Shifting Landscapes: Remapping the Writing Traditions of Islamic Southeast Asia through Digitisation

Studies of the languages and literatures of the Malay world of Nusantara have long been shaped by the collections of manuscripts held in western institutions, which strongly reflect the interests of colonial officials who formed them. A very different picture of the writing traditions of maritime Southeast Asia emerges from a survey of manuscripts still held in local communities digitised through the Endangered Archives Programme and DREAMSEA. Primarily concerned with Islamic topics and often written in Arabic, the study of these newly-accessible collections has the potential to lead to a remapping of the intellectual landscape of the region.


'IN THEIR OWN IMAGE': WESTERN COLLECTIONS OF MALAY AND INDONESIAN MANUSCRIPTS 1
For the past century, studies of the languages, literatures, and writing traditions of the Malay world -the largely Islamic archipelagic lands of Southeast Asia, comprising present-day Indonesia, Malaysia, Singapore and Brunei, and the southern parts of Thailand and the Philippines -have been fundamentally shaped by the collections of manuscripts held in western institutions, as well as those in the region founded during the colonial era. These collections, mainly developed from the early 19th century onwards, strongly reflect the preoccupations of those who formed them, predominantly European scholars and government officials. Influenced by Enlightenment ideals and ideas, the collectors avidly sought out original literary, historical and legal compositions in vernacular languages, primarily in Malay, but also in other local languages such as Javanese, Balinese, Bugis/Makassar and Batak. Conversely, Enlightenment neo-classical sympathies and an ensuing nostalgic empathy for a vanished Hindu-Buddhist 'classical' era manifested in a lack of interest in and even antipathy to works on Islam. As a result, there are relatively few Islamic texts from Southeast Asia in western public collections. The older British Library collections inherited from the British Museum and India Office Library contain five Malay manuscripts of undangundang, legal codes rooted in customary practices, but not a single Malay text on fiqh, Muslim jurisprudence.
These public collections were usually reasonably well catalogued and open to scholars, and the advent of digitisation as a staple library activity over the past decade has further broadened access to a previously unimaginable extent. Ever increasing numbers of Malay and Indonesian manuscripts, which for centuries had been held in locations far from the lands of their origin and from the majority of their potential readers can now be accessed freely and fully on the internet, by anyone with the inclination and linguistic skills. The British the archipelago, namely Malay writings in Sri Lanka, and Cham manuscripts in Vietnam. These 23 EAP projects are listed below, arranged in geographical clusters; 15 are currently accessible online, with a further seven yet to be published (indicated below by asterisks).

Sumatra (8)
EAP229 Acehnese manuscripts in danger of extinction: identifying and preserving the private collections located in Pidie and Aceh Besar regencies

A NEW WAVE OF DIGITISATION: THE ENDANGERED ARCHIVES PROGRAMME
In 2004, the Endangered Archives Programme (henceforth EAP) was founded at the British Library, funded by Arcadia, a charitable fund of Lisbet Rausing and Peter Baldwin. The aim was to digitise archives -understood in a very broad sense as records of pre-industrial societies, in various media -in danger of destruction, neglect or physical deterioration. The archives remain with their owners, but one set of digital images is sent to the British Library to be published online through the EAP website (https://eap.bl.uk), while further sets of digital images remain in the country of origin. Since 2004, over 400 projects have been funded in more than 90 countries all over the world, from Bolivia to Siberia in Russia, yielding over 7 million digital images and 25,000 sound recordings.
Among these are 18 projects based in Indonesia, in locations ranging from Aceh, West Sumatra, Jambi, Kerinci, and Riau, and through Java, Bali and Lombok to Makasar, Ambon, and Bima. Collections digitised include mosque and madrasah libraries, as well as archives belonging to private individuals and descendants of royal families. In addition, five further projects have digitised Austronesian-language materials located outside  Even the most cursory survey of these collections digitised through EAP and DREAMSEA reveals that the profile of manuscripts still held in the field in island Southeast Asia paint a radically different picture of writing traditions and reading cultures from that gleaned from manuscript collections held in western institutions. First and foremost, the vast majority of the manuscripts are Islamic in content in the broadest sense, while the second important and related finding is that a large proportion of these manuscripts are in Arabic. Another highly significant aspect of these digitised collections is that generally they have been captured as complete libraries, often still held in the places where they were originally created and consumed, visually and aurally. This fact has manifold implications for the study of the literary landscape of maritime Southeast Asia, as it enables us to see what kinds of texts were written and read together in particular locations. Outlined below are some aspects of manuscript studies where our understanding of intellectual life in Islamic societies in Southeast Asia may be tranformed by these digitised collections. As the DREAMSEA collections have only recently begun to be published online, most of the discussion will focus on EAP collections, which have been available publicly long enough now to have generated scholarly comment and publications.

ISLAMIC WRITING TRADITIONS OF MARITIME SOUTHEAST ASIA
Probably over ninety percent of manuscripts digitised by EAP in Indonesia are Islamic in content, encompassing a wide variety of texts. These include copies of the Qur'an, with commentaries (tafsir) and guides on recitation (tajwid); hadith collections of prophetic traditions; texts on aspects of theology (kalam) such the Oneness of God (tauhid) and His attributes (sifat), and faith (aqidah); multiple works on fiqh, with practical guidance for the observance of Islamic law (ibadat); Sufism and mystical aspects of Islam (tasawuf); prayers and supplications (doa), and poems in praise of the prophet (kitab mawlid); sermons (khutbah); and manuals for the learning of Arabic and its grammar (nahu). There are also many works on astronomy and divination (ilmu falak, ketika, primbon) based on prayers or using Qur'anic surahs (faal Qur'an); amulets (azimat, jimat) and mantras for protection based on Arabic words and letters, and magical diagrams (rajah); and works on medicine including efficacious prayers (kitab tib). Literary works are most commonly hagiographical stories of the prophets and other Islamic figures. Letters and documents are also found in some private collections.
The contrasting profile between these overwhelmingly Islamic manuscripts held in local collections in Indonesia and the Malay world, and collections of mainly literary manuscripts from this region in western institutions, is clear. Over the past two centuries, there has evidently been a distortingly circular process by which western manuscript collections were shaped by the literary proclivities of European collectors. In turn, the study and publication of these same collections informed the concept of a Malay-world 'Canon' of authoritative texts, which was then perpetuated in pedagogical structures. Thus a significant number of the cornerstone texts of university courses in Malay and Indonesian literatures and philology are only represented by manuscripts held in European collections. For example, the only two extant copies of the earliest Malay chronicle, Hikayat Raja Pasai, are both held in London collections, while the 17th century historical work Hikayat Aceh is found in a single manuscript in Leiden. In short, there was a conflation of literary culture with the culture of literacy in the Malay world.
Reflecting this approach, key works which underpinned academic Malay studies programmes for much of the 20th century, such as Hooykaas's Over Maleische literatuur (1937) One of the most insightful views of Malay literary creations within the context of writing in the Malay world comes from Teuku Iskandar's Kesusasteraan klasik Melayu sepanjang abad (1995). The most pioneering aspect of Iskandar's work is the focus in the second half of the book on regional centres of activity, with sections on Aceh, Palembang, Riau, Brunei, Banjar and Patani, which each consider the whole range of writings from each location, including works in Arabic. However, even Iskandar's innovative sketch of literary activity in the Malay world is focused on original compositions by Nusantara writers, and does not refer to the 'supporting cast' of works mostly of foreign origin which would have contributed to shaping their worldviews. A very helpful theoretical perspective, positioning literary works within a broader intellectual framework, was outlined by Braginsky in his System of classical Malay literature (1993), drawing on work by Russian scholars of medieval literature: 'They point out that the concept of literature in the Middle Ages includes practically all written texts. The centre around which the texts functioning in society are grouped is the Canon of sacred texts. The latter determines the Weltanschauung of the culture concerned, the place of human beings in it, the teleology of their activites, and the fundamentals of ethics, aesthetics, and so on. The other texts are regarded as being related to the Canon and, depending on the proximity of this relation, form a number of concentric circles, the number of which within the literary system decreases centrifugally. Therefore, it is quite natural that the core of medieval literature was formed by texts belonging to functional genres … As for the literary, i.e. non-funcational genres in the modern sense of the word, these are more remote from the Canon and nearer to the folklore shared by popular and elite culture … Medieval literatures are almost invariably bilingual (or sometimes multilingual). Certain texts, generally relating to the central spheres of the literary system, are written in a supraethnic, or sacral, language of one or another religion. The texts which are closer to the periphery are written in the language of particular ethnic groups.' (Braginsky 1993: 14).
Braginsky outlined this theory to evaluate its applicability to Old Malay literature during the pre-Islamic medieval period, when the Canon of sacred texts would have been written in Sanskrit, and thereafter, as noted above, his own interest was firmly on the 'non-functional', aesthetically-informed, literary genres. Nevertheless, the profile of manuscript collections extant in Indonesia and the Malay world into the 21st century suggests the continuing relevance and value of the theoretical model outlined above after the coming of Islam, with its sacral language of Arabic. The writing cultures of Islamic Southeast Asia, as reflected in the collections digitised through EAP, continue to be dominated by 'functional' works on Islamic matters, often in Arabic, accompanied by a very small minority of literary texts, wherein which there is a greater likelihood of encountering the vernacular.
One typical example is the collection of manuscripts held in the Islamic boarding school or pesantren at Langitan in Widang, Tuban, East Java, digitised by Amiq Ahyad in 2006 (EAP061/1), consisting of 133 manuscripts. These include eight copies of the Tafsir Jalalayn, two works of hadith, and ten manuscripts on the learning of Arabic, while almost all of the other manuscripts can be categorised as either works on theology, Sufism or jurisprudence, including several volumes specifically on inheritance law. There is just one literary work: a manuscript of Cerito Ambiya, tales of the prophets. The latter is one of only two works in Javanese, all the other Langitan manuscripts being in Arabic.

THE ROLE OF ARABIC
In general, the study of manuscripts in Arabic, the sacral language of Islam, in Southeast Asia has suffered twofold neglect. On the one hand, scholars of Arabic worldwide have tended to ignore Southeast Asia, regarding it as a peripheral backwater barely worthy of mention. In  (1977, new edition 2014) covered all Austronesian languages including Cham and Tausug, but not Arabic. It is very hard to tease out and identify any Arabic manuscripts from Southeast Asia held in often much larger collections of Arabic manuscripts in western institutions, save in the Netherlands, where the handlist of Arabic manuscripts was fortunately compiled by maestro P. Voorhoeve (1980), who was able to identify visually and document manuscripts probably originating from Indonesia. 3 Nonetheless, preliminary investigations suggest that, apart from in the Netherlands, the numbers of Arabic manuscripts from the Malay world held in other European collections are not large: from over 20,000 Arabic manuscripts in the British Library, only around twenty originating from Southeast Asia have been identified.
When Arabic manuscripts relating to the Malay world have been studied, naturally the focus of attention has been either on influential texts written for the benefit of a Jawi audience by scholars in the Hijaz such as Muhammad b. Fadl Allah Burhanpuri (Johns 1965) or Ibrahim al-Kurani (Fathurahman 2012), or on works by notable Southeast Asian ulama such as Syaikh Yusuf (Lubis 1996). More recently, utilising manuscripts digitised in EAP212, Andrew Peacock (2018) has written on literary activity in Arabic at the court of Buton, highlighting compositions by Sultan Muhammad Idrus (r. 1821-1851) and uncovering evidence of intellectual and spiritual links with the Yemen and even the Maghreb.
While it is natural that the focus of scholarly attention will be on original compositions such as these, it is equally important to be aware of the thousands of copies of the 'everyday' Arabic texts through which Islam was studied, taught and transmitted across the archipelago. This vast corpus of manuscripts is like a towering mountain, perenially veiled in clouds, but which constitutes an omnipresent backdrop to intellectual life in maritime Southeast Asia. 'It is highly unlikely, for example, that in early seventeenth century Aceh the Tuhfa of Muhammad b. Fadl Allah would have been studied, and that Shams al-Din would have written on the Wahda al-Wujud (The unity of being) without there also being available work on tafsir, hadith, dogmatics, fiqh, and Arabic grammar' (Johns 1996: 43). This quotation is from Antony John's 'In the language of the Divine: the contribution of Arabic', strategically positioned at the start of Illuminations: Writing traditions of Indonesia, the illustrated survey of Indonesian manuscripts published by the Lontar Foundation in 1996. This chapter also revealed that the least-studied corpus of manuscripts from Indonesia in fact contains the most beautiful manuscripts ever produced in the region, with graphically-dazzling works of grammar and fiqh, exquisite kitab mawlid, and, at the pinnacle of achievement in the Islamic book arts, numerous sumptuous illuminated copies of the Qur'an.

MULTILINGUAL WORLDS
While the role of Arabic certainly needs foregrounding, it should also be acknowledged that at every court and in every madrasahpesantren, pondok, surau or dayah -Arabic was taught, studied and read in an intellectual environment where other languages were also used and written, such as Acehnese, Minangkabau, Lampung, Sundanese, Javanese, Bugis or Wolio, and in nearly all cases, also Malay. The grouping in libraries of digitised collections in EAP thus also leads on to a consideration of the multilingual context within which the culture of writing flourished in Nusantara. Many of the collections are not just bilingual, but multilingual, and also multiscriptural; for example, Javanese manuscripts may be written in Javanese, Arabic and Roman characters.
In the study of manuscript cultures there is an urgent need for an increased awareness of the value of integrated multilingual philological studies. The traditional management of institutional library collections has militated against this possibility through the emphasis on the separate cataloguing of different languages. For example, manuscripts originating from a single collection in Banten now held in Leiden University Library are listed in three different catalogues, of Arabic, Malay and Javanese manuscripts, veiling their common origin and history. 4 Thus another area ripe for study through EAP collections is an examination of manuscripts in different languages within a single library, to analyse the factors which determined which languages were used, and within which spheres, in different parts of Nusantara, and at different times.

GENRES OF TEXTS
A natural corollary of the difference in content of manuscripts in EAP collections and those held in western public institutions is that there are whole genres of texts which are only now becoming easily available for study. These include the core canonical texts, and an essential component of Islamic studies: the manuscripts needed to support the learning of Arabic. 5 Among the most common works used for these purposes in Southeast Asia were the texts popularly known as the Ajurumiyya 6 , the 'Awamīl 7 and the Alfiyya 8 , and various commentaries thereon. Manuscripts of these works and other Arabic grammars are documented in numerous EAP projects from almost every part of Indonesia, from Aceh to West Sumatra and East Java, yet not a single such manuscript from the Malay world is held in the physical collections of the British Library.
Another genre of texts all but unknown in European collections is sermons, khutbah, designed to be read at the Friday congregational prayer or on Id al-Adha, the feast of the Sacrifice, and Id al-Fitr, at the end of the fasting month of Ramadan . In early Islamic states, the mention of a ruler's name in the sermon was one of the two prerogatives of a Muslim sovereign (the other being the right to mint coins). Elizabeth Lambourn (2008Lambourn ( & 2011 has recently highlighted how the offer to cite a ruler's name in the Friday sermon could be used as a bargaining tool in negotiations between the great Islamic empires and the coastal communities that fringed the Indian Ocean: khutbah were traded for cannon. In the late 19th century, the citation of the Ottoman sultan's name in sermons was used by Malay rulers to support claims to Ottoman overlordship and thence entitlement to protection against western colonial powers. 9 The British Library holds only one, recently acquired, khutbah manuscript, probably from Lombok (Or 15924). However, these texts features in many EAP projects, including three volumes from Aceh (EAP329), five from West Sumatra (EAP144), and, most significantly, 45 from Ambon (EAP276). Intriguingly, just like the Lombok sermon in the British Library, many are in the form of paper scrolls, which is actually a very unusual format for manuscripts in the Malay world. Many of these sermons are relatively recent, with one dated 2002. However, one sermon from the collection of Sarajudin Hatuina, Ambon (EAP276/11/15), cites the names of sultans of Ternate dating from the 17th back to the late 15th centuries, suggesting the preservation of a much older tradition, and one which will repay further study.

A CODICOLOGICAL TREASURY
The majority of manuscripts from the Indonesian and Malay world do not contain complete colophons with information on the identity of their creator and/or patron, and the date and location of their creation. This is either because the manuscripts now only exist in fragmentary condition, with the losses most usually occurring at the beginning or end of the book where such information on provenance is generally found, or because this information was never originally added to the manuscript. Some manuscripts may contain colophons with only tantalizingly brief details such as the time and day of month of completion of writing, but without the year, name of scribe, or the place where the manuscript was written. Since Malay was the lingua franca of the whole archipelago, and Arabic the language of learning of all Muslim communities, this means it is often impossible to identify the precise origin of Indonesian manuscripts held in foreign or national institutional libraries and archives. 10 For most manuscripts digitised through the EAP, there is a high likelihood that many of the manuscripts would have been copied in the region itself, and it can at least usually be assumed that these texts were read and studied in that location. The availability of a substantial corpus of Malay and Indonesian manuscripts where we have some certainty about place of origin is of enormous significance for many aspects of codicological studies, such as palaeography. A prerequisite for palaeography, or the study of common styles of handwriting which can be associated with certain places and certain time periods, is the assembling of a group of manuscripts with reliable evidence of their geographical origin. 11 EAP collections are thus ideal sources of material for such a study, and provide enough material to enable the further investigation of proposed categories such as a posited 'Acehnese religious book hand' (Gallop 2015: 35), or styles of handwriting in West Sumatran manuscripts, and possible distinctions between styles of writing Arabic and Malay in a particular location.
Another field of codicological study enabled by EAP collections is decoration or illumination in manuscripts. While certain distinctive regional schools of manuscript art have been identified, there are still many parts of the archipelago where little is known about artistic preferences. It is thus extremely useful to have evidence of decorated manuscripts from certain locations, sometimes to function as yardsticks against which other illuminated manuscripts could be evaluated. For example, the copy of Cerito Ambiya from Pesantren Langitan noted above has a decorated frame on its first page with very unusual free-form red foliate motifs (EAP061/1/95). In the event of other examples of such illumination being encountered, the confirmed East Javanese provenance of the Langitan manuscript will help to contextualise these manuscripts. On the other hand, a survey of two substantial digitised collections from Bau-Bau (EAP221 and DS10) with hardly a single example of polychrome decoration, except in one mystical manuscript, suggests that decoration was not a standard consideration within the manuscript culture of Buton.

MARKS OF MANY HANDS
Manuscripts are, of course, read primarily for the texts they contain. But we undoubtedly get a much richer understanding of these texts when the very manuscripts that contain them also bear witness to the responses and methods of reading of their users over the years, in the form of decorations, annotations, comments and even grimy marks indicating the most-read pages. Pen trials and doodles may shed a light on the other types of texts that the scribes worked on, and it is very common to find on preliminary and end pages sketches of calligraphic letter headings (kepala surat) and conventional opening lines of letters (bahwa ini warkat al-ikhlas wa-tuhfat al-ajnas) reflecting the stock-in-trade of scribes. In the absence of detailed colophons, promissory notes and records of debts scribbled in manuscripts may contain valuable clues on location and date. More pertinently, comments by readers may elucidate attitudes to the texts from the communities within which they were produced. The value of these 'marks of many hands' are vividly traced by Mulaika Hijjas (2017), who compared literary Malay manuscripts and Islamic kitab literature in both Arabic and Malay, and found a greatly heightened intellectual engagment with the text in the kitab milieu. As a case study, she examined a digitised Islamic compendium on Sufism from a surau in West Sumatra (EAP205/3/1), and identified multiple types and layers of annotations, over several generations, the latest written in roman script and dated 25/2/1996. In contrast, a considerable proportion of all manuscripts in some western collections, especially those of literary texts, are new copies commissioned by a European patron. Such manuscripts are typically clean-margined and totally devoid of any annotations, and thus of hints as to the genealogy of the manuscript itself. One high-profile example is the famous Raffles Malay 18 manuscript held in the Royal Asiatic Society in London, the only known surviving witness to the earliest version of the Sejarah Melayu, believed to date from the 16th century and edited in 1612. But the manuscript itself was copied for Raffles in Bogor in 1814, on English paper watermarked '1812', and there is no indication at all of where the original exemplar came from, and how it came to be available in west Java in the early 19th century. Other manuscripts commissioned by Europeans may be copied on just one side of the paper at the request of the patron, leaving plenty of room for annotations for a reader not yet fully proficient in Jawi. Also in this category of Malay manuscripts which came into being through European intervention is the beautiful Taj al-Salatin (British Library, Or 13295) copied in Penang in 1824, at the behest of Ralph Rice, evidently with no expense spared, for his 'bibliomanist' brother Rev. Charles Rice. It is important to recognize that the illumination of this exceptionally fine manuscript, created to adorn the shelves of a bibliophile in Brighton, is atypical for the Malay world and not necessarily a guide to traditional Malay book art of that period.

MANUSCRIPTS AS MATERIAL OBJECTS
Another drawback of studying manuscripts in western institutional collections is invasive past practices in preservation and conservation, which, with hindsight, have caused irreparable damage to manuscripts as authentic material objects, and have led to the permanent loss of precious physical attributes. In the 19th and early 20th centuries many manuscripts entering the British Museum (and now held in the British Library) were rebound in standard European bindings, either in full leather or half or quarter leather bindings with buckram boards, and the original covers were simply discarded. Only in the very late 20th century was this destructive practice decisively abandoned in favour of a minimum intervention approach, which seeks to preserve as much as possible of the original codicological features of manuscripts, whilst still ensuring that they can be made available to readers safely. In the National Library of Indonesia (Perpusnas) nearly all the manuscripts were rebound in the early 20th century in plain buckram or cardboard bindings, again with the loss of the valuable traditional covers.
Despite sounding like a contradiction in terms, digitised EAP collections are treasuries of information on manuscripts as material objects, as the photographs usually capture the manuscripts in the form in which they are stored within the communities. Thus a range of materials used as manuscript covers can be seen, from leather to cardboard, paper and cloth, and digital photographs may also show patterns of stitching of quires, and evidence of traditional repairs to pages. The range of cloths used to wrap manuscripts or as endpapers in volumes is of great interest, and may range from local production such as batik (a fine blue calligraphic batik binding is found on a kitab mawlid manuscript in Ambon, EAP276/7/32) and songket, to imported Indian chintzes and checked cotton sarongs (kain pelikat). In an inspiring study, an examination of bindings of Ethiopian manuscripts revealed some of the earliest dateable examples of Indian trade textiles encountered in Africa (Pankhurst 1980), and similar studies of the use of textiles in the bindings of Indonesian manuscripts may also throw up interesting data on the distribution of trade cloths. Traditional methods of storage may also be captured in  these images, such as woven baskets and wooden chests. Of particular interest is the widespread use of bamboo containers for documents and sermons in the form of paper scrolls, and even for storing old collections of lontar palm leaves.

COLLECTIONS OF MANUSCRIPTS, PRINT AND ARTEFACTS
In most modern libraries, a fundamental distinction is made between manuscripts and printed books, with implications for storage, handling, and cataloguing. When a manuscript may have been aquired together with a printed book from a particular source, that relationship is almost always irrevocably lost once the items have been accessioned into major institutions. However, digitised collections are able to capture traditional libraries in their entirety, with all the 'books' owned contributing towards a rounded picture of the intellectual heritage of that community. For example, the private collection of Raja Fahrul of Riau (EAP153/8) comprises 19 items, all written in Malay with parts in Arabic. Five are manuscript volumes, on theology, prayers, and charms, while nine are printed books on prayers, tafsir, theology and ethical guides, with two volumes of stories of the prophets and one didactic poem for children; in addition there are five documents, mainly land grants and a power of attorney.
These collections also provide valuable insights into the extent of distribution networks for printed materials. In a collection in Sri Lanka (EAP609/30/7) is found a rare copy of the Malay version of the Thousand and One Nights, Hikayat alf layla wa-layla, lithographed in Singapore in 1878-9 (Proudfoot 1993: 121), and with the distinctive decorated frames indicative of Keasberry's Singapore Mission Press, with ownership inscriptions from Penang and Colombo in 1890. In Kerinci in the highlands of Jambi, alongside manuscripts are found printed talismans and colourful maps of the holy cities, both these types of objects evidently brought home by pilgrims returning from the hajj. In collections from all over the archipelago, from Kerinci to Ambon, can be found printed copies of the Qur'an and kitab mawlid lithographed in Bomby alongside locally-copied manuscripts of the Holy Book.
While there is a fundamental divide between the management of manuscript and print in most institutions, an even bigger chasm gapes between books, traditionally housed in libraries, and objects, stored in museums. But in many regions, a community's cultural heritage may consist of certain written materials stored alongside artefacts of equally charged and sacral value. This is most evident in the EAP117 project in Kerinci, which set out to digitise 'heirloom collections' or pusaka belonging to communities, generally stored in the house of the headman, and only accessed on certain ritual occasions and after requisite preparations. All of these collections included manuscripts and sometimes printed items, but the project made the enlightened decision to capture digitally the complete collections, including all material objects, giving an unprecedented view of manuscripts within their cultural and material context. Thus for example, the Depati Panggar Gumi Tuo collection, EAP117/37, consisted of one manuscript -a buffalo horn inscribed in Malay in ulu script -stored together with 25 ceramic bowls, five other other vessels, seven weapons, three ceremonial items, one item of clothing and two musical instruments, all of which have been photographed.

CUSTODIANS OF COLLECTIONS
Finally, through the project documentation, we are often given information on the custodians of the collections themselves, and their lineages, and information on how they came to hold these collections, whether through inheritance, or through acquisition. This type of information, and in so much detail, has rarely been available before to scholars of writing traditions of Southeast Asia, and is thus of exceptional value.
In one project from West Sumatra (EAP144/5), the project team even managed to track down a collection originally belonging to a Syaikh of the Syattariyyah order in Malalo, who lived ca. 1850, and was known as Sheikh Limo Puluah or Uwai (grandfather) Limo Puluah. After his death and the decline of the surau as a centre of teaching, his manuscripts were dispersed amongst his family and students, and were hardly used any more. The project team managed to locate 79 manuscripts from this original collection currently in the hands of nine different owners. The collection consisted of copies of the Qur'an, commentaries (tafsir), works on tauhid, fiqh and tasawuf, and stories of Islamic figures, mirroring closely the profile of the library in the Pesantren Langitan in East Java.

CONCLUSION
Collections of manuscripts held in local communities throughout maritime Southeast Asia, which have recently been digitised and published online through the Endangered Archives Programme, tend to be very different in scope and content from the manuscript collections traditionally held in western institutions. This article has attempted to outline some new and fruitful avenues for scholarly investigation offered by these digitised collections. The emerging picture of intense pedagogical activity and engagement with texts in many corners of the archipelago is very different from the generally disparaging and moribund picture painted in most western accounts of the 19th century and later. The founder of Arcadia, Lisbet Rausing, described her view of the EAP as 'a library of history still waiting to be written'. It is hoped that the further study of the countless manuscripts from Indonesia and the Malay world now accessible for study through the EAP, and also DREAMSEA, might contribute to the writing of one of these as yet unwritten histories: the history of the writing traditions of the Islamic world of island Southeast Asia, and thereby lead to a remapping of the intellectual landscape of the region.

ENDNOTES
1) This article is based on two blog posts written for the British Library's Asian and African studies blog (Gallop 2014a(Gallop , 2018a, and further developed in a paper presented at the International Symposium on 'Literary Culture and the Culture of Literacy in Indonesia', organised by the Faculty of Cultural Sciences, Gadjah Mada University, Yogyakarta, 25-26 September 2018 (Gallop 2018b). I would like to thank Prof. Dr. Sri Margana, convenor of the Symposium, for permission to rework the paper for publication in Humaniora. All images are reproduced courtesy of the Endangered Archive Programme, British Library. 2) Brocklemann 2017. Peacock (2018: 45) notes that Brocklemann was largely dependent on published catalogues of the Arabic manuscripts held in the library of the Bataviaasch Genootschap, now held in the National Library of Indonesia in Jakarta. 3) There is one small window when a considerable number of Arabic and Islamic manuscripts from Indonesia arrived in Europe: this was during the Dutch war on