Reconstituting the Archive

Intending to write a book on law in the Mughal Empire, I began work in the National Archives of India in the year 2012, and was immediately overwhelmed by two problems, one practical and the other analytical. The practical problem consisted of the sheer bulk of the material available, disrupting the generally held view that Mughal archives have not survived. Archivists in the ‘Oriental’ section of the National Archives began by telling me about the roughly 130,000 Persian-language documents from the Mughal province of Golconda, which were generally single-page scraps in hideous shikasta or ‘broken’ scribal hand, and, as far as I could see, concerned with routine records related to taxation. When it became clear that I was making no headway at all, the archivists, taking pity on me, introduced me to the ‘acquired’ Persian documents series, which, as they explained, consisted of collections of documents from various families deposited with the National Archives. This was a much smaller series, but still substantial – around 5,000–6,000 documents – and were thankfully, very well catalogued, with detailed summaries of every document. Even so, how many of these should I read, and to what end? Taking a random and typical page from the catalogues, I struggled to see the connection between a letter about preparing tax records that referred to Bijapur, a copy of a parvāna that referred to Malwa, a bai ͑ -nāma (sale deed) executed in an unknown place and a rahn-nāma (mortgage deed) recorded in an unidentified location, all referring to different individuals. What was the connection of any of this with Mughal law? Where were the questions put to the muftī and his responses, the judgements of the qāz̤ī and the sijillāt or registers of judgements and deeds – all of which we have been taught to expect in the leading works of Islamic law? As the reader can see, this was the predominant empirical and interpretive structure that I wished to fit my material into, and found myself failing spectacularly in doing so.

I persisted with the catalogues, and plodded through the documents over several months, with the help of the most generous teacher one could ever have – Chander Shekhar, then Head of the Persian department at Delhi University – learning to identify the framing formulae, the logograms that compressed key words and the conventions of transcribing Indic person and place names in the Perso-Arabic script. From him, I learnt to read seals and to reliably identify the position and status of the persons using those seals, especially nobles and qāz̤īs; I learnt to tell the difference between an original and a copy; I learnt the joys of decrypting dates from calendars that combined a regnal year with an Islamic month. As I did so, certain names and patterns finally began to surface from the dump of scraps; I began to identify some people as they turned up again and again, doing various things in roughly the same cities and villages, and, after a predictable number of years, making way for other individuals who were clearly their descendants.

It is then that I understood how the ‘acquired papers’ were organised. Clearly, collections of specific families had been acquired by the National Archives, but the integrity of these collections had not been preserved in the catalogues. Instead, the cataloguers took all such documents as a single set of ‘acquired documents’ and organised them in a chronological order. The only way to re-establish the collections as they had once been would be to read through every single entry in the catalogues, pick out the shelfmarks of the documents pertaining to a single family, and order up and procure copies of those documents. I have told the story of how one individual in particular – Purshottam Das – forced itself on my attention. He was the principal protagonist in the majority of the eighty-four documents related to this family that were held the National Archives, whose dates stretched across the seventeenth century, straddling the reigns of the emperors Shah Jahan and Aurangzeb. The other documents referred to members of the family who inherited his estates after his death; the legacy he had created in the district of Dhar and its environs was clearly persistent. I ended that particular stretch of research with a diverse collection of documents – orders, tax contracts, deeds recording transactions, letters – all united by this individual or his descendants.

But what was the purpose of reconstructing that connection? Could this one obscure individual, not a religious scholar, not even a Muslim, dabbling in taxes and cutting a variety of deals in a corner of the Mughal Empire tell us anything about Islamic law in India? Deciding to suspend that question for a while, I decided instead to study the documents in minute detail, transcribing and translating every one of them in collaboration with my teacher, now research partner, and re-catalogue them. As I did so, I began to see further connections; I began to see the relations between the protagonists, I discovered how fortunes were made and passed on. I learnt about Purshottam Das’s father Mohan Das, and his adventures with robber catching,

As I became more familiar with the members of the family across generations, I began to put together a rough genealogy. I then began to see that offices and perquisites did not always pass from one generation to another: Mohan Das passed his title to his brother Chandar Bhan, who passed it to Purshottam Das, his nephew, rather than his own son, Suraj Bhan. Perhaps my most exciting discovery was that a persistently troublesome branch of the family derived from Suraj Bhan’s dalliance with a Muslim courtesan, and the son who was born of that relationship. I could see how that son was actively excluded by the patrilineage, but how he, and other such rivals, continued to circle the successful line across generations. There were also important external allies: imperial nobles with whom the family had long-standing connections, and who could sort out not just a mansion or two, but also difficult relatives from time to time. I began to see this as a dramatic narrative that could allow me to use the story of familial/dynastic formation in order to examine how the Mughal state was both substantiated and accessed – for offices, for resources, for selling services and for creating and recording entitlements. This began to accord well with my existing interest in families and in micro-history as a methodology, perhaps with law in practice, although I was still unsure what exactly this ‘law’ consisted of.

The process of gathering source material continued far longer than I had expected. It was as if every time I thought my story was complete, a completely new store of records would open up, drawing me further and further into a winding cave of wonders, but also despair. Through a process described in the introduction, I managed to find not one but two supplementary archives of material related to the same family, one in the Dar al-Athar al-Islamiyyah Museum in Kuwait and the other in the household of the family, whose descendants I managed to locate, and who chose to be extremely kind to me. Once again, the story of archiving pressed itself forward. Presumably, sometime in the 1950s, one or more members of the family had disposed of some papers from their collection. DAI Kuwait had acquired what appeared, on the face of it, higher-value manuscripts, nicely scribed and sealed, nearly monolingual in Persian. This is a collecting strategy that focusses on the artefact itself rather than the context from which it is derived. The National Archives, on the other, in an immediate post-colonial context, had accepted a more diverse body of material, perhaps in line with the effort to create a people’s history and corresponding archive. The residue of the most visually unattractive and linguistically provincial material remained with the family.

In reality, all three collections were part of the same pre-existing archive; together, they told a more complete and comprehensible story, one which not only throws light on how to get ahead in the Mughal empire, but also begins to offer some indication of writing and copying, of recording and archiving practices – all of which point towards the household rather than the state as the location par excellence of Mughal archives. With all the collections of family documents deposited in repositories all over India, Pakistan and Bangladesh, a small but significant portion of which has been edited and published, we have a tremendous archive of Mughal history. We have just been looking in the wrong place all along.

Documents as Historical Artefacts

As I began to present some of this work in various articles and conference papers, one question began to force itself upon me; the matter of authenticity. One interlocutor expressed particular concern about the possibility that all these documents may have been forgeries, generated at the cusp of transition to colonialism, when the British urge to stem leakage of revenue and record all entitlements led to the infamous ‘Inam commissions’ and the rush to produce, if necessary fabricate, records of pre-existing rights.

This pushed me to look wider, and confront the question of Mughal archives – their simultaneous ubiquity as well as absence. There are two tremendous collections of Mughal-era documents, mainly from the seventeenth century, and both related to the newest province, conquered and created in 1685 – Golconda. The first is the Inayat Jung Collection at the National Archives, Delhi, and the other is the series at the Telangana State Archives, Hyderabad. Together, they have around 250,000 individual documents or more, mainly related to taxation and military recruitment and pay. On the other hand, there are smaller collections held in several libraries and archives around the country, which appear to all derive from specific families or religious institutions. Comparable with the latter are sets of documents acquired by European corporations, including the English East India Company, again predictably related to their own rights and privileges. I discovered also that such collections were a combination of sealed originals and copies – the latter designating themselves as such – and learnt that historians of Mughal India have generally paid very little attention to the difference, and, despite periodic concern about forgery, been very trusting of the factual contents of an archive thus presented. My interlocutor’s question could potentially undermine the entirety of Mughal historical studies based on documents.

The documents have always been taken to be the same as their content. To take one of the best current exponents, Hasan treats a family collection deposited at the National Archives, New Delhi; ‘calendars’ or detailed catalogues of documents at the U.P. State Archives, Allahabad; an anonymously composed/copied collection of documents produced for a French Orientalist; and a munshāt all as equivalent sources.Footnote ³ Let us a see what difference a more careful reading might make. Nile Green’s erudite introduction to The Persianate World, published in 2019, presents a ‘Persian-Marathi Inam document of Maloji Bhonsle’ as an example of a Mughal secretarial bilingualism.Footnote ⁴ The image of the document shows that it is a chak-nāma, which was a document recording the measuring out of lands following a grant. Although chak-nāmas are a well-recognised Mughal documentary form, they were also used in the Marathi-writing areas. Maloji, Shivaji’s grandfather, had been a high-ranking officer in the Ahmadnagar Sultanate, the archenemy of the Mughals under Malik Ambar, the Abyssinian general-turned-king. If it is indeed an authentic document recording a grant of land being made by Maloji to a Muslim shaikh called Muhammad Darvesh Qadri, resident in the town of Chamargonda (Shrigonda, in Ahmadnagar), it would be more correct to call it an Ahmadnagar document rather than Mughal.

However, the authenticity of the document is worth questioning. The document is indeed in Persian and Marathi (in the cursive Modi script) and so presents evidence of (Ahmadnagari) bilingualism; it also bears a superscript saying ‘by the iqrār of Maloji Bhonsle, a naql (copy) according to the original’. The copied document, moreover, does not bear any validating seals, and even so, only a photocopy of this unvalidated copy seems to be held at its current repository. The handwriting suggests that the copy is no older than the nineteenth century, and if so, it may well be the case that the Marathi in Modi summary we have beneath the Persian text is in fact the addition of colonial clerks in western India rather than an instance of Ahmadnagari epistolary tradition.

These thoughts are of course speculative. The aim of this short discussion is pointing out the vast scope and need that remains for the precise reading of these documents, with attention to the entire document as a historical artefact – its formal, material and graphic features – together with constant attention to its archival history. Just as an archaeologist would not be satisfied with receiving a plaster cast of a decorative feature as sole evidence of its historical use, historians of the Mughal and Maratha empires could be both more careful and more exploratory in using such Persian or bilingual documents as historical traces.

Dealing with Data: a Historian Crunching Numbers

Although one hundred and eighty-eight documents may not seem like big data, when these documents are fiendishly difficult to read, and one is trying to make a story out of a soup of person and place names and seemingly unconnected small events, they do begin to look intimidating. As I moved through this research, I realised that I would have to find a way of searching through my collection, so that I could trace the connections between documents and between people. However, this process of cataloguing and creating a database led me to other difficulties and discoveries, too. The documents sourced from the National Archives had dates assigned to them; I had to establish the dates by reading (and interpreting) the documents from Kuwait and Dhar. This is much more difficult than it sounds, because only a minority of the documents have full Hijri dates, which can be converted will relative ease to dates from the Gregorian calendar (Common Era) using reliable online converters and smartphone apps. The added bonus is that through obvious processes of data mining and tracking of user activity that are ongoing on our phones and computers all the times, I am now regularly sent reminders of the need to pray. There are however more difficult combinations of dates – the Islamic month with the regnal year (julūs) – is one that has given me no end of trouble. A date such as 10 Rabi ʿII regnal year 29, for example, requires us to know, first of all which emperor’s reign is being referred to. Since none of our documents are imperial orders, we can only know the answer indirectly – by reading the seal of the issuing officer, and hoping that he would declare himself the banda/murīd (slave/disciple) of Emperor So-and-So. This of course worked only for high-ranking nobles – mansabdārs. For issuing authorities further down the line, for example, from a noble’s household, the seal may simply declare allegiance to the noble in question. In such a case, the connection with the emperor and his reign can only be established if the noble can be identified. With some nobles making repeated appearance in our documents, this was relatively easy (over time); it remained difficult for others.

Once at least roughly established, the dates allowed me to organise the documents in a chronological order, which revealed how documents followed each other, as copies, supplementary orders and so on; how events followed each other; and how people related to each other. Tracking Chandar Bhan, Suraj Bhan, Purshottam Das and Muhammad Asad on my spreadsheet, for instance, I could now understand much better why Suraj Bhan, the black sheep and Muhammad Asad’s father, was hounded out of his possessions by other members of the family. I could also track the role of certain officials in the activities of the family – Qāz̤ī Muhammad Mustafa, for example, became a familiar name to me, as he must have been for members of the family at the end of the seventeenth century, since he sealed so many of the documents they kept in their collection.

I began disaggregating the data – I separated out information in columns of a table that listed a document by its provenance, date, summary of contents, person and place names, seal, invocation and language. It soon became clear to me that if I wanted to really utilise the data, a table on Microsoft Word was not sufficient; I needed a spreadsheet. And so, to the great joy of my engineer husband, I transferred the data to a Microsoft Excel spreadsheet, discovering immediately that this allowed me to do much more. I could now utilise filters to discern patterns in the documents. For example, I could tell that around 20 per cent of the documents contained some non-Persian elements, and that if we removed the documents from after 1740, the percentage falls slightly lower. However, if we take the ‘truly’ bilingual documents – that is, with significant sections of non-Persian text in the main body of the document, as opposed to very brief notes in the margins or reverse – the percentage is even lower. The predominant language of giving orders, making contracts and transacting with other subjects of the Mughal empire, even in the villages, was Persian. The other languages played an important and visible, but inevitably limited, role. The language use was also, as I have said in Chapter 4, co-related to the type of document. Tax contracts and iqrārs recording sales or other transactions were most likely to have another Indian language, in this case Malwi/Rangri, albeit with a heavily Persianised vocabulary.

Managing and manipulating my spreadsheet also afforded me a highly demonstrable defence of my research method – of re-creating the archive. I could plot the documents in time, and I could add filters to show how the time distribution of documents appeared different at every stage of that archival reconstruction.

Each stage in this chart refers to a stage in the accumulation of these documents. Stage 1 refers to documents collected from the National Archives alone, Stage 2, to the collated documents from National Archives and DAI, Kuwait, and Stage 3, to all documents, from NAI, DAI and Dhar. The NAI documents on their own offer a picture whereby the family’s documentation is clustered in the late seventeenth century, with the biggest store of documents being created in the year 1684, the year that Purshottam Das died. Had I stopped my research at this stage, my conclusion would have been that generational boundaries were the causes of most profuse record-making, that is, the imperatives for legal documentation arose from within the family and its structures.

However, as the DAI documents were added, the time distribution shifts to earlier in time, with a peak related, not to Purshottam Das’s death, but Dara Shukoh’s. As the succession battled raged and then settled, the zamīndārs of Dhar, having dealt with Murad Baksh and then Dara Shukoh as jāgīrdārs, now had to ensure that their titles remained secure under the new emperor, Aurangzeb.

This peak of documentation, related to imperial politics rather than family dynamics was heightened with the addition of the documents from Dhar. The year 1660 was the one with the most documents now, with significant clusters around that year. The Dhar documents also showed that documentation continued well beyond the Mughal regime, in languages and forms that were familiar as well as novel.

There is of course no way to assess the completeness of the archive I have recreated. I rest my case, however, on the value of a methodology that is time and labour intensive, but which has the potential to open up archives that are all around us and yet invisible. I propose that we imaginatively restore the fragments to their household archives, and pay attention to the families which acquired and preserved such documents, for the family is the locale for the actuation of the pre-colonial state, activation of law and storing of its records.

Figure C.1 Time distribution of the number of documents, by stage of research

Law and the Uses of Documents

The last scholar to examine such documents on a sufficient scale was Muzaffar Alam. He, however, used such materials to fill out a story about the changing contours of the late Mughal state in two major provinces. More recently, B. L. Bhatia, in writing about Emperor Aurangzeb’s relations with Islamic scholars and judges, has made significant use of such documents, especially those from the U.P. State Archives, Allahabad. Farhat Hasan has used catalogue descriptions of the same collection to talk about the meanings of property in South Asia.Footnote ⁵ In other words, the documents are used to tell the stories of greater things, never of the people who are the principal protagonists of the documents themselves.

This book uses the archive to tell the story of the people who populate that archive. This is not a case of reading against the grain; it is in fact reading with the grain, hearing the story that those who created the archive wished to be heard. I reiterate: the purpose of that is not simply to add colour and detail to the bigger picture; it is about answering some of the biggest questions for the field, such as: ‘Where are the Mughal archives?’ The answer is: ‘At home’.

That answer would be less startling if we think of a parallel. There is no central repository of patient records in South Asia, every seeker of medical treatment archiving their own documents, often over years. That makes their grimy, tattered files with fading X-ray plates no less medical, and no less real. It may well be the same with law. If that parallel works, then one has to also keep in mind that what people archive depends on what they believe constitutes medicine, or at least routes to health (similarly, law, or routes to entitlement). Just as medical prescriptions and magical amulets might co-exist in a matriarch’s alcove ‘medicine-box’, parvānas and court orders sit comfortably with family history scrolls in a landed family’s archive. Reading them together should be the obvious thing to do, especially if we are aiming to find what ‘law’ really meant to middling people in Mughal India.

Awareness of the range of materials available, and the necessarily incomplete nature of any archive that one might discover or even painstakingly reconstitute, also pushes us to think of the archive and library as situated in a spectrum of repositories, language use, genre and even concepts changing dramatically from one end to another. The library-archive concept is Brinkley Messick’s update to his powerful schema of ‘textual habitus’; both concepts signal a diversity of texts, genres and users, but connected through ‘a set of acquired dispositions concerning writing and the spoken word, and the authoritative conveyance of meaning in texts’,Footnote ⁶ mainly by the mediation of expert legal mediators (the jurists), who may be a ‘local cosmopolitan’.Footnote ⁷ Although many of the documents that this book is based on would be easily recognised by scholars of Islamic law as ‘Islamic legal documents’, many others, such as the royal and sub-royal orders, the tax contracts and newsletters, would not make it to that trans-regional category. The models that these documents are based on are various too, and if they were indeed united in the Indo-Persian munshāts, as I have suggested, then the texts in the higher shelves of this imagined library would contains works of Islamic jurisprudence, stacked next to royal chronicles. In all probability, the mediators would not have been Islamic scholars, not even Muslims, but munshīs – Hindu Kāyasths like the protagonists of this story. To understand their mindscape, as likewise that of the protagonists, one has to also read family history narratives that emphasise Rajput codes of honour and valour rather than legal doctrine of any kind, and remain aware of a language and lexicon very distinct from that of the Persian legal documents. I have been thinking with a notion of circles – dāi’ra – to map and interpret the scope and jurisdictions of various levels of knowledge, authority and understanding of law present in the reconstituted archive. By thinking of many circles, not necessarily concentric, I have tried to sidestep a notion of ‘Islam-plus’ in constituting this version of the Islamicate. The library offers another metaphor, perhaps a more directly realistic one, to refer to a wider, heterogeneous, even discordant field of social life and language use; it also signals the limits of the investigative methodology that this book has proposed.

The Meaning of Law

Seeing all this material together also helped me resolve, finally, that question of what ‘law’ is, and what, therefore, are appropriate materials for writing a history of law, especially law in and of the Mughal Empire. As proposed in the Introduction, and as argued throughout the book, I decided that it is unproductive to re-fragment an archive, restored through the process described above, in order to achieve a coherent narrative. The meaning of law derives from the integrity of that collection; rather than imposing an external definition and typology, I have observed what functions the documents perform within the collection. Thus observed, a farmān, nishān or parvāna, while an order, can work as a ‘charter’ creating rights, that are then repeatedly examined and recorded, through the officiation, above all of the local qāz̤ī; but they are also disputed, re-evaluated and re-documented, by the qāz̤ī, but also by other authorities, including the locally resident nobles. And all the time, people transacted – they bought, sold, borrowed, returned, claimed, paid up and gifted; they recorded all this in documentary forms recognisable across the partially overlapping Islamic world and the Persianate world.

The ‘law’ in such a context derives from a sense of right – both objective (what should be) and subjective (somebody’s right) – that is so trenchantly expressed in this totality of documentation. That sense of right/s cannot be attributed to any single learned tradition, whether that be fiqh or dharmaśāstra, nor to an India-centric notion of Mughal rules, nor simply a purely plastic set of local custom which was a thin veil for social and political power. All these elements may be referenced, especially through a repertoire of terms recognisable trans-locally, but the meaning and content of ‘law’ has to be derived from what was clearly significant to the protagonists themselves. From this point of view, the family history document is as important a source for accessing that broader understanding as the Persian-language documents recording specific rights and obligations.

The word that is used most frequently in these documents in order to signal a concept of entitlement is dastūr. Read in connection with the dastūr al-ʿamal manuals we saw in Chapter 3, there emerges a sense of how things ought to be done, which, when done right, renders everyone their due. Twinned with sharīʿa, this was the other most common abstract term used to signal ‘law’. I propose that dastūr may be the window to understanding law in practice in the Mughal empire; the Mughal name for Islamicate law.

Micro-history and Family History: a Conversation

The family history document also forces us to contend with the implications of memory, especially memories in families about ancestors. This is where I realised that, despite my belief that by using the methodology of micro-history, I had demonstrated my interest in actual human protagonists, this interest was still objective; it was, in the final reckoning, concerned with understanding how social actors operated and a system functioned. I was interested in the abilities, motivations, interests and world-views of my protagonists, but what I discovered had no implication for my own sense of self, or so I thought, until I decided to have an extended conversation with Amit Choudhary, Mohan Das’s descendant and my constant interlocutor over the past three years, and his elder sister, Meenal Shrivastava, Professor of Political Economy and Global Studies at Athabasca University, Canada.

In this conversation, which we conducted over Skype, we discussed not just this book but also another ‘family history’, but of a very different type. That book is Amma’s Daughters, an imaginative, novelistic, and meticulously researched book based on the diary of an extraordinary woman who defied social conventions, joined Gandhi’s ashram in Sabarmati, entered an unusual marriage and raised two daughters alongside a punishing regime of work as a social-political activist in post-colonial India.Footnote ⁸ This woman was Meenal’s grandmother, Amma. Meenal’s mother and Amit’s were sisters. Two completely different families had been joined by a marriage in the mid-twentieth century and two historians had written stories of those two families.

In recalling the reasons that had brought her to write Amma’s Daughters, Meenal spoke of the pain and loneliness of being orphaned and simultaneously feeling herself in exile in Canada. She remembered having seized the writings of her grandmother as her one connection to her past, and having embarked upon her project in an effort to revive that connection. As the project entered the stage of archival research, however, the enormous significance of the material to hand dawned on her, and the book became an effort to tell the story of women in the Indian nationalist movement. For Amit, on the other hand, the desire to seek the past of the thākur family into which his mother married was more intuitive. Given the family’s long pedigree and obvious local significance, how could he not want to know? My arrival in Dhar was, he said, an opportunity too good to miss, to find out about his own family in ways that he could not do himself.

We thus had a subtle binary developing as the rationale for interest in a family’s past – emotional connection versus intellectual curiosity. And yet, as Meenal offered to Amit, his curiosity was also born on an emotional connection. Amit agreed, and reconfirmed what he had told me in the past, that his favourite character in the story I had uncovered was the shadowy Jayanti Das, present in one very early document as a grantee but a key protagonist in the Malwi/Rangri family history as the child-warrior who had taken on a rogue Rajput and a Mughal commander in a quixotic mission to revenge the murder of his elder brother. As for me, while I had no connection with the family, my determination to find out about its past had at some point transformed from intellectual curiosity into an emotional investment, in a past that was so very different from our present that it was hard to imagine today.

Were we all interested then, in outliers, misfits? Meenal was very strongly in favour of this explanation; she offered a poetic-ethical exposition of the essential outsiderhood of every person, and thus the need to celebrate outliers wherever one finds them. In doing so, said Meenal, we should learn from the techniques of storytelling that had been used by indigenous communities in North America as political protest as well as cultural activism. Both Meenal and Amit expressed their approval for my method of research, including potentially ‘embarrassing’ discoveries such as the episode with Muhammad Asad. They were looking, they said, for history, not eulogy. Besides this common affiliation to social realism, we found that all three of us had a shared cause in our joy in discovering the people who made up the past, people whom we may like and admire, or not, but people who give us access to other worlds, that were, that may have been. A shared ‘interest in the eclectic’, as Amit Choudhary put it, had brought us together, and for my commitment to the project of discovering the family’s history, Amit offered me the status of an honorary Choudhary. Gratefully receiving that emotional compensation, and on this conversational note, I end this story.