Note: This version of Robert of Sicily is represented by a 23-line fragment of linear verse on the fly-leaf of a collection of fourteenth and early fifteenth-century works, almost entirely in French, concerning the offices of the Earl Marshal and the laws of the admiralty. In the description below, the manuscript as a whole is taken up first, followed by details pertaining only to the fragment containing Robert of Sicily.

A. Dating of both the MS and fragment of Robert of Sicily in the early fifteenth-century is based on the handwriting and on a coat of arms [f. 6r] for the family Mowbray, Dukes of Norfolk and Earls Marshal. The manuscript is thought to have been written in the time of John Mowbray, Earl Marshal in 1405, Duke of Norfolk 1424, d. 1432. An English item added to the end of the manuscript, an Ordinance of the Duke of Clarence, is dated September 3, 1417.

B. With the exception of f. 1, a sixteenth-century table of contents in paper, the MS is vellum measuring 22.2cm. x 15.5cm. The Robert of Sicily fragment [f. 2] measures 21.4cm. x 15cm.

C. The manuscript is comprised of 61 folios bound so tightly that it is impossible to determine quiring. Of these leaves, 60 are vellum; 1 paper folio containing a list of contents is bound at the beginning of the MS. The Robert of Sicily fragment appears on what would have been the fly leaf of the original codex. Manuscript damage offers evidence that the paper folio was inserted prior to c. 1520 when the codex was bound in brown, leather-covered wooden board: several wormholes permeate the cover and most of the leaves. Wormholes scattered in the top and lower margins of the paper folio [f.1] correspond to those on succeeding leaves. In addition, f. 1 exhibits the same damage to its outer edges as much of the rest of the codex. The fact that it contains a table of contents in English in the same hand responsible for the English rubrication added throughout the volume raises the possibility that the insertion of f.1 corresponds to the date binding. The table of contents does not, however, mention the fragment of verse on the fly leaf, leaving the relative dating of Robert's appearance in question. The omission may perhaps be accounted for by the simple fact that the rubricator was not inclined to provide a title for an unfamiliar work.

The initial numbering in ink in the upper right hand corner of the leaves begins with the list of William the Conqueror's lineage [now f. 3] through f. 58. A final blank vellum leaf was left unnumbered, but again from evidence provided by wormholes, it is apparent that this leaf was included in the final quire. The fact that its ruling in red corresponds to that on f. 58 confirms this collation. New numbering in pencil begins on the recto of the paper Table of Contents so that the current foliation of contents that will be followed in this description extends from f. 1- f. 60, with f. 61 the blank, final leaf.

The codex proper appears to begin on f. 6r with the illuminated heraldic inital and elaborate border decoration. The first five folios represent a rather awkward beginning to what otherwise appears to be a carefully planned and executed volume. The irregular quality and scribal shift in the first item, the Norman lineage of William the Conqueror, suggests that the work was undertaken separately from the main project. Damage in the form of a large 6cm. x 8cm. stain and rust spots on f. 5v at the end of item 1, indicates that this gathering of a few leaves existed separately for a period of time. However, an imprint of the heraldic initial as well as some of the border decoration of f. 6r is clearly visible in the upper right hand corner of f. 5v.

The pricking, margins, and ruling of f. 5v also correspond to those of the carefully laid out f. 6r, the only difference being that the ruling on f. 5v is a pale brown rather than red. It would appear that f. 5v was either ruled as an afterthought to match f. 6r or that the ruling and decoration simply rubbed off after the two were gathered. Whatever the case, the main scribe takes over [f. 4v] in the middle of the Norman list. Whether or not these opening folios represent a rough or interrupted start--a form of quality control--to the overall project or an altogether separate undertaking, there appears to have been a lapse from the time of composition to compilation that accounts for the damage. For a fuller discussion of the implications of the manuscript's condition to its history see section (a) below.

The ultimate compilation of the codex with its current foliation is outlined as follows:

The final two items (see section G below) appear to have been added after the main manuscript was decorated. Signatures are not visible.

D. MS Appearance

(a) With the exception of localized staining, moderate damage to the edges, and wormholes throughout the cover and text, the manuscript is in good condition. Both front and back leather covers are stained and riddled with wormholes. This type of damage to the leaves and cover reflects problems related to damp storage conditions rather than to patterns of handling or use. Indeed, it is more likely that a manuscript of this type was considered too valuable to circulate beyond the confines of the library. Regardless of its relative value, the volume was stored under less than optimum conditions at some point in its history. BL Add. 34801 does, however, bear signs of damage that offer clues about its history.

A series of pale green stains corresponding in hue to the shade of green used in the border decoration beginning at f. 6r raise significant questions regarding the initial status of the preliminary leaves: the green staining covers ff. 1r and 2v, and is also discernable underneath the text of Robert of Sicily on f. 2r. The fact that the paper f. 1v escaped such staining is problematic as is the fact that the intervening leaves ff. 3 - 5 show no evidence of green staining. This suggests the possibility that f. 2 ( f. 2r now contains Robert of Sicily) might originally have been gathered with f. 6r--adjacent to its colorful decoration. If a paper bifolium were inserted for protection between the blank leaf [f. 2v] and decorated leaf [f. 6r]and later removed and inserted as f. 1, we could account for the disparity in staining on f. 1r and 1v. The existence of two such paper leaves appears possible in light of remnants of a paper fragment, corresponding in color and grain to that of f. 1, adhering to the vellum of f. 2v. Interestingly, the contents of this folio/fragment appear to be a rough draft of the table of contents on f. 1v. Taken together, this scenario indicates that ff. 3,4, and 5 containing the lists of Norman lineage were indeed added later, with the main scribe perhaps completing the list for the purpose of adding it to this volume.

There is one other bit of potentially significant damage in B.L. Addit. 34801: a large ink blot on f. 3r (Norman lineage) appears to match the ink of a blot on f. 60r, an ordinance of the Duke of Clarence, Grand Seneschal of England, concerning heralds' fees that is also comprised of a list of names. The scribe appears a bit self-conscious, striving for formality with a penchant for calligraphic flourishes and elaborate initials--not without a few pen-trials and blots, however. Considering the contents of the blotted folios, it is conceivable that the later scribe referred to the Norman lists as a cross reference, leaving his mark as he did so.

As might be expected, the flyleaf bearing the Robert of Sicily fragment has sustained a fair amount of damage. The vellum is darkened, bearing an almost collage-like pattern of soiling in addition to what appears to be the residue of another leaf adhering to this folio. A white film over the left side of f. 2r appears to be the result of a reagent.

(b) The leather cover of the codex was embossed with a panel and bands decorated with geometric patterns, floral designs, the figures of a dragon and gryphon. The binders' ciphers GW & IG place the date of binding at c. 1520. Remains of brass clasps are visible, with the back, uppper clasp missing.

The manuscript is notable for its illumination and decoration from ff. 6r - 58. A hearaldic seal measuring 3.5cm. x 4.5cm bears the arms of the Earl Marshal:

three gold leopards--lions passant guardant--on a background gules, or red, with an azure label couped with three points. It should be noted that this seal specifically represents the arms of the Earl Marshal, a title granted by Edward II in 1316 to his half brother, Thomas Brotherton, whom he had made the first Duke of Norfolk in 1312. The arms of the Duke of Norfolk include the Norfolk seal, a gold lion rampant on red background, quartered with the arms of the Earl Marshal. In 1366, Thomas Mowbray inherited the title through his wife. In 1394 Richard II made Mowbray Earl of Nottingham, and in 1397 Parliament recognized the Earl Marshalship of England as hereditary of his house. Thomas Mowbray's relationship with Richard II is significant in that he is reputed to have killed Gloucester; he was exiled and reported to have visited the holy land and to have died of the plague in Venice in 1399. His offspring with whom we are concerned include John Mowbray, second Duke of Nofolk, who was the probable patron of BL Addit. 34801 and John Mowbray, fourth Duke of Norfolk, whose dealings with the Paston family are of interest in the manuscript's history (see below).

The text itself is framed by gold borders intertwined with amethyst and blue decorative vines. The margins are filled with an English floral, leaf, and spray decorative motif in blue, amethyst, green, brown, and gold. This same pattern of decoration and colors is consistent throughout the French section of the MS. Smaller, decorated and illumined initials set off separate chapters or enumerated items. An enclosed and illuminated [Item] is employed in the Admiralty Laws. Illuminated initials are also accompanied by the same style of floral decoration although less dense than that found on f. 6r. Floral sprays surround even the more modest initials. When a new work is begun, as on ff. 19r, 28v, 36r, 40v, or 51r, the initial is larger, more elaborate, and accompanied by more extensive border decoration. Interestingly, the order of Oleron begins with little fanfare on f. 43r; it is the contents directly concerning the nobility that receive the artist's attention. The last two items in the MS are not decorated, but appear to be added to fill in the last quire.

Rubrication in English was carried out by a later scribe whose hand, consistently crossed out by yet another hand, also appears on the trial table of contents on f. 2v. He identifies the beginning of each new work with the English heading taken from the table of contents on f. 1v. Subsections of text are identified as Cap. 1, 2 etc., by numbering of items, or by marginal comments that serve as a reader's guides. Up to f. 15r, these comments are in French, after which they are offered in English. The comments all tend to deal with the issue of judgment at combat. However inconsistently applied, such rubrication appears responsive to the demands of the text or the needs of the reader.

(c) No fewer than eight hands appear to have been involved in the production or augmentation of BL Addit. 34801. Their respective stints are summarized below in the order they appear in the MS:

(d) The scope of this description does not permit full discussion of each of these hands. A detailed analysis of the main hand, for example, would be required to distinguish its mixed forms. Scribe 8, our "self-conscious scribe" reveals a bit of playfulness or perhaps naivete when he solemnly lists one "Launcelot de lake" in the list of nobility in an ordinance concerning heralds' fees. Regardless of the circumstances motivating the abbreviated, lineated verse scene from Robert of Sicily, the carefully executed Anglicana script suggests that for the scribe, it was a serious undertaking. For a discussion of the palaeographical features of this fragment please refer to the textual notes accompanying the transcription.

(e) With the exception of the preliminary leaves, the MS is pricked and ruled for a single column of 27 lines. Generous margins allow for elaborate decoration and measure 1.5cm top; 5.6cm. bottom and 4cm. outer. Folios 3-5 containing the Norman lists are not ruled (with the exception of f. 5v). Names are written in two columns, with the first scribe (scribe 5) fitting 23 names per column. When scribe 6 (the main scribe) takes over on f. 4v, he fits 29-34 names per column, as though concerned about running out of space. Ruling is visible on the first three lines of the Robert of Sicily fragment. Margins are not ruled, but with the exception of l. 11 which extends close to the edge, margins of .8cm. top, 1.5cm. left; 3.3cm. outer are maintained.

(f) With the exception of what appears to be a child's scribbling on the versos of ff. 1518 and ff. 23v, 25r, and 58v, marginalia are limited to a banner sketched in the upper right hand corner of f. 60r and a finger in bright red pointing to the text of the Office of Heraults, section 10, on f. 12v. An inscription in the top margin of f. 2v in dark brown ink in a hand different from others in the MS is only partially discernible: H_ _ _ _ Cril_ti_. Another name inscribed on the fly leaf at the end of the MS is no longer legible, but was deciphered in the Catalogue of Additions as "Anthony Widvale," Earl Rivers (1469-83). For a full discussion of the enormous significance of this inscription, see sections G and H below.

E. By virtue of its heraldic decoration, the provenance of the MS as a whole can be traced to Norfolk. The provenance of the Robert of Sicily fragment, however, is unknown and with so brief a sample, cannot be determined with any degree of certainty. One form does seem to exclude or at least limit the probability of a Norfolk provenance: [schall] appears five times in the fragment, lacks a distinguishing feature of the Norfolk dialect--[x] for [sch].

F. The fragment of Robert of Sicily is indeed out of its ususal context of religious miscellanies or mixed religious and secular materials. BL Addit. 34801 is purely secular, a manuscript of French texts whose few English items were apparently added later to fill out the volume. The nature of the volume's contents, its planned and modestly decorative presentation, and its rubrication by a succession of scribes suggest that it was primarily intended as a reference work, an historical document or archival record, for a somewhat limited, rather specialized audience. The items pertaining to the lists and lineage of England's noble families, the various titles of the Duke of Norfolk, the statutes of the Order of the Garter, the codes of chivalric duties and conduct, and the financial accounts of the heralds are precisely the kinds of documents one would expect to find in the library of the Earl Marshal. The manuscript offers an interesting mix of the ideal with the practical, the knightly pursuits of England's noble families being tempered by the attendant financial obligations. It is altogether likely that at some point the manuscript was read to evoke nostalgia for earlier times when royalty staged lavish tournaments for spectacle and entertainment. In the fifteenth century, such feats of arms were generally private productions, taking the form of judicial combats or duels of honor, but manuscripts of administrative accounts reveal that there was little difference in ceremony or staging between private or public productions--a matter of degree rather than substance.

G. Titles of the contents of BL Addit. 34801 listed below are taken from the MS Table of Contents as cited in the Catalogue of Additions: Where applicable other manuscripts containing the item are noted. The Addit. 34801 version of the admiralty laws has not been edited (Cat. Add. I: 90-92).

The contents of BL Addit. 34801 are significant in that they represent both the ideals of judicial combat embodied in the chivalric code and a related but distinctly practical application of precepts of honorable behavior in the commercial realm governed by the maritime code. The Roles of Oleron which form the basis of this international code evolved over a period of several centuries to facilitate commerce, particularly the wine trade. The impetus to put the codes into writing arose after the marriage of Henry Plantagenet and Eleanor of Aquitaine, with the code continually being revised from the thirteenth through the seventeenth centuries to accommodate changing circumstances. After studying many of the 37 extant manuscripts, Georges Peyronnet concludes that the code represents an exemplary achievement accomplished by experienced arbitrators: a clear articulation of all facets of the working relationship between employers and workers, in this case ship owners, merchants, captains, navigators, sailors, and laborers. Grounded in the practical realities of rapidly expanding markets and trade, the Roles d'Oleron contrast sharply with the nostalgic, allegorical nature of knightly tournaments.

Nevertheless, this unlikely mix of chivalric and commercial codes exists quite compatibly in The Black Book of the Admiralty. In Sir Travis Twiss's edition of this work, the French version is collated with its English counterpart in Lansdowne MS 285. Although the Black Book and the Lansdowne MS share several works with Addit. 34801, the versions reflect significant variation. Not surprisingly, Item 13, the ordinance of the Duke of Clarence, dated 1417 in Addit. 34801, lacks the opening address to Richard II that characterizes text of the Black Book (Anglo 189). Such discrepancies aside, the importance here is the link established between Addit. 34801 and the later (c. 1469) Lansdowne MS 285, which is not only reputed to be the "most valuable single source for the Tournament in fifteenth century England," but also "appears to be the very work, known as the Grete Boke, transcribed from various unspecified sources for Sir John Paston..." (Anglo 191).

The connection between Lansdowne MS 285 and Addit. 34801 is further strengthened by a series of letters, unique to the former, that describe the "courtly magnificence" of the feats of arms performed at Smithfield by Anthony Wydeville (or Woodville) Lord Scales, against the Bastard of Burgundy, Antoine Count of La Roche in June of 1467--two years after the challenge had been issued (Anglo 191). The following year (July, 1468)) Wydeville performed feats of arms at Bruges to celebrate the marriage of Edward IV's sister to the Duke of Burgundy (Charles the Bold) in a tournament so lavish, John Paston III likened it to Arthur's court. Wydeville must have been carrying on a family tradition: records and accounts attest to a similar display performed in Novermber of 1440 by a Sir Richard Wydeville and Peter de Vasques of Aragon (Anglo 186). On the basis of contents alone, Anthony Wydeville (also bearing the title Earl Rivers) appears a likely reader if not owner of the manuscript.

In the context of chivalric and maritime judicial codes, Robert of Sicily, especially the humiliation scene depicted in the fragment, appears almost comically out of place. Recalling the listing of "Launcelot de lake" among the English nobility listed in the ordinance of the Duke of Clarence [f. 60v], we might wonder if someone is having a bit of fun at our expense. There are, however, several thematic points where Robert of Sicily and its adopted context appear less incongruous. First, the chivalric contents appear most concerned with establishing the legitimacy not only of its patron the Duke of Norfolk, but the Plantagenet roots of English aristocracy itself--the purview of the Earls Marshal. In this regard, the tale of a proud prince humiliated would tend to put such worldly preoccupations in their proper perspective, a reminder of one's true state, of the allegiance owed the heavenly as opposed to the earthly king. Second, if we recall Robert's literary ancestry in Solomonic legend, the judicial context seems less foreign. It seems even less so when we recall also that while the Continental versions focus on the proud king's delinquency or neglect of judicial duties, the Middle English version casts King Robert as a type of Solomon--an "vmpeer" on each side, "without peer." Bearing both the title of King of Jerusalem and the renown of a medieval Solomon, the historical King Robert of Naples may well have evoked an association with judicial codes of this nature. Third, as we know, the Angevin King Robert ended his days in piety, choosing to be buried in the habit of a friar rather than the robes of a king. Perhaps the legends surrounding King Robert's wisdom and piety appeared also to reaffirm the caritas of the chivalric ideal. This is not so far-fetched as it may seem when we consider that records of fourteenth-century tournaments describe ceremonies in which knights process wearing masks or disguises: a chronicle from 1362 notes that the knights processed as the Seven Deadly Sins; another from 1394 chronicles the scene of knights entering the tournament attired as monks (Anglo 189n31). A fourth consideration is the matter of mistaken identity in which the remnants of Robert of Sicily were once again confused with the humiliation and penance of the romance's alter ego Robert the Devil. This tale of the demonic Robert, Duke of Normandy, provides some sort of link, however weak, to the tables of Norman lineage at the beginning of the manuscript. Tenuous as they might be, these thematic connections suggest only that Robert of Sicily may not be so out of its element as it might at first seem: the addition of the romance fragment may have been a considered rather than aimless choice. The association of Anthony Wydeville, Earl Rivers, with the manuscript (see below) tends to support the former.

H Nothing is known of the history of BL Addit. 34801 before it was purchased at Sotheby's July 2, 1895. Future studies concerning the volume's history should address the significant relationship between the Mowbray family, Dukes of Norfolk, and the Paston family of Norfolk, patrons and owners of the Grete Boke. reputed to be Lansdowne MS 285. This connection is established not only by the chivalric contents and references to Anthony Wydeville that Lansdowne MS 285 shares with Addit. 34801, but also by the evidence of direct communication between the 4th Duke of Norfolk and the Paston family during the years 1463 and 1465. Wydeville himself communicated with the Duke of Norfolk in a letter dated 1469. Together, the three family homes form a topographical triad capable of dominating East Anglian political and cultural affairs: the home of Wydeville, Earl Rivers, was in Middleton near King's Lynn in the upper northwestern corner of Norfolk; the Mowbray family, Dukes of Norfolk, claimed Framlingham in the lower southeastern corner of the county, in what is now Suffolk; and the Pastons were located twenty miles northeast of Norwich or near the coast (Davis xlii-xliii). Consumed as they were with the political turbulence of the late fifteenth century and constant challenges to one's property and position, it is a wonder any one of these families had energy for bibliographical pursuits. Yet we know that John Mowbray 2nd Duke of Norfolk most likely commissioned Addit. 34801 and presumably a good deal more and that John Paston II oversaw the production of a "grete boke" among many others. It is the third member of the triumvirate, Anthony Wydeville, Earl Rivers, who is the key to filling in the early history of Addit. 34801.

Three distinctive sets of historical evidence heighten both the probability and the significance of Anthony Wydeville's ownership of MS Addit. 34801: his reputation as bibliophile, translator and transcriber; his access to books through both friendships and royal affiliation; and the uncanny parallels of his life to that of King Robert of Naples.

His access to the volume would have been assured by his powerful position and the vicissitudes he endured during the War of the Roses. He was both victim and benefactor of the dynastic strife: in 1460 he and his father, fighting with Henry VI, had been captured in Calais by Warwick and the soon-to-be Edward IV. Warwick executed the elder Earl Rivers; a few years later the then Edward IV married Elizabeth Woodville, his daughter, and sister to Anthony Wydeville. Not surprisingly, Wydeville supported the Yorkists, and was made Lord Scales in 1461, Earl of Rivers in 1469, and protector of the Prince of Wales, Edward V. The Dukes of Norfolk supported the Yorkist cause and would likely have maintained communication, if not curried favor, with the highly placed Rivers. Norfolk's relationship with the inferior Pastons, however, was marred by a dispute over a property inheritance from John Fastolf, namely the Castle Caister, which Norfolk seized and held from 1469 to 1476 at which time John Paston II's claim was recognized (Davis xliv-xlvi, #334 546). John Paston III (brother of John II) who had served under Norfolk in 1462 fought on the Lancaster side in 1471. Wydeville appears to have maintained amicable ties with John Paston II throughout the turmoil. In a letter to her son John II after speaking with Lord Scales in Norwich, Margaret Paston remarks upon Wydeville's defense of the Pastons in the dispute: " Ye arn beholding to my lord of his good report of you in this country, for he reported better of you than I trow ye deserve. I felt by him that there hath be proffered him large proffers on your adversaries' part again you "(Davis #201 338-39). It is apparent that such loyalty arises from a longstanding friendship characterized by mutual interests and fostered perhaps by the manner of their ascendancy. At any rate, it is possible that in the early 1460s or the late 1470s before or after the Castle Caister dispute, BL Addit. 34801 was loaned to the Pastons for copying of or later collation with the chivalric content of the "grete boke." At any rate, Wydeville would likely have enjoyed access to the libraries of these parties, and hence to BL Addit. 34801.

When, in the final downturn of his fortune, Wydeville was executed by Richard III in 1483, he was said to be "the noblest and most accomplished of all Richard III's victims." What is notable is that he was also wearing a hair shirt, a symbol of the piety that preoccupied the latter years of his life, the period highlighted by his literary achievements. During this period he embarked upon several pilgrimages to Spain, notably to a number of holy places in southern Italy. Several circumstances and incidents combined to bring him to the attention of Sixtus IV who made him Defender of Papal causes in England. This preoccupation with piety, learning, and literature is reminiscent of the later years of King Robert of Naples with whose legend Wydeville was undoubtedly familiar. Considering the fickleness of his own fortune, the English earl also probably would have found much that was appealing and edifying in the tale of the fictional King Robert and may have acquired a copy of the romance for his collection. In a number of respects, then, Anthony Wydeville's interest in MS Addit. 34801 was at least as subjective as it was professional. Regardless of his means of acquiring either the manuscript or Robert of Sicily, the epilogue to his biography is the stuff of romance narrative: his niece Elizabeth was married to Henry VII, thus uniting the houses of Lancaster and York and ending decades of civil strife.

While these historical connections offer convincing evidence that at some time Wydeville was in possession of the manuscript and suggest possible motives and purpose for his interest, they do not shed light on his relationship, if any, to the Robert of Sicly fragment. It is quite one thing to establish why the romance itself may have been inserted in the volume and another to account for how it came to appear on the flyleaf in a fragmented state. Even more puzzling is the fact that the fragment is carefully rendered in a hand worthy of the volume as a whole. It appears a planned rather than a practice effort, but why the fragment begins and ends where it does seems inexplicable. Perhaps a household scribe, guilty of judging the book by its cover, took the codex to be damaged and, not understanding the French contents, proceeded to transcribe a section from a prose exemplar on the flyleaf. Practice copying from an available exemplar appears a more plausible explanation than a trial run at transposing from verse to prose, for the hand displays no evidence of the hesitation or self corrections we would expect in such a case. That is, of course, unless we consider the hand to be that of an accomplished translator the likes of Anthony Wydeville. But his role in this task is precluded by the dating of the hand as early fifteenth-century, well before Wydeville's literary period. Until other pieces of information fall into place, this version of Robert of Sicily will remain a mystery. Wydeville, Earl Rivers, died without heir, and his fellow bibliophile John Paston II had died several years before him. Thus, the volume may have fallen into hands less interested in its care and preservation.

I. For notices and descriptions regarding BL MS Addit. 34801, consult the following: