The Heroic Age

A Journal of Early Medieval Northwestern Europe

Founded 1998   |   ISSN 1526-1867

Issue 13 (August 2010)  |   Issue Editor: Larry Swain

The English Glosses in Eleventh-Century Anglo-Saxon Manuscripts

Kathryn Powell

University of Cambridge

© 2010 by Kathryn Powell. All rights reserved. This edition copyright © 2010 by The Heroic Age. All rights reserved.

§1.  In 2005, the Manchester Centre for Anglo-Saxon Studies concluded an AHRB-funded study of scripts and spellings in eleventh-century English manuscripts, directed by Professor Donald Scragg and Dr Alex Rumble. The aim of this project was to investigate the long-held assumption that a standard form of written English existed in the eleventh century by cataloguing a representative sample of spellings from manuscripts written at different centres throughout the period. The primary outcome of that project was the Manchester C11 Database, an online resource for the study of eleventh-century scripts and spellings. The database, which is freely available online at, provides users with a tool for studying and comparing spellings used throughout the period and correlating those spellings with the scribes who used them. This database currently contains spellings drawn from more than a thousand manuscript copies of texts in Old English, and new texts continue to be added. Because of constraints on time and technology, that initial project took the difficult decision not to investigate or catalogue occasional glosses and marginalia. Now, however, with two years of AHRC funding, a new MANCASS project, directed by Professor Scragg, has the opportunity to complete that work. This new project, "The English Glosses in Eleventh-Century Anglo-Saxon Manuscripts," aims to study this important body of late Old English and incorporate the results of this research into a larger study of late Old English spellings.

§2.  The original project focused on main texts and the spellings they present in eleventh-century manuscripts, and these constitute a relatively well-documented source of late Old English. Although not all editions of these texts include spelling variants from all available manuscripts, texts surviving in eleventh-century manuscripts generally have been edited and are widely available in some form. The same cannot be said of the large number of corrections, glosses and annotations written between the lines and in the margins of eleventh-century English manuscripts. For example, many editions do not represent spelling corrections in their apparatuses, especially where those corrections are not judged by editors to be "authorial." Likewise marginalia, if it is judged to be unrelated to the main text in whose margins it appears, are often left out of modern editions. Yet these brief snippets of text, taken together, stand as an important witness to late Old English writing and readership, and constitute the final frontier of unedited Old English. If collected, indexed, and made widely available, this body of work would undoubtedly alter our understanding of late Old English and our perception of the state of language and learning in the eleventh century.

§3.  The eleventh century in particular is a crucial period for the study of the development of English language and culture. The end of the tenth and beginning of the eleventh centuries saw the flowering of the Benedictine Reform movement begun by Æthelwold and his compatriots in the mid-tenth century. As a result of the teachings of these reformers, exponentially more people were reading and writing the vernacular than a century before, and they were writing in a fairly standardized and recognizable orthography that did not necessarily reflect local dialectical differences. But throughout the eleventh century, historical accident has numerous opportunities to disrupt what would otherwise be a natural continued development of this standardized language. First the series of Viking attacks that ultimately resulted in the Danish conquest of England, and later the Norman Conquest and the suppression of English's "official" status must have resulted in an influx of non-native speakers—and probably readers and writers—of English, and a necessary degree of linguistic change and diversity. We know all of this in theory, but without a reasonably complete record of the language as it was written throughout the century, we cannot know the true nature or extent of that change in practice. It is this gap in our knowledge that this project seeks to redress. The interest of the project is not narrowly linguistic; if we are able to draw a more accurate picture of the development of the language in this period, it may help us better understand the forces of historical and social change that influenced this development.


§4.  Because modern editions do not consistently record corrections, glosses and marginalia, this project must collect its data directly from the original manuscripts. To that end, I am spending two years looking at manuscripts which are known to contain, or which may contain, marginal or interlinear additions in English written in the eleventh century. Because of the difficulty of dating the compressed hands of such additions with any precision, the project defines "eleventh century" loosely, and its scope extends backward into the last decade of the tenth century and forward into the beginning of the twelfth. Eleventh-century additions can, of course, be made in manuscripts that were produced earlier, so it is necessary to consider all manuscripts written up to the end of the eleventh century. Thankfully, it is not necessary to actually view each one of these manuscripts. While catalogue descriptions cannot always be relied upon to tell one if and where eleventh-century glosses and marginalia occur, they do frequently and reliably state that no additions or corrections have been made to a manuscript. On this basis, it has been possible to draw up a list of manuscripts that need to be seen—85 in all. Each of these has to be examined page by page to ensure that any marginal and interlinear additions are recorded. In the first year of the project, I have viewed 36 of these manuscripts—many of them among the largest and most heavily glossed that need to be viewed—and have recorded approximately 5,600 marginal and interlinear entries in these volumes. These entries vary greatly in terms of their nature and extent, from single-letter spelling alterations made above the line of writing to marginal additions that continue over several pages. All of them nonetheless provide evidence of writers at work in the eleventh century, each with his own definite ideas about English spelling, responding to vernacular texts in a variety of ways.

§5. Currently, I am storing data about the annotations I find in spreadsheets which will be made available on the C11 Database website at the end of the project. For each correction, gloss, or marginal item I find in a manuscript, I record enough data to help anyone using these spreadsheets later to place that item in the context of its manuscript, in relation to the text it supplements or alters, and in relation to other work by the same scribe where such an identification is possible. For each entry, I record the text added (whether that is a single letter or a lengthy marginal passage), its immediate context in the manuscript (i.e., the word or words that it alters), its location in the word or line that it alters (if it is a single letter, for example, whether it is meant to be added initially, medially, or finally in the word it alters), its location on the manuscript page (e.g., above or below the line, in the top, bottom, left or right margin), and any information that can be provided about the scribe responsible. Every entry is indexed to the page and line numbers where it is found in its manuscript, to an item number in the description of that manuscript in Ker's Catalogue, and to a Cameron number and sense unit number in the Toronto DOE Corpus so that one may easily see what text is being altered and see if and how the alteration affects the sense of the text. So, for example, the data collected on a series of entries from London, BL, Cotton Caligula A. xv might look like this:

1 2 3
Ker Item # A. a A. 1 A. r
Page # 126r 131r 135r
Line # 1 31 12
Main Text earfoð hwylc hit Her forðferde eadward kyng
Annotation lice dæg 7 her com (abbrev.) willelm
Location on Page above line above line on line
Location in Word/Line finally between words after words
Hand main scribe main scribe later hand
Cameron # B23.3.2.1 B20.1.2.3 B17.2
Sense Unit # [000100 (1)] [000400 (8)] [002600 (1066.1)]

The entry in column one, for example, tells us that, in line one of folio 126r, the scribe has written the word "earfoð," and then later added the suffix -lice above the line. If one follows up the reference to Ker's Catalogue, one learns that this alteration is made to "a lunar prognostic" (Ker 1957, 174), and the reference to the DOE Corpus provides the context in which the altered word occurs (Förster 1912, 32). By way of contrast, the entry in column three shows that, on line 12 of folio 135r, the main scribe of this part of the manuscript has written a record of Edward the Confessor's death in 1066 and that a later reader felt that some mention should be made of William the Conqueror's accession and added "7 her com willelm" in a blank space at the end of the line. The reference to Ker tells us that the note is added to a set of annals mainly relating to Christ Church, Canterbury, that accompanies a table of years from 988 to 1193, extended at a later date to 1268 (Ker 1957, 175). By referencing the Cameron number, one can find the full text of these annals as edited by Liebermann (Liebermann 1879, 4).

§6.  The process of data collection so far has brought to light a few challenges for the project. For example, it was initially agreed that as I record annotations, I should include in each entry a note classifying the annotation as a correction, an alteration, a gloss, commentary, etc. While these types of annotation seem distinct enough in theory, in practice one finds significant overlap and grey area between them. In particular, I have found the distinction between corrections and alterations impossible to maintain. Corrections are a subset of alterations, and to call something a correction is to imply that the main scribe has in some way got it wrong—miscopied his source, misspelled a word, or deliberately departed from his exemplar. Sometimes, when the main scribe has made an obvious error such as a dittography, it is possible to make such a judgment based solely on a viewing of the manuscript in question. More often, however, it requires more detailed knowledge of a text's whole transmission history to know whether an annotation is correcting a text to match some other copy or version, or whether the annotator is simply altering the text to meet his own needs or personal preferences. Even deciding whether or not an annotator's substitution of one word for another is intended as a gloss can sometimes prove problematic. Particularly where an annotator writes as an apparent gloss a word which is not a synonym for the word it annotates, but which still makes some sense in context, one has to ask if this suggests that the annotator was struggling with the language of the text in a way that led him to gloss the word incorrectly, or if he was simply revising. Therefore, in an effort to avoid prejudicing the results of the project and preventing others from drawing their own conclusions about scribal intentions, I have tended to use the catch-all category of "alteration" to describe most annotations, unless the annotator's intentions are quite clear.

§7  The most troubling challenge faced by the project is posed by the difficulty of identifying scribes based on the compressed hands of marginal and interlinear writing. It is easy enough to see when a scribal hand changes in the main text of a manuscript; one has a large sample of writing that is typically of a uniform size upon which to base such a judgment, and the writing of the two scribes in question is usually adjacent. When alterations are made by one of the main hands of the manuscript, as in examples one and two in the table above, it is usually possible to identify the hand because one again has a large body of writing with which to compare it. Even here, however, the aspect of the scribe's writing and his choice of letter forms may change significantly when he is writing in cramped spaces between lines or in the margins of a page. When a number of later readers make annotations in a single manuscript, it can become difficult or impossible to sort out how many scribes are at work and precisely where each is working. If, for example, the only annotation on a particular page is an i added above the line, it will likely be impossible to identify that i with any particular scribe at work elsewhere in the manuscript. Marginalia tend to provide larger and more uniform writing samples than interlinear annotations, and the former can sometimes offer helpful clues to identifying the hands of the latter. Likewise, ink color and the use of marks of revision are often more helpful than letter forms for identifying hands in compressed writing. Nonetheless, except where the hands of annotations are quite distinctive, any scribal identifications made by the project are necessarily more suggestive than conclusive. This limits some of the claims the project will be able to make about exactly how many scribes are at work in the surviving manuscripts from the eleventh century, how uniform any particular scribe's spelling choices were, etc. Nonetheless, even being able to make tentative suggestions about the answers to these questions based on a broad survey of the evidence represents an advance in our understanding of eleventh-century writing and readership.


§8.  The project plans to disseminate its findings in a number of forms. As I've already mentioned, we hope to make the raw data that I collect publicly available through the C11 Database website. Originally, we hoped that it would be possible to integrate this data into the existing database, making it easy to search the data on glosses and marginalia and to compare the spellings used in them with those found elsewhere in the manuscripts. The loss of our database programmer from the earlier project, however, as well as a general lack of technical support from the University of Manchester has forced us to reduce our goals. We still plan to make the data available via the same website as the database, but possibly in the form of links from the manuscript descriptions currently in the database to PDFs of my spreadsheets describing annotations in those manuscripts. So, for example, if one looked up the description of Cambridge, Corpus Christi College 162 in the database, it would note that the manuscript includes numerous eleventh-century additions and alterations, and would include a link that would allow the reader to view a PDF of a spreadsheet that catalogued all of these annotations in a form similar to that of the table above. As PDFs, these spreadsheets would be searchable and could be downloaded or printed, so it is hoped that they will be useful in this form to a broad audience of early medievalists. From the point of view of studying the development of the language, it would obviously be more useful to be able to include this data in a larger, searchable database of eleventh-century English, but we hope that making the data available in any form will help to advance scholarship in this area.

§9.  In addition to the website, we will also disseminate our findings through the more conventional means of print publications and conference presentations. We have already presented preliminary findings at the MANCASS 2008 Easter conference, "Leaders of the Anglo-Saxon Church" and at the International Medieval Congress at Leeds 2008, and at the 2009 International Congress on Medieval Studies at Kalamazoo.

§10.  We also plan two substantial publications proceeding from the project. Donald Scragg is currently compiling a handlist of eleventh-century scribes writing English, including the scribes of main texts in manuscripts, annotations, and documents. Obviously, all of the difficulties of positively identifying, dating and localizing scribal hands complicate this project. Nevertheless, it would be difficult to write any account of the development of the language in the eleventh century without identifying and referring to the work of individual scribes, so some attempt to differentiate between them and devise a reference system for them seems necessary. Professor Scragg intends to compile such a reference work and to publish it in printed form, to be completed after the conclusion of the project in August 2009.

§11  At the same time, I am writing a monograph in which I examine glosses and marginalia in a selection of the manuscripts I am studying for evidence of how actual readers in the eleventh century received and responded to a variety of English texts. Marginalia, in particular, provide evidence of readers responding to texts by commenting directly upon them or by adding material of their own composition or drawn from other sources, so that we have a fairly direct window onto the scribe's thoughts about the text and, in many cases, the other texts he had available to him. Even minor corrections, by their mere presence, provide evidence that a text was being read at a particular date and, where a corrector has only marked up selected items from a manuscript, that some texts were not necessarily being read or revised. Taken collectively, such corrections can potentially provide us with a picture of a scribe's intellectual interests. Although I only plan to survey a small number of manuscripts in this study, I hope to be able to show readers responding to texts in surprising ways and thereby to demonstrate the usefulness of returning to the manuscripts for speculating about the audiences of early medieval texts.

§12.  The project will continue and will collect data from many more early English manuscripts—data which is likely to change the way we view late Old English and the history of the English language. Hopefully, this data will shed new light on the degree to which the Danish and Norman conquests of England affected the development of the language, and may perhaps provide us with some contemporary insights into the effects of these events on the lives of English people. Furthermore, even from the data collected already, it is clear that we will have to rethink our notions of how many people were reading and writing Old English in the eleventh century, what they were reading, and why they were reading it. Such insights should more than demonstrate the value of interlinear and marginal corrections and annotations, and will perhaps even persuade future editors of early medieval texts to accord them a slightly less marginal status.

Works Cited

Förster, M. 1912. Beiträge zur mittelalterlichen Volkskunde VIII. Archiv 129: 16–49.  [Back]

Ker, N. R. 1957. Catalogue of manuscripts containing Anglo-Saxon. Oxford: Clarendon Press.  [Back]

Liebermann, F. 1879. Ungedruckte anglo-normannische Geschichtsquellen. Strassburg: Karl J. Trübner, 3–8.  [Back]