UDK: 005.8:726.2 (497.6)Stručni rad

Karin Scheper

University Library Leiden

balancing conservation decisions - preserving material evidence of original structures and repair techniques in a yemeni manuscripts digitization project


Because of the war in Yemen, that has devastated effects on the cities and villages, the population and their cultural heritage, the manuscript collections in Yemen are in peril, and inaccessible to researchers. Therefore, a digitisation project was initiated by the German Arabist Sabine Schmidtke, professor at the Institute for Advanced Study, Princeton: ‘The Zaydi Manuscript Tradition: A Digital Portal’. A number of European and American collections of Yemeni manuscripts will be made freely available on the internet. It concerns relatively small but important collections of Yemeni manuscripts, mostly collected in the late nineteenth and early twentieth centuries by scholars, diplomats and travellers. Leiden University Libraries (UBL) is one of the participating libraries. The collection of circa 150 Yemeni manuscripts is in the process of being digitized, and this work was preceded by a physical survey and conservation project. A previous assessment of Leiden’s collection of manuscripts in Arabic script had pointed out that Yemeni books have characteristics that help to distinguish them from other traditions of the Middle East. In addition, the manuscripts from Yemen are in particular dire condition. How would interventive conservation serve the project and safeguard this specific Yemeni binding tradition at the same time? How can historic repairs be integrated in conservation treatments?

Though the Yemeni manuscripts are the focus in this talk, the paper reflects on the effect of digitisation on the use and value of the physical manuscripts, and the complexity of conservation decision-making in general. As the texts become accessible 24/7, the originals become all the more important for critical material evidence. The manuscript production in Yemen, like in many other areas, is characterised by several regional-specific materials and techniques. Conservation dilemmas also concern the native repairs. The local interventions provide information on what materials and knowledge on the craft of bookbinding was available, how aesthetics mattered, and they underline the importance and value of the artefacts themselves. The impact of the materiality will be illustrated using examples from the Leiden collection ranging from the 14th to early 20th century.

Keywords: Conservation, Decision Making, Codicology, Regional-specific Features, Materiality, Local mends, Digitisation; Yemeni manuscripts.

Digitising manuscripts from Yemen

Over the centuries, the oriental collections in the University Library of Leiden (UBL), the oldest university of the Netherlands, have accumulated manuscripts from all parts of the Islamic world. The collection comprises of more than 6,000 manuscripts in Arabic script, and around a 160 of these originate from Yemen. These Yemeni manuscripts were selected for an international digitisation project, The Zaydi Manuscript Tradition (ZMT), a joint project of the Institute for Advanced Study, Princeton, and the Hill Museum & Manuscript Library (HMML) in Minnesota. The sad incentive for this initiative is the ongoing civil war in Yemen, now in its fifth year. In addition to the huge human cost of the conflict, the country’s rich cultural heritage has been affected to a large extent as well. Among the damages, there are bound to be libraries, and losses of manuscripts, yet it is impossible to exactly say what the situation is. What is certain though is that even if collections are relatively safe, they are not accessible as they were before. In addition, manuscripts scholars from Yemen will have difficulties or will not be able to travel and access manuscripts abroad. The project’s aim is to make available a large part of Yemeni manuscript collections that are kept elsewhere; as the project’s website describes it: ‘the tasks at hand are threefold, namely “preserving” and “studying” the Zaydi manuscript tradition, and “democratizing” the access to these materials’.1 As one of the European institutes with a substantial Yemeni collection, the UBL was asked to participate.

It is good practice to have any manuscript digitisation project preceded by a condition assessment, and to secure the involvement of a conservator with a deep understanding of the material objects, in order to advise on specific handling requirements and necessary treatments, to optimise the quality of the digitisation and minimize the risk of damage for the manuscripts during the whole process. In the case of this specific project, the assessment proved particularly fruitful, as it also helped to gain more information about the material characteristics and physical appearance of these manuscripts produced in Yemen. A fair number of them had been included in a large survey, focussing on the physical characteristics of the manuscripts in the Arabic collection, to study the development of the composite materials and technical structures.2 However, they had not been studied together as a group, or sequentially, as products from one specific region.3 The digitisation project now offered the opportunity to do so, and this indeed led to a further understanding of the developments of the bookbinding tradition in this part of the Islamic world, although some findings also pose further questions. Certain characteristics appear to form a trend, such as the regional-specific endbands that are either made without a leather core, in which case a saw-cut can be found to secure a tread that serves as an anchor for the tiedowns, or a leather core is used, but then the extending sides have been secured to the gutter of the textblock with thread, instead of cutting the sides off as with the traditional endband.4 The oldest manuscripts from Yemen bear witness to the use of the traditional techniques for endbanding, but the later manuscripts show a consistent use of these adapted methods. The frequent use of paper cutting techniques to decorate the leather bindings is also noteworthy.

As the dates of origin of these manuscripts span a few centuries, they provide an insight into the development of methods and use of materials. The volumes also carry many traces of use, and the various and multiple local mends form an interesting collection of historic repairs. So, while the digitisation project necessitated the physical examination of the manuscripts because of conservation concerns, the opportunity was used to study and record their material appearance and characteristics; it is clear that these activities are intertwined and fall in the domain of the book conservator. The assessment resulted in an increased understanding of the physical particularities of this group of manuscripts, which in turn allowed for a better informed decision making process, firstly for individual items but certainly also for the collection as a whole.

Preserving past practices

Digitisation is often denoted as a means to conserve books. What is actually meant is that the content of the books is safeguarded via a different medium, and in addition, that the original item can be relieved of extensive consultation since many users will be sufficiently served by using the digital version. This may of course prolongate the life of the original. Digital copies have a few important advantages; on the screen one can zoom in, or increase contrast, which may improve the readability of certain parts. Also, when digitised manuscripts are put online, they can be accessed outside of the opening hours of the library, and from anyplace in the world. Nevertheless, a digitised copy can never replace the original, and for certain types of research the original is the only valid object. Much of the material information, which may assist in dating the object and putting together its historical context, can only be retrieved when the physical manuscript is studied. Watermarks in the paper for example, or the precise make-up of the sewing structure, the thread that was used, the exact compilation of the gatherings, the thickness of the boards, the presence of wastepaper pasteboard or the structure of the endbands, such features require the examination of the object in the flesh. The accessibility of the originals therefore remains crucial, even when they have been digitised. And therefore, it is equally crucial that the object is preserved with all its material details and characteristics, so that it continues to tell its historical narrative. It may well be that this new role of the original manuscript and our growing awareness of its material value is partly a result of the technical, digital developments. One of the major consequences is that the conservation approach of book collections has changed significantly over the past years. It has become clear that, although damaged objects need to be stabilised, treatments should be as minimally invasive as possible.

Given the poor condition of many of the manuscripts from Yemen, this proved quite a challenge. What to do with the dust, the incrusted dirt? What to do with the book structures that had been meddled with, especially those repaired by seemingly unprofessional users who meant to keep the pages together but interfered with the traditional structure underneath? What to do with repairs of board attachments that covered annotations on the inside of the boards? Or with repairs that were no longer useful as they had broken down themselves? In trying to find a coherent reasoning for the conservation approach of the individual objects, it proved to be helpful to break down the conditional problems in different groups while firmly keeping in mind the purpose of the conservation campaign. When the longevity and usage of the manuscript would be clearly impacted by the damage, a treatment would be denoted as ‘necessary’. However, given the fact that the digitisation would relieve the object of extensive use, and the time restraints we faced for this project, certain damages could be considered as acceptable and not ‘in strict need of conservation’; a loose but intact textblock with a detached but stable binding would qualify as such. Other damages could compromise the quality of the digital image, such as incrusted dirt or folded edges and corners; associated treatments could also be denoted as ‘required’ or at least as ‘desirable’, unless such a treatment would involve an action that would cause the loss of historical information or disturb the integrity of the object in another way. Other damages, such as losses due to old insect infestation, do not always cause further problems or risks, and as a consequence, a treatment is not always needed.

It also appeared to be extremely useful to define the information value of the repairs. Several repairs were hardly remarkable as they resembled other repairs that are commonly found in Arabic manuscript collections, but certain types of repairs were noteworthy. They seemed to form a group that represented a specific way of resewing, or applying endbands, or lining and board attachment, that I have not encountered with manuscripts from other parts of the Islamic world, at least not this consistently, and therefore they provide very interesting information. There were also repairs, however, that appeared to be rather random, carried out in a singular, accidental way. Qualifying the historic interferences this way allowed for a categorisation in ‘meaningful’ and ‘unmeaningful’ repair methods and materials.5 The meaningful and informative ones were to be preserved, of course, unless they would cause further damage or hampered access of the book, in which case they would be documented and, if possible, be removed and kept together with the treatment report. Most noteworthy is the frequent use of a specific method of repair sewing, using side-stitching or a stabbed connection. Ironically, precisely this type of repair interferes with the accessibility of the manuscript. Since the gatherings no longer flex from the spine-fold position, but from the position of the stabbed sewing holes which are placed in the paper margin, sometimes close to the text, the opening of the textblock is restricted. In many cases, the original thread from the link-stitch sewing could still be found in the spine-fold. Why the manuscripts were not resewn according to this traditional method is unknown; had knowledge about this technique eroded? Was it considered too time-consuming, or not worth the effort? Was the stabbed sewing with its shortcomings considered to be sufficient and could the manuscripts still fulfil their purpose? As long as we don’t have the answers to these questions, the objects are our primary sources for study as they bear witness to this practice. In those cases when the stabbed sewing restricted the opening of the manuscripts to an extreme extent, I cut the thread and secured the cohesion of the gatherings with a spine-lining (when most gatherings were still connected with the original sewing thread, of the unsupported link-stitch sewing); this way, the practice of stabbing remained clearly visible. In only a few cases was the manuscript resewn again, using the old sewing positions and the link-stitch sewing structure. After all, the extant stabbed holes of the previous repair structure is evidence to the use of this technique.

The uninformative repairs, such as pressure sensitive tapes and certain sewn repairs, could be replaced by more functional conservation materials. One manuscript, a Qur’an, had been lined with a poor quality machine-made paper and synthetic white glue, after a course stabbed sewing was applied, and in the process of ‘repair’ half of the illuminated opening pages had been displaced at the end of the textblock, upside-down and without taking care of creases and tension in the paper. The detached, damaged binding was re-attached with the same synthetic adhesive, but reversed, so that the board on the right side now extended in the envelope flap. It is hard to believe that this clumsy, if not careless repair was carried out by someone who understood the meaning of the manuscript, or still wanted to use it. As the binding was still severely damaged and the first and last folios of the textblock needed consolidation as well, it was decided to reverse all these changes to the textblock, structure and binding.

In a few other instances, the repairs were not be directly linkable to a regional practice, yet they display an attitude of respect towards the book, and a level of attention and perhaps frugality, which at least illustrates the relation between people and object. Such historic interventions were incorporated in the conservation treatment as much as possible.

In conclusion

Conservators often work with tight schedules, because projects are seldomly planned in such a way that the conservator is actually allowed to undertake all the work that she/he considers necessary. Quite the contrary, the development of many projects such as digitisation or exhibitions are often well under way, including the outline of a budget and time schedule, before the conservation requirements are taken into account. One of the consequences is that choices have to be made; prioritising is essential. Certain items will be selected for treatment while others will have to do as they are, or choices are even made within one object: the paper repair needs to be address for example, while the damage to the binding is considered fairly stable and will remain unattended. This may raise questions with future users, and this is one of the reasons why a treatment report is of value. It will explain the choices that were made.

Though many issues discussed in the above may relate to other manuscript holdings, it is not possible to create a decision-making tree that will be applicable for all digitisation projects of book collections or every individual manuscript, as different projects will have their own requirements and restrictions, and each collection needs its individual context to be taken into account. However, the ideas that informed the decision making process as it was developed for this particular project, may have relevance in other situations. I have tried to indicate which considerations and concerns we believed to be relevant. It is hoped that, as an example of a careful approach, this case-study may be of use for future digitisation projects elsewhere.

Projekat digitalizacije jemenskih rukopisa i uravnoteživanje odluka o konzervaciji – očuvanje materijalnog dokaza originalnih struktura i tehnike reparacije


Zbog rata u Jemenu koji ima devastirajući efekat na sela i gradove, na stanovništvo i njihovo kulturno nasljeđe, zbirke rukopisa u Jemenu su ugrožene i nedostupne istraživačima. Stoga je njemačka arabistica Sabine Schmidtke, profesorica na Institutu naprednih studija u Prinstonu, pokrenula projekat digitalizacije pod nazivom „Zaidska tradicija rukopisa: digitalni portal“. Izvjestan broj evropskih i američkih zbirki jemenskih rukopisa će tako postati dostupan na internetu. Radi se o relativno malim ali značajnim zbirkama rukopisa koje su uglavnom prikupljene krajem devetnaestog i početkom dvadesetog vijeka od strane učenjaka, diplomata i putnika. Univerzitetska biblioteka iz Leidena (UBL) učestvuje u ovom projektu. Trenutno je oko 150 jemenskih rukopisa u procesu digitalizacije. Ovom radu je prethodio projekat fizičkog pregleda i konzervacije. Prethodne procjene Leidenske zbirke arapskih rukopisa su ukazale na to da se jemenske knjige razlikuju od drugih tradicija Srednjeg istoka. Pored toga, rukopisi iz Jemena su u izuzetno lošem stanju. Kako bi interventna konzervacija poslužila projektu i istovremeno sačuvala specifičnu jemensku knjigovezačku tradiciju? Kako se historijska reparacija može integ­rirati u procese konzervacije?

Iako su jemenski rukopisi u fokusu ovog izlaganja, rad se osvrće na učinak digitalizacije na korištenje i vrijednost originalnih rukopisa kao i na složenost odlučivanja o konzervaciji generalno. Istovremeno sa pristupačnosti tekstova na portalu 24/7, originali postaju još važniji kao kritični materijalni dokaz. Proizvodnju rukopisa u Jemenu, kao i u mnogim drugim područjima, karakteriše nekoliko regionalno specifičnih materijala i tehnika. Dileme u pogledu konzervacije također se odnose na izvornu reparaciju. Lokalne intervencije nam pružaju informacije o tome koje znanje i koje knjigovezačke vještine su bile primjenjivane, koliko je estetika bila bitna, ukazujući na značaj i vrijednost samih artefakata. Značaj materijala će se ilustrirati primjerima iz lajdenske zbirke koja obuhvata period od 14. do početka 20. vijeka.

Ključne riječi: konzervacija, odlučivanje, kodikologija, regionalne specifične karakteristike, materijalnost, lokalne restauracije, digitalizacija.

1 https://www.ias.edu/digital-scholarship/zaydi_manuscript_tradition, see also https://www.ias.edu/idea/2017/schmidtke-zaydi-manuscript-tradition. (accessed 04-10-2019).

2 K. Scheper, “The technique of Islamic bookbinding. Methods, materials and regional varieties”. Brill, 2015.

3 Also, according to the selection criteria for the larger survey, volumes with extensive damage or heavily repaired items, which prohibited the examination of the structure were excluded from the survey; in fact, many of the manuscripts from Yemen fell into thedeselected category’. These manuscript could, however, still retain original details that reflect regional-specific techniques, but along the lines of the above mentioned study they could not be recorded.

4 Ive elaborated on this issue in K. Scheper, “The technique of Islamic bookbinding. Methods, materials and regional varieties”. Revised second edition, Brill, 2018, see pp. 301-304.

5 The repair in itself was obviously meaningful as it served a purpose, but the method does not represent a local tradition or practice, and holds in that sense no codicological information. The materials with which such repairs were carried out have some informative value, of course; for example, pressure sensitive tapes did not really come into use before the 1930s.