Code-switching in English of the Middle Ages

Päivi Pahta, University of Tampere

1. Introduction

English is well-known for its extensive contact with other languages during its history,
and the impact of these various contacts on the lexicon is one of the traditional key areas
in research on the history of English. The bulk of research on contact-induced change in
earlier English has been carried out in the frame of lexical borrowing. With an increasing
interest in linguistic phenomena pertaining to multilingualism in present-day speech
communities and to the processes of language change, language historians too have
begun to rethink the notion of contact and its implications from new perspectives.
One of the contact phenomena thematised in recent research is code-switching,
the alternating use of units from two or more participating linguistic systems within a
communicative event. Since the early 1990s, a small number of historical linguists have
examined code-switching in materials where English is one of the participating
languages, most often co-occurring with Latin or French (for surveys of research, see
Schendl 2002, Nurmi and Pahta 2004, Schendl and Wright forthcoming). These studies
supply considerable evidence of the lexico-grammatical, sociolinguistic and pragmatic
characteristics of language mixing in a variety of genres from different historical periods,
ranging from Old to Late Modern English, with most attention given to the medieval era.
It has become evident that a great number of mixed-language texts survive especially
from the Middle Ages, showing different types of mixture. Some texts bear witness to
something of a fused lect, where the participating languages are integrated to an extent
where they form a single grammatical system instead of two separate ones (see Wright
1998, 2002). At the same time, texts traditionally perceived as monolingual English have
also been shown to contain a substantial amount of language mixing, including linguistic
units whose position on the scale of foreignness from code-switched other-language
segments to domesticated lexical borrowings is difficult to define very precisely (see
Pahta and Nurmi 2006). Thus, these studies present a new outlook on written data
surviving from earlier periods, providing opportunities to rethink the role of contact and
mechanisms of contact-induced change in the history of English.
This article focuses on code-switching in medieval England, characterised by
inherent multilingualism, with an emphasis on the later Middle Ages. After this brief
introduction, section 2 discusses code-switching as a linguistic phenomenon and an
object of research, addressing some peculiarities in studying it in a historical perspective.
Section 3 examines the textual evidence of multilingualism and code-switching from
medieval England. Structural and functional aspects of code-switching receive attention
in section 4, while section 5 briefly thematises code-switching as a source of new lexicon.
Section 6 contains some concluding remarks.

1
2. Analysing code-switching

The term code-switching is variously defined in literature (see Bhatia and Ritchie 2004 or
Bullock and Toribio 2009). For some researchers it is a special subtype of language
alternation. It can be used for denoting only those changes of language that occur at a
sentence boundary as opposed to code-mixing, denoting alternation within sentence
boundaries. Alternatively, it can refer exclusively to changes of language that have local
meaning in the context in which they occur. I use it as a general term in a broad sense, as
a synonym for language alternation or language mixing, to refer to any identifiable use of
more than one language in the course of a single communicative event (Heller 1988: 1).
The communicative event as a unit of analysis can be defined in different ways for
different research foci: it can consist of a single text, e.g. a scientific treatise or a letter,
but it can also comprise a collection of texts, e.g. a recipe book, a manuscript codex
containing several texts, or a sequence of letters between two correspondents. Within
such communicative events, code-switching can occur in different forms and meanings,
reflecting the variety of ways in which speakers and writers make use of the linguistic
resources at their disposal (see section 4).
Code-switching is inherently related to individual, social and/or societal
multilingualism. As a widely attested discourse practice in bilingual communities it is
sometimes identified exclusively with bilingual speakers (see Bullock and Toribio 2009:
1-2). However, speakers traditionally characterised as monolinguals can also draw on any
foreign languages known to them in their communication, making use of their
multilingual resources in ways that resemble bilingual speakers. Furthermore, code-
switching is often associated with speech, and many scholars stress its nature as an
essentially conversational activity (Li 2005: 276). But code-switching occurs in writing
as well, which has only recently caught the attention of researchers on present-day
languages, partly as a result of historical approaches, and partly due to increasing interest
in mixed-language online writing (see Sebba et al. forthcoming). In all these cases, code-
switching can be seen as rational and motivated linguistic behaviour, a specific type of
language variation, where the variant pool contains items from different languages, not
just different varieties of one language.
Code-switching has been analysed in various theoretical and methodological
frameworks. Within linguistic approaches, analyses focusing on the grammar of code-
switching have aimed at describing the structural mechanisms of switching, with a
specific goal of identifying language-specific or universal constraints governing
switching. At the same time, functionally-oriented research has mainly sought to explain
switching as a meaning-making practice or to describe the indexical values of specific
codes or changes from one code to another on the macro-level of social functions and/or
on the micro-levels of interpersonal and identity functions. These main orientations can
also be found in historical linguistic approaches to code-switching in medieval England.
Research on structural aspects of switching has focused on the grammatical nature of
switched segments and switch sites, while research in the functional framework has
examined e.g. the textual, discoursal and pragmatic functions of code-switching, its genre
distribution and its social stratification.

2
 A number of challenges characterise historical code-switching research. While
studies on code-switching in medieval English by definition analyse the phenomenon in
writing, research to date is largely based on theories formulated in research on present-
day speech. Applying this analytical frame is not unproblematic, and most recent research
in fact emphasises the need to theorise code-switching in writing in its own right in order
to account for features that are specific to written language (see Schendl and Wright
forthcoming, Sebba et al. forthcoming). For example, phonological and phonetic clues
that in the analysis of speech have sometimes been used as criteria for separating code-
switched items from domesticated borrowings are not available in the analysis of
historical data. On the other hand, writing can also make use of discourse devices that are
not available in oral communication, and thus not taken into account in theories based on
speech. These features include the use of ambiguous graphic symbols, attention catchers,
headings, subheadings or other graphic devices such as underlining or a change of script
to structure discourse or to carry meanings that in face-to-face interaction are expressed
e.g. by intonation, pauses, gestures (see Pahta 2004, Nurmi and Pahta 2004). The
different production constraints on communication imposed by the two media also need
to be taken into account. For example, while spoken language in face-to-face
communication is situation-dependent, immediate and spontaneous, writing allows more
scope for planning, editing and self-censure, and ignoring this crucial distinction in the
analysis can lead to misinterpretation of findings. In general, pragmatic or discoursal
analyses of local meanings of code-switching practices in their social and situational
embedding are challenging in medieval data, where often little is known about the
language-external circumstances surrounding the communicative event (see Kytö and
Pahta in this volume).
Finally, as Schendl (2000: 71) points out, the systems of the languages in use in
medieval England are not adequately described, which makes it difficult to define what
precisely is meant by “code” and thus, inherently, “switching between codes”. While
code-switching research is built on the basic premise of different languages as neatly
separated systems, in medieval language data this is not always possible. The limits
between the languages are not watertight, as some words and morphemes cannot be
labelled as belonging unequivocally to one language only (see Rothwell 2000, Wright
2000, Trotter 2003). In addition, there is great internal variation in all languages in use,
which further complicates the distinctions (see section 3).

3. Code-switching across types of writing

Some words about the linguistic situation of medieval England are in order as a
background for the occurrence of code-switching in extant writings. England was a
multilingual language community throughout the Middle Ages. In the early medieval
period, English in its different varieties of Germanic descent (Old English dialects) was
used by the side of indigenous Celtic languages and, later on, varieties of Old Norse
spoken by Viking invaders in large areas of the country. Latin was used especially in the
Church. In post-Conquest England, the focus in this article, the three principal languages
in use were English, French and Latin, which were present in a variety of forms. English
before the rise of the national standard was spoken and written in innumerable regional,

3
local and idiolectal varieties, Latin existed in classical and various medieval varieties, and
French in forms derived from the dialectally distinct varieties of Norman and Central
(Parisian) French. In addition, Celtic languages were spoken in some western and
northern areas, Hebrew among Jewish communities, and e.g. Low German among
immigrant communities. The three main languages had partly overlapping social
functions that changed over time. After the Conquest, Latin and French were the prestige
languages used e.g. in the Church, administration and education, including the new
universities, while English remained primarily a spoken language. Towards the later
Middle Ages, English emerged as a written language in a gradual process of
vernacularisation, eventually gaining a full register-scope (for discussions, see e.g. Pahta
and Taavitsainen 2004, Wright 2007). It is obvious that in this complex multilingual
situation there were individuals and social groups fluent in more than one language.
Undoubtedly, the majority of the population was monolingual English, but for the higher
social strata, monolingualism must have been rare. As Rothwell (1991: 179) puts it,
generations of post-Conquest educated Englishmen passed daily from English into
French and back again in the course of their work; this must have also happened between
English and Latin, and French and Latin.
Against this background of endemic multilingualism and pervasive language
contact in medieval England, it is not surprising that language mixing is widespread in
extant written sources from the period. Bi- or trilingual manuscript codices containing
English, Latin and French in various combinations survive in multitudes from the
fourteenth and fifteenth centuries, providing ample evidence of polyglot scribes and
readers. No inventory of medieval mixed-language materials is available, but the
common nature of the phenomenon becomes evident e.g. in the handlists of the Index of
Middle English Prose, which catalogue mixed-language texts in various library archives
under “macaronic” texts. The ratios of languages vary according to the date of
composition and according to genres and registers. For example, a study of 178 late-
fourteenth- and fifteenth-century scientific and medical manuscripts and booklets
suggests that almost half of the medical books written in England at this time contained
texts in more than one language (Voigts 1989).
In general, there is great variation in the frequency of code-switching across
domain-specific registers and genres of writing. A quantitative corpus-linguistic analysis
of code-switching based on the Helsinki Corpus of English Texts, a diachronic multi-
genre corpus of c 1.6 million words c. 730 to 1710, suggests that texts representing the
domains of science and religion contain more code-switching than other types of writing
(Pahta and Nurmi 2006). Studies examining single genres and registers also bear witness
to the frequency of code-switching in texts from these domains. Within science, medical
writings in particular contain frequent code-switching within individual texts (Voigts
1989, 1996, Hunt 2000, Pahta 2003, 2004). In the light of present knowledge, language
mixing is also particularly common in business accounts (Wright 1998, 2000, Rothwell
2000), in sermons (Wenzel 1994, Schendl 2000), other types of religious prose and verse
texts (Machan 1994, Pahta and Nurmi forthcoming), and some types of administrative
and legal texts (Davidson 2005). Code-switching is also found in drama texts (Diller
1997/8) and personal letters (Nurmi and Pahta 2004). However, within practically all
registers and genres there is great variation between individual texts in the frequency of
code-switching, its structural manifestations, and the functions in which it occurs (see

4
Pahta 2004, Pahta and Nurmi 2006). Analysis of the social patterns of code-switching in
data from the Corpus of Early English Correspondence shows that code-switching can be
found in the language use of letter-writers from various social backgrounds. Some
correlation patterns with different social positions emerge, e.g. writers of the educated
social strata like clergymen and lawyers frequently switch from English into Latin and
back in their correspondence, while merchants tend to make use of French. Variation
within social groups, on the other hand, confirms that the explanations for variation do
not lie exclusively with demographic variables (see Nurmi and Pahta 2004).

4. Structures and functions in code-switching

The ways in which the resources of the different languages are used in medieval mixed-
language texts vary across and within genres. In most cases alternation takes place
between two languages, but there are also texts where English, French and Latin all co-
occur. The proportions of the languages and the patterns of switching vary over time and
across different types of writing. In general, English and/or French can occur in Latin
texts, French texts can include Latin and/or English passages, and English texts can
incorporate Latin and/or French.
An extreme form of mixing two or all three languages is often found in medieval
administrative and business documents, including accounts and inventories. These
macaronic texts studied by Rothwell (2000) and Wright (e.g. 2000, 2002) show a
frequent juxtaposition of lexical items and inflections from different languages, combined
with an extensive use of ambiguous abbreviated spellings that may belong to any of the
participating linguistics system. According to research that has accumulated on the
documents, the extensive lexical and morphosyntactical mixing is characteristic of the
genre, and seems deliberate and functional to such an extent that they may be considered
representatives of a special, supra-regional, fused language of business administration
and accountancy. Similar texts containing extensive language mixing are also found
among medieval sermons (Wenzel 1994) and in macaronic verse (Schendl 2001).
In writings where the languages are more clearly separable, switches can be
analysed on levels of grammatical or textual structure. The two levels overlap, as text-
level switches are also realised in grammar. In their grammatical structure, switches can
be classified into different types according to the morpho-syntactic nature of the switched
constituents and switch-sites. The three main types usually identified are intersentential,
intrasentential and extrasentential switches. In medieval data, intersentential switches are
best defined as switches that take place between sentences or independent clauses. As
Schendl points out (2001: 307), the syntactic relation between coordinated independent
clauses is closer to the relation between two sentences than to the relation between a
dependent clause and its matrix clause. Furthermore, in the absence of consistent
syntactic punctuation in medieval writing, it is often impossible to distinguish two
sentences from a single sentence consisting of two coordinated clauses. Extrasentential
switches consist of freely distributable categories, such as interjections or discourse
markers, which can occur in several positions in an utterance and do not depend on the
structural rules of a language to the same extent as other switches. Intrasentential
switches contain all other switches within a sentence or an independent clause, including

5
switches occurring between a dependent clause and its matrix clause. With respect to
structural complexity, intrasentential switches are thus more complex than intersentential
or extrasentential switches in that they require the matching of the syntactic rules of two
languages. It has been suggested that intrasentential switches tend to occur at points in
discourse where juxtaposition of elements from the two languages does not violate a
syntactic rule of either language, i.e. at points around which the surface structures of the
two languages map onto each other, but counter-examples to this tendency, known as the
“equivalence constraint”, have been presented in various language pairs, e.g. in Middle
English–Latin switching by Schendl (2000: 75).
Research has shown that the syntactic switching patterns and switch sites in
medieval writing mainly correspond to those found in modern code-switching (see e.g.
Machan 1994, Schendl 2002). Both intersentential and intrasentential switches are
common. Most intrasentential switches can be defined in terms of single sentence
constituents, but instances of unlimited switching are also found. The most common
switched sentence constituent is the noun phrase. Furthermore, frequency hierarchies of
switched constituents and switching points established for Middle English by Schendl
(2000) show considerable similarity with results on Modern English. Differences
between the two datasets are mainly seen in the relative frequencies of switching
strategies and may partly result from pragmatic and stylistic factors in the data.
Sometimes switches are more appropriately described in terms of text structure
than syntactic structure. On the level of text structure, a change of code occurs between
different text components or levels. This is seen e.g. in cases where a switch occurs at the
boundary between the title or preface and the body text, or separates a direct quotation or
the author’s metatextual comment from the rest of the text (see Pahta 2003, 2004, Pahta
and Nurmi 2006).
The pragmatic and discourse functions of code-switching have been addressed in
several studies analysing different types of writing. These studies show that the functions
and meanings of switching can vary e.g. according to the genre, purpose or audience of
the texts. Some genre-specific conventions have emerged, such as the use of switched
dates, greetings and leavetakings in personal letters (Nurmi and Pahta 2004). In poetry,
switched segments can be used to re-iterate or to contribute to speaker characterisation
(Machan 1994, Schendl 2001), but the same function can also be found in medical and
scientific texts (Voigts 1996). Switched quotations are found in a wide range of medieval
genres, serving more specific discourse functions such as structuring or organising a text
(Wenzel 1994, Pahta 2004) or lending authority via intertextual references to
authoritative sources of knowledge (Wenzel 1994, Machan 1994, Pahta 2004). Analysis
of personal correspondence, on the other hand, indicates that code-switching can also
serve to index and negotiate the identities or interpersonal relationships of medieval
letter-writers. It can specify a particular addressee as the recipient of the message,
establish group-membership, intimacy or social distance, or signal power or prestige (see
Nurmi and Pahta 2004, Pahta and Nurmi forthcoming).

5. Code-switching as a source of new lexicon

6
The potential of code-switching in lexical change, or in any other area of language
change, is illustrated by Thomason, claiming that “any feature that can be code-switched
from language A into language B can turn into a permanent interference feature in
language B”; furthermore, “any feature that can appear in a single person’s speech at any
time […] can turn into a permanent change in the entire language” (2003: 694-695). As
Thomason’s discussion shows, code-switching has been widely recognised as an
important mechanism of linguistic change. However, its role in language change in older
stages of English has so far received little attention. Schendl (2010) notes that the process
and mechanisms by which code-switching contributed to change in English of the Middle
Ages have largely remained hypothetical. The reason for this is that research on contact-
induced change has been carried out in the frame of borrowing, conceptualized as a
process of adoption rather than a process of change resulting from variation-like code
alternation. However, the consequences of this contact-induced process, i.e. the massive
amount of loan words that entered the language during the Middle English period, are so
profound that they are addressed in all standard discussions of the history of English
lexicon (see e.g. Burnley 1992).
Furthermore, according to Thomason (2003: 696), code-switching is perhaps the
major route by which loan words enter a language. In her view, it is impossible in
principle and in practice to draw and absolute boundary between code-switching and
borrowing; instead, the two phenomena are linked by a continuum, with a fuzzy dividing
line. A code-switched word or morpheme turns with more and more frequent use into a
borrowing, until it is a regular part of the recipient language and e.g. learned as such by
new learners. This view presents the continuum from code-switching to borrowing as a
process that follows the same path as any other process of linguistic change, beginning
with a single use, and spreading with the increasing use of more and more speakers
adopting the innovation. Accepting this view of code-switching as a source of new
permanent lexicon in English of the Middle Ages opens up new perspectives for research
on contact-induced lexical change in a variationist frame, emphasising the mechanisms of
change, to complement earlier research on loan words.

6. Conclusion

This article has provided a brief state-of-the-art survey on code-switching in English of

the Middle Ages, a research field first thematised in English historical linguistics in the
1990s. I hope that this survey, discussing a range of approaches to and results gained
from studies in historical code-switching, has illustrated some of the fresh views that this
new dynamic field of historical linguistics can offer on the multilingual nature of the
history of English and on the role of multilingualism and language contact in language
variation and change.

References

Bhatia, Tej K. and William C. Ritchie (eds.) 2004. The Handbook of Bilingualism. New
York: Blackwell.

7
Bullock, Barbara E. and Almeida Jacqueline Toribio (eds.) 2009. The Cambridge
Handbook of Linguistic Code-switching. Cambridge: Cambridge University Press.
Burnley, David 1992. Lexis and semantics. In Norman Blake (ed.), The Cambridge
History of the English Language. Volume II: 1066-1476. Cambridge: Cambridge
University Press.
Davidson, Mary Catherine 2005. Discourse features of code-switching in legal reports in
late medieval England. In Janne Skaffari, Matti Peikola, Ruth Carroll, Risto
Hiltunen and Brita Wårvik (eds.) Opening Windows on Texts and Discourses of the
Past. Amsterdam/Philadelphia: John Benjamins, 343–351.
Diller, Hans-Jürgen 1997/8. Code-switching in medieval English drama. Comparative
Drama 31: 506–537
Heller, Monika (ed.) 1988 Codeswitching: Anthropological and Sociolinguistic
Perspectives. Berlin: Mouton de Gruyter.
Hunt, Tony 2000. Code-switching in medical texts. In Trotter (ed.), 131–147.
Index of Middle English Prose. 1984–. Gen. ed. A.S.G. Edwards. Handlists I–.
Cambridge: D.S. Brewer.
Li, Wei. 2005. ‘How can you tell?’ Towards a common sense explanation of
conversational code-switching. Journal of Pragmatics 37: 375–389
Machan, Tim William 1994. Language contact in Piers Plowman. Speculum 69: 359–
385.
Nurmi, Arja and Päivi Pahta 2004. Social stratification and patterns of code-switching in
early English letters. Multilingua 23: 417–456.
Pahta, Päivi 2003. On structures of code-switching in medical texts from medieval
England. Neuphilologische Mitteilungen 104: 197–210.
Pahta, Päivi 2004. Code-switching in medieval English medical writing. In Taavitsainen
and Pahta (eds.), 73–99.
Pahta, Päivi and Arja Nurmi 2006. Code-switching in the Helsinki Corpus: A thousand
years of multilingual practices. In Nikolaus Ritt, Herbert Schendl, Christiane
Dalton-Puffer and Dieter Kastovsky (eds.), Medieval English and its Heritage:
Structure, Meaning and Mechanisms of Change. Frankfurt am Main: Peter Lang,
203–220.
Pahta, Päivi and Arja Nurmi forthcoming. Multilingual discourse in religious writings
from medieval and early modern England: A corpus approach to historical research
on code-switching. In Schendl and Wright (eds.).
Pahta, Päivi and Irma Taavitsainen 2004. Vernacularisation of medical and scientific
writing in its socio-historical context. In Irma Taavitsainen and Päivi Pahta (eds.)
Medical and Scientific Writing in Late Medieval English. Cambridge: Cambridge
University Press, 1–18.
Rothwell, William 1991. The missing link in English etymology: Anglo French. Medium
Ævum 60: 173-196.
Rothwell, William 2000. Aspects of lexical and morphosyntactical mixing in the
languages of medieval England. In Trotter (ed.), 213–232.
Schendl, Herbert 2000. Syntactic constraints on code-switching in medieval texts. In:
Irma Taavitsainen, Terttu Nevalainen, Paivi Pahta and Matti Rissanen (eds.)
Placing Middle English in Context. Berlin/New York: Mouton de Gruyter, 67–86.

8
Schendl, Herbert 2001. Code-switching in medieval English poetry. In Dieter Kastovsky
and Arthur Mettinger (eds.), Language Contact in the History of English. Frankfurt
am Main: Peter Lang, 305–335.
Schendl, Herbert 2002. Mixed language texts as data and evidence in English historical
linguistics. In Donka Minkova and Robert P. Stockwell (eds.), Studies in the
History of the English Language: A Millennial Perspective. Berlin/New York:
Mouton de Gruyter, 51–78.
Schendl, Herbert and Laura Wright (eds.) forthcoming. Code-switching in the History of
Earlier English.
Sebba, Mark, Shahrzad Mahootian and Carla Jonsson (eds.) forthcoming. Language
Mixing and Code-switching in Writing: Approaches to Mixed-language Written
Discourse.
Thomason, Sarah Grey 2003. Contact as a source of language change. In Brian D. Joseph
and Richard D. Janda (eds.) The Handbook of Historical Linguistics. Oxford:
Blackwell, 687–712.
Trotter, Derek A. (ed.) 2000. Multilingualism in Later Medieval Britain. Woodbridge: D.
S. Brewer.
Trotter, Derek A. 2003. Oceano vox: you never know where a ship comes from: On
multilingualism and language-mixing in medieval Britain. In Kurt Braunmüller and
Gisella Ferraresi (eds.) Aspects of Multilingualism in European Language History.
Amsterdam/Philadelphia: John Benjamins, 15–33.
Voigts, Linda Erhsam 1989. Scientific and medical books. In Jeremy Griffiths and Derek
Pearsall (eds.) Book Production and Publishing in Britain, 1375-1475. Cambridge:
Cambridge University Press, 345–402
Voigts, Linda Erhsam 1996. What’s the word? Bilingualism in late-medieval England.
Speculum 71: 813–826.
Wenzel, Siegfrid 1994. Macaronic Sermons: Bilingualism and Preaching in Late-
medieval England. Ann Arbor: The University of Michigan Press.
Wright, Laura 1998. Mixed-language business writing: Five hundred years of
codeswitching In Ernst Håkon Jahr (ed.) Language Change: Advances in Historical
Sociolinguistics. Berlin/New York: Mouton de Gruyter, 99–118.
Wright, Laura 2000. Bills, accounts, inventories: everyday trilingual activities in the
business world of later medieval England. In Trotter (ed.), 149–156.
Wright, Laura 2002. Code-intermediate phenomena in medieval mixed-language business
texts. Language Sciences 24: 471–489.
Wright, Laura 2007. The languages of medieval Britain. In Peter Brown (ed.) A
Companion to Medieval English Literature and Culture c.1350–c.1500. Oxford:
Blackwell, 143–158.

9
10