Merchants of Innovation: The Languages of Traders 9781501503542, 9781501511608

Citation preview

Esther-Miriam Wagner, Bettina Beinhoff, Ben Outhwaite (Eds.) Merchants of Innovation

Studies in Language Change

Edited by Cynthia Allen Harold Koch Malcolm Ross

Volume 15

Merchants of Innovation The Languages of Traders Edited by Esther-Miriam Wagner Bettina Beinhoff Ben Outhwaite

ISBN 978-1-5015-1160-8 e-ISBN (PDF) 978-1-5015-0354-2 e-ISBN (EPUB) 978-1-5015-0341-2 ISSN 2163-0992 Library of Congress Cataloging-in-Publication Data A CIP catalog record for this book has been applied for at the Library of Congress. Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http://dnb.dnb.de. © 2017 Walter de Gruyter Inc., Boston/Berlin Cover image: iStockphoto/thinkstock Typesetting: Konvertus BV, Haarlem Printing and binding: CPI books GmbH, Leck ♾ Printed on acid-free paper Printed in Germany www.degruyter.com

Acknowledgements We should like to extend a big thank you to all the delegates and participants of the ‘Merchants of Innovation’ conference, which took place on 7th–8th April 2014 in St John’s College Cambridge, for making it such an enjoyable occasion and for contributing to this intellectual endeavour. We owe a great debt to the Master and Fellows of St John’s College, Cambridge who graciously hosted our conference, and who contributed very generously to the costs of the workshop. Our heartfelt thanks go to Professor Peter Matthews in particular, who was instrumental in gaining financial assistance from the College. We are immensely grateful for the financial and administrative support from the Woolf Institute, Cambridge, and to Juni Hoppe and Magdalen Connolly for assisting us with the organisation of the workshop. Thanks also go to the Genizah Research Unit at Cambridge University Library for their financial contribution. Roger Wright, William Standing, Laura Wright, Geoffrey Khan, Terttu Nevalainen, Alexander Bergs and Jessica Tearney Pearce have to be praised for providing much needed support during the peer reviewing and editing process. An anonymous reviewer read the script in great detail and gave us very thorough feedback which helped greatly to improve the volume. Mouton de Gruyter have been fantastic publishers to work with, and our thanks go to Lara Wysong and the editors of this series who aided us throughout the emergence of the volume, and to our ever reliable production editor Wolfgang Konwitschny.

DOI 10.1515/9781501503542-202

Contents Acknowledgements 

 v

I Introduction Esther-Miriam Wagner and Bettina Beinhoff 1 Merchants of innovation: the languages of traders 

 3

II Literacy of traders and their agency as linguistic trendsetters Merja Stenroos 2 Like the coins when currencies are combined: contextualizing the written language of fifteenth-century English merchants   19 José Miguel Alcolado Carnicero 3 Bridges of innovation and change: the English language around the networks of the Mercery of London   40 Esther-Miriam Wagner 4 The socio-linguistics of Judaeo-Arabic mercantile writing 

 68

III Code-switching, loanwords and multilingualism Ivar Berg 5 Business writing in early sixteenth-century Norway 

 89

Laura Wright 6   Kiss Me Quick: on the naming of commodities in Britain, 1650 to the First World War   108 Samuli Kaislaniemi 7 The early English East India Company as a community of practice: evidence of multilingualism   132

viii 

 Contents

Agnete Nesse 8 Language choice in forming an identity: linguistic innovations by German traders in Bergen   158 Henrike Kühnert 9 From the synagogue to the market square: cardinal numbers in Older Yiddish   179

IV Mercantile linguistic communities Megan Tiddeman 10 Early Anglo-Italian contact: new loanword evidence from two mercantile sources, 1440–1451 

 217

Josh Brown 11 Multilingual merchants: the trade network of the 14th century Tuscan merchant Francesco di Marco Datini   235 Rachel Selbach 12 On a famous lacuna: Lingua Franca the Mediterranean trade pidgin?   252 Index 

 272

I Introduction

Esther-Miriam Wagner and Bettina Beinhoff

1 Merchants of Innovation: the languages of traders 1 Merchants and their spoken and written languages Traders around the world use spoken argots. Some of these started off as pidgins and developed into creole languages, others are mercantile registers which emerged within particular languages, often mainly lexical in nature but also displaying phonological and morphological variations. Aside from the practical needs to communicate across language boundaries, to hide commercial secrets from outsiders or to identify oneself as a member of a specific group or guild, the use of particular sociolects re-inforces ties between speakers and creates a sense of belonging, shared values and community, as the works of pioneer sociolinguists such as William Labov (1972) and James and Lesley Milroy (1985, 1987) have proposed. Famous spoken mercantile languages include Malay (Verstegh 2008: 175), the horse and cattle traders sociolect – related to other Yiddish-influenced argots such as Rotwelsh, Yenish and Masematte – that emerged in Germany (see Guggenheim-Grünberg 1954), or market argots used among Yemeni businessmen (Walter 2003) and Cairene goldsmiths (Khan 1995–1997). Lore describes the almost mythical historical Lingua Franca spoken by Mediterranean traders. In other parts of the world, we find mercantile bridge languages such as Swaheli (just one of dozens of lingua francas in Africa, see Heine 1970), or Basque-American Indian as used in North America in the sixteenth and seventeenth centuries (see Bakker 1989). Some trade argots, such as Malay and Swaheli, later developed into supra-regional means of communication. How homogeneous and grammaticalised many of these mercantile languages really were and whether we can talk about actual sociolects as opposed to “trading gibberish” (see Heine 1970: 19) is difficult to judge, as has been demonstrated by Selbach’s critical deconstruction of the Mediterreanean Lingua Franca in this volume. Yet conceptual expectations Note: We would like to thank the anonymous reviewer for his/her conscientious reading of the script and his/her helpful suggestions and recommendations. Esther-Miriam Wagner, Woolf Institute and University of Cambridge Bettina Beinhoff, Anglia Ruskin University DOI 10.1515/9781501503542-001

4 

 Esther-Miriam Wagner and Bettina Beinhoff

regarding the distinct peculiarities of traders’ languages have even made it into popular culture, as can be seen on the Star Wars wiki where it is stated that “trader’s argot is full of slang terms and grammatical errors”, to the degree that “not even a protocol droid could comprehend this intregration of languages used by the traders”.1 The Star Wars wiki entry shows that the popular idea of trade languages is related to issues of literacy, education and pragmatics: merchants are not the most well educated, but for pragmatic reasons need to learn a number of languages, which they only master to a certain, sometimes superficial degree, leading to heavily coded and perhaps intentionally secretive sociolects. For historical sociolinguists who deal mainly with written sources the question arises how such distinctive characteristics of trade languages manifest themselves in the writing of traders. An important part certainly is of lexical nature, but the studies presented in this volume suggest that the linguistic peculiarities of mercantile letters extend also beyond the lexical arena. For example, there appears to be a clear cross-linguistic tendency that merchant writings show a greater degree of language mixing and code-switching, along with, often, many more dialectal forms than are usually found in other text types. As such, the languages of commerce are a fruitful area of linguistic research, not just in the details and typologies of trade argots themselves, but also in what traders’ writings can reveal about the underlying dialects. Locker (1987: 40), who from a non-linguistic point of view, laments the use of jargon in business correspondence, admits that “many people think that business letters are supposed to use a special phraseology of mode of expression to be professional”. This jargon covers, however, only a small part of the linguistic inventory specifically used for business. Mercantile correspondence and commercial texts often display significant differences from the language used in other documentary writing, such as official, legal or private correspondence. Stenroos argues that the language of business and administrative texts developed quite differently from that of literary texts (see e. g. Benskin 1992: 75; Rissanen 2000: passim; Stenroos 2004: 275–81) as a result of the difference in text production: copying versus composition. Whereas copying of literary texts results in a perpetuation of existing linguistic traits, composition of documents, such as business letters, can lead to a more dynamic and innovative linguistic environment. In the world of the Cairo Genizah, a depository of manuscripts found in a synagogue in old Cairo, for example, this difference between literary and document production manifests itself in distinct classes of professional writers. Scribes who compose

1 http://starwars.wikia.com/wiki/Trader’s_argot, accessed on 19 November 2015.



Merchants of Innovation: the languages of traders 

 5

documents are called sofer or kātib, depending on the context of their writing.2 Copyists of books, however, are called nāsiḵ. There seems to be little overlap between the groups.3 Evidence also suggests that traders play an important role in promoting and spreading language standards. Their literacy is pragmatic, and their acts of writing fundamentally different from those of scholarly writers or scribes. As a consequence of their profession and their social and geographical mobility, they are more exposed to language contact than other groups of speakers and writers, and often form loosely-knit networks, which promote language-levelling and “writing against the norms” (Carnicero). They are agents in terms of innovation, i. e. the introduction of new variants into a language community, and in terms of supralocalisation, i. e., the increase in the relative frequency of forms with a wide geographical currency at the expense of local or regional form (Stenroos). On the other hand, in particularly close-knit networks we can observe complexification processes in spoken and written registers, leading to specific mercantile language forms (Wagner). This volume thus aims at placing the language varieties used by traders within a wider sociolinguistic context and examines their relationship to other registers of their respective languages. Questions posited include: What differences can be observed between mercantile registers and those of court or legal scribes? Do the traders’ texts show the early emergence of features that take longer to permeate into the “higher” varieties of the same language? Do they anticipate language change in the standard register or influence it by setting linguistic precedents? Closely connected to these questions, and rooted in commercial correspondence, the book will also examine bilingualism and multilingualism, for example Low German, Norwegian and Danish during the Hanse era (Berg and Nesse), Tuscan, Latin and Catalan in 14th-century Italy (Brown), or merchants who speak Yiddish and Arabic, but write Hebrew (Kühnert and Wagner). Similarly, semibilingualism is discussed, where writers have a range of languages to choose from, all understood by both writer and addressee. Reasons for code-switching in commercial correspondence and for the choice of particular languages over others will be explored. Finally, this volume uses evidence from a variety of mercantile networks to address questions about the writers themselves and their 2 See Goitein (1971: 183–185): sofer is the term used for the scribes writing in Hebrew, whereas kātib is used for the government clerks who write in Arabic. 3 While we have, for example, many hundreds of extant legal documents and letters written in one of the most familiar scribal hands of the Genizah, Ḥalfon ben Manasse, there appear to be hardly any literary sources in his handwriting, except for a quasi-magical text and an Arab commonplace book by a tenth-century Muslim author.

6 

 Esther-Miriam Wagner and Bettina Beinhoff

social environment. What differences can be observed in the language used within tightly-bonded mercantile groups and those in a wider business network?

2 Literacy of traders and their agency as l­ inguistic trendsetters The educational background of merchants and their level of literacy have been at the centre of much scholarly discussion. In his much cited study of literacies, Parkes (1973: 555) has described the linguistic behaviour of traders as “pragmatic literacy”, as opposed to literacy of a more scholarly or “cultivated” nature. Much of his assessment remains valid in general for mercantile language, but some additions should perhaps be made, as can be seen for example in the contributions by Stenroos and Wagner. Stenroos, on the one hand, argues that the concept of “pragmatic literacy”, with its assumptions of basic, and mainly vernacular, literacy skills does not necessarily give a fair idea of the writing competence of late medieval English merchants. She points out that a multilingual habit, as well as the perpetuation of local and regional forms, was brought from school education into business. The primary education of English traders was therefore of great influence to their business writing. According to Stenroos, English mercantile writing reflects a highly complex, bilingual literacy, combining the use of Latin as a first written language for work purposes with a highly innovative and regionally coloured use of English. Wagner, on the other hand, provides evidence supporting Parker’s assessment by showing that there seemed to be a consensus, at least in early medieval Middle Eastern Jewish mercantile networks, of what constituted an appropriate linguistic register of business letters. She demonstrates that when composing mercantile letters, educated writers adapted their writing style according to the style generally perceived as proper for business conduct. This adopted mercantile style indeed followed conventions of lower, less literary registers, i. e. what we could call a “pragmatic literacy”. When composing business letters, traders with higher education would suppress stylistic forms reminiscent of literary writing as well as those arising from a thorough religious training, such as Hebrew loanwords and phrases, which they would use in other types of correspondence. The linguistic register used for business correspondence would thus be inclusive for all traders, notwithstanding their varying educational backgrounds. An addition to Parkes’s assessment could therefore be that “pragmatic literacy” may be a choice of individual writers, rather than reflect a universal educational background and level of literacy of the writers of a particular professional class.



Merchants of Innovation: the languages of traders 

 7

This variation in literacy, linguistic writing ability and expression of literacy of individual writers within the medieval Middle East can perhaps help to explain the differing assessment of various scholars about the literacy of London merchants. While Thrupp (1948: 161) argues that most London merchants had some training in Latin, and Hanawalt (1995: 82) refers to the literacy requirements of fifteenth-century London guilds as “severe”, Graff (1987: 98) believes that hardly any guild members could read Latin and only few could read or write English. Various contributions in this volume show that the language of traders is closer to the spoken language than other contemporary correspondence: Berg demonstrates how Scandinavian texts of an economic nature use the vernacular; Wagner argues that colloquial forms appear to occur earlier and more often in mercantile Judaeo-Arabic writing than in other text types.4 In the Jewish Middle Eastern business correspondence investigated, a common source for the introduction of vernacular phonology, morphology and syntax and for the setting of linguistic precedents are, for example, colloquial sayings and phrases. These vernacular expressions are often employed in connection with emotional statements that create bonds between the merchants, or express dissatisfaction and anger. Even though the level of merchants’ literacy may not have placed them in the literary avantgarde in the sense of a linguistic elite, no matter whether this was due to educational backgrounds or self-imposed leveling efforts, they nevertheless played an important part as agents of language change in the linguistic history of various languages. Fennell (2001: 97), for example, suggested an active role for merchants in the development of written English, as “mobility was the key to the standardization of English, and (...) merchants, tradesmen and members of city government had a central role to play”. In the case studies of our book, this “standardisation” in connection with mercantile activity has to be understood as levelling, i. e., the elimination of widespread dialectal or sociolectal forms in favour of widespread forms, rather than the imposition of uniformity. Wagner’s analysis of Fatimid Jewish traders, for example, shows the tendency to dismiss markedly Jewish linguistic forms in favour of language commonly understood by all traders of all religious creeds. Stenroos states that merchants can be assumed to be particularly mobile, and she also describes them to have large “weak-tie” networks in the form of trade contacts; they might thus consequently be expected to form a relatively innovative group of speakers. Language innovation as a main feature of mercantile correspondence is reflected in various chapters of this volume: merchants’ writing shows language mixing, code-switching, and a higher incidence of dialectal forms than other text types. 4 For more examples, see Wagner (2013).

8 

 Esther-Miriam Wagner and Bettina Beinhoff

These assessments are supported by Carnicero’s analyses of the social networks of businessmen. He argues that the quality of social networks played a role in the process of adoption of English as the main language in the different records of the Mercers’ Company of London. Whereas groups with loose-knit social networks seemed to foster the spread and adoption of the English vernacular in late medieval London, close-knit networks tended to maintain the use of traditional languages of record such as Latin and French. Hence, professional groups with the greatest geographical mobility and thus the widest social networks are those that most easily adopt innovative linguistic forms, and, due to their reach, also are responsible for the spread of these forms. A part of Wagner’s contribution is dedicated to mercantile handwriting style. She shows that Judaeo-Arabic letters of the 11th century display a distinctly more cursive hand than comparable legal documents. Similar phenomena have already been described, for example, for Italian (Ceccherini 2009) and English  business documents. In England, secretary script, or rather a blend of Secretary and Anglicana, was the primary script used by English merchants, and it carried a sense of being the appropriate hand for business and general correspondence, see Iopollo (2010: 177). The difference in handwriting is due to the distinct writing circumstances encountered by traders as compared to other writers. Traders had to react quickly to incoming mail. They had to write large quantities of letters, often in multiple copies, in order to ensure safe arrival of their correspondence. Their writing style thus had to be suitable for swift writing, and necessarily needed to be written with greater cursivity.5 Particular handwriting therefore can be assumed to have developed within most networks of business writers. The contributors to the section on Literacy of traders and their agency as linguistic trendsetters in this volume thus provide case studies of this role of the merchants as a social group, by examining their agency in the creation of the dialectally mixed character of what became Standard English (Stenroos and Carnicero), and in the introduction and propagation of progressive linguistic forms into written Judaeo-Arabic (Wagner). Within Elspass’s (2005 and 2012) concept of language history “from below”, texts associated with business writing should therefore occupy a prominent space. Traders in many languages appear to set linguistic precedent by being the first innovators to cross particular linguistic thresholds and introducing colloquial forms. This innovative linguistic behavior may be caused by a combination of factors. First, the typical schooling associated 5 In the case of Middle Eastern Jews writing in Hebrew script, another factor could have been the familiarity of most traders with the much more cursive Arabic script, as Wagner (2017) has described.



Merchants of Innovation: the languages of traders 

 9

with the profession makes them much more likely to use non-standard linguistic forms that are close to the spoken language. In addition to this “pragmatic literacy”, which facilitated the introduction of less literary forms, the social relationship between addressees and senders of mercantile letters may have played an important role, at least in certain areas of Europe and the Middle East. Because of legal and political circumstances, medieval and early modern traders often needed to rely on cordial relations with their business associates. In their letters, the informal spoken and thus more cordial language enabled them to convey a sense of intimacy between them and their trading partners, important in a business world whose economy was largely built, to a much greater degree than the modern counterpart, on trust, and familiar and friendly relations. Thus, secondly, writing in the vernacular with emotionally-charged colloquial expression was advantageous for conducting business affairs. And thirdly, mercantile letters are less subject to linguistic constraints and prescriptive norms than other text genres because of the special writing circumstances (speed of writing, sheer number of produced texts) encountered in the business context.

3 Mercantile multilingualism and code-switching Due to their geographical mobility and their contact with other merchants conversant in many foreign languages, traders were exposed to linguistic contact to a greater degree than other professional groups, and were also often expected to be multilingual. The case studies in section 3 attest to the multilingualism inherent in the mercantile arena, and complement previous scholarship on the subject, such as Wright’s (1994 and 1998) and Trotter’s (2011) studies on multilinguistic business writing.6 Berg cites a famous passage from the thirteenth-century Konungs skuggsjá [The King’s Mirror], which states that a merchant should “learn all languages, and first of all Latin and French, because those tongues are most widely known”. The consequence of this multilingualism was often the introduction of new loanwords, borrowings and code-switching. A clear distinction is usually made between code-switching and borrowing, code-switching being “the alternative use by bilinguals of two or more languages in the same conversation” (Milroy & Muysken 1995: 7), whereas borrowing is generally understood to “refer only to the incorporation of foreign elements into the speakers’ native language” (Thomason 6 The “translingual” concept developed by Hsy (2013) is probably better suited for a literary approach to mercantile linguistic activity.

10 

 Esther-Miriam Wagner and Bettina Beinhoff

& Kaufman 1988: 21) and is thought to appear on the lexical level before it moves into other linguistic structures. Borrowed items tend to be absorbed into the native language of the speakers and take on specific characteristics of this language (e. g. morphological or phonological) and are available across large parts of the community (Poplack 2001). It is not necessary for a speaker to know the source language of a borrowed item, whereas this is a fundamental prerequisite for code-switching. Thus, multilingual merchants were well-placed to initiate borrowing based on their code-switching. Wright demonstrates how the global union of traders in the British Empire ­incessantly introduced new terminology that infiltrated everyone’s usage of everyday English. She shows, for example, how –pore and seer– signalled ‘brand-new, latest cloth imports’ and –oon the ‘latest type of firework’, and how the appeal of these morphemes was not merely phonological or grammatical but social, too. Kühnert explains how Yiddish under the influence of mercantile language developed an extremely complex system of cardinal numbers, with two, three or even four parallel lexeme variants for each numeral, borrowed from Hebrew and German. The practical nature of face-to-face linguistic contact between merchants is addressed by Berg and Nesse in the context of Scandinavian merchants. They describe how Scandinavians and (Low) Germans communicated through a form of receptive bilingualism, similar to the receptive multilingualism described by Braunmüller (2007, 25–29): each interlocutor spoke his own language, probably with some accommodation, and the other understood it – much the same way as Scandinavians communicate today, for example with English speakers. Nesse also explores the sociolinguistic reasons behind the emergence of this receptive, rather than active, Norwegian–German bilingualism during the Hanse era. The Hanseatic merchants were bachelors, meaning that no Norwegian–German families were formed that could establish a mixed identity. This predominantly masculine and exploratory form of mercantile conduct, either with young bachelors or families left behind in the country of origin, is perhaps another topic that should be explored in the context of business language. Nesse gives a convincing account of the pragmatic reasons why mercantile communities may shift towards particular languages, suggesting that following the breakdown of the Hanseatic League in the late 18th–19th centuries the change from using only German to writing Danish and speaking Norwegian happened mainly because the leading merchants “wanted to demonstrate a distance from the Hanseatic League in order to make sure they were seen and treated as equal with the Norwegian inhabitants in their potential for gaining power in the city”. Kaislaniemi explains that in the multilingual environments where trade was conducted, “profuse use of multilingual mercantile terminology, jargon, and code-switching was the norm”. In his article on the extensive correspondence



Merchants of Innovation: the languages of traders 

 11

of the East India Company, he suggests that their linguistic practices developed aboard Company ships and in factory outposts on the one hand, and in Company correspondence on the other. Kaislaniemi (2009) has already demonstrated the rapidity in the spread of codeswitches and borrowings in mercantile writing within written East India Company correspondence. In the present volume, he describes the emergence of particular stylistic rules within this “linguistic community of practice”. One of the results of his work is showing a change of pattern in the usage of code-switching that suggests that writers used less multilingual jargon when writing back to England. This demonstrates awareness of which register was considered appropriate in smaller – more specific – or greater – more general – communities of merchants, and how networks spanning large geographical areas may exhibit linguistic levelling as far as code-switching and multilingualism are concerned to a much greater degree than smaller networks. This might help to explain why the early medieval Judaeo-Arabic writing merchants in their pan-Mediterranean and multi-religious networks shied away from using Hebrew words, whereas the more regionally active Jewish Early Modern traders composing letters in Yiddish and Late Judaeo-Arabic deliberately used them, as described by Kühnert and Wagner.

4 Mercantile linguistic communities The Mercantile linguistic communities section of this volume is dedicated to particular mercantile communities, whose writing and speaking habits demonstrate the special linguistic circumstances of traders. The contributions show how foreign merchants exerted linguistic influence on their newly adopted environment, how the international nature of trade aided language contact and multilingualism, and where the language contact most likely occurred and manifested itself. The linguistic legacy of the Italian merchant presence in late medieval England is discussed in Tiddeman. She suggests that the strength of Italian banking and trading institutions in late medieval England and the extent to which Italians permeated English society as a whole led the Italian language to have an important effect on the trilingual bureaucracy of English trade. Her contribution examines the origin of previously known direct Italian loanwords, which are almost exclusively connected to trade, and suggests more than two dozen hitherto unrecognised potential loanwords from Italian, which all entered through Anglo-Italian mercantile exchange. Most crucially, she highlights the importance of commercial contact as opposed to that of elite culture in early Anglo-Italian lexical contact. Tiddeman also enlarges on the social circumstances

12 

 Esther-Miriam Wagner and Bettina Beinhoff

which may have contributed to lexical exchange: all alien merchants had to live with an English host (which raises further questions about the language of communication) and send bi-annual reports, detailing all exports and imports, to the Exchequer. In short, Italian loanwords which entered Middle English did so overwhelmingly in the context of trade and mercantile relations. Brown discusses the linguistic skills of traders within the framework of the Datini business network, which reveals illuminating insights into medieval expectations and realities about mercantile linguistic skills. 14th-century sources state that a merchant abroad had to be able to speak one or more foreign languages. This attitude to multilingualism is reflected in the database of the Datini Archive, showing that hundreds of documents are written in six different languages. The Datini merchants thus had to be versed in various languages to conduct their trade. However, it was only the letters between external merchants and Datini employees that were composed in a variety of languages. Internal correspondence with the head of the company, Francesco di Marco Datini, was always written in Tuscan, a unifying company language, which was perhaps part of the regional identity of the company. The contact between foreign languages and Tuscan thus occurred between the external merchants and the Datini employees, only. Internal communication between the different employees of the Datini networks was mainly composed in Tuscan and Catalan. The choice of the two languages displays confidence in using both vernaculars for written correspondence in a formal, business context, even with native users. The decision to adopt Catalan in correspondence towards merchants outside the Datini system may have indicated the traders’ willingness to accommodate their trading partners’ native tongue, and to create a common, intimate language with those traders who were unfamiliar with the regional Italian dialect of the company. Interesting in this connection is the preference of northern Italian merchants to write in Latin rather than vernacular; using competing Italian vernaculars may have touched upon competing regional identities, which could have been potentially damaging for business. By resorting to Latin, any invocations of inner-Italian divisions were avoided. Latin, which would have also brought connotations of formality was thus never used for internal mail but only for communication with external merchants of great reputation and prestige. In the final paper of the volume, Selbach carefully deconstructs one of the most famous presumed medieval and Early Modern spoken traders’ pidgins, the Lingua Franca. She finds no primary evidence that would support the traditional definition of Lingua Franca as a trade language, and expresses doubt that it really could have been an important trade pidgin in the Mediterranean over several centuries when no such evidence is available. Selbach also raises the important question whether traders really needed one uniform pidgin to communicate. The



Merchants of Innovation: the languages of traders 

 13

contributions of this volume show that this is usually not the case at all. Rather, the basis for mercantile communication was often a complex negotiation between a variety of different languages.

5 Output Kaislaniemi’s assessment of the East India Company as a “linguistic community of practice” should probably be assumed for all the merchant communities described in this volume. As a consequence of their profession and their social and geographical mobility, traders are more exposed to language contact than other groups of speakers and writers, and often form loosely-knit networks, which promote language-levelling. Because trade is ordinarily associated with urban culture, mercantile language forms would also have been prestigious and associated with progress; as Labov (1972: 299) has demonstrated, rural dialects are constantly being replaced with urban class dialects in mobile societies. Jucker & Kopaczyk (2013: 2) have argued that communities of practice are the loci of linguistic change. In comparison to other communities of writers, within the merchant communities this change was driven by more progressive linguistic policies, caused by multilinguistic environments, large networks leading to language levelling, and the particular, vernacular-based mercantile literacy. That merchants are trendsetters in terms of linguistic innovation and supralocalisation has been shown in various studies in this volume: traders’ language is closer to the spoken language than contemporary other correspondence, and colloquial forms seem to appear earlier and more often in mercantile writing than in other text types. Mercantile texts thus display significant linguistic differences from other documentary writing and from literary texts, suggesting a clear cross-linguistic tendency to display a greater degree of language mixing, code-switching, and a higher incidence of dialectal forms than other text types. Traders set these linguistic precedents because, for reasons such as their typical “pragmatic literacy” and their writing circumstances, they are often the first group to write in the vernacular, and because they are less subject to linguistic constraints and prescriptive norms than other text genres. This volume recommends, however, the extension of the term “pragmatic literacy”, as it has been previously associated with an assumption of basic, and mainly vernacular, literacy. Mercantile writing can reflect a highly complex, multilingual literacy, combining the use of various languages. The suppression of more literary and formal language forms in business texts may be a stylistic choice of individual writers, rather than reflective of the level of literacy of the writers of a particular professional class.

14 

 Esther-Miriam Wagner and Bettina Beinhoff

The quality of social networks plays an important role too: groups with loose-knitted social networks seem to foster the spread and adoption of vernacular and levelled forms, while close-knitted networks tend to maintain the use of traditional languages. Professional groups with the greatest geographical mobility and thus the widest social networks are those that most easily adopt innovative linguistic forms, and due to their reach, also are responsible for the spread of these forms. At the same time, the pattern of code-switching appears to vary in smaller and greater networks, with code-switching more frequent in the former and less frequent in the latter. Although small groups of traders around the world use regionally spoken argots, the existence of supra-regional trade languages, such as the Lingua Franca, is doubtful. Rather, the studies in this book show that transnational trade negotiated linguistic differences by relying on various strategies. We find for example receptive bilingualism, where interlocutors each spoke or wrote their own language, with some accommodation. Shifts towards particular languages often have sociolinguistic significance: to express belonging to a particular political or religious community. E. g., traders in Bergen changed from German to Danish and Norwegian after the breakdown of the Hanseatic system to distance themselves from the Hanse and show that they were a part of the newly independent Norwegian state. Italian traders accommodated their Catalan counterparts linguistically in order to improve business relations. Jewish traders preferred Arabic over commonly used Hebrew epistolary expressions to establish themselves as part of a wider Muslim-Christian-Jewish business community around the Mediterranean. To sum up, traders play an important role in promoting and spreading innovative language forms, as well as cursivity of handwriting, due to their pragmatic literacy and writing circumstances. Further research should provide answers to queries that could not be addressed in the framework of this volume. Concrete questions posited in future investigations should include: Through which channels do the innovative mercantile forms enter other registers? Who are the protagonists within the traders’ caste which set the linguistic standards? Are the family circumstances of traders responsible for particular linguistic developments? What sets traders’ letters apart from private correspondence and other “low” registers?

References Bakker, Peter. 1989. The language of the coast tribes is half Basque. A Basque-American Indian pidgin in use between Europeans and Native Americans in North America, ca. 1540–1640. Anthropological Linguistics 31. 117–147.



Merchants of Innovation: the languages of traders 

 15

Benskin, Michael. 1992. Some new perspectives on the origins of standard written English. In J.A. van Leuvensteijn & J.B. Berns (eds.), Dialect and standard language in the English, Dutch, German and Norwegian language areas, 71–105. Amsterdam: Royal Netherlands Academy of Arts and Sciences, North-Holland. Braunmüller, Kurt. 2007. Receptive multilingualism in Northern Europe in the Middle Ages. In Jan D. Ten Thije & Ludger Zeevaert, Receptive multilingualism, 25–47. Amsterdam: Benjamins. Ceccherini, Irene. 2009. Merchants and notaries: Stylistic movements in Italian cursive scripts. Manuscripta 53. 239–83. Elspass, Stephan. 2012. The use of private letters and diaries in sociolinguistic investigation. In Juan M. Hernández-Campoy & J. Camilo Conde-Silvestre (eds.), The handbook of historical sociolinguistics, 156–169. London: Wiley-Blackwell. Elspass, Stephan. 2005. Sprachgeschichte von unten: Untersuchungen zum geschriebenen Alltagsdeutsch im 19. Jahrhundert. Tübingen: Niemeyer. Fennell, Barbara A. 2001. A history of English: A sociolinguistic approach. Oxford: Blackwell. Goitein, Shelomo Dov. 1971. A Mediterranean society: The Jewish communities of the Arab world as portrayed in the documents of the Cairo Geniza. Vol II: The community. Berkeley: University of California Press. Graff, Harvey J. 1987. The legacies of literacy: Continuities and contradictions in Western culture and society. Bloomington, IN: Indiana University Press. Guggenheim-Grünberg, Florence. 1954. The horse dealers’ language of the Swiss Jews in Endingen and Lengnau. In Uriel Weinreich (ed.), The field of Yiddish, 48–62. New York: Linguistic Circle of New York. Hanawalt, Barbara. 1995. Growing up in medieval London: The experience of childhood in history. Oxford: Oxford University Press. Heine, Bernd. 1970. Status and use of African Lingua Francas. München: Weltforum Verlag. Hsy, Jonathan. 2013. Trading tongues. Merchants, multilingualism, and medieval literature. Columbus, OH: Ohio State University Press. Ioppolo, Grace. 2010. Early Modern handwriting. In Michael Hattaway (ed.), A new companion to English Renaissance literature and culture [2 vols], l.177–189. Chichester: Wiley-Blackwell. Jucker, Andreas H. & Joanna Kopaczyk. 2013. Communities of practice as a locus of language change. In Joanna Kopaczyk & Andreas H. Jucker (eds.), Communities of practice in the history of English (Pragmatics and Beyond New Series 235), 1–16. Amsterdam/Philadelphia: John Benjamins. Kaislaniemi, Samuli. 2009. Encountering and appropriating the Other: East India Company merchants and foreign terminology. In Arja Nurmi, Minna Nevala & Minna Palander-Collin (eds.), The Language of Daily Life in England 1450–1800 (Pragmatics and Beyond New Series 183), 219–251. Amsterdam/Philadelphia: John Benjamins. Kaislaniemi, Samuli. Forthcoming. Visual aspects of code-switching in Early Modern English manuscript letters and printed tracts. In Reading the page: Verbal and visual communication in Early English texts. Turnhout: Brepols. Khan, Geoffrey. 1995–1997. A note on the trade argot of the Karaite goldsmiths of Cairo. Mediterranean Language Review IX. 74–76. Labov, William. 1972. Sociolinguistic pattern. Philadelphia: University of Philadelphia Press. Locker, Kitty O. 1987. “As per your request”: A history of business jargon. Journal of business and technical communication 1. 27–47. Milroy, James & Lesley Milroy. 1985. Linguistic change, social network, and speaker innovation. Journal of Linguistics 21. 339–384.

16 

 Esther-Miriam Wagner and Bettina Beinhoff

Milroy, Lesley. 1987. Language and social networks (Language in Society 2), 2nd edn. Oxford: Blackwell. Milroy, Lesley & Muysken, Pieter 1995. One speaker, two languages: Cross-disciplinary perspectives on code-switching. Cambridge: Cambridge University Press. Parkes, Malcolm. 1973. The literacy of the laity. In David Daiches & Anthony Thorlby (eds.), The Mediaeval world, 555–577. London: Aldus Books. Poplack, Shana 2001. Code-switching (linguistic). In Neil J. Smelser & Paul B. Baltes (eds.), International encyclopedia of the social and behavioral sciences. Vol. 12, 2062–2065. New York: Elsevier. Rissanen, Matti. 2000. Standardisation and the language of early statutes. In Laura Wright (ed.), The development of Standard English, 1300–1800, 117–130. Cambridge: Cambridge University Press. Stenroos, Merja. 2004. Regional dialects and spelling conventions in Late Middle English: searches for (th) in the LALME data. In Marina Dossena & Roger Lass (eds.), Methods and data in English historical dialectology, 257–285. Bern: Peter Lang. Thomason, Sarah G., & Kaufman, Terrence 1988. Language contact, creolization, and genetic linguistics. Berkeley: University of California Press. Thrupp, Sylvia. 1948. The merchant class of medieval London 1300–1500. Chicago: University of Chicago Press. Trotter, David. 2011. Italian merchants in London and Paris: Evidence of language contact in the Gallerani accounts, 1305–08. In Dominique Lagorgette & Tim Pooley (eds.), Mélanges pour R. Anthony Lodge, 209–226. Chambéry: Presses de l’Université de Savoie. Verstegh, Kees. 2008. Non-Indoeuropean pidgins and creoles. In Silvia Kouwenberg & John Victor Singler, The handbook of pidgin and creole studies, 158–186. Malden: Wiley-­ Blackwell. Wagner, Esther-Miriam. 2013. Challenges of multiglossia: the emergence of substandard Judaeo-Arabic registers. In Esther-Miriam Wagner, Ben Outhwaite & Bettina Beinhoff (eds.), Scribes as agents of language change, 259–273. Berlin: de Gruyter-Mouton. Wagner, Esther-Miriam. 2017. A matter of script? Arabic and Judaeo-Arabic in the Genizah Collections. In Yousef Meri & Camilla Adang (eds.), Muslim-Jewish relations in past and present: A kaleidoscopic view, 115–136. Leiden: Brill. Walter, Mary Ann. 2003. Kalaam, Kalaarbaam: An Arabic language game in Hadramaut, Yemen. Paper given at the 10th Annual Symposium about Language and Society – Austin (SALSA). Downloadable at: http://faculty.wcas.northwestern.edu/~maw962/docs/salsapaper.pdf Wright, Laura. 1994. Early modern London business English. In Dieter Kastovsky (ed.), Studies in Early Modern English (Topics in English Linguistics 13), 449–465. Berlin and New York: Mouton de Gruyter. Wright, Laura. 1998. Mixed-language business writing: Five hundred years of code-switching. In Ernst H. Jahr (ed.), Language change: Advances in historical sociolinguistics (Trends in linguistics: Studies and monographs 114), 99–118. Berlin and New York: Mouton de Gruyter.

II Literacy of traders and their agency as linguistic trendsetters

Merja Stenroos

2 Like the coins when currencies are combined: contextualizing the written language of fifteenth-century English merchants Abstract: This chapter discusses the role of merchants as innovators in late medieval English writing. It presents two case studies of fifteenth-century school texts with merchant associations, and considers them as evidence of the education and literacy practices of late medieval English merchants. It suggests that a multilingual habit, as well as the perpetuation of local and regional forms, is brought from school education into business. The implication is that trade connections in the fifteenth century lead to regional mixing rather than the spread of a single standard, creating the basis for the dialectally mixed character of what became Standard English.

1 Introduction The present paper addresses the question of merchants as innovators in fifteenthcentury English writing, with reference to their literacy practices and educational background. It argues that the concept of “pragmatic literacy” (Parkes 1973: 555), with its assumptions of basic, and mainly vernacular, literacy skills does not necessarily give a fair idea of the writing competence of late medieval English merchants. It presents two case studies of mid-fifteenth-century school texts with merchant associations, both of which reflect a highly complex, bilingual literacy, combining the use of Latin as a first written language for work purposes with a highly innovative and regionally coloured use of English. Textbooks of the history of English often cite the emergence of the “middle classes” in the late medieval period as one of the formative factors in the development of English. However, they are not always entirely clear about what precisely this entails. Most writers seem to indicate a general contribution to the ongoing vernacularisation of writing, by providing a non-Latinate reading public (see e. g. Baugh and Cable 2002: 142–43; Fennell 2001: 96–97, 121). Fennell (2001: 97) also suggests a more active role in the development of written English: Merja Stenroos, University of Stavanger DOI 10.1515/9781501503542-002

20 

 Merja Stenroos

mobility was the key to the standardization of English, and (...) merchants, tradesmen and members of city government had a central role to play in the establishment of a standard code.

To what extent the development of late medieval written English should be discussed in terms of “standardization” is a controversial question that need not be dealt with in detail here. As the term carries associations of institutional control and prescriptivism, which are not necessarily appropriate for this period, it is not used in the present paper. Instead, the development of written language is here discussed in terms of innovation (the introduction of new variants into a language community) and supralocalisation (an increase in the relative frequency of forms with a wide geographical currency at the expense of local or regional forms). While these processes may be thought of as “standardizing” from a retrospective standpoint, they were not necessarily unidirectional or consistent; at the same time, they are at least to some extent objectively measurable within the limitations of the surviving material. Studies of present-day spoken communities often connect linguistic innovation with speakers characterized by “weak ties”; such speakers tend to be socially and geographically mobile and communicate with people belonging to several networks (Milroy and Milroy 1985: 363–70; Milroy 1992: 176–185; for a historical application of network theory see Bergs 2005). Of the groups listed by Fennell, merchants could be assumed to be particularly mobile, and to have large “weak tie” networks in the form of trade contacts; they might consequently be expected to form a relatively innovative group of speakers. It is not, however, self-evident to what extent the same assumption might work for the written mode. How far the written variation in medieval English may be expected to follow the general principles established for spoken variation is a major research question that requires much further study, including detailed investigations both of the texts produced by medieval writers and of their social and educational context. In late medieval England, merchants were one of the few sizeable groups of lay people who were to a large extent literate and produced an extensive amount of text. Their texts have, however, so far attracted the interest of few linguistic scholars: major exceptions are the work of Laura Wright on London business records (see e.g. Wright 2000, 2005, 2012) and studies of the correspondence of the Cely family (e. g. Hanham 1975, 1985; Raumolin–Brunberg and Nevalainen 1997; Rutkowska 2003). There are several reasons for this relative lack of interest. The texts are for the most part of a highly pragmatic nature, produced to record transactions, disputes or inventories, or to convey orders; they are generally not of great literary or even historical interest. They are seldom made available in editions, and the originals may be difficult to read and cumbersome to access. In addition, many of them are



Like the coins when currencies are combined 

 21

multilingual and so have been ignored by researchers who are working within the confines of one particular language. A large corpus of commercial texts, both monolingual and multilingual, is therefore a major desideratum in the study of the historical development of English. Apart from the commercial texts themselves, an appreciation of the role of merchants in the development of written English is based on our assumptions about this social group as writers and users of texts. Several scholars have discussed the educational background of merchants, especially those based in London; estimates of their level of learning range from the generous (Thrupp [1948: 161] argues that most London merchants had some training in Latin, and Hanawalt [1995: 82] refers to the literacy requirements of fifteenth-century London guilds as “severe”) to the dismissive (according to Graff [1987: 98], hardly any guild members could read Latin and only few could read or write English). The educational level of merchants both in London and elsewhere is likely to have varied greatly, and one should beware of firm generalizations from individual examples. At the same time, it may be useful to consider the direct evidence that survives. While the educational background of most individual Middle English writers is unknown to us, numerous school texts, both in Latin and English, survive from the fifteenth century. Two such manuscripts, both containing English, have been identified as relating to a commercial education; one of these was used by its owner as a notebook through a long career as a merchant. The present paper considers these two manuscripts as evidence of the education and literacy practices of late medieval English merchants, and attempts to draw some implications about the role of these merchants as innovators in the development of written English.

2 The education of merchants An English astrological treatise of the fifteenth century, The Thirty Days of the Moon, gives a useful description of merchants as they were viewed in late medieval society. It describes the lucky and unlucky features of each day of the lunar circle, including the prospects of the children born on each day. The treatment of the nineteenth day begins like this: (1) In the xix day I wotȝ wel yat / Isaac iacob hys sone be-gatȝ yat day is good for many ying / For bying for sellyng for ye se passyng And for to leren of marchaundyse / And of al maner werkys of prise (...) Quat chyld so yat day born is / he sal ben wys and trewe jwys wrchepeful and wys of lore / hys rychesse xun waxin euere more & more Quat manys chyld so it be / hardyly it sal wel the

22 

 Merja Stenroos

[On the 19th day I know well that Isaac begat Jacob his son. That day is good for many things: for buying, selling and crossing the sea and for learning about merchandise and all kinds of worthy occupations (...) The child who is born that day will certainly be wise and honest worshipful and knowledgeable; his wealth will grow ever more Whatever man’s child it may be, it will prosper well] (Thyrtty days of the mone, London BL Harley 1735, fols 9r–v) While the poem does not claim that all the points necessarily relate to merchants, it is instructive to consider the attributes here combined with “buying and selling”. First, there is the immediate connection with crossing the sea: merchants were highly mobile people with trade connections both within the country and overseas. Secondly, commerce was a route to wealth, for which one did not require a specific kind of social background; most merchants were of the “middling sort” (see Nevalainen 1996: 58). The most important point for the present purpose is, however, the reference to education: the poem makes clear that being a merchant was a specialized craft, requiring skills that had to be learnt. There are numerous records of sons who were put to “leren of marchaundyse” (see Orme 2006: 68–70; Moran 1985: 69). As with any trade, learning to be a merchant would involve many years’ training as an apprentice serving a master. In addition, however, a period of formal schooling seems to have been considered highly desirable, and many guilds required a certain level of literacy of new members (see Hanawalt 1995: 82–83). The contract for apprenticeship sometimes included an agreement that the master should provide the apprentice with schooling. An example of such an agreement is referred to in a letter of complaint held in the National Archives (PRO C 1/19/1–2); this letter, written in 1450 on behalf of an apprentice haberdasher, Thomas Bodyn, provides an interesting statement about the value of education for a mid-fifteenth-century merchant. Bodyn seeks redress because his master, Robert Chirche, “citizen and haberdasher of London”, had not honoured an agreement to provide him with schooling as part of his apprenticeship. According to his complaint, he had been promised eighteen months of grammar and half a year of writing, specified in a contract made when he was fourteen years of age: (2) that the said Robert shuld fynd to scole at hys awen costes and charge the said Thomas duryng two the furst yeres of the said terme that is to say a yere and half therof to lerne gramer and the resydue of the seyd two yeres which amounteth to half a yere to scole for to lerne to write



Like the coins when currencies are combined 

 23

[that the said Robert should sponsor the said Thomas to school at his own cost during the first two years of the said term of apprenticeship, that is to say, a year and a half thereof to learn grammar and the remainder of the said two years, which amounts to half a year, to go to school in order to learn to write] The complaint was filed eight years later, when Bodyn reports that his master had refused to honour the agreement, “to the grete hurte harme and losse of the said Thomas”. While it is to be expected that a complaint emphasizes the wrong suffered, the “great hurt, harm and loss” would not have been credible unless the kind of schooling described was considered important, or at least highly desirable, for a successful career as a haberdasher. It is therefore of some interest to enquire into the kind of education outlined here, and the level of learning Bodyn would have been expected to reach. Latin grammar formed the basis of the kind of education taught at what might be referred to, in modern terms, as the secondary level: it presupposed knowledge of “the letters”, that is, a basic reading competence. Writing was sometimes taught in separate schools; according to Moran (1985: 50) it seems to have been “considered part of a more elementary curriculum” compared to that of the grammar schools. It is possible that Bodyn’s schooling would simply have provided him with basic writing skills. However, as the wording seems to imply a chronological order where “writing” is preceded by grammar studies, one could ask whether a more advanced training might have been intended. A future merchant would certainly need to learn more than simply the production of letter shapes. Orme (2006: 68) lists the following subjects among the “arts of business”: – “dictamen” (the art of writing letters) – the methods of drafting deeds and charters – the composition of legal records – the keeping of financial accounts While letters were increasingly written in English in the fifteenth century, Latin was still the dominant language of legal documents and accounting. Accordingly, a future merchant had to study Latin before proceeding to the arts of business, and Thrupp (1948: 161) argues that most merchants in London in the fourteenth and fifteenth centuries had training in Latin. There is no way of estimating what proportion of all English merchants may have had a grammar school background. However, judging from the numerous

24 

 Merja Stenroos

references to the schooling of merchants, this must have been considerable at least among the wealthier merchant families, even though the boys may not necessarily have spent a very long time at school. Orme (2006: 130) suggests that “[a] couple of years of grammar might provide the basic skills required to follow a trade or business”. Some sons of merchant families did, however, spend considerably longer at school: Hanawalt (1995: 83) refers to fifteenth-century merchants’ wills that provide for their sons’ schooling up to the age of fifteen or sixteen. Thomson (1979: 15) also suggests that some schools were particularly oriented towards the kind of education required by merchants: The flourishing grammar school at Beccles (...) drew on the commercial prosperity of East Anglia and its merchant class, and [Cambridge University Library MS Add. 2830] gives a more practical and down to earth – and a times humorous – approach to grammar (...) than the systematic formality of [Cambridge, Trinity College MS 0.5.4]. Religious matter such as Drury’s De modo confitendi is not of course excluded from the teaching, but the orientation seems to be towards the pragmatic and “cultivated” literacy of the merchant rather than to the professional literacy of the clergy and academics.

Bearing in mind the explicitly stated order of the elements of Bodyn’s planned education, the contents of the second part of the study might have included the composition of business texts as well as training in the physical skill of handwriting. By 1450, his education would certainly have included reading, and perhaps writing, both in English and Latin. While the details of the education intended for Bodyn are lost to us, it may be possible to reconstruct something of its likely content and scope through the study of the two school manuscripts discussed below.

3 The manuscript evidence 3.1 The survival of school texts containing English By the fifteenth century, English and bilingual texts featured prominently at the earlier stages of grammar school education (see Orme 2006: 106–109). English school texts probably existed from at least the mid-fourteenth century (Thomson 1984: xiv), but no such early texts have survived; only from about 1400, the availability of paper made it possible for pupils to reproduce their school texts in a permanent form (Orme 2006: 111).



Like the coins when currencies are combined 

 25

Thomson (1979) catalogues 36 manuscripts that contain English grammar texts. By far most of these may be classified as “school texts” in the sense that they were compiled or produced by pupils for their own use; all of them are multilingual. Of the 36 manuscripts, 28 contain versions of the four major grammatical treatises in English that were circulating in the fifteenth century: Accedence (an English version of Donatus’ Ars Minor), Comparacio, Informacio and Formula. In addition, the method of teaching by means of translation exercises – “Latins” and “vulgaria” – ensured plentiful training in the writing of English, as well as models for spelling. Two of the catalogued manuscripts were identified by Thomson (1979: 18) as having a probable merchant context: Cambridge, St John’s College F 26 and Oxford, Bodleian Library, Rawlinson D 328. Both are connected to large grammar schools – St Albans and Exeter respectively – and seem to have been compiled by pupils who received training for a business career. Both these manuscripts thus provide direct evidence of the kind of education that future merchants received; in addition, Rawlinson D 328 also throws light on the writing habits and literacy practices of a grown-up merchant.

3.2 Cambridge, St John’s College F 26: Thomas Marchall Cambridge, St John’s College F 26 is a composite manuscript that was originally put together from several existing quires and then continued by several pupils at the St Albans grammar school. Thomson (1979: 18) suggests that this manuscript belonged to a merchant context: “as with Rawlinson – one of the Exeter manuscripts – we find legal and business formulae in St John’s which suggest that the literacy acquired would be used in commerce rather than the church”. The legal and business formulae are found in the third quire, which is part of the portion written by Thomas Marchall of Toddington, Beds. He is identified in a form of obligation (fol. 25r), dated to 1439: (3) Nouerint vniuersi per presentes me Thomam Marchall de Todyngton’ in Com’ Bed’ Clericum teneri & firmit’ obligari Johanni’ Marchall de eadm’ in xx. libr’ bone & legalis monete Soluend’ eidm’ Johi’ aut suo certo attornato ad Festu’ sci’ Michael prox’ futur’ Post data’ presencium Ad qua’ quidm’ soluconem bn’ & fidelic’ faciend’ obligo me heredes meos & excecutor’s meos. per presentes sigillo meo sigillat. Dat’ decimo die Mensis Maij Anno regni regis henrici sexti post conqm’ Septimodecimo

26 

 Merja Stenroos

[May it be known to all by these presents that I, Thomas Marchall of ­Toddington in Co. Bedford, clericus, am held and firmly bound to John Marchall of the same place in twenty pounds of good and lawful money, to be paid to the said John or to his certain attorney by the Feast of St Michael that follows next after the present date, for which payment well and truly to be made I bind myself, my heirs and my executors firmly by the present document sealed with my seal. Given the tenth day of the month of May in the seventeenth year of King Henry the sixth after the Conquest] The title clericus could in this period be used to refer to anyone who attended school, and many pupils were tonsured irrespective of their later career plans (Orme 2006: 135). Thomas seems to have had a gentry background; a John Marchall, gentleman, of Toddington appears as a defendant in the court of common pleas in 1466, accused of an armed robbery thirteen years earlier. The portion of the manuscript that is written by Marchall consists of 25 folios, gathered into two quires (fols 25–34 and 35–49 respectively). It contains the following main items (cf. Thomson 1979: 157): – Treatises on vowel quantity – Verses on Latin hexameters – A poem on cockfighting – De accentu by Mag. Thomas Sylton – A baccalaureate exercise treating a theme in prose, “rhythmus” and verse – Extracts from a grammatical treatise – Phrases and proverbs in English and Latin – An abridged version of Disticha Catonis – Another treatise on vowel quantity – Rhythmical verses on St Nicholas Most of these items are related to the composition of Latin verse, generally taught at the most advanced stage of a grammar school education (Orme 2006: 151–52). The poem on cockfighting (fol. 28v–29r) is provided with a note that it was produced by Marchall himself, presumably as an exercise in versification. It may be noted that both manuscripts discussed here contain a Latin poem on cockfighting; according to Orme (2006: 157), this was an extremely popular pastime at medieval English schools, especially on Shrove Tuesday. The main items are accompanied by a large number of scribbles and additions, including numerous legal formulae and drafts of legal documents. The latter involve a John Wayte of Toddington as well as Thomas and John Marchall.



Like the coins when currencies are combined 

 27

Almost all the contents are in Latin. This is also the case with the remainder of the manuscript, which consists of 189 folios and seems to have been put together by several consecutive pupils at St Albans; people who contributed to the manuscript include Robert Wulrey of London diocese and “Auncell”, two of which names became monks of St Albans in the latter part of the fifteenth century (Thomson 1979: 157). The only sizeable English text in the manuscript is the introductory grammar text Accedence, which is the first item in the manuscript (fols 1r–12r). In Marchall’s portion, the only English is found in the set of phrases and proverbs on fol. 44r, which include the following: (4) Syche as he is syche he wyl schewe [such as he is, so he will appear] Whan þu begynne a þyng þenk on þe endyng [when you begin a thing, think about the ending] Be gold neuer so red for bred it must go [no matter how red the gold, it will have to go for bread] A fastyng wombe makyþ no joye [a fasting stomach makes no joy] þe blynd etyth many a flye [the blind man eats many a fly] The English dialect of these sentences has a strongly regional character, suggesting an East Anglian provenance: forms include syche ‘such’, y-3oue ‘given’, maad ‘made’: chidryn ‘children’, xal ‘shall’. The third-person singular ending is regularly -yth and most past participles have the prefix y- (y-3oue ‘given’, y-3ete ‘eaten’). The language may be considered strongly regionally marked and contains some idiosyncracies, such as the introduction of into spellings such as knhyfe ‘knife’. The part of Marchall’s education that is recorded in the manuscript has clearly passed the initial stage of grammar training, and is focussed on two areas: versification and writing. In particular, the latter part of the portion written by him includes a large amount of handwriting practice, with letter forms, words and phrases (Amen; Est nichil) repeated across the margins. He seems to have been practising calligraphy and different script types, including textura. He also uses legal formulae for writing practice: so Nouerint universi per present’ nos Thomam Marchall (fols 25r, 40v, 43v) and Sciant present’ & futuri (fol. 40v). He frequently signs his work; the formula quod Marchall appears on fols 45v and 49r. It may be noted that the writing practice here carried out by Marchall clearly belongs to a fairly advanced stage of school education, rather than to elementary

28 

 Merja Stenroos

education. At this point, he appears to be concerned with the mastering of scripts that are appropriate for different purposes; the practice is combined both with training in versification and with the drawing up of business documents. It is unclear what eventually became of Thomas Marchall. He need not have become a merchant, but may have continued his studies: a Thomas Marchall was appointed vicar of Willington in Bedfordshire in 1463, although there may have been several of that name. Whatever the case, the manuscript reflects a type of education that combined relatively advanced Latin studies with the kind of clerical skills required of both merchants and administrative clerks: writing practice and the drawing up of legal documents, all at the same time and often on the same page.

3.3 Bodleian Library, Rawlinson D 328: the book of Walter Pollard Bodleian Library, Rawlinson D 328 is a manuscript of 194 folios. In the fifteenth century it belonged to the Plymouth merchant Walter Pollard. References in the text suggest that he attended Exeter city grammar school in the mid-­fifteenth century and produced a large part of the contents of the book while there (Thomson 1979: 301). According to a note on fol. 179r, he acquired the book, or parts of it, in 1444 or 1445: (5) Sciant presentes et future (sic) quod iste liber ad diem confeccionis presencium fuit in custodia Waltero Pollard qui ipsum habuit ex dono et concessione Thome Ismay clerice, hijs testibus Richardo Coke, Wyllelmo Payge, Iohanne Relond et alijs. Dat’ apud Plymmoth anno regnis regi Henrici sexti xxmo iijo. [May it be known to the present and future that this book from this day will be in the care of Walter Pollard who received it as a gift from Thomas Ismay, clericus, as testified by RC, WP, JR and others. Given at Plymouth in the 23rd year of the reign of King Henry VI] The manuscript contains a collection of grammar texts both in Latin and English, including versions of all the four main English grammatical treatises, as well as a range of other texts belonging to a school context, including word lists, collections of proverbs, lists of Latin verbs and practice verses. In addition, it contains numerous notes, scribbles, letter drafts and copies or drafts of business documents.



Like the coins when currencies are combined 

 29

The first quire of the manuscript (fols 1–6) was clearly originally separate; it contains a popular school text used for learning Latin, Disticha Catonis, written in a rather formal hand which may be characterized as a hybrid Anglicana, and which Thomson (1979: 290) dates to around 1400. The last folio of the manuscript (fol. 192) contains medical recipes and charms, in Latin and English, which are also written in a hand clearly different from the rest. The main part of the book, fols 7–191, contains texts and scribbles produced, it seems, chiefly by Pollard himself. There are a few other hands involved: fols 7r and 8–15 contain school texts written by another hand, and occasional short texts throughout the book appear in other hands, which seem to be those of fellow pupils and family members, including a John Pollard (fol. 191r). What makes Rawlinson D 328 of special interest is that it appears to have been continued to be used by Walter Pollard as his personal notebook over a long period of time. The latest dated entry in the book is from 1483, almost 40 years after Walter Pollard received the book as a schoolboy. Among the various hands that contribute to the book, it is possible to distinguish several that seem to represent Walter Pollard himself, at different times and doing different things: the majority of the manuscript, excluding the passages specified as being by the other hands, is in a group of hands including a rapid mixed bookhand (...) a version of this with more sharply angled strokes (...) a spindly mixed document hand (...) and a version of this with the ascender and descender coinciding (...) These hands possibly all represent the work of Walter Pollard and show his handwriting developing over perhaps 40 years (Thomson 1979: 290–91)

A detailed discussion of the paleographical and linguistic variation in the manuscript falls beyond the scope of this paper. However, there are two grounds for accepting Thomson’s suggestion that all these hands represent Pollard himself: firstly, they very clearly shade into one another, and, secondly, he identifies himself in all of them, whether as a first-person reference in a legal phrase (Be yt yn remembrans and see that I, Water Pollard of Plymmouth, hath ipaid, fol. 139r; also fols 7v, 133v, 150v, 191r) or a straightforward signature (Walterus Pollard, fols 67r, 74v, 120v, 175r, 190v, 193v); in addition, the formula quod Pollard appears (fols 106v, 123r), although it could, of course, in principle relate to other members of the same family as well. There is a general progression in the manuscript from school texts and letters, written by the first two hands, to business notes and accounts written by the two latter ones. However, notes in the late hands appear throughout the manuscript, suggesting that the grown-up Walter Pollard was adding scribbles wherever there was room.

30 

 Merja Stenroos

The first hand is very neat and rounded, and seems to be that of Pollard as a schoolboy, copying grammatical treatises and exercises: texts copied in this hand include the Formula, a poem on table manners known as Stans puer ad mensam, several bilingual proverb collections, translation sentences and word lists, as well as “a little book of doctrine for young gentlemen”, containing a list of so-called “terms of association” (a pride of lyons, a shrewdnes of apes). The second, a more angular and cursive hand, appears in letter drafts and writing exercises, at least partly still produced at school. Some of the English letters refer to Pollard’s apprenticeship: (6) Ryzth reuerent And Whell Be-lowhyd Fader J Ge grete yow whell y promyste yow att exceter that y whold be A Tome wt yow that whyke Fowluyng’ but y Feythe yt whas þwors-day nytthe whar Ar that my mayster came home than y preposyd to be wt yow on’ mon-day next wholuyn’ than y aykyt lewe of my mayster than he sayd ye must ryd wt my lade to excet[...] ... or hels ye must ryd wt me ynn to korwy[...] A hontyng’ ans (sic) so y fayt y chan haue no lazer to co[...] onto [rest illegible] (fol. 182v) [Most honourable and well-beloved Father, I greet you well. I promised you at Exeter that I would be at home with you the following week, but in faith it was Thursday night before my master came home, then I purposed to be with you on the Monday next following; then I asked leave of my master; then he said: you must ride with my lady to Exet[er] (...) or otherwise you must ride with me to Corwe[n] to hunt, and so in faith I can have no leisure to co[me] to (...)] This hand is also responsible for the version of Accedence described by Thomson (1979: 299) as “the first part only and in a very confused state”, with the opening lines (How many partes of Speche be ther. viij . Which viij., etc.) repeated and crossed out on several folios from fol. 120 onwards. These folios also include various other texts, including drafts of letter openings, proverbs and the like, by the same hand. It would appear most likely that the text of Accedence, as well as the other, miscellaneous texts, are not here copied because of their content but as handwriting exercises: this introductory grammar treatise would have been highly suitable for this purpose, assuming that it had been memorized at an earlier stage.



Like the coins when currencies are combined 

 31

If this is the case, it would seem that the two hands mainly belong to two different stages of Pollard’s upbringing and seem to represent two different kinds of writing. Hand A seems to be mainly copying texts from exemplars or dictation, with the purpose of recording them: Pollard is, basically, here doing the work of a manuscript copyist. Hand B, on the other hand, appears to be used for writing exercises and for the drafting of Pollard’s own letters. Corresponding to the difference in function, this handwriting appears bolder and more calligraphic, but not as neat as that of the grammar texts. The writing exercises appear to belong to a fairly advanced level of schooling, rather than to elementary writing practice, and include business formulae as well as parts of the Accedence. Like Marchall, Pollard seems to be practising different kinds of handwriting, although there are no attempts at decorative calligraphy or at a textura script. Finally, there is a quite different, spindly hand which would seem to represent Walter Pollard in his older age. The texts produced by this hand are mainly texts relating to business and finances, draft documents and business letters, both in Latin and English; a Latin bond on fol. 74v is dated to 1475 and signed “Walterus Pollard”. In addition, this hand has contributed two of the very few overtly religious texts found in the manuscript: a form of confession (fol. 130r) and an English verse text headed Inter diabolus et virgo (fol. 174v–175r). It also contributes some practical texts that would be highly appropriate for a grown-up merchant: sailing instructions to Berwick (fol. 183r) and a remedy for sore eyes (fol. 191v). A very large proportion of all the texts in the manuscript are mixed in language, containing both English and Latin. For the most part this is of course endemic to the genre: introductory books teaching a foreign language, such as grammar texts and vocabulary lists, are bilingual almost by definition. However, the bilingual practices in the manuscript are not restricted to such texts, but are found everywhere, including draft business documents and numerous casual scribbles. These include a little piece of insulting verse, perhaps penned by a fellow pupil, following Pollard’s formulaic Qui librum scripcit Walterus Pollard benedixit: (7) Walterus Polard non est but a dullard y saye that pollard ys none wery gullard Quod scripci scripi qd nolo negari et hoc verbum dubium sech Bilingual scribbles appear in many school books, and clearly reflect the linguistic context of school life, not just with regard to introductory studies but also to social life: at least in the better grammar schools, pupils were expected to speak

32 

 Merja Stenroos

Latin (Moran 1985: 36), and spoken communication outside the classroom was likely to be bilingual. More interestingly, perhaps, bilingual scribbles are still added by the late spindly hand, as on the otherwise blank verso side of fol. 150: (8) Be yt in reme’br’ that Ja Water pollard Quid abemus amplius ac vita nisi vapular’ Et Jocundare victum & vestitum veraciterque lugar’ No mor’ vnto at this tyme Butt good haue yow It is not obvious what language Ja is meant to represent, although it is clearly a form of the first person pronoun. The texts in all the hands that are identified with Walter Pollard are similar in dialect; however, there is slight variation both within and between the texts produced by each hand, especially in the relative frequencies of variants. All contain certain characteristic spellings, including thys ‘these’, byt ‘are’, goyt(h) ‘goes’, ys ‘his’ and the use of and where and would be expected. The dialect is strongly Southwestern and seems to be retained through life; there are no more signs of standardization in the later entries than in the earlier ones. Pollard uses regularly forms such as thay ‘they’ and them ‘them’, which were still mainly northern during this period; however, according to the maps in the Linguistic Atlas of Late Mediaeval English (McIntosh, Samuels and Benskin 1986) they appear commonly in southwestern texts as well. The angular hand stands out as being the most idiosyncratic in its spellings, including a very variable use of h, with numerous forms such as whell ‘well’, be-lowhyd ‘beloved’, whold ‘would’, a-tome ‘at home’, wholuyn ‘following’, chan ‘can’. These spellings may perhaps be related to the kind of texts produced in this hand. With the exception of short samples of grammar text, used as writing practice and probably memorized, these texts seem to be compositions, unlike the texts written in the rounded hand, which very clearly represent copying or writing from dictation. At least in the former case, the forms of the exemplar would partly override the scribe’s own forms, which might account for the less idiosyncratic spelling. The spindly, grown-up hand still produces the marked forms, but not nearly as frequently as the angular one; decades of literacy practices including correspondence and reading would presumably have changed his writing habits towards the mainstream. A draft bond on fol. 189 describes Pollard’s father, John Pollard, as gentylman and Walter himself as merchaunt. The reference to hunting in Pollard’s letter to his father, as well as some of the content of the school texts place Pollard into the social context of country gentry; at the same time the numerous references to business transactions identify him as a Plymouth merchant dealing in cloth. His schooling appears to have been of a relatively practical kind, and the proportion



Like the coins when currencies are combined 

 33

of English texts is very high compared to that of the St John’s manuscript. At the same time, the sheer quantity of text, including numerous long Latin works, suggests that Pollard must have spent quite some time at school. This is also suggested by some of the contents. Like Marchall’s book, Pollard’s includes a Latin poem on cockfighting (fol. 72r), written in three different metres. As Latin versification was generally learnt at the highest level of grammar school, the presence of the poem may suggest that Pollard’s education was, in fact, a fairly long one. Unlike Marchall’s poem, however, the one in the Rawlinson manuscript contains no note of authorship: Pollard may have copied it as a model, or simply because he found it entertaining.

4 The education of Pollard and Marchall: what kind of literacy did they achieve? In a highly influential paper, Parkes (1973: 555) classified medieval literacy into three major areas: professional, cultivated and pragmatic. There is a very clear distinction of levels between these categories. The “professional man of letters” was the scholar. Scholarly literacy took years of university study and was far more demanding than “cultivated literacy”, that of the aristocrat or the well-to-do merchant or lawyer who had enough spare time to engage in leisurely reading. Pragmatic literacy, finally, was the most basic of all and would not require expensive years of schooling: Although at this stage all accounts were drawn up in Latin, the ability to draw up an account would not necessarily require a knowledge of the Latin language as such. It would require only a knowledge of formulae. Latin terms found in the accounts could probably be regarded as the jargon of the trade (Parkes 1973: 559)

While Parkes is here referring to manorial account-keepers, the same would to some extent be true of the requirements of a merchant; however, there is no doubt that many merchants by the fifteenth century had a competence in Latin that went well beyond “a knowledge of formulae”. The main focus of Parkes’s paper is on reading habits and the shift from pragmatic to cultivated literacy: while he discusses briefly the production of documentary texts in connection with pragmatic literacy, his main interest is in literary texts and he deals with “the emergence of [the] rising middle class as a class of cultivated readers”. The categories clearly work best when applied to reading habits; they are somewhat more problematic when applied to text production. While the well-to-do merchants in Parkes’s classification might be “cultivated

34 

 Merja Stenroos

readers”, consuming vernacular literature in their spare time, this label does not give a fair representation of their work literacy, which was both pragmatic and to a large extent Latinate. Latin continued to be the dominant language of administration and business well beyond the Middle Ages. Even though English documents start appearing from the early fifteenth century, and English is adopted in some institutions and genres earlier than in others, in administrative texts overall English is still a minority language in the sixteenth. The slow adoption of English in business writing follows naturally from the role of Latin in education: grammar school education was largely about learning to speak and write Latin (Moran 1985: 39). For writers with a grammar school background, Latin would in effect function as the “first language” for writing purposes, at least within specific genres. All the terminology required for business texts was readily available in Latin, and there were established conventions of style, making Latin a far easier choice for the competent writer. At the same time, a much larger proportion of readers would be able to read and understand English, and the choice of language in late medieval English administrative writing probably reflects the competence of intended readers as well as of writers. The education at St Albans, as reflected in St John’s College F 26, was firmly Latin-based, even though Marchall also copies a set of proverbs with English renderings. Pollard, on the other hand, produces copious amounts of text both in English and Latin, as well as a considerable amount of bilingual, or “macaronic” writing. The English produced by both Marchall and Pollard is remarkably strongly dialectal, especially considering that they write at a time when, according to the compilers of A Linguistic Atlas of Late Mediaeval English (MacIntosh, Samuels and Benskin 1986, I: 3,), dialectal texts had “become rare” in the south. Strongly regional or local writing habits seem to be, indeed, typical of school texts of this period (cf. also Stenroos 2016). This regionality presumably reflects the fluidity and local character of the texts, which would generally derive from the personal copies of the schoolmaster, modified for his own use with a particular group of recipients in mind (Thomson 1979: 3–4; Orme 2006: 182–83). The kind of literacy achieved by Marchall and Pollard was thus bilingual, with Latin learnt systematically as the main medium of writing, and English developing as a second, regional and variable, but more reader-friendly, option. This would also have been the kind of literacy aspired to by Thomas Bodyn, even though he would not have reached the level of Latin versification. All three were trained in a period when the use of English in business writing was expanding, and when written English was going through a period of rapid development,



Like the coins when currencies are combined 

 35

marked by a strong tendency towards supralocalisation. It remains to consider what role writers such as Marchall and Pollard may have played in this development.

5 Concluding discussion: merchants as a force of innovation It has been suggested by several scholars that the language of business and administrative texts develops quite differently from that of literary texts in this period (see e. g. Benskin 1992: 75; Rissanen 2000: passim; Stenroos 2004: 275–81). Thengs (2013: 339–340) has recently suggested that this difference may have to do, at least in part, with different kinds of text production. The copying of literary texts results almost inescapably in a perpetuation of earlier tradition, even though individual scribes may be highly innovative. Business writing, on the other hand, is generally composed rather than copied out; its formulaic nature means in itself that a trained writer would need no models. This makes it potentially much more innovative when it comes to spelling and morphology. There are some indications of such a difference between copying and composition in the writings of Walter Pollard. The same distinctive dialectal and orthographic features are found in the different texts, but there are some suggestive differences of proportion. Pollard at all stages shows a tendency to use the spelling where we might expect or , suggesting a merger of the corresponding phonemes in his spoken dialect. However, these spellings are much more common in the letter drafts than in the grammatical treatises, where it may be assumed that the usage of the exemplar may have constrained him. They are least common in the last phase, suggesting that the more idiosyncratic spellings may have been worn off over decades of a mobile life, with the need to be understood by a broad range of readers. The English writing habits learnt at school would reflect the regional variability of written English. School books were copied by hand and usually based on the work of the local schoolmasters; it was only when printing got under way that schoolbooks began to contribute to more homogeneous writing habits. The schoolmasters themselves might have moved in from elsewhere, bringing along their own dialectal forms, which would then be transmitted to the pupils as part of their bilingual training. The strongly dialectal character of the school texts is very clear in both the manuscripts considered here. The English text written by Walter Pollard at all

36 

 Merja Stenroos

stages is strongly southwestern in its dialectal forms. For Thomas Marchall we only have a single bilingual text, which, however, shows a similarly strong East Anglian colouring. This could reflect the background of a schoolmaster or the provenance of Marchall’s exemplar; on the other hand, it could simply show that these forms were on the move. Once English had become the medium of instruction in schools, the mobility of schoolmasters would contribute greatly to the transmission of dialectal written forms to different parts of the country. As these forms were adopted by pupils who themselves came to use writing to communicate across the country, this created a highly favourable setting for linguistic change and supralocalisation. Merchants were, without doubt, among the most mobile of these writers. Some travelled regularly overseas, but the great majority travelled within England, either by road or by water. During the later medieval period the distances of travel tended to grow as the importance of bigger towns grew; this was the case, above all, with London but also with the large provincial towns. Baker (1976: 174) describes the development between 1334 and 1600 as follows: there was (...) a concentration of inland trade into fewer centres; everywhere agricultural traffic, for example, tended to be drawn away from the smaller markets, ports, and fairs, and to be centred in the larger provincial towns such as Canterbury and Reading. There were far fewer market towns and villages in 1600 than in 1334, probably less than a third (...) Moreover, the larger market towns began to specialise in particular types of products such as grain, cheese and butter, poultry, horses, sheep and leather products.

This meant that merchants would travel longer distances than before. Both their spoken and written linguistic habits would travel with them and their written correspondence would follow the same routes. It would be of great interest to enquire to what extent linguistic innovations might follow the geographical patterns of the trade routes. Bergs (2006) has attempted this kind of enquiry by matching data from the Linguistic Atlas of English (McIntosh, Samuels and Benskin 1986) to medieval trade routes. However, the geographical framework of the Atlas is not well suited for this kind of direct matching, as its localizations are based on dialect: the texts are fitted in relation to each other on linguistic grounds, and the localizations do not necessarily reflect the geographical provenance of the texts, which is usually unknown (McIntosh, Samuels and Benskin 1986, I: 9–12; Benskin 1991: 25; see also Stenroos and Thengs 2012). The Middle English Local Documents corpus (MELD) that is being compiled at Stavanger will be better suited for this kind of study, as its localizations are based on non-linguistic criteria, such as localizing clauses or the names of people and places. The West Midland part of the corpus was compiled and studied by



Like the coins when currencies are combined 

 37

Thengs (2013) in his PhD thesis; there, he shows that several innovative spellings in this period, such as land ‘land’ and theym ‘them’ (replacing lond and hom/hem in the West Midlands) appear first in urban centres (Thengs 2013: 334–337). It was pointed out above that the language of Walter Pollard includes elements such as the third-person plural pronoun forms thay and them, which at this point of time were still largely northern but seem to have been spreading especially around the urban centres of the southwest. Of these forms, of course, them ends up in Standard English, while thay does not; it is important to bear in mind that the adoption of innovative forms in this period does not necessarily go in the direction of a later standard. At the same time, the exchange of regional forms by writers communicating across the country, as merchants did, makes good sense of the regionally mixed character of what eventually crystallized as “Standard English”. This kind of development was described by Samuels (1981[1988: 92]) in a paper on Late and Post-Medieval spelling: Spellings which have hitherto been members of regional systems become like the coins when two currencies are combined; they have the same functional value as before, but they pass from writer to writer, or from writer to printer and back again, and their regional significance is lost.

Samuels was not referring specifically to merchants and trade routes; however, the comparison is particularly apt here. Merchants moved around the country over long distances, coming together in the major urban centres; they composed rather than copied, and they carried with them the regional English and multilingual writing habits developed at school. They were not the only ones, of course: the clerks of manorial lords did to some extent the same, and it seems reasonable that the texts produced by these generally neglected groups of writers should be of central interest when we study innovation and supralocalisation in late medieval written English.

References Baker, Alan R. H. 1976. Changes in the later Middle Ages. In Henry C. Darby (ed.), A New Historical Geography of England before 1600, 186–247. Cambridge: Cambridge University Press. Baugh, Albert C. & Thomas Cable. 2002. A history of the English language. Fifth edn. Upper Saddle River, NJ: Prentice Hall. Benskin, Michael. 1991. The fit-technique explained. In Felicity Riddy (ed.), Regionalism in late medieval manuscripts and texts, 9–26. Cambridge: D.S. Brewer. Benskin, Michael. 1992. Some new perspectives on the origins of standard written English. In J.A. van Leuvensteijn & J.B. Berns (eds.), Dialect and standard language in the English,

38 

 Merja Stenroos

Dutch, German and Norwegian language areas, 71–105. Amsterdam: Royal Netherlands Academy of Arts and Sciences, North-Holland. Bergs, Alexander. 2005. Social networks and historical sociolinguistics. Studies in morphosyntactic variation in the Paston letters (1421–1503). Berlin: Mouton de Gruyter. Bergs, Alexander. 2006. Spreading the word: patterns of diffusion in historical dialectology. In Markku Filppula, Juhani Klemola Marjatta Palander & Esa Penttilä (eds.), Topics in dialect variation, 5–30. Joensuu: University Press. Fennell, Barbara A. 2001. A history of English: a sociolinguistic approach. Oxford: Blackwell. Graff, Harvey J. 1987. The legacies of literacy: continuities and contradictions in Western culture and society. Bloomington, IN: Indiana University Press. Hanawalt, Barbara. 1995. Growing up in medieval London: the experience of childhood in history. Oxford: Oxford University Press. Hanham, Alison. 1985. The Celys and their world. An English merchant family of the fifteenth century. Cambridge: Cambridge University Press. Hanham, Alison (ed.) 1975. The Cely letters: 1472–88. London: Oxford University Press. McIntosh, Angus, Michael Samuels & Michael Benskin. 1986. A linguistic atlas of late mediaeval English. 4 vols. Aberdeen: Aberdeen University Press. Milroy, James. 1992. Linguistic variation and change. Oxford: Blackwell. Milroy, James & Lesley Milroy. 1985. Linguistic change, social network and speaker innovation. Journal of Linguistics 21. 339–84. Moran, Jo Ann Hoeppner. 1985. The growth of English schooling 1340–1548: Learning, literacy and laicization in pre-reformation York diocese. Princeton, New Jersey: Princeton University Press. Nevalainen, Terttu. 1996. Social stratification. In Terttu Nevalainen & Helena Raumolin-Brunberg (eds.), Sociolinguistics and language history: Studies based on the corpus of early English correspondence, 57–76. Amsterdam: Rodopi. Orme, Nicholas. 2006. Medieval schools from Roman Britain to Renaissance England. New Haven/London: Yale University Press. Parkes, Malcolm. 2008. Their hands before our eyes: A closer look at scribes. Aldershot: Ashgate. Parkes, Malcolm. 1973. The literacy of the laity. In D. Daiches & A. Thorlby (eds.), The mediaeval world, 555–577. London: Aldus Books. Raumolin-Brunberg, Helena & Terttu Nevalainen. 1997. Like father (un)like son: A sociolinguistic approach to the language of the Cely family. In J. Fisiak (ed.), Studies in Middle English linguistics, 489–511. Berlin: Mouton de Gruyter. Rissanen, Matti. 2000. Standardisation and the language of early statutes. In Laura Wright (ed.), The development of Standard English, 1300–1800, 117–130. Cambridge: Cambridge University Press. Rutkowska, Hanna. 2003. Graphemics and morphosyntax in the Cely letters (1472–88). Frankfurt am Main: Peter Lang. Samuels, Michael L. 1981. Spelling and dialect in the late and post-Middle English periods. In Michael Benskin & Michael L. Samuels (eds.), So Meny People Longages and Tonges: Philological Essays in Scots and Mediaeval English Presented to Angus McIntosh, 43–54. Edinburgh: Middle English Dialect Project. Reprinted in Jeremy J. Smith (ed.) 1988. The English of Chaucer and his contemporaries, 86–95. Aberdeen: Aberdeen University Press.



Like the coins when currencies are combined 

 39

Stenroos, Merja. 2004. Regional dialects and spelling conventions in Late Middle English: searches for (th) in the LALME data. In Marina Dossena & Roger Lass (eds.), Methods and data in English historical dialectology, 257–285. Bern: Peter Lang. Stenroos, Merja. 2016. Regional language and culture: the geography of Middle English linguistic variation. In Tim Machan (ed.), Imaging Medieval English Language Structures and Theories, 500–1500, 100–125. Cambridge: Cambridge University Press. Stenroos, Merja & Kjetil V. Thengs. 2012. Two Staffordshires: Real and linguistic space in the study of Late Middle English dialects. In Jukka Tyrkkö, Matti Kilpiö, Terttu Nevalainen & Matti Rissanen (eds.), Outposts of historical corpus linguistics: from the Helsinki Corpus to a proliferation of resources.(Studies in Variation, Contacts and Change in English, 10). Helsinki: VARIENG. Thengs, Kjetil V. 2013. English medieval documents of the North-West Midlands: A study in the language of a real space text corpus. Stavanger, Norway: University of Stavanger PhD thesis. Thomson, David. 1979. A descriptive catalogue of Middle English grammatical texts. New York/ London: Garland. Thomson, David. 1984. An edition of the Middle English grammatical texts. New York/London: Garland. Thrupp, Sylvia. 1948. The merchant class of medieval London 1300–1500. Chicago: University of Chicago Press. Wright, Laura. 2000. Bills, accounts, inventories: Everyday trilingual activities in the business world of later medieval England. In David Trotter (ed.), Multilingualism in later medieval Britain, 149–156. Cambridge: D.S. Brewer. Wright, Laura. 2005. Medieval mixed-language business texts and the rise of Standard English. In Janne Skaffari, Matti Peikola, Ruth Carroll, Risto Hiltunen & Brita Wårvik (eds.), Opening windows on texts and discourses of the past, 381–399. Amsterdam: John Benjamins. Wright, Laura. 2012. On variation and change in London medieval mixed-language business documents. In Merja Stenroos, Martti Mäkinen and Inge Særheim (eds.), Language change and development around the North Sea, 99–115. Amsterdam: John Benjamins.

José Miguel Alcolado Carnicero

3 Bridges of innovation and change: the English language around the networks of the Mercery of London Abstract: The fourteenth and fifteenth centuries were a period of profound change in England. Two domains particularly affected by innovative trends were the economy, with an emerging proto-capitalist economic system, and language, with a deepening development of vernacular English. This paper aims to explore connections between both of these spheres; more specifically, how the members of the premier livery company of the City of London, the Mercers, promoted the adoption of English as the primary language of record in their manuscripts. In order to do so, I apply a sociolinguistic model that attempts to allow analysis of the origin and propagation of innovations and changes through interpersonal connections: social network theory. The results show firstly that some of the individual mercers introducing the earliest shifts to written English belong to the same close-knit ego-centric social network (i. e. informants are related by master-­ apprentice ties), and secondly that the written variety of English is spread to the different records through members with loose-knit socio-centric social networks (i. e. informants have many connections outside their guild) who are also acting as bridges. Consequently, this indicates that social networks could provide a valid answer to explain the process of the language shift to English in the publications of the Mercers’ Company and even appear promising as a model for various companies with a similar set of networks (e. g. other contemporary London guilds).

1 Towards a better understanding of the shift to written English? Mercery, the term covering any business connected with the manufacturing and trading of dress and decoration goods, was one of the most lucrative sources of income for English people across class and society during the Middle Ages. Before Note: I am indebted to the editors for including me as one of the contributors to this volume, despite not being part of the list of speakers at the original conference. José Miguel Alcolado Carnicero, University of Santo Tomas, Philippines DOI 10.1515/9781501503542-003



Bridges of innovation and change 

 41

1304, Mercer was only recorded in the manuscripts as a byname for those carrying small wares on their backs to markets or those manufacturing and selling goods in their workshops in a personal capacity (Sutton 1998: 215, 2005: 17); subsequently, the term was used referring to any group of mercers acting as a permanent community (Veale 1991: 259). After that same date, the now affluent overseas traders from the Mercers’ Company also slowly became quite prominent actors in the writings of the metropolis in both its literary (e. g. “Le dit du mercier” [Sutton 2005: 15]) and administrative (e. g. Carta mercatoria [Sutton 2005: 31; 115–117]) dimensions. However, not only was the Mercers’ Company’s prominence mirrored in accounts kept by others, but also in its own increasing textual production from the fourteenth century onwards. In this respect, it is not unreasonable to assume that the administrative load issuing from the office of the Mercers’ Company during the late medieval period was directly affected by one of the most, if not the most, pervasive language changes ever taking place on English soil: the gradual replacement of Latin and French by English as the official language of record. This article endeavours to investigate the extent to which the change from Latin and French to English that is underway in the writings of the Mercers’ Company may be viewed through the lens of social network theory in order to model the way in which these small-scale changes could be explained. More specifically, the social network theory is expected to provide a framework for the attempt here to explain whether the interpersonal professional relations and other personal circumstances of possible language innovators and bridges may account for the actual dissemination of written English across the publications of this mercantile community.

2 Agency of language change within J. Milroy and L. Milroy’s social network approach Social networks had been treated in sociolinguistic literature on the same level as other classical independent variables such as class, age, or sex (e. g. Poplack 1979) until Blom and Gumperz (1972), Labov (1972), Gumperz ([1976] 1982), and Gal (1979) decided to conduct research focusing mainly on their role in language variation and change. But particularly influential to this field of research were J. Milroy’s and L. Milroy’s later publications as a result of their fieldwork around Belfast, Northern Ireland (e. g. L. Milroy [1980] 1987; J. Milroy and L. Milroy 1985; L. Milroy and J. Milroy 1992). The most distinctive conclusions drawn by J. Milroy and L. Milroy are those related to the identification of agents and factors responsible for language

42 

 José Miguel Alcolado Carnicero

maintenance and change within a speech community. J. Milroy and L. Milroy’s primary objective was to determine whether social ties were able to account completely for the strong presence of Belfast English vernacular features in the language of users from several key locations, despite the overpowering influence of Standard English. After months of fieldwork, the conclusion was reached that, regardless of their sex, age, and social class, informants who were strongly attached to their communities by belonging to close-knit networks tended to use non-standard variants more often than informants who were affected by mobility outside their communities and, consequently, were members of more loose-knit networks. Thus, a direct correlation between social network integration and language behaviour was finally proven: “[T]he closer an individual’s network ties are with his local community, the closer his language approximates to localised vernacular norms” (L. Milroy 1987: 179). The Milroys’ confirmed findings were summarised in the above hypothesis that, nevertheless, showed a certain bias towards only one side of the equation: the close relationship between network attachment and language conservatism. However, in the same way that network integration was shown to enforce language maintenance, network disruption should have been promoting language change. That almost exclusive orientation towards language stability was actually acknowledged by L. Milroy (1987: 190) and modified in J. Milroy and L. Milroy (1985). J. Milroy and L. Milroy (1985) did recognise the relevance of unstable networks within a speech community as promoters of language innovation and change by bringing to the fore research so far unnoticed in the sociolinguistic realm but which became quite influential afterwards: E. Rogers’s ([1962] 2010) process of diffusion of innovations and Granovetter’s (1973: 1983) analysis of weak ties. On the one hand, E. Rogers (2010) discovered that innovations were normally brought into and communicated through certain channels and, at a varying speed, went through a five-stage process: (i) knowledge, (ii) formation, (iii) decision, (iv) implementation, and (v) confirmation. Among the five, “decision” was supposed to represent the most critical phase, since the future of innovations would depend on the degree of innovation in the social system’s members (E. Rogers 2010: 37). Based on this rate of adoption, five adopter profiles were identified, going from those more prone to more reluctant to adopt the new features: (i) innovators, (ii) early adopters, (iii) early majority, (iv) late majority, and (v) laggards. Due to their venturesome personality, innovators tended to introduce innovations, whereas more influential early adopters motivated the wider adoption or rejection of said innovations. In the case of adoption, innovations would reach the rest of the members, and changes would be established; in the case of rejection, innovations would die out sooner or later.



Bridges of innovation and change 

 43

On the other hand, Granovetter (1973: 1983) did not focus on the whole process of adoption but on those channels that were at the origin of the transmission of an innovation. Granovetter attached greater importance to the individual’s interpersonal ties with low content, frequency, and intimacy since, contrary to what had been believed, these weak ties were fundamental in the transmission of new information. According to Granovetter (1973: 1364), only repetitive data could flow through (strong) ties between people who knew each other well or who would spend most time together; in contrast, weak ties could even function as “bridges”, channelling new or required information between different networks (see Lee 1969; Granovetter 1974). Combining both studies, J. Milroy and L. Milroy (1985) recognised two of the findings as highly valid for their social network framework. Firstly, language innovations were introduced by users who were not particularly attached to the language standards of any specific network but, rather, would show independent behaviour flouting vernacular norms of a particular network; further, they acted as bridges between different networks potentially transmitting linguistic items and practices from one to the other and vice versa. And secondly, innovations would be more likely to consolidate themselves as language changes in the speech communities, once early adopters adopted them or, at least, formed a favourable attitude to them. J. Milroy (1992) fully applied this model to earlier speech communities and, more specifically, to explain efficiently the contrast between (i) Icelandic, a very conservative language that had remained stable for centuries because its speakers had managed to maintain close-knit networks due to scarce external contact, and (ii) English, a very innovative language whose older varieties were vastly different from modern ones, several conquests having led to the disruption of social networks and an abundance of loose-knit networks connected through weak ties. The notions of mobility and instability at a more societal level were particularly highlighted as possible factors affecting the rate of adoption at the individual level. However, J. Milroy seems to have analysed the question from a more qualitative standpoint, insofar as he never mentions the explicit use of any of the two analytic tools that have made the Milroys’ series of studies be considered the first applications of networks as a concept subject to quantification, rather than as a simple metaphorical device, in the description of social relationships (L. Milroy 1987: 45). These two tools are (i) the different mathematical formulae – that proved defective later (L. Milroy 1987: 50, 51, 141) – to calculate the weakness and strength of the ties between the informant and the rest of network members (ego-centric perspective), and (ii), especially, the network strength scale proposed to measure the ties of the network types the informant is usually involved in (socio-centric perspective):

44 

 José Miguel Alcolado Carnicero

1. Membership of a high-density, territorially based cluster. 2. Having substantial ties of kinship in the neighbourhood. (More than one household, in addition to one’s own nuclear family.) 3. Working at the same place as at least two others from the same area. 4. The same place of work as at least two others of the same sex from the area. 5. Voluntary association with workmates in leisure hours. This applies in practice only when conditions three and four are satisfied. (L. Milroy 1987: 141–142) On this scale, the informant would be given one point as long as he/she fulfilled each criterion. The higher (closer to +5) the final score, the more close-knit his/ her networks were said to be; and lower final scores (closer to 0), were considered to indicate that he/she had more loose-knit networks. Bergs (2000) later argued that only societal-level applications of the social network theory to early communities worked effectively and without any kind of modification. By contrast, analyses of networks at the individual level had to be “developed and evaluated for each particular situation in which they [were] to be applied” (Bergs 2000: 241), since they tended to demand concrete quantification. Principally, modifications of the nature of the criteria in the original network strength scale and of the calculation in the score have been propounded in the literature to adjust the scales to the specific period under study (see Bax 2000: 282; Bergs 2005: 59; Section 4 for this research).

3 The change to English in the archives of the Mercery Linguistically speaking, the Norman Conquest of England in 1066 meant the adoption of French, the native language of the new ruling class, and the reinforcement of Latin, the traditional lingua franca during the Middle Ages, as official languages of record to the detriment of English, the native language of the conquered but still majority population. This situation of triglossia (Leith 1983: 27) is already perfectly visible in the business records kept during the earlier period (e. g. eleventh-century Billingsgate tolls of London, cf. Wright 1994: 463). Both the coexistence and the mixture of the three different languages within one single text are systematic in manuscripts from post-Conquest England and even extend to as late as the mid-sixteenth century with Wynkyn de Worde’s inventories for his printing shop (Wright 1996: 11):



Bridges of innovation and change 

 45

(1) Memorandum that I, John Lamberd, have resseyved of Richard Nedam le vi jour de Novembre l’ane xxxi – xl s. xl d. [Memorandum that I, John Lambard, have received from Richard Nedam on 6 November 31 [Henry VI] [1452] – 40s. 40d.] (Mercers’ Wardens’ Accounts 1452: 174v in Jefferson 2009: 714–715)1 Even though code-switching falls outside the scope of this research, this accounting entry (and others later) perfectly illustrates the intricate and enduring level of language contact also in the Mercers’ Company’s records. In this short excerpt from the Mercers’ Wardens’ Accounts in 1452, the three different languages are still present in less than twenty words, with a borrowed Latin-origin formulaic expression – equivalent to English-origin “be it known” – at the beginning, a longer part in English as the probable base language, and a final date in French. What gradually changed throughout the five centuries during which mixed-language business manuscripts appear to have existed was the extension of the extracts mixed with English. In the earlier centuries the presence of English was limited to single lexemes never larger than words (e. g. nouns, verbal roots, or adjectives), but as time went by, more and more function words, clauses, and sentences in English began to be incorporated into the manuscripts; in the end, the final stage of this evolution was the engrossing of business records completely in English (Wright 1998: 101–109). Cooper (1984) classifies the archives of the City of London livery companies and other related organisations into six different categories, according to the internal matters being recorded: (i) trade and craft regulation, (ii) apprenticeship and membership, (iii) estates and charities, (iv) constitution and operation, (v) governing bodies, and (vi) finances. Following this division of administrative records, the Mercers’ Company managed to author as many as four types of these manuscripts – amounting to around 150,000 words – during the fourteenth and fifteenth centuries and in a language mixture that deserves specific attention here. The four types of administrative documents available in the Mercers’ Company are the Records of estates and charities, the Records of constitution and operation, the Records of governing bodies, and the Records of finances.2 1 Since I have not worked directly from the original manuscript, I have followed Jefferson’s (2009) editorial method for expanding abbreviations and suspension signs through italicisation. 2 Records relating to trade and craft regulation and Records relating to apprenticeship and membership are the two types of records whose language analysis cannot be carried out here for different reasons. As for the former, the Mercers’ Company never compiled such records, insofar as this livery company never had a craft to supervise. As for the latter, the earliest separate register providing that kind of information began to be composed around 1527–1528 by John Coke, then clerk to the companies of Merchant Adventurers and Mercers (Hardy and Page 1915).

46 

 José Miguel Alcolado Carnicero

In the Records relating to estates and charities, the Renter Wardens’ Accounts constitute the group of manuscripts in which the Mercers’ Company recorded the management of its properties. The compilation began immediately after the inheritance by the company of Richard Whittington’s estate in 1442–1443 (Sutton 2005: 175). Due to the need for a more organised handling of property accounts,3 a book for fair copies came into use eight years later (1449–1450). Except for a few scattered rentals kept in Latin at the beginning, the Renter Wardens’ Accounts were completely written in English. Within its Records of constitution and operation, the ordinances and oaths of the Mercers’ Company can be found.4 The Mercers’ Company kept different groups of rules and obligations for its current, and prospective, members over several fiscal years (1347–1348, 1376–1377, 1404–1405, 1407–1408, 1409–1410, 1410–1411, and 1416–1417) before a ledger, the Book of Ordinances, was explicitly created for such a purpose around 1436–1437 (Parker 1980). Of the earlier seven series, originally recorded in the Mercers’ Wardens’ Accounts and most of them restated later in the main ledger, five are kept in French and, quite interestingly, two in English, namely, 1407–1408 and 1416–1417. Except for the recopied French regulations, the Book of Ordinances was kept entirely in English, with some oaths appearing in its preliminary pages. It is nevertheless not possible to say when exactly the ledger, and consequently English as its main language, was in actual use, insofar as both these oaths and the first nine ordinances in English are unfortunately undated;5 the first dated ordinance is from 1456–1457. In the Records of its governing bodies, the Mercers’ Company’s decisions concerning legislative and executive powers were compiled in a book called Acts of Court that, in spite of being considered inaccurate from a chronological perspective (Sutton and Hammond 1978), had been in use since approximately 1453–1454. All of the records, many of them dealing with the governance of the Merchant Adventurers’ Company, were also composed in English. Finally, the Records relating to finances of the Mercers’ Company are represented by the Mercers’ Wardens’ Accounts. This main book of the livery company was kept every accounting year by the four different wardens in office. In this particular case, the Mercers’ proper financial accounts extend from 1390–1391 to 1463–1464. These types of records in the Mercers’ Company tend to reflect, perhaps better than the rest, the above-mentioned complex language reality in 3 John Abbot’s and William Eastfield’s estates were bequeathed soon after Richard Whittington’s, in 1444–1445 and 1446–1447 respectively. 4 Even though Cooper (1984) also includes royal charters, I have decided to leave them out, as they were not kept by any member of the Mercers’ Company but by the King’s secretaries. 5 Parker (1980: 14) proposes 1448–1449 as a probable date of composition.



Bridges of innovation and change 

 47

late medieval London, since they were discontinuously written in the three languages used on English soil (Latin, French, and English). In fact, French is the main language of records from 1390–1391 to 1457–1458, except for a brief shift to Latin in 1391–1392 and another shift to English from 1449–1450 to 1451–1452. From 1458–1459 onwards, records in English finally prevail. The entire above-mentioned records share one common feature: They were internal documents kept for the better functioning of the livery company, i. e. they were produced and could only be accessed by members of the Mercers’ Company. But the Mercers’ Company also composed external documents, i. e. texts intended for, quite often powerful, personalities outside the auspices of the guild. Beyond any doubt, the most popular within this type of text has always been “The Folk of the Mercery” (R. Chambers and Daunt 1931: 33–37), a petition presented to Parliament in the second half of the 1380s denouncing the intimidating lobbying by Nicholas Brembre’s entourage against John of Northampton and his supporters, most of them mercers. The popularity of this document, at least among historical linguists, is due to the fact that the Mercers’ was the first petition to be entered completely in English into the Rolls of Parliament (Fisher 1996: 45).

4 In search of innovators and bridges of the English language J. Milroy and L. Milroy’s social network framework seems to provide a convincing sociolinguistic explanation to the riddle of the actual process of propagation and spread of eventual language changes in speech communities, either contemporary or ancient, and, as such, has an important position in the literature (e. g. Chapter 3 in Coulmas [1997], Chapter 22 in J. Chambers, Trudgill, and Schilling-Estes [2003], or Chapter 18 in Hernández Campoy and Conde Silvestre [2012]). In a similar vein, J. Milroy and L. Milroy’s characterisation of the innovator as that prototypical carefree language user who adopts a language innovation elsewhere and brings it later as new material into the social networks he/she interacts with, acting so as a bridge as well has gained a foothold in academy, despite some doubts casted by Labov (1980: 261; 2001: 364). The main social network analysis in this article puts a special emphasis on potential innovators as well as bridges and tries to check whether similar processes of propagation and spread may have happened at the written dimension. In fact, speaking and writing, albeit two essential and interconnected skills of language use, have always been considered different processes (Crystal 1995). In

48 

 José Miguel Alcolado Carnicero

the context of historical linguistics, since the study of past states of a language has to rely exclusively on written sources, this distinction acquires more relevance here and even has a detrimental effect on the efficacy of analysis, namely, the impossibility of gaining access to entire dimensions of linguistic information and to data from all members of a community (J. Milroy 1992: 45). Historical linguists have mostly mentioned the constraints brought to their discipline by the issue of working with written data (e. g. Labov 1994: 11) and have rarely focused on written texts’ strong potential. That initial attitude has recently begun to change; for example, Nevalainen and Raumolin-Brunberg (2003: 27) mention one main characteristic in which historical research has an advantage over modern research: depth in time, i. e. diachronic study can be undertaken from original emergence to final outcome. In the specific case of London livery companies, the fact that this type of corporation decided to record, by hand, as many activities and movements as possible on a rather strict basis provides important features needed by most diachronic analyses of the impact of social networks on language innovation and change. Firstly, it is feasible to assign a rather precise date to textual innovations (e. g. the term for “box” in English): (2) Item, resçu de Thomas Roos, mercer, pur le encrese de la box – iii li. vi s. viii d. [Item, received from Thomas Roos, mercer, as an increment to the communal funds – £3 6s. 8d.] (Mercers’ Wardens’ Accounts 1422–1423: 82v in Jefferson 2009: 342–343) For example, it can be seen above in which accounting year the lexeme referring to the “strongbox for money or valuables, or its contents; specif., a fund (as of a guild or other corporation)” (Kurath and Kuhn [1958] 1971, B.5: 1102) was written in English (box) for the first time: 1422–1423. Secondly, it is also possible to trace the evolution of textual innovations until they seem to become (or do not become) proper written changes. For instance, the French variant of the previous lexeme (buiste) can still be found in accounts kept 12 years later (1434–1435), which could mean that the change to written English for this particular lexical variable had not yet been completely accomplished and/or that the transfer of active innovations from the oral dimension to the written data was inconsistent: (3) Item, Johan Wood doit al buiste – v s. [Item, John Wood owes to the communal funds – 5s.] (Mercers’ Wardens’ Accounts 1434–1435: 118r in Jefferson 2009: 474–475)



Bridges of innovation and change 

 49

In fact, this might also be the reason why the word box could not be yet considered a written loanword into Anglo-French (accomplished change) but still a switch to English (fluctuating innovation). The textual occurrence of the French variant buiste alongside the English variant box referring to the variable “strongbox for money of a guild” may actually be pointing to the (possible) process of borrowing of the latter into the word-stock of the French of England as not being completely finished by 1434–1435; it could be said to be completed once the French variant does not appear anymore in written texts. Having said that, an important circumstance affecting these first two features should be taken into consideration: the different rate at which the oral and the written spheres evolve. Raumolin-Brunberg (1996: 17) claims that “an inevitable temporal gap between the introduction of new forms in speech and their first recordings in written texts” has to be assumed in historical sociolinguistic research, and Rothwell (1991: 183) confirms that, during post-Conquest England, hundreds of words would have been already in daily spoken use for generations without necessarily being committed to parchment. Consequently, any possible evidence from this study – and any other, mainly focused on the written use during the medieval period – should not be employed to draw conclusions on the oral use (Kristol 2000: 38). Thirdly, the informants’ language use could be investigated over decades, recalling so the depth-in-time advantage mentioned by Nevalainen and Raumolin-Brunberg (2003: 27). The case of John Olney is paradigmatic, since he served the office of wardenship six times in five different decades: 1419–1420, 1426–1427, 1432–1433, 1437–1438, 1442–1443, and 1454–1455 (Sutton 2005: 556–557). It is to be clarified here that the mercer-wardens have been assumed to be the sole authors of the Wardens’ Accounts; it is true that the Mercers’ Company did employ other mercers as scribes and even hired external scriveners to make fair copies from scratch manuscript notes (Sutton 2009), but the conception here is that these scribes limited themselves to reproducing original writings verbatim without modifying any extract, at least, as far as language mixes were concerned, insofar as cases of scriveners copying the accounts long after the fiscal year in question have been reported (Jefferson 2009: 14). Finally, both the true identity of informants and their links with other guildsmen could be investigated. The possibility of undertaking a piece of social network research directly depends on the availability of these data. Ego-centric networks have been reconstructed in this article by resorting to the three types of ties attested in the Mercers’ Wardens’ Accounts: (i) master-apprentice tie, i. e. every time a new apprentice was admitted into the livery company, his master’s name was written down beside, (ii) successor tie, i. e. at the end of each year, the four wardens leaving office used to elect their successors for the coming term,

50 

 José Miguel Alcolado Carnicero

and (iii) fellow-warden tie, i. e. mercers shared office with three more wardens.6 Whereas the first type of link has been considered to be always strong, because apprenticeship extended as many as seven years, the second and the third types of links have been considered weak, unless attested more than once, since these kinds of contact seem to have been much more sporadic and less dense and multiplex. As for socio-centric networks, their membership of the Mercers’ Company of London meant that these informants were quite influential people in London and, fortunately, their presence in ancient records of any kind was recurrent. Consequently, it is fairly easy to compile the data needed to fill in the criteria in the network strength scale built ad hoc for the study of socio-centric networks in the Mercers’ Company of London. For its part, when no mention is made of a mercer-warden doing something, it has been assumed that he never did it and, as such, scores have been calculated in the socio-centric network strength scale (see Table 1). Obviously, this assumption runs counter to what Wang (1969: 21) calls “the platypus metaphor” in historical linguistics: “[W]e cannot prove that a platypus does not lay eggs with photographs showing a platypus not laying eggs”; in other words, it should never be said that somebody did not do something in the past only because no records of him/her doing it have survived. Notwithstanding that and as a result of dealing with very poor data, the assumption of absence of proof meaning proof of absence had to be inevitably made, because there was not any other possibility. Table 1: Socio-centric network strength scale for the wardens of the Mercers’ Company Criteria

Close-knit (–1)

Balanced (0)

Loose-knit (+1)

Marital status Place of origin and/or of living Business destinations Offices Contacts

Married Village and/or Mercers Local Mercers and/or None Low prestige

Widower Village and/or London

Single London and London National and/or Overseas Local and/or National High prestige

This scale is partly based on Bergs’s (2005: 59) proposal for late medieval England, particularly on the final network score, insofar as, unlike L. Milroy (1987: 141–142), the higher the score (as high as +5), the more loose-knit the networks would have 6 Only professional-related ego-centric ties established by each informant have been traced. Their links outside the Mercers’ Company have not been taken into consideration for the egocentric perspective.



Bridges of innovation and change 

 51

been and vice versa, i. e. the lower the score (as low as –5), the more close-knit the networks. The provisional results obtained from each criterion have to be summed, and this final result (ranging from –5 to +5) would be the socio-centric network strength score of the mercer-informant under study. Negative scores are then incorporated into this scale and indicate close-knitness of the socio-centric networks in question; by contrast, positive scores indicate socio-centric loose-knitness. Discussing criteria a bit more in detail, previous research has shown that whereas remaining single and living in big cities, like London at that time, implied an involvement in much more open networks full of weaker ties (Nevalainen and Raumolin-Brunberg 2003). Being married an entire life and living in smaller more isolated communities more frequently equalised closer networks and stronger ties. Consequently, these aspects are also included in the “marital status” and “place of origin and/or of living” criteria of this scale, but eventual combinations of both ends (i. e. being married but widow later, and living initially in a village but later in the big city) are further introduced into the scale and represented by a balanced (0) score. As for the other three criteria, they are focusing on suspected circumstance; for example, it is deemed almost impossible to find, in the case of “business destinations”, a mercer of London travelling to the Low Countries who had not been engaged in local business before, in the case of “offices”, a mercer being appointed Member of Parliament who had not been warden of the Company before and, in the case of “contacts”, a mercer frequenting aristocratic circles who had not been previously (or even simultaneously) in contact with less influential people from much lower classes. As a result, had the same type of score calculation been applied to the three last criteria, finding informants combining the criteria of both ends and scoring 0 in the network strength scale was going to occur quite straightforwardly all the time but, by contrast, identifying mercer-wardens with network scores other than 0 (especially, positive) was going to be very difficult. In order to avoid this, the important difference in the calculation of these three criteria is that practices leading to the loosening of ties (and expressed by +1) are supposed to have more relevant implications than tie-closing practices (represented by –1) for the informants’ socio-centric networks. Regardless of his previous or simultaneous involvement in close-knit practices, as soon as the informant comes into contact with loose-knit business destinations, offices, and contacts, he is assigned +1 for each one of these three criteria.

4.1 John Whatley and William Cantelowe The earliest group of internal records in which a shift to English occurs are those of constitution and operation; more specifically, one series of ordinances kept in

52 

 José Miguel Alcolado Carnicero

1407–1408. During this fiscal year, the mercer-wardens, among whom one of the potential innovators at least could be found, are the following: (4) 1407 John Woodcock, John Middleton, John Whatley, John Eton. (Sutton 2005: 556) Following the Milroys’ postulates on agency of language change, the possible innovator should belong to the group of mercers administering the guild in 1407–1408, since the earliest ordinances in English were introduced during their wardenship. In addition, his social networks should not be extremely closed, but weak ties should be prevalent. Analysing more in depth the socio-centric networks of the mercer-wardens of 1407–1408, it can be seen in Table 2 how one of them (John Whatley) scores extremely high (+5) in the network strength scale because, unlike the rest, he fulfils positively the five criteria. Table 2: John Woodcock’s, John Middleton’s, John Whatley’s, and John Eton’s socio-centric network strength scale scores Warden Woodcock Middleton Whatley Eton

Marital status

Place of origin and/or of living

Business ­destinations

Offices

Contacts

Network score

–1 0 +1 –1

0 0 +1 0

+1 +1 +1 –1

+1 –1 +1 –1

+1 –1 +1 –1

+2 –1 +5 –4

John Middleton’s and John Eton’s scores are negative, meaning in both cases that their socio-centric networks must have been closer and, consequently, more conservative, linguistically speaking, than Woodcock’s and, especially, Whatley’s. This is the main reason why Middleton and Eton are discarded here as possible innovators. John Middleton was married but became a widower later, since it is clearly recorded that, after his wife’s death, he preferred to wear black or grey clothes rather than the livery (Mercers’ Wardens’ Accounts 1410–1411: 53v in Jefferson 2009: 234–235); no mention is made of him remarrying any other woman later (0 for “marital status”). Although the surname Middleton could have been adopted by this mercer-warden to make reference to his place of origin, as sometimes did happen, no further references to any activities carried out or lands purchased in (or near) that place close to Manchester make it difficult to believe that Middleton actually hailed from there. However, the fact that several eminent mercers



Bridges of innovation and change 

 53

surnamed Middleton came from Hertfordshire (Burgon 1839, 1: 21), together with the fact that John Middleton is recorded taking part in some kind of land transactions in that same county (I. Rogers 2009), leads one to believe that he may also actually come from Hertfordshire (0 for “place of origin and/or of living”). Aside from in England, John Middleton seems to have been also involved overseas, since letters exist nominating his attorneys in Ireland for three years (I. Rogers 2009) (+1 for “business destinations”). Finally, Middleton is never recorded holding any relevant office in London or England (–1 for “offices”) or having any kind of contacts with very prestigious social networks (–1 for “contacts”). John Eton was married to a certain Mabel, who survived him (Barron and Sutton 1994: 50) (–1 for “marital status”). John Eton is recorded with a certain Richard Eton, probably a relative, in the settlement of an action against Isabel Heton, perhaps also related to them by blood and deforciant of one messuage, 80 acres of land, 24 acres of meadow, and 40 acres of pasture in Duns Tew and Aston, Oxfordshire. Similar surnames, probably revealing a family dispute on inherited lands, and the existence of a hamlet called Eaton also in Oxfordshire and not far from Aston could be pointing to John Eton hailing from this very county (I. Rogers 2010) (0 for “place of origin and/or of living”). Apart from his involvement in land transactions in his presumed county of origin, John Eton’s name does not appear in any other major business negotiations, either nationally or internationally; consequently, he seems to have focused more on his role as a mercer in London (–1 for “business destinations”). Nor does he appear to have held any office (–1 for “offices”) or to have got into contact with high prestige networks (–1 for “contacts”), since no mention is made of these in any kind of records. As for the two other mercer-wardens during 1407–1408, John Woodcock was married to Felicity, daughter of fellow-mercer Thomas Austyn, by 1390–1391 (–1 for “marital status”). Woodcock was a Yorkshireman by birth; more specifically, he was son of William Woodcock of Doncaster, who later migrated to London (0 for “place of origin and/or of living”). Even if no clear evidence of Woodcock’s involvement in overseas markets has survived, he was active in businesses all over England buying lands (e. g. Kent, Surrey, and Essex), and especially collecting customs (e. g. Ipswich, Gravesend, and Tilbury) (+1 for “business destinations”). But Woodcock held many more offices during his lifetime: from alderman of London wards (Coleman Street and Cripplegate) to Member of Parliament for a London constituency in 1404 or Lord Mayor of London during 1405–1406 (+1 for “offices”). Furthermore, Woodcock’s activities brought him many powerful connections from very early on; for instance, he had been directly supplying the wardrobe of the future Henry IV, at least from 1387 to 1398 (+1 for “contacts”) (Roskell, Clark, and Rawcliffe 1993, 4: 896–899).

54 

 José Miguel Alcolado Carnicero

For his part, Whatley seems to have remained single for his entire life; no mention whatsoever of any wife or child is made in any records (+1 for “marital status”). Contrary to common practice at that time, Whatley as a surname does not refer to any possible place of origin other than the City of London. Nor does Whatley appear to have acquired any land in the countryside; rather, he preferred to concentrate his property acquisitions around the City of London (e. g. one tenement in the parish of St Christopher and another in St Mary Colechurch’s parish) (+1 for “place of origin and/or of living”). Whatley prospered as a mercer engaged in national and especially in overseas trade; he is known to have rapidly expanded his trading network in the Low Countries, where he received a papal indult to improve the performance of his commercial activities (+1 for “business destinations”). He also held important offices both at local and national level; for example, he was tax collector of London in March 1404, warden of London Bridge as late as September 1418, and, more importantly, Member of Parliament for the London constituency in 1421 (+1 for “offices”). Finally, Whatley began to have contacts with the royal household quite early (by 1405) perhaps as a linen supplier, which lasted at least until 1420 (+1 for “contacts”) (Roskell, Clark, and Rawcliffe 1993, 4: 825–826). John Woodcock’s socio-centric network score is positive and, in theory, he has been fulfilling the criteria for acting as the innovator so far. Nevertheless, in terms of chances to be the eventual innovator, Woodcock’s fall short in comparison with Whatley’s due to the latter’s exceptionally higher degree of loose-knitness in his socio-centric networks. According to the modification of the Milroys’ model proposed by Bergs (2005) and followed by me, Whatley’s score of +5 means that the socio-centric networks in which he was involved were extremely looseknit and, consequently, very prone to language innovation and change. Furthermore, Whatley’s ego-centric network (see Figure 1) consisting of the wardens with whom he, anchor in bold, had attested professional contacts within the Mercers’ Company is also of foremost relevance for the present research for two main reasons.

Figure 1: John Whatley’s ego-centric social network



Bridges of innovation and change 

 55

Firstly, it is interesting to see the prevalence of weak ties over stronger contacts during his entire life as a mercer-warden. And at the beginning of his career during the early 1400s, some stronger links (Melreth’s and Middleton’s) would have been weak as well. Whatley served his office as third warden in 1407–1408 after being directly appointed by Walter Cotton at the end of the previous fiscal year (Mercers’ Wardens’ Accounts 1407: 45v in Jefferson 2009: 200–201). He became warden for the second time in his life, but this was the first time that he was chosen as first option at the wardens’ election ceremony; in 1400, Whatley had to replace William Sonningwell, who had died suddenly and who was John Middleton’s first option (Mercers’ Wardens’ Accounts 1400: 31r in Jefferson 2009: 148–149). And secondly, one of Whatley’s particular ties deserves special attention here. Whatley was strongly linked with William Cantelowe through a master-­ apprentice tie; Whatley was Cantelowe’s master. Thus, Cantelowe’s behaviour regarding language has to be highlighted as well. He was first warden when the initial shift to English in the Mercers’ Wardens’ Accounts took place in 1449–1450. The venturesome and innovative attitude towards the language of his master seems to have been transmitted to Cantelowe. Indeed, Cantelowe had a social profile quite similar to that of Whatley, with a wealth of weak ties in his ego-­ centric network (see Figure 2) and a positive socio-centric network score overall (see Table 3).

Figure 2: William Cantelowe’s ego-centric social network

Table 3: William Cantelowe’s socio-centric network strength scale Marital status –1

Place of origin and/or of living

Business destinations

Offices

Contacts

Network score

+1

+1

+1

+1

+3

56 

 José Miguel Alcolado Carnicero

William Cantelowe was married to Margaret Barry, and both had as many as seven children, 4 sons and 3 daughters; he might have been married once again to a certain Elizabeth, but apparently no more children were born from this new marriage (Sharpe 1894, 1: 328; Richardson and Everingham [2005] 2011, 1: 565) (–1 for “marital status”). Cantelowe seems to have been born in Kentish Town, but lived between Milk Street, St Mary Magdalen, and St Clements Danes, London; in addition, he may have owned some lands in Faversham, Kent, but nothing points to this county as his place of origin (Baildon 1904, 6: 44; Wedgwood and Holt 1936: 152) (+1 for “place of origin and/or of living”). He also became a prominent stapler around Calais for a time (+1 for “business destinations”). There is practically no office that Cantelowe did not hold during his lifetime; for example, he was alderman of Cripplegate Ward, London, from 1446–1461, and Member of Parliament in 1453–1454 and 1455–1456 (+1 for “offices”). Cantelowe had good relations with the monarchy; for instance, he is recorded conveying money over sea to bring Queen Margaret, wife of King Henry VI, to England (Wedgwood and Holt 1936: 152). These strong contacts may have been what earned him the honour of being knighted on 27 June 1461 (Maxwell-Lyte and Bland 1910: 286) (+1 for “contacts”). But aside from the presumed linguistic influence from his master, there is another important factor that could have exerted an influence on Cantelowe’s decision to adopt English as the main language of the Mercers’ Wardens’ Accounts. In late 1449, with Cantelowe still serving as first warden of the Mercers’ Company, a separate ledger for keeping the records relating to estates and charities was created: the Renter Wardens’ Accounts (1v in Jefferson 2009: 686–687). This book emerges out of the need to better organise the increasing load of paperwork the Mercers’ Company had been coping with since the inheritance of Richard Whittington’s estate in 1442. In fact, the record book already gathered the accounts for the rental years from 1442–1443 to 1448–1449 (1v–7r in Jefferson 2009: 578– 581, 596–599, 610–611, 622–623, 634–635, 646–649, 1018–1025). Linguistically speaking, these pre-1449 rental accounts are important because they were mostly written in monolingual English from the very beginning.7 When collecting and gathering the manuscripts kept before 1449 in order to enter them into the new ledger, the wardens in office may have become aware of the enduring use of English in the Renter Wardens’ Accounts, and Cantelowe, at least, might have decided to transmit this Englishness to another texttype. In sociolinguistic terms, Cantelowe would have acted as the prototypical bridge through which innovations flow rather easily. Further examples of other

7 The only texts kept in a language other than English, in this case, Latin, were a series of rentals of 1444–1445 or 1447 (1v–3r in Jefferson 2009: 1018–1025).



Bridges of innovation and change 

 57

mercer-wardens in a rather similar situation and with rather similar behaviour (i. e. contact with English via the Renter Wardens’ Accounts as well as later use of English elsewhere) could help to support the hypothesis of mercer-bridges’ documents as potential motivators for the wholesale adoption of English (see Sections 4.2 and 4.3). Overall and going back to Whatley’s analysis as the possible innovator, it has been proved that, according to the social networks of the four candidates, Whatley appears to be the most likely introducer of the earliest groups of ordinances originally kept in 1407–1408 in the Wardens’ Accounts. Notwithstanding this evidence, the Mercers’ Company later devoted a book specifically for composing all its records of constitution and operation: the Book of Ordinances. The ledger is likely to have been bought during 1436–1437, as the payment made by William Haxay, then rent-collector to the Mercers, is fully recorded in the Mercers’ Wardens’ Accounts (126r in Jefferson 2009: 506–507) but, more importantly for the present research, the parchment appears to have remained unused for years. Parker (1980: 10, 16) is very doubtful about its immediate use and prefers to suggest 1448–1449 as a tentative starting date. It is true that the Book of Ordinances is written entirely in English, but the oaths and ordinances kept in its first pages are undated. It could also be that quires recording texts in monolingual Latin or French from 1436 to 1448 have been lost. These two facts lead to a common point of interest: inherent limitations when dealing with ancient textual data. In this case, searching for the users and making any further interpretations regarding their language choice would result in a quite unproductive outcome for this new ledger.

4.2 Geoffrey Boleyn Geoffrey Boleyn appears to be another promising potential bridge here. The introducer of English into the rental accounts of the Mercers’ Company was a person outside the guild: John Mortham.8 After John Carpenter’s death, Mortham also handed to the four wardens in office together (John Olney, Geoffrey Fielding, Geoffrey Boleyn himself, and John Burton) the earliest rental accounts detailing Richard Whittington’s estate in English. Olney, Fielding, Boleyn, and Burton consequently may have had direct contact with English as the main language of this type of records.

8 The amount of extant biographical data on Mortham is too scanty to even attempt to sketch his social networks. He is only known for being John Carpenter’s rent-collector.

58 

 José Miguel Alcolado Carnicero

Very little can be said about Boleyn’s ego-centric networks within the Mercers’ Company. He did not enter the guild by apprenticeship, as most of the members did, but rather by redemption in 1429–1430 (Mercers’ Wardens’ Accounts 1435– 1436: 118v in Jefferson 2009: 476–477); therefore, he had no master attached to the Mercers. None of his 16 apprentices served an office of wardenship either. In contrast, Boleyn himself acted as warden to the Mercers’ Company twice, once in 1442–1443 and again in 1453–1454, but it is not recorded in the Wardens’ Accounts who was the warden appointing him to office on either of the two occasions; Boleyn’s only recorded wardens’ election is from 1443, when he elected William Oliver as his successor in office. As for Boleyn’s socio-centric networks, much more information is available, since he was a prominent man in late medieval London. Boleyn wed a Denise but, after being widowed, he immediately married one Anne (née Hoo), with whom he seems to have had several children (Wedgwood and Holt 1936: 90) (–1 for “marital status”). He was born in Norfolk, a region of origin of many mercers, where he also purchased the estate of Blickling from Sir John Fastolf in 1460 (0 for “place of origin and/or of living”). Boleyn’s role in the release of licences to evade commerce in Calais is well documented (Sutton 2005: 257) (+1 for “business destinations”). Sheriff of London (1446), Member of Parliament for London (1449), alderman of Castle Baynard Ward (1452–1457) and of Bassishaw Ward (1457–1463) in London, or Lord Mayor of London (1457–1458) are some of the offices held by Boleyn (+1 for “offices”). Finally, Boleyn was a convinced Yorkist and devoted much of his time to pay good service to Richard, Duke of York; this is perhaps the main reason why Boleyn was exempted from being put on assizes and juries as soon as the duke died in battle (Maxwell-Lyte 1897: 42) (+1 for “contacts”). Boleyn, then, scores +2 in the network strength scale. Focusing now on Boleyn’s language behaviour after his presumed contact with monolingual English records in 1442–1443, the next and final time he served in office is in 1453–1454. During this fiscal year, a new record book, the Acts of Court, was created. Intended initially for keeping the records of the governing body of the Mercers’ Company, manuscripts kept in the first decade deal only with the functioning of the Merchant Adventurers’ Company of England (Lyell 1935; Imray 1964). That guild united all English merchants travelling overseas, mostly to the Low Countries, for commerce. During the fifteenth century, the Merchant Adventurers’ Company became directly dependent on the Mercers’ Company of London; for instance, mercers such as John Warren or William Caxton were governors to the Adventurers, and both companies used the Mercers’ Hall as headquarters (Creaton 1976: 99). But undoubtedly more important for this research, and probably the reason why the Acts of Court are full of records from the Merchant Adventurers in its first pages, is the fact that the Mercers’ clerk used to make up



Bridges of innovation and change 

 59

the Adventurers’ accounts as well (Carus-Wilson 1933: 160). Furthermore, the Merchant Adventurers’ manuscripts were written in English from the beginning; as were the earliest writings from the Mercers’ Company’s Court of Assistants in 1463. As first warden of the Mercers, Boleyn might have had a say in the language choice of the newly-created book and, perhaps influenced by his loose-knit networks, could have opted for choosing the same main language of the Renter Wardens’ Accounts he came into contact with in 1442: English. In relation to this, it is also significant to note that the first warden in office in 1463–1464, when the proper records from the Mercers’ governing body of the Court of Assistants began to be engrossed, also in English, was Ralph Verney, who was second warden in 1453–1454 as well. Verney’s potential relevance warrants some discussion here.

4.3 Ralph Verney Ralph Verney is actually the other mercer-warden whose sociolinguistic behaviour may show strong evidence of language users in the Mercers’ Company acting as bridges for innovation and change in the main language of the records. Verney’s above-mentioned wardenships, 1453–1454 and 1463–1464, corresponded to his second and fourth terms in office; his third wardenship during 1458–1459 provides further relevant information: A second shift to English occurs in the Mercers’ Wardens’ Accounts after six consecutive years (from 1452–1453 to 1457– 1458) of French as the main language of record. In this particular case, it was Verney who could have been responsible for following the example in the Acts of Court’s ledger and for shifting to monolingual English again in the financial records. Like Geoffrey Boleyn, Verney’s socio-centric networks are characterised by what has made language users potentially act as bridges: a positive score (see table 4), suggesting then that formative networks were generally quite open. Table 4: Ralph Verney’s socio-centric network strength scale Marital status –1

Place of origin and/or of living 0

Business destinations +1

Offices Contacts +1

+1

Network score +2

Verney was married twice; firstly, to Eleanor, daughter of Geoffrey Pole and mother of three of his children, by around 1435 and, subsequently, to Emma Pickering, with whom he had one more daughter (–1 for “marital status”). Verney was ­apparently born in Claydon, Buckinghamshire, but might have moved to London

60 

 José Miguel Alcolado Carnicero

with his family soon after his birth due to his father being appointed Serjeant-­atMace (Wedgwood and Holt 1936: 906–907) (0 for “place of origin and/or of living”). However, he seems to have maintained strong connections with his county of origin: Verney was granted lands in Aylesbury and Barton, Buckinghamshire, which had been forfeited by William Wandesford in 1467 (Maxwell-Lyte 1900: 33); he was also a witness to the quitclaim of the manor of Hurdwyke by Margaret Hungerford in 1468 (Bird and Ledward 1953: 31) (+1 for “business destinations”); and he was elected Member of Parliament for the constituency of Buckinghamshire in 1473 (White 1980: 213) (+1 for “offices”). Finally, he was knighted in 1471 for his participation in the defence of London (White 1980: 213) (+1 for “contacts”). Verney’s sociolinguistic profile then may go some way to confirming the correlation between loose-knit socio-centric networks, represented by positive scores in the scale, and increasing use of written English, which seems to have been slowly spreading among members until reaching the whole livery company.

5 “The Folk of the Mercery”: The origin of textual language transmission? English has been introduced as main language of the different internal documents authored by the Mercers’ Company of London by mercer-wardens with loose-knit socio-centric networks and, even in some cases, with shared ego-­ centric networks. These innovative informants have sometimes acted as bridges transmitting the use of English as main language from one text-type to another. It seems interesting to analyse now if similar sorts of connections can be found between these internal documents and “The Folk of the Mercery”, as the earliest (external) document ever penned in monolingual English by the Mercers’ Company of London. Paying attention firstly to the exact date when “The Folk of the Mercery” was written by the Mercers, one difficulty immediately emerges: It is not possible to assign a precise date to “The Folk of the Mercery”, insofar as, unlike the internal records, the petition does not provide any specific date. A further obstacle for the necessary identification of the earliest users of written English in the Mercers’ Company of London is encountered as well: It is not possible to know which mercer-wardens presented the petition or even which were in office at that date.9 9 The fact that the text represents a harsh attack on one of the most influential men in London at the time (Nicholas Brembre) seems to be the perfect explanation for the complete absence of any name alluding to members of the Mercers’ Company of London, even in the document itself.



Bridges of innovation and change 

 61

As for the first limitation, in the scholarly literature three different dates of composition have been proposed for the Mercers’ parliamentary petition, but none of them has been totally conclusive: 1386 by R. Chambers and Daunt (1931), 1387 by Connolly and Mooney (2008), and 1388 by Giancarlo (2007). The authorship of the petition could be attributed to the mercers in office from 1386 to 1388, if their names had survived, but the only extant (and yet uncertain) list of wardens closest to these years dates from 1384–1385: (5) 1385 Jan., John Bosham, John Shadworth, William Sheringham, Thomas Austin, John Lovey (?) (Sutton 2005: 555) Here another tie found in Whatley’s ego-centric network acquires special relevance. Whatley, interestingly, learned the skills of mercery as an apprentice to William Sheringham, one of the wardens in office during 1384–1385, i. e. very close to the moment when “The Folk of the Mercery” might have been submitted to Parliament. A doubt inevitably arises here; in the same way as Whatley is hypothesised to have transmitted the use of written English to his apprentice, Cantelowe, could not Sheringham have done the same with Whatley, his apprentice? These three members could be demonstrating that an important written language influence was exerted through the master-apprentice tie and that, at least in the Mercers’ Company, the masters might be instructing their apprentices in accounting skills in English (Barron 1996: 224). Furthermore, a positive answer to the previous question would also imply that Sheringham did participate in the composition of the parliamentary petition in the vernacular, either because he was again in office on any of the three dates proposed in the literature, or because “The Folk of the Mercery” was actually presented to Parliament during 1384–1385. Nevertheless, despite belonging to the same ego-centric network and evolving normally in loose-knit socio-centric networks (+3 in network strength scale),10 Sheringham may not have acted as a bridge in the same way as Whatley

10 William Sheringham married a Maud, who did not give him any children and who died before him; Sheringham seems to have remained a widower for the rest of his life (0 for “marital status”). His surname indicates that he could have been born in Norfolk, land of great merchants; certainly, Sheringham migrated to London later, where he invested some money in land, mainly, by acting as a feoffee in property transactions of premises in the City (0 for “place of origin and/ or of living”). Sheringham appears in records as an active exporter of wool to the Low Countries as early as in 1371 (+1 for “business destinations”). He held important national offices such as sheriff of London and Middlesex (1395–1396) and, above all, Member of Parliament (1391) (+1 for “offices”). Finally, he was also a regular supplier of the future Henry IV’s wardrobe with luxury goods (+1 for “contacts”) (Roskell, Clark, and Rawcliffe 1993, 4).

62 

 José Miguel Alcolado Carnicero

and Cantelowe – i. e. by independently introducing shifts to English into some internal records of the Mercers’ Company –, but only by bringing the innovation from outside, i. e. from an external document, and spreading it among his apprentices.11 It is impossible to provide here a proven identification of any one innovator among the group of mercers submitting their complaint to Parliament, and this is the reason why more far-reaching disquisitions on the petition would be too risky to make. Nevertheless, the possibility has been offered that a warden in office during the years close to its submission (William Sheringham) might have been (i) a composer or, at least, an instigator of the content and, more importantly, of the form of the petition, and therefore also (ii) one of the primeval sources of the initial uses of English in the records of the Mercers’ Company of London later introduced by one of his apprentices (John Whatley) and his apprentice’s apprentice (William Cantelowe).

6 Conclusions In line with Li (1994), L. Milroy and Li (1995), and Li and L. Milroy (2003), this article has confirmed the meaningful applicability of the Milroyian model to situations of multilingualism, but has taken a step further, transporting the model’s refined application to earlier societies (Bergs 2005), thereby enhancing a promising marriage with historical sociolinguistics. Following this explanatory model for the propagation of the shift to written English in the Mercers’ Company, innovators and bridges have been proposed for all the external and internal documents produced by this livery company. Table 5 proposes an organised timeline of the different language changes.

11 Sheringham had another apprentice serving as a mercer-warden in 1397–1398: Nicholas Hamme. However, Hamme’s sociolinguistic behaviour does not seem to have been as innovative as Whatley’s; the main reason may mainly reside in his negative score at the socio-centric network strength scale (–3), revealing rather closed social networks. Hamme was married to a certain Joan (–1 for “marital status). He seems to have been a Londoner, since he is not recorded being involved in matters of property inheritance in any county or even in Ham, the place near London (at that time) his surname could be actually revealing (+1 for “place of origin and/or of living”). Nicholas Hamme’s name exclusively appears in manuscripts dealing with land transactions in different parts of London (e. g. St Pancras Soper Lane or St Martin Pomary), pointing to him as mainly focused on London local affairs (–1 for “business destinations”). Finally, Hamme does not seem to have held any important offices (–1 for “offices”) or performed any duties at the service of the high nobility (–1 for “contacts”) (I. Rogers 2004).



Bridges of innovation and change 

 63

Most mercer-wardens with more loose-knit socio-centric networks – scoring higher in the network strength scale – have usually tended to adopt vernacular deviations away from the traditional written language norms more easily (Geoffrey Boleyn and/or Ralph Verney). The restriction “most” applies, because it is not for the mere fact of having a positive network score that informants might have shown innovative preference for the English variant (John Woodcock). In addition to this, mercer-wardens have also tended to be part of ego-centric networks where the use of the written vernacular was promoted and, consequently, they might have previously come into contact with English as a primary language of documentation within their apprenticeship training as future mercers. It has even been proved that some original innovators of English had a master-apprentice relationship. William Sheringham (a possible composer of “The Folk of the Mercery”) is the master of John Whatley (introducer of English in the ordinances), who, in turn, is the master of William Cantelowe (warden in office when the Wardens’ Accounts were recorded in English for the very first time). The evidence strongly hints that social networks, both ego-centric and socio-centric, possibly played a role in the process of the adoption of English as the main language in the different records of the Mercers’ Company of London. Whereas loose-knit networks seem to have fostered the spread and adoption of the English vernacular in late medieval London, more close-knit networks have tended to maintain the use of traditional languages of record such as Latin and French (e. g. John Middleton and John Eton). Table 5: Chronological shift to English in the records of the Mercers’ Company of London. Text circulation

Type of record

Date(s) of shift(s)

Innovator(s) and/or bridge(s)

External

Parliamentary petition Constitution and operation

Between 1384–1385 and 1387–1388 1407–1408 (1st) Between 1436–1437 and 1448–1449 (2nd) 1442–1443

William Sheringham

Internal

Estates and ­charities

Finances Governing bodies

1449–1450 (1st) 1458–1459 (2nd) 1452–1453

John Whatley (1st) ? (2nd)

John Mortham Geoffrey Boleyn (contact in 1442–1443) William Cantelowe (contact in 1449–1450) William Cantelowe (1st) Ralph Verney (2nd) Geoffrey Boleyn Ralph Verney (contact in 1453–1454)

64 

 José Miguel Alcolado Carnicero

Needless to say, this research is far from being complete. On the one hand, much more extensive social data about the informants is required so that the reconstruction of their networks does not depend on mere supposition on some occasions. In the end, the identification (and exclusion) of mercers-wardens as innovators and bridges has also been made paying crucial attention to informants’ social networks and in terms of likelihood; but obviously, this does not mean that there are no other alternative proposals such as, for example, the possibility of innovators’ fellow-wardens (e. g. John Woodcock) accepting (or even introducing themselves) the shift to the new language. On the other hand, it should be tested to see if the tendencies suggested here constituted an isolated phenomenon only present in the Mercers’ Company or if, rather, the pattern of guided change also existed in the networks of members of other late medieval London livery companies, probably as part of a more general trend. Further investigation along the promising line of research proposed in this article would definitely lead to a better understanding of the process of shift to written English, not only in the manuscripts kept by the livery companies but perhaps also in the writing of other day-to-day business activities of late medieval London overall.

References Baildon, William P. 1904. The Home Counties Magazine, vol. 6. London: F. E. Robinson & Co. Barron, Caroline. 1996. The expansion of education in fifteenth-century London. In John Blair & Brian Golding (eds.), The cloister and the world: Essays in medieval history in honour of Barbara Harvey, 219–245. Oxford: Clarendon Press. Barron, Caroline & Anne F. Sutton (eds.). 1994. Medieval London widows, 1300–1500. London & Rio Grande: The Hambledon Press. Bax, Randy. 2000. A network strength scale for the study of eighteenth-century English. In Ingrid Tieken-Boon van Ostade, Terttu Nevalainen, & Luisella Caon (eds.), Social network analysis and the history of English, 277–289. Bergs, Alexander. 2000. Social networks in pre-1500 Britain: Problems, prospects, and examples. In Ingrid Tieken-Boon van Ostade, Terttu Nevalainen, & Luisella Caon (eds.), Social network analysis and the history of English, 239–251. Bergs, Alexander. 2005. Social networks and historical sociolinguistics: Studies in morphosyntactic variation in the Paston letters, 1421–1503 (Topics in English Linguistics 51). Berlin & New York: Mouton de Gruyter. Bird, William H. & Kenneth H. Ledward (eds.). 1953. Calendar of Close Rolls, 1468–1476. London: Her Majesty’s Stationery Office. Blom, Jan-Petter & John J. Gumperz. 1972. Social meaning in linguistic structure: Code switching in Norway. In John J. Gumperz & Dell Hymes (eds.), Directions in sociolinguistics: The ethnography of communication, 407–434. New York: Holt, Rinehart, & Winston. Burgon, John W. 1839. The life and times of Sir Thomas Gresham, vols. 1–2. London: Jennings.



Bridges of innovation and change 

 65

Carus-Wilson, Eleonora M. 1933. The origins and early development of the Merchant Adventurers’ organisation in London as shown in their own medieval records. The Economic History Review 4(2). 147–176. Chambers, Jack K., Peter Trudgill, & Natalie Schilling-Estes (eds.). 2003. The handbook of language variation and change (Blackwell Handbooks in Linguistics 11). Malden: Blackwell. Chambers, Raymond W. & Marjorie Daunt (eds.). 1931. A book of London English, 1384–1425. Oxford: Clarendon Press. Connolly, Margaret & Linne Mooney. 2008. Introduction. In Margaret Connolly & Linne Mooney (eds.), Design and distribution of late medieval manuscripts in England, 1–6. York: York Medieval Press. Cooper, Chris R. 1984. The archives of the City of London livery companies and related organisations. Archives 16(72). 323–353. Coulmas, Florian (ed.). 1997. The handbook of sociolinguistics (Blackwell Handbooks in Linguistics 4). Oxford: Blackwell. Creaton, Heather (ed.). 1976. The Wardens’ Accounts of the Mercers’ Company of London, 1347, 1391–1464. London: University of London MPhil dissertation. Crystal, David. 1995. Speaking of writing and writing of speaking. Longman Language Review 1. 5–8. Fisher, John H. 1996. The emergence of Standard English. Lexington: The University Press of Kentucky. Gal, Susan. 1979. Language shift: The social determinants of linguistic change in bilingual Austria. New York: Academic Press. Giancarlo, Matthew. 2007. Parliament and literature in late medieval England. Cambridge: Cambridge University Press. Granovetter, Mark. 1973. The strength of weak ties. The American Journal of Sociology 78(6). 1360–1380. Granovetter, Mark. 1974. Getting a job: A study of contacts and careers. Cambridge: Harvard University Press. Granovetter, Mark. 1983. The strength of weak ties: A network theory revisited. Sociological Theory 1. 201–233. Gumperz, John J. 1982 [1976]. Social network and language shift. In John J. Gumperz (ed.), Discourse strategies (Studies in Interactional Sociolinguistics 1), 2nd edn., 38–58. Cambridge: Cambridge University Press. Hardy, William J. & William Page (eds.). 1915. List of members of the Mercers’ Company from 1347. London: The Mercers’ Company of London. Hernández Campoy, Juan M. & Juan C. Conde Silvestre (eds.). 2012. The handbook of historical sociolinguistics (Blackwell Handbooks in Linguistics 68). Oxford: Wiley-Blackwell. Imray, Jean M. 1964. The Merchant Adventurers and their records. Journal of the Society of Archivists 2(10). 457–467. Jefferson, Lisa (ed.). 2009. The medieval account books of the Mercers of London: An edition and translation. Farnham & Burlington: Ashgate. Kristol, Andres. 2000. L’intellectuel “anglo-normand” face à la pluralité des langues: Le témoignage implicite du MS Oxford, Magdalen Lat. 188. In David Trotter (ed.), Multilingualism in later medieval Britain, 37–52. Cambridge: D. S. Brewer. Kurath, Hans & Sherman M. Kuhn (eds.). 1971 [1958]. Middle English dictionary, part B.5. Ann Arbor: The University of Michigan Press.

66 

 José Miguel Alcolado Carnicero

Labov, William. 1972. Language in the inner city: Studies in the Black English Vernacular. Philadelphia: University of Pennsylvania Press. Labov, William. 1980. Locating language in time and space (Qualitative Analyses of Linguistic Structure 1). New York: Academic Press. Labov, William. 1994. Principles of linguistic change: Internal factors (Language in Society 20). Oxford: Blackwell. Labov, William. 2001. Principles of linguistic change: Social factors (Language in Society 29). Oxford: Blackwell. Lee, Nancy H. 1969. The search for an abortionist. Chicago: University of Chicago Press. Leith, Dick. 1983. A social history of English. London: Routledge & Kegan Paul. Li, Wei. 1994. Three generations, two languages, one family: Language choice and language shift in a Chinese community in Britain (Multilingual Matters 104). Clevedon: Multilingual Matters. Li, Wei & Lesley Milroy. 2003. Markets, hierarchies, and networks in language maintenance and language shift. In Jean-Marc Dewaele, Alex Housen, & Wei Li (eds.), Bilingualism: Beyond basic principles – Festschrift in honour of Hugo Baetens Beardsmore, 128–140. Clevedon: Multilingual Matters. Lyell, Laetitia. 1935. The problem of the records of the Merchant Adventurers. The Economic History Review 5(2). 96–98. Maxwell-Lyte, Henry C. (ed.). 1897. Calendar of the Patent Rolls, A. D. 1461–1467. London: His Majesty’s Stationery Office. Maxwell-Lyte, Henry C. (ed.). 1900. Calendar of the Patent Rolls, A. D. 1467–1477. London: His Majesty’s Stationery Office. Maxwell-Lyte, Henry C. & Alfred E. Bland (eds.). 1910. Calendar of the Patent Rolls, A. D. 1452–1461. London: His Majesty’s Stationery Office. Milroy, James. 1992. Linguistic variation and change: On the historical sociolinguistics of English (Language in Society 19). Oxford: Blackwell. Milroy James & Lesley Milroy. 1985. Linguistic change, social network, and speaker innovation. Journal of Linguistics 21. 339–384. Milroy, Lesley. 1987 [1980]. Language and social networks (Language in Society 2), 2nd edn. Oxford: Blackwell. Milroy, Lesley & James Milroy. 1992. Social network and social class: Toward an integrated sociolinguistic model. Language in Society 21(1). 1–26. Milroy, Lesley & Wei Li. 1995. A social network approach to code-switching: The example of a bilingual community in Britain. In Lesley Milroy & Peter Muysken (eds.), One speaker, two languages: Cross-disciplinary perspectives on codeswitching, 136–157. Cambridge: Cambridge University Press. Nevalainen, Terttu & Helena Raumolin-Brunberg. 2003. Historical sociolinguistics: Language change in Tudor and Stuart England. London: Pearson Education. Parker, John (ed.). 1980. The Book of Ordinances of the Mercers’ Company of London. London: University of London MPhil dissertation. Poplack, Shana. 1979. Sometimes I’ll start a sentence in Spanish y termino en español: Toward a typology of code-switching. Centro de Estudios Puertorriqueños Working Papers 4. 4–83. Raumolin-Brunberg, Helena. 1996. Historical sociolinguistics. In Terttu Nevalainen & Helena Raumolin-Brunberg (eds.), Sociolinguistics and language history: Studies based on the Corpus of Early English Correspondence, 11–38. Amsterdam: Rodopi.



Bridges of innovation and change 

 67

Richardson, Douglas & Kimbal G. Everingham (eds.). 2011 [2005]. Magna Carta ancestry: A study in colonial and medieval families, vol. 1, 2nd edn. Baltimore: Genealogical Publishing. Rogers, Everett. 2010 [1962]. Diffusion of innovations, 5th edn. Glencoe: Free Press. Rogers, Ian (ed.). 1975–. Fifteenth-century biographical index. http://www.girders.net Rogers, Ian. 2004. Hamme, Nicholas, (fl. 1410). In Ian Rogers (ed.), http://www.girders.net/ index.php?dir=Ham%2F (accessed 10 September 2012) Rogers, Ian. 2009. Middleton, John, of London, (fl. 1400–1416). In Ian Rogers (ed.), http://www. girders.net/index.php?dir=Mi%2F (accessed 28 August 2012) Rogers, Ian. 2010. Eton, John, of London, mercer, (fl. 1411). In Ian Rogers (ed.), http://www. girders.net/index.php?dir=Et%2F (accessed 20 September 2012) Roskell, John S., Linda Clark, & Carole Rawcliffe (eds.). 1993. The history of Parliament: The House of Commons, 1386–1421, vols. 1–4. Stroud: Alan Sutton Publishing. Rothwell, William. 1991. The missing link in English etymology: Anglo-Norman. Medium Aevum 60(2). 173–196. Sharpe, Reginald R. 1894. London and the kingdom, vol. 1. London: Longmans, Green, & Co. Sutton, Anne F. 1998. The silent years of London guild history before 1300: The case of the Mercers. Historical Research 71(175). 121–141. Sutton, Anne F. 2005. The mercery of London: Trade, goods, and people, 1130–1578. Aldershot: Ashgate. Sutton, Anne F. 2009. Fifteenth-century mercers and the written word: Mercers and their scribes and scriveners. In Julia Boffey & Virginia Davis (eds.), Recording medieval lives: Proceedings of the 2005 Harlaxton Symposium (Harlaxton Medieval Studies 27), 42–58. Donington: Shaun Tyas. Sutton, Anne F. & Peter Hammond. 1978. The problems of dating and the dangers of redating: The Acts of Court of the Mercers’ Company of London, 1453–1527. Journal of the Society of Archivists 6(2). 87–91. Tieken-Boon van Ostade, Ingrid, Terttu Nevalainen, & Luisella Caon (eds.). 2000. Social network analysis and the history of English. [Special issue]. European Journal of English Studies 4(3). Veale, Elspeth. 1991. The “Great Twelve”: Mistery and fraternity in thirteenth-century London. Historical Research 64(155). 237–263. Wang, William. 1969. Competing change as a cause of residue. Language 45. 9–25. Wedgwood, Josiah C. & Anne Holt (eds.). 1936. History of Parliament: Biographies of the members of the Commons’ House, 1439–1509. London: His Majesty’s Stationery Office. White, William J. 1980. Further Yorkist connections in Bucks: John Heton, Richard Fowler, and Sir Ralph Verney. The Ricardian 5(69). 212–215. Wright, Laura. 1994. Early modern London business English. In Dieter Kastovsky (ed.), Studies in Early modern English (Topics in English Linguistics 13), 449–465. Berlin & New York: Mouton de Gruyter. Wright, Laura. 1996. Sources of London English: Medieval Thames vocabulary. Oxford: Clarendon Press. Wright, Laura. 1998. Mixed-language business writing: Five hundred years of code-switching. In Ernst H. Jahr (ed.), Language change: Advances in historical sociolinguistics (Trends in linguistics: Studies and monographs 114), 99–118. Berlin & New York: Mouton de Gruyter.

Esther-Miriam Wagner

4 The socio-linguistics of Judaeo-Arabic mercantile writing Abstract: This article sets out to investigate sociolinguistic phenomena in the Judaeo-Arabic mercantile correspondence of the Cairo Genizah. The difference between mercantile and scribal handwriting constitutes the first subject of examination. As in other languages, mercantile Judaeo-Arabic is written in a much more cursive hand than, for example, contemporary legal documents, and we can thus establish a Judaeo-Arabic “mercantesca”. In the second section, I investigate three separate corpora from the early medieval, the late medieval, and the early modern period, and show how the written language used by merchants was different from that used in other types of correspondence, such as private letters, and letters used for community purposes. Several distinct developments can be discerned: in the early medieval period, traders avoid linguistic forms marked as “Jewish” by shunning Hebrew and Aramaic vocabulary in their writing, and tend to use a register of Arabic that would have been very similar to that used by their fellow Christian and Muslim merchants. In sociolinguistic terms, this may suggest that medieval Jewish traders perceived themselves as “merchants” in the first place, and that their Jewish identity was secondary. The language used in the letters thus grounds documentary commercial Judaeo-Arabic in the more syncretistic common tongue of Middle Eastern merchants, and follows the style and norms of contemporary Muslim society to a larger degree than community or religious correspondence. In the late medieval and early modern period, mercantile writing continues to be composed in the spoken language, Arabic, whereas other contemporary Jewish documents and correspondence are written only in Hebrew and Aramaic. In addition, the Judaeo-Arabic used contains many more colloquial expressions than other contemporary text genres. In fact, many of the forms encountered in mercantile correspondence appear first in these letters and only later in other text sorts. Places where crossovers from the vernacular into the written language were particularly prevalent appear to be emotionally charged sentences, and colloquial sayings and adages, which are often used in the expressive language of business. The post-medieval merchants are thus unknowing agents of language change in the written medium,

Esther-Miriam Wagner, Woolf Institute and University of Cambridge DOI 10.1515/9781501503542-004



The socio-linguistics of Judaeo-Arabic mercantile writing 

 69

and contribute to changing written norms by introducing and propagating new linguistic forms.

1 Introduction: Business correspondence in the Cairo Genizah The great linguistic source presented by the materials of the Cairo Genizah offers us insights into a plethora of socio-linguistic phenomena.1 The documents allow us to conduct research into language contact and language change, the interaction between colloquial and written language, and gender-specificity of documentary writing. Letters sent between business men (and much fewer women) make up a large part of the epistolary inventory of the Cairo Genizah. The mercantile language mainly used by the Arabic-speaking Jewish traders of the Middle East was Judaeo-Arabic2, Arabic written in Hebrew characters, although they also used Hebrew.3 The literacy of the Jewish traders is still the subject of research, and probably varies considerably in the different time periods. Under the Ayyubids and Mamluks the economic downturn certainly affected education. We know little about the 14th–17th century because of the scarcity of documents discarded during that time as the deposition history of the Cairo Genizah is relatively uneven. Business letters from the medieval period come mainly from the 11th and 12th centuries, whereas there are much fewer sources from the following centuries. At the end of the 18th century, more documents were deposited again, and there is a large number of mercantile letters written between 1800 and 1830. In the following, a number of sociolinguistic phenomena pertaining to traders’ letters will be investigated. At first, I will focus on the difference between mercantile and scribal handwriting in the Cairo Genizah. Subsequently, I will

1 For a more detailed overview of the Cairo Genizah, its history and documents, see Goitein (1967: 1–28), and for a summary of most of the letters used as a basis for this article, Wagner (2010: 8–24). 2 The term Judaeo-Arabic is used in this article to designate Arabic written in Hebrew script, as suggested by Khan (2007: 526). The term is not used for spoken Jewish varieties of Arabic. 3 The comprehensive editions by Gil (1997) provide over a thousand of such letters. Further letters with mercantile content can be found in Gil (1983). For the historical context, see Goldberg (2012).

70 

 Esther-Miriam Wagner

investigate three separate corpora of mercantile correspondence to demonstrate how the language used by merchants was different from that used in other types of correspondence, such as private letters, and letters used for community purposes. What becomes obvious from the analyses of the material is that distinct differences are observable between letter writing pertaining to trade issues, and letter writing, for example, for private purposes and for community affairs.

2 Handwriting of traders One of the most obvious features of traders’ letters in the Genizah is the distinction in handwriting that marks mercantile correspondence. Whereas scribes of legal documents and community dignitaries appear to employ a comparatively square hand, business correspondence is often composed in a much more cursive style. The reason for the cursivity of business letters is obvious: traders often have to react very quickly to incoming letters to be sent out again on the next ship or relay urgent insider knowledge, therefore the act of composing a letter is distinctly different from that of composing a legal document or a well thoughtout community missive. In his classification of literacies, Parkes (1973: 555) has described the linguistic behaviour of traders as “pragmatic literacy”, owed to the circumstances of their acts of writing and the education of the merchants. They have to write very expressively and precisely to make themselves understood. In addition, we know from the Genizah sources that business letters were often sent out in four or five different copies in different ships to ensure their safe arrival. The practises of copying are described in great detail for example in the letter T-S 13J17.3 (published in Hebrew by Gil 1997: II 528-532). The writer mentions how he sent four copies of the same letter in four different ships, five copies of the same letter to his uncle in Qayrawan, and how he copied the recipients’ letters to be forwarded on to others. Thus writers had to copy their own letters a number of times, which must have additionally contributed to a more cursive style. The following figures 1–4 show the difference between contemporary 11th-century legal documents (figures 1 and 2, T-S 10J28.2 and T-S 24.18) and business ­correspondence (figures 3 and 4, T-S 13J15.9 and T-S 10J9.2). The images are representative for most of the letters and legal document from the period as found in the Cairo Genizah. Scribal hands (figs. 1 and 2) versus contemporary mercantile hands (figs. 3 and 4):



The socio-linguistics of Judaeo-Arabic mercantile writing 

Figure 1: T-S 10J28.2 (dated 1052 CE). Reproduced with kind permission by the Syndics of ­Cambridge University Library.

Figure 2: T-S 24.18 (dated 1055 CE). Reproduced with kind permission by the Syndics of ­Cambridge University Library.

 71

72 

 Esther-Miriam Wagner

Figure 3: T-S 13J15.9 (dated ca. 1053 CE). Reproduced with kind permission by the Syndics of Cambridge University Library.



The socio-linguistics of Judaeo-Arabic mercantile writing 

 73

Figure 4: T-S 10J9.2 (dated ca. 1054 CE). Reproduced with kind permission by the Syndics of Cambridge University Library.

As is obvious from the images, the handwriting used in correspondence is distinctly more cursive than the handwriting used by contemporary scribes in legal documents, and this cursivity can be observed in most of the mercantile correspondence. The same phenomenon has been described by Ceccherini (2009) for Italian traders, who developed their own style of handwriting, mercantesca, distinct from the contemporary chancery hands cancelleresca, by the 14th century. Iopollo (2010: 177) describes the blend of Secretary and Anglicana used by English merchants. In similar fashion, distinct mercantile and legal styles of handwriting can be assumed in the Genizah material of the 11th century, and I would like to suggest a Judaeo-Arabic mercantesca for the traders of the Fatimid Empire, in particular those active in the 11th century.

74 

 Esther-Miriam Wagner

3 Mercantile writing and language change 3.1 The dichotomy of Arabic written language and vernacular Although no language is spoken as it is written, the case of Arabic is particular. As a consequence of the Islamic conquests in the 7th century, Arabic spread from the Arabian peninsula over a vast area in relatively short time, leading to rapid language change due to contact with a large number of vernaculars originally spoken in the conquered areas. From this, various dialects of spoken Arabic emerged in the different countries of the Islamicate, yet only the variety in which the Qur’an and the religious literature were codified in the first centuries of Islam was used as a normative basis for the emerging written standard language of Arabic, and thus the prescribed written variety remained the same for all Arabic speakers. It is a religious requirement for Muslims to follow, as much as they can, its relatively uniform norms when they write Arabic. This dichotomy between written and spoken Arabic is one of the most striking features of Arabic, and was termed diglossia by Ferguson (1959), which is still, although critically viewed, the most widely used term among Arabists and linguists. Christians and Jews are bound to a lesser degree to the religious prescription of the standard language, in particular when they use other alphabets to write Arabic, and that is the reason why colloquial forms can often be found first and in larger quantities in the writing of those minorities. When we observe progressive language change in the medium of written Arabic, it is usually related to colloquial forms finding their way into writing. Because of the circumstances of composing business letters and the linguistic outsider status of Jewish traders, colloquial forms seem to appear earlier and more often in mercantile Judaeo-Arabic writing than in other text types. This occurrence and usage of colloquial forms is, however, as the examples below show, dependent on the social integration of the Jewish merchants into the common market place. Another topic discussed in the following is the use of particular Jewish language forms, often borrowed from Hebrew and Aramaic, the languages of liturgy and law, in business writing of the Jewish Middle Eastern merchants, and how the use of markedly Jewish forms differs from that in other documents. Wagner and Connolly (2017), for example, were able to show that the extent of code-switching between Hebrew and Arabic is largely dependent on the purpose of a letter. I am presenting three different case studies from the early medieval, the late medieval and the early modern period to demonstrate how mercantile writing was distinct in comparison to other sorts of texts, including other documents, and how



The socio-linguistics of Judaeo-Arabic mercantile writing 

 75

mercantile writing set precedents and incorporated progressive language forms before other text genres.

3.2 Case study A: Early medieval Genizah traders The medieval mercantile letters from the Genizah were written by traders who were part of networks spanning much of the Mediterranean, and which stretched, in particular in the 12th-century, all the way down to Yemen and eastwards to India. The writers of the letters were trading in the Fatimid Empire, which had a Shiite dynasty ruling over Sunni subjects. This meant that the Islamic majority by itself was not monolithic but quite diverse in nature, which in turn led to a relative tolerance toward the religious minorities, Jews and Christians, and their employment in government functions and at the court. At least until the middle of the 11th-century, the economy was prosperous, facilitating the emergence of a wealthy Jewish middle class. This middle class was not particularly segregated from other groups in the country and in close contact with Muslim authorities, for which they worked in the chanceries and other government offices, although their religious status also entailed the paying of a poll tax levied upon Jews and Christians. Jewish merchants conducted business with Muslim and Christian traders on a daily basis, with some traders dealing exclusively with Muslim traders in their business partnerships. Since the traders were in everyday contact with their Muslim counterparts, they must have been very familiar with Muslim mercantile writing. Yet, the medieval Jewish business letters as they have been preserved in the Genizah are, although composed in the Arabic language, to an overwhelming degree written in Hebrew script. This may be the result of the deposition history of the Genizah; perhaps, Arabic language script found their way into the Genizah on a less common basis, as suggested by Goitein (1971: 346) and Khan (1993: 138). We do, however, find a number of Arabic script letters sent by Jewish traders. In many of these cases we can probably assume that the addressees of the letters were not fellow Jewish traders, but Muslim or Christian merchants, as can be seen in a comparable collection, the Arabic letters of the Prize Paper Collections in the National Archives London currently prepared by myself and Mohamed Ahmed (under the working title: From Tuscany to Alexandria: Arabic and Hebrew mercantile letters in the Prize Paper Collections). Other letters may have been composed by Karaite traders, who preferred to write their spoken language in Arabic, not in Hebrew script. The evidence concerning the general ability of Jewish merchants to read and write in Arabic script is conflicting. Certainly, the main purpose of primary

76 

 Esther-Miriam Wagner

education was to learn to read the Bible, so children would have been in the first place been instructed in Hebrew script, which they then subsequently used to write other languages, too. In the Egyptian context, we additionally find evidence that boys were instructed in Arabic calligraphy and how much this was treasured (Goitein 1971: 193). Social class certainly played an important role as to whether a Jewish Egyptian learned the Arabic script or not. It was a sign of higher education in the upper echelons of society (see Goitein 1971: 346), and a necessity for particular professions. Jewish scribes working for government offices would have been trained in Arabic just as their Muslim colleagues. Doctors who treated patients of all religious confessions would have to be able to write their prescriptions in Arabic. Yet how widespread the knowledge of Arabic was within the mercantile profession is still unclear. The Genizah sources tell us about Muslim trading partners of Jewish merchants, so many businessmen would have needed to be able to read and write Arabic for the purpose of correspondence. However, some writers explicitly request not to receive letters in Arabic script because they cannot read them.4 Or, an 11th-century Jewish trader would write a note in Hebrew letters to his Muslim business partner and ask a Jewish acquaintance to read it out to him (Goitein 1971: 294). To complicate matters, reading and writing were two entirely different issues in the medieval period (see Wagner, Outhwaite & ­Beinhoff 2013: 5), and men of moderate learning may have been able to read various alphabets, but unable to write in them. The following statements will focus on the Judaeo-Arabic/Hebrew script letters alone, and although the use of Hebrew script may have been in some instances a purposeful barrier maintained among Jewish merchants, we will assume that the use of the Hebrew alphabet in mercantile had largely to do with primary education in Hebrew script and general literacy standards. Statistical analyses of Judaeo-Arabic mercantile correspondence presented by Wagner and Connolly (2017) clearly demonstrate that code-switching to Hebrew enjoyed the highest prestige in Genizah correspondence pertaining to religious matters. In such letters, the percentage of Hebrew and Aramaic could range between thirty and forty percent. This is not unusual as Hebrew was the liturgical language of the Jewish communities, while Aramaic was the language used in legal writing and legal documents. Since Judaism is a religion in which belief and law are intricately linked (religious identity is to a large degree derived from following the religious law), the frequent citation of Aramaic and Hebrew in letters with religious content is only natural. Those letters with mainly political content use Hebrew and Aramaic

4 See for example Mosseri IV.45, line 3 where the writer states: ‘you should write in Hebrew script so that we understand’.



The socio-linguistics of Judaeo-Arabic mercantile writing 

 77

less frequently than religious letters, with the proportion of Hebrew/Aramaic between 6 % and 27 % of the total letters. In mercantile correspondence, however, there is a sharp drop of Hebrew/Aramaic content, and the few Hebrew phrases do not normally exceed 2 % of the total word count in the letters, with some showing no Hebrew/Aramaic forms at all. This is a very stable number, which can be found across the majority of purely mercantile Judaeo-Arabic letters of the 11th century. Those Hebrew forms that do appear are often formulaic, and are part of a restricted repertoire. For example, the final goodbye greeting ‫‘ עקב ושלום‬reward and peace’, or the pious formulae ‫‘( ננ‬his soul is rested’, i. e. ‘deceased’) given after names occur relatively frequently. We also find the abbreviated expression ‫ כל ישר‬for ‘all of Israel’, the word goyyim used for ‘Muslims’, and šalom ‘peace’.5 In sum, Wagner and Connolly’s work demonstrates that there was an emphasis to use the language employed by all traders in their mercantile correspondence – Arabic. Their important identity as traders perhaps located even Jewish merchants firmly within Arabic culture, and therefore in their mercantile writing they shied away from using phrases and expressions that were seen as too markedly Jewish. This meant that, compared to other documents of the time, the traders’ correspondence contains much fewer instances of codeswitching between Arabic and Hebrew than comparable letters written for non-commercial purposes. Wagner and Connolly (2017) also show that particular individuals vary the Hebrew and Aramaic content of their letters according to the content of the correspondence. Through a case study of letters composed by the dignitary Daniel b. Azaryah, they demonstrate that individual writers were aware of which registers were appropriate for which audience and for which purpose of the letter, and varied their style accordingly. While Daniel b. Azaryah uses up to 40 % Hebrew/ Aramaic phrases in his letters discussing religious matters, he avoids any markedly Jewish forms in his letters concerned with trade. The fact that traders avoid using Hebrew phrases in their mercantile writing may have had a number of causes. A very important reason may be the emotional immediacy that is connected with using the spoken vernacular of the time in writing. The economy of the Fatimid Empire was largely built on mutual trusts and cooperation, so using more formal language forms such as Hebrew may have been counterproductive for the relations between individual traders. Also, on a pragmatic level, it would have been much easier to quickly write a 5 Goitein (1967: 271) also mentions that in certain periods hazardous or not permissible commodities were expressed in Hebrew to avoid detection by government agencies. Unfortunately, he cites no letters or classmarks to corroborate where such linguistic behaviour can be found. I have not come across this in medieval Judaeo-Arabic mercantile writing, although I have seen examples in Yiddish letters from Europe and in Late Judaeo-Arabic writing.

78 

 Esther-Miriam Wagner

reply to an urgent matter in Arabic, in particular as all the common trade vocabulary between Jew, Muslims and Christians would have been Arabic. This is a point specific to trade the Middle East in the Fatimid era, when there was much mixing between the different faith communities in the mercantile arena. At a later period in the Middle East, or in contemporary Europe, where the Jewish communities had become more segregated and where there were exclusively Jewish economic networks, the use of Hebrew/Aramaic forms may have been identity reinforcing. It is therefore extremely important to understand the social circumstances of any particular time period to evaluate the language used in mercantile correspondence. Segregation of faith minorities in the case of Jewish traders can potentially lead to the use of markedly Jewish forms, such as Hebrew words and phrases, in traders’ languages, in attempts to disguise knowledge specific to the trade to outsiders or to reinforce group identity. We see examples of this in Yiddish or Yiddish-derived Handelssprachen (trade languages) in Europe, where even non-­ Jewish traders adopted Hebrew words and phrases into their professional argots. We find, however, the opposite in the traders of the Fatimid Empire. Because trade was a multi-faith endeavour, and the trading networks included Jews, Muslims and Christians, the linguistic register of trade was levelled, and a common language was used that differed from that written in other types of correspondence and documents, and literary texts. This links with Algolado Carnicero’s observations in this volume: members of open mercantile networks write against the norms, i. e. in the case of Fatimid Jewish traders, there is a tendency to dismiss markedly Jewish linguistic forms commonly used in other types of documentary writing or literature in favour of a language ­commonly understood by all traders, be they Jewish, Muslim or Christian.

3.3 Case study B: Late medieval Genizah traders Whereas the Genizah has preserved hundreds and thousands of mercantile letters from the 11th and 12th centuries, there are substantially fewer letters from the 14th and 15th centuries. Empirical studies on the late medieval period are thus difficult to carry out due to the scarcity of the material. Two phenomena encountered in letters from the period will therefore serve as examples to demonstrate the trend-setting of the mercantile language of the time. The examples in the following come from a letter written at the end of the 15th century to the eminent Alexandrian community leader and trader Moses b. Judah by a certain Isaac Beit ʿAṭṭān, which deals with business matters relating to the trade of books and kosher cheese. The letter, kept in the Bodleian Library



The socio-linguistics of Judaeo-Arabic mercantile writing 

 79

in Oxford, shows an adverbial connective baš, which appears in final clauses, expressing the concept of ‘in order to’.6 ‫ואפתרקנה מאע ואחד אלבארכה באש נמורו פי סירקוסא‬ w-ʾftrqnh mʾʿ wʾḥd ʾl-bʾrkh and-depart.PAST.1PL with.PREP one (of-)the.ART-boats(?) bʾš nmwrw fy syrqwsʾ in.order.to.CON go.PRES.1PL to.PREP Syracuse Bodl Heb.72.c.18,4–5 ‘We departed with one of the ships (?) in order to go to Syracuse.’ . . ‫נזל ר צוריאל ור מצליח באש יאשתריו אל כובז ואל נביד‬ nzl R. Ṣwryʾl w-R. Mṣlyḥ bʾš and-go.down.PAST.3SG R. Ṣuriel and-R. Maṣliaḥ in.order.to.CON yʾštryw ʾl ḫwbz w-ʾl nbyḏ buy.PRES.3PL the.ART bread and-the.ART wine Bodl Heb.72.c.18,5–6 ‘R. Ṣuriel, together with R. Maṣliaḥ, went down in order to buy the bread and wine.’ ‫ועמרו זוג מראכב באש יאכדו אל מרכב‬ w-ʿmrw zwj mrʾkb bʾš yʾḫḏw and-approach.PAST.3PL pair ships in.order.to.CON take.PRES.3PL ʾl        mrkb the.ART     ship Bodl Heb.72.c.18,7 ‘a pair of ships approached in order to seize the ship.’ The examples appear to contain the earliest recorded occurrence of baš as a connective. In this meaning, it cannot be found in any of the historical dictionaries of Arabic or in any other medieval text. It does appear, however, in glossaries of modern spoken North-African vernaculars, such as Harrell and Sobelman (2008: 5–6), with the translation of ‘in order to’ and temporal ‘since’.7 While baš is elusive in any other written Muslim texts, I found it used as a connective in an early 20th-century Judaeo-Arabic printed book, which is in my private possession. The book was printed in Jerba in 1917 and bears the title Ora ve-simḥa (‘Happiness and joy’). In the book, baš serves as the complementizer ‘that’, as final and consecutive

6 The following three examples are taken from Wagner (2014: 148–149). 7 Wahrmund (1898: 258) has the Classical Arabic equivalent of bi-mā for ‘so dass (so that)’. The construction appears to be a colloquial form of bi-mā, whereby the Classical Arabic for mā is replaced by colloquial ayš, which is then shortened to aš. The form aš is a typical Maghrebian form, see Blau (1981: 62).

80 

 Esther-Miriam Wagner

connective ‘in order to’ and ‘so that’, and as causal connective ‘because’. In fact, baš appears in various examples on most pages of the 250 page pamphlet. As opposed to the usage indicated in Harrell & Sobelman, it seems to have taken over most of the functions of a complementizer and adjectival connective: ‫ויסבב לנפסהו באש יכון נאקץ מן אלכיר‬ w-ysbb l-nfs-hw bʾš ykwn and-quarrel.PAST.3SG to-self-PRONSUFF.3SG.M because.CON be.PRES.3SG nʾqṣ mn ʾl-ḫyr missing.PART.M from.PREP the.ART-good p. 25b, line 17 ‘and he would quarrel with himself because of missing all the good things’ ‫וכרגו גמיע נאס אלבלאד באש יעארצוהם‬ w-ḫrjw jmyʿ nʾs ʾl-blʾd bʾš and-walk.out.PAST.3PL all people the.ART-region in.order.to.CON yʿʾrḍw-hm honour.PRES.3PL-PRONSUFF.3PL p. 15b, line 8–9 ‘and all people of the region would go out to honour them’ Thus the occurrence in the mercantile letter of the 1480s appears to show syntactical connectives of a kind that can only be found again, as far as I am aware, in the colloquially influenced texts of the early 20th century, setting early linguistic precedents. Connectives are not the only instances of progressive forms in late medieval  mercantile writing. Another example displaying the first examples of ­particular forms before they occur in other types of writing can be found in GW XXX, a 15th-century letter from the Freer collection. This document was written by a  man originally from Old Cairo, who was trying to establish a business in Damascus. The phenomenon in GW XXX concerns numeral constructions. In Modern Egyptian Arabic, a t-infix appears in numeral constructions of numbers 3-10 when the following noun begins with alif. Not every noun starting with alif shows this t-linking but it commonly occurs with nouns of the pattern af‘āl. These constructions have the form ḫamast iyyām ‘5 days’ or talatt ̓alāf ‘3,000’, i. e., the short form numeral is followed by -t before the counted item, see Mitchell (1956: 62-63). In GW XXX, we find some of the earliest examples of this t-infix in numeral ­constructions. The following examples show a t-prefix attached to the counted: . . ‫ קעת ה תאיאם‬GW XXX/13 ‘I stayed 5 days’; ‫ עמלת ג תנצאף‬GW XXX/11 ‘I made three . half-dinars’; ‫ עמלת ה תנצאף‬GW XXX/10 ‘I made 5 half-dinars’; ‫ והבה עשר תנצאף‬GW XXX/24 ‘and a gratuity of 10 half-dinars’.



The socio-linguistics of Judaeo-Arabic mercantile writing 

 81

As with the example of the connective baš, mercantile correspondence appears to be the environment in which these numeral construction forms emerge for the first time. The traders in their writing thus seem to set precedents by introducing forms normally used in the vernacular into their writing.

3.4 Case study C: Early modern traders Economic worsening followed the end of the Fatimid Era, and slowly, we can see a shift in the documents of the Genizah away from Arabic towards Hebrew.8 By the 16th century, most letters and legal documents are composed in Hebrew. Yet, again it is the traders who defy this development. The early modern corpus of mercantile Genizah letters is distinct from other documents of the time as they are composed in Judaeo-Arabic, not in Hebrew and Aramaic as found in contemporary documents, correspondence and legal documents. The merchants thus show a preference for the spoken language of the time, which re-confirms what we have seen in the medieval corpus too: immediacy of language by use of the spoken language is of great importance in the sociolinguistic arena of business relations, which is built on mutual trust and co-operation. And again, we find traders writing against norms. The peculiarities of Late Judaeo-Arabic business correspondence have been the subject of study in Wagner (2013), which showed how early 19th-century traders developed their own linguistic register by incorporating both colloquial and very literary Classical Arabic forms. Similarly to the letters from the 15th century in the previous section, the 18th- and 19th-century business letters display a large number of forms which appear in writing for the very first time.9 These forms include vernacular phonology and verbal morphology, dialectal spelling of pronominal suffixes, syntactical innovations such as the negation particle š at the end of verbal nation phrases, and the introduction of a lexicon that is distinctly spoken and not ordinarily part of the written language. In fact, some of these forms are only now being introduced into Arabic writing, after the social media, such as Facebook and Twitter, facilitated the development of Latin and Arabic script writing systems for Arabic dialects. Vernacular forms appear particularly prevalent in emotionally charged phrases, i. e. when traders express their despair, exhaustion or anger about 8 For more details on this development, see Wagner & Outhwaite (2018). 9 A volume with 50 editions of Late Judaeo-Arabic business letters is currently collated by Prof Geoffrey Khan and myself, which will contain detailed linguistic analyses of mercantile correspondence of the early 19th century.

82 

 Esther-Miriam Wagner

something or someone. In the following example, the writer is upset about dodgy scribal or financial practices in Alexandria. The sentence is particularly rich in colloquial linguistic forms. The negation mā, normally a separate word in Standard Arabic, is shortened and attached physically to the verb it precedes. The 1st person plural verb niʿrifū shows the vernacular ending -ū, and another negation particle (-š) attached at the end that is only part of the spoken language. We also find a vernacular wh-word in ʾayš. ‫ומנערפוש איש מין סגולה עמלו מעהום פי נאאמון‬ w-m-nʿrfw-š ʾyš myn sgwlh ʿmlw and-not.NEG-know.PRES.1PL-NEG what from document make.PAST.3PL mʿhwm           fy Nʾʾmwn With-PRONSUFF.3PL    in Alexandria L-G Misc. 24,15–16 ‘and we do not know what sort of document they made out in Alexandria’. Another important point in connection with setting precedents and moving vernacular phrases into the written languages appear to be vernacular adages and popular sayings. These phrases are an essential part of expressive business writing, and we can speculate that they were introduced particularly for the purpose of establishing a connection to the reader and to create immediacy in the writing. In these adages we often find the most progressive spellings and morphology in the letters. The Genizah letter T-S NS 99.23 for example shows peculiar spelling of wallāh, originally ‘by God’ (used here for ‘nor’), colloquial lexicon in the words for ‘we’, ‘one-armed’ and ‘blind’, and a usage of the negation lam in an unusual syntactic environment: ‫לם אחנה כותע וואלה אחנה מבצרווייה‬ lam ʾḥnh kwtʿ ww-ʾlh not.NEG we.PRON one-armed nor (lit. and-God) mbṣrwwyyh blind T-S NS 99.23,23 ‘we are neither one-handed nor blind’.

ʾḥnh we.PRON

None of the words in the above example could be found in this form or function in a contemporary normative Standard Arabic text; sayings and adages perhaps cannot be “translated” into standard language and provide an opportunity for colloquial forms to enter the threshold of the written language, and to become codified for the first time. This phenomenon is not restricted to Judaeo-Arabic business correspondence. In a bundle of mainly Christian Arabic mercantile letters from the year 1759, which are part of the Prize Paper collections in The National Archive in London,



The socio-linguistics of Judaeo-Arabic mercantile writing 

 83

we also find a number of sayings in which the writers codeswitch from a language that mostly follows Standard Arabic conventions to a colloquial register.10 Expressions found include for example ‫‘ لسا متل ماهو‬still as it is’ and ‫‘ وقوي حصل لنا غم‬and we were badly upset’, colloquial sayings that feature typical vernacular words and expressions. As already stated, it could thus be argued that sayings often act as precedents for the introduction of forms; there is no way to express a colloquial saying in Standard Arabic and no point to “translate” it into a higher register as this would defeat the point of using it in the first place. Hence the frequent colloquial sayings employed in business correspondence could have acted as a means for the introduction of colloquial lexicon, and as a consequence also for the setting of precedents for vernacular phonology, morphology and syntax. To sum up, the Late Judaeo-Arabic mercantile correspondence follows the patterns we have seen in the late medieval letters. The traders use progressive forms in their correspondence that only appear in literary writing at a later stage. Emotional statements and colloquial sayings and adages may have played an important role in introducing vernacular forms for the first time, thus advancing the vernacularisation of the written language, and leading to language change in the written medium.

4 Conclusions As in the Italian documents described by Ceccherini (2009), we find that Judaeo-Arabic letters of the 11th century display a distinctly more cursive hand than comparable legal documents. The style in handwriting is due to the different writing circumstances encountered by traders as compared to other writers, for example those who composed legal documents. Traders had to react very quickly, writing large quantities of letters, often in multiple copies, dependent on outside factors such as leaving messengers, boats, or changes in markets and pricing. The fact that their writing would display greater cursivity is thus not surprising. Within the social group of traders, beautiful handwriting may also have enjoyed less prestige than in other groups of writers, and it was certainly outweighed by business concerns, that relied more heavily on speed of action.

10 The following examples are from the above mentioned project “From Tuscany to Alexandria: Arabic and Hebrew mercantile letters in the Prize Paper Collections”, which I am currently preparing together with Mohamed Ahmed, in which we aim to publish a volume with editions of all the Arabic Prize Paper letters from The National Archives in London.

84 

 Esther-Miriam Wagner

Several distinct developments in the linguistics of Jewish medieval and early modern mercantile correspondence can be discerned. Individuals vary the Hebrew and Aramaic content of their letters according to the content of the correspondence, and in the early medieval period, traders avoid linguistic forms that are marked as “Jewish” by shunning Hebrew and Aramaic vocabulary in their writing. In sociolinguistic terms, this suggests that medieval Jewish traders perceived themselves as “merchants” in the first place. This also suggests a greater integration of the Jewish population into the Egyptian Middle class than in later time periods, when we see an increase in Hebrew vocabulary, even in mercantile letters. In a way, the language used in the early medieval business letters grounds the contemporary documentary Judaeo-Arabic in the more syncretistic common tongue of Middle Eastern merchants, and follows the style and norms of contemporary Muslim society to a larger degree than community or religious correspondence. In the late medieval and early modern period, on the other hand, the segregation of the Jewish population makes them more likely to use words and phrases marked as Jewish as part of their business identity, which means we see an increased use of loanwords and phrases from Hebrew. This demonstrates that there is great variation within mercantile writing, and that we have to be careful to avoid generalisations about code-switching in mercantile language: while Jewish trade languages typically show a high proportion of Hebrew content, such as in Yiddish or in Judaeo-Arabic from the 13th-century onwards, this may not be the always be the case, as can be seen in early medieval material. The most unusual phenomenon in the late medieval and early modern period is the continued use of Arabic in mercantile letters when all other Jewish epistolary and documentary writing has long switched to Hebrew. This is probably due to the communicative function of using an immediate, i. e. spoken language, for communication in the business arena, where trust is an important foundation of business. The use of Arabic (in Hebrew script) is a phenomenon restricted to business correspondence, which also means in turn that the only documentary writing of the time is that of mercantile nature, which makes the traders the only force in the vernacularisation of the Judaeo-Arabic, vis-à-vis the linguistically more conservative literary texts. Judaeo-Arabic used in mercantile writing contains many more colloquial expressions than contemporary literary text genres. In fact, many of the forms encountered in mercantile correspondence appear first in these letters and only later in other text sorts. Places where crossovers from the vernacular into the written language were particularly prevalent appear to be emotionally charged sentences, and colloquial sayings and adages, which are often used in the expressive language of business. Merchants are thus, perhaps unknowingly, agents of



The socio-linguistics of Judaeo-Arabic mercantile writing 

 85

language change in the written medium, and contribute to changing written norms by introducing and propagating new linguistic forms.

Linguistic glossary ART article CON connective M masculine NEG negation particle PART participle PAST past PL plural PREP preposition PRES present PRON pronoun PRONSUFF pronominal suffix SG singular

References Blau, Joshua. 1981. The emergence and linguistic background of Judaeo-Arabic. 2nd Edition. Oxford: Oxford University Press. Ceccherini, Irene. 2009. Merchants and notaries: Stylistic movements in Italian cursive scripts. Manuscripta 53. 239–83. Ferguson, Charles. 1959. Diglossia. Word 15. 325–337. Gil, Moshe. 1983. Palestine during the first Moslem period (634–1099). Tel Aviv. 4 volumes. [in Hebrew] Gil, Moshe. 1997. In the Kingdom of Ishmael. Jerusalem: Mosad Byalik. 4 volumes. [in Hebrew] Goitein, Shelomo Dov. 1967. A Mediterranean society; the Jewish communities of the Arab world as portrayed in the documents of the Cairo Geniza. Volume I: Economic foundations. Berkeley: University of California Press. Goitein, Shelomo Dov. 1971. A Mediterranean society; the Jewish communities of the Arab world as portrayed in the documents of the Cairo Geniza. Volume II: The community. Berkeley: University of California Press. Goldberg, Jessica. 2012. Trade and institutions in the Medieval Mediterranean. Cambridge: Cambridge University Press. Harrell, Richard & Harvey Sobelman. 2008 (Reprint). A dictionary of Moroccan Arabic. Georgetown: Georgetown University Press.

86 

 Esther-Miriam Wagner

Ioppolo, Grace. 2010. Early Modern handwriting. In Michael Hattaway (ed.), A new companion to English Renaissance literature and culture [2 vols], l.177–189. Chichester: Wiley-­Blackwell. Khan, Geoffrey. 1993. On the question of script in medieval Karaite manuscripts: New evidence from the Genizah. Bulletin of the John Rylands University Library of Manchester LXXV. 133–141. Khan, Geoffrey. 2007. Judaeo-Arabic. Encyclopedia of Arabic language and linguistics. II 526–536. Leiden: Brill. Mitchell, Timothy. 1956. An introduction to Egyptian Colloquial Arabic. Oxford: Clarendon Press. Parkes, Malcolm. 1973. The literacy of the laity. In David Daiches & Anthony Thorlby (eds.), The Mediaeval World, 555–577. London: Aldus Books. Wahrmund, Adolf. 1898. Handwörterbuch der neu-arabischen und deutschen Sprache. Giessen: J. Ricker’sche Verlags-Buchhandlung. Wagner, Esther-Miriam. 2010. Linguistic variety of Judaeo-Arabic in letters from the Cairo Genizah. Leiden: Brill. Wagner, Esther-Miriam. 2013. Challenges of Multiglossia: The emergence of substandard Judaeo-Arabic registers. In Esther-Miriam Wagner, Ben Outhwaite & Bettina Beinhoff (eds.), Scribes as agents of language change, 259–273. Boston/Berlin: De Gruyter Mouton. Wagner, Esther-Miriam. 2014. Subordination in 15th- and 16th-century Judeo-Arabic. Journal for Jewish Languages 2. 143–164. Wagner, Esther-Miriam & Magdalen Connolly. 2017 (accepted). Code-switching in Judaeo-Arabic documents from the Cairo Geniza. In Päivi Pahta, Janne Skaffari & Laura Wright (eds.), A multilingual approach to language history: Studies in historical codeswitching. Berlin: De Gruyter Mouton. Wagner, Esther-Miriam & Ben Outhwaite. 2018 (in press). “These two lines”: Hebrew and Judaeo-Arabic Letter-writing in the Classical Genizah Period. In Jennifer Cromwell & Eitan Grossman (eds.), Beyond free variation. Oxford: Oxford University Press. Wagner, Esther-Miriam, Ben Outhwaite & Bettina Beinhoff. 2013. Scribes as agents of language change. In Esther-Miriam Wagner, Ben Outhwaite & Bettina Beinhoff (eds.), Scribes as agents of language change, 1–18. Berlin: De Gruyter Mouton.

III Code-switching, loanwords and multilingualism

Ivar Berg

5 Business writing in early sixteenth-century Norway Abstract: This article discusses the linguistic make-up of business writing in late medieval / early modern Norway and addresses questions of language choice and language mixing. The most important foreign language was Low German, and the written culture also depended on Latin; these multilingual influences are evident in business records and accounting manuscripts from the early sixteenth century. A case study of language choice in international communication sheds light on the status of Latin as opposed to the vernaculars, as well as the relationship between the various vernaculars. Official documents are normally written in one language; however, in the more informal register of administrative and economical notes we find Latin and Low German words and phrases (code-switching and loanwords) in texts otherwise written in Scandinavian. This language mixing is markedly different from that of more formal writings and provides valuable insights into the general linguistic competence of those involved in trade, either as merchants or scribes. It is suggested that the code-mixed informal texts represent an unmarked, functional code for bilingual scribes.

1 Introduction The present contribution addresses Norwegian business writing mainly in the first half of the sixteenth century, albeit with some remarks on the earlier situation. For reasons outlined in Section 2 on the source material, there are very few written traces of merchant activities from this period. The title is thus phrased to cover business in a slightly wider sense, including notes and accounts by Norwegians dealing with merchants who were themselves of foreign origin. Norway entered a union with Denmark in 1380 (Sweden was also intermittently included in a pan-Scandinavian union from 1397 to 1523), with Denmark being the stronger and dominant partner. Union matters concerning Norway were written in Norwegian until 1450; from that year onwards official proclamations from the king were only issued in Danish. This initiated a large-scale language shift to Danish that started in high social layers and spread socially downwards. The language shift was not completed in documents written in remote rural areas until Ivar Berg, Norwegian University of Technology and Science DOI 10.1515/9781501503542-005

90 

 Ivar Berg

the latter half of the sixteenth century, yet most Norwegian writing after around 1500 adhered to Danish models, i. e. the language of leading Danes and the Danish chancery, although some Norwegian traits usually remained. The differences between Norwegian and Danish were – and more arguably still are – of course dialectal, and in what follows “Scandinavian” is used as a superordinate term. The Low German of the Hanseatic League had become the most important language of commerce in the North Sea area during the Late Middle Ages. One of the four main Hanseatic trading stations was the Kontor in Bergen, where they maintained a physical presence for several centuries. The long-term language contact with Low German had a profound impact on the Scandinavian languages and especially on the local dialect of Bergen (see e. g. Nesse 2012; Jahr 1999). The old international language Latin remained in use as the language of learning and the church, and was generally used in communication with the British Isles. As will be shown, the vernaculars were by this time replacing Latin in many domains, for instance trade. This is the general backdrop against which two questions will be addressed in Sections 3 and 4, respectively: 1. What languages did those involved in trade know, and in what situations did they use them? 2. How did different linguistic varieties interact in the written output of mercantile activities? Section 3 examines the languages used by traders in the North Sea area, mainly through a case study of an incident that brought a number of merchants and mariners of Scots, English, and Dutch origin to Norway. Section 4 looks more closely at the linguistic make-up of manuscripts written by Norwegians in their dealings with merchants, focusing on the relationship between Scandinavian and the important foreign languages Latin and Low German. As both “code choice and code-switching in multilingual societies are dependent on similar sociolinguistic factors” (Schendl 2012: 523–524), it makes sense to compare code choice in Section 3 with code-mixing in Section 4. The two perspectives are brought together in a concluding discussion in Section 5.

2 Sources Preserved documents from medieval Norway are published in the series Diplomatarium Norvegicum (DN), so far in 23 volumes. Most of them deal with legislation and a fairly restricted trade – land ownership. These documents were more



Business writing in early sixteenth-century Norway 

 91

important as titles to the land than as part of the trade act itself, and are thus connected to legal language and follow established legal formulae. Excavations in Norwegian towns, especially Bergen, have unearthed many runic inscriptions on wood dealing with matters of trade, pointing to the development of merchant literacy (Hagland 2011; Johnsen 1987). Many inscriptions contain names, obviously used as ownership labels, and a large number of tally sticks found in Bergen also point to commercial activities (Hagland 2011: 35). Simple tally sticks used for counting have a long tradition, and the line skáro á skíði ‘cut in wood’ in the Eddic poem Vǫluspá (stanza 20) probably refers to it. Johnsen (1987) takes the evidence of runic inscriptions as proof that merchants used runes on wood for their purposes, whereas texts in the Latin alphabet on parchment were restricted to legal matters and the church. On the other hand, the number of inscriptions suggests that literacy – or “runacy” – was indeed widespread among merchants and city dwellers in general. As trade became more complex, one would expect that merchant literacy developed beyond the limits of runic writing. There are, however, few traces of material written by merchants in the Norwegian sources. One would for instance expect the larger merchants to make accounts of incomes and expenses. There is no doubt that accounting was of growing importance in official administration, as we have many documents either asking for an account or stating that one has been duly made, although the actual accounting documents have rarely been preserved. The same applies to writing in connection with foreign trade: there is the occasional receipt, but not much more. It is telling of the source situation that most information on Norwegian trade and other connections with the British Isles during the Middle Ages is found in English Patent Rolls and similar sources (published in vols. XIX–XX of DN). Foreign trade increasingly came under the control of the Hanseatic League and to a lesser degree British merchants, leaving Norwegians to conduct only petty domestic trade, where there was probably less need for writing in any case. Norwegian merchants sailing to England mostly disappear from the English records around the mid-fourteenth century, albeit with late exceptions such as Oluf Henriksson from Tønsberg, who sailed to Hull in 1392 (DN XIX, no. 617). The source situation changes around 1500. Most of the documents printed in DN dealing with trade belong to the Munich Collection, the archives of King Christian II (reigned 1513–23) and Archbishop Olav Engelbrektsson (1523–37). Both fled to the Low Countries because of political conflicts and brought their archives with them; the two archives were merged and later surfaced in Bavaria and hence got their present name. Because of this special situation, drafts, notes, accounts, transcripts etc. have all been preserved until today, whereas in other cases documents of limited interest for posterity have been lost in the course of

92 

 Ivar Berg

time. Archival practices must be at least partly responsible for the lack of written material “related to the activities carried out by merchants and tradesmen” from the Late Middle Ages (Hagland 2011: 29). A few medieval cadastres are preserved, registers of land and revenue. This text type, though not necessarily dealing with trade as such, shows similar linguistic traits as other written material serving an administrative or economical purpose: stereotypical form (lists) and elements of language mixing not found in letters or official writing. Such texts will thus be briefly mentioned below.

3 Language choice in international c­ ommunication A famous passage from the thirteenth-century Konungs skuggsjá [The King’s Mirror], an educational book in the European specula tradition, states that a merchant should “learn all languages, and first of all Latin and French, because those tongues are most widely known”.1 During the High Middle Ages, Latin also served as the administrative language of most European states as well as the language of the Catholic Church. From the Late Middle Ages the vernaculars rose to prominence as suitable international means of communication, although Latin retained its position in some domains (see Burke 2004, especially chapters 2 and 3, for a discussion of the relationship between Latin and vernaculars in Early Modern Europe). The Protestant Reformation even had vernaculars replace Latin as the language of the church in Northern Europe. As Latin lost its position as an international lingua franca, those engaged in international trade had to develop some sort of bilingualism to be able to conduct their business.

3.1 Communication and (receptive) bilingualism It is often assumed that Scandinavians and (Low) Germans communicated through what was essentially a form of receptive bilingualism (or semi-communication): each interlocutor spoke his own language, probably with some accommodation, and the other understood it – much the same way as Scandinavians

1 My translation from the original Old Norse: næmðu allar mallyzkur en alra hælzt latinu oc valsku. þviat þær tungur ganga viðazt (Holm-Olsen 1970: 19).



Business writing in early sixteenth-century Norway 

 93

communicate today (Braunmüller 2007). On the basis of this, it has been argued that Scandinavian and Low German were dialects rather than languages, i. e. mutually intelligible; nonetheless, I have previously (Berg 2016) argued that semi-communication was no automatic process, as there are several written statements from the early sixteenth century pointing to lack of understanding. It is also very hard to keep this phenomenon apart from bilingualism in the written records. Irrespective of the exact nature of the language contact, the preserved texts show at least that Low German was read and understood among Scandinavians. In any case, no such receptive bilingualism was possible with the British. One of the parties would have had to learn the language of the other, or gain access to the language through interpreters. The written sources give evidence of both practices. A charter issued in Marstrand (then southern Norway, present-day Sweden) in 1486 tells of a raiding English ship that had an interpreter who spoke ‘both Norwegian and German’ (han talede bode norske och tyske) in dealings with the townsmen (DN V, no. 930).2 A trader who may have been bilingual is Johnne Crethonne Scotts man serwand to ye archbischop of Noroway, who is mentioned in 1528 (DN XI, no. 487) and himself issued a letter in Scandinavian at the same time (DN XI, no. 486). The language, either his own or the work of a secretary, shows both ­Norwegian and Swedish traits, which may be expected of a Scotsman who had learned ­Scandinavian to do business. He was quite likely the man identified as Jon Skot in two charters from 1529 (DN VIII, nos. 589, 600), both issued in Bergen by a merchant from Newcastle, and it is a fair assumption that John had acted as i­ nterpreter. An example of a multilingual German trader is Wilhelm Frank. He is mentioned in at least 23 documents from 1523–1530, first in a pass issued by the Elector of Brandenburg to ‘our servant’ (vnsern diener) Wilhelm Frank in High German (DN XIII, no. 199). During the following years he was in the service of the exiled King Christian II of Denmark–Norway, and issued receipts in Low and  High  German as well as Danish, and wrote letters in Danish and High German (e. g. DN IX, no. 559; X, nos. 501, 615). His Danish is peculiar, even considering the possible orthographic variation of the era, which most likely means that it was a second language. However, such cases of attested multilingualism are rare (cf. Berg 2016).

2 “German” was used to refer to all varieties of High and Low German, including Dutch, during the Middle Ages. I use “Low German” as a cover term in the following, including texts with Dutch dialect features.

94 

 Ivar Berg

3.2 Different languages in use The politically unstable situation of the late 1520s offers a case study of interaction between different languages in Northern Europe. The archbishop of Nidaros (present-day Trondheim), as much concerned with practical and economic matters as with the afterlife, sent out armed ships to protect trade and seafarers in the North Sea in April 1527. He issued a charter in Scandinavian stating their mission and asking people to aid them; there was also a parallel Latin version (DN I, no. 1072). Whether the captains acted independently or on orders, they appear to have been mere privateers and ostensibly attacked the wrong ships. This led to a lengthy quarrel, and several Scottish, English, and Dutch merchants came to Norway to seek compensation for their stolen property. In this connection charters were issued in various languages, both during the negotiations and as proof of the final agreements for economic compensation. In addition to documents in Scandinavian, Low German, and Latin, there is also an interesting example of a trade contract in Scots.

3.2.1 Latin In late July 1527 King James V of Scotland sent an official complaint concerning the capture of Scottish ships to the Dano-Norwegian king’s commander at Bergen castle (DN VIII, no. 561). The letter was written in Latin except for the signature ‘James’, whereas the king is called ‘Jacobus’ in the main text. King James sent another complaint as well, but only the one that was forwarded from Bergen to Trondheim is preserved. The commander at Bergen castle returned an apology to Edinburgh in late September, also in Latin, according to a copy sent to Trondheim (DN VIII, no. 565). The official high-level communication was thus conducted in Latin. In the summer of 1528 two envoys from Scottish merchants in Dysart came to Norway with a letter of introduction in Latin (DN VIII, no. 573) from their superiors. The following year an Englishman, Roger Dischaunt, arrived on behalf of merchants in Newcastle, bringing with him an introductory letter in Latin (DN VIII, no. 590). Both these Latin letters are marked as ‘erroneous’ (feilfuldt) by the editors of Diplomatarium Norvegicum, indicating that the knowledge of Latin among merchants was not the best. Roger’s introductory letter was sent to the archbishop along with a letter in Scandinavian issued by Roger, presumably written on his behalf by a scribe at the archbishop’s residence in Bergen. We may also note in passing that ships were assumed to carry some kind of pass or



Business writing in early sixteenth-century Norway 

 95

introductory letter; medieval sources tell of foreign ships being asked to produce letters stating their origin and mission upon arrival (see e. g. DN V, no. 930, and IX, no. 415, Marstrand 1486 and 1493). The merchant David Falkoner of Leith sent a formal letter in Latin (signed Dauid Falkner / wyt my hand in the vernacular) about the ongoing negotiations to the archbishop in 1529 (DN VIII, no. 591); Falkoner is also known to have used French in his trade (DN XIII, no. 481, a pass for a ship he sent to Denmark in 1528). The communication across the North Sea is the only known use of Latin in this case. When it actually came down to doing business and documenting it, the vernaculars were used.

3.2.2 Scots Following up the exchange of letters the previous year, two Scotsmen came to Trondheim in March 1528: the captain of the captured ship Peter and a scribe; the ship was owned by said David Falkoner. The arrangement made was that the archbishop bought, or rather paid for what his men had already taken, both the ship and its cargo. This deal is particularly interesting because the statement of the transaction was written down in both Scots (DN VIII, no. 569) and Scandinavian (DN VIII, no. 570), both versions issued 10 March 1528. The Scots version is signed by Captain Thomas Gardner and Thomas Huchesson, and was surely written by the latter – he is called Thomas Hughon scriffwer ‘scribe’ in another charter from the same year (DN VIII, no. 580; notice the variable spelling of his surname). Both charters have the same content, yet the Scandinavian version follows the usual style for such charters in Norway at the time, whereas the Scots version is slightly different. It thus seems that neither was a translation of the other, but that both were written independently at the same time. The independence of each text is shown by the Norwegianised name Haakenson in the Scandinavian version (with a Norwegian intervocalic ‹k› where Danish would have ‹g›; the name is written Haagenson in DN VIII, no. 571). Later in March, the same two Scotsmen issued a receipt (DN VIII, no. 571) for part of the payment for the ship’s cargo, including a remark that the archbishop had promised in writing to pay the rest in Bergen. The receipt is in Scandinavian, except for a signature similar to the one in the two previous charters: Thomes Houchesonn wytht my hand at/on ye pen (with some minor orthographic variation), all clearly in the same handwriting. This practice of issuing receipts in Scandinavian, possibly with a signature manu propria, is the usual one, as we see from subsequent dealings in Bergen.

96 

 Ivar Berg

3.2.3 Scandinavian During the summer of 1528 several Scotsmen got their business settled in Bergen, and a series of receipts were issued (DN VIII, nos. 574–576, 578, 580–581). All of these were written in Scandinavian, despite being nominally issued by Scotsmen who did not sign, but set their marks on several of them. Most of these texts are of a modern form, where the one issuing the charter states that he has received the agreed amount. Only the closing of this particular case, a charter issued in Bergen in 1529 (DN VIII, no. 611), is written in the traditional Norwegian style with several witnesses attesting that such and such agreements were made and duly carried out. Thus we also see a transition from the older style of witnesses attesting an agreement to the modern style of issuing a personal receipt; the written text becomes the constitutive legal act, rather than just proving a previously reached oral agreement. Although written in Scandinavian, DN VIII 611 has the phrase myn hand ath pennen ‘my hand at the pen’ that is not used in other Scandinavian charters; this is clearly a calque of the Scots expression found in the signatures mentioned above (DN VIII, nos. 569–571) and a witness of foreign influence. This document also states that setting a mark is køpmens sydhwane ‘merchants’ custom’. Roger Dischaunt from Newcastle, who came to Bergen in 1529, also received the money he wanted and issued a receipt in Scandinavian in Bergen (DN VIII, no. 600), with a Latin signature: per me Rogerum Dychaunt. / manu mea propria ‘by me, Roger Dischaunt, with my own hand’. He is the only one of the British merchants who actually came to Norway that shows any sign of Latin competence. In the other charters under discussion, signatures are in the vernacular; even the Scottish king signed as James, not Jacobus.

3.2.4 Low German The man sent out by the archbishop in 1527 was named Paall Jonson (DN I, no. 1072). In two undated documents (DN X, nos. 540–541, probably from the summer of 1527) he admits to having property belonging to two Dutch merchants and promises to compensate for it. Based on the language of these charters he was probably of Dutch origin. One of these merchants was Claes Oedszon from Amsterdam, who came to Bergen in 1528 to get his promised compensation. He made an agreement with the archbishop’s representative in Bergen. The charter stating the deal (DN VIII, no. 579) was probably written by a scribe there, as it is in Scandinavian. The



Business writing in early sixteenth-century Norway 

 97

merchant, however, signed in Dutch: Jtem een kennys daer waerheyt soe hebbe ic Claes Oedszon van Amsterdam myn eeyghen hant hyer onder gheset myn merck ‘And to witness the truth have I, Claes Oedszon from Amsterdam, with my own hand set my mark here below’. The skipper Allert Tomaszon also signed ­(similarly phrased) and put his mark on the charter.3 The deal was settled a few days later, and Claes Dodessen issued a charter stating that he was fully satisfied (DN VIII, no. 609; despite the spelling it must be the same person as the mark according to the editors is the same as in DN VIII, no. 579). This time he probably did the writing himself, as the text is in Dutch, and this was ostensibly legally valid. For comparison I shall mention an unrelated deal made in Trondheim in March 1532. A Dutch seafarer promised to take a ship to the Netherlands and back to Trondheim on behalf of the archbishop (DN VII, no. 692), and stated that he had received some goods and money (DN VII, nos. 693–694). The contract is in Low German, the two receipts in Dutch. The hand in 693–694 also signed Jan Heynryck zon in 692; this handwriting is markedly different from the hand in 692, which is quite similar to that of Scandinavian texts from the archbishop’s chancery during this period. Both linguistic differences and palaeographic analysis suggest that 693–694 were written by the skipper himself, whereas 692 was written by a local scribe (Berg 2013: 183). The most plausible explanation for the language choice here is that the skipper could not understand Scandinavian, and the Norwegians were able to write Low German in dealings with Dutch merchants. The scribe himself may of course have been German; we know for certain of one Henrick tysk ‘German’ working as a scribe in Trondheim a few years later (Seip 1936: 67, 88, 114; cf. DN XII, no. 571). A contemporary note (DN VIII, no. 678) lists the goods given to the skipper; the list was made for internal administrative purposes and written in Scandinavian. It appears that Low German was a marked choice only used in dealings with foreigners.

3.3 Patterns of language choice As we have seen, only high-level international communication was conducted in Latin. Its use at a lower level is restricted to the merchant David Falkoner and the introductory letters some of the merchants brought with them on their mission to Norway. However, according to the editors of the Norwegian diplomatarium, the 3 This is a common practice also found in other documents, see e. g. DN XXII, no. 276, Bergen 1535, in Scandinavian with signatures by Dutch merchants.

98 

 Ivar Berg

language of the latter was “erroneous”, pointing to only mediocre knowledge of Latin. Recall also that the archbishop issued a similar letter to his skipper in two versions, Scandinavian and Latin, although the captain himself appears to have used Dutch in two preserved documents. Latin seems to have been the preferred choice for such documents. However, none of the charters stating the final agreements are in Latin. The contracts between the British and the archbishop (or his representatives), as well as the receipts issued by the former, are mostly in Scandinavian. We must assume that they were written by scribes employed by the archbishop in Bergen. In the 1530s, when we have more information due to preserved salary lists for the archdiocese (published in Seip 1936), it appears that approximately five scribes were stationed there. In these arrangements the interlocutors must either have used some common language or alternatively had interpreters. As the knowledge of Latin among the merchants seems to have been rather basic, it is doubtful whether that was the language used, and I assume that the negotiations were conducted through interpreters. On the single occasion where a charter was issued in Scots, we see that a parallel version in Scandinavian was needed. This is an exception to the usual practice, and we can only speculate about the reasons. It was an important deal, as the item at hand was an entire ship, and it was also the first agreement made in these negotiations, which perhaps made it especially important for the Scotsmen to have a contract in an intelligible language. It was ostensibly no alternative to use Latin as a shared language for this practical purpose. The overall picture we get is that the importance of Latin was clearly in decline  by the sixteenth century. In everyday economic matters Latin does not appear to have been a possible choice, and only the vernaculars are used in that  domain. The examples here illustrate the decline of Latin as an international  language of communication, being reduced to a language of learning exclusively; for the clergy and in scholarship Latin continued to be the common language. But it was never used for simple business deals, not even when it might have been a neutral choice as an alternative to parallel versions in different vernaculars. There is a striking contrast between the Scots example of parallel versions and a few contracts and receipts in Low German without translations. Because of the familiarity with that language, Low German could be used in mercantile matters and must thus have been considered valid for legal purposes. As mentioned, the archbishop had a German scribe in his service in the 1530s, and a list of staff at the royal castle in Bergen around 1531 (DN XIII, no. 582) includes both a scribe and specifically a German scribe (Tysk schriffuere).



Business writing in early sixteenth-century Norway 

 99

4 Language mixing in business writing The more official documents, be they letters between merchants or documents closing a transaction, are written within the normal levels of variation in premodern texts. There is not much – if any – language mixing. The only significant exceptions are signatures, for reasons easy to understand. We will now look at more informal documents, lists and accounts done only for administrative purposes, which have a much more varied linguistic mix-up. It seems to be a general tendency that code-switching is mainly found in unofficial notes and administrative writings (Schendl 2012: 527). Because such notes had no lasting value, they have rarely been preserved; in Norway it is also possible that the use of runes on wood for such practical purposes may at least partly explain the absence of such texts (cf. § 2 above). As mentioned initially, cadastres have some of the same characteristics and will also be mentioned here. The main language of these texts is Scandinavian, with substantial amounts of Latin and Low German inserted.

4.1 Merchant notes One of the oldest known merchant notes in Norwegian sources is a list of imported goods and how they were paid in money and natural goods, written during 1491– 96 (DN XXI, no. 671). The scribe is anonymous, but he dealt with a merchant from Lübeck called Bertolt (also mentioned in the contemporary DN IX, no. 428). The text is mainly in Norwegian and differs from Danish in several respects: the use of ‹p, t, k› where Danish had ‹b, d, g›, word forms like miol ‘flour’ (Da. miel or meel), honom ‘him’ (Da. hannem), vikunne ‘the week.dat’ (Da. ugen), diphthongs (e. g. ein ‘a/one’ for Da. en), unstressed a (e. g. saltad ‘salted’ for Da. saltet). As Danish had an ever increasing influence on the written language in Norway during the Late Middle Ages, the abundance of Norwegian dialect features in this text is noteworthy, although not uncommon in texts of this type (cf. § 4.3). The scribe uses quite a lot of Latin, mostly in forms of date referring to the closest feast day as was customary, but also more freely in phrases like hiis annis ‘in these years’. There are also some Latin entries in the list not connected to time or date, and it all flows freely in the text, as shown in (1), with Latin parts in boldface:4 4 Expanded abbreviations are in italics. The examples are given with fairly verbatim translations as a compromise between glossing and translation; untranslatable units of measurement are in italics.

100  (1)

 Ivar Berg

a. jtem xvi marker lagde han fram domino johanni for brefuenn [item 16 mark laid he forth on Saint John’s day for the letters] b. … førde berulf in xv hudelag tercia die pasce [brought Berulf in 15 hides on the third day of Easter] c. dominus naruo commodauit mek ix marker [sir Narve gave me 9 mark]

(1a) has correct case endings in the ablative (domino) and the genitive (johanni), which by most definitions would be considered code-switching. It is, however, possible that such phrases were learnt as lexical units in exactly this form – as they only occur in dating, these are the only relevant case forms. The same goes for (1b). (1c) is particularly interesting, as the Latin is not connected to time or date, but includes the verb with a switch before the indirect object. Although the sentence starts in Latin, the word order (verb second) indicates that Scandinavian is the matrix language; cf. Wright (1999: 109–110) for an example where most lexical items are Latin, but the underlying word order English. There are also other examples of Latin outside temporal expressions, such as the phrases given in (2): (2) a. solui ei, reddidi ei, misi ei [I paid him, I repaid him, I sent him] b. sic credo, dubito [I think that, I doubt] The expressions in (2a) were probably technical terms in accounting; nonetheless, it is harder to explain e. g. dubito (2b) in that way. Another word which is rare in normal prose, but also occur in the account discussed below, is the Low German deker ‘an amount of ten’. The word vordor (pl.) used for an amount of fish in this document is known from later dialects, yet apparently a hapax legomenon in the medieval sources. Similar use of Latin technical terms in accounting is also found in some of the other extant business notes (e. g. DN VIII, nos. 582–585, c. 1530). A few other phrases are given as (3). (3) a. summa (summarum/lateris), in toto, minus, soluit [sum (of sums/of page), in total, less, paid.3sg] All the mentioned words could be labelled loanwords; nevertheless, instances such as number agreement with facit.3sg ∼ faciunt.3pl ‘make’ (used when adding sums) indicate real Latin knowledge.



Business writing in early sixteenth-century Norway 

 101

The Narve mentioned several times in the text, e. g. in (1c), was probably a canon regular in Oslo: either Narve Thorersson, canon in Hamar 1487 (DN V, no. 931) and archpriest in Oslo 1498 (DN I, no. 993), or Narve Jonsson, canon in Oslo 1489 (DN XVII, no. 751). This may indicate that the scribe was also a cleric or connected to a clerical institution, which fits nicely with the extensive use of Latin. This is no surprise; the church was also concerned with earthly matters. Most of the tithe and taxes that the church collected were in natural goods which had to be traded, and many clerics took advantage of their tax exemption and participated eagerly in trade (Kolsrud 1958: 320). A note from a royal servant in Bergen complains strongly about this, as it deprived the king of tax income (DN XIII, no. 183, undated c. 1520).

4.2 Accounts The involvement of the church in trade is also evident from Archbishop Olav Engelbrektsson’s account books, several manuscripts from 1532–38 published by Seip (1936).5 They are mostly written in Scandinavian (Danish with Norwegian interference, as was usual at the time), with some Latin. Some of the scribes also use Low German and must have been bilingual, at least to some degree. Salary lists, both for regular employees and Dutch mariners hired 1536–37, make up much of the material, but the most interesting manuscript in our context is the 1536 account of the archbishop’s estate in Bergen (NRA dipl. München papir, no. 4297; Seip 1936: 131–152). Already the heading of this manuscript is bilingual: Anno Christi mdxxxvj / Jnt jaar xxxvj. Pages 17–31 in the manuscript (134–40 in the edition) deal specifically with various merchants, sorted into goods in and goods out (not incomes and expenses as we are used to today), and the same names mostly appear in both sections. The miscellany of the following few pages also mentions dealings with merchants. Most of the text is in Scandinavian with the occasional Latin phrase, but there is also some Low German tossed into the linguistic mix. The entry for each merchant ends with a phrase that clearly derives from Low German, yet often shows code-switching. This has been overlooked in previous research on the contact between Scandinavian and Low German, and for instance Jahr (1999: 134) claims that “no examples of code-switching have been found to date”. Consider the forms in (4), with Scandinavian in boldface. 5 This is a diplomatic edition where expanded abbreviations are in italics, if expanded at all. References follow manuscript pagination as given in the edition.

102 

 Ivar Berg

(4) a. Jtem gereckendt all ding dodt met Kleine Pauell dat myn heer blyfft em skyldiig xliij bg. g. (pag. 26) [item accounted all things clear with Kleine Paul that my lord owes him 43 bergergulden] b. alting klar geregnet met Hans Køne dett min herre bliffuer honum skuldich effwen jc smalt woge fisk (pag. 19) [all things clear accounted with Hans Køne that my lord owes him exactly a small hundred voge fish] c. Jtem gerekendt medt Roleff Røwekamp all ding dodtt saa dat he blyfft minnum here skyldug viij voger fiisk oc ij ste miøll (pag. 22) [item accounted with Rolf Røvekamp all things clear so that he owes my lord 8 voge fish and 2 stykke flour] (4a) is in Low German except for the spelling skyldiig (‘indebted’), and the number and currency are inconclusive (this applies throughout). (4b) switches to Scandinavian from min herre, except for LG effwen and the spelling of skuldich, an apparent hybrid of LG schuldich and Sc. skyldug/-ig. Skuldich is this scribe’s ­preferred form with a total of four tokens, although he also has one token each of the normal LG and Sc. forms. The scribe in (4a) has only Scandinavian spellings of  this word, also in this example otherwise completely in LG. (4c) has a clear switch to Scandinavian from minnum.6 The form geregnet in (4b) is also a hybrid,  with Danish ‹g› in regnet despite the LG prefix ge-. Such intermediate orthographic forms show clearly that the scribes were under influence from conflicting codes. The regular use of item and noch ‘additionally’ is typical of such accounts. This kind of code-switching is usually termed “tag insertion”. Item is found in all kinds of lists, and also as a paragraph marker in letters, whereas Low German noch is specifically connected to the manuscripts in Seip (1936) and a few similar documents (e. g. DN XII, no. 570, written by one of the scribes in the account books). The same goes for the Low German entfangen ‘received’, used as a technical term. Many of the Latin terms mentioned in § 4.1 are also found in account manuscripts, e. g. summa lateris/in toto, or others in the same vein, e. g. restat ‘remains’. Similar texts also have longer stretches of Latin, as shown in (5a), taken from another manuscript in the same edition (Seip 1936: 19), and (5b), taken from a cadastre (Jørgensen 1997: 47), with Latin in boldface.

6 The spelling ‹saa› in the subjunction saa dat ‘so that’ must be due to Scandinavian influence, since Low German as far as I am aware only had ‹so›; in Scandinavian ‹aa› was used for /ɔ/.



Business writing in early sixteenth-century Norway 

 103

(5) a. Fic pro anno preterito et anno futuro ein skød om iij lod [got for last year and next year a spoon of 3 lod] b. Odals breffuet fek oss Birgita vxur seu relicta viri prefa7 [the allodial letter gave us Birgita, wife or widow of said man] The Latin of accounts is usually restricted to temporal expressions and register-specific technical terms, whereas cadastres like the one quoted in (5b) in addition to this also use the language more generally. Comments about ownership such as this one are frequently written in Latin (more examples in Berg 2013: 172–174). The account manuscript and others that deal with imported goods have many lexical borrowings from Low German, particularly connected to the semantic fields of weapons and naval terms, clothing and fabric, and food and beverages (Berg 2013: 124–125). Many of these have remained in the language as loanwords, whereas others have disappeared. Some of the latter instances may be considered code-switching, depending on how broad one’s notion of the concept is. Since both languages had lost most of their case morphology by this time, inflectional endings cannot serve as a means of identification of single forms as either borrowing or code-switching. Whereas abbreviations made it possible to render a written form in both English and Latin in English mixed-language business writing (Wright 2011: 203–204), this was no issue for the simultaneous use of Scandinavian and Low German lexical items, a fact that probably made borrowing easier.

4.3 Norwegian dialectal forms As mentioned in the introduction, the written language in Norway was Danish from around 1500. However, many texts that can loosely be described as “administrative documents”, including accounts such as those discussed above, have more clearly Norwegian features than what is usually found in more formal texts (Berg 2013: 116–124). The document from Section 4.1 may again serve as an example. It was written in the part of Norway closest to Denmark, where the dialect is most similar to Danish, yet has many unmistakably Norwegian forms. It is also notable that new dialect forms that emerged by language change during the Late Middle Ages were in many cases hidden by the written tradition as long as one existed. After the language shift to Danish, however, such forms may surface as dialectal forms in the written language. 7  marks the editor’s conjecture.

104 

 Ivar Berg

There are several possible explanations for this pattern. Administrative texts use a vocabulary in many respects different from that of official letters and proclamations. This would leave Norwegian scribes with no other model for such words than the pronunciation in their own dialect. However, I also think that less rigid adherence to norms of writing and in general less careful language is a property of administrative writing, and this facilitates code-mixing not found in other situations.

4.4 Patterns of language mixing The patterns of code-mixing in Norwegian texts closely mirror those described by Kopaczyk (2013) in records of Scottish expatriates living and trading in seventeenth-century Poland. She divides the code-switches to Latin into three types: dates, text cohesion, and register-specific expressions.8 Latin dating was shown in (1) above, and (5a) showed a more elaborate Latin temporal expression; the category that Kopaczyk terms “text cohesion” is best represented by the ubiquitous item, and Low German noch also has a similar function in some texts; both (2) and (3) above give examples of register-specific expressions. The Low German in Norwegian accounts may be compared to the Polish of Kopaczyk’s Scottish merchants. She found code-switching to Polish in names, objects of trade, and currency. The last category may be extended to units of measurement more generally, and so conforms to e. g. deker mentioned above (§ 4.1). In the account manuscript (§ 4.2) we find e. g. the Low German schôf ‘bundle’. This puts the linguistic practices of Norwegian business writing into a wider European context. The patterns described here may be compared to the relationship between Latin and the vernacular in official charters and letters (Berg 2013: 164–177). Clerics in particular wrote certain formulaic parts of their vernacular letters in Latin: the name of sender and addressee, the initial greeting (often followed by another greeting in the vernacular), and the concluding information on where and when the document was written. Occurring within the same stretch of text, this may be termed inter-sentential code-switching, yet is very regular and not as interesting to the linguist as the intra-sentential code-switching of the administrative writings discussed above. To the degree that intra-sentential code-switching between Scandinavian and Latin does occur, it is found in letters of a more private character and not in official documents (Berg 2013: 171–172). 8 I believe she uses the term “code-switching” too freely, and for instance anno should probably be classified as a borrowing.



Business writing in early sixteenth-century Norway 

 105

The significant difference in text type thus seems to be between official letters and charters on the one hand, and unofficial notes and administrative texts, including accounts and cadastres, on the other. The latter conforms, in all their variety, to some genre-specific norms that allow code-mixing not found in more formal texts.

5 Conclusion The evidence of the scant Norwegian sources does allow some general conclusions. First and foremost it appears that official letters and charters generally stick to one language. These may be in Latin for international communication, both in letters such as the one from the merchant David Falkoner of Leith to the archbishop and between King James V of Scotland and the commander at Bergen castle, and in introductory letters or passes, such as the one of Paal Jonson as well as those brought to Norway by British merchants. However, all purely economic matters (contracts, receipts) were written in a vernacular; it was even preferable to make different versions in Scots and Scandinavian rather than using Latin as a common language. And we have seen that financial deals could be concluded by written agreements in Low German, and that Low German receipts were similarly accepted and ostensibly legally valid. There is no code-mixing in these cases, except that signatures for obvious reasons may be in a language different from the main text. On the other hand, in more unofficial texts like administrative notes, accounts and cadastres, we do find extensive code-mixing – even if we disregard loanwords, which abound in these texts, as many of the Low German loans were connected to trade and new goods and technologies. The many Latin words can be broadly divided into two categories – expressions of time and date and genre-specific terms in accounting. Nonetheless, there are also phrases that hardly fit under even such broad notions, e. g. sic credo and dubito as comments to entries in the merchant note discussed in § 4.1. Examples of Latin additions or comments are even more plentiful in cadastres. The account for the archbishop’s estate in Bergen discussed in § 4.2 showed similar use of Latin, and also included Low German entries. The phrase used for closing the entry of that year with individual merchants comes in a wide range of mixed linguistic forms, including code-switching between Low German and Scandinavian. A last thing to note concerning the linguistic form of such texts is the occurrence of Norwegian dialectal forms after the language shift to Danish.

106 

 Ivar Berg

Code-mixing is usually assumed to be socially meaningful in some sense. However, the administrative texts under discussion have no clear sender or addressee, and it is thus hard to imagine a communicative function of the code-­ mixing. On the contrary, it appears that the scribes avoided code-mixing altogether in formal writings, except for the strictly regulated use of Latin formulaic phrases. The explanation must then be that the code-mixed texts represent an unmarked, functional code for bilingual scribes. Laura Wright has discussed the mixed language of business writing in England (see Wright 2011 with references to previous studies) and claims that “[t]his business variety was a functional written code, and nobody’s mother tongue” (Wright 1999: 114). In my opinion, we are dealing with the same situation here, and scribes who were accustomed to dealing with texts in several languages exploited their whole linguistic repertoire when there were no restrictions on the output. As such, the written forms may not directly reflect spoken language; nonetheless, a large part of the Low German loanwords that entered Scandinavian during the Middle Ages are related to commerce and trade, and the people dealing with these activities were probably a major inlet for loanwords. The scribes’ multilingualism is evident from their ability to read and write documents in several different languages, and on another level by mixing the same languages in ways that make perfect sense in other documents. Both these manifestations of their linguistic competence were governed by genre-specific expectations, in accordance with more general European patterns.

References Berg, Ivar. 2013. Eit seinmellomalderleg skrivemiljø: Nidaros erkesete 1458–1537 [A late medieval scribal community: Nidaros archdiocese 1458–1537]. Trondheim: Norwegian University of Science and Technology dissertation. Berg, Ivar. 2016. A note on the relationship between Scandinavian and Low German. Journal of Historical Sociolinguistics 2(2). 189–210. Braunmüller, Kurt. 2007. Receptive multilingualism in Northern Europe in the Middle Ages. A description of a scenario. In Jan D. ten Thije & Ludger Zeevaert (eds.), Receptive multilingualism: Linguistic analyses, language policies, and didactic concepts, 25–47. Amsterdam: John Benjamins. Burke, Peter. 2004. Languages and communities in Early Modern Europe. Cambridge: Cambridge University Press. DN = Diplomatarium Norvegicum I–XXIII. 1847–2011. Christiania/Oslo. Online version: http:// www.dokpro.uio.no/dipl\_norv/diplom\_felt.html. Hagland, Jan Ragnar. 2011. Literacy and trade in late Medieval Norway. Journal of Northern Studies 5(1). 29–37.



Business writing in early sixteenth-century Norway 

 107

Holm-Olsen, Ludvig (ed.). 1970. Konungs skuggsiá. Utvalgte stykker [The king’s mirror. Selected parts]. Bergen: Universitetsforlaget. Jahr, Ernst Håkon. 1999. Sociolinguistics in historical language contact: The Scandinavian languages and Low German during the Hanseatic period. In Ernst Håkon Jahr (ed.), Language change. Advances in historical sociolinguistics, 119–139. Berlin & New York: Mouton de Gruyter. Johnsen, Ingrid Sanness. 1987. Die Runeninschriften über Handel und Verkehr aus Bergen. In Klaus Düwel, Herbert Jankuhn, Harald Siems & Dieter Timpe (eds.), Der Handel der Karolinger- und Wikingerzeit (Untersuchungen zu Handel und Verkehr der vor- und frühgeschichtlichen Zeit in Mittel- und Nordeuropa 4), 716–744. Göttingen: Vandenhoeck & Ruprecht. Jørgensen, Jon Gunnar (ed.). 1997. Aslak Bolts jordebok [Aslak Bolt’s cadastre]. Oslo: Riks­ arkivet. Kolsrud, Oluf. 1958. Noregs kyrkjesoga I: Millomalderen [Norway’s ecclesiastical history I: The Middle Ages]. Oslo: Aschehoug. Kopaczyk, Joanna. 2013. Code-switching in the records of a Scottish brotherhood in Early Modern Poland-Lithuania. Poznan Studies in Contemporary Linguistics 49(3). 281–319. Nesse, Agnete. 2012. Norwegian and German in Bergen. In Lennart Elmevik & Ernst Håkon Jahr (eds.), Contact between Low German and Scandinavian in the Late Middle Ages. 25 Years of research, 75–94. Uppsala: Kungl. Gustav Adolfs Akademin för svensk folkkultur. Schendl, Herbert. 2012. Multilingualism, code-switching, and language contact in historical sociolinguistics. In Juan Manuel Hernández-Campoy & Juan Camilo Conde-Silvestre (eds.), The handbook of historical sociolinguistics, 520–533. Chichester: Wiley-Blackwell. Seip, Jens Arup (ed.). 1936. Olav Engelbriktssons rekneskapsbøker 1532–1538 [Olav Engelbrektsson’s account books] (Noregs Riksarkiv). Oslo: Dybwad. Wright, Laura. 1999. Mixed-language business writing: Five hundred years of code-switching. In Ernst Håkon Jahr (ed.), Language change. Advances in historical sociolinguistics, 99–117. Berlin & New York: Mouton de Gruyter. Wright, Laura. 2011. On variation in medieval mixed-language business writing. In Herbert Schendl & Laura Wright (eds.), Code-switching in Early English, 191–218. Berlin: Mouton de Gruyter.

Laura Wright

6 Kiss Me Quick: on the naming of commodities in Britain, 1650 to the First World War Abstract: This chapter looks at some historical product-names, positing that regardless of origin, they were words like any other and subject to the same kinds of phonological, morphological and semantic changes that affected the rest of the lexicon. I focus on four linguistic processes which show merchants’ innovations: the development of bound morphemes, conversion from one part of speech to another, semantic extension, and sociolinguistic innovation. My data comes from commodity-names retrieved from the British Newspapers 1600–1950 database, and two main areas of multilingual borrowing emerge: the names of commodities brought to Britain from abroad via trade, and names arriving in Britain as the result of defense policy when the Empire was threatened. A multilingual trading perspective reverses the traditional relegation of commodity-names to the sidelines of linguistic history, foregrounding instead everyday domestic conversations. This everyday lexicon directly reflected Britain’s global political and economic partnerships with speakers of other languages.

1 Introduction Textbooks on the history of the English language rarely devote much space to words which were the names of commercial products – too ephemeral, invented by traders and therefore somehow spurious, perhaps perceived as peripheral to the rest of the language which had evolved “naturally”. Not proper English, in short. This chapter refuses to recognize this distinction, positing that for the speaker, product-names were words like any other, needing to be discussed (in pre-supermarket days) between customer and shop-assistant. A multilingual trading perspective reverses the traditional relegation of commodity-names to the sidelines of linguistic history, foregrounding instead everyday personal, domestic experiences such as reading newspaper advertisements, walking along urban streets, looking at labels on shop-goods and talking to shop assistants.

Laura Wright, University of Cambridge DOI 10.1515/9781501503542-006



Kiss Me Quick: on the naming of commodities in Britain 

 109

This everyday lexicon directly reflected Britain’s global political and economic partnerships with speakers of other languages. Certain types of product advertised in British newspapers over the period 1650 to the First World War – including household items such as cloth, grocery, snuff, wine, card-games, tea – increasingly included foreign words in their advertising columns, many of which bore little relation to their phonological source. Since the British maritime trading empire began, British newspaper-readers were subject to innumerable borrowings. Commodity-names retrieved from the British Newspapers 1600–1950 database are found to have behaved like other parts of the lexicon, subject to reinterpretation and semantic development, and conversion from one part of speech to another. Words started life denoting one kind of commodity but then underwent semantic extension and became applied to others – depending on the decade, garibaldi was the name of a type of blouse (1860s), hat (1880s), fish (1880s), and also biscuit (1890s to the present).1 Two main areas of multilingual borrowing are considered here: the names of commodities brought to Britain from abroad via trade; and names arriving in Britain as the result of defense policy when the Empire was threatened (the latter was a particularly fruitful source of inspiration for traders). I focus on four linguistic processes which show merchants’ innovations: the development of bound morphemes, conversion from one part of speech to another, semantic extension, and sociolinguistic innovation. I make no pretence at exhaustion; these four processes have been chosen in order to show both unconscious developments, such as happened when merchants fitted unfamiliar foreign names into their phonological systems; and conscious developments, as when names were invented ab initio for commodities new to the market.

2 Bound morphemes The British Empire is more or less coterminous with the British East India Company, which began trading in 1600. Wright (forthcoming) considers seventeenth and early eighteenth century cloth terms imported by the Company, many of which were the names of villages in Bengal where the cloth was produced.

1 OED Garibaldi, n. All commodity names discussed hereafter have been retrieved from the British Newspapers 1600–1950 database unless specified otherwise. I especially thank Alex Bergs for comments on an earlier draft.

110 

 Laura Wright

A word-final cloth-name element which repeats in British newspaper advertisements is –pore: (1) Cargoe of the Colchester and the Fame from the Bay of Bengall, the Loyal Merchant from Bombay, and the Wentworth from China, arrived the 11th Inst. Viz. Atlasses 2576, Atcheenbannees 1196, Alibannees, 704. Aubrowahs 1942, Baftaes 414,. Ditto broad, 13600, ditto narrow 6286, ditto broad blew 3140, ditto Baguzzees 480, ditto Ponabagguzzees 1980, Brawles 240, ­Chellanumfaree 120, Cabooter Coupee 109, Chints 44271, Chinachurrees 437, Cossaes 9146, Conches 168, Chowcarea 2400, Choradaries 3196, ­Chowtars 925, Chucklaes 1436, Cuddumpholeys 537, Cuttanees 1514, Damasks 3418, ditto Birds-eyes 727, Deribands large 1300, ditto small 2800, Dimities 856, Doreaes 7950, Dungarees 4000, Elatches 7422, Flower’d Cloth 336, ­Geelongs 160, Ginghams coloured 1239, Guinea Stuffs 21160, Gorgorons 402, ­Gungallary 298, Gurrahs 6594, Humhums 1603, Hummerlees 717, Jamwars 861, Jamdannees 588, Long Cloth brown 1790, Luckhowries 646, Lungees 1655, ditto Herba, 1903, Mamoodies 480, Mahumudhiattees 2670, Mulmuls 18476. Musters 296. Neckcloths 1321. Nillaes 1134. Niccannees large 1600. Ditto small. 6720. Pallampores 1158. Paunches 6103. Ditto coloured 2452. Pelongs Nankeen 2972. Ditto Canton 2818. Peniascoes 1277. Pegue 822. Photaes 1216. Phooleys 94. Poses 1730. Rehings 1139. Romalls 6574. ­Rungebangulpoor 380. Sallampores brown 2020. Sailcloth brown 440. Sannoes 2289. Sattins strip’d 523. Seersuckers 2142. Seerhudcanns 105. Seerbands 211. Silks flower’d 63. Sovaguzzees 10824. Soosannees 25. Sooseys 7399. Sorts Nunsaree, &c. 565. Ditto Dungom 990. Taffaties plain 10130. Ditto Raw 556. Ditto Herba 1312. Ditto Chequer’d 196. Ditto strip’d 1977. Ditto Silver 344. Ditto Flower’d 230. Tanjeebs 12264. Tapsiels broad 800. Tannah Stuffs 144. Terrindams 756. Tunsooks 82 Pieces. London Post with Intelligence Foreign and Domestick, August 11, 1701 – August 13, 1701. Pallampores (line 14) and rungebangulpoor, sallampores (line 17) fit into the pattern of other East India Company cloth-terms ending in –pore/poor, such as callawapores, monepore, bettellees gunapore, chints birampore. Most of these contain the Sanskrit place-name element pur- ‘suburb, settlement’: rungebangulpoor is rang ‘muslin’ from Bhagalpur, sallampores were a type of cloth from Salampur, monepore was cloth from Manipur, bettellees gunapore were muslins



Kiss Me Quick: on the naming of commodities in Britain 

 111

from Gunupur and chints birampore were chintzes from Birampur.2 However, the word-final element of pallampore and callawapore is likely to be derived from Early Modern Persian pōš ‘cover’.3 These two names have been assimilated by analogy to the –pore group, not by speakers of the original donor languages, but by speakers foreign to the place of manufacture along the trade-route to Britain. Similarly, seersuckers, seerhudcanns, seerbands (lines 18–19) show a word-­initial cloth-name element seer-, fitting into a larger group of East India Company clothnames including seerbettees, seerhaudconnaes (possibly the same as seerhudcanns) seershauds, seerbafts, seerpaws. The predominant word-initial element is Persian sar, sir ‘head’, entering English through Urdu, prefixing the names of cloths that were used for turbans. However, seersucker contains Persian shīr o shakkar, ‘milk and sugar’, in reference to its stripes. Seersucker joined the seer- group by analogy, the meanings of ‘head-cloth’ and ‘milk and sugar’ being irrelevant to English-speaking merchants. Again, this development must have happened away from the place of manufacture, at a point where the original languages were no longer understood.4 Both –pore and seer- came to act as bound elements in English clothing vocabulary, but the very success of East India cloth in Britain caused an abrupt downfall. British cloth-manufacturers (including Huguenot weavers, the foreign traders of earlier generations) saw their livelihoods threatened, and petitioned parliament to grant them security. As a result, retailing East India cloth in Britain was outlawed, and by 1720 it was even illegal

2 See Wright (forthcoming) for discussion of these terms; “Rang is a muslin which resembles jhuna in its transparent gauze or net-like texture” Yule and Burnell (1886 sub Piece-Goods); OED Salempore, n. “of unascertained origin”, “A blue cotton cloth formerly made at Nellore in India, and largely exported to the West Indies, where it was the usual slave cloth”; OED beteela, n. ‘kind of muslin’, etymology obscure; OED chintz, n., from Hindi chint ‘variegated’. There are currently thirteen villages named Salampur in India, including two in West Bengal (http://villagesinindia.in/odisha/jagatsinghapur/tirtol/salampur.html); Bhagalpur was a cloth manufacturing centre in Bihar, Manipur is in eastern Bangladesh, Birampur is in western Bangladesh, and Gunupur is in Odisha. 3 OED palampore, n. “Hindi palaṅg-poś ‘bedspread, coverlet’, from palaṅg ‘bed’ from Sanskrit palyaṅka + early modern Persian pōš ‘cover’. The palampore forms perhaps arose as an inferred singular of perceived plural palampores, palempores, which might have been attempts to represent the Hindi pronunciation (with long o) of the singular, with voicing of the final consonant after Portuguese. The medial -m- is also likely to be after Portuguese pronunciation”. See Wright (forthcoming) for a discussion of callawapores (callaway poos (1678), callaway poose (1688)), where I identify the second element as Persian pōš ‘cover’ as above, but the first element is unknown – possibly a settlement name, or possibly after two East India Company employees named Calloway. 4 See OED seersucker, n.; OED seerpaw, n.; OED baft, n.; Yule and Burnell (1886 sub PieceGoods) seerbands, seerbetties. Seerhaudconnaes/seerhudcanns/seershauds remain unidentified. Hindi ser ‘weight measurement’ may perhaps also be relevant (OED seer n.2). Word-final is another spelling convention of East India cloth texts.

112 

 Laura Wright

to wear it.5 Cloth-terms such as the above, which had been established over a hundred years of trading, became largely abandoned. Moving to another set of commodity-names that was very visible in press advertisements from around 1750–1850: the newspaper-reading public would have repeatedly seen the names of fireworks during the summer firing seasons. Fireworkers published the names of their fireworks week in, week out, year after year, in the “Order of Firing” timetables that publicized their attractions at the 200-odd pleasure-gardens in and around London. Expertise in fireworking at this time came  from France, Switzerland and Italy, and French and Italian-speaking fireworkers dominated the London industry. The suffix –oon developed from the sixteenth to the eighteenth centuries in words borrowed from French with the suffix -on, especially where French had borrowed those words from Italian words with the suffix -one (OED -oon, suffix), and -oon became particularly productive in this semantic field. The prototype is likely to have been balloon, which is first attested in French in 1549 in the sense ‘hollow spherical firework’ (OED balloon n.): (2) Twelve Water Rockets. 3. Twelve Water Genouillieres. 4. Six Water Balloons. 5. Twelve Water Fountains (from M. Caillot’s Order of Firing at Ranelagh, Morning Chronicle, July 17, 1794) (3) A GRAND FIRE-WORK, Under the Direction of M. CAILLOT. ORDER OF FIRING. 1. A Salute of Twenty-One Maroons 2. Two dozen Water Rockets 3. Twelve Half-pound Sky-Rockets 4. An Air-Balloon illuminated 5. A large Verticle Wheel in brilliant Fire of different Colours, which ­terminate with Aigrettes in Chinese Fire (Advertisement for Ranelagh, The Times, June 2, 1791) (4) 2 Jerbs. 3 Tourbaloons. 1 Fourloon. 3 Rockets (Advertisement for Marylebone, Public Advertiser, July 16, 1770) Maroons first appeared in the London press as marons (e. g. London Gazetteer, April 15, 1749), derived from French marron ‘chestnut’ and its etymon Italian marrone (OED maroon, n.1 and adj.1 A. n.1 2.a. “A firework designed to make a single loud report like the noise of a cannon (often with a bright flash of light) 5 See O’Brien, Griffiths and Hunt (1991: 398); Lee-Whitman (1982: 39).



Kiss Me Quick: on the naming of commodities in Britain 

 113

… the firework makes the noise of a chestnut bursting in the fire”). Although the French names of fireworks could have been translated – maroons could have been called ‘chestnuts’, genouillieres could have been ‘knee-pieces’, in reference to how they were fired, aigrettes could have been ‘spangles’, and jerbs could have been ‘wheatsheaves’ – the French and Italian names were kept in London advertisements, presumably to add foreign allure. A tourbillon (French for ‘whirlwind’) was a kind of firework which spun as it rose, and the second and third syllables, opaque in meaning to English fireworkers, were interpreted as containing the element balloon (OED tourbillion, || tourbillon 3.), although a tourbillon was not a balloon type of firework. The –oon suffix was also extended to fourloon, a kind of firework which has at least eighteen highly variant spellings in the London press (Wright 2011b: 130), probably ultimately from the Italian verb frullare ‘to whirr’.6 This is a conflation of something that was productive elsewhere in English at the time (Italian –one and Spanish –ón > French > –on > English –oon), together with a perception that the -oon morpheme was suitable for that semantic field, just as –pore and seer- had become perceived as fitting components of the East India cloth-naming word-stock.7 The process would have occurred within the speech of industry-insiders, the merchants who dealt in cloth and the fireworkers who made and sold fireworks, rather than that of the general public. Similarly, the process known as Hobson-Jobsonizing (turning a foreign phonological string into whatever word/s it approximates in English, however meaningless the result) is something that would have happened in the speech of English fireworking apprentices, on hearing technical terms from their French and Italian masters. The French firework known originally as an aigrette ‘spangle’ became anglicized as air greets: 6 wheels of frueli, friulonies, friuloni, furiloni wheel, fruilioni, fruiloni wheels, furilloni wheels, flurious wheels, flurioni wheels, flurione wheels, thurioni wheels, fourloon, furilonii wheels, fourlony wheels, furlony wheels, fourloney wheels, forlony, frailona wheel; provenances in Wright (2011b: 129–30). 7 The spelling of the suffix in maroon, air balloon, water balloon, tourbaloon/tour balloon, fourloon had settled down to (predominantly) by the second half of the eighteenth century, from variants . It is complicated by multiple input: some words in –oon are borrowings into French from Spanish and Italian augmentatives (balloon, poltroon, cartoon, festoon); some seem to come direct from Spanish (picaroon); some are native French (harpoon), some are French borrowings (dragoon, shalloon); and some were formed in English (spittoon). To cloud matters further there were two opposing semantic thrusts: the Spanish and Italian augmentative (mantoon ‘large cloak’, doubloon ‘coin double the value of a pistole’) versus the French diminutive –on (mandilion ‘type of coat’, marmiton ‘type of minion’, minion). Pronunciation is unclear: OED –oon, suffix: “French -on /ɔ̃/ seems to have been identified by Englishspeakers with the close rounded back vowel /uː/ in the early modern period (and as late as the early 18th cent. …); but the considerable variation in spelling of the element between -oon, -on, -one, and even -oun suggests that other pronunciations also occurred”.

114 

 Laura Wright

(5) Two Balloon Wheels illumined and decorated with Pots d’air Greets. (Advertisement for Ranelagh, Public Ledger, June 30, 1766) I have counted twenty-four spelling variants between 1747 and 1798 in London advertisements for this particular firework, showing the process of oral transmission from French-speaking master to English-speaking assistant, whether fireworker or printer.8 The general public could not have been expected to decode the original meaning of callowapore, thurioni wheel or pow de grate, or even to have reliably related them to other visual manifestations of the same commodity, such as calloway poes, frailona wheel or pots daignettes. This kind of variation is the result of oral transmission without translation, where someone along the trading chain did not understand the term in the original language. Meaning, however, was less important when it came to point of sale than other, more valuable, social, information conveyed by foreign-looking words. The overall meaning (a type of cloth, a type of firework) could not be mistaken in the context of a press advertisement, and the names’ very impenetrability conferred an exotic appeal. How new, how modern, how very à la mode, to go dressed in atcheenbannee to view the latest flurious wheels in London in 1701. The selling-point was a sense of up-to-dateness and exclusivity – which novelty and exclusivity would have been reduced had the customer been able to translate.

3 Conversion I move now to personal names, and the nineteenth century. The nineteenth century was a time of much innovation in the field of applied chemistry, resulting in new substances marketed as drugs, dyes, perfumes and disinfectants. A common naming procedure for new products was to add suffixes to the name of the ­inventor in order to give nouns, verbs, modifiers and verbal nouns. For example, the words kyanised/kyanising and burnettised/burnettising, ‘wood that has been/is being treated with patented preservative’, invented by John Howard Kyan and Sir William Burnett respectively, were in use from the 1830s to the 1860s: (6) In the Great Western it is proposed to use kyanised timber only. The Standard, August 20, 1837 8 pot a aigrette, pots d’airgrets, pots d’aigrettes, pots de airgrets, pots d’airgretts, pots degritts, pots de airgreet, pot de airs grets, pots des aigrettes, pots d’air greets, po’s dair greets, pots d’aigrets, pots d’agrettes, pot d’egret, pots de’airgrets, pots de air-grets, pot d’agret, pots daignettes, pots dargrettes, pow de grate, pots d’aigrete, pots de grats, pots de grete, pots d’aigrette; provenances in Wright (2011b: 127).



Kiss Me Quick: on the naming of commodities in Britain 

 115

(7) The Admiralty have given directions that the Burnetising process is to be brought into full operation. The Standard, June 8, 1848 Kyanising, and burnettising which replaced it, were superseded by changes in technology (treating wood with creosote) so the products and the words fell out of use in the 1860s, but during their thirty-year lifetime they were high-frequency items much in the news. Faraday spoke upon kyanising in his inaugural lecture at the Royal Institution, and Birkbeck lectured on kyanising at the Society of Arts. An Act of Parliament was passed for the building of great kyanising tanks at Grosvenor Basin, at the docks at Rotherhithe, and at the City Road canal basin. The palings round the Inner Circle of Regent’s Park were kyanised, and small brass plates were attached at intervals announcing the fact. The Gentleman’s Magazine in 1841 reported that kyanising had practically become a craze, to the extent that a man had had his coffin kyanised. Like other crazes, it was satirized by Charles Dickens. In his Kyan’s Patent – the Nine Muses, – and the Dry-Rot, the Nine Muses on Mount Parnassus discuss anti-dry-rot treatment (and sing a song with rather nice rhymes in the burden which go a shy ‘un, a shy ‘un / a dry ‘un, a dry ‘un! / a-nigh ‘un, a-nigh ‘un / with Kyan, with Kyan!).9 Kyanised, kyanising do not equate to present-day proprietary names for wood-treatment products; rather, they would have been a contender for the Victorian British equivalent of the Linguistic Society of America’s Word of the Year.10 Kyan+suffixes was actually not Kyan’s own coinage. His preferred term for his wood-preservative was corrosive sublimate, but it did not take with the general public, who preferred a simple conversion of his name. I move now to an inventor’s name + suffix which almost certainly was deliberately chosen by the inventor himself, pulhamite, marketed as a type of artificial stone and invented some time in the mid-1800s by James Pulham of Broxbourne, Tottenham and Spitalfields. Pulham took advantage of the craze for having one’s own fernery, which started in the 1840s and reached a heyday in the 1870s (shrubberies had been in vogue for the preceeding hundred years; rockeries, crazy-paving and gnomes were yet to come). Fern-collectors were known as pteridophilists, ‘lovers of ferns’, and mid-century pteridophilists would buy different types of ferns and try to grow them. However British domestic back gardens tended to be lawned and flat, whereas ferns grow in damp crevices in rocks.

9 ‘Boz’ (1837: 93–98). The Nine Muses sing whilst enjoying a celebratory dinner at the Macclesfield Arms in the New (now the Euston) Road, having triumphed over dry rot. 10 For references and further illustration, including the use of kyanised metaphorically in literature, see Wright (2013).

116 

 Laura Wright

What was really needed was a cliff, or a ravine, preferably with a waterfall. But urban pteridophilists were not to be daunted (many of them being doughty Victorian ladies) and, with the kind of determination that ruled the British Empire, they went out and ordered themselves scenery. Importing a ravine into one’s backyard was no small undertaking, so James Pulham invented a process of covering builders’ rubble with coloured cement which he and his firm artistically textured and shaped to resemble natural rock. As with kyanising, during its lifetime, the word pulhamite was ubiquitous: James Pulham and Son built pulhamite ferneries in suburban gardens in Beckenham, Brixton, Clapham, Croydon, Dulwich, Forest Hill, Upper Norwood, Lower Clapton, Ruislip, Isleworth, Hayes, Hendon, Norbiton, Tottenham, Peckham, Streatham, Sydenham, West Wickham and Winchmore Hill.11 There may have been more, possibly still extant in back gardens. The Queen has a pulhamite fernery in hers at Buckingham Palace, and pulhamite can be viewed in the lake at St James’ Park, where Pelican’s Island was built of pulhamite between 1895 and 1899, at the “gothic ruins” in Gunnersbury Park, built for Baron de Rothschild in 1876, and at the Boating Lake at Battersea Park, with its nearby “crags” and “caves”, the grandest being known as the Victorian Cascade and over which water used to flow, built by James Pulham and Son in the late 1860s. Not a few Victorian seaside councils ordered pulhamite features for their gardens and promenades too, such as Albion Place Gardens, Victoria Parade Gardens, Royal Parade and Madeira Walk, all in Ramsgate, and Cliff Garden, Felixstowe. Some seaside councils went further, placing pulhamite rocks on the actual seafront itself, to make it look more rocky – Blackpool seafront was thus decorated in the 1910s. At Lower Leas, Folkstone, a whole series of zigzag paths and tunnels was built from pulhamite at the same date, in order to get holiday-makers down the cliffs in properly scenic fashion. Pulhamite was used at Winterstoke Undercliff, Ramsgate, to help visitors transition to the beach via something more interesting than steps against a flat sea-wall (Folkstone and Ramsgate in particular splashed the town budget on pulhamite). The –ite suffix was one of many available in English for forming nouns, becoming productive in the nineteenth century attached to names.12 It was especially used for the names of newly-discovered minerals, usually attached to the name of the discoverer, or an honorand: wernerite (1811), humboldtite (1823), brewsterite (1843), darwinite (1860),

11 This information is taken from the work of master pulhamite detective Claude Hitching, Hitching (2012) and at http://www.londongardenstrust.org/features/Pulham2012.htm; http:// pulham.org.uk/where/london/. See also http://en.wikipedia.org/wiki/Pulhamite. 12 See OED -ite, suffix1 2. b. Min.



Kiss Me Quick: on the naming of commodities in Britain 

 117

whitneyite (1861).13 Pulhamite, of course, was not a newly-discovered mineral – but it sounded like one, and enjoyed the cachet.

4 Semantic extension and exploitation of borrowing I now return to borrowings, to which traders gave new life beyond their original meanings. Beginning with banyan, a London tailor’s bill of 1737 reads: (8) Nov ye 9 For takeing ye spots out of young Masters Searles Banyon for 2 doz & 4 brest buttons at 8d p(er) doz for a Velvet coller lining for silk & silk twist for the holes for Buckram & Canvas stayes for work done to the Banyon C. Hoare & Co. archive, Richard Hoare’s bill, 1737

£0 0s 9d £0 1s 7d £0 1s 0d £0 1s 4d £0 0s 9d £0 2s 6d

Banyan was a common eighteenth-century English borrowing meaning a loose gown, from “Portuguese banian, probably from Arabic banyān (16th cent.), from Gujarati vāṇiyo man of the trading caste, ultimately from Sanskrit vaṇij merchant”, first attestation in English a1597, first attestation in the clothing sense 1725 (OED banian, n.) OED banian, n. C2. banian-tree n.: “1638 T. Herbert Some Yeares Trav. (rev. ed.) 122 A Tree (or rather twenty Trees, the boughs rooting and springing up a whole aker together)..namd by us the Bannyan Tree, from their adorning and adoring it with ribbons and streamers of varicoloured Taffata”. A banian meant a trader or a broker in the Indian languages, but Europeans saw banians sitting under specific trees out of whose boughs they had formed a pagoda, wearing specific clothing, and transferred the name to both tree and clothing. Banian only signified a type of gown or tree in European languages, not Indian ones. The  sense-developments are, however, linked to the source; in context, a banian(-type gown) or banian(-type tree) did not need to be fully expanded. This is unlike the uses to which traders put the names of contemporary battles, where there really was no link, semantic or otherwise. In Wright 13 OED wernerite, n., humboldtite, n., brewsterite, n., darwinite, n.2, whitneyite, n. Pulhamite has no OED entry. After the fernery craze died down, the term pulhamite lived on, as the company of James Pulham and Son reapplied the name to a stone-coloured terracotta used for ornamental garden furniture. From the late nineteenth century pulhamite also referred to a less theatrical garden product – sundials, balustrades, birdbaths, urns, and so on.

118 

 Laura Wright

(2011a;  see references therein) I discuss the pink-mauve colour-term magenta. Early names for this newly-­developed aniline dye were roseine (1859), rubine (1862), fuchsine (1865), aniline red (1865), rosaniline (1872), usually named by the inventor. The successful name, magenta (1860), was the name of the town in Northern Italy after which a battle was named in 1859, where the Austrians were defeated by the French and S ­ ardinians in the Franco-Piedmontese War or Second War of Italian Independence. The colour, it has been claimed, is semantically linked to the battle, e. g. Collins English Dictionary: “named after Magenta, Italy, alluding to the blood shed in a battle there”. However, this cannot be the case on two counts. The colour-term magenta (despite its synonyms rubine, aniline red) has always designated a pink-mauve colour unlike that of blood. In order to ascertain this I tracked artists’ colour-tint cards from the second half of the nineteenth century, and although the dye itself might have changed colour over the intervening 130-odd years, the labels have not. Magenta was classified with the violets and mauves, not the reds: (9) Magenta. Cobalt violet. Mineral violet. Mauve. Permanent violet. Permanent mauve. Mineral violet, 2. Mauve, 2. Specimen Tints of Winsor & Newton’s Artists’ Oil and Water Colours, p.1878–a.1882. This is Winsor and Newton’s earliest tint card showing the magenta strip.14 Secondly, there was another similarly-coloured dye, the name of which did not catch on, named solferino. Oberthur and Dautheny et al (1905) is a dictionary of European colour terminology, as by 1905 the numerous patented dye-names in the various European countries had proliferated and overlapped to such an extent that tables of comparisons were needed. Oberthur and Dautheny et al (1905: 182, 169) list under the headword “Magenta rougeåtre Origine: Combinaison du Magenta et du Solférino” the French synonyms “Solférino de Lorilleux (impropre). Solférino violacé”. Solferino, attested as an aniline dye-name in English in 1861,15 was also the name of a Northern Italian town near which a battle was fought in the Franco-Piedmontese War of 1859. These newly-synthesised dyes were named after towns which were repeatedly in the newspapers from June 1859 onwards for the next couple of years. In trade, novelty was key, and battles and their participants’ names lent an immediacy, a dernier cri modernity, regardless of whether there was any link at all with the goods. Names such as Trafalgar,

14 There are copies at Winsor and Newton’s Wealdstone factory, the National Art Library and the British Library. 15 Scientific American 4: 150, antedating OED solferino, n. c1865.

Kiss Me Quick: on the naming of commodities in Britain 



 119

Waterloo, Wellington, Raglan, Cardigan, Sandwich and so on were applied to furniture, dyes, food, confectionary, clothing and all sorts of other commodities. Battle-names in particular were taken up in Britain by speculative builders: Table 1: British addresses named after theatres of war, retrieved from http://www.royalmail.com/ UK street name

No. of UK addresses

Battle location and date

Waterloo Alma Kimberley Trafalgar Mandalay Portobello Ladysmith Inkerman Odessa Mafeking Lucknow Magdala Balaclava Crimea Plevna Kandahar Khartoum Sebastopol Varna Abyssinia Omdurman Spion Cop Shipka Meerut Ulundi Taku

3702 3657 3519 2365 1191 525 521 338 291 266 221 180 161 90 85 78 75 56 54 33 14 7 7 5 5 4

Belgium, 1815 Crimea, 1854 S. Africa, 1899 Portugal, 1805 Burma, 1887 Panama, 1739 Natal, 1899 Crimea, 1854 Crimea, 1853–6 S. Africa, 1899 India, 1857 Ethiopia, 1868 Crimea, 1854 Crimea, 1853–56 Bulgaria, 1877 Afghanistan, 1879 Sudan, 1885 Crimea, 1854–55 Bulgaria, 1877 Ethiopia, 1868 Sudan, 1896 Natal, 1900 Bulgaria, 1877 India, 1857 Zululand, 1879 China, 1859

Practically all of these were residential streets, named by the speculators who built the houses and named to underscore the impeccable newness of the property rather than invoke the country or battle in question, although patriotism did have an appeal for some prospective buyers – the 1856 census lists children named, e. g., Wellington Charles Inkerman, Walter Alma Peace Inkerman, Alma Odessa, Rebecca Sebastopol, Canrobert Raglan.16 16 General Canrobert led the French and Lord Raglan led the English at the Battle of Inkerman.

120 

 Laura Wright

5 Exploitation of sociolinguistic information The final traders’ linguistic technique considered here is that of sociolinguistic exploitation, for which I turn to the selling of perfume. Scented water is now considered to be a non-essential, luxury commodity, but prior to the First World War it was regarded as rather more functional, having the property to ward off airborne germs and included as “toilet vinegar” in first-aid kits. Before 1710, London perfumers puffed eye-water, hair-water, gout remedies, tooth-ache remedies, dentifrice, kidney stone remedies, balsam for preventing apoplexy, sudden death, &c., scorbutick water, oils and elixirs, cordial drops, spirits and treacles against arthritis, liquid to cure the bite of a mad dog, essence against colds and coughs, hysterics and convulsions – but not perfume. Perfume was not advertised to any extent until the 1810s, and when it began, the predominant language was English. It was only in the 1820s that French perfumiers started to become successful in London  – and, conversely, English perfumes became fashionable in Paris. I have noted the following multilingual names in British newspaper advertisements up to 1857: (10) Aqua Mellis; or, the King’s Honey-Water (Charles Lillie, 1721); ­Konigsrauch, or Royal Perfume (S. James & Co., 1821); Atkinson’s Persian Bouquet de Rose, Genuine Eau de Cologne, from Farina of Cologne, and Chardin Hubigant, of Paris (1828); Esprit de Lavende aux Mille-Fleurs; Perfume Bouquet du Roi G. I.V.; Eau de Camelia et Vitiver (Delcroix, 1828); ­Concentrated Essence of Otto of Roses aux millefleures (Atkinson, 1829); Jean Devereaux’s Esprit de Lavande au Portugal, Eau de Camelia, Esprit Bouquet du Militaire, Esprit de Rose, Muguet, Marchalle, Rezeda, ­Chevrefeuille, Portugal, Mousseline, Bouquet des Dames, &c. (Devereaux, 1830); Esprit des Fleurs de la Violette Napolitaine (Hendrie, 1834); Bouquet de Haut Ton (Dickens, 1835); Extrait de Magnolia (Atkinson, 1837); Blount’s Royal Eau de Bouquet (1837); La Bouquet d’Isabelle (Drew, Hayward & Co., 1838); Parfum D’Arabie (S. Poole, 1839); Bouquet de la Famille Royale ­d’Angleterre (Delcroix, 1839); Bouquet du Prince Albert de Saxe Coburg (Delcroix, 1840); Esprit de Violette, Melarosa Florenti (Hendrie, 1842); Bouquets De la Reine Victoria, du Prince Albert, du Prince of Wales, de la Princesse Royale (Delcroix, 1843); Bouquet d’Arabie (Best Ede, 1843); La Reine des Alpes (Ross & Sons, 1847); Rimmel’s Bouquet de Jenny Lind (1847); Viner’s Blumenhauch. – The Breath of Flowers (1848); Val d’Andorre Bouquet (L. T. Piver & Co., 1850); Breidenbach’s Extraits D’Odeurs (1853); Bouquet de la Reine Victoria (Rigge, Brockbank, Rigge, 1854); Eau de Berlin (Hoffman & Co. 1855); La Duchesse Perfume (Phillipson & Co., 1857)



Kiss Me Quick: on the naming of commodities in Britain 

 121

One perfume-title is Latin, two are German, one Italian, and the rest French, starting with Atkinson’s importation of Farina’s Eau de Cologne. The oldest sociolinguistic technique was to invoke socially elevated personages, starting with royalty: The Princely Perfume (Daily Courant, June 30, 1708); the King’s Honey Water (Post Boy, August 26, 1721 – August 29, 1721).

Plate 1: Charles Lillie’s trade-card, 1736. The medieval perfumers’ sign at the top of the ­cartouche is the sign of the civet cat, the scene underneath depicts the then-current name of the building, the City of Barcelona. Note: no French. On the dorse is a bill: “June 16 1736 Rec(eive)d for 6 pounds of Sup(er)fine Rice’d Powder six shill(ing)s p(ar) Ch: Lillie”. British Museum number 1910,1208.14, AN1123604

122 

 Laura Wright

The first non-generic mentions of royalty were Perfume Bouquet du Roi G. I.V. (Delcroix, 1828); His Majesty’s Perfume, Bouquet du Roi G. IV. (Jean Devereaux, 1829); Her Majesty Queen Adelaide’s Refreshing Perfume (Price & Gosnell, 1830); Royal Adelaide Perfume (W. Brewster, 1832); The Royal Adelaide Perfume (Robert Low & Son, 1833); Adelaide Bouquet, King William Perfume, Victoria Bouquet (Rigge, Brockbank and Rigge, 1834). In particular there was a flurry of codeswitching into French around Queen Victoria’s ascension to the throne and her marriage to Prince Albert in 1840: Bouquet du Prince Albert de Saxe Coburg, prepared expressly in honour of the approaching happy event (Widow J. Delcroix & Son, 1840). Like Albert, Adelaide was also German and lived in Germany until her marriage to William; the British royal family was the House of Hanover, and there were German perfumers in London, one of the main innovators of perfume titles being Breidenbach, although he mostly advertised in English and French. Delcroix and Devereaux, two French perfumers (and their families) who lived and worked in London were highly influential in introducing French, rather than German, as the predominant other language in this semantic field. Eau de Cologne was the most popular perfume, sold everywhere and ousting the previously-popular Hungary Water – but it was sold in London via the French language, not the German of Cologne, nor the Italian of Giovanni Maria Farina, its inventor. By the 1840s perfumers were running short on British royals: Royal Bouquets: viz:- De la Reine Victoria, du Prince Albert, du Prince of Wales, the Queen Dowager’s, de la Princesse Royale (Widow J. Delcroix & Son, 1843) and got started on foreigners: The Emperor of China’s Perfume (Robert Best Ede, 1846); The Napoleon Perfume (Emily Dean, 1850); The Empress Eugenie’s Nosegay (Breidenbach, 1853); Eau de la Reine de Hongrie (Piesse & Lubin, 1856). Perfumes were not yet restricted to the female market, neither were they worn on the skin: Scents for the Waistcoat-pocket and Reticule: Rose, Violet, Musk, Tonquin, Mareschal; Essences, &c., in Verbena, Geranium, Forget-me-not, Patchouli (Viner, 1846). In the 1850s an upper-class lifestyle began to be invoked: (11) Guards’ Club Bouquet (L. T. Piver & Co., 1850); Jockey Club Bouquet (Rigge, Brockbank, Rigge, 1854); Yacht Club Nosegay, and the Royal Hunt Bouquet (Breidenbach, 1854); The Oxford and Cambridge Bouquets (Metcalfe, Bingley & Co, 1857); The Belgravia. – a new and aristocratic Perfume (L. T. Piver, 1857). But aspirant-lifestyle titles aping the pursuits of the upper classes were quickly superseded by hot-news, latest-craze titles:



Kiss Me Quick: on the naming of commodities in Britain 

 123

(12) La Reine des Alpes, a beautiful new perfume (Ross and Sons, 1847); The Chobham Camp Nosegay. – A new and exquisite perfume of rich and ­delicate fragrance, from very rare flowers (Breidenbach, 1853); Medjid, or Royal Turkish Perfume (Rigge, Brockbank, and Rigge, 1854); New Crystal Palace Bouquet (James Lewis, 1854); Little Dorrit’s Nosegay (Piesse & Lubin, 1855); Honour to the Brave. – The Victor’s Bouquet, and entirely New Perfume with emblematic device representing the glory and success of our heroes in India (G. T. Jerram, 1857); Postage Perfume (Piesse & Lubin, 1871), Suez Canal, stilled from the Sacred Lotus Flower of Egypt (Piesse & Lubin, 1875); Pyramid Perfume (October 25, 1882). La Reine des Alpes of 1847 is a very early trend allusion as it was not until 1852, when Mr. Albert Smith’s Ascent of Mont Blanc opened at the Egyptian Hall, Piccadilly, on a stage resembling a Swiss chalet, that British tourists realised they actually wanted to go to the Alps. The Chobham Camp Nosegay, advertised in July 1853, refers to the camp of eight thousand soldiers mustered at Chobham in Surrey between June and August in training for the Crimean War. Large crowds, including royalty, went to watch them practice. Medjid is in reference to the Sultan of the Ottoman Empire, Abdulmecid, who was fighting alongside Britain in the Crimean War against Russia on May 31, 1854, when the perfume was first advertised. Five days earlier the Times had reported “By order of the Minister of War, the English airs, God save the Queen and Rule Britannia, and the march of The Grand Sultan Abdul Medjid Khan” (which had been hastily composed for the event) “shall in future form part of the repertory of all the bands of the army”.17 The New Crystal Palace Bouquet, advertised June 29, 1854, was in reference to the Crystal Palace which had housed the Great Exhibition of 1851 in Hyde Park and which had been removed to Penge Common, reopening in June 1854. Little Dorrit was first anticipated in the press in early November 1855,18 the first instalment was out by December 2, and Piesse & Lubin advertised Little Dorrit’s Nosegay on December 16.19 Honour to the Brave, advertised on December 16, 1857, is in reference to the Second Battle of Cawnpore in India’s First War of Independence of 1857, which led to the dissolution of the East India Company in 1857–8 and the foundation of the British Raj. Pyramid Perfume, advertised on October 25, 1882 17 The Times, May 27, 1854: 10. 18 The Lady’s Newspaper, November 3, 1855. 19 Bell’s Life in London and Sporting Chronicle, December 16, 1855.

124 

 Laura Wright

was probably in reference to a drawing in The Graphic of October 21, 1882, of the Duke of Connaught ascending the Pyramid of Cheops, whilst in the vicinity fighting the Anglo-Egyptian War. Sociolinguists usually focus on social groupings – communities of practice – that are interactive, whereas merchants and traders target customers (e. g. collectors of toy soldiers, people who want to go on a cruise) who do not form a necessarily interactive group.20 Dissemination of vocabulary may be one-sided: a reader could have been persuaded to buy a ticket to see fireworks by dint of repetitive advertising, without ever actually uttering the words fourloon, tourbaloon, or indeed even ever hearing them spoken aloud. The fireworking terminology presented in the advertisements nevertheless formed part of the newspaper-readerships’ passive language repertoire. Traders target customer-bases that cut across the traditional social groupings of gender, class or age, revealing more specific social commonalities. The Bank Holidays Act of 1871 served to institutionalise an older set of social behaviours whereby lower-middle and upper-working class women could approach men – even those unknown to them – on certain designated days of the year (Easter Monday, Whit Monday, August Bank Holiday Monday, Boxing Day). Far from wanting to emulate an upper-class lifestyle, ladies of the upper and upper-middle classes could only listen enviously to accounts of Bank Holiday freedoms enjoyed by women further down the social scale. The Bond Street firm of Piesse and Lubin capitalized on this phenomenon, and aided, if not actually caused, the shift away from perfume worn on the handkerchief by both sexes to repel unpleasant odours and germs, to perfume worn on the skin by women to enhance allure: (13) Kiss Me Quick (1858), Stolen Kisses (1858), Box His Ears (1860), Tom-Boy, Pop-Kiss, Follow Me, Lads!, Kiss Me, and Let Me go; Kiss Me, You Dare! (1876), Jolly Dogs (1876). The song Kiss Me Quick and Go, words by Silas Sexton Steel, music by ­Frederick Buckley, had appeared two years earlier in 1856 in the sheet-music Boosey’s Christy’s Minstrels Album containing their Twelve Most Popular Songs. However the phrase kiss me quick had been in use before: kiss me quick was the name of

20 A definition of “community of practice” is “groups of people who share a concern or a passion for something they do and learn how to do it better as they interact regularly” (Lave and Wenger 1991).



Kiss Me Quick: on the naming of commodities in Britain 

 125

a small bonnet standing far back on the head in the early 1840s,21 and Kiss me Quick had been the name of several racehorses.22 The perfume Kiss  Me  Quick seems to have been particularly successful as other perfumers also used the title.23 The Google Ngram Viewer shows a discrete usage of the phrase “kiss me quick” between 1818–1831, and another usage from 1846 to the present, with a huge peak between 1865–1875, its heyday as a catch-phrase. Frederick Buckley also had a song entitled Stolen Kisses, published in 1859 but possibly performed in the music halls a few months previously.24 The phrase stolen kisses had been apparent in the press in 1855 as a quip from the New York Ladies’ Repository and was syndicated in regional British newspapers in June, 1855: “Stolen kisses are said to have more nutmeg and cream than other sorts … A stolen kiss is the most agreeable.” This quotation continued to circulate in British newspapers for over two years.25 Jolly Dogs was another popular song, current from 1865. It was written by Harry Copeland and sung by The Great Vance, amongst others. The Jolly Dogs in question are young men who drink wine and go out a-larking, with the chorus: They always seem so jolly oh, so jolly oh, so jolly oh They always seem so jolly oh, wherever they may be They dance, they sing, they laugh ha ha They laugh ha ha, they dance, they sing, what jolly dogs are we Fal la la, Fal la la, Fal la la, Fal la la, Fal la la, Fal la la

21 OED kiss-me-quick, n. 1., first attestation: 1852 “G. W. Bungay Crayon Sk. (1854) 372: She wears..a Kossuth hat instead of a ‘kiss-me-quick’”; however, this is antedated by the Liverpool Mercury April 7, 1843: “A Popular fashion. – ‘Kiss me quick!’ is the name of a new fashioned bonnet worn by the ladies somewhere, and invented by somebody. The gentlemen find them very convenient, and the ladies exceedingly pleasant. The demand for them is becoming very great. – New York Union.” The same date, April 7, 1843, the Vermont Phoenix published: “A Laughable Scene. – On Saturday morning several young ladies made their appearance in our markets with ‘Kiss me-quick mother’s coming’ bonnets on their heads, and some of the butchers indulged in a little sport at the expense of these votaries of the top of the fashion. “Kiss me quick! Kiss me quick!” escaped the lips of the butchers, and as it fell upon the ears of the young ladies, a blush of deep crimson mantled their cheeks. … – Philadelphia Sentinel.” 22 E. g. The Times May 25, 1846. 23 E. g. “Felix Sultana’s Golden Casolette, which unceasingly emits a delightful fragrance, 1s.; the Fairy Fountain, six different perfumes in box, 1s.; Queen Dagmar’s Cross, a jewel for a lady’s neck, deliciously perfumed, 5s. 6d.; a bottle of Jockey Club, Wood Violet, and Kiss Me Quick, in case, 4s. 6d.; genuine otto of roses, in original bottles, 3s. 6d. All post free. Felix Sultana, Royal perfumer, 23, Poultry, city; and 210, Regent-street, London.” (The Times, March 6, 1865). 24 http://otbrass.com/PublishedMusic/Notes/D_Page.html. 25 E. g. Liverpool Mercury, June 26, 1855; Trewman’s Exeter Flying Post, August 6, 1857.

126 

 Laura Wright

Fal de the ral, de the ral lal li do Slap, bang, here we are again Here we are again, here we are again Slap bang, here we are again, what jolly dogs are we.26

In 1861 Piesse and Lubin advertised their Fountain Finger Ring, with which ladies could squirt prospective gentlemen at balls.27 The Fountain Finger-Ring was part of a wider practice of squirting people with odiferous liquid. In London, Easter Monday, Whit Monday and the August Bank Holiday had become associated with outings to fairgrounds on heaths and commons, and certain Bank Holiday customs arose, for evidence of which I turn to journalism: (14) Up the road, in a word, come boys and girls, men and women, old and young, in rags and in finery, married and single, with babies and without; and all the way by the roadside vendors of “ladies’ tormentors”, long feathers known as “ticklers”, penny bagpipes and tin trumpets, stand contributing to the general uproar. (Adcock 1903: 116).28 The definition of ladies’ tormentor is found under OED tormentor n. 3. f. “A device used to annoy at pleasure-fairs (freq. a device for squirting liquid)”, first attestation f.1891, last attestation 1912. The liquid in question did not smell pleasant. Ladies’ tormentors are found in fiction of the period:

26 Written and composed 1865 by Harry Copeland, performed by Frank Hall, Alfred Vance and Tom MacLagan (http://monologues.co.uk/musichall/Songs-S/Slap-Bang-Here-We-Are-Again. htm). 27 “FOUNTAIN RING. As a means of carrying scent about the person, the FOUNTAIN FINGER-RING has recently become famous. The delight of all who have seen this little conceit is most gratifying to its inventor. It is at once useful and ornamental. By the least pressure, the wearer of the ring can cause a jet of perfume to arise from it at any time desired — thus every one can carry with him to a ball, concert, or sick chamber, enough scent, so refreshing! for the time being. The practical application of this invention causes a good deal of merriment and laughter. A gentleman who abhors perfume, unless it be snuff, ‘squeezing’ a lady’s hand, will receive a shower of the eternal frangipanni or kiss-me-quick, much to the delight of all present at being thus sweetly ‘found out.’ The rings can be filled with perfume with the greatest ease — thus: Press the ball at the back of the ring nearly flat, pour scent into a cup and dip the ring into it; the elasticity of the ball will then draw the perfume into the interior till full.” (The Lady’s Newspaper, November 2, 1861; Piesse 1879, 4th ed.: 303). 28 Arthur St. John Adcock (1864–1930) was a Fleet Street journalist, editor and main writer of The Bookman (http://en.wikipedia.org/wiki/Arthur_St._John_Adcock). His editor, George Robert Sims (1847–1922) was also a London journalist (http://en.wikipedia.org/wiki/George_Robert_Sims).



Kiss Me Quick: on the naming of commodities in Britain 

 127

(15) There is no other fair like Whit Monday’s on Wanstead Flats. Here is a square mile and more of open land where you may howl at large; here is no danger of losing yourself as in Epping Forest; the public-houses are always with you; shows, shies, swings, merry-go-rounds, fried fish stalls, donkeys are packed closer than on Hampstead Heath; the ladies’ tormentors are larger, and their contents smell worse than at any other fair. Also, you may be drunk and disorderly without being locked up – for the stations won’t hold everybody – and when all else has palled, you may set fire to the turf. Hereinto Billy and Lizerunt projected themselves from the doors of the Holly Tree on Whit Monday morning. But through hours on hours of fried fish and half-pints both were conscious of a deficiency. For the hat of Lizerunt was brown and old; plush it was not, and its feather was a mere foot long and of a very rusty black. Now, it is not decent for a factory girl from Limehouse to go bank-holidaying under any but a hat of plush, very high in the crown, of a wild blue or a wilder green, and carrying withal an ostrich feather, pink or scarlet or what not; a feather that springs from the fore-part, climbs the crown, and drops as far down the shoulders as may be. Lizerunt knew this, and, had she had no bloke, would have stayed at home. But a chance is a chance. As it was, only another such hapless girl could measure her bitter envy of the feathers about her, or would so joyfully have given an ear for the proper splendor. Billy, too, had a vague impression, muddled by but not drowned in halfpints, that some degree of plush was condign to the occasion and to his own expenditure. Still, there was no quarrel; and the pair walked and ran with arms about each other’s necks; and Lizerunt thumped her bloke on the back at proper intervals; so that the affair went regularly on the whole: although, in view of Lizerunt’s shortcomings, Billy did not insist on the customary exchange of hats. Everything, I say, went well and well enough until Billy bought a ladies’ tormentor and began to squirt it at Lizerunt. For then Lizerunt went scampering madly, with piercing shrieks, until her bloke was left some little way behind, and Sam Cardew, turning up at that moment and seeing her running alone in the crowd, threw his arms about her waist and swung her round him again and again, as he floundered gallantly this way and that, among the shies and the hokey-pokey barrows. Morrison (1894: 23–24) Tales of Mean Streets.

128 

 Laura Wright

(16) Circus and menagerie, swing-boats, roundabouts, shooting-galleries – all were gone. The whole area lay trampled and bare, with puddles where the steam-engines had stood, and in the puddles bedabbled relics of paper brushes, confetti bags, scraps torn from feminine flounces, twisted leaden tubes of “ladies’ tormentors” cast away and half-trodden into the mire; the whole an unscavenged desolation. Quiller-Couch (1909: 21) True Tilda. From pictorial art we learn of another custom: that of Bank Holiday tumbling. In Modern London; being the History of the Present State of the British Metropolis (Phillips: 1804), Richard Phillips commissioned artist Edward Pugh to provide drawings to be engraved as illustrations. Pugh’s view entitled Greenwich Park with the Royal Observatory on Easter Monday was engraved by J. Pass, and shows a crowd of people, some of whom are tumbling down the hill (Phillips 1804: 477). Barrell (2012: 188) quotes the commentary: “The hill […] is the principal attraction to the merry-making folks. It is extremely steep, and usually thronged; and, every now and then, a group of young men and women, locked hand in hand, rush down this path at full speed; the grand jest and enjoyment of the scene consisting in the falls that happen to the females as well as the males in this slippery enterprise.” Barrell presents other views dated between 1740 and 1811 (2012: 188–91) which also depict holiday tumbling at Greenwich, showing that this custom had quite a long history. He has collected together references to Whit Monday tumbling in drama and verse (2012: 188, 190, 192), and his article makes the point that Pugh’s drawings show the middle and the lower classes enjoying their Bank Holiday outings together; that Bank Holiday behaviour (such as swapping hats, squirting ladies and tumbling down the hill) was enjoyed by the middling sort as well as the working classes. Returning to Piesse and Lubin’s Box His Ears, Tom-Boy, Pop-Kiss, Laughing Water, Follow Me, Lads!, Kiss Me, and Let Me go, Kiss Me, You Dare!, and the Fountain Finger Ring: these names invoke the kind of flirtatious behaviour permitted to middle and working-class women on Bank Holidays. It was horse-play, a roughand-tumble, appealing to youth. There was a physicality about Bank-Holiday licence, where men squirted women and pulled them violently down a hill. Piesse and Lubin also named a perfume Leap Year Bouquet, advertised with the quotation ““In leap year they have power to choose; Ye men no charter to refuse.”  – Chaucer”; and a perfume named St. Valentine’s.29 Piesse and Lubin’s humorous 29 The Times March 12, 1860 and February 11, 1896 (1860 and 1896 being leap-years); Morning Post, August 30, 1862.



Kiss Me Quick: on the naming of commodities in Britain 

 129

flirtatious perfume-titles had a purpose beyond the obvious appeal to holiday fun. Kiss-Me-Quick, Tom-Boy, Pop-Kiss, Box-His-Ears, Follow Me, Lads!, Kiss Me, You Dare, sold women the fantasy of taking the initiative in matters of romance, in a society which did not allow it, apart from on the three Bank Holiday outings of the year, Valentine’s Day, and leap-years. Socially-elevated women were prevented by their social class from participating in Bank Holiday get-togethers, but all, apart from the poorest, could be reached by advertising.

6 Conclusion This survey of traders’ influence on English leads me to conclude that traders’ contributions to the English language are as worthy of inclusion in historical dictionaries as hapax legomena from learned sources. Trading terminology is just as informative for understanding linguistic change as words from other domains. If firework names are omitted from a history of the development of the suffix –oon, for example, then the social dimension of the popularity of the suffix remains hidden – fireworks being essentially a matter of fashion. It is true that nobody nowadays refers to callawapores or kyanising, but they did once, and frequently and in great numbers, and such names carry information about society at the time. Commodity-names such as Suez Canal and Box-His-Ears reveal traders’ expectations of prospective customers, so long as they are viewed from the wider social context (an inference that customers thought that the perfume actually smelled of the Suez Canal would in all likelihood be inaccurate). Before the nineteenth century, non-essential commodities were sold on an appeal to social betterment, but after that date, royalty shifted goods less than novelty. Novelty, as an inducement, started early: the appeal of morphemes –pore, seer- and –oon was not phonological or grammatical but social, –pore and seer- signalling ‘brand-new, latest cloth imports’, and -oon signalling, in context, ‘latest type of firework’. The British Empire, in sum, was a global union of traders, whose seemingly endless conveyor-belt of latest terminology infiltrated everyone’s usage of everyday English.

References Adcock, A. St. John. 1903. Bank holiday London. In George R. Sims (ed.), Living London. vol 2, 114–120. Fuller: London. Boosey’s Christy’s minstrels album containing their twelve most popular songs. 1856: New York: Firth, Pond & Co.

130 

 Laura Wright

‘Boz’ (Dickens, Charles). 1837. Bentley’s Miscellany, 1: January. Hitching, Claude & Jenny Lilly. 2012. Rock landscapes: The Pulham legacy. Woodbridge: Garden Art Press. Lee-Whitman, Leanna. 1982. The silk trade: Chinese silks and the British East India Company. Winterthur Portfolio 17(1) (Spring). 21–41. Morrison, Arthur. 1894 [1912]. Tales of mean streets. London: Methuen & Co. Ltd. Lave, Jean & Etienne Wenger. 1991. Situated learning: Legitimate peripheral participation. Cambridge, Cambridge University Press. O’Brien, Patrick, Trevor Griffiths & Philip Hunt. 1991. Political components of the industrial revolution: parliament and the English cotton textile industry, 1660–1774. Economic History Review, XLIV, 3. 395–423. Oberthur, René, Dauthenay, Henri, Mouillefert, Julien, Harman Payne, C., Leichtlin, Max, Severi, N. & Cortés, Miguel. 1905. Répertoire de Couleurs pour aider à la détermination des couleurs des Fleurs, des Feuillages et des Fruits. Société Française des Chrysanthémistes. Rennes: Imprimerie Oberthur and Paris: Librarie Horticole. Piesse, G. W. Septimus. [1857] 1879. The art of perfumery, and the methods of obtaining the odours of plants. 4th edn. London: Longmans, Green, & Co. Quiller-Couch, A. T. 1909. True Tilda. Bristol & London: J. W. Arrowsmith; Simkin, Marshall, Hamilton, Kent & Co. Ltd. Winsor & Newton. 1878–1882. Specimen tints of Winsor & Newton’s artists’ oil and water colours. London: Winsor & Newton. Wright, Laura. 2010. Eighteenth-Century London Non-Standard Spellings as Evidenced by Servants’, Tradesmen’s and Shopkeepers’ Bills. In Nicholas Brownlees, Gabriella Del Lungo & John Denton (eds.), The language of public and private communication in a historical perspective, 161–190. Cambridge: Cambridge Scholars Publishing. Wright, Laura. 2011a. Semantic shift of the colour-terms maroon and magenta in British Standard English. Revista de Lengua para Fines Específicos 17. 341–376. Wright, Laura. 2011b. The nomenclature of some French and Italian fireworks in eighteenth-century London. The London Journal 36(2). 109–39. Wright, Laura. 2013. On Dickens and Kyanising, Joyce Grenfell and Vassarettes, Barbara Pym and Kestos – or, historical brand names: nouns, evidence and usage. Babel – The Language Magazine 2 (February). 16–21. Wright, Laura. In press. On the East India Company vocabulary of the island of St Helena, South Atlantic, 1676–1720. World Englishes. Yule, Henry & Arthur C. Burnell. 1886 [1993]. Hobson-Jobson: A glossary of colloquial Anglo-Indian words and phrases, and of kindred terms, etymological, historical, geographical and discursive. London, reprinted Ware: Wordsworth.

Online Resources British Library Newspapers 1600–1950 (Gale Cengage British Newspapers 1600–1950). http:// gale.cengage.co.uk/british-newspapers-16001950.aspx (03 March, 2015.) Chronicling America: Historic American Newspapers. http://chroniclingamerica.loc.gov/ (03 March, 2015.) Collins English Dictionary. http://www.collinsdictionary.com/ (03 March, 2015.)



Kiss Me Quick: on the naming of commodities in Britain 

 131

Early English Books Online (EEBO). http://eebo.chadwyck.com/home (03 March, 2015.) Eighteenth Century Collections Online (ECCO). http://gale.cengage.co.uk/product-highlights/ history/eighteenth-century-collections-online.aspx (03 March, 2015.) Google NGram Viewer. https://books.google.com/ngrams (03 March, 2015.) Oxford English Dictionary Online. www.oed.com (03 March, 2015.) Royal Mail database. http://www.royalmail.com/ (03 March, 2015.) Times Digital Archive. http://archive.timesonline.co.uk/tol/archive/ (03 March, 2015.)

Samuli Kaislaniemi

7 The early English East India Company as a community of practice: evidence of multilingualism Abstract: The initial target of the English East India Company was Southeast Asia. As a result, for the first few decades of the 17th century, the Company’s trading posts were relatively isolated micro-communities who lived and traded within the complex multilingual trading world of maritime Asia. Thus the early records of the English East India Company (EIC) provide excellent material for the study of a community of practice, and can reveal the effects of language contact in real time. This article looks at multilingualism in the early correspondence of EIC merchants, namely the records of the EIC trading post in Japan, 1613–1623. The focus is on direct evidence of language contact seen in the form of code-switching and lexical borrowing in the letters. The study shows that code-switching and borrowing do provide evidence of an EIC community of ­practice.

1 The early English East India Company The English East India Company (henceforth EIC) was founded in London in 1600 in order to trade into the East Indies – meaning all maritime Asia, from the Persian Gulf to the Spice Islands in the east and to Japan in the north. The primary commercial targets of the English were the spice markets of India and of what is now the Indonesian archipelago, but also the textile markets of the Indian subcontinent, and the silk and silver produced by China and Japan. As is well known, the focus of the EIC later shifted to the Indian subcontinent, but for the first thirty years or so, they were actively engaged in trade in Southeast Asia, and their primary base in Asia was at Bantam on the island of Java (near modern-day Jakarta). Although the first EIC voyage only departed in 1601, by 1620 the Company had sent near thirty expeditions and had established some two dozen trading posts across the East Indies. Each trading post was staffed by a small number of Englishmen, and by local staff making up the servants, cooks and interpreters. For many trading posts, ships visited as rarely as once or twice a year, and much

Samuli Kaislaniemi, University of Helsinki DOI 10.1515/9781501503542-007



The early English East India Company as a community of practice 

 133

of the time the English merchants at individual factories (as the trading posts were called, after the Portuguese feitoria) were relatively isolated. The primary means of communication across the vast regions the EIC operated in was, of course, correspondence. Yet long-distance correspondence presented a grave problem. Communication with far-flung outposts took months if not years. This meant that verification of information through Company-external means was difficult, if not ­impossible, and thus the Company required a degree of documentation higher than commercial organizations with less extensive networks. The distance of communication was reflected in the length and contents of the letters: EIC employees were required to write extensive letters home, and received lengthy instructions in turn. The fact that communication exchanges took years also means that the Company’s employees in the Indies wrote lengthy minutes of their meetings and reports of their decisions, many of which had to be made at short notice under circumstances not envisaged by the Company directors in London. But one thing the Company directors were well aware of was that conducting commerce in the Indies was a multilingual affair. Therefore in the early seventeenth century, an excellent qualification for finding employment with the English East India Company was knowledge of languages. This is not particularly surprising in itself, as the world of commerce has always been international, and merchant communities have always been multilingual (see e. g. Wright 2002 and Trotter 2003). Potential EIC servants would have learnt Latin in grammar school, and one or several European languages during their apprenticeship, as merchant apprentices were regularly sent abroad for training. They also made use of the dictionaries, grammars and language manuals being published for merchants and other middling-sort Englishmen. Thus East India Company employees sent to the East Indies could often speak Portuguese or Spanish, and some knew French, Dutch, or Italian. Most of them soon picked up local dialects or lingua francas, and all were quick to learn the large mercantile vocabulary necessary for conducting trade in the Indies. This vocabulary is a prominent feature in the letters sent home from the East Indies. It consisted of not only terms for all kinds of merchandise, currencies, weights and measures, but also words necessary for living and trading in maritime Asia: words for interpreters, types of ships, official documents, and also for innkeepers, entertainers and prostitutes. As a casual perusal of the Hobson-­ Jobson will deftly illustrate, this was a highly multilingual jargon, with words borrowed from all languages encountered. But while many of the terms remained close to their regional source, some words came to be used by EIC employees across the Asian seaboard from the Red Sea to Japan.

134 

 Samuli Kaislaniemi

2 What code-switching means in this study This study uses the “common-sense approach” to code-switching as described by Gardner-Chloros (2009: 7–9). Code-switching is here used as a non-restrictive umbrella term for all occurrences of foreign items within a text. This is in part a theoretical stance, for I agree with those who question the validity of seeing languages as discrete entities in the language-users’ repertoire (cf. Gardner-Chloros 2009: 1; Fischer 2001: 105–110; and esp. Rohde et. al 2000). In other words, I am not convinced it is meaningful to try to differentiate between code-switching and borrowing. As many have noted, “one person’s code-switching is another person’s borrowing” (Gardner-Chloros 2009: 170).1 Yet my attitude is certainly in part conditioned by the material I work on. In historical texts, even when there is a desire to make the distinction, it can be difficult to distinguish between code-switches, borrowings and loanwords, especially in the case of switches consisting of a single word or a short phrase.2 Indeed, most of the foreign items in my material occur as single words, and arguably they can be classified as borrowings (see Kaislaniemi 2009a). Yet another reason why I will use the term code-switching is because I am building upon the research of Arja Nurmi and Päivi Pahta (i.a. Nurmi & Pahta 2004, 2012; Pahta & Nurmi 2006, 2007a, 2007b) – since I draw from comparable material, as explained below – and follow their use of terminology. In this study, I am applying their rough three-fold categorisation of codeswitches to analyze my findings (Nurmi et al 2014): – Conventionalised elements – Prefabricated chunks – Free switches In this division, “Conventionalised elements” refers to established borrowings that have not lost their foreign flavour: examples include expressions like terra firma and entre nous. “Prefabricated chunks” is used to mean proverbs and quotations, as well as formulaic expressions (e. g. Alea iacta est). As Pahta and Nurmi (2007a: 41) write, Prefabricated switching “is perhaps the easiest form … [and]

1 This is not to say that the distinction between code-switches and borrowings is not worth debating. For recent discussions of theories of code-switching, see Bullock et al (2014); see also Myers-Scotton (2006: 239ff.) and Gardner-Chloros (2009: 7–19, 165–180). 2 For a discussion of code-switching versus borrowing in historical texts, see especially Pahta (2004: 77–80), but also Grzega (2003) for an excellent summary (and critique) of the classical typologies of borrowing developed by the likes of Haugen (1950) and Weinreich (1966). On historical code-switching, see Trotter (2000), Schendl & Wright (2011a, 2011b) and Sebba et al. (2012).



The early English East India Company as a community of practice 

 135

does not require any fluency in the second language”. Finally, “Free switches” refers to “real” code-switching, where the switch indicates some level of skill in the switched language; but Free switching also includes direct quotes. That is to say, a quotation from the Vulgate would be a Prefabricated switch, but reported speech or a quotation from another letter count as Free switches. These categories are by no means clean-cut, and it can be difficult to establish whether a foreign element is a Conventionalised or Prefabricated item. As an aide to analysis, the description of findings is below supplemented with a close reading of the socio-historical context of the foreign items.

3 A community of practice in speech and writing As defined by Wenger (1998: 72–85), a community of practice has three defining characteristics:3 – mutual engagement, – joint enterprise, and – shared repertoire. The East India Company would appear to fulfil all these criteria. By definition, joint-stock companies consist of groups of individuals who are mutually engaged in a joint enterprise – in this case, trade (what is more, according to rules and principles that are jointly negotiated). And to begin with, they had not one, but several shared textual repertoires: as Englishmen; as English merchants; and as merchants engaged in overseas trade – as described in section 1. Over time, the EIC also developed its own textual styles, standards, and practices, but in the period under investigation their practices were still emergent and clearly unfixed. Their shared textual repertoires not only affected style and practical conventions (such as how to draw up letters of credit), but (as already intimated) also text structure and lexical choices (cf. Locker 1985; Kaislaniemi 2009b). Where the EIC formed a community of practice within the broader (discourse) community of English merchants, it can be seen to have developed in four different (if overlapping and interweaving) spaces within the Company’s sphere: (i) in London (and the British Isles overall), (ii) aboard ships, (iii) in factories (and later in settlements), and (iv) in correspondence (and other texts). All of these formed different spheres of mutual engagement, requiring the (re)negotiation of objectives and practices, and each required slightly different 3 See also Jucker & Kopaczyk (2013), Kopaczyk & Jucker (2013) and Meyerhoff & Strycharz (2013).

136 

 Samuli Kaislaniemi

linguistic repertoires. In the multilingual environments where trade was actually conducted, profuse use of multilingual mercantile terminology, jargon, and code-switching was the norm. The linguistic practices of the EIC developed aboard Company ships and in factory outposts on the one hand, and in Company correspondence on the other. This latter is the critical point for this study: although EIC ships actually spent weeks if not months at each port of call (thus allowing for extensive periods of interaction between EIC employees at various factories and those travelling between them or working on the Company ships), the usual format of Company-internal communication was correspondence. This was conducted in huge numbers, and since factory communities were so distant from the linguistic and textual world back in England, influences between writers were stronger and innovations were adopted rapidly. In this context, it is not surprising that features from spoken language made their way into written texts, as will be seen below. The distance also eventually allowed for the textual sphere of administration and trade to become normative – but this was not yet the case in the early seventeenth century.

4 Material: Early letters of the East India Company This study is based on material in the Corpora of Early English Correspondence (CEEC), which are corpora of Early Modern English personal letters compiled at the University of Helsinki. The CEEC were designed for historical sociolinguistics, for testing present-day sociolinguistic theories and methodologies on historical material. Personal letters were chosen as suitable material for two reasons. First, as Biber and Finegan (1989, 1992) have shown, of all written genres, personal letters are in many ways the closest to spoken language. Particularly relevant here is the point that one of the first text types language change manifests in are letters; therefore when looking at historical texts, the language of letters is often more progressive than that of other genres. Secondly, letters commonly come with the kinds of background information required for sociolinguistic studies: details about when and where the text was written, and who wrote it. This allows for the creation of socially stratified historical corpora, which enable the study of language change in real time, including the comparison of various parameters such as social rank and education. The entire CEEC spans 1403–1800 and contains c. 5.2m words in 12,000 letters written by some 1,200 writers. I have used several bespoke subcorpora of CEEC, which are described as they are referenced below. But my primary material



The early English East India Company as a community of practice 

 137

consists of the Factory collection, which contains a selection of the correspondence of the East India Company trading post in Japan, published in The English Factory in Japan, 1613–1623 (Farrington 1991).4 The East India Company reached Japan in 1613, and established a factory on Hirado, a small island off the west coast of the westernmost main island, Kyushu. The Hirado factory existed for a turbulent ten years before being closed down as unprofitable, having failed to establish trade into China in order to produce the mountains of silver expected by the EIC board of governors in London (the story of the Hirado factory is best told in Massarella 1990). But the Hirado factory left behind a large corpus of material: correspondence, letter-books, a factory diary, ships’ logs and journals, accounts and bills, and some other documents such as wills and contracts (most now held in the India Office Records at the British Library; all extant documents have been published, mainly in Cocks 1978–1980 and Farrington 1991). These documents are of course primarily concerned with business, but the diary and correspondence in particular reveal much about the daily life of the English merchants in Japan, and are unparalleled in containing the only contemporary descriptions of Japan written by Europeans other than Catholic missionaries. The Factory collection consists of 220 letters spanning 1613–1622, containing 198,497 words by 21 writers. The bulk of these are letters written by members of the Hirado factory, primarily its head, Richard Cocks, and its senior merchants, Richard Wickham and William Eaton. Together with the letters of local resident William Adams – generally credited as being the first Englishman in Japan (see e. g. Milton 2002, but cf. Farrington & Massarella 2000) – the letters of these writers make up 85 % of the collection. In a previous study using this material, I discovered that a useful way of dividing the Factory material into further subcorpora is to group them according to the origin and destination of the letters (Kaislaniemi 2009a: 226). Such groupings correlate with smaller communities of practice within the East India Company itself. I have abstracted the spaces of the EIC community of practice (see section 3 above) into 5 “directions” of letters within my sources: 1. Indies local 2. Indies general 3. Indies to England 4. England to the Indies 5. Other 4 For information on CEEC, see www.helsinki.fi/varieng/CoRD/corpora/CEEC/index.html. For a list of texts contained by CEEC, see e. g. the Appendix in Nurmi, Nevala and Palander-Collin (2009).

138 

 Samuli Kaislaniemi

Direction 1 refers either to letters sent between members of a single East India Company factory (including their servants and associates, as well as local merchants etc.), or to letters sent within a relatively small geographical region. Due to their relative isolation, single factories can be said to have formed communities of practice in themselves. Such discourse communities can usually be roughly located to towns, countries, or small regions. Direction 2 refers to letters sent between more distant places within all of the East Indies. The Indies being a vast region, this Direction outlines a community of practice where practices need to be supralocal, and local conventions may not be suitable for communication. Directions 3 and 4 refer to letters sent between the East India Company board of governors in London and the Company’s servants in the Indies. It can be theorized that these letters require the use of less local jargon than those in categories 1 and 2, in order to make themselves understandable. This is not only a concern for in-letters (letters sent to England), for in communicating with employees spread over dozens of wildly divergent cultural spheres, the Company board in London needed to make sure not to misapply terms and concepts learnt from letters received. Finally, Direction 5 lumps all other letters together; in practice however it means letters written within England. Such letters are at the furthest remove from the multilingual environments of the East India Company factories in the East Indies, where the Company actually conducted its day-to-day business, and thus we would expect to find least instances of jargon and foreign items within these documents. Although I have listed 5 Directions, Direction 4 in fact does not occur in the Factory corpus, and there are only 2 letters from Direction 5. These Directions are thus absent from the discussion below, but they may be useful for future studies.

5 Language contact as seen in epistolary code-switching That the multilingual environments the early EIC employees lived and worked in are evident in the texts they produced is hardly surprising. I have explored this topic in two earlier studies: the first was a micro-level analysis of lexical borrowing in its socio-historical and discourse contexts in the Factory records (Kais­laniemi 2009a); the second charted variation within a semantic field in early EIC records (Kaislaniemi 2009b). But neither study discussed code-switching in Factory records in general terms. In this section I first compare the Factory



The early English East India Company as a community of practice 

 139

collection to other Early Modern English merchant letter collections, and then present my data and discuss my findings. My analysis is based on the three-fold division of code-switching types given in section 2, the Direction-based discourse community division given in section 4, and the distribution of languages in the texts themselves.

5.1 Multilingualism in Early Modern English merchant letters Although this article looks specifically at code-switching in the Factory collection, a helpful way of setting the scene is to briefly compare the code-switching in the data to that in Early Modern English merchant letters in general (a similar comparison is made in Kaislaniemi [2009a: 228–229]). Being a part of CEEC, the Factory collection has text-level annotation marking foreign words and passages (see e. g. Raumolin-Brunberg & Nevalainen 2007: section 4.1). This encoding not only allows for the quick retrieval of foreign items and passages in the corpus, but also means that the results of such searches are comparable (e. g. with work by Nurmi and Pahta noted in section 2), and thus comparing Factory to the CEEC in general is feasible. Table 1 shows the amount of code-switches (henceforth also CS), with normalised figures (switches per 10,000 words), in two subcorpora extracted from CEEC, together with the figures for the Factory collection. The first CEEC extract contains all the letters in the corpus (excluding the Factory collection) between 1580–1639, framing the overall context of letter-writing within which the Factory collection letters fall. The second extract contains only the merchant letters in the corpus; since merchants are only one of many social groups represented in the corpus, here I expanded the period to cover 1540–1700 in order to have more data. As can be seen in the table, merchants in CEEC do not code-switch any more than average (indeed, perhaps even less than average), but the Factory collection contains 5–6 times as much code-switching as the CEEC extracts. In other words, something sets the Factory material apart from letters in the CEEC in general. Table 1: Code-switches in CEEC subcorpora and the Factory collection compared Collection CEEC extract 1580–1639* CEEC extract 1540–1700* Factory 1613–1623

All Merchants All

Words

CS

CS /10,000 words

996,496 195,826 198,497

851 141 889

8.54 7.20 44.79

* Excluding the Factory collection; the figures include writers living in England and abroad

140 

 Samuli Kaislaniemi

One hypothesis is that when writers live in multilingual environments (i. e. abroad), this is reflected in their letters – what could be called an “expatriate effect”. It makes sense that daily language contact would be reproduced as an increased number of code-switching in texts (as well as in speech). This hypothesis is explored in Table 2: this table reproduces code-switching in three merchant letter collections in CEEC. The top three rows contain the data from all the letters in these collections: as can be seen, the normalized frequencies of code-­ switching do not deviate from those for the CEEC given in Table 1. In both cases, the data includes writers living in England and writers living abroad. Yet when we separate letters written abroad, as shown in the bottom three rows of Table 2, the normalized frequencies of code-switches multiply by two or three. Table 2: Code-switches in CEEC merchant collections with expatriate writers CEEC Collection All

Expatriates

Cely Johnson Marescoe Cely Johnson Marescoe

Years

Words

CS

CS /10,000 words

1474–1488 1542–1553 1668–1680 1476–1488 1543–1552 1676–1680

51,478 191,695 21,807 10,191 31,505 9,835

2 141 18 1 77 16

0.39 7.36 8.25 0.98 24.44 16.27

The low number of code-switches in the Cely collection is difficult to explain, and suggests that domicile abroad cannot be taken as more than a general indicator of increased code-switching. On the other hand, the textual and letter-­writing practices of the late 15th century may be categorically different from those in the early 1600s. In any case, in the light of Tables 3 and 4, it would appear that the expatriate factor may indeed be one explanation of the prevalence of code-switching in the Factory collection, and the assumption that people living in multilingual communities reflect their environment in the texts they produce seems to be a reasonable one.

5.2 Code-switching in the Factory collection: Figures As seen above, there is more code-switching within the Factory collection than in other comparable and contemporary material in the CEEC. The amount of CS naturally varies within the Factory material. This study does not delve into differences between writers (but see Kaislaniemi 2009a), but Table 3 presents the data divided into Directions (discourse communities). It is not surprising that



The early English East India Company as a community of practice 

 141

the most active code-switching occurs within the tightest community of practice: the Hirado factory. In contrast, it is perhaps slightly unexpected for there not to be more code-switching in letters written between different factories in the East Indies than in those written from the Indies to the EIC directors in London. As seen in Table 3, division by Direction immediately shows how the distance of communication is reflected in length, as mentioned in section 1: letters written from the Indies to England (Direction 3) are on average over thrice as long as letters written locally (Direction 1). Note also how, as mentioned in section  4, Direction 4 is absent in this data and there are only two letters in Direction 5. Since the latter contain zero CS, they are also omitted from the calculations below, but zero-switch letters from other Directions are included for calculating normalized frequencies. Table 3: The Factory collection in the CEEC, by Direction Direction

Letters

Words

Average words per letter

CS

CS /10,000 words

1. Within Japan 2. Within the East Indies 3. East Indies to England 5. Within England

136 38 44 2

80,894 33,263 83,600 740

595 875 1,900 370

574 93 222 0

70.96 27.96 26.56 0

TOTALS

220

198,497

902

889

44.79

This section explores these results in more detail, giving a breakdown according to type of code-switch, and also by languages switched to. Table 4 shows the amount of CS per Direction, divided by CS type, as percentages of CS per Direction. This reveals some interesting variation between Directions: in line with CS being most frequent and common in the tightest-knit (as well as most isolated) discourse community, Direction 1 letters have the highest percentage of Free CS. But Free CS is also very frequent in letters sent back to England, and instead, it is the letters sent within the East Indies that contain the most Conventionalised and Prefabricated switches. This may be explained by the functions of the letters, as discussed below after the examples. When we divide the code-switches by CS type and group them by language, we find that Conventionalised and Prefabricated switches occur in only three languages, while there is Free switching into eight languages. As seen in Table  5, over half of the Conventionalised CS and all but one of the Prefabricated CS in the Factory letters are in Latin. As is illustrated below, this is due to medieval ­letter-writing and mercantile (accounting) conventions, which include a

142 

 Samuli Kaislaniemi

substantial number of highly conventionalised words and expressions stemming from Latin. The Conventionalised switches in Italian (or multivalent Romance) and French fall into the same category. But as can be seen in Table 6, there are some Free CS in Latin, Italian/Romance and French in the Factory letters as well. Table 4: Code-switches in Factory letters: CS type, percentages within Directions CS type

Conventionalised Prefabricated Free TOTAL

Direction 1

2

3

ALL

38.0 5.2 56.8

63.4 6.5 30.1

51.8 0.5 47.7

44.1 4.2 51.7

100.0

100.0

100.0

100.0

Table 5: Conventionalised and Prefabricated CS in Factory letters: Languages, absolute frequencies Languages

CS type Conventionalised

Prefabricated

TOTALS

Latin Italian/ Romance French

212 67 113

36 1

248 68 113

TOTALS

392

37

429

Over half of the code-switches in the Factory collection are Free CS and in various languages. Unsurprisingly, since most of the letters were written by Englishmen living in Japan, Japanese is the most frequent language. It is followed by Spanish, Malay, and Portuguese. Spanish was not only familiar to many of the Hirado factory members from their career before joining the EIC (Farrington 1991: 1542–1578), but was also spoken by the many Iberian missionaries (and traders) in Japan and across the East Indies, and thus valued as a medium of communication between Europeans. Malay and Portuguese, on the other hand, were de facto lingua francas spoken across the Asian seaboard: Portuguese from the Iberian Peninsula all the way to Japan, and Malay from the east coast of the Indian subcontinent to the Philippines and further to Formosa and Japan (see e. g. Ansaldo 2009: 52–80). Arabic, on the other hand, was of course a lingua franca used in the Islamic sphere of influence – from North Africa to the Indonesian Archipelago – which the EIC merchants came into contact with on their way from England to



The early English East India Company as a community of practice 

 143

Japan. Finally, living in the same town as the English in Japan there was a factory of the Dutch East India Company; accordingly, the English were in regular contact with them. Table 6: Free CS in Factory letters: Languages vs Directions, absolute frequencies Languages

Directions 1

2

3

TOTALS

Japanese Spanish Malay Portuguese Arabic Italian/ Romance Dutch Latin French

230 28 35 15 8 7

14 7 3 3 1

60 21 9 11

304 56 47 29 9 8 4 2 1

TOTALS

326

28

106

2 1

1 4

460

The fact that there is this much Free CS in the Factory collection is interesting in itself. However, as is typical for code-switches and borrowings in language contact situations, and as explored in Kaislaniemi (2009a, 2009b), many of the CS and borrowings in these letters are nominals, titles and terms – the staples of mercantile jargon. The interesting question is the extent of non-jargon, of words and phrases that hint at or reveal the writer’s real linguistic competence. The code-switches described in the tables above are illustrated in the examples in the next section.

5.3 Code-switching in the Factory collection: Examples In the examples below, I have highlighted foreign elements in bold, and inserted glosses in [square brackets] where felt necessary. The abbreviations ‘w’ch’ for ‘which’ and ‘w’th’ for ‘with’ are left unexpanded. Citations give the writer of the letter from which the passage is taken, the name of the collection, and the page number in the source edition (Farrington 1991). Further examples of codeswitches and borrowings in the Factory collection and in the early letters of Richard Cocks can be found in Kaislaniemi (2009a, in preparation).

144 

 Samuli Kaislaniemi

Conventionalised chunks: Letter-writing conventions Early Modern English letter-writing conventions include a range of conventional terms and expressions. Most of these stem from Latin, which explains why most of the Conventionalised CS in the Factory letters is in Latin. The primary terms relate to address, intertextuality and text structure. A common, if by this time old-fashioned, pious salutation is Laus Deo (‘praise God’) in (1), used as the opening phrase of the letter. Since the delivery of letters was at best uncertain in the Early Modern period, intertextuality plays an important part in any correspondence: almost every letter contains references to previous letters sent by the writer and received from the recipient. This required the frequent mentioning of dates; repeated references were commonly curtailed by using words such as ultimo, ‘last (month)’, in (2). And finally, although it was at this stage developing the Anglicised variant postscript, quite often the postscript was headed with the eponym in Latin, as in (3). (1) Laus Deo. Firando in Japon le 5th December 1615 (Ralph Coppendale 1615; Factory 346) (2) Loving frend Mr Wickham, since my dep’ture [departure] from yow I have written yow 3 other lettrs besides this, viz. one dated in Shrongo the 21th ultimo, left with our host Stibio to be convayed unto yow; the second from a towne called Odouar, sent per a m’rchant [merchant] of Fushami whoe tould me he knew yow, & was dated the 24th ditto; and the therd letter from this cittie of Edo, dated the 28th ditto, sent per a man of the kyng of Firando’s; (Richard Cocks 1616; Factory 482) (3) Post scriptum. The bongew is not as yit retorned from Firando, which maketh us to wonder of his soe longe staying. (William Eaton 1616; Factory 425) In addition to Latin, French and Italian contributed to these conventional letter-writing items. French is highly specialised and occurs really only as the definite article le in datelines (1), a somewhat redundant mercantile tradition: cf. the use of the English the in (2).5 The equally highly frequent Italian term in (2), ditto, ‘the same’, is requisite when discussing dates. The Latin abbreviation viz., for videlicet, ‘namely’, as seen in (2), is also very frequent particularly in mercantile correspondence, but it was also used in Early Modern English texts generally, much like those survivors into Present-Day English, i. e. and e. g.. 5 I am unaware of any studies of ‘le’ in Early Modern English, but see Wright (2010).



The early English East India Company as a community of practice 

 145

Prefabricated elements: Latin Prefabricated elements are infrequent in the Factory collection. These typically originate from other texts, and thus those occurring here appear to arise from Latin texts read in grammar school and, of course, after. Some of the Hirado factory members take up Latin in the valedictions in their letters, as in (4). This appears to be a means of expressing group membership (28 instances occur in Direction 1 letters, 5 in Direction 2, and none in Direction 3). (4) I praye forget me not in your love to Walter Carwarden and Edward Sares, whome upon my life you shall finde dutyfull, honnest and true boath to the Companye and yourselfe. Vale. (John Saris 1613; Factory 122) Another typical example of Prefabricated CS are proverbs. As seen in (5), Cocks is quite fond of sayings and maxims, such as nemo sine crimine vivit, lit. ‘no one lives without crime’, i. e. is without faults. This comes from the fifth disticht in the first book of the Distichts of Cato, a highly popular schoolbook which Cocks clearly was taught in grammar school (see Chase 1922, esp. pp. 1–11, 16–17). (5) But, as the saying is, nemo sine crimene vivet. Yow must p’don [pardon] me yf I speak falce Latten &c. (Richard Cocks 1620; Factory 766)

Free code-switching 1: Borrowed names (supralocal items) In contrast to Conventionalised and Prefabricated CS, where there are large numbers of a restricted set of foreign elements, in Free CS there is great variation of types but respectively less of each token. Despite the category being called Free CS, in fact most of the words and phrases in this category, too, are if not conventionalised, nonetheless restricted. The biggest subtype are what constitutes the specialist’s jargon: borrowed terms and names. Some of these words later established themselves in the English lexicon. For instance, cagaroches (cockroaches) in (6) is a loan from Spanish which was only coming into English at this time (s.v. OED1). Similarly, lascars and caffros in (7) derive from Persian and Arabic, respectively, but were being borrowed into English through the lingua franca Portuguese (s.v. OED and Hobson-Jobson). Lascars refer to East Indian sailors; caffros to black Africans.6 6 The OED3 entry for Kaffir has a first attestation from 1577, and records the generic sense “black person, esp. one from southern Africa” from 1607 from a letter or diary of EIC captain William Keeling (n.2b, noting it is offensive in PDE). The Hobson-Jobson gives instances from

146 

 Samuli Kaislaniemi

(6) The mapes for the most parte weare all lickwise spoyled w’th the cagaroches. (William Eaton 1616, Factory 535) (7) som wick after the Amacan ship dep’ted [departed] from Langasaq’ [Nagasaki] their fell an extreme storme w’ch sunck a bark they toed after them, wherin were 2 horses w’th 6 laskeras & 2 caffros (Richard Cocks 1616; Factory 427) Other words are restricted to context or sphere of usage. As shown in Kaislaniemi (2009b), Early Modern English did not have a generic word for ‘interpreter’; jurebasso in (8) is the Malay word jurubahasa, ‘language-master’ (s.v. Hobson-Jobson). Naturally, the first words travellers would learn in local languages and lingua francas would include the word for ‘interpreter’. The use of local referents makes communicative sense: in the usage context, everyone knows what is being talked about, and these terms are only opaque outside the broader discourse community. ‘Jurebasso’ was used by Englishmen throughout the East Indies, but other words were more restricted regionally. Keeping to titles, two Japanese examples are bongew in (3) and tono in (17) and (21). Bongew is the Japanese word bugyō 奉行 meaning an administrative official (Farrington 1991: 1593) – with whom the English merchants in Japan naturally were in constant contact. Tono 殿, on the other hand, is not a title as such but an honorific roughly equivalent to ‘Sir’ or ‘Lord’, applied to one’s feudal superiors (Farrington 1991: 1606; see also Kais­ laniemi 2009a). Thus neither bugyō nor tono have simple correlates in English (although the English merchants saw the daimio (or daimyō 大名), the feudal rulers of regions, as petty kings, and often glossed tono as ‘king’). (8) My jeurebasa, binge a Bingalla, brought mee to the Chenesa’s house in the pareyane where all the Cheneses cepes thear marr’s [merchandise] all the tyme the jounks staye theere (Edmund Sayers 1618; Factory 699) Words are of course borrowed for new referents. A famous case in this material is chaw(e), cha 茶, ‘tea’, as in (9). The Hirado factory members were among the first Englishmen to taste (Japanese) tea, took a liking to it, and actively sought out good-quality tea. They also found Japanese lacquerware to their liking – the maky or varnisth woorke in (9) – of which they primarily commissioned writing-desks, which they invariably called scrittores (10). Actually makie 蒔絵 refers to a specific type of lacquerware, in which the item is painted by sprinkling with silver and/or gold dust (Farrington 1991: 1600) – this type of borrowing, where a word or phrase is used in technically the wrong sense, is of course very common since 16th-­century Spanish and Portuguese, and cites a 1614 letter by Richard Cocks as the first English attestation.



The early English East India Company as a community of practice 

 147

language contact situations always contain lots of ambiguity (see the discussion of tatami in Kaislaniemi 2009a). An interesting case is the combined form makeman in (10), in which the Japanese word is given an English suffix – this is the only word in the Factory collection showing word-internal code-mixing. (9) I much marvayle that you write nothing of the receyt of maky or varnisht woorke, with certeyne jarrs of chawe w’ch I sent by the barke of Leamon Dono with my hoste’s man of Sacay. (Richard Wickham 1616; Factory 479) (10) I have sent by this berer 17 sondry p’ssells [parcels] of contores and scrittores marked w’th R.W. The frayt of them I pray pay to the m’r [master] how much it is. My man Jeinkich will sartyfy you how much. I hav bin at Meaco and talked w’th the makeman, who hath promysed that in short tym hee will a-dooun. (William Adams 1617; Factory 640) Contores (10) is an odd case, for it appears Spanish or Italian in form, yet the word does not occur in Spanish, Portuguese or Italian, rather being the English ‘counter’, of French origin, and meaning a bureau, or a table or desk (“for counting money, keeping accounts, etc”, s.v. OED1 counter, n.3 II.3–4).7 Scrittores in (10), elsewhere also scretores and scritorios, on the other hand, is from Portuguese (or Spanish) escritorio, meaning a writing-desk.8 Despite its attraction as a luxury item and the high quality Japan remains renowned for, lacquerware was not a major trade item for the East India Company. The merchants’ jargon did of course contain borrowings related specifically to trade. Some of these were from Portuguese, as ‘partudo’ in (11) – partido, meaning originally a part of a cargo, shipment or inventory of merchandise, but used also to mean the entire cargo or inventory. Others were local, such as goshon in (24): this was Japanese go-shuin 御朱印, literally ‘(honorific)-vermillion seal’, meaning a trading pass or licence granted by the Japanese government (Farrington 1991: 1598). (The technical term was shuinjō 朱印状, ‘vermillion seal document’). Each trading voyage originating in Japan required such a license, and as the English 7 Irwin (1953: 193) suggests contore is from “contemporary Portuguese contador, a cabinet on legs”. The -es endings of the bolded words in (10) may be editorial expansions of the word-final es-graph, which can also stand for just -s (Tannenbaum 1931: 70, 73–74). 8 The glossary in the source edition suggests contore is used for low writing-desks, “sometimes with drawers”, while scritorie (taken to be ‘French, éscritoire’) is used for “cabinets, nests of drawers and writing desks” (Farrington 1991: 1595, 1604). The OED1 has scrutoire (from French, first attestation 1678, but notes the spelling ‘scrito(i)re’ in 16th-century texts) and escritoire (from French, first attestation 1707), defined as a ‘writing-desk constructed to contain stationery and documents; in early use, often one of a portable size”. Irwin (1953) argues that scrutores refer to small cabinets of drawers, and that the form was actually German: Japanese craftsmen were given ­European cabinets to copy; see his article for more on the early English trade in Japanese lacquerware.

148 

 Samuli Kaislaniemi

merchants engaged in trade from Japan to Southeast Asia, they needed to apply for licenses often. As with tono mentioned above, some of the writers gloss goshuin with the English ‘pass’ or ‘passport’ (Farrington 1991: 319, 352, 830, 911). (11) I would wish I weare at Ferando agayne untyl I had a better p’tudo [partido] to make sayle of. I have 2 cattaberas a-dying for your daughter & have sent you one of ben of the new fation, as my hostes sayeth (Richard Wickham 1616; Factory 412) A large number of the borrowings from Japanese in the Factory letters are for everyday items. It is somewhat unclear whether the English merchants in Hirado dressed in native clothes. At any rate, they formed long-term relationships with local women, some even fathering children (see Leupp 2002; Lewis 2004), and many of their letters to each other refer to gifts and purchases relating to their families. Thus in (11), Richard Wickham writes that he has two cattaberas, or katabira かたびら, unlined kimono used in summer (Farrington 1991: 1599), being dyed for William Eaton’s daughter. He is also sending one of ben, or one already dyed red, beni 紅. If they liked tea, the English merchants also enjoyed drinking. Morofack in (12) refers to morohaku 諸白, “a superior quality Japanese rice wine” (Farrington 1991: 1601). Much of this had to be imported to the provincial town of Hirado, which required chartering foyfones or haya-fune 速舟, ‘fast-ships’, and dealing with their sinde or sendo 船頭, masters.9 (12) My last unto yow was per the sinde or m’r [master] of our foyfone, dated yisterday, per whome I sent Mr Nealson’s tyn box or syfon, w’th the greate silver tankard that was Capt’ Jourden’s. And presently after your letters came to my hands, dated in Firando the xviiith currant, w’th the 30 barills morofack & 14 barills of tuny fesh (Richard Cocks 1620; Factory 769) Occasionally there were more juicy stories to relate, as in (13). While kabuki 歌舞伎 today means the style of Japanese drama where all the parts are played by men, in the early 1600s it referred to a type of dance theatre performed by women (see e. g. Liscutin 1991). Thus, the cabuques in (13) indicate dancers or actors. The soldier and nobleman mentioned in (13) would have both carried a wackadash and a cattan as part of a samurai’s attire. The katana 刀 was the primary weapon, a long single-edged sword (that has become iconic in popular culture); and the waki-zashi 脇差 was a short sword used as a back-up weapon, and commonly for committing hara-kiri (or more correctly, seppuku). 9 A short note on Japanese pronunciation as reflected in the spellings in these examples: Early Modern Japanese /f/ became /h/ in present-day Japanese (except before /u/), cf. (12); and vowels were nasalized before /g/, cf. ‘bongew’ in (3) and ‘cattanki’ in (23).



The early English East India Company as a community of practice 

 149

(13) one other souldear of good acco’ [account] having stolen a cabuque from Meaco and being found out in the aprehention would have cut the woman’s throt (w’th her consent), giveing her to wounds but not mortall, & making acco’ [account] the woman had her pasbord [passport], he cut his belley w’th his wackadash; & lastly Micaonacama, the nobellman that gave me my cattan, hath caryed away a cabuque from Meaco & hath payed her master 10,000 tayes in oban. I would I had the money & it makes no matter who hath the woman &c. (Richard Wickham 1616; Factory 404–405)

Free code-switching 2: Quoted use The borrowings and switches in the examples above are mostly words and phrases used more than once, being established in the lexicon of the English merchants. Yet since most of them are words for new referents, some may have been adopted without an understanding of their sense in their source language. Such evidence is very difficult to ascertain in historical texts, yet the following examples provide some window into the linguistic competences of the English merchants in Japan. Although the English and Dutch in Hirado were neighbours and in frequent contact, there is very little Dutch in the Factory letters. The instance in (14) is in the context of reported speech. Farrington (1991: 674) surmises that startmon is Dutch staart-mensch, ‘a man with a tail’, meaning a homosexual – but staartman, ‘tail-man’, is in fact an old Dutch insult for Englishmen which implies they are descended from the devil, who has a tail (Staffell 2000). (14) certen Englishmen having taken pocession of an iland in the Kynge’s Ma’tie’s [Majesty’s] name, where the Hollanders had nothing to doe, they pulled them out per force, shooting at our cullers, geving out many reviling speeches against the Kinge’s Ma’ti’s [Majesty’s] p’son [person], calling hym startmon & saying that yf he had byn at the place where the cullers were, that they would have shot at hym. (Richard Cocks 1618; Factory 667–668) In texts produced in a multilingual environment, the habit of using words and phrases as they occur in the actual communication situations is quite common. That is to say, the writer reports of events and conversations that took place in other languages, and does not attempt to translate everything. Another instance of this can be seen in (8) in the word pareyane: Farrington (1991: 700) suggests that this is Portuguese praia, ‘beach’. In the letter, Edmund Sayers is reporting incidents that occurred in Cochin China (present-day Vietnam), where he went on a trading voyage. Since he was there alone of the Englishmen, the language he would have used for communication was Portuguese, and this appears to be reflected in (8).

150 

 Samuli Kaislaniemi

There is also some evidence of the English merchants’ proficiency in Japanese. One example can be seen in (15), where Cocks refers to a city in Cochin China as Miaco. Here Cocks is actually using the Japanese word miyako 都, ‘(imperial) capital (city)’, for what in the source edition is identified as the city of Huế (Farrington 1991: 196, 199) in central Vietnam (Edmund Sayer does the same, cf. Farrington 1991: 698, 700). This suggests that the English are reporting communications with the Japanese – who refer to the local capital in Cochin China as ‘the capital’, rather than by its proper (local) name. At this time, there were Japanese communities living in many places in South-East Asia, usually called nihon-machi or ‘Japan towns’.10 Since the Hirado merchants mostly chartered Japanese junks for their trading voyages to Cochin China, it would make sense for the crew of these vessels to have sought out the Japan towns at their destinations. The English merchants, too, would probably have felt comfortable staying with these Japanese expatriate communities, where they could at least to some degree understand the spoken language. (15) Soe now w’th greefe of mynde I write you of the ill hap and death of our frend Mr Tempest Peacock in Cochin China, where he arived in saffetie, as the Duch did the lyke, and sould their goodes to the kinge, whoe gave order they should come to his cittie of Miaco to receave payment, but forestald them and sett upon them in their retorne and kild all that were in company, both Duch, English & Japans, their followers. (Richard Cocks 1614; Factory 196) A similar case is presented by (16), where Richard Pitt, writing in Ayutthaya in Siam, uses the Japanese word for uncooked rice, okome お米 (thus commonly written: etymologically the o- is the honorific prefix 御, but 御米 is rare), instead of the English word. Pitt was not a member of the Hirado factory, so it is curious that he used this word at all. Presumably he was reporting information gained from Japanese (or Japanese-speaking) sources – there was a nihon-machi in Ayutthaya as well. (16) in Champa you may build as great a junke as the Duche’s is for the vallue of tenn cattes, allwaies provided that theair may be too or three Einglesh carpentors sent alonge w’th them to give derections, w’th sawes, exes & other toules for the country peopell. Nailes, boults & ocome must be brought alonge w’th them. (Richard Pitt 1618?; Factory 709–710)

10 For more on nihon-machi, see Massarella (1990: 135–137, and especially p. 395 n. 20 for the standard works on the subject).



The early English East India Company as a community of practice 

 151

In contrast, it can be surmised that Richard Wickham was using a well-known term when in (17) he wrote of Nambagins, ie. nambanjin 南蛮人, ‘southern barbarians’, meaning Europeans. This term was derived from Chinese conceptual models of the peoples of the world: ‘southern’ in this context is explained by the simple fact that due to prevailing winds, most of the traffic to Japan came from the south, including the Europeans. (17) the tonos have order given them per the Empe’ [Emperor] that no Nambagins’ goodes shalbe p’mitted [permitted] by them to be sent upp (Richard Wickham 1616; Factory 520)

Free code-switching 3: Phrases and expressions The final category in these examples are free phrases and expressions. Out of all instances of language-mixing in the Factory letters, these best illustrate the linguistic competence of the writers. First, the fact that much of their Latin was Conventionalised or Prefabricated, does not mean that the merchants could not use Latin more freely. Up to a point, anyway: in (18), Richard Wickham refers to the local administrative grouping as decemvirs11 – a reference to Roman history and a period of oligarchy, drawn from grammar-school texts (or later reads of Plutarch and/or Livy). An example of more free usage is (19), where William Nealson forms a cryptic puzzle of the name of Richard Cocks’s mistress, Matingio, but luckily provides a key to solving the cypher – in rather simple Latin. (18) poore Dicke Hudson is like to be turned out of doors by the decemvirs of the street where he liveth in Meaco (Richard Wickham 1616; Factory 505) (19) Read this reversed, ad dextro ad sinistro: OIGNITAM. (William Nealson 1614?; Factory 132) The English merchants appear to have known Spanish somewhat better. Richard Cocks lived for some ten years on the borders of France and Spain (Kaislaniemi in preparation), and Richard Wickham had spent two years as a prisoner of the Spanish and Portuguese, which gave him time to perfect his linguistic skills (Farrington 1991: 1576). Thus his asides in (20) and (21) appear to be true free 11 Towns in Japan were divided into groups that oversaw local order: these were called gonin-­ gumi 五人組, lit. ‘five-person unit’ but meaning five household(er)s (Farrington 1991: 1598).

152 

 Samuli Kaislaniemi

code-switches rather than reported speech: although there is a chance that (20), el más ladino in la lengua Japón, ‘the most fluent in the Japanese language’ (Farrington 1991: 320), is a comment by ‘Jacob the drunkard’ and Wickham spoke with him in Spanish, there is no question that pasa tiempo (‘pass the time; for diversion’) are Wickham’s own words. (20) Jacob the drunkard or caulker of Edoe sayeth that one of the fathers, el mas ladino in la lingua Japon, is close prisoner in Edoe for his goodnes there. (Richard Wickham 1615; Factory 320) (21) Yesterday I visited Simy Dono, whoe gave me kinde entertaynment. Yet I doubte I shalbe troubled farther with him for he hath promised to come to my lodging for to passa tiempo. I had rather he would spend his time somwhere else. Nevertheles I am bownd to see it. He tould me that the tono of Firando was kindly entertayned of the Emp’ [Emperor] & will shortly come downe for Firando. (Richard Wickham 1616; Factory 400) To the crucial question, did the English merchants learn Japanese properly, there is no easy answer, only subtle hints. The records of the Hirado factory contain numerous instances of set phrases like caska matteca in (22) – probably kashikomatta 賢まった, ‘sitting upright, in a formal position’ (Farrington 1991: 547). As another example, one often-repeated phrase in the letters as well as Cocks’s diary is nefone cattanki in (23) – elsewhere also nifon catange, being nihon katagi 日本 気質 (かたぎ), glossed by Cocks in his diary accurately as ‘Japon fation’ (i.a. Cocks 1883: I. 264, 288, 305). But the most revealing example of all is that in (24), where Cocks, having found out bad news, exclaims in Japanese: Warry, warry, warry, or Warui, warui, warui 悪い 悪い 悪い, ‘Bad, bad, bad!’ (Farrington 1991: 632). One senses that this expression was needed all too frequently in the daily life of a foreign trader in Japan. (22) I saw a hudge greate colosso or image of copp’ [copper], standing in the open filds, siting (caska matteca as they terme it) crose-legged lyke a tailor, w’th his handes one in another. (Richard Cocks 1617; Factory 543) (23) Yisterday heare arrived Gonrock, whom I visited and carryed him 2 barrels of wyne & some fish, nefone cattanki, and I had some speches w’th him (William Eaton 1619; Factory 761) (24) Yisternight came your letters dated in Firando the 8th & 9th ultimo, accompanid w’th the goshon, w’ch came in good tyme (I instantly sending it to the cort, where there was much enquiring for it). Soe we gott out our goshons, but the privelegese as they were the last yeare. Warry, warry, warry, &c. (Richard Cocks 1617; Factory 630–631)



The early English East India Company as a community of practice 

 153

There is also some indirect evidence of the English merchants’ knowledge of the Japanese language. For instance, the merchants bought Japanese books, as recorded in Richard Cocks’s diary (Cocks 1978–1980: I. 232) – and as suggested by Richard Wickham’s will (Farrington 1991: 729–736). More intriguing is the encounter at the shogun’s court mentioned by Cocks in a letter in 1620, where Cocks apparently is able to not only understand a conversation in Japanese, but also to butt in with a sarcastic comment (Farrington 1991: 778–779). For a longer discussion of these points, see Lewis (2004: 104–148), and see also the interesting points made by Kusakabe (1955, 1956).

6 Conclusions: East India Company communities of practice In this article I have shown that the early letters of the East India Company are excellent material for the study of the development of a linguistic community of practice. I have previously noted (Kaislaniemi 2009a: 246–247) that the amount of foreign items in the Factory collection is not unusual as such, considering that the letters fall within the context of multilingual interaction as required for conducting international trade. But also, keeping in mind the restricted linguistic communities formed by the EIC factories in the East Indies, the proliferation of foreign items is balanced by the shared understanding that most of the words are for familiar concepts, usually those relevant to conducting business. This is reflected in the fact that different spheres of discourse – the Directions discussed above – show different patterns of usage of code-switching. This suggests that there were already by this time (largely implicit) stylistic rules, such as that one should use less multilingual jargon when writing back to England. That is to say, different spheres of the EIC community of practice demonstrably behaved in different ways. In previous work, I have shown that early EIC correspondence can be used to study the spread of code-switches and borrowings in real time, and that such change can be rapid (Kaislaniemi 2009a). More recently, Jucker & Kopaczyk (2013: 2) have argued that communities of practice are the loci of linguistic change, and that understanding language change requires the study of communicative situations between individuals and communities, and between different communities (also Kopaczyk & Jucker 2013). As seen in this current study, early EIC correspondence fits the bill perfectly. If a study of ten years can reveal this much, extending the scope to a century or more may potentially yield great insights into language change. There is no shortage of material, as some 14 shelf kilometres of East India

154 

 Samuli Kaislaniemi

Company records survive in the India Office Records at the British Library (Moir 1988: ix), much of which have been edited and printed and can be found in digital form. Yet surprisingly, these have been virtually unexplored by linguists, although a handful of exploratory studies have demonstrated their value for historical linguistics (Minagawa 1974–1976; Chaudhary 2009), and for language contact and multilingualism (Kaislaniemi 2009a, 2009b; but see also Kaislaniemi and Bolton, in preparation).12 I hope this article will draw more attention to the value of the records of the English East India Company for linguistic studies.

References Ansaldo, Umberto. 2009. Contact languages. Ecology and evolution in Asia (Cambridge Approaches to Language Contact 8). Cambridge: Cambridge University Press. Biber, Douglas & Finegan, Edward. 1989. Drift and the evolution of English style: A history of three genres. Language 65 (3). 487–517. Biber, Douglas & Finegan, Edward. 1992. The linguistic evolution of five written and speechbased English genres from the 17th to the 20th century. In Matti Rissanen, Ossi Ihalainen, Terttu Nevalainen & Irma Taavitsainen (eds.), History of Englishes. New methods and interpretations in historical linguistics, 688–704. Berlin/New York: Mouton de Gruyter. Bullock, Barbara E., Lars Hinrichs & Almeida Jacqueline Toribio. 2014. World Englishes, code-switching, and convergence. In Markku Filppula, Juhani Klemola, & Devyani Sharma (eds.), Oxford Handbook of World Englishes. Oxford: Oxford University Press. Chase, Wayland Johnson. 1922. The Distichts of Cato. A famous medieval schoolbook (University of Wisconsin Studies in the Social Sciences and History 7). Madison, WI: University of Wisconsin, Madison. Available in the Internet Archive, archive.org/details/ distichsofcato00chas. Chaudhary, Shreesh. 2009. Foreigners and foreign languages in India: A sociolinguistic history. New Delhi: Cambridge University Press. Cocks, Richard. 1883. Diary of Richard Cocks, cape-merchant in the English factory in Japan, 1615–1622, with correspondence (Hakluyt Society First Series 66–67), 2 vols. Ed. by Edward Maunde Thompson. London: Hakluyt Society. Cocks, Richard. 1978–1980. Diary kept by the head of the English factory in Japan: Diary of Richard Cocks, 1615–1622, 3 vols. Ed. by The Historiographical Institute, University of Tokyo. Tokyo: University of Tokyo Press.

12 The EIC records have also been used to study the development of Language for Special Purposes, namely business writing (Locker 1985, 1987). There remains some hope for the late Kitty O. Locker’s unfinished monograph, The Development of the Faceless Bureaucrat: The Emergence of Bureaucratic Style in the Correspondence of the East India Company, 1600–1800, to be published posthumously (Thralls 2007: 10).



The early English East India Company as a community of practice 

 155

Corpus of Early English correspondence (CEEC). Compiled by the CEEC project team under Terttu Nevalainen at the Department of English, University of Helsinki. See www.helsinki.fi/ varieng/CoRD/corpora/CEEC/. Farrington, Anthony. 1991. The English factory in Japan 1613–1623, 2 vols. London: The British Library. Farrington, Anthony & Massarella, Derek. 2000. William Adams and early English enterprise in Japan. STICERD discussion paper IS/2000/394. sticerd.lse.ac.uk/dps/is/IS394.pdf (accessed 30 September 2014). Fischer, Andreas. 2001. Lexical borrowing and the history of English: A typology of typologies. In Dieter Kastovsky & Arthur Mettinger (eds.), Language contact in the history of English, (Studies in English Medieval Language and Literature 1), 97–115. Frankfurt am Main: Peter Lang. Gardner-Chloros, Penelope. 2009. Code-switching. Cambridge: Cambridge University Press. Grzega, Joachim. 2003. Borrowing as a word-finding process in cognitive historical onomasiology. Onomasiology Online 4. 22–42. onomasiology.de (accessed 8 October 2014). Haugen, Einar. 1950. The analysis of linguistic borrowing. Language 26 (2). 210–231. Hobson-Jobson = Yule, Henry. 1903. Hobson-Jobson: A glossary of colloquial Anglo-Indian words and phrases, and of kindred terms, etymological, historical, geographical and ­discursive, 2nd edn., ed. by William Crooke. London: J. Murray. Available on Digital ­dictionaries of South Asia, University of Chicago. dsal.uchicago.edu/dictionaries/. Irwin, John. 1953. A Jacobean vogue for Oriental lacquer-ware. Burlington Magazine 95(603). 193–195. Jucker, Andreas H. & Joanna Kopaczyk. 2013. Communities of practice as a locus of language change. In Joanna Kopaczyk & Andreas H. Jucker (eds.), Communities of practice in the history of English (Pragmatics and Beyond New Series 235), 1–16. Amsterdam/­ Philadelphia: John Benjamins. Kaislaniemi, Samuli. 2009a. Encountering and appropriating the Other: East India Company merchants and foreign terminology. In Arja Nurmi, Minna Nevala & Minna Palander-Collin (eds.), The language of daily life in England 1450–1800 (Pragmatics and Beyond New Series 183), 219–251. Amsterdam: John Benjamins. Kaislaniemi, Samuli. 2009b. Jurebassos and Linguists: The East India Company and Early Modern English words for ‘interpreter’. In R. W. McConchie, Alpo Honkapohja & Jukka Tyrkkö (eds.), Selected Proceedings of the 2008 Symposium on New Approaches in English Historical Lexis (HEL-LEX 2), 60–73. Somerville, MA: Cascadilla. Available online at www. lingref.com/cpp/hel-lex/2008/index.html. Kaislaniemi, Samuli. In preparation. Material communities of practice. The letters of Richard Cocks, English merchant and intelligencer, 1603–1609. Helsinki: University of Helsinki PhD thesis. Kaislaniemi, Samuli & Kingsley Bolton (eds.). In preparation. The East India Company and language. [Special issue]. World Englishes. Kopaczyk, Joanna & Andreas H. Jucker (eds.). 2013. Communities of practice in the history of English (Pragmatics and Beyond New Series 235). Amsterdam/Philadelphia: John ­Benjamins. Kusakabe, Tokuji. 1955. The orthography and pronunciation of the diary of Richard Cocks, cape-merchant in the English factory in Japan, 1615–1622 (1). 京都學藝大學學報 A (文科) [Bulletin of the Kyoto Gakugei University ser. A, Education, social sciences, literature and arts] 7. 1–16.

156 

 Samuli Kaislaniemi

Kusakabe, Tokuji. 1956. The orthography and pronunciation of the diary of Richard Cocks, cape-­merchant in the English factory in Japan, 1615–1622 (2). 京都學藝大學學報 A (文科) [Bulletin of the Kyoto Gakugei University ser. A, Education, social sciences, literature and arts] 9. 1–7. Leupp, Gary P. 2002. Interracial intimacy in Japan: Western men and Japanese women, 1543–1900. London: Continuum. Lewis, James. 2004. ‘Nifon catange or Japon fation’: A study of cultural interaction in the English factory in Japan, 1613–1623. Sheffield: University of Sheffield PhD thesis. Liscutin, Nicola. 1991. Review of Die Entstehung des Kabuki: Transkulturation Europa-Japan im 16. und 17. Jahrhundert. Monumenta Nipponica 46(2). 268–272. Locker, Kitty O. 1985. Earliest correspondence of the British East India Company. In George H. Douglas & Herbert W. Hildebrant (eds.), Studies in the history of business writing, 69–86. Champaign, IL: Association for Business Communication. Locker, Kitty O. 1987. ‘As Per Your Request’. A history of business jargon. Journal of Business and Technical Communication 1. 27–47. Massarella, Derek. 1990. A world elsewhere. Europe’s encounter with Japan in the sixteenth and seventeenth centuries. New Haven/London: Yale University Press. Meyerhoff, Miriam & Anna Strycharz. 2013. Communities of practice. In J. K. Chambers & Natalie Schilling (eds.), The handbook of language variation and change, 2nd edn., 428–447. Chichester: Wiley-Blackwell. Milton, Giles. 2002. Samurai William: The adventurer who unlocked Japan. London: Hodder & Stoughton. Minagawa, Saburō. 1972–1974. Erizabesu-chō Nichi-Ei Kankei Bunken to Gengo エリザベス朝日英関係文献と言語 [A linguistic study of Elizabethan Anglo-Japanese ­relations], 2 vols. Tokyo: Shinozaki Shorin. Moir, Martin. 1988. A general guide to the India office records. London: The British Library. Myers-Scotton, Carol. 2006. Multiple voices: An introduction to bilingualism. Oxford: Blackwell. Nurmi, Arja, Minna Nevala & Minna Palander-Collin (eds.). 2009. The language of daily life in England 1450–1800 (Pragmatics and Beyond New Series 183). Amsterdam: John Benjamins. Nurmi, Arja & Päivi Pahta. 2004. Social stratification and patterns of code-switching in Early English letters. Multilingua 23. 417–456. Nurmi, Arja & Päivi Pahta. 2012. Multilingual practices in women’s English correspondence 1400–1800. In Mark Sebba, Shahrzad Mahootian & Carla Jonsson (eds.), Language mixing and code-switching in writing. Approaches to mixed-language written discourse, 44–67. New York: Routledge. Nurmi, Arja, Päivi Pahta, Anna Petäjäniemi & Jukka Tyrkkö. 2014. The social embedding of multilingual practices in Late Modern English: A corpus-based analysis. Paper presented at the Historical Code-switching: The Next Step conference, University of Tampere, 11–13 June. OED = Oxford English Dictionary Online. Oxford University Press. Available to subscribers at www.oed.com. Pahta, Päivi. 2004. Code-switching in medieval medical writing. In Irma Taavitsainen & Päivi Pahta (eds.), Medical and scientific writing in Late Medieval English, 73–99. Cambridge: Cambridge University Press. Pahta, Päivi & Arja Nurmi. 2006. Code-switching in the Helsinki Corpus: A thousand years of multilingual practices. In Nikolaus Ritt, Herbert Schendl, Christiane Dalton-Puffer & Dieter Kastovsky (eds.), Medieval English and its Heritage: Structure, Meaning and Mechanisms of Change, 203–220. Frankfurt: Peter Lang.



The early English East India Company as a community of practice 

 157

Pahta, Päivi & Arja Nurmi. 2007a. Negotiating interpersonal identities in writing. Code-­ switching practices in Charles Burney’s correspondence. In Arja Nurmi, Minna Nevala & Minna Palander-Collin (eds.), The Language of Daily Life in England (1400–1800), 27–52. Amsterdam/Philadelphia: Benjamins. Pahta, Päivi & Arja Nurmi. 2007b. ‘What we do con amore’: Structures of code-switching in ­eighteenth-century personal letters. In Javier Pérez-Guerra, Dolores González-Álvarez, Jorge L. Bueno-Alonso & Esperanza Rama-Martínez (eds.), Of Varying Language and Opposing Creed’: New Insights into Late Modern English, 401–420. Bern: Peter Lang. Raumolin-Brunberg, Helena & Terttu Nevalainen. 2007. Historical sociolinguistics: the corpus of Early English correspondence. In Joan C. Beal, Karen P. Corrigan & Hermann L. Moisl (eds.), Creating and digitizing language corpora, vol. 2, Diachronic databases, 148–171. Houndsmills: Palgrave Macmillan. Pre-print available at www.helsinki.fi/varieng/CoRD/ corpora/CEEC/generalintro.html. Rohde, Ada, Stefanowitsch, Anatol & Kemmer, Suzanne. 2000. Loanwords in a usage-based model. LAUD Series B 296. Essen: University of Essen. Schendl, Herbert & Laura Wright. 2011a. Code-switching in early English: Historical background and methodological and theoretical issues. In Herbert Schendl & Laura Wright (eds.), Code-switching in Early English, 15–45. Berlin: De Gruyter Mouton. Schendl, Herbert & Laura Wright (eds.). 2011b. Code-switching in Early English. Berlin: De Gruyter Mouton. Sebba, Mark, Shahrzad Mahootian & Carla Jonsson (eds.). 2012. Language mixing and code-switching in writing. Approaches to mixed-language written discourse. New York and London: Routledge. Staffell, Elizabeth. 2000. The horrible tail-man and the Anglo-Dutch wars. Journal of the Warburg and Courtauld Institutes LXIII. 169–186. Tannenbaum, Samuel A. 1931 [1930]. The handwriting of the Renaissance. London and New York: Routledge and Columbia University Press. Thralls, Charlotte. 2007. A tribute to Kitty Locker. Journal of Business & Technical ­Communication 21(1). 8–11. Trotter, David (ed.). 2000. Multilingualism in Later Medieval Britain. Cambridge: D. S. Brewer. Trotter, David. 2003. Oceano vox: You never know where a ship comes from. On multilingualism and language-mixing in medieval Britain. In Kurt Braunmüller & Gisella Ferraresi (eds.), Aspects of Multilingualism in European Language History (Hamburg Studies on Multilingualism 2), 15–33. Amsterdam/Philadelphia: John Benjamins. Weinreich, Uriel. 1966 [1953]. Languages in Ccontact. Findings and problems. The Hague/ London/Paris: Mouton de Gruyter. Wenger, Etienne. 1998. Communities of practice: Learning, meaning and identity. New York: Cambridge University Press. Wright, Laura. 2002. Code-intermediate phenomena in medieval mixed-language business texts. Language Sciences 24. 471–489. Wright, Laura. 2010. A pilot study on the singular definite articles le and la in fifteenth-century London mixed-language business writing. In Richard Ingham (ed.), The Anglo-Norman ­Language and its Contexts, 130–142. York: York Medieval Press and The Boydell Press.

Agnete Nesse

8 Language choice in forming an identity: linguistic innovations by German traders in Bergen Abstract: In this article, the sociolinguistic practise in the city of Bergen in Norway is presented in view of the choices made by the local merchants who were immigrants or sons of immigrants from German speaking areas. The innovations that took place are analysed as being driven by economic and strategic interests, and emphasis is placed on the practical, pragmatic attitude towards language among the merchants. We first follow the language shift of the merchants from German and into (written) Danish and (spoken) Norwegian. Then we have a brief look at their attempts at developing a New Norwegian language during the era of national romanticism. During the main period of German settlement in Bergen, from the middle of the 14th and until the middle of the 18th centuries, the German merchants were not integrated in the Bergen society. This was a consequence of the policy of the Hanseatic League, to which they belonged. The Hanseatic merchants would not benefit from learning Norwegian actively; thus, receptive bilingualism prevailed. However, as a new, local trade organization took over from the Hanseatic League in the middle of the 18th century, the merchants were soon to make use of language as part of a renewal of both trade and identity. As a result of this they chose, at least officially, Danish/Norwegian, and the decline of the use of German in Bergen started. However, the merchants were pragmatic, using both languages according to what gave the best results. A hundred years later, German was no longer the mother tongue for merchants in Bergen; they all spoke Norwegian and wrote Danish. Still, by the end of the 19th century, language became important once more, as debates of linguistic tradition and innovation became vivid in Norway. Many of the Bergen merchants of German ancestry chose innovation by joining in establishing the first language organisation of Norway, The Society of the Westmen. This society favoured a New Norwegian standard based on Old Norse and contemporary rural dialects. After an introduction giving the background for the bilingualism in Bergen, the role of the Bergen merchants as linguistic innovators is illustrated through three case studies.

Agnete Nesse, Bergen University College DOI 10.1515/9781501503542-008



Language choice in forming an identity 

 159

1 Introduction Language shift is always a consequence of individual choices that in sum lead to a shift for a whole speech community. The reasons why these choices are made vary and need not be the same for each individual. In order to understand why any particular language shift takes place, it is necessary to analyse both these individual choices and the social and political conditions on a community level. As a general pattern, we can assume that there are certain factors in the society that start the process and that the different members of the group in question relate to these factors in certain ways. Several of the important studies of language shift deal with indigenous or minority languages ceased being used in favour of the majority language in the area (Haugen 1953; Bull 1996; Dorian 1989). This happens either in certain domains or in all domains, in the latter case resulting in language death. These shifts can be the result of determined policies from the authorities, or they can be the result of more subtle group pressure. Sometimes the latter process can be more effective than the first. Einar Haugen’s investigations of the fate of Norwegian in the USA give an example of that: Even if the school politics during the 19th century was strongly in favour of English only, the lack of teachers and lack of control with the linguistic practises of the teachers enabled many immigrants to be schooled in their home language instead of in English (Haugen 1953: 39–40). After the two World Wars, however, when there was a general scepticism in the society to anything un-American, the use of Norwegian in immigrant families dropped dramatically (Haugen 1953: 255–258). Haugen’s well known model of language shift through five stages illustrates how language shift can be a process going from monolingualism in the immigrant language via different types of bilingualism until new monolingualism – in English – is reached. For the language shift situation from Norwegian to English in the USA it must also be taken into account that immigration lasted through a whole century, from approximately 1820 and until 1920. This means of course that the possibilities for the immigrants to maintain their home language were great during the immigration, as newcomers benefitted from the fact that they could be met and helped by people speaking their own language. This is also important for the bilingualism that existed in the West–Norwegian city of Bergen during the period from the 14th and until the 18th century. The language shift that will be discussed in this article has some resemblance to the situation in the USA, and at the same time, there are important differences that can illuminate both the general pattern of language shift situations, and the way in which each situation is unique. In our case, the language shift took place in a group of merchants who had for centuries been able to use their mother tongue

160 

 Agnete Nesse

in the new country. When the shift finally happened, it was not due to explicit demands from the authorities, but rather to the desire to find a Norwegian identity.

2 Bilingualism during the Hanse era (1350–1750) The city of Bergen on the west coast of Norway was a Norwegian and German bilingual city for four hundred years. This article deals with the language shift at the end of this period, when the process of going from bilingual to monolingual started. The different choices made by the different speakers and writers during the period of language shift rely to a large degree on the sociolinguistic conditions in the city during the era of bilingualism. Therefore, it is necessary to give an introduction to these sociolinguistic conditions, and to the political and commercial frames that formed them. The contact between Low German and the different Scandinavian languages during the late middle ages has been thoroughly examined during the last decades; an overview can be found in Elmevik and Jahr (2012). Bergen was founded in 1070 as a commercial port, and as a residence city for the Norwegian kings. The harbour of the city was deep and well sheltered, and this, in combination with the position mid-way between the rich fisheries in the northern part of Norway and the markets in other parts of Europe, led to thriving business, based on export of dried fish and import of salt, grains and other products. The powerful, German-based Hanseatic League took an interest in this trade, and was granted privileges by the Norwegian kings (Burkhardt 2009). This led to the formation of a trade settlement called Kontor ‘office’ during the latter half of the 14th century (Helle 1982). The Hanseatic League had similar trade settlements in London, Novgorod and Brugge, but none of these was as long lived as the one in Bergen, which existed until the middle of the 18th century. There are several reasons for the long lasting existence of the Kontor in Bergen; only two of these reasons will be mentioned here: The Norwegian dependency on imported grains, and the late forming of a local bourgeouisie (Fossen 1979). The latter is of great importance for the language shift process. How bilingualism emerged in the city and how long it took before the sociolinguistic pattern that we know from the 14th century and onwards emerged is hard to tell. Nevertheless, there must have been experiments with both learning, accommodation and linguistic negotiations before the speech community settled in what seems to have been a very stable, receptive bilingualism. Studies of letters, treaties and other texts from Bergen during the Hanse era show a remarkable lack of metalinguistic comments (Nesse 2002: 133). Even though such silent sources must be used with care, there is a clear contrast between the high level of conflict



Language choice in forming an identity 

 161

on economic matters and the lack of conflict on linguistic and cultural matters. The languages that met in the early years were Old Norwegian and Middle Low German. These were closely related languages, and it is reasonable to assume that the speakers fairly easily were able to learn “the other” language receptively in relatively short time. Already in the early 1990s attempts were made to measure the closeness between the late medieval Scandinavian varieties of Swedish and Danish and Low German, both theoretically (Braunmüller 1995) and empirically (Warter 1995). Closeness of the languages may, however, not be a prerequisite for the development of receptive – rather than active – bilingualism. In modern Norway, for example, we see that immigrants with English as their mother tongue often do not learn to speak Norwegian actively; they get by with a receptive competence. Thus, a conversation in a restaurant can be bilingual in that the waiter speaks English and the guest speaks Norwegian. The reason is that all Norwegians between 10 and 80 years of age know some English. Learning English is an integrated part of being Norwegian today, and research on the Bergen speech community, from early in the 20th century (Brattegard 1932) as well as the research from 1990 onwards (summed up in Elmevik & Jahr 2012), suggests that during the Hanse era, learning Low German was an integrated part of being from Bergen. However, the terminology has changed from “Mischsprache” (Brattegard 1932: 303), via “Semicommunication” (Braunmüller 1995) to passive/receptive bilingualism (Nesse 2002; Rambø 2010).1 As English today is the lingua franca of large parts of the world, Middle Low German was the lingua franca of Northern Europe in the High and Late Middle Ages. The most important reason for the emergence of receptive, rather than active, Norwegian–German bilingualism during the Hanse era is how the Kontor was organised. The Hanseatic merchants were bachelors, which means that no Norwegian–German families were formed that could establish a mixed identity. Thus, the Kontor was a male, German “island” in the middle of the Bergen trade area. This area was called Bryggen ‘the wharf, pier’ in Norwegian and de Brugge ‘the pier, street’ in Low German. The recruitment of new traders was secured through the arrival of German apprentices. These were teenage boys who came to Bergen from the North German areas, and who after their apprentice years either went back to Germany, or stayed on as foremen and later as merchants at Bryggen. Just as it must have been a part of the linguistic upbringing of Norwegian children

1 The indicators that receptive, rather than active, bilingualism dominated Bergen are discussed in Nesse (2002). See also Braunmüller and Zeevaert (2001) for a discussion of the terms semicommunication and receptive bilingualism.

162 

 Agnete Nesse

in Bergen to learn to understand Low German, one of the numerous things the young apprentices from Germany had to learn upon arrival in Bergen – or during the first year as an apprentice – was to understand Norwegian. We do not have direct information of this process from Bergen, but Winge (1992: 274) cites a Danish source from the 18th century where apprentices of both languages meet, and how after only a few weeks a kind of basic understanding of the other language is established. In a situation where both groups learn the other group’s language and communicate by the means of both languages, it is no wonder that similarities emerge. Loanwords were exchanged, and during the centuries a core vocabulary was developed, covering lexical fields that were important for the co-existence of the two groups. This common vocabulary consisted of both German and Norwegian words, such as the Old Norwegian klefi ‘small room’ used as kleff in Low German, and the Low German Doorschlag ‘sieve’ used as dørslag in Norwegian. In the local dialect of Norwegian, the grammar went through important changes as the result of the many accommodations that took place (Nesse 2002: 2003). Even if the groups communicated well with each other in their two languages (which became more similar as the centuries went by) there was a strong understanding among the groups that they spoke two different languages. The identity of being German or being Norwegian was strong and tightly linked to language. Language choice enabled them to distinguish the two groups that otherwise were very similar: The North-Germans and the Norwegians had the same religion (they changed from Catholic to Lutheran Christianity at about the same time), more or less the same culture, and the same appearance. The important difference between the Hanseatic merchants and the rest of the population in Bergen, namely the privileges granted by the kings stating which group was allowed to do what, could not be seen or heard at first glance. Therefore, it seems that the languages, even if the differences were not the largest, and even if they all understood each other, became the most important sign of identity. Thus, speaking German meant several things: It meant belonging to the Hanse, being a man, dealing with trade, being a bachelor, living at Bryggen, going to the German church and following Hanseatic jurisdiction. This presentation of the sociolinguistic conditions is, of course, simplified. The Hanse had their Kontor in Bergen for four hundred years and it is obvious that changes took place during that time (Nedkvitne 2014). One important aspect is that Norway entered a union with Denmark, as the weaker partner. Bergen grew, and the number of Hanseatic merchants decreased, especially after the great fire of 1702, when the whole area where they lived and worked burned down. Most important for the sociolinguistic conditions was, however, the emergence of a local bourgeoisie. As already mentioned, a bourgeoisie did not fully



Language choice in forming an identity 

 163

develop in Norwegian cities until the 16th century (Fossen 1979). Whether this was because the Hanse had too much power, or the Hanse was powerful because of the lack of Norwegian traders, has been disputed by historians (Ersland 1999; Nedkvitne 2012). From 1570 on, the city of Bergen had a proper council, and citizenship became a formalised institution. The civil book, Bergenn Byes Borgers Bogh (D)2 1551–1751, is the oldest account of citizenship in Norway, and since place of birth is listed for each new member of the bourgeoisie, the international orientation of this group in the society becomes evident. Interestingly, it is also in exactly these years in which the first history of Bergen is written, in a book called Bergens Fundas (Sørlie 1957). In this book, the hero is the King’s representative, who frightens the Hanseatic merchants and makes them respect the local power. By the middle of the 16th century, the city council in Bergen became the main, local counterpart for the leaders of the Hanseatic Kontor. Earlier, the king’s representative in the city held this position, now the power was divided between the state and the city. The new situation meant that there were merchants belonging both to the Hanse and to the city. Whereas the Hanse merchants were all German, the city merchants were of different origin. The civil book shows that the two largest groups consisted of Norwegians and Germans, but there were also Danes, Englishmen and Scots who took citizenship in Bergen during the 16th century (Fossen 1979: 43). The Germans who took citizenship were to a large part Hanseatic foremen or merchants who decided to break their ties with the Hanseatic League, and settle as independent, Norwegian traders in Bergen. This meant less security for them since they were no longer backed by the Hanseatic League and the League’s economic and military muscles, but it meant the possibility to marry and form important networks in order to be powerful in the city. They were often prosecuted by the Kontor, since they had broken the oath to the Hanseatic League (Fossen 1979: 45, Ersland 2011: 125). Ersland reads the anti-Hanseatic tendency in texts written by these ex-Hanseatic merchants as a way to legitimize their choice to leave the Kontor.

3 After the Hanse era: a new identity as German Norwegians The group of merchants in Bergen who were German born, but listed members, not of the Hanseatic League, but of the Bergen bourgeoisie, are sociolinguistically

2 Since examples in this articles are in German, Danish and New Norwegian, the language of the quotes is indicated by the letters G, D, NN respectively.

164 

 Agnete Nesse

interesting. On the one hand, they benefited from the traditions in the speech community; the well-established receptive bilingualism in Bergen (Jahr 1999; Nesse 2002; Rambø 2010) must have been a strong argument for the German-born merchants who had left the Hanseatic League, to continue to use German. On the other hand, since being German and speaking German for so long had been connected with the Hanseatic League, there might be a conflict of identities. If they had chosen to leave the Hanseatic Kontor in order to become citizens of Bergen, and thus were seen as traitors by the Kontor’s leaders, which language choices were available to them? Did they have to abandon their mother tongue and learn Norwegian? These merchants had some opportunities that the Hanseatic merchants did not have, namely to join the city council, be part of a Norwegian church congregation, and marry. At the same time, our sources show that it was convenient for the non-Hanseatic merchants to “surf” on the receptive bilingualism of the city. Their informal writing was in German (a group of protocols are analysed in terms of language choice in Nesse 2012), and it seems that before 1754 the language choice of the independent merchants was pragmatic: Those who felt a need to learn Norwegian actively did so, the others did not bother. However, after 1754, this changed. At this point, the Hanseatic Kontor was down to one single man, as opposed to 1500–2000 members during the heydays of the 15th and 16th centuries. The reason for the decline will not be discussed here, but the fact that the Hanseatic League was founded during the Middle Ages and can be characterized by a mediaeval way of trading explains some of it at least. But the idea of a trade organisation, where several firms agreed on prices and benefited from each other, had not become outdated. The merchants saw the strength of an organisation based on some of the same ideas as the Hanse but on a small, local scale. At the same time, it was important for them to keep their position as good Bergen citizens. As a result of this, on 7 October 1754 the Bergen trade Kontor was established. Of the founding fathers of this new organisation, two were born in Germany and at the time of the foundation they had lived in Bergen 13 and 14 years respectively. A third was the son of a German father (Nicolaysen 1878: 178, 183; Wiberg 1932: 234). Their linguistic choice was clear: As a symbol outwards, to the population and leaders in Bergen and the king in Copenhagen, they used Danish only. In 1768, a Norwegian clerk was hired to make sure of this (Nicolaysen 1878: 200). Internally however, in bookkeeping and letters among the Bergen merchants, the new Kontor continued the pragmatic attitude to language that the German, non-Hanseatic merchants had practiced earlier. Analyses of texts such as letters, protocols and family books (Nesse 2007 and 2012) show that first generation Germans wrote predominantly in German, the second generation wrote in both languages, and the third seemed to prefer Danish. Regarding the letters, there is not a pattern of accommodation to the language of the receiver. On the contrary,



Language choice in forming an identity 

 165

there seems to be an expectation that all receivers would understand, regardless if you wrote in German or in Danish. There is only one single example (analysed in Nesse 2002 and 2007) where language choice is used to emphasize the power relations during a conflict. The leaders of the Bergen Kontor refused to accept a letter in German from a colleague, but insisted that he should translate it into “the letter of the country” – Danish. This in spite of the fact that the leaders were born in Germany or sons of Germans, men we know from other texts used German in most of their writing. Before I present some of the sources from the period of language shift in Bergen, I will briefly explain the relationship between Low German, High German, Norwegian and Danish in the city, since this may be somewhat confusing for those not familiar with the history of the Scandinavian languages. When the first Germans arrived in Bergen in the 13th century, Old Norwegian and Middle Low German were both written codes as well as spoken varieties. During the era of the Hanse Kontor in Bergen however, both languages almost ceased to be used in writing, and were restricted to the oral domain and very informal written domains such as comedy and poems for weddings etc. The reason why Norwegian and Low German both lost the prestige that goes with the official, written domain was not the Hanseatic League. Rather, the loss was due to political and sociolinguistic processes that happened both in Scandinavia and in the German areas, independently of one another. Norwegian was replaced by Danish in writing, later also in formal speech, and Low German was in the same way replaced by High German. The result was a double diglossia: The Norwegians in Bergen used Norwegian as L-variety and Danish as H ­ -variety, and the Germans in Bergen used Low German as L-variety and High German as H-variety. It is not clear if these parallel changes carried much meaning to most people, as the dialects were only slowly influenced by these changes. Even the literate part of the society may not have felt this developing process of a state of diglossia as a big step. The shift from Norwegian to Danish writing happened gradually through mixing, so that texts in the 16th century can be seen as Norwegian with Danish influence or as Danish with Norwegian influence. The shift from Low to High German happened approximately one century later, and by 1700, the Bergen texts were written in either Danish or High German. There are differences, however, in how the varieties were labelled. The German varieties are always called “German” (dudesch [G], deuthsch [G], tydsk [D] etc.), in the sources, both by the Germans themselves and by the Norwegians. On the other hand, the labelling of the language of the Norwegian inhabitants in Bergen is a more complicated matter. “Norwegian”, “Nordic” and “Danish” (Nordesch [G], Norsk [D], Dansk [D]) are used, and in spite of much research on the matter (Janson 1997; Sandøy 2011), there seems to be no clear pattern as to what

166 

 Agnete Nesse

is called what before the national romantic period of the 19th century. Even in the Norwegian constitution of 1814, the (Danish) language is called “Norwegian”. In the case of the constitution, the distinction between Norwegian and Danish did not matter, what mattered was that the language should not be Swedish, since Norway in 1814 was released from the union with Denmark only to be included in a new union with Sweden. The same logic applies to the labelling of languages in Bergen: It had to do with which distinctions were important and which were not. In Bergen before 1850, German vs. Norwegian/Nordic/Danish was important, not Norwegian vs. Danish. In discussing the written sources of the 18th and the early 19th centuries I will use the terms “German” and “Danish” for the sake of simplicity, since these were the languages the sources are written in; knowing that most of the speakers spoke a Low German or a Norwegian dialect.

4 Case study 1: trying to establish Danish as the default language The first source that will be discussed is a book published on the same day that the Bergen Kontor was officially established (Wiberg 1945). This book was meant to be read by the new apprentices that came to work with fish trade at Bryggen in Bergen. It listed the requirements the apprentices had to meet in order to pass their foreman’s exam, and it contained local rules and regulations. The full title of the book is (1) Kong Friderich den femtes allernaadigste Anordning og Articler for De Contoirske i Bergen Dateret Christiansborg slot den 7 October 1754, med modstaaende oversættelse paa Tydsk, til deres tieneste som ey maatte forstaae det Danske (D) [King Friderich the fifth’s most gracious regulations and articles for the members of the Kontor in Bergen, dated the castle of Christiansborg [in Copenhagen] the 7 October 1754, with facing translation in German, for the service of those who might not understand Danish.] The title of this text book shows three important characteristics of the Bergen trade Kontor: First that many of the apprentices were German, and did not know Danish when they arrived in Bergen, they had to learn the language before they could take part in the receptive bilingualism that prevailed in the city. Secondly the title is a declaration of loyalty to the Danish-Norwegian king, this is very much contrary to the Hanseatic Kontor, who had the king as its counterpart in negotiating privileges and responsibilities. Last, but not least the insistency on Danish



Language choice in forming an identity 

 167

being the default language and German being reduced to a sort of exception. This gives a clear signal of the language policy of the Bergen Kontor. However, there is reason to doubt the truth of the last statement in the title, where it says that the book was translated into German. Was this really a book written in Danish and then translated into German? On the other hand, it may have been written in German, and then later translated into Danish. In that case, what made the authors of the title insist on the contrary? There are many facts that speak in favour of the assumption that the book was based on earlier German, not Danish versions. Wiberg (1945: 71) states that the articles are excerpts from older, Hanseatic, articles, adapted to the requirements of the new Kontor. In the book from 1754, there are frequent references to the 1672 version of the “articles of the Kontor” – in this case the Hanseatic Kontor, which only used German in their texts. There are also frequent references to old customs: On page 32: efter den maade som ved de tydske contoirske stuer ald sin tiid haver været brugeligt (D) ‘in the way that at the German firms belonging to the Kontor always has been customary’. This shows that even if the new Bergen Kontor in 1754 were eager to show their affiliation to the city and the state (the king), and not to the Hanse, they did actually also see their organisation as a kind of descendant to the Hanseatic Kontor. Another interesting fact in this text is that the Danish version has so many more Romance loanwords than does the German text. This was at a time when Romance, and especially French loanwords were in fashion, and they were marked graphically using different fonts, so there is a clear indication that if not the writers themselves then clearly the printers were aware of the loan status of these words. Table 1: Romance loan words Danish text

German text

(English translation)

Dependerer Chicanere exempler extendere leverancen logemente necessaire piquanterie presserende provianten public resterende stipulerede

abhängt beunruhigen beyspiele ausdehnen ablieferung herberge nothwendigsten grollübel trifftiges speise-vorrath gemein wesen rückständigen festgesetzt

depends bother / worry examples expand delivery lodging necessary harassment important provisions public remaining stipulated

168 

 Agnete Nesse

Many of the Romance loanwords that were used in Scandinavian came via German. This means that if the book had been translated from Danish and into German, it would have been likely that also the German text had these loanwords. Thus the use of Romance loan words in the Danish version of the text, where the German version of the text has German words, indicates that the German version is the eldest one. The reason why the leaders of the Bergen Kontor chose to express it differently must have to do with their need to reassure the society around them that even if many of them were German and they formed a Kontor, and even if they lived and traded in the area of the town where the Hanseatic Kontor had had their quarters, they were not Hanseatic. On the contrary, they were loyal subjects to the city and to the king. Given the sociolinguistic history of Bergen, an obvious way to show their loyalty was to choose the local language for their organisation. On the other hand, these people were merchants, and as such used to finding practical solutions in order to achieve their goals. The book was meant to inform the new apprentices what to expect when they came to the Kontor. Since so many of the new apprentices came from German speaking areas, this information must be given in German. Therefore, the book is designed pedagogically, with every uneven page written in Danish, and every even page written in German – as if it was a language learning text book (and to some extent, it may have served as that as well). The educational system of the Hanseatic Kontor and later the Bergen Kontor was an important part of the recruitment of German speaking men to the city. The Hanseatic Kontor only recruited German boys; the Bergen Kontor recruited boys from German, Danish and Norwegian areas. It seems from our data (see case study 2 below) that the local boys generally came when they were 11–12 years old, whereas the boys who came from further away were older, 17–18 years of age when they came to Bergen to serve as an apprentice. After some years of apprenticeship, any boy had the possibility to advance and get a position as a foreman, either in the same house where he had served as an apprentice, or in another trade house. But first he had to go through an examination and give an oath to never deal in spoiled fish, and he had to pay for his testimony. All this is thoroughly described in our bilingual book from 1754. Another point to take into consideration is the degree of literacy among the apprentices. If this book should be of any use to the new apprentices, we must assume that they all could read on their arrival to the Kontor. And indeed this seems to be the case for some of the boys, but it also seems that there were boys who needed to learn the skills of writing and elementary mathematics – both essential skills in a merchant’s life. In one passage in the book, it says:



Language choice in forming an identity 

 169

(2) De af saadanne Drenge, som forinden de komme til Contoiret, ikke tilgavns kunde skrive og regne, skal Hosbonden være forbunden til at lade gaae i skole, i særdeleshed om Vinteren, naar mindst ved Contoiret falder at forrette, og de som haver lærdt det før, skal ved alle Leiligheder tilholdes, at de jefnlig øve sig derudi (p. 88) 3 (D) [These boys, who before they come to the Kontor, cannot write and calculate well enough, the merchant has a responsibility to send to school, especially in the winter, when there is least to do at the Kontor. Those who have already learnt this, must at all times be imposed, that they regularly practice these skills]. In the mid-18th century, it was still possible – perhaps even common – in Norway that children learned to read but not to write (Apelseth 2007). After the Christian Confirmation was introduced in 1736, it became obvious to the authorities that the children needed to be able to read, so that they could learn the Christian curriculum required to be granted the confirmation. Therefore, in 1739, school became obligatory for all. This early school could last just a couple of weeks during the winter, when the parents could manage without the working hands of their children. This means that the differences in what was actually achieved in the schools around the country varied greatly according to geography and social group (Høverstad 1918: 6–12; Skard 1972: 77–78). The fact that all apprentices at the Kontor had to learn to write as well in addition to reading means that the boys educated there had an advantage over many other Norwegian children from the lower classes.

5 Case study 2: the language choice of the apprentice testimonies Apprenticeship lasted – according to the bilingual book discussed in the section above – for 6 years. In reality, however, it lasted until someone, either the merchant that the boy already was connected to, or another merchant, was willing to hire him as a foreman. This could be less than six years or more than six years, depending on the age and the abilities (and the luck?) of the boys. We have two interesting books that give us information about these matters and on the language of the testimonies that the boys received after their exam (the books are also used by Wiberg 1945). One of these books is called Drengenes bog (D) ‘The 3 All examples are given in their original spelling.

170 

 Agnete Nesse

book of the apprentices’, it includes entries from 1755 and until 1837. The other book is called Examinations protocol hvor udi gesellernes navne reverser og testimonia indføres (D) ‘The examination protocol where the names, declarations and testimonies are written’, which also includes entries from 1755 and until 1837. Where the first book gives information about what language the boys wanted their testimony to be written in, the second book has copies of the actual testimonies, which were written in either Danish or German. In both books the place of birth of the boys is given, together with their father’s name and occupation, the age of the boy and how many years he had served at the Kontor. Both of these books are written in Danish, in line with the official policy of the Bergen Kontor of 1754. Likewise, the declaration that the boys signed before they received the testimony is always in Danish, no matter what language was used in writing the testimony. These declarations were in most cases signed by the boys themselves, but written by someone else. From the expressions used in the Apprentice-book, it seems clear that the boys could decide for themselves which language they wanted their testimony to be written in. This may come as a surprise, since we have discussed how strongly the leaders of the Bergen Kontor were in favour of using Danish as much as possible, to state their loyalty to the city and to the king. In such a sociolinguistic environment, looking at the choice of the apprentices can shed light on how they perceived the situation. Did the apprentices go along with the linguistic policy of their leaders? Apparently they did not. The general pattern when it comes to the language choice of the boys is as follows: Boys born in Germany (this includes one boy born in “Swedish Pomerania” an area that today lies on both sides of the Polish-German border along the Baltic Sea), wanted their testimony in German. Boys born in Norway outside Bergen (this includes one Danish boy), wanted their testimony in Danish. This comes as no surprise. The group that acts less predictable is the one consisting of boys born in Bergen. Quite a number of these boys wanted their testimony in German, even though there is no reason to believe that they did not have Norwegian as their mother tongue (the language of the families, especially the role of the mothers, is discussed in Nesse 2007). These boys had in common that they were sons or grandsons of German immigrants, either merchants or handicraftsmen. For example, we find the merchants Mowinckel and Dwerhagen and the herringcontroller Kehlenbeck among the fathers of the Bergen-born apprentices who wanted testimonies in German. For these boys, we must assume that the language choice was caused both by respect for their fathers and their backgrounds, and by an idea that a German testimony would work to their benefit if they chose to seek employment in a German city, like many young men from Bergen did. The leaders of the Bergen



Language choice in forming an identity 

 171

trade office had extensive contacts in the German cities, and were in favour of the next generation being able to nurse these contacts. They therefore applied linguistic pragmatism when writing the testimonies, even if they could be linguistic purists towards their colleagues, as shown earlier. A last point to be made about the exams, before we turn to the language ideologies of the next generation of merchants, is the language of the actual exam. The exam was oral, and partly practical, so language choice must have been an issue here too: (3) ... von uns examineret worden, und beydes durch mundlich Beantwortung und Handgreiflichs Proben sein Geschicklighkeit dergetalt beweisen hat dass wir ihn für tüchtig befunden haben als gesell zu dienen, Er auch eins Kunftige einnen Contoirischen Handel als Handels Verwalter vorstehen könne ... (G) [... by us is examined, and both by oral answering and practical tests has proved his capability, that we consider him fit to serve as a foreman, and also that he in the future can lead a trade house at the Kontor ...] The oral part had to do with knowledge of fish and trade. The practical exam had to do with hands-on knowledge of fish and other products. The boy should for example know how much salt one needed for the preservation of herring (Wiberg 1945: 74). The commercial side of the foreman’s work was equally important, and by examining a dried cod, the boy should be able to tell where in Europe it could be sold; knowing the preferences of the different markets was an important part of the knowledge acquired through the apprentice years. It is also interesting to look at the names given to the different varieties in the light of the research about what languages are called in this period before the national romantic ideology tied language to the alleged national personality of a group. In one case the apprentice is supposed to have asked to bergensk at examineres (D). In today’s language, this actually means that he was to be examined in the Bergen dialect, but we must assume that bergensk in this case does not mean dialect, it simply means “not German”. So far we do not know if the linguistic pragmatism4 went as far as to the actual examinations. The question would then be if the boy could choose not only which language he himself would speak, but also which language the examiners would speak. Winge (1992: 244–245) cites an example from 18th century Denmark where an oral exam is in fact bilingual, so the questions are asked in Danish, and the answers are in German. Most likely, such bilingual exams were also common in Bergen.

4 Linguistic pragmatism in one special text type, namely the Neighbours’ books, is discussed in Nesse (2012).

172 

 Agnete Nesse

There are not many examples of boys born in Germany who decided to have their testimony in Danish. The very last entry in the examination protocol from 1837 gives an indication that this choice did not exactly go together with ambition and a quick advancement in the system. The 37 year old Godtfried Kuhnle, born in Beutelsbach in Würtenberg as son of a wine gardener, served as an apprentice for 20 years before he had the opportunity to take his exam and work as a foreman. This German-born man wanted his testimony in Danish. He worked for a widow and may have had hopes to be able to marry into the business and spend the rest of his life in Bergen. Many widows who owned business married their foreman; the foreman would then advance to merchant.

6 Case study 3: new Norwegian patriotism In the long run, the official policy of the leaders of the Bergen Kontor won. A number of apprentices with German ancestors who wanted testimonies written in German, did not stop the change from German to Danish as the major written language among the traders in Bergen. One obvious reason for this was the demographic pattern. The Norwegian population grew during the 19th century and local boys outnumbered the German boys at the Kontor. They were younger, and had to be paid less. Also, new working markets had opened up to adventurous youth from both Germany and Norway in the USA. Following this, the population of Bergen became increasingly monolingual Norwegian (spoken)/Danish (written). German was still used in church and in those trade firms where there was a German merchant, but as the group became smaller, the opposition in the city to bilingualism became stronger (the problems with maintaining German in church is discussed in Nesse 2007). In addition to a general assimilation of the inhabitants of German ancestry, it seems that the pragmatic merchants decided that being Norwegian was favourable on many levels, and their patriotism to their hometown Bergen grew strong. During the 19th century, Norwegian nationalism grew all over the country, inspired both by the independence from Denmark in 1814 and by the ideas of romanticism. The 19th century also saw a strong increase in the forming of organisations of social, cultural and political types. The merchants in Bergen were in front in many of these organisations, for example in forming musical and theatrical organisations, shooting companies and others, some founded already in the 18th century (Ertresvaag 1982). In 1868 the so-called Vestmannalaget ‘The society of Westmen’ was formed. This was a language organisation, promoting the study of Old Norwegian and



Language choice in forming an identity 

 173

the (formal and poetic) use of a New Norwegian based on Old Norwegian and the dialects of the Norwegian countryside. I will not go into the history of the forming of New Norwegian here; this history is well documented, for example by Hoel (2011). My interest is in the merchants who joined in the forming of the society of Westmen. The leader of the new language society was Henrik Krohn, an editor whose ancestors had emigrated from Germany as merchants. Of the 37 founders of Vestmannalaget, 10 were merchants and four were listed as foremen (Hannaas 1918: 17–18). This means that one third of the founders were employed with some sort of trade. Many of the founders, also those that were not merchants, had surnames that are well known from the immigrant history of the Kontor, such as Krohn, Ellerhusen, de Lange, Elbrecht and Lehmann. Our question is why practical, pragmatic merchants, and men with German ancestors would get so involved in this New Norwegian language project, a project that included a strong ideology of purism directed against German loanwords and German word formation elements. One of the goals in the founding of the organisation, was – in line with the romantic nationalism – to fremja Nordmannskapen i Landet (NN) ‘strengthen the Norsemanship in the country’ (Hannaas 1918: 19). It is likely that the story of the founders of the Bergen Kontor in 1754 and the story of the founders of Vestmannalaget in 1868 could shed light on one another. Linguistic innovation combined with tradition had proven to work well for the Bergen Kontor. In 1754, the founders of the Bergen Kontor can be seen as innovators in their insistency on the use of Danish, while they followed the traditional trade methods of the Hanseatic League. In 1868, Vestmannalaget can be seen as innovators in their insistency that the country needed a language that was not Danish. They were at the same time traditionally oriented, in that the New Norwegian should not be a modern language, but a traditional one. Just as Danish carried the important meaning “not German” in 1754, New Norwegian carried the important meaning “not Danish” in 1868. And just like pragmatism towards bilingualism led the way to monolingualism in the period 1754–1837, as we saw in the case studies, the men behind Vestmannalaget had the idea that pragmatism and bilingualism would lead the way in the end of the 19th century as well. Therefore, they studied how to write and speak New Norwegian, and they wrote poems and published papers in the language. In their everyday lives, however, they wrote Danish and spoke their Bergen dialect. For them, New Norwegian was an effective way to state their patriotism, and they combined the linguistic work with an interest in folk art such as costumes and paintings. Since they also had quite a lot of economic strength, they contributed to things like the rebuilding of the mediaeval hall at the Bergen fortress (Håkonshallen), which had been in decay for centuries (Hannaas 1918: 69–75).

174 

 Agnete Nesse

This time, the strategy of pragmatic bilingualism did not work that well. Winge (1992: 310) writes that “problemlose Mehrsprachigkeit” ‘Multilingualism without problems’ was no longer possible in Denmark by the end of the 18th century, but that this situation did not arise in most other European speech communities until after the Napoleonic Wars. In Norway, that is certainly the case. Time had come to make sure one nation had one language. And the New Norwegian movement in Christiania,5 which was larger and stronger than the one in Bergen, had a different, more monolingual attitude to the language question: Whereas the Bergen-based Vestmannalaget saw the prestige in the creation of a conservative language with ties to Norway’s glorious past, many spokesmen in Christiania wanted a New Norwegian based on the rural dialects and not so much on the Old Norwegian. These men in Christiania were also more specific in their demand that the New Norwegian should be the only language in Norway, not just an exotic language to enjoy in certain clubs. One of the most famous texts from the debate between the New Norwegian spokesmen in Christiania and Bergen was written by one of the Christiania men, Olaus Fjørtoft. In his article Nokre Or til Bondevenne og Maalmenn (NN) ‘Some words to the friends of farmers and language’ from 1871, he writes: (4) ...men Ulukka hev vel vore, at Mesteparten av disse Bergjensarane alder hev snakka Norsk, daa dei va smaa, Mesteparten e Storfolks Søne, som hev gjort Bykse mæ eingang fraa Salongmaal og te Bondemaal og dæfør hev dæ gaatt som dæ hev gaatt mæ. Og so e dæ dæ merkjelege mæ alle disse bergensiske Maalmenn, enten snakka dei i Live ompass kav Dansk elde so knota dei eit kunstigt Normalmaal kvar Gong dei stend opp og ska halde Tala i et Maal som korkje finst paa Himel elde Jor, korkje inne elde ute (NN) (from Torp and Vikør 2003: 173–176) [...but the disaster must have been, that the larger part of these men from Bergen have never spoken Norwegian when they were little. Most of them are the sons of big people, who have jumped at once from the language of the Boudoir to the language of the peasants, and that is why it has gone the way it has. And then it is the strange thing about all these New Norwegian men from Bergen: Either they speak naturally almost half Danish, or they mix an artificial New Norwegian every time they stand up to give a speech in a language that does not exist, neither in Heaven or on earth, neither inside or outside].

5 The city Christiania had the name changed to Kristiania in the late 19th century, and to Oslo in 1925.



Language choice in forming an identity 

 175

Unlike the tradition a hundred years previously, when terms like Norwegian, Danish and Bergen(ish) all meant “the language of the country” or “not German”, for Fjørtoft the distinction between the term Norwegian and Danish was crystal clear. For him, Norwegian and Danish were no longer synonyms, but opposites. His advice to the posh members of Vestmannalaget in Bergen was that they should start listening to the language of their servants in order to learn how to speak the new language in a natural way. Natural, however, does not seem to have been the goal for the variety preferred in Bergen. The merchants wanted to learn to speak like the saga heroes, not like the men loading and unloading their ships or the women washing their houses. Fjørtoft had a rural background, he was an anarchist who wrote in an almost orthophone language. For him, the New Norwegian movement was all about social equality and the rising of the status of the poor. This was not the project for Vestmannalaget, who saw the new language as yet another opportunity to rise socially in a patriotic way, and to define what was to be fashionable in speech and in writing. Fjørtoft and Vestmannalaget have later been viewed as the extreme opposites in the New Norwegian movement in the 19th century. The creator of what eventually became the official New Norwegian, Ivar Aasen, held a position in the middle, partly social, partly national in his direction (Venås 1996; Walton 1996). Eventually the Bergen merchants left the organisation they helped to create and Vestmannalaget was taken over by eager teachers and other learned men from the countryside. In the end, the Bergen merchants did what most Norwegians did; they settled in the Dano-Norwegian language that was created during the 20th century as a koiné between written Danish and Norwegian city dialects. This soon became the majority language, and is now – under the name Bokmål ‘book language’ – used by 90 % of the population of Norway.

7 Conclusion In Bergen, a town that from its foundation was concentrated around trade, the merchants were both innovators and carriers of tradition. Their approach to new ideas and – in our case – new linguistic practices, seem to have been steered by pragmatism: If it served the position of trade as such, or groups of merchants, it was worth trying out, either at the work place (case study one), in the education system (case study two) or as a leisure activity (case study three). Our first case study deals with the shift from German to Danish (written) and Norwegian (spoken), and strongly suggests that this happened mainly because the leading merchants in the late 18th and the early 19th centuries wanted to

176 

 Agnete Nesse

demonstrate a distance to the Hanseatic League, in order to make sure they were seen and treated as equal with the Norwegian inhabitants in their potential for gaining power in the city. With this, they succeeded. The German-­ speaking merchants, who had played such an important part in the city since the 14th century, were becoming old fashioned. The modern merchant used – at least officially – the language of the country. The second case study shows how the bilingual merchants applied a step-by step- method in order to secure both this new Norwegian-ness at the same time as making sure the boys examined were able to use their testimonies in order to apply for positions in Germany. It is no surprise that Bergen-born boys took this opportunity, and chose German as the language of the testimony; this shows that it takes time from the launch of a new language programme to the day it is achieved. The third case study shows how merchants in the 1860s again were at the linguistic forefront by cultivating what they perceived as a future Norwegian. However, whereas the change from German and into Danish a hundred years previously had happened first in the public domain and later in the private domain, the work to enhance a New Norwegian language took place in the private domain, as cultural activity alongside other cultural interests. It was hardly used for business writing at all. A fourth case study might fit in here, as it is clear that the business domain has been innovative in the ongoing shift from Norwegian and into English during the last 20 years. Luckily, we do not know the outcome of this yet.

Data The city archive of Lübeck: Archiv der Hansestadt Lübeck, Alte senatsakten, Externa Danica, 844. The city archive of Bergen: Bergen byarkiv, Det norske kontor, nordlandshandel, A-0585 and 585 H:4

References Apelset, Arne. 2007. Gamlelæraren frå Vinje. Allmugen og skrivekunna. [The old teacher from Vinje. The commoners and literacy]. In L. J. Larsen, P. A. Michelsen & D. Orseth (eds.), 18.06.46. Festskrift til Sveinung Time på 61-årsdagen, 166–179. Bergen: Bergen University College. Brattegard, Olav. 1932. Über die Organisation und die Urkunden des hansischen Kontors zu Bergen bis 1550. Bergen historiske forenings skrifter 38, 237–285. Bergen: Beyer.



Language choice in forming an identity 

 177

Braunmüller, Kurt. 1995. Semikommunikation und semiotische Strategien. Bausteine zu einem Modell für die Verständigung im Norden zur Zeit der Hanse. In K. Braunmüller (ed.), Niederdeutsch und die skandinavischen Sprachen II, 35–70. Heidelberg: Universitätsverlag C. Winter. Braunmüller, Kurt & Zeevaert, Ludger. 2001. Semikommunikation, rezeptive Mehrsprachigkeit und verwandte Phänomene. Eine bibliographische Bestandaufnahme. Arbeiten zur Mehrsprachigkeit, Volge B, 19. Bull, Tove. 1996. Språkskifte hos kvinner og menn i ei nordnorsk fjordsamebygd [Language shift by woman and men in a North Norwegian Sami village]. In E. H. Jahr & O. Skare (eds.), Nordnorske dialektar. [North-Norwegian dialects], 185–200. Oslo: Novus. Burkhardt, Mike. 2009. Der hansische Bergenhandel im Spätmittelalter. Handel – Kaufleute – Netzwerke. Köln: Böhlau. Dorian, Nancy C. (ed). 1989. Investigating obsolescence. Studies in language contraction and death. Cambridge: Cambridge University Press. Elmevik, Lennart & Ernst Håkon Jahr (eds.). 2012. Contact between Low German and Scandinavian in the Late Middle Ages. Uppsala: Kungl. Gustav Adolfs Akademien för svensk folkkultur. Ersland, Geir Atle. 1999. Inn i unionen [Into the union]. In G. A. Ersland & H. Sandvik (eds.), Norsk historie 1300–1625 (Vol. II in Norsk historie 800–2000), 15–138. Oslo: Samlaget. Ersland, Geir Atle. 2011. Byens konstruksjon. Varige spor i byens landskap. [The construction of the city. Lasting traces in the landscape of the city.] Oslo: Dreyers forlag. Ertresvaag, Egil. 1982. Bergen bys historie. Bind III. Et bysamfunn i utvikling 1800–1920. [The history of Bergen. Volume III. A developing city 1800–1920]. Bergen: Universitetsforlaget. Fossen, Anders Bjarne. 1979. Bergen bys historie. Bind II: Borgerskapets by 1536–1800. [The history of Bergen. Volume II: The city of the bourgeiouse, 1536–1800]. Bergen: Universitetsbiblioteket. Hannaas, Torleiv. 1918. Vestmannalaget i femti aar. [The Society of Westmen for fifty years]. Bergen: Chr. Madsens boghandel. Haugen, Einar. 1953. The Norwegian language in America: a study in bilingual behavior. Philadelphia: University of Pennsylvania press. Helle, Knut. 1982. Bergen bys historie. Bind I: Kongssete og kjøpstad. Fra opphavet til 1536. [The history of Bergen. Volume 1: Seat of kings and trading city. From the beginning and until 1536].Bergen: Universitetsforlaget. Hoel, Oddmund Løkensgard. 2011. Mål og modernisering 1868–1940. [Language and modernization 1868–1940]. Oslo: Samlaget. Høverstad, Torstein. 1918. Norsk skulesoga. Det store interregnum 1739–1827. [Norwegian school history. The great interregnum 1739–1827]. Kristiania: Steenske forlag. Jahr, Ernst Håkon. 1999. Sociolinguistics in historical language contact: the Scandinavian languages and Low German during the Hanseatic period. In E. H. Jahr (ed.), Language Change. Advances in Historical Sociolinguistics, 119–140. Berlin/New York: Mouton de Gruyter. Janson, Tore. 1997. Språken och historien. [The languages and history]. Falun: Norsteds. Nedkvitne, Arnved. 2012. A post–national perspective on the German Hansa in Scandinavia. In L. Elmevik & E. H. Jahr (eds.), Contact between Low German and Scandinavian in the Late Middle Ages, 17–38. Uppsala: Kungl. Gustav Adolfs Akademien för svensk folkkultur. Nedkvitne, Arnved. 2014. The German Hansa and Bergen 1100–1600. Köln: Böhlau. Nesse, Agnete. 2002. Språkkontakt mellom norsk og tysk i hansatidens Bergen [Language Contact between Norwegian and German in the Hanse era of Bergen]. Oslo: Novus.

178 

 Agnete Nesse

Nesse, Agnete. 2003. Written and spoken languages in Bergen in the Hansa era. In K. Braunmüller & Gisella Ferraresi (eds.), Aspects of Multilingualism in European Language History, 61–84. Amsterdam/Philadelphia: Jonn Benjamins. Nesse, Agnete. 2007. 1750–1850: The disappearance of German from Bergen, Norway. In S. Elspass, N. Langer, J. Scharloth & W. Vandenbussche (eds.), Germanic language histories ‘from below’ (1700–2000), 423–436. Berlin/New York: Walter de Gruyter. Nesse, Agnete. 2012. Four languages, one text type: The neighbours’ books of Bryggen 1529– 1936. In M. Stenroos, M. Mäkinen & I. Særheim (eds.), Language contact and development around the North Sea, 81–98. Amsterdam/Philadelphia: John Benjamins. Nicolaysen, Nicolay (ed.). 1878. Bergens Borgerbog 1550–1751. Kristianina: Werner & Co. Rambø, Gro-Renée. 2010. Historiske og sosiale betingelser for språkkontakt mellom nedertysk og skandinavisk i seinmiddelalderen: et bidrag til historisk språksosiologi. [Historical and social conditions for language contact between Low German and Scandinavian during the late Middle ages: a contribution to historical language sociology]. Oslo: Novus. Sandøy, Helge. 2011. Frå tre dialektar til tre språk. Språklig og ideologisk. [From three dialects to three languges. Linguistically and ideologically]. In G. Akselberg & E. Bugge (eds.), Vestnordisk språkkontakt gjennom 1200 år [West Nordic language contact through 1200 years], 19–38. Tórshavn: Fróðskapur. Skard, Vemund 1972. Norsk språkhistorie. Bind II, 1523–1814. [The history of the Norwegian language. Volume II, 1523–1814]. Oslo/Bergen/Tromsø: Universitetsforlaget. Sørlie, Mikjel 1957. Bergens Fundas. Bergen: Beyer. Torp, Arne & Vikør, Lars S. 2003. Hovuddrag i norsk språkhistorie. Main lines in the language history of Norway]. Oslo: Gyldendal. Venås, Kjell. 1996. Då tida var fullkomen: Ivar Aasen. [When the time had come]. Oslo: Novus. Walton, Stephen. 1996. Ivar Aasens kropp. [The body of Ivar Aasen]. Oslo: Samlaget. Warter, Per. 1995. Computersimulation von Wortverstehen am Beispiel mittelniederdeutsch-skandinavischer Sprachkontakte. In K. Braunmüller (ed.), Niederdeutsch und die skandinavischen Sprachen II, 71–124. Heidelberg: Universitätsverlag C. Winter. Wiberg, Christian Koren. 1932. Hanseaterne og Bergen. Forholdet mellem de kontorske og det bergenske bysamfund. [The Hanseatic merchants and Bergen. The relationship between those from the Kontor and the civic society]. Bergen: John Grieg. Wiberg, Christian Koren. 1945.Gesellstanden og geseller i Bergen. [The foremen in Bergen]. Bergen: John Grieg. Winge, Vibeke. 1992. Dänische Deutsche – deutsche Dänen. Geschichte der deutschen Sprache in Dänemark 1300–1800 mit einem Ausblick auf das 19. Jahrhundert. Heidelberg: Carl Winter Universitätsverlag.

Henrike Kühnert

9 From the synagogue to the market square: cardinal numbers in Older Yiddish Abstract: Older Yiddish texts, from the 15th to the early 18th century, display an astonishingly large array of expressions for cardinal numbers. They descend from the two main language components of Older Yiddish – German and Hebrew-Aramaic. Mercantile language probably played an important role for the development of this rich well of synonyms. The present article identifies (socio)linguistic factors that influenced the flexible system of cardinal use: language registers as mercantile or literary language, the dialect and the educational level of the writer. After an overview of the text genres and registers of the corpus under investigation, the distribution of the numeral expressions in the language registers of Western Yiddish will be presented. The findings suggest that many of the number digits have to be read as Hebrew elements, not as German-derived cardinals as in Modern Yiddish. The final section examines the influence of the mercantile register, which enriched common Western Yiddish.

1 Introduction Commerce is one of the most interesting sociolinguistic aspects of Yiddish history. The impact of mercantile Yiddish has been so strong that in popular opinion it is reduced to just that part of the language or even bracketed with thieves’ argot (Elyada 2012: 99, 104, 109–110). At the end of the 18th century, the activists of the Jewish enlightenment (Haskalah) introduced the term “Jargon”, with strongly pejorative connotations, to refer to Yiddish (Jacobs 2005: 54). However, in certain areas, the Western Yiddish commercial register was regarded positively not only by Jews but also by Christians. The most famous example is the slang of the Franconian village Schopfloch, which was the common trade language of Christians and Jews, based on mercantile Yiddish (Klepsch 2004: 4, 148). To this day Jewish diamond traders speak Yiddish, and in Antwerp, the metropolis of diamond commerce, that holds for business deals regardless of the ethnicity of the traders: “Jews and Indians have been working so closely together that many Indian traders have learned to speak Yiddish” (Siegel 2009: 101).

Henrike Kühnert, University of Jena DOI 10.1515/9781501503542-009

180 

 Henrike Kühnert

An important aspect of mercantile Yiddish, not to mention thieves’ argots and slang, is its connotation as a secret language. Traditionally, the religious lexicon and the vocabulary of trade had remained Hebrew-Aramaic (Lötzsch 1990: 6) and, as a rule, it was difficult for the Jews’ non-Jewish neighbours to understand that component of Yiddish. Numbers, for obvious reasons, tend to be secret elements, and form an important part of mercantile language. As Althaus (2004: 43–60)1 writes in his paper about Yiddish mercantile lexemes in German: “Daß Preisangaben intern mit jidd. Zahlen angegeben wurden, war nichtjüdischen Geschäftspartnern oft ein Dorn im Auge” (Althaus 2004: 43) [That prices internally were indicated with Yiddish numbers was often a thorn in the flesh of the non-Jewish business partners]. However, although the numerals used in secret argots and in the livestock dealers’ language (Klepsch 2004: 1474) originate from Hebrew, the use and frequency of Hebrew-derived numeral elements is not clear in common Yiddish. In Modern Yiddish only German-derived numbers are known. In older texts, numbers appear both as lexemes and as symbols in digits. The latter leaves ambiguity about the lexeme that lies behind it. Whereas ordinal numbers occur most frequently in dates, cardinal numbers play a significant role in the indication of quantities – essential for trade as in nearly every business transaction a sum of money is mentioned. This is why the present paper focuses on the Older Yiddish cardinal numbers. Examining many sources, I have conducted an analysis to determine from which language component – German or Hebrew-Aramaic – the cardinals in the main text types originate, with particular attention paid to the utilitarian prose. The corpus consists of Western Yiddish texts from the 15th to the early 18th centuries. This excludes the modern Eastern Yiddish period, which starts from the middle of the 18th century when Western Yiddish began to die away.

2 Text types, registers and styles of the corpus For this investigation, two groups of Western Yiddish sources are examined, namely literary and utilitarian prose. The data show a widely divergent use of numeral expressions between the two. This difference is related to the proportion of Hebraisms in each text type, since in texts with a large Hebrew component,

1 More about Yiddish mercantile lexemes in Althaus (2002: 38, 179–81).



From the synagogue to the market square: cardinal numbers in Older Yiddish 

 181

numbers tend to be borrowed from Hebrew, whereas in German-based sources they are derived from German. Therefore, the following representation of the corpus under examination considers not only the text types and language registers,2 but also goes into detail about the Hebrew component. Katz (1983: 1032) lists three types of stylised  language that proceed in different ways with the Hebrew-Aramaic element of Western Yiddish (Michels 2013: 26). In the first type, the Hebrew-Aramaic element is repressed and Hebrew and Aramaic words are actively avoided (germanised style), in the second type the proportion of Hebrew-Aramaic is very high (chancery style), and in the third type, in biblical translations, the Hebrew-Aramaic element structurally influences word-order and the choice of word stems. The literary register around 1600 – called “Literary language I” by Max Weinreich (1980) – is located between two of those stylised forms: the language of biblical translations and the germanised style, i. e. the Early New High German writing tradition in Hebrew script (see 2.2). The literary language, being the register of ideas, of religion, narration, history, fiction and poetry, differs considerably from utilitarian prose, which not only reflects the written language of everyday life but also relates to specialist areas such as law, society, commerce and medicine. Among such specialist text types, laws and decrees, directly drawn up by Jewish courts and institutions show linguistic peculiarities: they are influenced to a remarkable degree by  Katz’s third form of stylised language, the chancery style, which is characterised by a large Hebrew component (for the chancery style in general see Weinreich 1958;3 Kahan-Newmann 1990; Michels 2013: 26, 28). Furthermore, the chancery style had an impact on private and colloquial writing (Weinreich 1958 according to Klayman-Cohen 1994: 12–13) as well as on the written dimension of the mercantile language, though little is known about its effects (Michels 2013: 28). Finally, the mercantile register ranges from the specialist argot of professional traders and lenders to the colloquial language of everyday business on the market square. However, it is to be supposed that the most important part of mercantile life is played out in speech. This is the reason why in the compilation of the corpus under examination the colloquial text types have priority.

2 The term “register” will be used as synonymous for “(specialist) language”. 3 Klayman-Cohen (1994: 11–13) gives a précis of the Hebrew article.

182 

 Henrike Kühnert

2.1 Utilitarian prose In contrast to Early New High German linguistics, which can rely on a standardised text corpus (the Bonner Frühneuhochdeutschkorpus), the linguistics of Older Yiddish, up to now, cannot count on such a corpus.4 Generally, only a few older Yiddish texts are extant, but colloquial utilitarian writings are especially rare and no purely mercantile sources are available.5 Thus, the corpus compiled for this study comprises 16 texts and text collections that contain mainly almost all published private letters from before 1650 (Timm 1994: 449–50, see Tim) and witness statements (see 6.2).6 The earliest usable writings go back to the beginning of the 15th century and the memoir of the businesswoman Glückel of Hameln around 1700 (GlH) is the latest source. From the 18th century onwards, Jews in the West switched gradually to German, and Eastern colloquial writings seem not to be published. All the named text types show examples of traders’ language. However, the private letters probably are the nearest to speech. This text type has the advantage that the author, the exact writing date and the location are known. The letters are written by persons of both sexes and different levels of education. The handwritings, with rare exceptions, differ from letter to letter: these are not dictated texts but autographs, and have come down to us as originals. For the case of English, it has been proved that early private letters are much closer to orality than fictional genres and essays. In time, all genres evolve steadily towards oral styles, but letters show more preference for colloquial language in its different aspects (Biber and Finegan 1989: passim, esp. 508–510). With respect to Yiddish, it could be suggested that letters are less “authentic”, i. e., further from colloquial language, than English examples. This goes back to Uriel Weinreich, who in his 1958 article about the chancery style surmises the existence of a “prohibited” Yiddish-Hebrew language mix. It contains various grammatical elements from Hebrew, numbers among others, which are not usually found in literary sources. In Weinreich’s view, these forms got into private documents from the 16–18th centuries due to the influence of the

4 A survey of Older Yiddish text types can be inferred from Frakes (2004), but the texts are neither transcribed nor long enough for the present examination, and not available for electronic investigation. 5 The recently published “Jiddische Handschriften der Niederlande” (Yiddish manuscripts of the Netherlands) gives testimony about Yiddish account books belonging to the Jewish community as well as private ones (Michels 2013: 269–292). 6 Much of the utilitarian prose had to be romanised from the Hebrew alphabet, some from the non-published manuscripts, see 6.1.



From the synagogue to the market square: cardinal numbers in Older Yiddish 

 183

community scribes, who also taught writing (Klayman-Cohen 1994: 12–13). Uriel Weinreich generally explains the large amount of Hebraisms in private utilitarian prose as the impact of chancery style (Klayman-Cohen 1994: 68, 11). Thus, the prohibited elements would never have been part of the vernacular. But Klayman-Cohen’s investigation shows that Hebrew grammatical elements and  “numbers in any case” (1994: 68) belonged to the Hebrew component of Yiddish and were used as alternatives to synonymous German-rooted words (see section 4). Another text type is regarded as being very close to oral speech: it has been said that the witness statements within the Responsa literature (Rub) represent a verbatim record (Rubashov 1929: 116–117, Timm 1987: 373, Kahan-Newmann 1990: 35–36). They were recorded by the shammes, a beadle, who was a messenger of the court and took down the depositions (Weinryb 1972: 69). However, the syntax of the witness statements is not as close to the assumed vernacular as the letters are, which is due to the beadle’s instruction not only in Biblical, but probably also in Rabbinic and Mediaeval Hebrew (Kühnert & Wagner 2014: 133–134). It is to be supposed that also the numeral usage of the recorded witness statements was influenced by the chancery style (see 3.3.1 and 4). In her memoirs, Glückel from Hameln uses different registers, a stroke of luck for Yiddish linguistics. Some parts belong to edifying literature, but her language is more colloquial when she writes about business and everyday matters, which makes the text unique within the utilitarian corpus. Those text types, as they are close to orality, fall between two extremes: on the one side, texts as the laws of the Krakow Jewish community (KrT) are drawn up in the chancery style mentioned above. The Hebrew component in the Yiddish parts is substantial. Furthermore the texts are characterised by real intrasentential code-switching between Hebrew and Yiddish. The Krakow laws contain extracts of mercantile language, and this data can be used to make meaningful comparisons with the data gained from the previously mentioned sources (letters, Rub, GlH). On the other side, sources that follow the German-marked tradition of literary non-religious texts contain only few Hebraisms. The remedy books from 1474 and 1509 (RezB) are early examples. Timm (2013: 128*–129*) mentions that much of the content of RezB 1474 is adapted from a German-speaking vicinity.7 For instance, names of plants which were common in German are represented in their dialectal form. The remedy books do belong to the genre of utilitarian prose,

7 Language mix and style of the manuscript, however, are not homogeneous, and it probably goes back to different sources (Neuberg 2014).

184 

 Henrike Kühnert

but, like edifying literature, they give instructions and are not close to spoken language. Those texts were introduced in the corpus to obtain literary-influenced control data.

2.2 Literary texts Early Yiddish narrative prose, following old writing traditions, shows very few Hebraisms, with a maximum proportion of two percent. This holds also for rhymed sources of Jewish origin (Timm 1987: 374). In the course of the 16th century, the share of Hebraisms increased continuously. In belles-lettres and historiography, the average is 4–8 percent at the end of the century (Timm 1987: 372–373). Therefore, the oldest texts of the corpus were chosen from this time. Bible paraphrases and edifying literature contain more Hebraisms, and even the syntactic structure is occasionally influenced by Hebrew. This also applies to religious literature written for women and uneducated men. Those sources are valuable for linguistic investigation, as they show the higher social level of common language. These text types represent the centre of the literary language and form the chosen corpus of literary prose, which comprises a total of 20 sources, dating 1579–1834 (see 6.2). The later time period of the literary corpus compared to that of the utilitarian texts is justified as earlier literary writings probably do not show any Hebrew numbers. To understand the linguistic differences of the edifying and the narrative texts, one has to realise that they – though forming the core of written common Yiddish – are subject to contrasting influences: one of Hebrew-Aramaic, the other of German origin. The first is the stylised language of the Bible translations. Sometimes it even adjoins the biblical paraphrase, preceding it directly. But the influence of Hebrew on the language of Bible translation is not reflected in the share of Hebraisms. In fact, there are no Hebrew loanwords: the roots of the lexemes are all German. The tradition did not allow the translation of a Hebrew word by a Hebraism even if it was well-known in Yiddish (Timm 2005: 74–76). However, the syntactic structure of the translation corresponds in large part to Hebrew word order, forming non-grammatical sentences (Leibowitz 1931 passim, Timm 2005: 59–60). As for the vocabulary, the selection of the German stem was influenced by the Hebrew original: an important number of modern Yiddish words appear German without actually being German. Thus, the influence of Hebrew-Aramaic is important, but exclusively related to the semantics (Timm 1991: 59–63, 68–74; Timm 2005 passim, especially 61–69).



From the synagogue to the market square: cardinal numbers in Older Yiddish 

 185

German texts in Hebrew script dated not later than 1600 (Timm 1987: 372)8 show the second of the influences mentioned above. Whereas the belles-lettres of Jewish origin differ from German in syntactic and semantic features, many of the early narrative and rhymed sources are copies of German texts with few changes. They form the other extreme of non-usable sources for the linguistic examination of Older Yiddish, as they are very close to German. Apart from the transcription into Hebrew letters, they are barely adapted: they contain almost no Hebraisms, only a Judaisation took place and anti-Semitic and Christian contents were deleted (Friderichs 1981: 25).

3 Cardinal numbers in the corpus of Older Yiddish 3.1 Cardinals and language registers Three main types of cardinal expressions appear in the corpus: The first are German-rooted numbers as ain, zwaiʾ, dreiʾ; the second Hebrew eḥed ʻoneʼ, šnajim ʻtwoʼ, šloše̱ ʻthreeʼ; and the third are numbers homonymous with the names of the letters of the Hebrew alphabet, for instance bess ʻtwoʼ, caf ʻtwentyʼ. Furthermore, the combination of an alphabetic numeral9 with Hebrew meʾess ʻhundredsʼ could occur. In writing, these numbers can be represented fully spelled or as digit, i. e. abbreviated alphabetic numeral (see table 1): Table 1: Numerical expressions for ʻtwoʼ and ʻtwo hundredʼ and their writing conventions (last column) German-derived lexeme

Hebrew-derived lexeme

Hebrew alphabetic numeral

Hybrid

Digit

zwaiʾ zwaiʾ hunde̍rt

šnė šnė meʾess

bess reš

– bess meʾess

b` r` and b` meʾess

8 In Germany in about 1800, this tradition had a new impact, which is an influence of the Haskalah (Lowenstein 1979: 188 and passim). For instance, BoM 1798 is quite germanised and even the Pinkasim (record books) of the Jewish Community in Frankfurt/Oder have characteristics of the surrounding dialect (Litt 2007: 534). 9 With the term “alphabetic numeral” I always refer to the fully spelled-out letter number, not to the digit.

186 

 Henrike Kühnert

In the case of the hundreds theoretically four possibilities of oral expression existed. However, there is no evidence for further variants, for instance the combination of a Hebrew-derived cardinal or a alphabetic numeral with German-rooted hunde̍rt or touse̍nd. The opposite occurrence, the German single-digit cardinal + Hebrew meʾess, is also lacking, although this does not hold for the number one, which is homonymous with the indefinite article: ain meʾo existed (Lin 1562 I.1,8). The rule of phrase formation seems to be the same as generally in Older Yiddish:10 a Hebrew word could syntactically depend either on a Hebrew or a German lexeme, whereas a German word could depend only on a German one. Therefore, the more “lexical” or content word in a mixed phrase will be the Hebrew, and the “grammatical” or function word the German lexeme. From the data of dictionaries and glossaries, the following hypothesis (see figure 1), about which numeral expression is used in which language register, was drawn up:

Figure 1: Assignment of language registers to number expressions (hypothesis)

For the literary register, the Röll (2004) Glossary of four 16th-century Western Yiddish texts shows that German numerals are the standard, Hebrew cardinals an exception. Thus, only German cardinals appear in literary texts. Modern Yiddish is said to be the language of the prose authors of the 19th century, such as Sholem Aleichem. If they followed the Western Yiddish literary tradition, they had to choose the Germanisms because there were no other options. Actually, only ­German-rooted cardinals are used in Modern Yiddish (Niborski 1999). There is also an important written feature distinguishing literary from utility sources in general: the absence of number signs. In the colloquial language, as Friedrich (1784) and – for the Hebrew component – also Klayman-Cohen (1994) show, all numerical expressions are used. Dialectal and sociolectal characteristics have to be specified. 10 See also Klayman-Cohen (1994: 14, 15, 68).



From the synagogue to the market square: cardinal numbers in Older Yiddish 

 187

For business, alphabetic numerals are characteristic, as can be seen from Friedrich’s invoice (see 3.3.3). The same goes for digits in the written commercial language. In Reitzenstein’s (1764) horse dealer glossary the Hebrew numbers are included; thus they were probably part of the mercantile argot, too. As for the German-rooted numerals, traders certainly knew them although they probably did not use them when they had another choice. In the following section, two examinations considering text types will be made: First, I shall compare the use of the cardinal ʻtwoʼ in utilitarian prose and literary texts. The aim is 1) to test for the exclusive appearance of German-rooted numbers in the literary language, and 2) to quantify the distribution of the different cardinal expressions in utilitarian prose. Secondly, all fully written cardinals within the utilitarian prose will be counted and I shall discuss the diachronic, dialectal and sociolectal characteristics of the numeral usage.

3.2 The number ʻtwoʼ in literary and utilitarian texts The search for the cardinal ʻtwoʼ also included compositions such as ʻtwenty twoʼ, ʻtwo hundredʼ, etc. (see table 2). Table 2: Spellings/romanisations of ʻtwoʼ in the Older Yiddish corpus German-derived lexeme

Hebrew-derived lexeme

Digit

Alphabetic numeral

zwaiʾ, zwaiʾe̍n, zwėn, cvayʾ etc.

šnė, še̍ni, šeni, še̍nė, šny, šty

b`, 2`

bėss, bess, byt

In the corpus, different transcriptions for the spellings and morphological varieties of the German-rooted zwaiʾ, the Hebrew-rooted šnė and the number digit “bet  + geyresh (abbreviation sign)” are used. The fully spelled letter bess as a cardinal was not found (see table 3). Table 3: The expression of ʻ2ʼ in literary texts (1579–1834) and utilitarian prose (1400–1719) Text type Literary Utilitarian

zwaiʾ

šnė

b`

Total

1150 = 99 % 32 = 14 %

4 = 0,3 % 32 = 14 %

8 = 0,7 % 160 = 72 %

1162 224

188 

 Henrike Kühnert

In the literary texts, the German-rooted zwaiʾ appears more than 1150 times. Šnė is used only four times as a cardinal (and a few more times as an ordinal), especially in texts from the 18th century. And there are eight instances of the  digit “bet  + geyresh”. They date also from the early 18th century (Slt 1706, ShN 1707/1727) with the exception of three b/b` me̍ʾess ʻtwo hundredʼ in Mab 1602. no. 223. In verse, sometimes unusual words and forms are chosen because of the rhyme, but in the Hebrew word list of two large rhymed books from the 16th century (Timm 1996: 209–11, Schumacher 2006: 327–29) there are no Hebrew cardinals nor letter names with cardinal meaning. Similarly, excerpts from a calculation book contain exclusively German-rooted numbers (Arye Levi 1699, cited in Frakes 2004: 815–822). Not surprisingly, the complete data confirm the hypothesis that the vast majority of numbers in literary language, 99 percent, are German-rooted. This can be explained by the huge influence of the language of Bible translation on the Yiddish writing tradition. Similarly, the texts that were copies of German sources may have encouraged the perception that German numbers were the only appropriate option for the literary register. In utilitarian prose, the digit b` occurs much more often than in literary sources: 156 times, and in all texts with the exception of the short Gö and Dub. Certain parts of the German influenced RezB 1474 (nos. 508–608) do not contain digits, which coincides with other linguistic peculiarities of those remedies (Neuberg 2014). But generally, even the remedy books show b` (see also RezB 1474. 772, 774, 797/2, 810, 910, 917, and RezB 1509. 110r, 107v): (1)

loś štėn b` let stand 2 ʻleave it for 2 daysʼ RezB 1509. 110v/5r,7

jomim days

Thus, Hebrew number signs are a characteristic of Western Yiddish utilitarian prose. Zwaiʾ appears only 32 times, the same as šnė. There is a difference of 84 percent with respect to the literary texts. The vast majority, over 70 percent, are digits. Although it could be possible that the digit should be interpreted generally  as German-rooted zwaiʾ, this is not probable. It is to be supposed that Hebrew ­cardinals appear at least in the same proportion as German ones. Besides, alphabetic numerals existed as an alternative. Thus, it is even imaginable, that the abbreviations always had to be read as Hebrew-rooted numbers or as acronyms. Further details about the number usage depending on context,



 189

From the synagogue to the market square: cardinal numbers in Older Yiddish 

time, place and circumstances of origin are required before there can be a deeper examination.

3.3 Distribution of cardinals between the German and the Hebrew component in utilitarian prose All cardinals in the utilitarian texts that are written in full were investigated and assigned to the German or Hebrew component (see table 4). Only the number ain(s)/ ‫ אחד‬eḥed ʻoneʼ was excluded, as ain before a noun is identical with the indefinite article in Western Yiddish. Table 4: Fully written cardinals in the utilitarian corpus of Older Yiddish11 Text

Year

Short form

Germanderived lexemes

Hebrew-­ derived lexemes

Alphabetic numerals

 1. Remedies 1474/1509  2. Charge sheet 1518  3. Testimonies