Before Typographic Man

If printing was known about in some Indian circles in the seventeenth and early eighteenth centuries, then how do we explain its failure to spread at this time? With technology and personnel present on Indian soil as early as the sixteenth century, there is no obvious reason why Europeans should not have been recruited to serve Indians as printers in the same way that they served them as cannon-makers.Footnote ¹ In the 1670s this is precisely what did happen in the case of one entrepreneur, Bhimji Parekh, and his hired English hand, Henry Hills, but their experiment did not last and is not known to have been imitated during the next century.Footnote ² Before 1800, the argument that the spread of printing was delayed by the hold of bazaar copyists over the book market, and the relative cheapness of their product compared with the initial capital required to set up a printing press and sell a large enough number of copies to turn the enterprise to profit, holds some merit.Footnote ³ This, after all, seems to have been the reason for the abandonment of Parekh's Devanagari Press. When indigenous printing did eventually develop in Indian and other Islamic settings, in economic terms the key (if long unrecognized) enabling factor was the invention in 1800 of the mass produced iron hand press.Footnote ⁴ Even so, the late development of printing in India was not only a question of the mechanical and economic aspects of the supply of presses. It also related to the cultural aspects of the demand for books and through this to the uses of books during the period of ‘print absence’. Assessing the factors shaping the demand for books requires us to ask what work books did in early modern India.Footnote ⁵ If Christopher Bayly has characterized the later eighteenth century as a period in which more things were being written down in India than before, then it is worth turning to the early part of that century to explore the possible background to this development.Footnote ⁶ In partial answer to this question, this paper addresses the relationship between books and the social capital of knowledge in the later Mughal era.

The following pages argue that among many of the Muslim religious circles which constituted some of the most important potential markets for books in early modern India, in practical as well as in conceptual terms, knowledge was located primarily in persons rather than in books. Books were not considered independent sources of knowledge, but were appendages to the personal pedagogical relationships through which knowledge was transferred and within which writing served to provide only one dimension of the knowledge being transferred. Even where books were used for religious learning, this occurred under the personal instruction of a master who closely directed his students’ reading and placed it within larger non-textual programmes of acquiring knowledge through prayer, dreaming, ascesisFootnote ⁷ and the service rendered as an apprentice. Books worked in the service of an anthropocentric mode of knowledge, as mnemonic aidsFootnote ⁸ and adjuncts to the bodily incorporation of words in the person of the authoritative master and through him to his students. Correspondingly, those in search of knowledge looked for a master rather than a bookshop or library, with libraries in any case generally being private collections accessed through personal affiliation to a teacher.

In the literary circles under discussion, the social capital of knowledge was transferred through pedagogical relationships within which books were used but independent of which their perusal had little formal status. Even in the book-intensive circles of legal study, ijāza certificates were not testaments of the simple reading of books, but of having gained the understanding of them that could only come through having their ‘correct’ meanings explained by a master. Among the Sufi circles that included ‘ulamā as well as other professional groups, the epistemological value of books was lesser still, even as they produced and used them as part of larger learning programmes. In an epistemological tradition in which a person could not access or possess knowledge by the mere purchase and self-study of books, there was a correspondingly lesser incentive for individuals to buy them. This limited demand may have helped constitute the smaller book market that acted as a disincentive to the development of printing. This left instead an economy of bazaar copyists and private scriptoria sufficient to supply the men who possessed and passed on knowledge through practices of shared reading and ‘incorporative’ book use.

Incorporating Books

This conceptual location of knowledge in persons becomes apparent when we begin to examine the ways in which books were used in traditional patterns of Muslim learning. Traditions of Quran and Hadith-learningFootnote ⁹ had long lent a pedagogical emphasis to their memorization by a small ‘literate’ class.Footnote ¹⁰ Such literacy was in many cases only a tool to mnemonically master the Quran and a few other basic texts; as we see below, literacy had more value for the bureaucratic purposes of gathering ‘information’ than of acquiring ‘knowledge’. The wider social reach of these texts came not through widespread book-use among the general population but through the texts’ mnemonic internalization and spoken performance by learned remembrancers (huffāz). As its name designated, the Quran was a text ‘recited’ in speech, preferably from memory, and the learning habits that surrounded it affected wider attitudes towards book-learning.Footnote ¹¹ The self-replicating traditions of Quran learners were long lasting and widespread, inculcating attitudes among the religious classes that valued internalizing books over owning them. In a period of hand-produced books, readers seem to have read more deeply than widely: as one observer phrased the matter with regard to Iranian savants, “The Persians have few books; but what they read, they read thoroughly, and can repeat much of by heart”.Footnote ¹² Outside India, inventories of Damascene book collections show that even by the nineteenth century surprisingly few ‘ulamā possessed copies of the Quran, carrying its words within themselves instead.Footnote ¹³ Hadith were similarly committed to memory, their ‘sound’ (sahīh) authority founded on a ‘science of men’ (‘ilm al-rijāl) that developed through a corresponding epistemological faith in truthful persons rather than documentary evidence. In the context of jurisprudence, Quran and Hadith were certainly sometimes supplemented by written compendia of sound judgments (fatāwā), but here too judicial authority remained in the person of the faqīh who had mastered the texts.Footnote ¹⁴ Knowledge was that which was committed to memory, which books served to aid and supplement, but not to correct.Footnote ¹⁵ When a ‘standard’ printed edition of the Quran was issued in Egypt in the early twentieth century, its accuracy was checked not against codicological evidence but against human memory.Footnote ¹⁶

The ‘ulamā and their modes of knowledge loomed large in literate learning throughout the medieval period, their institutions providing the basis for bureaucratic no less than for clerical training as it emerged in the Turkish sultanates of Persia and North India.Footnote ¹⁷ As a consequence, their modes of book-use influenced the development of other fields of written learning. Among these other fields were the major genres of medieval and early modern Persian writing in India. The most obvious example is the tazkira, quite literally the book of ‘memory’ (zākira), the purpose of which was to supplement or correct the failing recollection of its users.Footnote ¹⁸ As the hardware of cultural memory, a tazkira might contain the lives of royal and saintly heroes or otherwise the anthologised verses of different poets, typically of more obscure poets lacking a dīwān of their own. In such tazkira books the written form of stories or verses served as aides-mémoire: what was deemed important would be committed to memory. From there it would reach its larger and main audience through spoken performance. As the theorist Walter Ong has noted, ‘written texts all have to be related somehow to the world of sound, the natural habit of language, to yield their meanings’.Footnote ¹⁹ The mnemonic function can also be observed in a range of other Persian genres in India. Historiography, or tārīkh, had the purpose of recording the ‘dates’ that its name suggests as part of the commemorative enterprise that made historiography so central to book-writing at court. These functions became particularly important to the great burst in the sponsorship of historiography at the Mughal court, in which Afghan, Mughal, Rajput and other diasporic elites sought to record collective memory as it dwindled under the affects of movement, dislocation and language change.Footnote ²⁰ Biographical books (tabaqāt) provided performative moral models of emulation for the pious; poetic dīwāns a choice of verses for the dandy to memorize and reproduce from his own person. The more widespread genres of the courtly or religious etiquette book involved even more patently physical modes of memory, knowledge of them testified not through recitation or quotation but through the public modification of bodily habits and comportment.

Here I would hesitate to pursue Mary Carruthers’ model of the ‘book of memory’ because while the notion of memory may be useful up to a point, we should be wary about projecting modern psychological models of the mind and its parts back in time.Footnote ²¹ Certainly, memory (hafz) was recognised as an important human ability, but to substantiate memory as a part of the brain, or even mind, may not only be anachronistic but may also be to miss the trick of what such acts of memory were seen as doing.Footnote ²² To commit knowledge to memory was not merely to store it in one's mind in a mentally internalised ‘book’ of memory, but also to incorporate it into one's whole person, moral, metaphysical and social. To return to our focus on authoritative religious knowledge, it is in this sense that we should understand the various things that Indian Muslim readers did with the words they read: they ingested words written on paper or breathed on water; they inhaled words that accompanied the chanting exercises of zikr; they inscribed words on their internal organs through techniques of visualization; they wore words as talismanic shirts and amulets.Footnote ²³ Words—indeed writing as such—had diverse locations, the most authoritative of which were not those on paper (which few could access or read) but those incorporated into persons through the mnemonic and other incorporative means already mentioned. Before the mass proliferation of books with the spread of printing, the manuscript codex was only one of a range of locations for the written word, and its individual purchase and reading in the form of a book only one of its uses.

If this has sketched a general model of how knowledge, and the authority it bestowed, were situated primarily in the person rather than in the book, it is now time to present evidence for its existence in the period of ‘print absence’ with which this paper is primarily concerned. Since the intention here is not to present static models of ‘Islamic’ reading habits, ways are suggested in which this one mode of book praxis was challenged and arguably displaced by the later Mughal era. The particular time and place for this investigation is the opening decades of the eighteenth century, when at the same time that the Saxon printer Bartholomaeus Ziegenbalg was printing Tamil books in Tranquebar, in his takiyya on the other side of the Deccan in the erstwhile Mughal capital of Aurangabad, Shāh Muhammad Musāfir was using his books in a distinctly anthropocentric mode.Footnote ²⁴

The Uses of Books in the Aurangabad Takiyya

In testament to the patronage Aurangabad received when it was briefly capital of the Mughal Empire, and subsequently the founding capital of the ‘Hyderabad’ Nizams, the takiyya of Shāh Musāfir is one of the most notable examples of late Mughal architecture.Footnote ²⁵ At its centre lie the tombs of two Central Asian migrants, Shāh Sa‘īd Palangpōsh (d.1110/1699) and Shāh Muhammad Musāfir (d.1126/1715), members of the Naqshbandī order of Sufis with a large following among the city's imperial elite.Footnote ²⁶ Reflecting the ethnic as much as the religious affiliation of its followers, the takiyya was chiefly associated with Central Asian (turānī, moghol) soldiers in the service of Ghāzī al-dīn Khān Fīrōz Jang (d.1122/1710) and later his son, Nizām al-Mulk (d.1161/1748), the subsequent founder of Hyderabad State. Shāh Musāfir's affiliations with these servants of empire brought fruitful reward in the construction of his takiyya complex, which comprised a mosque, khanaqah, mausoleum and residential quarter, as well as a book collection which included several works from the imperial library.Footnote ²⁷ It was here that Shāh Musāfir's first successor, Shāh Mahmūd (d.1175/1762), wrote the Malfūzāt-e Naqshbandiyya in commemoration of Shāh Musāfir and the takiyya's activities in his lifetime.Footnote ²⁸ In this work we find a series of detailed accounts of the uses to which the takiyya's books were put in Shāh Musāfir's circle. Lest it be suggested that this was an unlearned, or book despising place, and so a poor choice of case study, it is important to state that, if anything, the takiyya was an uncommonly bookish place. Its library attracted the major literary figures of the region, including the foremost literato of the age, Ghulām ‘Alī Āzād Bilgrāmī (d.1200/1786), who in the 1740s spent seven years living there. Given that Bilgrāmī's chief contribution to Indian letters was the compilation of anthologies of poetic biography and verse, the texts produced by literati no less than Sufis point to the links of writing and embodied memory.Footnote ²⁹ As we see in more detail below, even in what was in its period an usually book-familiar circle, the uses of other books were similarly constrained.

Amid the cache of circumstantial detail preserved in the Malfūzāt-e Naqshbandiyya, we find evidence that the use of books was limited in the circle of the takiyya's founder, Shāh Muhammad Musāfir. Shāh Musāfir had attended a maktab school as a boy before spending several years as a teacher in Ghūrī (present-day Afghanistan) during his early manhood, so we know that he was fully literate before he reached Aurangabad.Footnote ³⁰ The Malfūzāt-e Naqshbandiyya's description of his subsequent dervish apprenticeship under his own master Shāh Palangpōsh points towards a mode of acquiring certain knowledge (ma‘rifat) that had no place for any kind of book-learning at all and drew by contrast on a programme of extreme self-abnegation reached through destitution and servitude.Footnote ³¹ It was mastery of the soul and body, rather than of books, that gave access to such knowledge.

However, when Shāh Musāfir came to teach apprentices himself, his own mode of instruction did include a role for books. But his use of them was closely circumscribed and they were only introduced into his curriculum within a context of oral face-to-face tuition.Footnote ³² One revealing anecdote from the Malfūzāt-e Naqshbandiyya describes him reading to his students from a biographical compendium of other holy men, the Tazkirat al-awliyā, during which he was interrupted by an argument between his students over the meaning of the segment under scrutiny.Footnote ³³ Here several characteristics of book use emerge that deserve attention. First is the oral context within which the book was used, such that written texts were woven into ex tempore commentaries and discussions that situated authority in the living master who guided and controlled the reading session. There was no independent transfer of knowledge between author and reader. Second, the book in question was a collection of the sayings and deeds of the saints and as such a medium of exemplary behaviour, concerned like the Hadith with embodied action more than abstract principle. It was a book that served the authority of men worthy of emulation rather than of ideas or even other books. We also witness in the account of the reading-group the operation of a tripartite model of book use.Footnote ³⁴ For when the argument broke out, after some debate among the apprentices themselves, it was the shaykh Shāh Musāfir who intervened to provide the authoritative answer to the issue in dispute. Book knowledge was thus transmitted through a continuum of the author's voice, readerly opinion and the final but decisive interpretation of the living shaykh. Textual authority—even an author's control over his own words—was relegated to the authority of the master in and as person.

Books were no more easily accessible and open to unguided reading in other spheres of religious learning, where access to books was itself controlled through social mechanisms made possible by the limited availability of books in the private libraries of the manuscript ecumene rather than the later free market of print.Footnote ³⁵ In such contexts, book knowledge could be mediated through the authority of the person, who was able to decide at which point his students were ready to access a particular book, books which in the case of libraries such as that at the takiyya could be accessed only with the shaykh's permission.Footnote ³⁶ Group reading strategies thus sought to control the written word in much the same way that a master decided when or whether to transfer knowledge through the spoken word. Books—and their readers—were not considered autonomous and the master remained at the centre of his particular ‘field’ of knowledge.Footnote ³⁷

The Malfūzāt-e Naqshbandiyya provides little evidence of reading activity that was not organised by means of such directed reading groups. Reading was a social activity, and its sonorous and collectively intensified performance lent it correspondingly powerful results. A shared recitation of the Masnawī of Mawlānā Rūm (d.672/1273) at the takiyya led one participant to fall into raving ecstasy (hāl). The physical and socially administered act of reading aloud summoned a psycho-physical state in which knowledge was received into the person directly from God rather than through the mediation of writing.Footnote ³⁸ Here the book did not contain knowledge as such but was an instrument for provoking the conditions in which knowledge could be received from beyond it. Ironically, the book of the Masnawī functioned as a technology that rendered itself, and other books, epistemologically obsolete.Footnote ³⁹

When other books are mentioned in the Malfūzāt-e Naqshbandiyya as being used by the takiyya's residents and visitors, they consisted mainly of what was by the early 1700s a canon of classical Persian poetry and Naqshbandī sacred biography.Footnote ⁴⁰ Epistemological anthropocentrism also underwrote the construction of these biographies, which frequently quoted persons in preference to books as their sources, echoing the older perceptions we have seen that while the reliability of a human informant might be reasonably assessed, written knowledge had a propensity towards dissimulation and forgery. Like similar works of its type, the Malfūzāt-e Naqshbandiyya (itself a tazkira in spite of its latter day title) thus presented itself as a repository of spoken knowledge, providing the names and biographies of its ‘trustworthy’ informants before recording the details of what they remembered and said about Shāh Musāfir's life.Footnote ⁴¹ Even in the 1730s, when the text was composed, books still in these ways attempted to capture the reliability of direct spoken communication, the importance of which was itself a reflection of an anthropocentric model of knowledge. Nor was this a peculiarly Sufi attitude, for among ‘ulamā book users in the Central Asian tradition from which so much Indo-Islamic scholarship descended, the ‘hearing’ (masmū‘āt) method of mastering books was considered superior to the ‘reciting’ (maqrū’āt).Footnote ⁴² While the doctrinal risāla did enjoy some success in Mughal India, the Malfūzāt-e Naqshbandiyya also suggests that such treatises were by no means a staple of every library. It is by no means clear by whom and in what circumstances such books of abstract and theoretical learning were employed.Footnote ⁴³ I would suggest that it was not only rather less than many scholars have imagined, but that such works can be seen within the same epistemological nexus as ideological armoury defending the magisterial person as the locus of knowledge. It is after all in such doctrinal works that we find the fullest elaboration of an ideology of ‘friendship’ and ‘proximity’ (wilāyat, qurbat) to God that championed the knowledge of such friends of God over the uncertain ‘acquired knowledge’ (‘ilm-e husūlī) gained from reading.

Although numerous books are mentioned in the Malfūzāt-e Naqshbandiyya as being present at the takiyya, there was a clear sense of poetry as the paragon of letters, especially the Masnawī of Rūmī and the lyrics of Hāfiz. But again, poems were not conceived nor even, in most cases, received as written texts. Writing was only one stage in the social life of the poem, the stage before it was committed to memory and then circulated, performed and appreciated through spoken or sung recitation.Footnote ⁴⁴ Writing was not only tangential to the human locations of poetic speech, it was also comparatively expensive. In the manuscript era, memorising borrowed or recited books was inevitably cheaper than buying them and potentially more efficient than copying them, a relationship that arguably did not change until time (spent memorising) gained a greater financial value than the (by then printed) book in question. At several points the Malfūzāt-e Naqshbandiyya does draw attention to the financial value of books, for example, when the takiyya's library was hidden during a raid by the Marathas, and on another occasion when a form of arrow-spinning divination was used to discover the thief of an especially valuable book.Footnote ⁴⁵ So to claim that knowledge was not seen as independently located in books—was not ‘bibliocentric’—is not to say that books did not have value as commodities. Yet here too the Malfūzāt-e Naqshbandiyya points to the variant status of different genres: when a handler of stolen manuscripts was punished by having his loot strangely turn to dust, only the Masnawī-ye Ma‘nawī, Tazkirat al-Awliyā and Nafahāt al-Uns remained intact.Footnote ⁴⁶ For the readers at the takiyya at least, we have a supernatural iteration of the relative value of different books: poetry and saintly biography—the most anthropocentric of books—were the most esteemed. For the same reasons, the most highly valued documents outside of the takiyya were letters written in the royal hand, because ‘letters were deemed to preserve the charisma of their writers’.Footnote ⁴⁷

This sense of poetry as the most valued element of Persian literary culture was more widely shared and the ways in which this was manifested show how the attitudes in question did not solely pertain to religious knowledge. After all, a Mughal gentleman was not judged on the futile measure of how many poems he had perused, but on how many verses (or, for other kinds of ‘book’ specialists, legal rulings or hadith) he had committed to memory and so absorbed into his person. In numerous other settings beyond Aurangabad, it was books of poetry that were given greatest value by more patently financial means than the supernatural valuations given at the takiyya. More than any other type of book, it was poetic works that were copied out by the finest calligraphers and bound between the finest covers.Footnote ⁴⁸ As a result, among the diverse holdings of the Mughal imperial library, it was books of poetry that were reckoned to have the greatest financial value.Footnote ⁴⁹

It is not, however, the intention of this paper to present the relationship between books and their users in terms of fixed modalities and unchanging attitudes. If, in the 1730s, the Malfūzāt-e Naqshbandiyya showed the continued elevation of persons over books as the locations of knowledge, other texts from Mughal Aurangabad point towards more nuanced and perhaps transitional attitudes. Examples can be seen in the circle of Shāh Musāfir's contemporary, the Chishtī master Nizām al-dīn Awrangābādī (d.1142/1729), whose khānaqāh in Aurangabad served as no less a repository for public book use.Footnote ⁵⁰ Here too we find expressions of controlled access to knowledge, and his ‘recorded conversations’ (malfūzāt) describe not only parallel group readings to those seen at the Naqshbandī takiyya but the same awareness of the embodied locus of authoritative knowledge in the text's attempt to use writing to recreate the verbal presence of the master.Footnote ⁵¹ As Nizām al-dīn's own teacher Kalīm Allāh Dihlawī (also d.1142/1729) had taught him, the instruction of a living master provides ‘an assistance impossible to get by the mere reading of books’.Footnote ⁵² Paradoxically, though, this advice had been made in a written letter. Here we see a tension emerging from the increasingly geographically-dispersed relationships between teachers and their disciples, relationships which were conceived as being face-to-face but which, as a result of increasing individual movement (in Nizām al-dīn's case as a result of imperial expansion into the Deccan), needed to be vicariously recreated through the exchange of letters. Paper—and the personal paper of the intimate letter—thus affected a geographical repositioning of the master's voice and in Nizām al-dīn's case this meant through letters carried from Delhi by the imperial postal dāk.Footnote ⁵³

Nizām al-dīn's own book, the Nizām al-Qulūb (‘Order of the Hearts’), was a no less peculiar kind of text, which struggled to describe in words the peculiar vocal sounds and bodily practices appropriate to different forms of meditation.Footnote ⁵⁴ In Jack Goody's sense, it was a characteristically bookish book, an attempt to catalogue, classify and codify all of the physical breathing, chanting and prostrating practices its author had ever heard and seen.Footnote ⁵⁵ The Nizām al-Qulūb is a testament of the transformation of bodily into bookish knowledge, of corporeal and oral knowledge into the written record. As one of the earliest works to commit to writing practices of meditation that had previously been only directly transmitted between master and disciple, its very composition was a sign of changing times.

Like other innovators, Nizām al-dīn was only a few paces ahead of his time. Nonetheless, elsewhere at this time other Muslim knowledge-masters were also reformulating the relationship between knowledge and writing. One such figure was Nizām al-dīn's Syrian contemporary, ‘Abd al-Ghānī al-Nabulusī (d.1143/1731). In several of his writings, notably the Takmīl al-Nu‘ut fī Luzūm al-Buyūt (‘Perfecting the Attributes in Remaining at Home’), ‘Abd al-Ghānī wrote of the superiority of the books of the dead against the words of the living as a source of knowledge, a position which stood in juxtaposition to the epistemological tradition of fellow Naqshbandīs like Shāh Musāfir.Footnote ⁵⁶ While it has to be said that the context of Ottoman Syria was very different to that of the Mughal Deccan (by ‘Abd al-Ghānī's period there already existed a highly bureaucratized and centralized system of religious learning in which the authority of paper documents had arguably made much more headway than in India), he was still forced to defend his position against contemporaries who claimed that at a certain point of the epistemological journey, book knowledge had to be abandoned.Footnote ⁵⁷ In India itself, probably the most obvious comparative example of such a ‘bibliocentric’ shift was Guru Gobind Singh's substitution of the living guru for the book of the dead gurus’ words in 1708. Over the following decades, under the designation Gūrū Grānth Sāhib the book gradually came to be regarded—seated, treated, even termed—in the manner of a living master: the book had replaced the person. There remains the possibility, then, that comparative research will detect larger shifts towards a growing faith in books over persons that preceded and paved the way for the shorter technological ‘moment’ that was the birth of indigenous Indian and Islamic printing around 1820.

From Books to Bureaucracies: Knowledge as Information

If we are to detect such an ideological shift in the location of knowledge in early modern India—from persons to books, from voices to charters—it may be towards the governmental sphere that we should look for its chief sponsors.Footnote ⁵⁸ Here we consider the spread and increasing valorisation of written learning (and hence of literacy more generally) within the context of the growth of bureaucracy. This points us to the realm of government and statehood, which was in any case never more than a short step from the circles of Aurangabad's Sufis. It is significant that the period of Shāh Musāfir and Nizām al-dīn also saw the centralising attempts of the Mughal state to depersonalise and thence control Islamic legal knowledge. The Fatāwā-ye ‘Ālamgīrī, the ‘standardised’ handbook to Hanafī juridical practice sponsored during Shāh Musāfir's lifetime by the emperor Awrangzēb, was compiled to diminish the role of personal legal reasoning (ijtihād) in favour of written juridical precedent.Footnote ⁵⁹ While the text has often been seen as a symptom of the emperor's personal piety, it makes more sense to view it as a response to the great expansion of the empire's geography and personnel in Awrangzēb's reign. This bureaucratic push towards a bibliocentric relocation of knowledge was also seen in Shāh Musāfir's own takiyya, where outside of the direct circle of his own disciples the main function of book learning there was to train civil servants in the more practical skills of literacy. Thus, while Shāh Musāfir gave his followers access to knowledge as ma'rifa, in the same takiyya the resident teacher Mīr Muhammad Yūsuf taught the many resident orphans of the Mughal armies to read and write and so be equipped to manage the paper government of imperial bureaucracy. Many of the orphans apparently did find such jobs, one becoming secretary to the new state-founder, Nizām al-Mulk.Footnote ⁶⁰ It would be more correct to view the orphans’ literate education in terms of the learning of a marketable skill or trade rather than as the acquisition of authoritative knowledge. Here we are clearly dealing with a different kind of social capital, a mode of knowledge used by a different social group for a different purpose. While the Sufi and ‘alim sought unquestionable truth (haqq), the state sponsors of the munshī sought the more malleable category of the datum or ‘happening’ (wāq'a). The religious and bureaucratic apprentice thus had different purposes in his book use and correspondingly different attitudes to books as a result. But rather than seeing these attitudes as constituting parallel separate modes of knowledge, we might ask whether they were beginning to converge in the period in which Shāh Musāfir's takiyya flourished. The takiyya's greatest book producer, Āzād Bilgrāmī, points towards such a convergence, for while posterity remembers him as a man of letters, he in fact spent much of his career in the civil service of Nizām al-Mulk.

At the takiyya, as elsewhere, the general promotion of literacy was achieved as part of the expansion of bureaucracy rather than as part of an alternative praxis linking writing with access to knowledge per se. The latter, especially in the higher sense of ma'rifa, was controlled through the social mechanisms of initiation (bay'at) at the hands of a living master and the controlled pursuit of book learning as part of a larger pedagogical programme of ascesis, prayer and service. Clearly, such comprehensive programmes were unnecessary, even counter-productive, to the bureaucratic purposes of empire and so, as in the madrasas of the Saljuq Empire beforehand, we find the takiyya also providing a stripped-down programme of more basic instruction in functional literacy. We should not be surprised. The expansion of Mughal power saw the development of the largest and most complex political entity in India's precolonial history, a development that required a larger and more sophisticated system of bureaucracy than had ever previously existed. It is therefore in the bureaucratic sphere of information rather than the epistemological sphere of knowledge that we should seek the emphasis and demand for writing and thence for literacy. If such more efficient modes of pedagogy had the danger of compromising the moral education from which the larger anthropocentric mode was inseparable, then in the growing numbers of adab etiquette texts we can see a subsequent attempt to use book-reading as a form of moral self-fashioning in the absence of a master.Footnote ⁶¹ Such a perspective would connect the shifts we can observe in Aurangabad to a larger political context of imperial state-building and in turn connect them to the signs of ‘early modernity’, as well as to recent discussions of the connection between bureaucratic expansion and the growth of history-writing.Footnote ⁶² It would perhaps in turn help explain what appears to be a more general expansion of Persian book production under the Mughals, as testified in the vastly growing number of manuscripts that survive from around 1600 onwards.Footnote ⁶³

Persian documentation from the later Mughal period provides further evidence for such changes in modes of knowledge, for this expansive bureaucratic conscience had its own effects on the organisation of writing. These ranged from the practice of keeping valuations and inventories of the Mughal imperial library and the development of the governmental news report (akhbār) to the growing obsession with paper learning certificates (ijāzāt), the growth of epistolary (inshā’) manuals and the appearance of new bureaucratic tropes in a range of other literary spheres.Footnote ⁶⁴ In Aurangabad, Shāh Musāfir's contemporary Nizām al-dīn was thus reported to have given charity to the poor by stamping chits with an official seal lent by the imperial chancery, while the shrine of Shāh Nūr Hammāmī (d.1104/1692) was patronised by members of the Mughal civil service who there sought miraculous solutions to their bureaucratic dilemmas.Footnote ⁶⁵ When Awrangzēb's minister Diyānat Khān left an important dossier in Agra, Shāh Nūr was said to have mysteriously transported him to that city to collect it in good time.Footnote ⁶⁶ The karāma or grace brought by the shaykhs’ personal knowledge of God was thus in itself seen as serving governmental purposes. From around the seventeenth century, the dispersal of Sufi personal charisma or baraka itself increasingly relied on such bureaucratic means, from the genealogical document (shajara-nāma) to the dispersal of Sufi lodges along the tracks of state representatives.Footnote ⁶⁷ It is no coincidence that the main patrons and many of the followers of Aurangabad's major Sufis were imperial servants, whose ranks included men like Āzād Bilgrāmī and the recorder of Nizām al-dīn's conversations, Kāmgār Khān, who committed their shaykhs’ lives and words to writing.Footnote ⁶⁸ In such ways, civil servants paid back their dues to their miraculous and moral benefactors.Footnote ⁶⁹

It is fitting, then, that Bahā’ al-dīn Hasan ‘Urūj (d. c.1230/1814), the main biographer of the Aurangabad shaykhs at the end of the eighteenth century, earned his living as a teacher of the art of official letter-writing (inshā’).Footnote ⁷⁰ In the bigger picture of Persian and thence Urdu book culture, this explosion of bureaucratic book production would also manifest itself in the production of ever-longer commemorative tazkirāt. Often written by civil servants like ‘Urūj, in the eighteenth century these works flourished on an unprecedented scale, cataloguing the saints, scholars and poets—as it were, the religious and cultural resources—of a given town or region in a manner that drew on the book-keeping habits of the state.Footnote ⁷¹ An early example of these literary adaptations of the language of bureaucracy was the Dabistān-e Mazāhib (1657), the great early modern typology of India's religious groups which took the model of the munshī's ‘school’ (dabistān, contraction of dabīristān) as its organising trope. Moving beyond such medieval models as Dawlat Shāh Samarqandī's Tazkirat al-Shu'arā, these books drew on Mughal imperial gazetteers such as Ā’īn-e Akbarī and began to fill multiple volumes of encyclopaedic scope. In an echo of their provenance, many bore titles that reflected an increasingly bureaucratic mentality, with titles designating their contents a ‘treasury’ (khazīna), ‘storehouse’ (makhzan) or ‘market’ (ganj) of knowledge.

As the state created increasing numbers of Persian-writing literati in the seventeenth and eighteenth centuries, there arguably emerged what we might see as a literary culture of ‘accounting’ in which the bureaucratic organisation of knowledge seeped into numerous fields of writing. The process may be observed beyond the direct reach of Persian in the development of the Marathi historiographical genre of the bakhar, in which the Maratha past was lent shape under the influence of the Mughal news report or khabar.Footnote ⁷² In some respects, these developments were reflected in other parts of Asia. It has been argued that the huge expansion of writing (in this case, in print) in Tokugawa Japan was part of a new culture of ‘classification’ seeing knowledge broken down into manageable components that could be organised in new written formats for the recording of information.Footnote ⁷³ As Mary Elizabeth Berry has written in connection with the rapid rise of the urban ‘gazetteer’ in early modern Japan, such writings formed an attempt ‘to examine and order the verifiable facts of contemporary experience for an open audience of consumers’.Footnote ⁷⁴ If in Mughal India there was still no printed ‘open audience of consumers’, we may still be able to observe a relocation of knowledge by a series of bureaucratic increments, from the person to the archive, from flesh into paper. The pressure of civil service examinations in Ming China similarly served to transform knowledge into digestible portions. Until the emergence of the alternative literary culture of the shishang (‘merchant-literato’) with the massive rise of printing under the late Ming, bureaucracy ‘structured the economy of linguistic exchange, reproducing the power relations among the users of linguistic systems by privileging the “language of the official” (guanhua)’.Footnote ⁷⁵ It seems possible that the increasing scale of the Mughal bureaucracy, at its peak in the years of Shāh Musāfir's takiyya, encouraged a conceptual relocation of knowledge from persons to written documents, that is, from larger programmes of textual and non-textual learning under the close guidance of a master to the more autonomous and bibliocentric patterns of education epitomized by the private reading, and copying, of the epistolary manual. Such a shift need not therefore be identified with the rise of either writing or printing as such. The shift can be identified in the particular ways in which written (in this case, hand-written) documents were used in a period that saw the development of a large-scale manuscript bureaucracy and its axiomatic tendency towards locating knowledge in writing and in turn identifying knowledge with information.

Locating the beginnings of this shift of epistemological paradigms in the period of Mughal rule makes greater sense of the notion of an early modernity in India that is not dependent on the arrival of print as it was in Europe, China and Japan. Bureaucracy and governmentality in general now begin to take centre stage in this picture of the reconfiguration of knowledge and its gradual seepage into the religious no less than governmental spheres of writing. If this nexus was by no means entirely new by 1700, it does seem that the scale of the infiltration of bureaucratic into religious book use was. One thing that this suggests is the analytical limitations of conceiving of discretely ‘religious’ and ‘political’ spheres as opposed to a more unified field of textual and linguistic operations.Footnote ⁷⁶ But it also helps us to lay the ground for answering our initial question concerning the absence of print by moving away from a framework of momentous epistemic breaks towards one of gradual epistemological collusions and convergences. The emergence of ‘modern’ attitudes to book use in India therefore becomes less a question of revolutionary moments (‘When was the first book printed in India?’) than of the gradual adoption of new epistemologies, reading practices and learning methods that instigated new attitudes towards the uses of books. In the course of the sixteenth and seventeenth centuries, a related transition was made with the passage of the Deccan's vernacular languages into writing.Footnote ⁷⁷ But the larger transition we have identified, by which books replaced persons as the conceptual and practical locations of knowledge, was a much slower process and no such tipping point would be reached until perhaps the early or mid-1800s. What seems clear is that by the great age of the Indian madrasa in the second half of the nineteenth century, religious no less than secular learning was being practised through a ‘bibliocentric’ mode favouring fixed curricula, examinations and the intensive mastery of books.Footnote ⁷⁸ If, like the printing press, these were also ‘colonial’ adaptations, then what we have seen would suggest that these changes too had precolonial roots.

At present, it is difficult to assess the degree to which these developments can be connected to the price and availability of paper or to the possibility of easier access to it by civil servants.Footnote ⁷⁹ With regard to Aurangabad, we know of the existence of a papermaking industry around ten miles away in the village of Kaghazipura (literally ‘Paperville’) that owed its original existence to the brief relocation of the Delhi Sultanate to nearby Daulatabad, again placing bureaucracy as a promoter of paper-based knowledge. Production there seems to have continued under the Mughals.Footnote ⁸⁰ Yet in Kaghazipura too, it was the end of the eighteenth century that saw the great expansion of its existing paper industry, when the vast demands of the Peshwa daftar arranged for the founding of forty new factories.Footnote ⁸¹ Only research on the precolonial paper industry can equip us to fully address the question of the rise of bureaucracy and its relationship with written knowledge in India before printing. In this paper, as in other studies of precolonial literary culture, paper and other physical media remain the material missing link.