Two Millennia of Making, Using, and Keeping Written Texts for Readers: Then, Now, and in the Future

FOR WRITTEN TEXTS to be legible, there are various conditions that need to be met. First, a text needs to be written in a language the reader can understand. This does not only refer to the linguistic guise of the written text; it also has to do with the register in which a text is written. Secondly, a reader must know the alphabet and the script in which the text is written. They need to be able to read both ‘uppercase’ and ‘lowercase’ letters, and any abbreviations used. Thirdly, there is punctuation with the other ‘paratextual’ signs which help the reader to give voice to the text – even if that voice often be interior only.1 Fourthly, the spaces between the words are important. And fifthly, there is the way the text is laid out on the page. The way these five conditions of legibility changed over time, from papyrus via parchment to paper and, lately, the ‘electronic page’, have consequences for our ability to read ancient texts nowadays. Their makers simply had different ways of doing things than the makers of modern printed or electronic texts. We can of course make excellent digital copies of medieval manuscripts, but as modern users of written texts we expect to see things on the page that are different from the expectations of our medieval predecessors. This means that, although we may be capable of appreciating the beauty of manuscripts as objects, the availability of digital copies of the same is not enough to allow modern readers to make sense of the manuscripts’ contents. And it will be hardly surprising if readers in the future may find themselves in a similar position when they are confronted with the texts made for us today – texts which we have no difficulty in reading.

Reading Then

Let us start with an example of how a twelfth-century monk had to deal with the ‘grammar of legibility’ of some of the old texts he found among the centuries-old documents he tried to make sense of.

Around the middle of the twelfth century, Eberhard, a monk of Fulda, received the assignment to collect the old charters of his monastery in a cartulary. He finished this huge work, in two volumes, between 1150 and 1160. In a short introduction to the second volume, he wrote that he found among other things “some testaments written by those who because of their age could hardly be read”.2 This was not only because of their wear and tear, but also because these small documents were written in old writing which was “unknown to modern [readers] and hardly legible”.3 They were also composed with “wondrous simplicity”.4 Fulda had been founded in the eighth century, when it had been the monastery where the ancient Boniface had retired before his last fateful mission to Frisia and his martyrdom in 754. Documents written in what is called ‘insular script’ could be found among the ancient parchments Eberhard had to work with’.5 Already in the ninth century, texts in this script (then often named scriptura Scottica, ‘Irish script’, after the group of insular scribes who had made the deepest impression on the contemporary imagination) had been copied in the more legible Caroline minuscule. To quote Eberhard once more, this time from the introduction to the first volume, “not all charters had been easy to read because they had been so old and because of lack of experience with Irish writing and bad quality of the letters”.6

Eberhard clearly had problems with reading texts in outmoded script. This does not mean that he was a bad reader of manuscripts, who was not up to the task he had been set by his abbot. Far from it, witness the cartulary he produced, which is today kept in the Marburg archives as the two-volume Codex Eberhardi.7 From the ninth century onwards, monastic scholars had allowed the reuse of the parchment of texts in ‘Irish script’, for instance as binding materials for new manuscripts in the then current Caroline minuscule.8 If texts contained in the manuscripts to be reused had been deemed valuable, a copy of them was made in a script the readers of the time could actually read – even if there might be some mistakes in a copy, due either to the form of the letters of the exemplar or to the use of abbreviations that were no longer current.

Eberhard’s problems had been caused by changes in what is called the ‘grammar of legibility’. This notion has been developed, among others, by the palaeographer Malcolm Parkes.9 To be legible to a reader, the five visual aspects of written texts using an alphabet which we mentioned at the beginning of this article are important: the language of the text; the forms of the letters of the alphabet (including any abbreviations and ligatures that may have been used); punctuation;10 the use of space between words;11 and the layout of the text. Each of these five aspects has a history of its own; together, these histories form the history of legibility.

In the twelfth century, Eberhard experienced difficulties in understanding the early documents of Fulda mainly because of the unfamiliar forms of the letters of the alphabet used for ‘Irish script’; possibly the use of insular abbreviations did not help either. Of minor importance was the Latin of the documents: he was surprised by their “wondrous simplicity”,12 and possibly also by what he may have considered usages bordering on mistakes. He did not comment on the other three aspects. Insular scribes had invented word separation, for the excellent reason that the texts they were interested in were almost exclusively written in Latin; texts written in insular vernaculars were very rare indeed. To understand the Latin of their exemplars, they needed to know where one word ended and the next word began, in order to make sense of the language of the texts. In Antiquity, word separation in Latin texts was not necessary, as most readers were speakers of Latin who could make sense of written texts also without word separation. Nevertheless, there had been some problems when a text was meant to be performed in public. For this reason, a prior reading (a praelectio) would be necessary; for the actual ‘reading aloud’ the text would have been memorised. Latin texts would also be provided with a modicum of punctuation; this would be developed in the early Middle Ages, and Eberhard would have been helped in his reading of the old documents of Fulda by the punctuation symbols he would have encountered. As for layout, finally, he was able to understand when he was dealing with charters, presumably because their layout offered him some clues. In the case of the royal charters he copied in the second volume of his cartulary, he carefully drew the signa (i.e. the monograms) of the kings which he could see at the bottom of the parchments.13

It is not too difficult to work out which of the visual aspects of the documents Eberhard needed to copy were posing problems. His own grammar of legibility can be deduced from the cartulary he produced. It is well organised, with each of the two volumes being preceded by a table of contents and a very short introduction. The layout shows artwork of some quality.

From Reading Then to Reading Their Texts Now

After Eberhard, the grammar of legibility continued to develop and change. For instance, Eberhard considered Latin to be the language most suitable for the documents he copied. Nowadays, we prefer to read texts in our mother tongue – or at least in a modern language we have mastered during our education. We want to understand what we read, and one prerequisite of understanding is knowledge of the language of the texts we want to read. In the Middle Ages, there had been a distinction between reading (legere) and understanding (intellegere). One could be considered capable of reading a text when one was able to voice it. That is why of the very few primers surviving from late medieval England, booklets meant for instruction in reading that were given to ordinary children of a tender age – and unavoidably were destroyed by them – two out of four are in Latin.14 These primers, as well as the deluxe copies meant for royal or at least noble children, invariably start with an ABC, sometimes showing the various forms some of the letters might take (e.g. ‘s’ next to ‘ſ’) followed by a few letters that were added to the Latin alphabet to render some of the sounds of English not known to Latin (e.g. ‘ð’, ‘þ’, or ‘ƿ’). This is followed by the Lord’s Prayer, the Ave Maria (Hail Mary), and other prayers. When someone proved able to give voice to these texts, the learning of reading was finished. If someone had to understand the texts read as well, in Latin, then attendance at a school teaching Latin was required. Clearly, more people were able to read than to understand what they read.

In our modern grammar of legibility, reading, the ‘voicing’ of the text on the page, is hardly considered reading, unless one is able also to ‘understand’ the text one reads. And to get access to the latest knowledge and insights on a topic, Latin is hardly ever necessary. We use a (selection of) vernacular(s). We also prefer the texts we read to use only the letters of the alphabet we are familiar with. In our texts, the use of capitals and lower-case letters is regulated, albeit with minor variation from one vernacular to the next.15 We would, just as Eberhard, be very unhappy with texts in which word separation was not rigorously applied. We have become used to divisions and subdivisions of longer texts that are much more standardised than they were in the twelfth century. We have also become used to numbered pages in our printed materials, and to indices at the end of our printed books in which we can find out at a glance on which page information of interest to us can be found. But the development of indices from the thirteenth century onwards is deemed less necessary by us when we can use a searchable digitised copy of a text online. That is the reason why we no longer use concordances of texts such as the Bible – indeed, most in our age group might be at a loss to define what a ‘concordance’ is (or was). And finally, if, at the foot of the page in our printed books (or in the left or right margins of our digital texts), we find numbered annotations, these may harken back to the interlinear and marginal glosses in medieval manuscripts (roughly from the ninth century onwards). We will recognise texts in which numbered annotations occur, either as footnotes or as endnotes, as scholarly or scientific texts. The layout of these texts has become so standardised, that we may recognise them even without reading a single sentence.16

The changes that the grammar of legibility has undergone in the past two millennia have consequences for the way we have to deal with its earlier phases when we want to make texts from the past available to present-day or future readers. This problem has been around ever since scholars started to think about providing their contemporaries with reliable editions of earlier texts. It is becoming more urgent now that digital means of reproduction enable us to encounter any texts in manuscript form on the screen of our computers – texts whose grammar of legibility, however, was different from our own.17And yet, these digital images seem, quite often, to adapt unwittingly to our present-day grammar of legibility, when they ‘flatten’ the image of parchment pages which never are wholly flat, or when they adapt the colour of the parchment of documents or manuscripts from different origins to a ubiquitous hue of whitish yellow (or yellowish white). Possibly, however, this falls under a yet to be defined ‘grammar of making texts’ rather than under the grammar of legibility.18

Whenever scholarly editions are made, editors have to come to grips with the differences between their own grammar of legibility and those of the makers of the (handwritten) texts their editions will be based on. The choice of which features of the grammars of legibility they come into contact with are to be included in their editions, is problematic, as the medieval manuscripts they have selected for use in their editions show perceptible differences in this respect. The rules imposed by the editorial boards of the better-known series of source publications suggest adaptations to the modern grammar of legibility, both with respect to the normalisation of spelling, punctuation, the use of capitals, and layout. Word separation is not mentioned, it is taken for granted that it will be used.19 These choices are understandable, as the object of the exercise is providing a present-day audience with a legible text. It is also inspired by the kind of edition that aims at reconstructing, as much as possible, the text an author once wrote, on the basis of all extant manuscripts of this text, using what is called the ‘stemmatic method’.20 As the surviving manuscripts of a text may be of widely divergent origins and dates, they may have considerable differences in their grammar of legibility as well. It will be impossible to retain the specific habits of the scribes of all manuscripts deemed important in the making of the edition, which is printed as a single text that is the result of the textual criticism applied by the editor. All that may be done to remedy this somewhat, is to provide photographs of some pages of the manuscripts that have been used. For the reconstructed text, the choice of the editor’s own grammar of legibility therefore suggests itself with some urgency.

The stemmatic method is, of course, by no means the only method for making editions. The palaeographer Bernhard Bischoff finished the “Introduction” of his Latin Palaeography: Antiquity and the Middle Ages with the words “[e]very manuscript is unique. Our aim should therefore be to recognise that uniqueness, to consider the manuscript as a historical monument and to be sensitive to its beauty, especially when its script and illumination are of a high aesthetic order”.21 This invitation to consider all manuscripts to be equally worthy of attention has led to editions of texts which are preserved in a single manuscript only. In such cases, reproducing the text as it occurs in its manuscript is a possibility. Indeed, there is a trend to publish not just scholarly editions of individual texts, but integral manuscripts, with all texts exactly as they occur in the manuscripts chosen for this treatment. This idea is in fact an old one, going back at least to the eighteenth-century edition of Domesday Book.22 The series of diplomatic editions of Middle Dutch manuscripts which were meant as material units by their makers and which contain at least two texts (making them verzamelhandschriften, ‘collective manuscripts’) presents texts as they occur in the manuscripts, with only very minor changes. As the manuscripts lack editorial introductions or commentaries, so the editions do without them as well. Nor is there a translation in a modern language (modern Dutch would have been helpful to some, if not most potential readers of the editions in this series).23 Although the grammar of legibility of the makers of the manuscripts is on the whole respected, only specialists can use the volumes published in this series, as the grammar of legibility of modern readers is not catered for. Under the circumstances, an annotated digital facsimile of the manuscripts might have done the job equally well, if not better.

Even digital facsimiles of texts, however, are defective if they do not provide both an edition and a translation in a modern language, because the knowledge of languages of modern readers requires it. One cannot assume that modern readers know enough Latin, or Middle Dutch, or Old Norse, or whatever the language of the texts in a digitised facsimile may be. Without at least a diplomatic edition (in which letters which were current at the time a manuscript was written, but are no longer in general use today, can be easily rendered thanks to Unicode) and a translation, only specialists will be able to use our present-day efforts.

Maybe we ought to take a leaf out of the editorial habits of runologists. They tend to provide a photograph of the runes they edit, with, underneath, a rendering of the runes as they read them, followed by a transliteration of the runes, followed by the text according to the scholarly conventions for rendering Old Norse, and finally followed by a translation in a modern language (usually English).24 But runic inscriptions tend to be very short, different from most texts that survive on parchment and paper…

Reading in the Future?

We do not know what future readers of medieval handwritten texts will require of their editions. In all likelihood their grammar of legibility will change due to their increasingly reading texts in a digital format on screen. We can only hope that the editions we can make nowadays to satisfy the requirements of contemporary readers will be legible for a considerable time to come.

Bibliography

Primary sources

Codex Eberhardi, Hessen, Staatsarchiv Marburg, K425 <https://arcinsys.hessen.de/arcinsys/digitalisatViewer.action?detailid=v3353926>.

Codex Eberhardi, Hessen, Staatsarchiv Marburg, K426 <https://arcinsys.hessen.de/arcinsys/digitalisatViewer.action?detailid=v782853>.

Domesday Book, ed. as Domesday Book; seu, Liber Censualis Willelmi primi regis Angliae, inter archivos regni in Domo Capitulari Westmonasterii asservatus..., 4 vols. (London: Record Commissioners, 1783-1816).

Published secondary materials

Bischoff, Bernhard, Paläographie des römischen Altertums und des abendländischen Mittelalters, 2^nd edn. (Berlin: Erich Schmidt Verlag,1986), trans. as Bernhard Bischoff, Latin Palaeography: Antiquity and the Middle Ages, trans. Dáibhí Ó Cróinín and David Ganz (Cambridge: Cambridge University Press, 1990) <https://doi.org/10.1017/CBO9780511809927>.

Bunčič, Daniel, and others, Biscriptality: A Sociolinguistic Typology (Heidelberg: Universitätsverlag Winter, 2016) <https://doi-org/10.1515/jhsl-2016-0022>.

Clanchy, Michael, Looking Back from the Invention of Printing: Mothers and the Teaching of Reading in the Middle Ages (Turnhout: Brepols Publishers, 2018) <https://doi.org/10.1353/pgn.2020.0011>.

Hunnisett, R.F., Editing Records for Publication (London: British Records Association, 1977).

Kwakkel, Erik, Books before Print (Leeds: ARC Humanities Press, 2018) <https://doi-org/10.1017/9781942401636>.

Mertens, Th., ‘Richtlijnen voor de uitgave van Middelnederlandse Verzamelhandschriften uit de Nederlanden’, in Het Geraardsbergse handschrift: Hs. Brussel, Koninklijke Bibliotheek Albert I, 837-846, ed. by Marie-José Govers et al. (Hilversum: Verloren, 1994), pp. 173-191.

Mostert, Marco, ‘Celtic, Anglo-Saxon or Insular? Some Considerations on “Irish” Manuscript Production and Their Implications for Insular Latin Culture, c. AD 500-800’, in Cultural Identity and Cultural Integration: Ireland and Europe in the Early Middle Ages, ed. by Doris Edel (Dublin: Four Courts Press, 1995), pp. 95-119.

Mostert, Marco, ‘Latin Learning and Learning Latin: Knowledge Transfer and Literacy in the European Middle Ages’, in Theory and Practice of Knowledge Transfer: Studies in School Education in the Ancient Near East and Beyond, ed. by W.S. van Egmond and W.H. van Soldt (Leiden, 2012: PIHANS 121), pp. 25-37.

Parkes, M.B., Pause and Effect: Punctuation in the West (Aldershot: Scolar Press, 1992) <https://doi.org/10.4324/9781315247243>.

Parkes, M.B., ‘The Contribution of Insular Scribes of the Seventh and Eighth Century to the “Grammar of Legibility”’, in Grafia e interpunzione del Latino nel Medioevo: Seminario internazionale, ed. A. Maierù (Rome: Edizioni dell’Ateneo, 1987), 15-30; reprinted in M.B. Parkes, Scribes, Scripts and Readers: Studies in the Communication, Presentation and Dissemination of Medieval Texts (London and Rio Grande: The Hambledon Press, 1991), pp. 1-18.

Raaijmakers, Janneke, The Making of the Monastic Community of Fulda, c. 744-c. 900 (Cambridge University Press, 2012) <https://doi.org/10.1017/CBO9781139030366>.

Reynolds, L.D., and N.G. Wilson, Scribes and Scholars: A Guide to the Transmission of Greek and Latin Literature, 2^nd edn. (Oxford: Oxford University Press, 1974),

Richtlijnen voor het uitgeven van historische bescheiden samengesteld in opdracht van het Nederlands Historisch Genootschap en van de Rijkscommissie voor Vaderlandse Geschiedenis, 6^th revised edn. (The Hague: Nederlands Historisch Genootschap, 1988).

Saenger, Paul, Space between Words: The Origins of Silent Reading (Stanford: Stanford University Press, 1997).

Sources chrétiennes: Directives pour la préparation des manuscrits, 2^nd edn. (Lyons: Institut des Sources Chrétiennes and Les Éditions du Cerf, 1978).

Traube, Ludwig, ‘Perrona Scottorum: Ein Beitrag zur Überlieferungsgeschichte und zur Palägographie des Mittelalters’, in Kleine Scxhriften, ed. Samule Brandt (Munich: C.H. Beck, 1920), pp. 95-119.

Zilmer, Kristel, and Jesch, Judith, Epigraphic Literacy and Christian Identity: Modes of Written Discourse in the Newly Christian European North, (Turnhout: Brepols Publishers, 2012) <https://doi.org/10.1484/M.USML-EB.6.09070802050003050402090402>.

Marco Mostert is Professor of Medieval History at Utrecht University. In his research, he devotes attention to the history of medieval literacy and communication. He edits the series Utrecht Studies in Medieval Literacy, published by Brepols, in which more than fifty monographs and collected volumes have been published so far.