One of the treasures of the State Library of Victoria is the Pilgrimage of the Lyfe of the Manhode, an English medieval manuscript produced in Lincolnshire in the mid-fifteenth century.1 Acquired in 1936 through the Felton Bequest for £600,2 the volume is comprised of two companion works, the Pilgrimage of the Lyfe of the Manhode and the Pilgrimage of the Sowle, bound together at an early date and the only known instance of this coupling.3 Signed in several places by the scribe, and also possibly illustrator, Benett, the Pilgrimage is in many ways a modest item. Written in English in a Lincolnshire dialect rather than Latin, on thick, inferior sheets of vellum and without the use of gold leaf or elaborate decoration, it stands in marked contrast to the sumptuous productions of the medieval monastery. The text is an adaptation in prose by an unknown translator of two well known works composed in French a century earlier, Le Pèlerinage de la Vie Humaine (1332) and Le Pèlerinage de l'Âme (1358) by the Cistercian monk and author Guillaume de Deguileville (born c.1294). A contemporary of Dante, Guillaume's allegories relate in the form of a dream the spiritual journey of a pilgrim, guided by a beautiful woman, Grace Dieu, as he partakes of the sacraments, confronts his own bestial nature in the form of the seven deadly sins and is received at the end of his journey into Paradise.4 Illustrated throughout with extraordinary allegorical scenes closely following Guillaume's text and often in lurid detail,5 the Pilgrimage is a unique item.6

Like Guillaume's pilgrim the manuscript itself has travelled its own long journey. When acquired in 1936, it was bound in a plain, undecorated late seventeenth-century binding of full calf leather over laminated pasteboards. The textblock had been resewn on five thongs of tawed skin and tanned leather wound together and with single folded paper endleaves. Fragments of dark blue thread at head and tail, indicating sewn endbands now lost, were visible. The textblock appeared fortunately to have survived untrimmed but no clear evidence of earlier or contemporary bindings could be seen.

It was brought to the attention of the Conservation Department in 1992 when it was proposed that a complete microfilm copy be made. The condition of the seventeenth-century binding was very poor: the sewing had broken down, the animal glue on the spine had become brittle, the textblock was loose and unstable and the spine leather friable with large pieces lost. The decision was made to disbind as the volume in its damaged state could be neither handled nor displayed safely. In 1992, the Pilgrimage was removed from its binding, disbound into loose quires and a complete microfilm copy made. The seventeenth-century cover and other loose fragments, including 'evidential dirt' from the gutters, were kept and boxed in line with the conservation practice of retaining all original material. An archival rare book box was also made to safely house the unbound manuscript.

Repairing the vellum leaves and rebinding the volume was discussed many times over the following years. With a range of possible approaches, the library's book conservators were for a while unsure of how to proceed. The disbound volume was shown on numerous occasions to visiting bookbinders and conservators for comments or advice.

The impetus to rebind the Pilgrimage came some years later and was two-fold. Firstly in 2006, Nicholas Pickwoad, an English book conservator and scholar, visited Melbourne for a Rare Book Summer School, presenting an exhaustive and minutely detailed overview of early European bookbinding. Dr Pickwoad examined the Pilgrimage with Conservation staff and made quite detailed recommendations as to how it could be rebound. This advice was invaluable. Secondly, a major exhibition of illuminated manuscripts from Cambridge University (as well as Australian and New Zealand collections) entitled The Medieval Imagination was planned for March 2008 at the State Library of Victoria. Exhibiting the rebound Pilgrimage at this important event became a priority.

A major treatment of this kind is undertaken only after a great deal of thought, research and preparation. A range of issues were carefully considered: the best approach to repairing the damaged vellum folds; humidification to flatten stiff, distorted leaves; the appropriate style of sewing and sewing supports; which of multiple holes to use; endleaf construction; the extent to which contemporary (that is, fifteenth-century) materials or techniques should be replicated; the physical functioning of the binding, its ease of opening and the action of the spine; the thickness and shape of the boards; the outward appearance of the volume, the covering material, the colour;7 the reversibility of the treatment and the ethical implications of this kind of major intervention. A series of experimental mock-ups was made to trial various techniques and to observe the behaviour of materials before work on the actual manuscript commenced.

A literature search was carried out into the treatment of medieval manuscripts. Journal articles describing related treatments (notably by Anthony G. Cains 'The Bindings of the Ellesmere Chaucer', Huntington Library Quarterly, Vol. 58, No. 1, 1996, pp. 127-157) were very useful. An examination of related items in the library's Rare Books Collection was also undertaken.

Of central importance was The Archaeology of Medieval Bookbinding by J. A. Szirmai published in 1999. Szirmai's work is an exhaustive, highly technical survey of early bookbinding covering a vast array of detail. Particularly useful in Szirmai's survey was a series of tables showing the frequency with which technical and stylistic details, such as board shape, lacing path, disposition pattern etc., occur in the volumes which have survived. Broken down further according to period and region, it was possible to piece together in some detail a picture of how the Pilgrimage was in all likelihood originally bound.

Conservation practice, however, is not merely concerned with historical exactitude and the crafting of historical facsimiles. The priority of conservation is to preserve into the future, in this case an artefact already over five hundred years old, rather than duplicate past practice for its own sake. Historical materials and techniques can be replicated, but not if function or longevity are compromised.

After much discussion, it was decided to rebind the manuscript following many of the suggestions made by Dr Pickwoad as well as much of the historical information compiled by Szirmai. It was decided to rebind in alum tawed goatskin, a common medieval covering material, over thick shaped boards laced to the textblock. It was also decided the binding would be non-adhesive (that is, with no glue applied to the vellum spine or any other part of the original) making the treatment completely reversible. The volume would be bound using stable archival materials and in a style sympathetic in appearance to English Gothic bindings. The decision was made to create a plain, white binding without tooling, titles or adornment in keeping with the modest character of the original. Metal clasps, a common feature on medieval books to hold the hygroscopic (reactive to moisture in the atmosphere) vellum textblock tightly closed when not in use, were considered and rejected as an unnecessary, even gaudy complication. A well fitting box would effectively perform that function.

A treatment of this kind raises its own special aesthetic and ethical issues. It was important to appreciate that a binding in an archaic style, using historical materials and techniques, is nevertheless still a new binding, lacking patina and 'character', with a stark quality, and that the meeting of the old and the new is inevitably something of a clash. It was decided that the manuscript would be best served by a restrained, unobtrusive binding which did not compete with or overwhelm the original, which was sympathetic in style but at the same time clearly distinguished between the medieval and the modern.

In this spirit, treatment began in March 2007 and was carried out as follows:

Firstly, remnants of animal glue were removed mechanically from the backs of the vellum bifolia, that is, using scalpels and erasers rather than solvents or moisture. Surface dirt was also carefully removed from the leaves. A detailed 'map' of the multiple sewing holes in the folded sheets was carefully made to record evidence of previous sewing. It was decided that humidification to flatten distorted leaves was unnecessary: the undulation in the leaves was minor and a natural part of the character and age of the item. Damaged folds were repaired with Japanese tissue and starch paste. Single loose leaves were guarded onto laminated Japanese tissue guards and wrapped around adjoining quires. These guards were trimmed to about 30mm and left visible in the sewn textblock as they indicate in most instances that a blank vellum leaf has been removed.

At this stage, prior to sewing, the opportunity was taken to create a high-resolution digital copy of the manuscript. To ensure safe handling during digitisation, the quires were secured individually with loose tacket sewing. Also, a non-destructive, X-Ray fluorescence analysis of the pigments was carried out which revealed the presence of azurite, vermilion and tin-lead yellow in the text and illustrations.

The repaired sections were then sewn on five double linen cords through single, existing sewing holes with 'packed' sewing, that is, adding several extra twists of thread per section around the cords to create solid, flexible bands across the spine. The text block was also sewn through a continuous Japanese tissue concertina guard providing an extra layer of protection for the backs of the sections. Double folded archival paper endleaves were attached and plain, white endbands wrapped around a thick linen core, with a back bead, were sewn on. These techniques were all trialled on the experimental mock-ups prior to work on the actual manuscript.

The decision was made to use 12mm thick boards created by laminating several pieces of dense archival bookbinding board rather than traditional wooden boards. It was decided that laminated board (with alternating grain direction) was a more stable and predictable material than wood, which can warp if insufficiently seasoned or split along the grain. Although the sewn textblock was irregular and asymmetrical, the boards were cut with true right angles creating squares of around 4mm extending beyond the block. Historically, bindings of this period had very small squares (often none at all) but it was decided that replicating historical practice in this instance would not provide sufficient protection for the irregular edges of the textblock. The boards were chamfered on four sides to create a shallow cushion board shape typical of English gothic binding as indicated by Szirmai8. Holes and channels for lacing were drilled and chiselled out: five primary holes for the double cords drilled at a 45° angle disposing inside the boards to four vertical holes with a VVVV lacing pattern also indicated by Szirmai as common practice in fifteenth-century English binding.9 Single holes and channels for lacing the endband cores were also drilled and chiselled out. Again, these technical details were trialled first on the mock-ups before working on the manuscript itself.

The boards were laced and drawn on, giving the textblock a gentle rounded shape and the cords adhered into the channels variously with cornflour paste and animal glue. The lacings were pegged with wedges of alum tawed skin, pasted out, drawn tightly through the holes and trimmed flush with the board. The holes were further filled in with a 'putty' made from linen cord fibres and paste to create an even surface for the skin. A loose, non-adhesive linen spine lining was attached to the insides of the boards covering the disposition channels and passing over the spine without adhesion.

The binding was then covered with full thickness, unpared, white alum tawed goatskin adhered to the boards with cornflour paste. Caps and tongue-mitred corners were formed and the damp non-adhesive spine tied up over the sewing cords to create pronounced raised bands. The goatskin turn-ins were trimmed and the insides of the boards filled with a thin card to create an even surface. And lastly, the endpapers were glued onto the boards with cornflour paste.

An archival phase box was made to safely house the rebound Pilgrimage. The treatment was completed on 12th October 2007.

A complete resewing and rebinding of this kind provides the opportunity to scrutinize a manuscript in very close detail. Intriguingly, it was observed that the first leaf of the Pilgrimage of the Sowle (f. 96r-v) is in fact a transcription of a previous first leaf now lost.10 In a different hand, noticeably not that of Benett, and in a paler brown ink (and with space left for a rubric which was never added), the text is written on one of the blank leaves at the end of the Pilgrimage of the Lyfe of the Manhode. One can conjecture that the first leaf of Sowle was spoiled or damaged at an early date and was copied, before being cut out and discarded, onto one of the blank sheets fortuitously located before it. The text of Manhode ends on the left hand page of the innermost bifolium of quire 10 (f. 95v), leaving five blank leaves of vellum. The transcription of the lost first leaf of Sowle was made on the fourth of these blank leaves and the remaining four leaves around it removed. Cutting out individual pages from folded sections of course creates loose adjoining sheets and it was necessary to guard these leaves onto Japanese tissue extensions in order to sew them into the textblock. All of this history is now apparent and can be 'read' from these guards which have been left visible in the sewn textblock to indicate the cut out pages. But there is an intriguing further element: the copy of the lost leaf is imperfect and its last four words on f. 96v appear to be repeated at the top of f. 97 in the hand of Benett. The repeated words have been struck through in red ink to correct the mistake. It is hard to see how a scribe working from the damaged sheet before him could create this duplication as one imagines the transcription would have been made faithfully line by line. Conjecture about a second exemplar (that is, the sheet from which the scribe works) is tempting but the issue remains inconclusive.

In 2007, over five and a half centuries after its creation, the Pilgrimage entered into the next phase of its long life. The short phase before had been spent unbound, boxed and largely inaccessible in the safe of the State Library Rare Books department for fifteen years. Prior to that, for three hundred years it survived in the seventeenth-century binding which had finally broken down by the end of the twentieth century. Of its life during the quarter millennium before that, virtually nothing is known other than the name of an early owner, Sir John Roucliffe of Cowthorpe, South Yorkshire (who died in 1531) written in a flamboyant hand at the bottom of folio 215v.11 Multiple holes in the vellum folds would indicate one or more previous sewings but beyond that, little can be stated with certainty.

But today, conserved in a new, flexible and well functioning binding (which can be easily and safely reversed by future custodians if desired) the manuscript is accessible again to readers and viewers into the future.