Why code-switch in writing? Findings from sociolinguistic research into classical and modern texts
Alex Mullen, All Souls College, Oxford

Studying Classics has always entailed an appreciation of Greek-Latin bilingualism and bi-culturalism and yet it is only recently that we have extended our view to engage fully with modern bi- and multilingualism theory and practice (Adams 2003, Biville, Decourt and Rougemont 2008, Cotton et al. 2009, Mullen and James 2012). This has led to demands for increasingly sophisticated analyses, in particular, of written code-switching. All our direct sources of linguistic code-switching are written and Classicists have been at pains to demonstrate that terminologies, models and concepts created by modern bilingualism research, which has traditionally had speech as its primary focus, can be applied with caution to textual evidence. Differences clearly do exist, however, between spoken and written code-switching, for instance concerning intentionality and the context of interaction (reader versus interlocutor), and we still have to define these comprehensively.
We need to understand how and why code-switching occurs in written media. How variable are the functions, types, constraints and frequencies of code-switches across different written genres and how should we explain the variances we encounter? By analysing the socio-cultural context of the multifarious practice of code-switching in a range of well-known genres, for example code-switching in the vast Ciceronian corpus, Roman comedies and inscriptions, we might be able to form views about the motivations to code-switch and be more confident in identifying code-switching, as opposed to other processes which result in mixtures of languages. This information might then come to our aid when we are treating less-well contextualized material which has undergone complex multilingual and multi-author transmission processes.

Adams, J. N. (2003) Bilingualism and the Latin language. Cambridge.
Biville, F., Decourt, J.-C. and Rougemont, G. (eds.) (2008) Bilinguisme gréco-latin et épigraphie. Lyon.
Cotton, H. M., Hoyland, R. G., Price, J. J. and Wasserstein, D. J. (eds.) (2009) From Hellenism to Islam. Cultural and linguistic change in the Roman Near East. Cambridge.
Mullen, A. and James, P. (eds.) (2012) Multilingualism in the Graeco-Roman worlds. Cambridge.

Bilingualism and text transmission in mediaeval texts: bilingualism in mediaeval England
Laura Wright, University of Cambridge

As the aim of the project is to trace the linguistic and cultural backgrounds of language switching, my contribution will be to outline the several text-types which show this phenomenon in medieval England, as detailed in Schendl and Wright eds (2011: 22). Before 1066, language switching is found mainly in non-royal charters and poetry and is mainly intersentential. After the Anglo-Norman administration took over in 1066 however, language switching is found in sermons, accounts, inventories, wills, scientific and medical texts, religious texts, poetry, drama and verse, and business accounts – texts both formal and informal. The switch points vary from larger syntactic units to within the word. The main languages are English, Anglo-Norman French and Medieval Latin, but switching with other languages occurs too, e.g. with Welsh in Wales, and with Middle Low German in documents produced by Hanseatic merchants in London.
I shall then consider mixed-language business accounts, which are trilingual (that is, the matrix language is either Medieval Latin or Anglo-Norman, and the switched language is English, but the Latin and French are influenced by – receive borrowings from or into – each other). This text type does indeed show chronological differences in the incidence of language switch, which may be due, in part, to the demise of naturally-acquired Anglo-Norman. It also shows borrowing and reborrowing between Medieval Latin, Anglo-Norman, and Middle English, to the extent that one cannot safely distinguish between these languages at a lexical level.

Herbert Schendl and Laura Wright, eds. 2011. Code-switching in Early English. Berlin: Mouton de Gruyter.

Bilingual texts in mediaeval German
Elvira Glaser, Deutsches Seminar, Universität Zürich

In my talk I shall start with an overview on the corpus of bilingual texts and the different types of „bilingual text transmission“ in the mediaeval German-speaking area. I shall discuss the criteria used for a classification of these text types (from glosses to code-switching as two extremes in this connection). In a second part I shall focus on the rich corpus of mixed prose (Mischprosa) of Notker III from St Gall analyzing the specific use of Latin and Old High German together with a discussion of the different methods used for the analysis of the Latin elements. I will present some results of a grammatical analysis inspired by recent contact linguistic research on code-switching, which has been rarely taken into consideration up to know. A a conclusion, the motivation of Notker’s language use will be discussed taking up several proposals put forward by the traditional Notker research.

The Altalemannischen Psalmenfragmente and other Germanic Interlinearversionen of the Psalter
Alderik H. Blom, Corpus Christi College, Oxford

The last two decades have seen the radical re-evaluation of the early ninth-century Old High German Interlinearversionen. Whereas older scholarship tended to regard these continuous vernacular glosses written between the lines of several key Latin texts within monastic culture (notably the Althochdeutsche Benediktinerregel and the Murbacher Hymnen) as more or less successful attempts at translation, scholars such as Nikolaus Henkel, Lothar Voetz and Ernst Hellgardt have recently argued for a rather different approach.
According to them, the extant Interlinearversionen only reflect the written part of a complex process of Texterschliessung which included oral communication as well. Rather than incomplete or faulty translations, they should instead be regarded as classroom texts, providing grammatical and lexical clues for the understanding of the Latin target language. Thus, they should no longer be seen as translations existing independently from their Latin originals, but instead as integrated bilingual text-ensembles, in which every vernacular lemma is directly bound to the Latin word, or word-group, written immediately underneath it. These considerations have placed the Old High German Interlinearversionen squarely within the context of monastic education in the Carolingian period.
Still, the extant fragments of an Old High German Interlinearversion of the Psalter, the so-called Altlalemannischen Psalmenfragmente, have as yet received little attention and still await renewed analysis within the light of recent scholarship. Moreover, as the text of the Latin Psalter was of pivotal importance within early medieval monastic life, not only for its use in the daily liturgy, but also as a means for the teaching and learning of Latin, these Psalmenfragmente provide a unique window unto the teaching, learning and understanding of the Latin Psalter in the ninth century and the function(s) of the Old High German vernacular within this process. Furthermore, the Psalter is one of the few Latin texts which have received comparable Interlinearversionen in Old Saxon and Old English, and the Altalemannischen Psalmenfragmente thus uniquely enable us to compare and contrast similar text-ensembles in neighbouring languages.

Explaining Oengus in Style: The use of Irish and Latin in the commentary on the Félire Óengusso, a preliminary case study.
Nike Stam, Utrecht University

The textual commentary on the Félire Óengusso is very complex and is still surrounded by many questions. For one thing, this commentary, that appears in nine different manuscripts, has never  satisfactorily been edited or dated. However, one thing about the text is crystal clear: Latin and Irish were used side by side and it did not seem to bother anyone. Or did it?
This paper will present a case-study of the present state of affairs within the project ‘Medieval Irish Bilingualism’. Examples will be compared of code-switching in the commentary on the month of January in four manuscripts (Brussels 5100-4, Brussels 5057-5059, Leabhar Breac and UCD A7) to provide a chronological perspective. This is specifically aimed at examining whether, at any stage of the transmission of the text, the different scribes copied the switches differently. Furthermore, this paper will focus on some specific bilingual phenomena; which of these bilingual encounters are more likely to reflect a certain jargon and which might be considered as ‘proper’ code-switching as defined by modern contact linguistics?

Degrees of Irish-Latin bilingualism in the Leabhar Breac
Tom de Schepper, Utrecht University

The adaptation of modern bilingualism theory to historical sources often focuses on code-switching criteria. While this is a valid perspective, it is only one part of the socio-linguistic spectrum that the medieval manuscript encompasses. Within the study of language there are other matters that can be considered, such as interference of Irish on the Latin language and the policies of translation and paraphrasing. Looking at Latin citations and their Irish equivalents, there is more at work than mere transference; moreover, language divisions are blurred by the conscious use of both suspensions and abbreviations. It is also profitable to view diverse avenues in related disciplines such as codicology,  the production and use of manuscripts; the characteristics of or differences between various genres; and the linguistic state of affairs in sources and parallel texts. All of these angles can be explored in investigating the potential extent of bilingual ability in historical societies such as medieval Ireland.
The bilingual Leabhar Breac displays all the linguistic ranges within the bounds of the book. Changing between the two languages occurs from the level of the morpheme to the order of texts in a quire. In considering the composition of the collection it becomes clear that the scribes intended to make a conscious choice of language (which is partly disturbed by the present binding). By looking at the genre of the homily in the Leabhar Breac and its parallels and sources in other manuscripts, the linguistic ability and agenda behind the book are manifested. In particular, additions and glosses can informs us on differences between several textual traditions. It is also interesting to note that the boundaries between genres cannot always be drawn that easily. In addition to linguistic insights, the socio-historical significance of this bilingualism speaks volumes on the scientific setting of Ireland. This preliminary paper explores the many degrees of Irish-Latin bilingualism in the Leabhar Breac.

Code-switching in the Würzburg glosses?
Jacopo Bisagni, Classics, National University of Ireland Galway

The large majority of Medieval Irish glossing is characterised by the alternate use of Latin and Old Irish. In particular, the Würzburg glosses to the Epistles of Paul constitute one of the most interesting collections, from the point of view of the level of linguistic interaction between the two languages in question (in many cases, the linguistic switch takes place within a single clause). It is of course very tempting to describe this phenomenon as a written form of code-switching. This, however, raises a number of interesting questions: are we really dealing here with a behaviour essentially comparable to oral code-switching? If yes, are current explanatory models of code-switching suitable for studying this particular case? And if on the contrary this is not code-switching, could the fluctuation between Latin and Old Irish in written sources be due to other factors, such as borrowing or textual transmission? This paper will address these and other related questions, in the hope of providing new insights on the nature of Early Medieval Irish élite bilingualism.