Introduction^{Footnote 1}

Although orality has been deeply embedded into Ethiopian cultural discourse, Ethiopians esteemed the written word and believed in its authority for several centuries.^{Footnote 2} The oldest indigenous script and literary language, Ge’ez, which developed in the fourth century AD, the long-lasting scribal culture, and the vast collections of manuscripts and documents of all kinds that Ethiopians left behind in church collections are emblematic of this. From the early eighteenth century onwards, major changes in recordkeeping and the variety and quantity of documents emerged. In towns with churches and surrounding countryside, more and more people demanded registrations of all sorts of contracts they entered into and deposited them in church archives for safekeeping. Consequently, writing facilities held by churches developed into centers for registering and preserving conveyances of property under the supervisions of officials and specialized writers. By the middle of the twentieth century, church archives and writing centers had developed into chanceries in every accepted sense of the term.

In Christian Ethiopia, churches served as the exclusive custodians of documents and manuscripts for many centuries. Royal annals, hagiographies, and liturgical and doctrinal works formed the core of Ethiopia’s manuscript collections. The purpose of annals and hagiographies was to advertise, promote, display, and perpetuate the power and influence of a secular ruler or a religious saint. The records considered in this paper, however, are legal and administrative documents which are for the most part concerned with landed property such as charters, registers, cartularies, inventories, wills, contracts, and deeds. Churches placed their manuscripts, sacred objects, and vestments in a treasury as a place of safekeeping. These treasuries are called eqabét in Amharic. Eqabét also functioned as scriptorium and depositories of public and private deeds, contracts, and other documents relating to individuals, the state and churches. The production of documents was a highly organized and standardized process. As a norm, clerics with specialized knowledge and scribal training exclusively and permanently attached to eqabét were responsible for the composition of both private and official documents. It might be a stretch to call these eqabét of churches, libraries or archives and chanceries, pure and simple. Nevertheless, in terms of practice, function and staffing, eqabét can be appropriately considered the institutional equivalent of archives and chanceries.^{Footnote 3}

Chanceries were developing since the eighteenth century when the variety and quantity of documents increased and churches began long-term recordkeeping. The dramatic and steady growth of land market, which inspired the proliferation of documents and new rules of documentation, use of vernacular in legal records, and the practice of compiling registers and inventories were fundamental changes in this period.^{Footnote 4} The change in documentation may be explained by the allied change in the land tenure system of Ethiopia which inspired new ways of redacting material transactions, to which point I shall return later. Despite the increase in records, however, written documents did not supplant oral arrangements. Business transactions continued to be made by word of mouth as before and literate royal and ecclesiastical administration did not develop until the twentieth century. With the habit of recording property transactions in writing expanding both socially and geographically, a complex interdependence between the oral and the textual emerged in eighteenth-century Ethiopia and thereafter. Documents were commonly used to commemorate and reinforce oral transactions. Nevertheless, it is evident that documents also played practical roles. In anticipation of detailed discussion later, this paper argues that written records were created primarily for future administrative use. It will explore the development of written records and chanceries and the purposes to which archives were put to use.

As a historically significant phenomenon in itself, the circumstance of creation and preservation of documents has intrinsic interest. There has been a considerable amount of research done by scholars of the medieval west in particular on the development of literacy, as well as its social, political, cultural, and cognitive implications.^{Footnote 5} While the preservation and transmission of documents and the allied topic of literacy in the medieval west garner much attention from scholars, in the case of Ethiopia, scholarly publication on the subject has been very few and far between. A harbinger of historical works to come was the Italian scholar Carlo Conti Rossini, whose work involved the edition and publication of charters. Particularly noteworthy are the corpus of charters and historical documents that Conti Rossini edited and published from the ancient churches of Šemäzana (1901) and Aksum (1909–1910) titled Liber Axumae. Yet Conti Rossini does not offer detailed treatments about the creation, preservation, and transmission of charters themselves per se. The corpus of documents in Liber Axumae was not historically constituted as an archive by the church of Aksum. Instead, Conti Rossini pulled the charters and other historical documents together from diverse collections and reorganized them into a coherent and chronological framework.^{Footnote 6} This chronological and logical framework, entailing as it does reconstruction and compression, flattens the original (dis)order in which documents were organized by the institutions that created them and does not advance our understanding of Ethiopian archival practice.

Far more perceptive and far more sophisticated in its analysis of Ethiopian documentary practice is Ignazio Guidi’s Gli archivi in Abissinia (1906), where he published a complex series of legal acts and charters from the eighteenth and nineteenth centuries. Guidi’s approach contrasted with that of Conti Rossini. The documents compiled in Gli archivi had actual organic existence as marginalia of a gospel manuscript belonging to the, now nonexistent, church of Hämärä Noh established in 1711. Guidi did not attempt to bring order to the formless Ethiopian records. He edited and published documents in the original order organized by the church, which provides evidence to the documentary practices undertaken by the church. More relevant to our discussion is Guidi’s observation that Ethiopian church records were the result of organized and conscious recordmaking and recordkeeping activity. He therefore rightly concluded that the various legal acts which studded the marginalia of liturgical manuscripts held by Ethiopian churches had to fall into the category of archives.^{Footnote 7} While a significant analysis of Ethiopian archival practice, Guidi’s contribution has never been appreciated by scholars until very recently.

Since Conti Rossini and Guidi our documentation has increased astronomically and reached a new level of maturity with the works of the late Donald Crummey. Crummey’s research on land and society in Ethiopia brought to light a prodigious amount of new sources.^{Footnote 8} He has been mining land related documents to great effect for essays on social history over the last three decades, but his early works on land and society leave out discussion of recordmaking and recordkeeping. In 2006 he published the only article on the social, political and institutional use of writing, in which he argued that in Christian Ethiopia, as was the case elsewhere, “literacy was power”, and “writing and written materials possessed great authority.”^{Footnote 9} In his more recent work on the subject, Crummey examines the production, reproduction, and preservation of charters in eighteenth century Ethiopia. Put briefly, for Crummey, the writing of a charter could go through distinct stages of evolution through a process of reenactment and repetition spanning several years. While referring to a single transaction, the reproductions were not the exact replications of a preexisting exemplar upon which they were based. Although they borrowed certain clauses and phrases from it, charter drafters in eighteenth-century Ethiopia radically refashioned an underlying template and in that way produced original texts, which, in turn, served as antecedents to yet other, original or authoritative, versions. These versions were made for archival purposes and placed by recipients of grants in prominent churches in the country. According to Crummey, the act of recycling, rearranging, and altering charters discernible in the eighteenth century was an old established archival practice, harking far back to the Ethiopian past.^{Footnote 10} This paper follows and lends support to the views and interpretation of Crummey’s work on recordmaking. Nevertheless, unlike Crummey’s work, it does so while emphasizing the existence of a significant change in Ethiopian documentary culture in the eighteenth century that was inspired by an allied change in legal culture.

Besides Crummey, there have also been other scholars who have investigated the questions of how records began to be kept and where they were deposited. Examples of such scholarship include works by Anaïs Wion, and Claire Bosc-Tiessé and Marie-Laure Derat. Wion has examined the organization, personnel and activities of scribes at the ancient Cathedral of Aksum in northern Ethiopia, where a royal chancery began to manifest itself in the sixteenth century. The church of Aksum produced, copied, and archived charters given by various Ethiopian kings in the name of other churches in the country. The very fact that these charters came down to us exclusively via Aksum’s archives indicates that the church had a functional writing office and served as a regional central repository. While it was certainly true that a writing organ of sorts existed at Aksum, Wion concedes that there is no clue to the existence of specialized competencies of scribes charged with the task of writing and keeping legal documents.^{Footnote 11} Through the perspectives of these scholars, we gain new insights into the history of archival formation in Christian Ethiopia. Yet the bureaucracy which produced documents and the legal institutions that inspired them is still poorly understood. The current literature ignores the existence of change in recordmaking practice and the language, forms and properties of documents themselves in tandem with changes in social and legal institutions in the country. In addition to a technical analysis of the documents themselves, reconstructing the history of archival formation in Ethiopia requires understanding the legal, social and political contexts within which documents were created, preserved, and categorized by churches. Let us begin then with an overview of the substance and forms of the documentation in church archives to gain a sense of the significant changes in scribal practice during the eighteenth and nineteenth centuries. Further, the legal and social institutions that inspired the new outpouring of documents will be offered. This will be followed by a discussion of the development of chancery offices and archives, and the uses of documents.

Types of Records and Proliferation Documents

The Gondärine period of Ethiopian history, which lasted between 1636 and 1769, witnessed a revolution in diplomatics since the thirteenth century. New genres of documents appeared and the structure and language of familiar records changed during this period. So much of the creativity in scribal practice occurred with respect to land charters or serit, the oldest of which dates to the early thirteenth century and remained the most widespread forms of legal documents until the eighteenth century. Charters from the pre-Gondärine period are streamlined and rarely offer details about the estates being granted and the legal rights of beneficiaries. They focus on the witnesses, beneficiaries, and contemporary royal and church officials and little else.^{Footnote 12} But from c. 1700 onwards, unlike those from the preceding centuries, charters came to have unusual features and began to be involved and address a new range of issues.^{Footnote 13} One of their novelties is a clause on the formula of land division between beneficiaries of grants and previous occupants of the land. Rulers granted land, technically known as rim, to the clergy often by displacing the previous cultivators’ hereditary rights of access over two-thirds of their agricultural land, only leaving them with one-third of their original land. Numerous charter drafters recorded the two-thirds and one-third formula of land division between the clergy and rural farmers conscientiously. The second novel aspect and a stock phrase of charters is the judicial clause. The clergy were commonly granted jurisdictional rights over the people who worked and resided on their rim land. A typical charter issued in 1766 to the church of Moṭa Giyorgis describes the jurisdictional right of clerical landlords over their subjects as follows: “The däbtära (clergy) would preside in judgment over their zéga. However, if they (the zéga) are implicated in cases involving murder, adultery, theft, and killing of animals, the case shall be seen by the aläqa and the liqätäbäbt.”^{Footnote 14} That meant that clerical landholders had the right to collect rent from their laborers, to administer them and the land in different ways, and to dispense justice in case of dispute. As a whole, Gondärine charters began to address in impressive detail the terms under which land was held, the rights and duties of beneficiaries of grants and rural farming people.

One can see also signs of the evident trend towards a new documentary tradition in Ethiopia the type of which was first seen in the mid-seventeenth century and became fully developed in the 1730s. This new type of document is known as mäzgäb, or land register or inventory. Land registers and inventories were typical accompaniments of Gondärine grants and drawn when churches took control of new agricultural land belonging to rural farming people. Records of this type offer detailed information about the lands involved in grants as well as their occupants. Compiled in the 1730s and 1740s following a “survey,” the two volumes register of the church of Däbrä Ṣäḥay Qwesqwam in Gondär, founded by Queen Mentewwab (reigned 1730–1769), are some of the most remarkable documents of this type. They provide much legal, social and geographical statistical information.^{Footnote 15} In a much quoted text, the contemporary royal chronicler informs us that three royal “chancery” officials were directly involved in the production of the registers:

The work of the three officials included writing descriptions of the topographical and physical details of every parcel of land available at the time of the survey and the list of the inhabitants of the land. In total the Qwesqwam registers give the name of 260 clerical holders of rim land and 3,899 tenants and serf-like laborers known as zéga.^{Footnote 17}

Besides illustrating the ways in which registers were created, the above cited instance demonstrates the growing concern with the physical and topographical details of property in eighteenth-century Ethiopia.

The eighteenth century also witnessed a further change in the variety and quantity of documents. We can see this in the appearance of private acts and contracts which were known under the generic name of däbdabé (lit. “letter”), issued by individuals in church archives beginning in the 1730s. In many instances, these acts are written down in the margins of parchment manuscripts, especially Gospels. In other instances, private acts are recorded in manuscripts created solely for the purpose of registering legal documents. Prior to the twentieth century, manuscripts consisting exclusively legal documents, including private acts, were referred to as yäbaher däbdabé (lit. “sea of letter”). In the twentieth century, they were categorized by churches as yärest mäzgäb or “register of inheritance deeds.”^{Footnote 18} Large bodies of private acts deal with purchases, or sales of residential plots in towns and agricultural land in rural areas. From 1730 until 1974, when the Ethiopian revolution abruptly ended private property, there is an unbroken series of contracts of sale in church records throughout northern Ethiopia. Written contracts of sale had not been the practice hitherto and their appearance during the eighteenth century was sudden and revolutionary. Other kinds of documents dealing with marriage settlements and wills appeared as early as the mid-eighteenth century. Finally, the late nineteenth century saw the appearance of documents of manumission and adoption related inheritance in the records of the churches of Gojjam province in particular. Documents of the latter kind made their way in large quantity to the archives in the first half of the twentieth century.^{Footnote 19}

The novelty of the Gondärine period was not limited to the increasing recourse to written agreements. The development of the vernacular as the language of record was part of the larger changes in documentary culture. Ge’ez was the language of legal records until the Gondärine period, during which Amharic had largely replaced it. With few exceptions, since the mid-seventeenth century, inventories, registers, private acts, and charters were written almost exclusively in Amharic supplanting Ge’ez. The pressure on Ge’ez from Amharic intensified in the nineteenth century when even chronicles and official and private correspondences began to be written in the latter including not only in the Amharic-speaking provinces but also elsewhere in northern Ethiopia.^{Footnote 20}

This significant change in the variety and quantity of documents is symptomatic of corresponding change in scribal practice and a growing confidence in the safety of church archives. But how do we explain the flood of unusual documentation and the change in the language and content of familiar records in Gondärine Ethiopia? Several factors account for the documentary revolution of the era. The innovation in documentation during the era in question must be viewed primarily as the reflection of a change in legal culture. The Gondärine period had brought a new dynamic to the system of landholding. In this regard the introduction and growth of rim property throughout northern Ethiopia in the late seventeenth century and thereafter is worthy of note.^{Footnote 21}Rim had brought new ways of material transactions and documentation. Typically, registers and inventories are the kind of records where the new property terminology of rim first emerged. As we have seen above for the Qwesqwam church, title to rim land was established after a royal official(s) had travelled to a locality to divide the land among the clergy and the rural farmers, write down the land, and lay out its boundaries. The instance of Qwesqwam demonstrates the intimate link between registers and rim property. What is more, rim was closely tied to the sudden appearance of the land market in the 1730s. Since it was the first to enter the market, so far as I know, the birth of land market in the eighteenth century and the accompanying proliferation of documents was driven by rim property.^{Footnote 22} Despite its relevance for the topic, this visible connection between rim property and the new genre of documents that emerged in the Gondärine era have not attracted any attention from those who write about Ethiopia’s archival history. My argument therefore is that in reading the charters and the accompanying private acts, registers, and inventories, we may see the relationship between new documentation and changes in the land holding system of Gondärine Ethiopian society.

In sum, the variety and number of documents in church archives of northern Ethiopia, the issues treated and the social background of the people involved in transactions is quite impressive. The majority of the people who created these archives are the ruling class, both clerical and lay. But the use of documents spread downward. Kings, queens, nobles, priests, famers, merchants, servants, slaves, and artisans all appear as parties to contracts in these documents. All kinds of property ranging from slaves to residential sites to houses to land to coffee to livestock are treated in these documents.

Churches preserved charters, registers and other documents because they were beneficiaries of land grants and needed documentary proof for their continued ownership. Some lay courts also came to have their own archives and writing offices staffed by professional scribes and managed by appointed heads. However, private persons, including noblemen and women, tended to deposit their records in church archives because churches enjoyed royal protection and were believed to be the safest places to keep documents. Many documents of private nature housed in church archives were created by people who were illiterate. Thanks to the writing offices and archives held by churches, which were available to anyone, the lack of literacy did not prevent individuals from documenting their transactions. Records were usually drafted by the clergy and monks of the Ethiopian Orthodox church because they were the only literate class prior to the twentieth century. As more and more individuals saw the value of protecting their property by means of carefully recording their transactions and depositing them in churches for safekeeping, Ethiopia would witness the development of permanently organized writing facilities and specialized groups of scribes across regions and churches. Examples of the kind of bureaucracy and procedure of production and authentication of documents that would have existed elsewhere in Ethiopia is found in the extensive records of the churches of Gojjam province in northwestern part of the country. Below I will show the gradual development of archives and writing organs by focusing on representative institutions from Gojjam.

The Development of Church Archives and Chanceries in Gojjam

The development of organized chanceries in the region and churches of Gojjam was very gradual. The trend towards full-blown chanceries emerged in the late eighteenth century where many churches in the province began keeping long-term record of written agreements and registers. During this early stage of their development, the number of scribes in most writing offices appeared to be low. Further, the task of drafting documents was often performed by unnamed scribes and their level of competency and specialized knowledge is unknown. But, during the late nineteenth and early twentieth centuries, the number, social standing, and privilege of chancery officials tremendously increased along with the growth of organized chanceries across the region. In this final stage of their evolution, royal and church chanceries in Gojjam were not only staffed by professionals, some of whom were distinguished church scholars and royal advisors, but had also developed to a point where they could be viewed as bureaucratic organs.

The church of Saint Giyorgis in the town of Moṭa was one of the first permanently organized archives in Gojjam for which we have information. As steps toward the development of archives, the church’s foundation charter issued in 1767 by its founder, Wälätä Isra’el, proved to be decisive. In the charter, Wälätä Isra’el, who was the daughter of Queen Mentewwab (reigned 1730–1769), devised rules that would remain fixed for two centuries for the clergy of Moṭa Giyorgis church and the people who live in the land under the church’s administration on a whole range of topics. One of the provisions in the charter with a bearing on issues of documentation is quoted below:

Although not directly stated in the charter, Wälätä Isra’el asked that registrar(s) levy registration fees on transacting parties in return for the service of drafting their written agreements and entering them into the church’s central registry. Part of the fees was used to contribute for the upkeep of the official directed by the gäbäz, who was in charge of the internal affair of the church. Written agreements were made before church officials and the church’s treasury store assumed the function of a writing office where documents were often, and still are, deposited together with sacred objects. The charter and the thousands of documents that exist in the archives of the church provide evidence that Moṭa Giyorgis possessed an organized writing office since at least 1767. Written down between the 1760s and 1974, the church’s register preserve both the records of the church itself and thousands of documents of personal nature.

The register was never bound together. Instead its leaves are organized into a total of fifteen separate quires. The quires are unequal size and the number of folios ranges from five to twenty. The texts of private acts are written often in two columns and in a variety of professional hands (see illustration 2). Unlike those dating from the nineteenth and twentieth centuries, documents drawn in the eighteenth century are for the most part recorded in the register without any regard for chronology.^{Footnote 24} Unfortunately, there is very little clear evidence about how the work of producing documents was actually organized within the writing office at Moṭa Giyorgis church. Even the names of scribes are not handed down to us as a general rule.

Illustration 1. A folio from the Qwesqwam register. Illinois/IES, 88. I.8, Mäzgäb.

Illustration 2. A folio from the register of Moṭa Giyorgis church. Yärest Mäzgäb, picture # 4089.

Although there is no direct evidence about the organization and specific duties of the chancery staff at Moṭa Giyorgis church, the process involved in the making of documents can be best seen by analyzing the structure of documents they produced. Some of the documents housed in Moṭa Giyorgis church are a sentence long while others are involved and extended. As a whole, however, documents were produced in standard form and their structure remained remarkably consistent in the period 1770s–1974. First, documents do have a dating formula and most are dated by reigning church officials as well as by the regnal years of kings and regional lords. Consequently, documents almost always begin with the list of contemporary church officials and/or kings, followed by the subject matter of the document. Church officials appear in the documents because they were an assumed part of the writing staff. Another common element of documents is the list of witnesses and guarantors to transactions. Typically, documents in the Moṭa Gyorgis archives were drafted in the presence of witnesses because property transactions customarily required witnesses to be valid. Finally, land sale documents give the names of the parties to the transactions, the price of the land and its general location. The exact physical dimension and precise location of the land is always left vague, demonstrating that it was not important at all for scribes and transacting parties. With the exception of some minor changes in the twentieth century, such as in the dating clause, the structure and the phraseology of documents remained formulaic throughout the period 1770s–1974.

The evidence gleaned from Moṭa Giyorgis’s foundation charter and archives raises two interrelated issues. First, legal culture in eighteenth-century Ethiopia required that registrations of transactions needed to have archives. The purpose of writing offices consisted of not only producing documents, but also of preserving and administering records held at the archives. Church archives were established as much to preserve material relevant to a specific church as to store and manage the records of private individuals and thereby serve as public archives. As the above evidence indicates and actual practice confirms, Moṭa Giyorgis church permitted anyone to use its writing facilities upon the payment of a registration fee. Second, the standardized production of documents is evidence that those who worked in Moṭa Giyorgis’s writing office had received scribal training. It is very easy to see a similar purpose and reason for the development of writing offices and archives for other churches in eighteenth century Ethiopia where written transactions appeared. A lively wave of composition of inventories, charters, and private transactions came a few decades later after our first evidence of the new type of documents in Moṭa in the 1770s. Such is the case with the churches of Märtulä Maryam, Däbrä Wärq, Yawiš Mika’el, and Qäranyo Mädhäné Aläm. In the latter, the issuer of its foundation charter and governor of Gojjam, Ras (lit. “head,” a politico-military title below a king) Märed Haylu (reigned 1796–1800), instituted 135 clerical personnel, including one unnamed scribe.^{Footnote 25} Yet the size of archives, their personnel, and organization are not at all clear until the late nineteenth century, when elaborate rules articulating the bureaucracy associated with recordmaking and preserving became available for many churches and lay courts in Gojjam.

Writing came to be used more extensively and in a more organized fashion than ever before during the time of King Täklä Haymanot (reigned 1874–1901), the hereditary ruler of Gojjam. Täklä Haymanot administered Gojjam from his capital Däbrä Marqos, which was named after the church he built in the 1880s. His reign is remarkable for the creativity in scribal art and artisanal works, church building and painting, and renovation it brought to Gojjam. Information about the writing activities and the personnel and organization of Täklä Häymanot’s court come from numerous charters issued by the king, and the eye-witness accounts of Täklä Iyäsus Waqǧera (hereafter Täklé). Täklé, one of the most conscientious members of the clergy, was a self-trained historian, skilled painter, and craftsman in the period.^{Footnote 26} Täklä Haymanot was the first attested lay court to have a permanent chancery and scriptorium. The king had at least eight clerks in his service who were organized into three different groups of scribes. The first group of clerks was comprised of four expert copyists or calli-graphers Afäwärq, Ešäté, Mänker, and Tägäññä. Based on information provided by Täklé, the head of the calligraphers and translators of scriptures was Wäldä Giyorgis, a highly learned church scholar and distinguished courtier of Täklä Häymanot.^{Footnote 27} As we will see below, Ešäté doubled as head of a chancery office and ṣähafé te’zaz (lit. “scribe of order”) or private secretary of Täklä Haymanot. Likewise, Tägäññä was appointed and served as the registrar and head of the Däbrä Marqos church in the period 1890s–1930s. The duties of these clerks included writing and copying parchment manuscripts in a scriptorium. They wrote in the beautiful and distinctive script called qum ṣehuf, which was learned by training or experience. All held the honorific title of aläqa (“leader”) and, according to Täklé, were served with special drinking cups made from buffalo horns at the king’s banquet hall to match their prestigious positions as scribes.^{Footnote 28}

Besides the above expert copyists, an organized chancery writing staff existed at the court of Täklä Haymanot. Three clerks, named Asrat, Dästa, and Sebhat Lä’ab, were recruited by the king from Däbrä Marqos as the chancery’s writing staff. The three scribes had to write documents originating in the court, expressly called däbdabé, a term which denoted both letter and legal acts during this period. The three däbdabé writers were directed by Ešäté, whom we have met above in his capacity as expert copyist. Ešäté served as royal secretary and head of the chancery office until January 1888 when he was killed in the battle of Sar Wuha, near the north-east shore of Lake Tana, fought between the army of King Täklä Haymanot and a Sudanese army. We do not know who succeeded Ešäté as head of the chancery office. We know however that Sebhat Lä’ab was appointed as the ṣähafé te’zaz of Ras Haylu (reigned 1901–1932), son and successor of King Täklä Häymanot, and served in that capacity until his death in 1925.^{Footnote 29} It is very likely that the king’s letters and the charters he issued in the 1880s and 1890s were produced in this chancery headed by Ešäté. Perhaps the greatest surviving work of Ešäté and his colleagues is the cartulary now stored in the church of Däbrä Marqos. Truncated copies of royal edicts, court judgments, charters, registers, and field inventories belonging to churches throughout Gojjam dating from the reign of Täklä-Haymanot are copied in this cartulary. The compilers abbreviated the inventories and charters issued by the king for archival purpose while the churches to which the grants were given held the extended versions.^{Footnote 30}

The demand for documents at the church and town of Däbrä Marqos also resulted in a special measure to be taken by Täklä Haymanot to establish a writing office to cater to the needs of the local population. As mentioned above, the king founded the church of Däbrä Marqos in the 1880s and renamed his capital Mänqorär after the new church. Täklä Haymanot built the church of Däbrä Marqos to serve both as a suitable place of worship for the local population and his necropolis where he was later buried in 1901. A formal chancery existed at Däbrä Marqos church since the late 1880s with the church’s treasury house serving as its chancery office. Special title for the head of the writing office was ṣähafé te’zaz (“scribe of order”), which was also a title held by royal secretaries. The first holder of the office of ṣähafé te’zaz of Däbrä Marqos church was Tägäññä, who, as we have seen above, was also expert copyist and writer of manuscripts. Tägäññä was a skilled craftsman and the favorite courtier of Täklä Haymanot as well as a personal friend of Täklé. The latter wrote extensively about the ingenuity and skills of Tägäññä and himself:^{Footnote 31}

Undoubtedly, Tägäññä and Täklé were the most creative and respected clerics of Gojjam, and both men were active participants of the extensive writing activity during the reign of Täklä Haymanot. Tägäññä served first as chief registrar of Däbrä Marqos church since 1880s and later as its head in the 1910s and 1920s. He ensured that his appointment to the office of registrar by Täklä Haymanot was recorded in a charter spelling out his duties and rights. He was charged with two primary duties. First, Tägäññä would draft land transactions initiated by individuals and record them in the central registry of Däbrä Marqos church. Second, he would record and keep track of the moveable properties of Däbrä Marqos church. To compensate for his writing service, Tägäññä was granted rim land in lieu of salary.^{Footnote 32} By the end of the century, a well-organized writing office was in place at the church of Däbrä Marqos. Transactions, exchanges, and agreements originating throughout Gojjam found their way to the archives of the church. Evidence for the activity of the writing office staffed by Tägäññä began in the 1890s, but its output would increase after the 1910s.^{Footnote 33}

As the chanceries became highly organized and staffed by professionals, an allied change was underway in scribal practice. The growing professionalism of recordkeeping and recordmaking in nineteenth-century Gojjam culminated in the establishment of new models of behavior in terms of the authentication of documents. The innovation in scribal practice lies in the use of seals to validate legal documents. Täklä Haymanot and his court officials were the first to adopt personal seals. Although seals were still a novelty for nineteenth-century Gojjam, they were already the objects of forgery in the first decade of the twentieth century. As skilled craftsmen with common interest in engraving, Tägäññä and Täklé were very likely responsible for the design and manufacturing of seals for Täklä Haymanot and his officials. As we will see below, early in the twentieth century, both men were charged with forging the seal of Afä Negus (lit. “mouth of the king,” a title given to chief judges) Yegzaw, King Täklä Haymanot’s confidant and chief justice.^{Footnote 34} Among other factors, it was the foundation of the church of Däbrä Marqos that induced this novel practice in the authentication of legal documents with seals. Täklä Haymanot concluded the process of founding and endowing Däbrä Marqos church in the late 1880s with one general charter authenticated by his unique and magnificent seal, which appears on top of the document. The new church also received three confirmation charters from archbishop Péṭros, Ečägé (title of the abbot of Däbrä Libanos monastery) Téwoflos, and Emperor Yohännes IV (reigned 1872–1889). These charters are authenticated by seals that bear the names and mottoes of their issuers (see illustrations 3 and 4).

Illustration 3. The charters of King Täklä Haymanot and Archbishop Péṭeros and Bishop Luqas. Illinois/IES, 89. XXI. 18–19, Wanna Mäzgäb.

Illustration 4. The charters of Yohännes IV and Téwoflos. Mäzgäb, Däbrä-Marqos, picture # 202.

Yohännes IV’s and Téwoflos’s charters survived in two versions. The first versions do not bear their respective seals and exist in the Däbrä Marqos church cartulary as one continuous document. Those bearing their seals were the latest to be prepared and written. In terms of design and the royal symbol of the Lion of Judah embossed in it, Yohännes’s seal is identical to Täklä Haymanot’s. Another, unembellished, seal appears in all Yohännes’s official documents, including his solemn diplomatic correspondence with European heads of states. It is safe to say that like that of Täklä Haymanot’s, the emperor’s seal was cast in Gojjam and presumably kept by the chancery officials in the king’s court at Däbrä Marqos. Yohännes might have given his permission to cast a seal in his name so as to authenticate his confirmation charter. The need for devising special seals for the king and emperor was therefore linked to the foundation of Däbrä Marqos church.

The use of seals and later signatures represented a new and distinct phase in the long-term evolution of archival practice. Gradually but steadily, the use of seals expanded outside of a small circle of church and government officials while many individuals adopted personal signatures to authenticate written agreements in their own names. Although not a universal practice yet, in some churches, documents came to be considered valid only if they bore the seals or signatures of participants in transactions. We obtain a very instructive picture of the authentication and preservation of documents and the personnel and organization of writing offices in the churches of the Zägé Peninsula near Lake Ṭana. Two of such churches where competitive recordkeeping system appeared to have existed were Ura Kidanä Mehrät and Mähal Zägé Giyorgis. Public archives with official supervisors of registration and writers had been established in both churches from at least the nineteenth century onward. The writing was made at these churches’ treasury stores, where professional scribes assigned for the task regularly worked. A July 1968 official document recording a new archival rule for Mähal Zägé Giyorgis church referred to its writing personnel as “yärest wel säratäñoč” or “Registrars (lit. “workers”) of Inheritance Deeds.”^{Footnote 35} At Ura Kidanä Mehrät church, four scribes can be shown to have been regularly employed for the purpose of recordmaking for the period 1938–1975. The writers were successively Bäyänä Tägäññä (1938–1952), Abäjä Haylä Iyäsus (1952–1956), Fäntahun Ṣagayä (1956–1968), and Dägu Ṣaḥay (1968–1975). All of them held the title ṣähafi or secretary and included their name in the documents they drafted. The intense activity of the chanceries of the two churches resulted in the production of four registers consisting of several thousands of documents.^{Footnote 36}

The churches of Ura Kidanä Mehrät and Mähal Zägé Giyorgis were among the early initiators of the use of signatures as a means to validate documents. As early as the 1920s the procedure for producing documents at Ura Kidanä Mehrät compelled that the actors to contracts, church officials, and witnesses approve the document with their signatures. In the case of Mähal Zägé Giyorgis church, the July 1968 rule of documentation required members of the writing personnel, who were drawn from the officials of the church, to authenticate documents being enrolled into the church’s register by their signatures. Prior to the new legislation, it was customary for only transacting parties and witnesses to sign documents. The signature of church officials became important to bring the church’s system of registration “in line with modern practice (lit. ‘civilization’).”^{Footnote 37}

Here it is interesting to take note that church officials regarded signing documents as a symbol of modernity rather than a simple necessity to validate documents. After 1968, documents entered into the Mähal Zägé Giyorgis’s archives were commonly authenticated, as was required, with the signatures of its writing personnel besides the transacting parties and witnesses. Illiterate persons authenticated legal acts by putting the impression of their fingers on documents. A folio from the register of Ura Kidanä Mehrät recording two contracts of sales from the early 1930s authenticated by seals and fingerprints are reproduced to illustrate the changes in scribal practice (see illustration 5).

Illustration 5. A folio from Ura Kidanä Mehrät’s register. Yärest Mäzgäb, picture # 051.

It is apparent that it was the need to obviate fraud and ensure the permanence of their archives that led many churches to adopt the use of seals and signatures. This is clarified in a new regulation of archives written on 3 March 1953 by the church of Täklä Haymanot in the Zägé Peninsula. The church had a corporate or communal seal and the new regulation pays much attention to the manner in which it ought to be used. By the new regulations of 1953, the church’s seal was to be kept by three trustworthy elder men, who were to be chosen by officials of Täklä Haymanot church and the people who lived in the land held by the church. If there was urgent and regular need for it, the church’s treasury store was the only valid repository of the communal seal. For further security the rule explicitly prohibited the use of the seal without the authorization of the head of the church (mämher) and the church’s administrator next in rank to the head (liqäräd) as well as the deputy judge (mesläné). The register, which the church kept, was also to have a single place of deposit at the treasury store, where it had to remain at all times. If and when the evidence of a document was required for any reason, the rule stipulated that the register should not be carried around. Instead, the register should be searched and a copy of the document in need should be made for anyone who might demand it. As additional precaution, any unauthorized register was not permitted to be deposited in the treasury store.^{Footnote 38} This regulation was perhaps intended to prevent fraud and the tampering of the register. In this way, the church devised a systematic rule of producing and authenticating documents, and retaining them in its archives.

Churches were as much interested in encouraging documentation as in collecting registration fees. The most insightful archival rule concerning the intimate connection between documentation and fees was devised by Ura Kidanä Mehrät church. Issued on 17 July 1960, the rule requires charging higher registration fees than before and dividing the fees among the various church officials who supervised the registration of documents. According to the new ordinance, transacting parties had to pay one Ethiopian dollar each for the store keeper, the scribe, the head of the church, the liqäräd, the mesläné, and the local level administrative official (čeqa šum). Registration fees also contributed to the upkeep of the holders of the offices of raq masäré and the gäbäz, who were in charge of the internal affairs of the church. In fact the idea of revising the procedure of registration itself came so that these two church officials gained a right to a share of the fees and through that alleviate their financial difficulties, which we are told they were facing. They received one Ethiopian dollar each from the registration fees collected by the church.^{Footnote 39}

As a whole, the bureaucratic reforms and rules of registration were instituted at the churches of Zägé for fast recordkeeping, the collection of fees, and the safety and permanence of archives. The procedures of documentation described for the churches above spread to several other institutions of Gojjam.

The conclusion to draw from the evidence presented above is that church records were the result of purposive recordmaking and recordkeeping activities. The chanceries and the bureaucracy of recordmaking and recordkeeping that was put in place in many churches in the twentieth century grew gradually and had at least two centuries of development behind it. Significant advance in the institutional means to produce, store and record documents resulted from the growing appreciation for documents, requiring a more regular presence of specially organized group of writers and prescribed rules of authentication and preservation of documents. Churches systematized their established procedure of registration to put all important transactions into writing and to preserve and transmit legal acts as well as to control what entered into their archives. As a result, writing offices at several churches were transformed into mature bureaucratic organs during the twentieth century. The above documentary and bureaucratic background provides the context to appreciate and understand the functions of the documents within courts and within the society that produced them.

The Uses of Documents

James McCann has argued that Ethiopian church documents are basically oral acts and have largely symbolic and commemorative importance.^{Footnote 40} I acknowledge that documents played commemorative role and have oral character. The fact that individuals were routinely called upon to witness transactions and serve as guarantors to transacting parties lends support to the oral character and symbolic role of church documents. Guarantors provided security against violation of agreements while witnesses ensured the validity of documents.

However, documents are not mere transcriptions of oral transactions. It must be noted also that the existence of witness list in legal documents does not necessarily indicate the oral character of legal acts. Witnesses to transactions were often church officials and their names are included partly because they were assumed to be part of the writing staff and partly because as a matter of habit. Records in church archives were valued by contemporaries primarily for their future administrative and legal use rather than for their commemorative purposes. The existence of many partially and completely deleted documents in church archives points to the same conclusion. The huge volume of documents in church archives leaves little doubt that a significant change in the standards of proof was occurring in Ethiopia as early as the eighteenth century.

Many individuals had a respect for documentation and it was the written nature of the transactions in which they were involved that was very important to them. This is most clearly attested, among countless other instances, by the property records of the families of Abbéto (a title given to the hereditary nobility) Gälawdéwos Wäldä Giyorgis and Ṣämru Asägehaññ. The property dealings of Gälawdéwos and his descendants in particular is the best studied and known subject. Written between 1706 and 1825, wills recording the transfer of the office of family leadership from one generation of this family to the next have been found in the archives of several churches.^{Footnote 41} One common theme running through the property records of this family is disputes among members, with documents being written at every dispute, often in several copies. In 1758, for instance, a violent dispute broke out between Daǧǧazmač (lit. “commander of the gate,” a politico-military title given for noblemen) Ešäté and his brother-in-law, Abbèto Bätru, over the office and the legacy of Gälawdéwos. Ešäté was married to Wälätä Rufa’el, who was the granddaughter of Gälawdéwos and Bätru’s sister. The complex negotiations between Bätru and Ešäté led to the redaction of twenty-seven copies of the dispute settlement. Yet the contest for inheritance and office between the various branches of this family continued unabated throughout the last half of the eighteenth and early nineteenth centuries. Even though written agreements and wills were not always effective to prevent adverse claims, Ešäté and his successors continued to write down their transactions.^{Footnote 42}

Writing became an obsessive concern particularly for Däǧǧazmač Haylu, Ešäté’s son and main inheritor. Haylu had compiled a major chronicle and his family genealogy in the 1780s and 1790s.^{Footnote 43} Then shortly before his death in 1809, Haylu designated his daughter Merṣit to be his principal heir and family leader. But Merṣit’s property did not pass unchallenged, for we find her engaged in a bitter dispute with her brothers Abbéto Gošu and Abbéto Wäldä Täklé. The quarrel between Merṣit and Wäldä Täklé was especially violent. The latter was disowned and excluded from the share of inheritance by Haylu, yet he forcefully took cattle and other properties from Merṣit. Under the pressure of relatives and elders, Merṣit agreed to readmit her brother into the family. Also, Wäldä Täklé received shares from the inheritance under the pain of losing it and irreversible rejection from the family if he mistreated Merṣit ever again. The negotiations between Merṣit and Wäldä Täklé were written in four copies and “deposited at [the churches of] Moṭa Giyorgis, Nazrét Iyäsus, Dima Giyorgis, and Däbrä Wärq.”^{Footnote 44} These churches are located within Gojjam, some of them hundreds of miles from each other. The recourse to multiple documents at every stage of the dispute to record a single transaction involving the descendants of Gälawdéwos is rare, but not isolated. Their faith in the written word was shared by Ṣämru whose father, Balambaras (a title given to low level administrative officials) Asägehaññ, was an active buyer of land at the town and church of Märtulä Maryam in Gojjam during the nineteenth century. In total, about thirty documents recording the property transactions of Asägehaññ and his descendants have survived via the archives of the churches of Märtulä Maryam and Däbrä Marqos. In 1899, long after the death of his father, Ṣämru required copies of his family records to be drawn up; and he then asked the permission of King Täklä Haymanot to deposit them, as he did, at the archives of Däbrä Marqos. The scribe copied the original documents accurately without eliminating the witness list and the dating formula.^{Footnote 45}

The evidence above points us toward some of the reasons why people wrote agreements in several copies and handled their documents with care. Undoubtedly Ešäté, Merṣit, Ṣämru and others like them understood documents as a primary proof of the legitimacy of their ownership. Whether drawn up centuries ago or few years in the past, documents of wills, contracts of sale and charters still had validity in the eyes of the law. Living witnesses could be used to attest that a transaction had taken place on the occasion of legal battles. However, the witness list ceased to be relevant once the witnesses are dead. If witnesses were what gave documents their legal force, then Ṣämru’s records would have been dead letters since most of the witnesses to his father’s transactions would have already died by 1899, when his records were copied and deposited at Däbrä Marqos church. Furthermore, the records settling the dispute between Merṣit and her brothers do not contain witnesses, reinforcing the fact that documents did not derive validity solely from witnesses. Once a document enters the archive of a church through the accepted and normal procedure, its validity is permanent and cannot be diminished by the length of time.

The instances presented above also offer answer to the question why individuals register and deposit their documents in several institutions. Undoubtedly, copies of their respective property documents were strategically placed in different churches by Ṣämru, Merṣit, and Ešäté in order to minimize the risk of the destruction of the originals. In the eighteenth and nineteenth centuries, church archives were not sufficiently secure and documents were vulnerable to theft, loss, removal, tampering, and destruction. One commonly comes across doctored and partially or completely deleted documents. In rare cases, even personal seals were forged. Two examples, among several others, illustrate that the threat of theft, forgery, and manipulation of documents was present and real. In the first instance dated to 1887, one Mäšäša Gäbrä Hiwät had deleted the names of his rival Grazmač (lit. “commander of the left,” a lower level political-military title) Märša from a document issued earlier and inserted his instead. The deletion was soon detected, and Mäšäša was brought to recognize his guilt at the court of King Täklä Haymanot.^{Footnote 46}

The second instance concerns the forgery of a seal. One of the most notorious, high profile, forgery cases was staged by Ras Haylu, who, as we have seen, was the son and successor of King Täklä Haymanot. Haylu had a rocky relation with officials of his deceased father. In particular, he suspected the former chief justice of Gojjam, Afä Negus Yegzaw, for conspiring against him with his powerful rival, Ras Mängäša Atikäm, the governor of Damot subprovince. Haylu ordered Täklé to forge Yegzaw’s seal. Then the seal was applied to a fake letter from Yegzaw to Haylu. Then Haylu secretly sent the fictitious letter to Mängäša, leading to the arrest of Yegzaw and other alleged conspirators. The letter characterized Mängäša as spineless and urged Haylu to attack the latter. What followed was a tremendous political chaos and scandal in Gojjam. When the news of this scandal reached him, Emperor Menilek II (reigned 1889–1913) had Haylu, Mängäša, Yegzaw, and several other individuals summoned into his court. Yegzaw claimed his innocence and the forgery would be revealed later. Among clerics implicated in the scandal and held responsible for the forgery were the registrar of Däbrä Marqos church, Tägäññä, and Täklé, both of whom we have met above. Tägäññä was cleared sooner while Täklé admitted to casting Yegzaw’s seal, but he defended his act on the ground that he was ignorant of its intended purpose and that he was simply “obeying the order from his lord,” Haylu. Menilek II exonerated Täklé from the charge while Haylu was found guilty and sentenced to prison and temporarily stripped of his governorship of Gojjam.^{Footnote 47} Although seals were harder to forge and churches generally avoided fraudulent documents, the two instances illustrate that forgery and deletion of documents could and did occur. Multiple documents could therefore provide security against such acts of forgery, deletion and mutilation.

Records were readily accessible to individuals who needed them and documents were created and stored in church archives to settle dispute in later time. A note concerning the use of the Däbrä Marqos church cartulary is very revealing of the intended function of its archives: “If the evidence of the register was required for any matter,” declares the note, “it shouldn’t be removed from the premise of the church. Instead it can only be consulted within the premises of the church.”^{Footnote 48} This note indicates that the provisions in the cartulary were used as a written code of law and any person interested in obtaining information in it had the right of search. There is also evidence of registers duly searched and being used as evidence in courts. For instance, an official document written on 13 December 1959 informs us that the two registers of Ura Kidanä Mehrät church were carried to Addis Ababa to serve as evidence, notably, in the Crown Court, the highest court of appeal presided by Emperor Haylä Sellasé I (reigned 1930–1974).^{Footnote 49} Why the evidence of the register was physically required rather than a copy of the specific document is not known to me. Whatever the reason, these instances underline the importance of documents in legal life.