Introduction
I’ve recently been pondering the ways that we categorize and analyze liturgy. What are the various bits, bobs, and elements that we can use to make sense of liturgical texts that retain meaning across time? To put a finer point on it, how would we go about creating a taxonomy for liturgical texts that could apply equally well to a 9th century Anglo-Saxon liturgical miscellany and the ’79 Book of Common Prayer? Specifically, can one be created that can be used for the analytic markup of liturgical materials for digital use?
The best way to begin is to not reinvent the wheel. Other, wiser, and better-informed minds than mine have worked around this issue even if the scope was not entirely the same. People who work on manuscripts have already had to do a lot of this work, as have the pioneers who began the work of digitizing the material that’s available on the web in a variety of forms. Whenever we encounter someone who has done both, we know that we’re in the right company! As a result, the best conversation partner for this work is Andrew Hughes, author of the magisterial Medieval Manuscripts for Mass and Office: A Guide to their Organization and Terminology (MMMO) and the ground-breaking Late Medieval Liturgical Offices (LMLO) digital project which is only partially online through this search resource.
The MMMO presents a set of abbreviations and symbols used both within it and in the LMLO. These are presented and discussed on pages ix-xxv. The two questions we need to answer are these: 1) do his temporal restrictions (the years 1200-1500) invalidate his scheme for this proposed general use that hopes to expand out to include material produced 4 centuries in each direction? 2) to what degree can his material be adopted and where must it be adapted?
I’d suggest, first, that the bones of the schema here can be preserved with the important caveat that this taxonomy is only appropriate to Western liturgies that have their basis in the Latin-speaking tradition. That is, I think it’s helpful and useful as long as we don’t try to apply it to Eastern Orthodox liturgies or various modern protestant liturgical traditions that eschew the historic liturgy. With this caveat in mind, I do believe that there is sufficient continuity within the historic western liturgy that question 1 can be answered satisfactorily.
Moving to the second question, a certain amount of adaptation will be required. While Hughes uses abbreviations that sometimes overlap and must be distinguished by typeface or other methods, abbreviations are no longer necessary in a scheme intended for use within a modern markup situation. Indeed, not only are they not necessary, they’re not desirable either as their use hampers the natural-language legibility of a marked-up document. (That is, you have to refer to a key instead of just looking at the source text and understanding what’s going on…)
Too, as is perfectly appropriate for his project, Hughes uses technical liturgical terms appropriate for the high medieval period. Given the much wider span that we’re looking at, not all of these terms may possess or retain the particular technical definitions they had in Hughes’s span. As a result, while I think most can be used as is, it’s worth considering where we might need to build in tolerances to handle this situation. (This is one I’m holding in reserve—I can’t think of any cases like this at the moment but reserve the right to discover some…)
When I read through Anglican prayer books, I see a number of elements that don’t seem to fit properly within the elements that he offers. In order to maximize meaningful tagging within these documents, we will need to add elements that go outside the scope of the texts Hughes was considering. For instance, an enduring heritage of the 1552 book is lengthy addresses by the presider to the congregation; I don’t recall these in earlier liturgies nearly to the degree that we have them now.
One of the most helpful aspects of Hughes’s scheme is that it offers, by means of several lists, different levels of elements necessary to distinguish between whether you’re talking about a specific section of a book or a service itself or about an event within a given service. This aspect definitely needs to be retained and will be reflected in how we present and order our proposed elements.
Proposals
I envision this categorization scheme at use within the TEI P5 guidelines for xml markup of humanities texts. Thus, the section-level and service-level elements would appear as “type” descriptors within <div> elements; intra-service elements would be noted as types within <seg> elements tied to <interp> and <interpGrp>
sections.
Section Level Elements
This level would identify sections within a book and be located within <div> tags. The following list is a modification of Hughes’s List 1e with the inclusion of some items from 1f.
- Aspersion
- Common of saints
- Dedication feast
- Hymnal
- Invitatorium
- Kalendar
- Ordo
- Ordinary of the Mass
- Kyriale
- Psalter
- General rubric (as a section rather than a specific direction)
- Sanctorale
- Temporale
- Votives
- Tonary
- Prayers (as a collection/grouping of prayers of the same sort)
Service Level Elements
These labels would identify the specific rituals or services being described and would operate at the <div> level. This is a substantial expansion of Hughes’s List 1c.
- Office
- First Vespers
- Compline
- Matins
- Prime
- Chapter
- Terve
- Sext
- Nones
- Vespers
- Morning Prayer
- Noon Prayer
- Evening Prayer
- Lamp Lighting
- Eucharist
- Baptism
- Confirmation
- Confession
- Unction
- Marriage
- Burial
- Procession
- Exorcism
- Rite (Thinking specifically, of the ’79 BCP, the alternation between Rites I & II is a service-level phenomenon and needs to be captured at that point.)
- Service (This is a generic catch-all for anything else not covered. Differentiation would appear in a subtype.)
Liturgical Elements
These items would appear within interpretive <seg> elements. This is a modification of Hughes’s List 1a, List 1b, and 1f
- antiphon
- invitatory
- gospel canticle
- psalm
- lesson
- chapter (a brief, generally one sentence bit from Scripture in an Office)
- first lesson
- second lesson
- prophecy
- epistle
- gospel
- announcement
- conclusion
- psalm
- invitatory
- canticle
- invitatory
- gospel
- dialogue
- versicle
- response
- intercession
- petition
- response
- litany
- section
- petition
- section response
- orison (the English term for an oratio—basically any kind of a number of relatively short prayers typically found in early medieval mass-sets and most commonly encountered in Anglican books as collects. However, not all so-called collects fit the formal criteria for collects; preserving “orison” as a general-use term conveys the concept even when the contents fall short)
- collect
- secret
- post-communion
- sermon
- hymn
- anthem
- canon
- proper preface
- consecration (not just of Eucharistic elements but of baptismal water, etc.)
- distribution (again, could be Eucharistic elements, baptismal water, ashes, etc.)
- fraction anthem
- agnus dei
- creed
- communion chant
- gloria
- doxology
- gradual
- introit
- kyrie
- offertory chant
- sanctus
- tract
- responsory
- verse
- repetenda
- doxology
- benediction
- rubric
- prayer (this is a generic catch-all when the presider addresses God when none of the above seem to fit.)
- address (this is a generic catch-all when the presider speaks to the people and nothing above seems suitable.)
- response (this is the generic catch-all when the people speak and nothing above seems suitable.)
Concluding Thoughts
I believe this lays out a useful initial plan for liturgical tagging. That having been said, I’ve done enough liturgical research and enough coding to know that this represents only a tentative beginning. As well defined as things seem to be at the beginning, the true usefulness of the scheme is revealed in the coding itself. As a result, this provides a framework. Now some coding actually needs to be attempted to determined where this works, where it fails, and where the framework needs to be reconceived.
I studied with Hughes, and recently had a conversation with him about a project that I have in early medieval (10-12C) liturgical music, and he would be the first to say that MMMO and LMLO are _late_ medieval in orientation, and shouldn’t be forced to earlier periods, where they only fit “in general.”
As much as I like the idea of a TEI-like XML encoding, every time I’ve tried to encode an (early-) medieval manuscript, I run into a philosophical/aesthetic problem. The ontology (in the sense of the names we use for elements) is fairly easy to get pretty close, but the ontology (in the sense of being) is problematic. TEI is barely two-dimensional (it can encode trees, but not networks) and it is functionally impossible to encode an implied three- or four-dimensional structure in any meaningful way. That doesn’t mean that a TEI schema for liturgy is impossible, or more so, that it wouldn’t be useful, but it’s clearly going to be limited and incomplete.
Two things on the Hughes system—1) yes, it’s for the late medieval which is why I introduced some modifications into it. 2) sometimes “general” is better than overly precise especially with a system as extensible as TEI xml where greater precision can be added as needed.
I’m a little unclear on what you’re getting at with your dual use of ontology… In what way does it fail to capture the sense of being?
I’ve not encoded liturgical materials with xml, but I have done some early medieval homilies that way and was pleased with the amount of data of varying types that could be encoded—from the material particulars of the manuscript to the content itself and variations between manuscripts. My chief interest in the liturgical taxonomy is setting up a system that is functional for not just identifying but assembling liturgical elements into coherent services. Based on my breviary and related projects, I have a pretty good idea of what’s required and I think it’ll work.
That having been said, I’d love to hear suggestions for what would work better!