can be seen a crowned figure. This detail demonstrates how different moments in the narrative are combined. Earlier in the text, Boccaccio tells how Olympias, who had been spurned by her husband Philip for committing adultery, had urged Pausanias, her lover, to kill Philip. After Pausanias had been executed for the assasination, Olympias had his body placed over the remains of King Philip. The three miniatures, thus, visualize the story in the same way even to the point of conflating different episodes of the account. But comparison of corresponding figures in the different manuscripts suggests no evidence of dependence on a common visual model. The correspondences evident in the Olympias miniatures are characteristic of the relationship between the miniature cycles in the three copies.
Even when the text does not articulate a particular scene, the conception of the miniatures is the same. This is exemplified by the illustrations opening the text dedicated to Minerva. The text enumerates Minerva's various gifts: the art of wool (how it should be cleaned, softened with a comb, placed on a distaff and spun with the fingers, and woven); the pressing of olives for oil; the invention of the cart; how to make iron weapons; the use of armor; the strategy for soldiers and rules of battle; the use of numbers; and the creation of the flute. Each miniature shows Minerva seated on a throne with figures illustrating her various gifts encircling her. Comparing the list of gifts enumerated by the text to those illustrated reveals significant variations. None of the miniatures illustrate either the invention of the cart or the strategy and rules of battle. The miniatures conflate the creation of iron weapons and the use of armor by showing a man beating a helmet on an anvil, and all visualize the use of numbers by showing a money-changer at his counting desk. There are differences between the miniatures. In MS fr. 12420, Minerva is represented without a crown and halo, and the art of wool is represented by a single figure separating wool. In contrast, both MS fr. 598 and MS 9509 represent Minerva as crowned and haloed, and the art of wool is represented by a woman combing wool and a man weaving. These differences, however, do not override the striking similarities between the miniatures, and, as will be argued later, these variations possibly reflect a revision in the program of the illustrations.
Differences in the placement and poses of the figures in the corresponding miniatures argue against the use of a common visual model. The relationship between the different miniature cycles is best explained by their dependence on a common written model. It was a common practice to provide miniaturists with written directions for miniatures. In margins adjacent to miniatures in a number of manuscripts erased instructions can be found. We also have the famous case of the pictorial program that Jean Lebègue wrote for the illustration of Sallust's Conspiracy of Catiline and Jugurthine War. Another example is presented by what Gilbert Ouy has identified as a maquette or manuscript model for luxury copies of the text. This is a paper copy of Honoré Bouvet's Somnium super schismatis, where marginal notes provide instructions for the addition of miniatures.
Although in most cases the miniatures correspond well with the texts they illustrate, cases can be found where particular miniatures are inconsistent with the text and disagree with other versions of the pictorial cycle. An example of this is the miniature in MS fr. 598 accompanying the story of Polyxena. The text describes how the son of Achilles, Neoptolemus, revenges his father's death which had innocently been caused by Polyxena. The story recounts how Neoptolemus slays Polyxena by the tomb of Achilles. The miniatures in both MS fr. 12420 and MS fr. 598 show the son wielding a sword over the figure of Polyxena who kneels by a tomb. In MS fr. 598, however, instead of showing a closed tomb which would have been consistent with the text, an empty tomb is depicted with its lid ajar. Without knowledge of the text it would be natural to assume that the coffin was intended for Polyxena. We can imagine this type of assumption being made by a miniaturist who did not consult the text but followed instructions that called for Polyxena to be represented kneeling by a tomb.
The miniature of Veturia in MS fr. 12420 presents another instance of a miniaturist's probable misinterpretation of written instructions. The narrative tells how Veturia's son, Coriolanus, had been exiled from Rome and fled to the Volscians. In time, he became the commander of the Volscian army which he led in a war against his native Rome. After Volscian victories, the Romans sent three embassies to Coriolanus to sue for peace. In each case, Coriolanus rebuffed the Roman appeals. The Roman matrons turned to Veturia in hopes that she could persuade her son. Veturia, along with Coriolanus' wife and his children, went to the Volscian camp, and there made an impassioned appeal to Coriolanus. In general terms, all three miniatures visualize the confrontation between Veturia and Coriolanus in the same way. The miniatures are divided in two groups, one showing Coriolanus before his army and the other having Veturia as the principal figure. However, MS fr. 598 and MS 9509 are much closer to the details of the narrative. The text describes Veturia as being old and specifies that she was accompanied by Coriolanus' wife and his children. While MS fr. 598 and MS 9509 are consistent with the text in these details, MS fr. 12420 does not represent Veturia clearly as an older woman and her companions are not clearly specified as Coriolanus' wife and children. But most significantly, the miniature in MS fr. 12420 misinterprets the confrontation between Veturia and Coriolanus. It is Veturia's speech which is the central point of the account. In both MS fr. 598 and MS 9509, the gestures indicate that Coriolanus is listening to his mother's appeal. In these miniatures, Veturia uses a pointing gesture while Coriolanus' hands are raised with his palms placed outwards, the conventional gesture of someone listening. In the case of MS fr. 12420, however, the gestures are reversed. Coriolanus is shown ticking points off on his fingers, while Veturia has her hands raised as if listening to Coriolanus' speech. Anyone familiar with the text would not make this critical error. Again, a miniaturist misinterpreting written instructions or imprecise instructions explain such inaccuracies. An instance like this strongly argues against the possibility mentioned by others that the miniatures of the later copies of the text were based on the miniatures of MS fr. 12420. Again, the dependence on a common model, verbal as opposed to visual, is the most satisfactory explanation.
Although misinterpretations of verbal instructions can account for many of the inconsistencies between the different miniature cycles, there are instances where differences do occur between corresponding miniatures which can not be explained by miniaturists misrepresenting the instructions. This is exemplified by the miniatures in MS fr. 12420 and MS fr. 598 which accompany the story of Opis. In MS fr. 12420, Opis sits in a building with three men kneeling before her, while in MS fr. 598 Opis, seated in a temple, gestures to three statues placed on pedestals. Comparison to the text suggests that both miniatures are consistent. The text tells how an image of Opis was brought to Rome, and was placed in a noble temple. The Romans and Italians honored the image for many years in a variety of manners and ceremonies. This clearly explains the miniature in MS fr. 12420, while earlier in the text, it is noted that the fame of Opis is largely due to her rescue of her three sons, Jupiter, Neptune, and Pluto, and their father Saturn from Titan's plans to kill them. This linking of the veneration of Opis with that of her sons is represented in MS fr. 598 by having Opis gesture to the three statues, presumably representing Jupiter, Neptune, and Pluto.
MS fr. 598 image is not available
Another example of this type of disagreement between the miniature cycles is represented by the second to last miniature in the two Paris manuscripts. The story of Camiola tells how after a sea battle a certain Roland was held as a prisoner. While Roland's brother refused to pay the ransom, Camiola, a woman of great wealth, took pity on Roland languishing in prison. She realized to have him released she needed to marry him and pay the ransom. The text tells how Camiola secretly sent someone to Roland to seek his consent to her proposal. With his approval, Roland and Camiola were married by proxy through the mediation of a procurator with the pledging of a ring. Camiola immediately sent the ransom of two thousand ounces of silver to Roland to have him released. The text does not indicate that Camiola and Roland met until after they were married and the ransom had been paid. The miniature of MS fr. 12420 represents Roland inside the prison holding the hand of Camiola who stands outside while a servant offers the ransom to another man. Although disagreeing with the letter of the text which refers to Camiola and Roland being married by proxy, this miniature is still consistent with the spirit of the text with the joining of hands signifying the marriage while Roland was in prison. In the case of MS fr. 598, the miniature is much more consistent with the letter of the text. In this illustration Camiola is represented standing alone on the right side of the miniature while her servant, holding the ransom offers Camiola's proposal to Roland inside the prison.
Since the Opis and Camiola miniatures are consistent to at least the spirit if not the letter of the text, the variations can not be attributed to miniaturists misinterpreting verbal instructions. Also our knowledge of book making practices suggests that it would be extremely unlikely that the miniaturists made these revisions independently. These variations most probably reflect a revision in the pictorial program. One can imagine a planner, once MS fr. 12420 was completed, deciding to make minor revisions in the pictorial program. Being aware of the inconsistency between the text and miniature in the Camiola miniature in MS fr. 12420, he could have revised his instructions to produce the corresponding miniature in MS fr. 598. At the same time, the planner might have made minor revisions in other miniatures. The crown and halo could have been added to the Minerva miniature at the same time adding in the same miniature the woman combing wool and the man weaving. The type of revisions in miniature cycles found in the Boccaccio manuscripts parallels the variations Sandra Hindman has identified in her analysis of the early copies of Christine de Pizan's Epistre Othéa. As Hindman's study clearly reveals, miniaturists did not work independently, but worked under close supervision.
Before leaving the miniature cycles, I would like to introduce a possible example of how the illustrations enhanced a reader's understanding of the text. This is suggested by Christine de Pizan's use of these manuscripts. It is well known that Boccaccio's Des femmes noble et renommées served as a principal source for Christine's Livre de la Cité des Dames, and Hindman has already determined how Christine used miniatures in the Boccaccio manuscripts in the creation of the pictorial cycle for her Epistre Othéa. At least one example can be identified where a miniature in our manuscripts was probably used as a source in Christine's writings. This is represented by the miniatures illustrating the account of Sempronia which show her embracing a man while behind her are placed a variety of string instruments (Fig. 16). The text recounts the richness of Sempronia's intellectual accomplishments. It notes her mastery of Latin and Greek and her ability to write verses. She was also an accomplished singer and dancer, and knew how to play all instruments (...Elle savoit de tous instrumens jouer....). Boccaccio goes on to note how Sempronia used these abilities as instruments of sensuality and turned to wantonness. Christine includes an account of Sempronia in her Cité des dames. Her account closely follows Boccaccio's listing of the abilities of Sempronia except for a subtle, but I think a significant difference. Whereas Boccaccio states that Sempronia was accomplished in all instruments, Christine is more specific and states that Sempronia "...played all string instruments so skillfully that she won every contest" (emphasis added). The only source of which I am aware which specifies her mastery of string instruments as opposed to musical instruments in general is represented by the miniatures of our Boccaccio manuscripts. This suggests that Christine de Pizan did not base her borrowings simply on the text of Boccaccio, but she, as has already been demonstrated by Hindman, had access to one of the early copies of our French translation. Christine- perhaps used both the text and the miniatures as sources for the composition of her text.
All stages of production from the transcribing of the text, through the painting of the secondary decoration, to the creation of the miniature cycles are closely interconnected in the three earliest extant copies of the Boccaccio text. All the scribes and the majority of the decorators collaborated on multiple copies, and that, although different miniaturists were responsible for the pictorial cycles in the respective books, their work was apparently based on the same, albeit at times revised, written instructions. Such interconnection raises important questions about the nature of the Paris book industry. Undoubtedly the various specialists were working under common supervision. Can it be determined who was responsible for coordinating production? The question of the working relationship of the various makers also needs to be examined. Were they working within the context of a workshop, or were they working independently?
Patrick de Winter has speculated about the role the financier Jacques Rapondi might have played in the creation of these Boccaccio manuscripts. De Winter has gone as far as to say that, at least in the case of MS fr. 12420, Jacques Rapondi directed the project, and he has suggested that Rapondi provided Jean de la Barre with the copy of the text which the latter presented to Jean de Berry. Further study of the Burgundian accounts attest to Rapondi's frequent dealings with the book industry. Examination of the other extant manuscripts clearly associated with Rapondi do suggest interesting correspondences. For example, Scribe A (Fig. 17) and Decorators B (Fig. 17) and E (Fig. 18) contributed to the production of the Livre des propriétés des Choses (Brussels, Bibliothèque royale, MS 9094) which Jacques Rapondi sold to Philippe le Hardi in 1402, while the A Scribe (Figs. 19-21) and the A (Fig. 19), D (Fig. 20), and E (Fig. 21) Decorators appear to have collaborated on a copy of Hayton's Fleur des histoires de la terre d'Orient (Paris, Bibliothèque nationale, MS fr. 12201), known to be one of three copies Rapondi had completed for Philippe le Hardi in May of 1403. These instances of repeated collaboration support de Winter's conclusion that Jacques Rapondi played a central role in the production of these manuscripts, but I question whether Rapondi, considering his other activities, would have been directly responsible for coordinating production. He perhaps served as a financial backer for the project, and left the task of oversight to someone else.
The dealings of another financier Bureau de Dampmartin present possible parallels to the activities of Jacques Rapondi. Dampmartin served as the patron of Laurent de Premierfait, who, while living in Dampmartin's Paris "hotel", worked on a French translation of Boccaccio's Decameron. This translation was subsequently dedicated to Jean de Berry. The inventories of the Duke of Berry's collection indicates that Dampmartin was an important source for new additions to the library. François Avril has discussed the role Dampmartin played in the production of at least one of these acquisitions. The Berry inventory of 1413 refers to a "livre de Troye la grant" that Jean de Berry bought from Bureau de Dampmartin in April of 1402. Avril has identified this manuscript as MS fr. 301 in the Bibliothèque nationale which, along with another copy, was copied from a fourteenth-century Neapolitan manuscript of an Histoire ancienne jusqu'a Cesar now in the British Library (Royal 20.D.I). This London manuscript contains the following intriguing note on folio 8v: "Ci faut le secont cayer que maistre Renaut doit avoir, qui baillé à Perrin Remiet pour faire l'enluminure de l'autre cayer." Avril has convincingly identified the "maistre Renaut" as Regnault du Montet, an important Parisian "libraire" of the beginning of the fifteenth-century, and as Avril notes, Perrin Remiet was an active "enlumineur" in Paris at the end of the fourteenth and beginning of the fifteenth-century. As Avril has argued, in order to make copies of the London manuscript, it was unbound so that different portions could be distributed to the scribes and miniaturists involved in the project. Apparently, Regnault du Montet served as the coordinator of the production of the copies, and when the original manuscript was being rebound after the copies were completed, the second gathering was found to be missing. Unfortunately there is no clear documentary evidence about the relationship between Regnault du Montet and Bureau de Dampmartin in the production of BN fr. 301. It has been assumed that Dampmartin purchased the manuscript from Regnault du Montet, but considering the significant financial outlay involved in producing such a manuscript, the question should be raised whether Regnault du Montet acted independently. Did Montet carry out the project with the financial backing of Bureau de Dampmartin? The roles of merchants like Bureau de Dampmartin and Jacques Rapondi as middlemen between authors and/or manuscript makers and patrons deserves closer study.
If Jacques Rapondi served as the financial backer of the project, this leaves open the question of who took direct responsibility for supervising production. A possible candidate could be the anonymous author of the translation. Recent studies of the activities of Christine de Pizan have convincingly demonstrated the strong control she had over the production of copies of her work. Reno and Ouy have demonstrated that Christine frequently acted as a scribe in her own manuscripts, and Sandra Hindman's study of the Epistre Othéa has convincingly demonstrated the control Christine had over the creation of miniature cycles that illustrate her texts. The author of the translation would be the most logical person to have had the responsibility of devising and then revising the pictorial program. Study of various copies of Christine's texts shows that she regularly corrected and revised her work as different examples were produced. As Hindman has observed, the comparative importance of creating a miniature cycle is suggested by noting that when we can identify the creator of a miniature cycle it is usually a highly educated person. For example, Ouy, on the basis of paleographic evidence, has identified the creator of the maquette of Bouvet's Somnium super schismatis as Jean Gerson, the principal Parisian theologian of the early fifteenth-century. Likewise, the humanist and royal functionary, Jean Lebègue, devised the pictorial program for the work of Sallust.
Rouse and Rouse discussion of Raoulet d'orleans in the Vaudetar bible.
Another possible candidate could be one of the scribes. Documents reveal that scribes, or écrivains, regularly were given the responsibility of coordinating production. Here Scribe A is a possibility. Scribe A was responsible for the beginning portions of each copy of the Boccaccio text, and he also completed the rubrics and running-titles for Scribe B's stints in MS fr. 598. It is significant to recall that the work of Scribe A can be identified in the two other extant manuscripts which Jacques Rapondi provided for Philippe le Hardi (Brussels, Bibliothèque royale, MS 9094 and Paris, Bibliothèque nationale, MS fr. 12201). Since he was the sole scribe of MS fr. 12201 and was responsible for the opening portions of MS 9094, he was likely the principal scribe for these other projects. Scribe A could have been Jacques Rapondi's principal contact with the book industry. Having the responsibility of coordinating the work of the different specialists, Scribe A could also have worked with the anonymous translator, who is the likely candidate responsible for the creation and revision of the written guides for the miniature cycles.
Another question which needs to be explored is the relationship of the various makers of the Boccaccio manuscripts to each other. Were the artisans members of an équipe , or were they independent craftsmen hired on an ad hoc basis to make their contributions? The appearance of different miniature styles in the respective copies suggests that the miniaturists at least were independent, but what about the relationship of the scribes and decorators? The collaboration of a number of the makers in other projects for Jacques Rapondi has already been noted, and the work of these scribes and decorators can be found in a number of other manuscripts. Text Figure 4 charts out the appearance of these specialists. While there are examples of repeated collaboration, there is no consistent pattern of collaboration. The examples of repeated collaboration probably suggest more about the common supervision of different projects than the possible existence of a single workshop. For example, Scribe C and Decorators A and B along with other specialists shared in the responsibility of creating three copies of the Bible historiale (Brussels, Bibliothèque royale, MS 9024-9025; Paris, Arsenal, MS 5057-5058; Paris, Bibliothèque nationale, MS fr. 159). These correspondences probably indicate that these manuscripts were produced under the guidance of the same coordinator. Although a more conclusive answer about the relationship of the makers needs to await a more detailed examination of a larger group of manuscripts, the evidence to date suggests that the different specialists did not work in a shared workshop, but rather that they worked independently.
Other study has suggested that artisans did work independently and were called in on an ad hoc basis to contribute to particular projects. There is also little documentary evidence for the existence of large workshops that would employ the number of scribes and decorators involved in the production of the Boccaccio manuscripts. In fact, documents suggest that workshops were composed of members of a family with the addition of maybe a couple of assistants. It seems that the book industry depended on the close physical proximity of the shops of the various practitioners. It is well known that artisans involved in book production lived in the neighborhood near the university, especially on the streets around St. Severin. The Pont-du-Notre Dame and the rue Neuve de Notre Dame, adjacent to the cathedral, were also popular locations for workshops involved in the industry. The collaboration of Regnault du Montet and Perrin Remiet in the production of copies Histoire ancienne jusqu'a Cesar demonstrates how the centering of the industry in specific neighborhoods facilitated the participation of several shops in the making of individual manuscripts. Rent records for Paris from 1450-51 list as a previous owner of a house on the rue de la Parcheminerie a Pierre Remiot, perhaps identifiable as Perrin Remiet. Significantly this house was adjacent to one which had belonged to Regnault du Montet. Standardization of book making practices such as types of scripts and decoration also facililated the coordination of production shared by independent shops. The manuscript industry, thus, demanded a clear articulation and standardization of the discrete specialties. The success of a practitioner depended on the ability to conform to the standards of the trade and to work effectively with the network of interdependent shops.
With full awareness of the conjectural nature of the following conclusions, a hypothetical account of the process of production of the Boccaccio manuscripts can be proposed. Jacques Rapondi, being aware of the growing interests of members of the Valois family in such new texts, could have served as the patron of the anonymous translator just as Bureau Dampmartin was the patron of Laurent de Premierfait while the latter was producing the French translation of the Decameron. By giving such a deluxe manuscript to the Duke of Burgundy as a New Year's present, Rapondi was aware that he would not only receive a monetary gift in return, but as a token of his allegiance to Philippe le Hardi, such a gift would further his other commercial interests as well. This is implicit in the record of Philippe le Hardi's gift of 300 francs to Jacques Rapondi for not only the presentation of the Boccaccio manuscript but also for "...les bons services qu'il [Jacques Rapondi] lui faiz chascun jour et espere que face ou temps avenir...." De Winter has already suggested that the motivation of Jean de la Barre in giving a copy of the text to Jean de Berry was directly connected to his hope for a positive resolution of the legal difficulties he was facing.
In completing the translation, the anonymous author in all probability produced a maquette comparable to the one Gerson produced for the Bouvet text. This model, possibly written on paper, would have provided the basic lay-out of the text with spaces defined for the large initials and miniatures that introduce each account. In the maquette's margins adjacent to the spaces designed for miniatures could have appeared the guides for the pictorial program. The author could then have either parcelled out the work, or could have turned the project over to an écrivain, possibly our Scribe A, who would have had the responsibility of overseeing the work of the different specialists. In either case, the continued involvement of the author in the production of the different copies is suggested by the revisions we have observed in the pictorial cycles. The maquette seems to have been divided into separate scribal units which were distributed to various scribes who probably worked concurrently. When the writing of a group of gatherings was completed, they could then be distributed to the decorators. Once the decorators completed their work, the gatherings could then be given to the miniaturists along with the pictorial guides. These guides could have been in the form of the maquette created by Gerson, or they could have been on separate sheets like those composed by Jean Lebègue for the Sallust text. The separate stages of production could have been carried out concurrently so that while the scribes were working on the later portions of the text, the decorators and miniaturists could have already been at work on their contributions.
Although considering our current state of knowledge significant questions about the nature of the manuscript industry need to be left without conclusive answers, the type of integrated study of the different stages of book production presented in this paper can provide valuable clues to the resolution of these questions. This type of research appears to be particularly rich in its implications. For the literary historian, it can provide a better picture of the creation and production of new texts. The art historian can gain a better understanding of the nature of the creative process involved in painting a new pictorial cycle. The economic historian can gain insights into the workings of a major medieval industry. While the historian can learn much about the role patronage played in the social and political life of the period.