Breakthroughs in Coding the Diversity of Medieval Liturgical Traditions

The liturgy was a web of numerous and highly orchestrated elements in the Middle Ages. It was a colorful theater both of extreme conservatism and creative power. It was lasting, costly, and involved—expressing and influencing the life of entire societies. It was a highly standardized field of activity, meaning that it dealt with a precisely defined, extensive, but restricted code of elements. This code comprised texts, melodies, objects, spaces, persons, and institutions that were diverse, but still conformed to inclusive conceptual categories. As such, they are ideal objects of digital experiments that try to build up webs of information without losing the uniqueness of their data.

Yet what we know of medieval liturgy mostly relies on written sources: service books that primarily contain the texts that must have been performed, to a lesser degree their melodies, and to an even lesser degree the instructions of how the ceremony had to be conducted.

The surviving books pose the opposite problem. On the one hand, there are too many of them. Tens of thousands have survived, ranging from tiny booklets with a few hundred pages to volumes of about a thousand folios. They far exceed the number that an individual or even the most diligent and well-provided research group could ever process. The other problem is that their quantity notwithstanding, the proportion of the surviving evidence is extremely uneven. Some churches and regions are overrepresented, while others remain blind spots. Some periods are documented, but others are almost inaccessible. Some types of ceremonies are recorded, yet others are obscure.

All this means that if we try to get a comprehensive and organized picture of liturgical affairs in the Middle Ages, we must narrow the source material on the one hand and supplement it on the other. The challenge is to select a limited but representative sample of evidence that a collaborating group of researchers can process within a reasonable time span. This sample, however, must counterbalance the contingencies of source preservation. It must contain balanced information about every involved institution, age, and ritual. It must not indulge in the well-documented ones but must explore those that are less documented. This is what Usuarium, a digital library and database for the study of Latin liturgical history in the Middle Ages and Early Modern Period, sets out to do.[1]

The Apropos: Experiences with the Chartvirgus Pontifical

To me, this was not originally a theoretical problem. I wrote my doctoral thesis on the few ordinals, a rubrical type of liturgical book, surviving in Hungary.[2] It was then that I realized that a condensing approach is not limited to the ordinals and the normative texts of the High Middle Ages, but it is already present in pontificals around the first millennium. Luckily, two such eleventh-century pontificals open the series of extant Hungarian service books.[3] When I compared them to my fifteenth-century ordinals, mostly printed, I was struck by the similarity of those rites that are the most interesting and detailed among the ceremonies common to both book types, ordinals and pontificals.

These are the extraordinary rites of the ecclesiastical year: Candlemas, Ash Wednesday, and, first of all, Holy Week. Ordinals are burdened with an important but, at first sight, tedious list of incipits (opening words) that is missing from pontificals, while pontificals contain many extra ceremonies like the dedication of churches or the ordination of priests that are rarely documented by any ordinal. So, the aforementioned rites are the common section of the two clusters, and they, surprisingly, prove to be basically the same in Hungary, or more precisely, in its archiepiscopal see, Esztergom, for the five hundred years between the eleventh and the sixteenth centuries.

This congruency between historically distant sources of identical lineage encouraged me to think in liturgical uses—the assumption that the same ecclesiastical institution possessed a distinct and continuous tradition of how rites were performed in its central church, providing a model for several subordinate churches as well. This sort of tradition is also called rite and custom, but it is most frequently labeled a use, and I preferred to apply this latter term because I associated the rite with a larger, structural category, and the custom with a smaller category, a local reification of the use.

Thinking in uses was neither a purely empirical discovery nor an absolutely theoretical idea. Liturgical use is an original, innate medieval concept, although it only began to be articulated with the emergence of ordinals, and only became ubiquitous by the late Middle Ages. Articulation, however, only means that service books explicitly refer to their uses in their titles, prefaces, or colophons. It does not mean that the phenomenon itself was contemporary with its first mentions. On the contrary, my eleventh-century pontificals long preceded the books that first mentioned their uses. In Hungary, it did not happen before the late fifteenth century.

There was, however, an anomaly. The next generation of Hungarian service books, those that stem from the twelfth and thirteenth centuries, deviates from that line of tradition that I could identify as continuous throughout more than five centuries. If the early pontificals had not survived, one would be inclined to think that the use of Esztergom was a product of the fourteenth century, from which point we find an uninterrupted continuity. Yet the eleventh-century pontificals demonstrate that the twelfth and thirteenth-century books do not differ from the Esztergom lineage because they are earlier, but because they belong to different, although still Hungarian, traditions.

Therefore, the challenge lay in two things. One was to demonstrate the continuity (or to interpret the change) of uses by comparing sources from different ages but belonging to the same tradition. The other was to prove the distinctiveness of their lasting features. Of course, every use has many constant peculiarities, but most of them come from the shared Roman heritage of the Latin churches. Only those features are rightly interpreted as distinctive marks that are common to their specific tradition but are unique in a wider context, or at least belong to a well-identifiable minority. It was here that I had to depart from the Hungarian source material and assess the objective range of western liturgical diversity.

This happened in the early years of the past decade, an epoch when a British colleague anticipated that the digitization of sources will have a particularly significant impact on the liturgical field.[4] It indeed did. Until that time, liturgical scholarship primarily drew on authoritative editions that suggested a far more centralized picture of liturgical affairs than what can be justified according to an unbiased consultation of sources. It was the same time when a large number of digital copies became freely accessible online. I started to collect them on my private drive and to use their information as a background for interpreting the continuity and distinctiveness of Hungarian sources.

It meant no database-building at all, but I was immediately faced with the problem of systematically describing the sources and their contents. In order to find the relevant books, I had to develop a standardized nomenclature of origins, book types, and ages. As the precise contents of pontificals can hardly be predicted, I also needed an indexing system that registered the types of rituals actually contained in particular books. For example, the extraordinary rites of Palm Sunday may equally occur in pontificals, missals, rituals, ordinals, customaries, and so on, and may be missing from any of them. To spare much time and effort, I labeled the chapters of my books with uniform tags that indicated which rites I can or cannot expect to find in them.[5]

There soon emerged two problems. Again, one was of a restrictive and the other of an expansive nature. My collection began to grow larger than what could be easily handled with the usual facilities of a private computer. I asked a programmer friend of mine, Károly Gossler, to store the collection in a safe place and enable me to browse the file names comfortably. This resulted in a digital library, the predecessor of the current Usuarium database. On the other hand, I was unhappy with the range of sources that I could access freely. I could not reconcile myself with the deficiency that while I had countless sources from the Staatsbibliothek in München or the Bibliothèque Nationale in Paris, I had scarcely any from Italy, Britain, or Southern France. Consequently, I applied for academic grants that could provide the financial support to acquire a proportionate sample, regardless of the costs. Fortunately, I won what I could and became a wholesale customer of digitized service books all over the world.[6] Another benefit of the financial support was that I could employ my colleagues and students and process a much more extensive amount of sources.

The Principles: Mapping and Printed Service Books

If we had to sum up Usuarium in a single expression, I would say that it is a map-view of liturgical diversity. The basic idea behind the collection of sources, their selection and ranking (the schedule and the actual level of their processing), and the classification of the records is the vision of a map. It represents the web of ecclesiastical institutions: cathedrals and monasteries with their circles of influence. Even centralized, international religious orders are assigned to geographical sites: their headquarters or places of origin. Mapping was my principal discovery that enlightened the whole process and helped to model the concept of uses: a long time may pass away, book types may change, and rituals may vary, but the places where all this happens do not move. You can better understand the significance of the mapping model if you contrast it with other visual models that have been popular in liturgical scholarship. One could be a timeline, emphasizing historical development, and the other a tree diagram, emphasizing textual transmission from book to book.

No source can be uploaded to Usuarium unless it is assigned to geographical coordinates. Books of doubtful origin get the coordinates of their earliest preserving institution or, in really desperate cases, the rough coordinates of a region or a country. If we discover something about the source’s closer affiliation, we can easily change its attribution to different or more precise origins. Depending on the depth of processing, thousands of records can belong to each source, yet every record is linked to the coordinates of the book’s origin. Consequently, any kind of information obtainable in Usuarium can be visualized on a map.

The second most important principle is the precedence of printed sources, an offshoot of the mapping concept. For such an approach, the information value of a source is determined by the reliability of its location. The more firmly a book belongs to a certain spot, the higher it ranks in our hierarchy. Scholars of the Middle Ages typically neglect or even despise service books of the early printing press period as if they marked the end of the Middle Ages. The tendency is strengthened by the fact that, in the past decades, musicological research did the most in describing and understanding medieval liturgy, and manuscripts preserved their precedence in musical transmission. Notes seldom appear in printed books, and if they do, they mostly accompany texts that are less interesting for musicians.

It is true that prints were the outcome of a revolutionary paradigm shift, justly associated with the Renaissance, the Protestant Reformation, and the early modern age. Yet they were very effective in realizing the liturgical ideal of the High Middle Ages that manuscripts had been unable to perfectly achieve. They could disseminate the use of a cathedral over the subject diocese or the established use of an order in its religious houses.[7] Bishops and general chapters did not use the new technology for innovation. Their principal goal was not to change the liturgy but to regulate it. In this sense, early printed service books are not the first fruits of a new paradigm but the synthesis of the old.

Most frequently, authorized persons or ecclesiastical bodies commissioned the printed editions. At least, they selected the best manuscript at their disposal and sent it to the printer. Under more advanced circumstances, they carefully revised the master copy and collated it with other reliable sources. We may rightly suppose that these prints preserved the best of the late medieval manuscript tradition, but even if they did not, they have an undeniable advantage: the direct reference to their origins.

This was the reason why I opted for a reverse chronology. In contrast with earlier scholarship, we did not start our work with the earliest surviving books, but systematically selected one book for each use, preferring self-defining sources, by default prints. Prints almost cover the full geographical extent of the Latin liturgy, thus, with their help, we can layer a typological grid on the map. We can pose any question, check the distribution of any phenomenon, investigate patterns of variation, and this method enables us to assess the degree and nature of liturgical variation at least in the late fifteenth and early sixteenth centuries. Even the most skeptical medievalists will agree that Usuarium offers reliable snapshots of the situation as it was at the end of the Middle Ages. By doing so, however, we can easily find distinctive traits and trace their presence back to earlier periods. Wherever we can prove continuity and distinctiveness, we can extend the validity of the late medieval snapshot to earlier centuries.

There are, certainly, some limitations. Neither mapping nor prints are wonder weapons that solve every problem once and for all. Sure, in a few cases prints were instrumental in introducing reforms, but this was not typical. More difficulty comes from the deficiency of prints in specific realms like Italy and Britain, where most of the dioceses abandoned their uses by the printing press period and adopted the use of Rome and Sarum respectively. We encounter a similar problem when researching the monastic uses of particular monasteries. Only a few of them published prints, while their majority conformed to an international Benedictine practice. The Balkans and the Baltic region have no prints at all. And, at the earliest (in the tenth century), we totally lose the vestiges of uses. This can be explained by the fewness of the sources and their uncertain origin, but it seems that the concept of uses itself was the product of the second millennium.

Yet, admitting all the limitations, Usuarium is a great step forward. A typology that is based on a geographically proportionate grid of uses can effectively assess the range of diversity and highlight those traits that distinguish one use from the other. What I initiated to better understand my Hungarian sources is now available to anyone.

A Report: Achievements, Ongoing Projects, Plans

My lonely work evolved into a funded research project in 2013. First, we wanted to test and refine our methods through a couple of preliminary projects. We knew that it is a common error of liturgical scholars that they generalize their theories. Findings in one specific field do not necessarily apply to further fields. With this precaution before our eyes, we worked with different types of sources and rituals each year until 2018.

First, we entered the temporal Mass propers, the regularly changing texts of the Eucharistic liturgy of the annual cycle that conforms to seasons and weeks, nicknamed the “Missalia Project.” Its principal novelty was that it equally comprised the chants, the lessons, and the prayers of the Mass.[8] All three types of propers had relevant editions and some databases, but they had not been united in a coherent system.[9] Moreover, earlier works focused on the earliest source material. Our emphasis on uses and the corresponding re-evaluation of late sources helped to discover many unknown items and to interpret variation more in terms of typology than chronology.

In the subsequent years, we worked with the Psalter parts of breviaries that regulate the basic cycle of the Divine Office, with calendars and sanctorals that listed the dates and the feasts that were celebrated in particular institutions, with occasional ceremonies of the annual cycle, recorded in rituals and pontificals besides missals, and, finally, with the non-Roman sources of the Ambrosian, Mozarabic, Gallican, and Beneventan traditions. Some of my colleagues are still working on side projects like the dedication of churches, the coronation of kings, monastic initiation, baptismal, matrimonial, and funeral rites, etc.

Four years ago, we obtained the best academic grant that is available in Hungary. We chose to go back to the Missalia Project, but with three extras. First, we extended our scope to the full content of the missals, including not only the temporal parts but sanctorals, commons, and votives as well. Second, we realized that we have no right to exclude any content that figures in our sources. The information system we use must not be selective: we must develop a system that can include any sort of liturgical information. Third, we did not consider the Missalia Project as the final goal. It is only the first step towards a comprehensive store of liturgical knowledge, including the divine office and the occasional rites. In sum, the aim is threefold comprehensiveness: comprehensiveness concerning geography and institutions, as I detailed above; generic comprehensiveness, meaning the totality of liturgical rites; and historical comprehensiveness, i.e., the incorporation of ancient evidence up to the first extant service books and norm texts of the first millennium.

Parallel to the synthesizing work, we are engaged in editing the liturgical heritage of our homeland. It is always good to stay in close contact with the intricacies of one single tradition and so to resist the tendencies of oversimplification. From Hungary, we publish critical and practical editions of all types of service books: missals, ordinals, pontificals, sacramentaries, rituals, graduals, and notated breviaries.[10]

In quantitative terms, we have now about 1,500 sources, of which more than 1,000 have an index. Figuratively, the Missalia Project stands at the slopes of the Pyrenees, i.e., we are almost ready with every European missal outside of the Iberian Peninsula; some southern French and Breton sources are still underway. This means the full missals of more than 180 institutions with more than a million records.

In qualitative terms, the database does not merely provide facilities for querying texts. It provides tools for querying according to assignments, i.e., liturgical positions that can be filled by various texts. This is the basic tool to assess and interpret variation and it is helped by statistics and mapping functions.

Theoretically, the chief result consists of the systematization of liturgical subjects, actions, genres, and texts. We have identified the actors of the liturgical theater as ecclesiastical institutions and compiled a structured list of those who left behind service books. On an empirical ground, we have compiled a list of ceremonies, containing every type of liturgical activity recorded in our sources, from the basic ones like the mass and the office to the rarest blessings and processions. We have defined the generic structure of the rituals, serving as a skeleton for the actual texts, and we are working hard on the standardization of the texts themselves, which will connect variants of essentially the same texts. We have both the intact evidence and the highly standardized code for its interpretation, and this is, I think, what digital technology can offer to liturgical research.