Codification in the shadow of standards: ideologies in early nineteenth-century metalinguistic texts on Luxembourgish

ABSTRACT

Inspired by the ideological and multilingual turn in ‘third wave’ language standardisation studies (McLelland 2020, Walsh 2021), this paper demonstrates the value of these perspectives for historical analysis by exploring the implications of language ideologies for the early codification of Luxembourgish. As a ‘late’ standardised language (Vogl 2012), Luxembourgish provides a valuable case study for evaluating how existing powerful standard language regimes (Gal 2006) ideologically influence the discursive construction of a ‘late’ standard language, especially in multilingual borderlands. Ideologies of linguistic differentiation (Irvine & Gal 2000; Gal & Irvine 2019) are inherent in the standardisation process of Luxembourgish which sits between the Romance and Germanic language spheres of influence. The analysis focuses on metadiscourses of three early texts on Luxembourgish (Meyer 1829, Meyer & Gloden 1845, De la Fontaine 1855) in their discussions and proposals for codification. The diverse labelling of Luxembourgish in the texts forms part of a metadiscourse of differentiation and hierarchical contrast. Other core emergent discourses foreground affinities with Standard French and Standard German respectively but in differing ways that evoke the ideological notion of erasure. The final part of the analysis identifies further discourses that, in contrast, frame Luxembourgish as unique and different from other languages.

KEYWORDS:

• Luxembourgish
• standard
• standardisation
• ideologies
• codification
• metalinguistic
• discourse

The ideological and multilingual turn in language standardisation studies: applying the ‘third wave’ to historical analysis

This article analyses language ideologies in metadiscourses emerging from the early codification of Luxembourgish. The approach is informed by the recalibration of perspectives in language standardisation studies that have been collectively heralded as a ‘third wave’ in the discipline and are distinguished by a renewed ‘research agenda that seeks to understand how language standardisation ideologies and processes are discursively constructed and enacted in multilingual contexts’ (McLelland 2020, 9, emphasis in the original). By drawing attention to the influence of standard language ideology (Milroy & Milroy 2012, 18) on ‘non-standard’ language varieties, Walsh (2021, 773) demonstrates how the standardisation of such linguistic varieties takes place in the ‘shadow’ of other standards. According to Vogl’s (2012, 24) typology, Luxembourgish is categorised as a ‘late’ standardised language, because the process of standardisation for Luxembourgish did not gain momentum until at least the nineteenth century. The processes that give rise to the ‘late’ standardised languages are largely influenced by mirroring idealised models of ‘early’ standard languages, for example Dutch and Spanish, or ‘middle’ standard languages, such as Finnish and Norwegian (Vogl 2012, 22–4). This is in contrast to phases of standardisation in Early Modern Europe when Latin and Classical Greek were held aloft as highly prestigious linguistic forms to be emulated (Burke 2004, 89).

Each process of language standardisation inherently cultivates a hierarchisation of linguistic varieties, beyond which Vogl (2012, 26) identifies a further meta-hierarchical structure between the emergent standardised languages themselves. Languages undergoing standardisation towards the later or more recent end of the spectrum are more likely to be ‘disputed’ standard languages (for example Standard Macedonian and Standard Moldovan) rather than the longer-established ‘undisputed’ standards (such as Standard Dutch). This raises questions about the degree of legitimacy (Bourdieu 1991, 46–9) and authority (Gal & Woolard 2001, 7) bestowed on certain standard varieties, leading to feelings of linguistic insecurity amongst its speakers that can undermine the prestige and status of a comparatively recent standard. Language situations that are particularly fertile for these late standardisation processes are the ‘language borderlands’, where there has recently been an increased research focus (McLelland 2020, 10) in order to establish more concretely the roles of standards and ideologies for multilingual speakers. Intersections between culture and language along boundaries and borders have been a productive area for research that explores the discursive and ideological construction of languages in these frequently multilingual, linguistically fluid zones (for example, the edited volumes and collections by Peersman et al. (2014), Rutten et al. (2017) and Lane et al. (2018)). Against this background, this paper will focus on the early codification of Luxembourgish, which is situated in a territory that has historically been ‘subject to constant influences from the Romance and the Germanic languages, their cultures and politics’ (Gilles & Moulin 2003, 304).

While ‘third wave’ developments in the domain of language standardisation studies have largely centred on present-day processes, this paper also aims to demonstrate the value of applying these lines of enquiry to the historical analysis of standardisation. Specifically, the paper examines the metalanguage of codification in the early 19th century, a period ‘regarded as the cradle of standard languages and, most importantly, of the spread of the so-called standard language ideology’ (Krämer, Vogl, & Kolehmainen 2022, 13) and builds on an established body of research tracing the standardisations of languages in the 19th century within the ‘shadows’ of other contemporary standards in Europe, including Germanic languages such as Faroese (see Svabo Hansen et al. (2003, 160–1)), Romance languages such as Galician (see González Seoane and Rei-Doval (2019, 73–4)) and Slavic languages such as Macedonian (see Gjurkova (2016, 320)). The Euro-centric nature of these illustrative linguistic situations is also acknowledged and is not intended to imply that they are globally representative (see Smakman & Barasa (2016, 25) for criticisms of such generalisations). Sociohistorical approaches are valuable for understanding present-day standard language cultures and the concomitant ideologies, as Rutten and Vosters (2021, 86) exemplify for Dutch.

In order to examine the ideological construction of Luxembourgish through patterns of discourses in the texts that compare and contrast the nascent standard with Standard French and Standard German, the analysis is also informed by Gal and Irvine’s (2019) work (see also Irvine & Gal 2000) on the semiotic processes of differentiation. This approach helps to understand how the language ideologies surrounding early codification efforts for Luxembourgish are linked to the discursive differentiation of Luxembourgish from other standard languages. By bringing together fresh insights into ‘late’ standardisations in their respective multilingual and ideological contexts, based especially on the aforementioned conceptual developments by Vogl (2012), Gal & Irvine (2019), McLelland (2020) and Walsh (2021), a historical lens taking Luxembourgish as a case study will help discern the underlying ideological conditions that have given rise to recent standard language cultures in late modernity. After providing an overview of the linguistic situation in Luxembourg in the next part (Section 2), the methodology will be explained in Section 3. This will be followed by the discussion of the data in Section 4 and then concluding in the final Section 5.

The three selected metalinguistic texts on Luxembourgish (1829, 1845, 1855)

The following overview is intended to provide key historical context for introducing the metalinguistic texts and is not meant to serve as an exhaustive account of the development of Luxembourgish during this period, because comprehensive surveys have already been undertaken (see for example Moulin (2006), Péporté et al. (2010) and Gilles (2020)). Since they are amongst the earliest codification guides for Luxembourgish and have been particularly prominent in discussions about Luxembourgish (cf. for example Moulin (2006, 318–21)), the analysis will focus on the following three texts:

• Meyer (1829): book of poetry containing the first published metalinguistic comments on Luxembourgish and language guide

• Meyer & Gloden (1845): literary work containing a more in-depth and elaborated commentary on Luxembourgish

• De la Fontaine (1855): an authoritative proposal for Luxembourgish spelling rules

With Belgium, France and Germany as neighbouring states, the Grand Duchy of Luxembourg is located at a linguistic intersection between the Romance and Germanic language areas, which have historically exerted political and sociocultural influences on the region and continue to do so. Luxembourgish shares many structural linguistic features with the Moselle Franconian dialects of German (Gilles 2019, 1039) and has considerable lexical influence from French. The borders of present-day Luxembourg were determined by the Treaty of London in 1839 when Luxembourg gained full independence (Péporté et al. 2010, 5), precipitating a gradual rise throughout the 19th century in the sense of autonomy and identity. This growing national consciousness gradually became reflected in the regional linguistic variety spoken by Luxembourgers (Gilles & Moulin 2003, 304), materialising in the emergence of metalinguistic works on Luxembourgish from the 1820s onwards (Newton 2000, 136). During the period that the three texts under analysis were published, written communication was undertaken in Standard German and Standard French. Luxembourgish was primarily used for spoken communication Gilles (2020, 110–11). Literacy was taught by means of Standard German. Standard French was used for administrative functions and was also an important language of education in Luxembourg.

The first metalinguistic text included in this analysis is by Antoine Meyer (1801–1857) who was born in the city of Luxembourg and became widely regarded in the field of mathematics, especially in Belgium after studying the subject in Liège and later obtaining Belgian nationality (Moulin 2006, 318). His (Meyer 1829) poetry collection, E Schréck op de Lëtzebuerger Parnassus [A Step up the Luxembourg Parnassus], written entirely in Luxembourgish (including metalinguistic comments), spans 53 pages and is regarded as a founding work of Luxembourgish literature. The book begins with a four-page foreword outlining eleven points on Luxembourgish spelling and pronunciation to facilitate reading the poetry. The collection finishes with a nine-page afterword consisting of comments on Luxembourgish, including further guidance on grammar and spelling. Meyer’s use of Luxembourgish for metalinguistic commentary marks this work out as unique amongst contemporary works that have sections on Luxembourgish to contextualize their literary material. Others are written in German or occasionally in French, an example of the latter being Félix Thyes’s Essai sur la Poésie Luxembourgeoise [Essay on Luxembourgish Poetry] (1854), which similarly comments on Luxembourgish as part of a broader discussion of poetry. Although Meyer’s (1829) work did not initially receive much resonance or praise, the pioneering nature of the work cemented his status as a founding figure in Luxembourgish literature (Péporté et al. 2010, 234). The earliest ideas for a Luxembourgish spelling outlined by Meyer in this text are predominantly to facilitate reading the poetry in his collection (1829, iii), which ‘is traditionally considered the first “true” work in Luxembourgish’ (Péporté et al. 2010, 234).

Meyer later expanded on his 1829 work with a more substantial 204-page collection, Luxemburgische Gedichte und Fabeln [Luxembourgish Poems and Fables] (Meyer & Gloden 1845) which is also included in the analysis for this paper. Alongside poems and fables by Meyer, it contains lengthier metalinguistic comments and guidance for Luxembourgish written in German and provided by Heinrich Gloden (1804–1894). Born in Eech, Luxembourg, Gloden lived most of his life in Belgium, similar to Meyer (Muller 2021). Besides his work on this volume together with Meyer, one of Gloden’s most important works is a book of German poetry Eichenblätter (Muller & Weber 2022) undertaken together with his brother, Hubert, reminiscing on their childhood in Eech (Muller 2021).

The 1845 book by Meyer and Gloden begins with a 38-page initial section entitled ‘Grammatisches [matters of grammar]’. Comprising eleven chapters on Luxembourgish, this preliminary section respectively covers pronunciation, declination (of nouns, demonstrative pronouns and possessive pronouns), number, verb tenses, adverbs, prepositions, contractions, conjunctions and interjections. The explanations are in German. After the main section of poetry and fables, there is a 21-page glossary of ‘Einige Verdeutschungen [Some Germanisations]’. Their codification suggestions are intended to counter negative comments that disparage Luxembourgish, as Meyer & Gloden (1845, v–viii) explain in the preface. They add that an aim of the book is also to provide a grammar and spelling of Luxembourgish for the poetry contained in the collection, as well as to help with documenting Luxembourgish expressions. These metalinguistic sections are all covered in more detail in Section 4 of this article.

The first Constitution of Luxembourg in 1848 formally recognised French and German as official languages, although the lack of mention of Luxembourgish in the constitution itself indicates the pervasive perspective of the time that it was not yet widely acknowledged as a discrete language (Péporté et al. 2010, 255). With the codification of the Luxembourgish variety still in its initial stages, Edmund De la Fontaine (1823–1891, nom de plume: Dicks) introduced his influential Versuch über die Orthographie der luxemburger deutschen Mundart [Essay on the Orthography of the Luxembourg German Dialect] (De la Fontaine 1855) and this is the third text to be examined in detail in the article.

De la Fontaine is celebrated in traditional narratives recounting the origins of Luxembourgish literature, which include him in an influential trio, together with Michel Lentz (1820–1893, author of the Luxembourgish national anthem) and Michel Rodange (1827–1876, writer of Renert, a cornerstone of Luxembourgish literature). De la Fontaine, whose father was the State Minister for Luxembourg, Theodore De la Fontaine, studied in Heidelberg in 1846–1847, where he was influenced by contemporary views on linguistics from the Brothers Grimm and Karl Joseph Simrock (Péporté et al. 2010, 237–9). Amongst his activities and interests, De la Fontaine joined the Historical Section of the Grand-Ducal Institute, collected Luxembourgish sayings and published ethnographic works. A lawyer by profession, De la Fontaine also wrote plays and poetry in Luxembourgish (Muller & Weber 2022). At this stage, early writings on Luxembourgish were generally written by, and intended for, members of the educated middle-class (Gilles 2020, 113). He participated in discussions surrounding spelling rules for Luxembourgish and criticised some of the earlier proposed orthographies, including the rules suggested by Meyer and by Gloden (Meyer & Gloden 1845) because of the large number of diacritics required for that spelling system (Moulin 2006, 320–1).

De la Fontaine published his Versuch über die Orthographie der luxemburger deutschen Mundart in 1855 and expresses his intention in the preface (De la Fontaine 1855, 2) that he prefers a more phonetic basis for Luxembourgish orthography. He discusses Luxembourgish in the context of dialects and of ‘new’ languages (De la Fontaine 1855, 2) and prefers a spelling system that is not built on an etymological basis, which is a criticism he makes of the previous suggestions from Meyer and Gloden. De la Fontaine’s spelling proposal was immediately influential, being adopted by contemporary writers, including Michel Lentz. Written predominantly in German (using blackletter typeface for German, rather than Antiqua), his fifteen-page spelling guide consists of seven principal sections, followed by examples in the final three pages to illustrate the proposed orthography. The seven sections consist respectively of preliminary remarks, general rules, vowels, consonants, elision, harmonious letters/extra syllables and appendix.

A methodological framework for the historical analysis of ideologies in standardisation discourses

Research on the ideological factors that shape the conceptualisation and construction of languages, dialects or varieties throughout the processes of standardisation and ‘language making’ (Krämer, Vogl, & Kolehmainen 2022, 3) has drawn extensively on discursive approaches, for example the discursive construction of Scots (Unger 2013), Portuguese (Moita-Lopes 2014) and the Berber language, Tamazight (Khalid & Said 2023). Such research explores epistemological questions about the nature of expertise and authority on language (Gal & Woolard 2001, 32), observes the enregisterment of linguistic varieties (cf. Agha 2006, 190) and determines semiotic processes underpinning the indexical meanings embedded in linguistic differentiation (Gal & Irvine 2019; Irvine & Gal 2000).

Vogl (2014, 65) also adopts tools from discourse analysis ‘to explore and make explicit the language ideologies informing’ four texts (from 1992 to 2003) that give differing accounts of the history of the Dutch language. She investigates the ‘[s]tandard language ideology and the history of Romance – Germanic encounters’ (Vogl 2014, 1), making use of discourse ‘models’ (Gee 2014, 175) to identify the ‘typical stories’ that are narrated in the histories of Dutch that she examined. A discursive approach to analysis of language ideologies in this context is implemented by Breda and Krämer (2021, 128) who discern ‘discursive patterns’ in their corpus of online comments made on news articles published by Spanish outlets (2019–20) reporting on legislation regulating the use of the Basque language. The patterns they identify are based on the concept of ‘argumentation schemes’, used by Reyes (2019, 19) to analyse ideological language debates surrounding spelling reforms for the Spanish language proposed by the Real Academia Española (Spain’s official institution regulating the Spanish language). The argumentation schemes are in turn inspired by methods from discourse analysis and feature as part of the ‘Discourse Analysis as Ideology Analysis’ framework of Van Dijk (1995, 29).

Focusing similarly on overarching metalinguistic discourses in the texts, the analysis in this paper is also informed by ideologies of linguistic differentiation (Irvine & Gal 2000; Gal & Irvine 2019, 19) to discern how a standard for Luxembourgish is brought into being through semiotic processes of ideological representation by means of French and German. The next section (Section 4.1) will address first of all the labelling of Luxembourgish throughout the selected 19th-century language guides, taking into consideration how the practices of naming form a key component of the discursive construction of Luxembourgish, especially in terms of creating ‘a metadiscourse of comparison and hierarchical contrast’ (Gal 2018, 222). In Section 4.2, the analysis then moves onto discussing how features of Standard French or Standard German respectively are foregrounded in the discourses of the texts in their proposals for the codification of Luxembourgish, for example in terms of orthographical choices. The final part of the subsequent discussion in Section 4.3 draws attention to contrasting discourses in the texts that conversely frame Luxembourgish as unique and different from other languages, therefore unveiling further complexity and ideological tensions arising in the descriptions of Luxembourgish in these early codification works.

Discussion and analysis

Labelling and hierarchies: Luxembourgish as a dialect, language, ‘our German’

The practice of naming linguistic varieties is inherently ideological (Davies 2012, 56–7), whether it is through the act of acknowledging a form of language as a discrete entity or inferring some (high or low) status to a variety via the meaning and connotations of the chosen label (Wiese 2015, 344–5) (consider, for example, the different indexical values of ‘dialect’, ‘language’, ‘slang’, ‘standard’ or ‘Spanglish’ and so forth).

In Meyer (1829), the earliest of the three selected texts for the analysis, the naming of Luxembourgish is very varied and includes the labels ‘onz Dialekt [our dialect]’ (Meyer 1829, 1), ‘ons Sproch [our language]’ (Meyer 1829, 46), ‘vun onzer Mondaart [from our dialect]’ (Meyer 1829, 45), ‘lezeburger Deitsch [Luxembourg German]’ (Meyer 1829, 46) and ‘Volleks-Sproche [people’s language]’ (Meyer 1829, 47). Since ‘Dialekt’ and ‘Mundart’ are largely used interchangeably (Löffler 2003, 1–8), both terms have been translated in this article as ‘dialect’ in English. Meyer’s (1829) poetry collection is unique both in terms of writing all metalinguistic commentary in Luxembourgish and being the only work of the three analysed in this study that refers to Luxembourgish as a ‘language’ (for example, ‘Sproch’). Even accounting for potential fluidity in Meyer’s conceptualisation of language and dialect, as well as historical ambiguity surrounding these terms (see van Rooy (2020) for a comprehensive historical study of this distinction), Luxembourgish is not named as a ‘language’ in the two later metalinguistic works. In Meyer & Gloden (1845), Luxembourgish is labelled as: ‘unser Dialekt [dialect]’ (Meyer & Gloden 1845, vi), ‘unserm Idiom [our idiom]’ (Meyer & Gloden 1845, vii), ‘die lux. Leseart [reading Luxembourgish]’ (Meyer & Gloden 1845, viii) and ‘im Luxemburgischen [in Luxembourgish]’ (Meyer & Gloden 1845, xvi). De la Fontaine’s (1855) spelling guide is similar to Meyer & Gloden (1845) in that Luxembourgish is referred to as a dialect: ‘unsere Mundart [our dialect]’ (1855, 6), ‘der luxemburger deutschen Mundart [of the Luxembourgish German dialect]’ (1855, 2) and ‘Gebrauch unsers Dialekts [usage of our dialect]’ (1855, 2).

The examples establish that, besides the widespread use of ‘dialect’ for Luxembourgish throughout these three works, all three guides make a distinction between ‘our’ and other varieties of ‘German’, which was a common way of alluding to Luxembourgish (‘our German’) during this period (Gilles 2020, 113). Usage of the first-person plural pronoun and its counterparts (‘we’, ‘us’, ‘our’) typically form part of the repertoire for discursively constructing national consciousness through an imagined shared set of affinities (Billig 1995, 106; Anderson 2006, 6; Wodak et al. 2009, 45–7). In this case, the shared characteristic of ‘speaking Luxembourgish’ is established through distinguishing this variety from other types of German, through a processes of differentiation (Irvine & Gal 2000; Gal & Irvine 2019, 20). Luxembourgish is also distinguished from other varieties of German in these examples through being labelled ‘Luxembourg(ish) German’ or ‘the Luxembourgish German dialect’. Whereas languages and grammars in the European context of the late Middle Ages were largely discussed in relation to classical languages, such as Latin or Greek (Vogl 2012, 9), ‘late’ standard languages are described in respect of other dominant contemporary standardised languages, which, in the case of Luxembourgish, gives rise to naming practices that contextualise Luxembourgish within the ‘family’ of German. The act of naming the linguistic variety constitutes a preliminary stage in the standardisation process of Luxembourgish (Gilles & Moulin 2003, 307–8; Moulin 2006, 313) and recognising the complexity surrounding labels for Luxembourgish in these early stages helps with understanding the enduring multiplicity of terms used for Luxembourgish, as well as the range of perspectives towards its status as a language that continue to the present-day (Bellamy & Horner 2018, 339; Vari & Tamburelli 2020, 14).

Furthermore, the terms used for Luxembourgish in these guides suggests a hierarchization of varieties. Although Meyer (1829) occasionally refers to the Luxembourgish ‘language’ in his metalinguistic comments, the guides generally label Luxembourgish a ‘dialect’ or ‘our German’. A ranking of linguistic varieties along the lines of standard language ideology emerges, where culturally dominant standardised ‘fully-fledged’ languages (Vogl 2012, 24) are at the top (such as the Standard French and Standard German languages) and other varieties are positioned further down the scale (such as the Luxembourgish dialect). As Gal (2006, 171) observes: ‘[c]ontrary to the common sense view, standardisation creates not uniformity but more (and hierarchical) heterogeneity’.

Other metalinguistic remarks contribute to this creation of a linguistic hierarchy. Meyer (1829, 53) concludes the afterword with a lament of ‘den Aarmut vun onser Sproch [the poverty of our language]’ because of what he perceives to be the large number of homonyms which can hinder clarity, unless adequately distinguished by intonation. The notion of a deficit in Luxembourgish reappears in Meyer & Gloden (1845, xxvii) when the ‘beschränkter [more limited]’ verb tenses of Luxembourgish are characterised as a type of ‘Sprachmangel[s] [linguistic deficiency]’ compared to Standard German. De la Fontaine (1855, 1) opens with a comment that a ‘Mundart [dialect]’ has ‘keinen allgemeinen literarischen Werth [no general literary value]’ in contrast to ‘Schriftsprache [written language]’. Such evaluative comments resemble the contrasting pairs suggested by Gal (2018, 233) in the context of standardisation, minority languages and modernity, where standard language is seen as ‘literate’, ‘universal’ and representing ‘progress’, whilst ‘dialects, patois, minority and indigenous languages’ are conversely regarded as ‘oral’, ‘particular’ and ‘backward’.

The repeated comparisons drawn between Luxembourgish and the contemporary existing standard language cultures have been a consistent thread running through the history of Luxembourgish and still form the basis of current language ideological debates, for example controversies over Luxembourgish language-in-education policy (Weber 2016, 190–200) and continuing uncertainties surrounding writing Luxembourgish (Bellamy 2021, 696). In connection with this, Section 4.2 will explore more thoroughly how Luxembourgish has been continually constructed in the ‘shadow’ of other standards (Walsh 2021), emphasising shared linguistic characteristics in some respects (‘sameness’) or nurturing linguistic individuality in other aspects (‘difference’).

(In)visibility of features shared by Luxembourgish with French or German

In both the foreword and afterword of his poetry collection, Meyer (1829) mentions a number of languages, including accents of Swedish (Meyer 1829, 47) and the ‘sannefter [softer]’ consonants of Dutch (Meyer 1829, 50), although the overwhelming majority of comparisons are made with French and above all German. When introducing spelling rules, Meyer (1829, iv) confirms that ‘Ech haale’ mech sò vill als meglech oien der Orthographie vun den hòhdeitschen Mondaarten [I am keeping as much as possible to the orthography of the Standard German dialects]’. Standard German (‘hòhdeitsch’, literally ‘high German’) serves as the primary model for making comparisons with Luxembourgish in Meyer’s early work and this includes many references to German to explain spelling, pronunciation and aspects of grammar. The implied elevated status of the Standard German variety through the description ‘Hochdeutsch’ (in German) or ‘hòhdeitsch’ (in Luxembourgish), which can both be translated as ‘high’ German, has been shown to have undergone ‘a reinterpretation from a geographic characterisation to a qualitative ranking’ under the influence of standard language ideology (Wiese 2015, 345–6).

Meyer & Gloden (1845) also describe Luxembourgish mostly through comparison with Standard German. A representative example is the opening line of chapter two on ‘Mitlauter [consonants]’, which begins: ‘Wir übergehen b, p, c, d, t, f, w, k, r, s, ss, x, z als gleichlautend mit denselben hochdeutschen [We shall skip b, p, c, d, t, f, w, k, r, s, ss, x, z as they have the same sounds as Standard German]’ (Meyer & Gloden 1845, xiii). Since Luxembourgish is largely framed as similar to German, the remainder of their chapter focuses on any sounds and spellings that are of interest precisely because they differ from Standard German, an approach typical for most of the concepts explained in their guide.

In De la Fontaine’s (1855) proposed spelling rules, a close affinity between Luxembourgish and German is already given prominence by naming Luxembourgish in the title as the ‘luxemburger deutschen Mundart [Luxembourgish German dialect]’ (cf. Section 4.1). When explaining the different typefaces used for German and Luxembourgish (more details on the choice of typeface in Section 4.3), De la Fontaine (1855, 3) adds that ‘[d]ie Aussprache dieser Buchstaben behält nichts destoweniger ihren ursprünglichen deutschen Charakter [The pronunciation of these letters nonetheless maintains their original German character]’. In describing the proposed spelling rules, comparisons are drawn principally with Standard German, for example the explanation of consonant doubling: ‘Als wirkliche Verdoppelungen gelten und ebenfalls, wie im Hochdeutschen, ck für kk und tz für zz [ck for kk and tz for zz are, just like in Standard German, regarded as genuine doublings]’ (De la Fontaine 1855, 8). Likewise, the rules on splitting syllables, initial letter capitalisation, hyphenation, punctuation, spelling of proper names and spelling of foreign words are all described as ‘dieselben wie im Hochdeutschen [the same as in Standard German]’ (De la Fontaine 1855, 13).

Luxembourgish is, then, largely described in terms of close kinship with Standard German. The similarities are given prominence and hence more visibility in the descriptions, whereas less attention is drawn to differences between the two languages. Foregrounding linguistic homogeneity in this way by overlooking difference is an ideological semiotic process of erasure (Irvine & Gal 2000, 28–9; Gal & Irvine 2019, 20–21) and resembles the prevailing contemporary view in the 19th century that Luxembourgish was a dialect of German and therefore ‘was clearly seen as being directly dependent on and dominated by the German standard language’ (Gilles & Moulin 2003, 306–7). Although a national consciousness had been growing since the formation of the Luxembourg state in 1839, the strength of national identity amongst Luxembourgers in the first half of the 19th century is still a matter of debate (Péporté et al. 2010, 234). At this stage there was little intent to construct Luxembourgish in opposition to Standard German, since ‘Luxembourgish was considered a dialect of German’ (Gilles & Moulin 2003, 307). It is not until the 20th century that efforts gathered significant momentum to differentiate Luxembourgish substantially from German in the elaboration stage of the standardisation process (Moulin 2006, 313).

Parallels with French are drawn in codification proposals by the texts but fewer than with German overall. One of the few examples where Meyer (1829) positions Luxembourgish within the context of French is in point 11 of the Foreword: ‘Den ïe get wé de franzéschen ie, am Wuurt fièvre, ausgespracht; den oie sprecht mer aus graad wé d’fransécht substantif, une oie [The ïe is pronounced like the French ie, in the word fièvre [fever]; we pronounce the oie like the French noun, une oie [a goose]]’ (Meyer 1829, vi). In Meyer & Gloden (1845), Luxembourgish is also occasionally described in terms of its relationship with French, although again less often than the resemblances which are drawn with German. One such example emerges when discussing the tenses in the subsequent section ‘Bemerkungen [notes]’ (Meyer & Gloden 1845, xxix–xxx), nearly all the comparisons are with Standard German, with the exception of the verbs ending in the Romance-influenced suffix ‘éeren’, such as ‘absolvéren [to absolve]’, which is explained as ‘vom französischrn [sic.] hergeleiteten [derived from French]’.

A decade later, De la Fontaine (1855), whose first language was most likely French (Péporté et al. 2010, 238), explains Luxembourgish in relation to Standard German (Hochdeutsch), with only occasional mentions of French (Französisch) alongside the examples from Standard German. For example, the pronunciation represented by the character <ü> is described as sounding like ‘dem hochdeutschen ü und dem französischen u [the Standard German ü and the French u]’ (De la Fontaine 1855, 8). There are scarcely any places where De la Fontaine refers exclusively to French when presenting his proposed orthography for Luxembourgish. One example where this does occur is: ‘g und j nehmen in einigen Fällen die französische Aussprache an [g and j adopt in some cases the French pronunciation]’ (De la Fontaine 1855, 7). This likewise emerges in the section on consonants: ‘Ausnahmsweise finden wir ch in Wörtern französischen Ursprungs, wie charmant [As an exception, we find ch in words of French origin, like charmant]’ (De la Fontaine 1855, 9). However, in both of these last examples the role of French is downplayed with the remarks ‘in some cases’ and ‘as an exception’. Despite the lack of specific mentions of French in explaining the codification of Luxembourgish, French is still visible on a graphemic level. Indications of this are the integration of diacritics around the character e, such as the circumflex <ê> and accents <é> and <è>, into the Luxembourgish orthography to indicate certain long vowels (De la Fontaine 1855, 6–7) which are clearly not derived from Standard German orthography. These characters were also already present in the spelling proposal by Meyer & Gloden (1845).

From the perspective of standard language cultures and their influence on ‘late’ standardisations, the choice of letters and diacritics is another indication of how the codification of Luxembourgish has been guided by the contemporary dominant standard language cultures of French and German and the powerful influence of their respective orthographies. By introducing these characters, the guides’ orthographies foreshadow the continual back-and-forth with Luxembourgish spelling over time, which has gravitated towards German in some periods and then contrastingly veered towards French in others, in line with contemporary sociopolitical developments. One of the most notable later examples is the end of occupation by Nazi Germany during the Second World War, when an alternative spelling system for Luxembourgish was introduced. The Margue-Feltes system, introduced through legislation in 1946, was deliberately designed to expunge iconic orthographical features of German (from the former Welter-Engelmann system since 1914) as a means of creating cultural distance from Germany and the German language in the immediate post-war period. ‘German’ characters such as <ä>, <ß> and <sch> were removed from the official Luxembourgish spelling and the letters <ê> and <è>, originally in De la Fontaine’s (1855) proposal, were reinstated (Gilles 2019, 1051). This constitutes another facet of ideological linguistic differentiation (Sebba 2012, 4) where writing systems become iconic through their indexical relationship with groups and social characteristics (Irvine & Gal 2000, 37).

Constructing Luxembourgish as unique and individual

Having analysed the ways in which Luxembourgish is constructed in the texts as similar to existing standard languages, this section will turn now to other, seemingly contrasting, narratives which highlight the individuality of Luxembourgish and position the language as unique.

Although in Meyer (1829) Luxembourgish is generally described in relation to German and, to a lesser extent, French, there are also instances where he remarks on ‘der lezeburger Individualitét [the Luxembourgish individuality]’ (Meyer 1829, 46) and ‘daat, waat onser Sproch eegen as [that which makes our language unique]’ (Meyer 1829, 47). When discussing verb tenses, Meyer acknowledges both the similarities and the differences between Luxembourgish and Standard German: ‘Och waat Zéitwiirder oie’belangen, hoiet ons Sproch eng séer gròs Aenlechkeet mat der Hòhdeitscher; nemme’, waat d’Hellef-Zeitwiirder oie’belangt, se’ mer vil verschiden, besonnesch waat d’Zoiel vun dezen oie’geeht [As far as verbs are concerned, our language bears a very similar resemblance to Standard German; as far as the auxiliary verbs are concerned, they differ considerably, especially with regard to the quantity of them]’ (Meyer 1829, 53). So, in addition to the resemblances that Meyer outlines between Luxembourgish and other (standardised) languages, there are also places where Luxembourgish is differentiated from these other dominant standard varieties and its perceived distinct qualities are emphasised.

In Meyer’s (1829) guidance notes and in De la Fontaine’s 1855 spelling guide, German is distinguished from French and Luxembourgish by being presented in a blackletter typeface, whereas in Meyer & Gloden (1845) an Antiqua typeface is used for all text, including all words in French, German and Luxembourgish. The choice of typeface is significant because there are ideological implications for using blackletter for German, where conventionally ‘[l]oanwords and even loan morphemes were typeset in roman, while everything that was considered “proper German” was typeset in blackletter’ (Spitzmüller 2012, 262–4). This is the case in Meyer (1829) and in De la Fontaine (1855), where Luxembourgish is typeset in roman (as is French) but Standard German in blackletter. Furthermore, a contemporary debate was taking place (Newton 2000, 191; Spitzmüller 2012, 264) over the most suitable type to be used for German, and the ‘Fraktur-Antiqua dispute’ grew particularly intense in the 19th century, with observable ramifications for these metalinguistic texts. The usage of an Antiqua typeface for Luxembourgish in Meyer (1829) and De la Fontaine (1855), alongside a blackletter typeface for German, is another factor that frames Luxembourgish as different from German. In fact, this has the consequence that Luxembourgish instead shares the same typeface as French, drawing a parallel therefore with French rather than with German.

As in Meyer (1829), the later language guide by Meyer & Gloden (1845) also suggests the uniqueness and individuality of Luxembourgish at various points. The title page of the book includes a statement explaining that the book tackles the ‘dem Dialekt mehr oder weniger eigenartigen Ausdrücke [more or less idiosyncratic expressions of the dialect]’. The final paragraph of the Foreword (Meyer & Gloden 1845, viii) posits that reading Luxembourgish can present ‘manche Eigenthümlichkeit [some peculiarities]’ as well as ‘Schwierigkeiten [difficulties]’. Another similarity with Meyer’s (1829, 50) description of Luxembourgish is the description of the consonants as different from German, in the sense that they are ‘gemildert [softened]’ (Meyer & Gloden 1845, xxiv). Meyer (1829, 50) likens Luxembourgish consonants to the ‘sannefter [softer]’ sounds of Dutch, rather than German. Another sound depicted as distinctive in Luxembourgish is the one represented by <ó>: ‘Scharf, dem luxemburgischen Dialekt eigen [Harsh, unique to the Luxembourgish dialect]’ (Meyer & Gloden 1845, xii). Such exoticisation of ‘non-standard’ varieties was not unusual in 19th-century efforts to document dialects in Europe, and such comments on their ‘peculiarities’ constitute another facet of the powerful normative authority of standard languages in this period (Vogl 2012, 21). Similarly, such contrasts are evocative of the dimension of differentiation suggested by Gal (2018, 233): the axis of modernity which Gal discusses in relation to standardisation and minoritised languages. The discourses around Luxembourgish convey the meanings of ‘particular’, ‘various’ and peripheral’, which are discursively in opposition to modernist values attributed to standard languages such as ‘universal’, ‘homogenous’ and ‘central’ (Gal 2018, 233).

The representation of French in the guides is particularly ambiguous because the language is presented as similar to Luxembourgish in some respects (as shown in previously) but as foreign(ising) with regard to Luxembourgish in others, especially in the later two guides by Meyer & Gloden (1845) and De la Fontaine (1855). Chapter 3 in the Meyer & Gloden (1845, xxiv) guide begins by presenting the declination of nouns in Luxembourgish. When remarking on ‘Die fremden Ausdrücke mit folgenden Endungen [the foreign expressions with the following endings]’ (Meyer & Gloden 1845, xxiv), the listed examples with ‘foreign’ endings are linguistically Romance-oriented in nature: ‘z. B. Musekant-en, Stodent-en, Cadett-en, Cokett-en, Ballong-en, Cosöng-en, Zaldot-en, Notär-en, Mär-en, etc. [e.g. musician, student, chap/lad, flirt, balloon, cousin, soldier, notary, mayor, etc.]’. However, in other parts of the 1845 guide, the assumed uniqueness of Luxembourgish is in fact described in terms of tendencies that are more French-leaning. The sound represented by the character <é> is presented as ‘dem luxemburgischen Dialekt eigenthümliche Aussprache [a pronunciation that is specific to the Luxembourgish dialect]’ (Meyer & Gloden 1845, x) and yet ‘Hat viel ahnliches mit dem französischen é in parlée, trainée [Has much similarity to the French é in parlée, trainée]’ (parlée and trainée are feminine singular forms of the past participles: parler, to speak, and trainer, to drag). The symbolic link to French through the iconicity of <é> has already been discussed to some extent above, but here it is portrayed as another peculiarity of Luxembourgish. In the same section on vowels, a further parallel is drawn between a unique aspect of Luxembourgish and its relationship with French: ‘ue ein unserm Idiom eigenthümlicher Laut hat etwas ähnliches mit dem französischen oi [ue an idiosyncratic sound in our idiom that has a similarity with the French oi]’ (Meyer & Gloden 1845, xii).

In describing his orthographical rules, De la Fontaine (1855) generally positions French as linguistically more distant from Luxembourgish in comparison to the close relationship he identifies between Luxembourgish and German. When criticising the unsuitability of spelling variants that had until then been used for the sound represented by the combination <uo> in his spelling guide, he remarks:

So findet man dafür oi und ue. Diese Orthographie ist ganz zu verwerfen, weil oi zu französisch dem Geiste der Sprache fremd ist, und ue, das ohnehin den Laut nur höchst mangelhaft wiedergibt, zu leicht mit ü verwechselt werden könnte. [So we find oi and ue used for it. This spelling should be discarded entirely because oi is foreign, being too French for the spirit of the language, and ue, which in any case is a terribly inadequate reproduction of the sound, could be too easily confused with ü.] (De la Fontaine 1855, 8)

These seemingly divergent representations of Luxembourgish, on the one hand described as similar to Standard French and Standard German, yet also sufficiently different from them, discursively reflect the multiple and not always mutually compatible conceptualisations of Luxembourgish at the time:

In contrast to the pluricentric languages of their German and French neighbours, Luxembourgish was a ‘language’ that was unique to Luxembourg. It is characteristic of 19th century descriptions to recognize the dependence of Luxembourgish on German and, on the other hand, to emphasize that Luxembourgish has developed into an important factor of national identification and integration. (Gilles & Moulin 2003, 307)

Whilst discussing the tensions around the development of Luxembourgish in the early 19th century, especially in the sense of both belonging to, and being separate from, the Romance and Germanic cultural spheres in which it is situated, Moulin (2006, 309) notes how they are prescient of the subsequent rise of the Mischkultur (mixed culture) concept that emerges later in the work of Luxembourgish writer, Batty Weber (1909). The Mischkultur approach embraces the idea that Luxembourgish national identity is founded on bridging the geographical, cultural and linguistic intersections between France and Germany, rather than being defined through other nation-building traditions that promote independent cultural homogeneity and distinctiveness (Hobsbawm 1992, 51). This is one of the strands of the historical ‘Luxembourgish master narrative: the balance of the German and the French languages (often associated with German and French culture)’ (Péporté et al. 2010, 261). The opposing perspective of particularism counters the Mischkultur view by prioritising instead the ‘uniqueness’ of Luxembourg. The particularism narrative therefore tends to link up instead with discourses that ideologically position the Luxembourgish language as both distinctive and an exclusive feature of Luxembourgish national identity (cf. ‘one nation, one language’ ideology (Woolard 2008, 4).

Tensions between metaphors of hybridity (such as Mischkultur and identity through multilingualism) and purity (such as the ethnolinguistic primacy of a single language in the Herderian sense) commonly form part of the metadiscursive ideologies and practices for imagining language and tradition in the context of modernity (Bauman & Briggs 2003, 17). There are many such contradictions of standard language (Gal 2006, 167–73) and research on language ideologies amongst multilingual communities reveals that such seemingly conflicting concurrent perspectives are common, where authenticity, purity and hybridity are amalgamated into a complex language ideological assemblage (Kroskrity 2018, 10). Although such assemblages might be highly multilingual, there can be ‘a special semiotic investment in one language within the linguistic repertoire as emblematic of the group’ (Kroskrity 2018, 10), which in the particularist narrative would be Luxembourgish.

Conclusion

This paper has demonstrated the value and insights gained by applying central concepts from third-wave standardisation studies to further our historical understanding of the ideological complexity surrounding ‘late’ standardisations in multilingual contexts (McLelland 2020). By considering closely through this lens the early codification efforts for Luxembourgish in the early 19th century, multiple contrasting discourses have been identified in the guides that construct Luxembourgish within the ‘shadows’ of the contemporary dominant standard languages, French and German (Walsh 2021). These include ideological processes in the context of naming practices, hierarchisation of linguistic varieties, language differentiation, visibility, exoticisation of ‘non-standard’ varieties and competing national narratives around tensions between hybridity and purity. The analysis suggests that the present-day discussions about the framing of Luxembourgish linguistic identity (Vari & Tamburelli 2020, 14), especially along the multilingual-monolingual continuum within the respective language ideological assemblage, go back further than the Mischkultur movement at the turn of the twentieth century and stretch as far as the earliest published metalinguistic works on Luxembourgish beginning in the 1820s.

More broadly, the findings here complement and expand on a range of developments in related fields. In the context of language making (Krämer, Vogl, & Kolehmainen 2022, 3), we have seen the effective application of concepts from language making to historical sociolinguistic enquiry, including glossonyms (labels for varieties), standardisation, hierarchies and ideologies. The discursive formation of difference along the lines described in this paper intersects with insights from research on the semiotic processes of differentiation (Irvine & Gal 2000; Gal 2018; Gal & Irvine 2019) which has been instrumental for discerning the various ways in which Luxembourgish has been constructed in the context of other standard languages in the discourses of codification. In the context of the Luxembourgish language and national identity, the enduring implications of the ideological tension between hybridity and purity have been recognised in recent debates and policies about the language (cf. Horner (2007, 4) on the hybridy/purity debates; Bellamy & Horner (2018, 338–9) on the mixed language metaphors). Close consideration of pioneering metatexts on Luxembourgish in the early 19th century has revealed where the origins of these tensions and discourses can be identified in metalinguistic comments on the language, helping us to trace more thoroughly the development of the multiple narratives over the centuries and therefore to understand more comprehensively the background of competing present-day discourses on Luxembourgish in the shadow of neighbouring standard language cultures.

Acknowledgments

This work was supported by the AHRC-funded MEITS project (Multilingualism: Empowering Individuals, Transforming Societies) under award number AH/N004671/1. Thank you to Wendy Ayres-Bennett, Nicola McLelland, Gabriel Alejandro Rivera Cosme, Olivia Walsh and the two anonymous reviewers who provided helpful feedback on the drafts. Any remaining errors are my own.

Disclosure statement

No potential conflict of interest was reported by the author.

Data availability statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This work was supported by the AHRC-funded MEITS project (Multilingualism: Empowering Individuals, Transforming Societies) under award number [AH/N004671/1].

Notes on contributors

John Bellamy

John Bellamy is a Senior Lecturer in Linguistics at Manchester Metropolitan University. His areas of research include sociolinguistics, discourse analysis, language ideologies and multilingualism.