Show Less
Open access

Linguistic Construction of Ethnic Borders

Edited By Peter Rosenberg, Konstanze Jungbluth and Dagna Zinkhahn Rhobodes

This volume focuses on the linguistic constructs involved in ethnic borders. Ethnic borders have proven themselves to be surprisingly long-lived: in nearly all European countries and beyond, border demarcation, exclusion of foreigners, and minority conflicts are some of the most persistent challenges for nations and societies. Which linguistic factors play a role in the formation of these borders, especially those drawn along ethnic lines? Which linguistic constructs contribute to the negotiation, establishment and maintenance of ethnic groups and identities? Under which conditions can processes of linguistic convergence, hybrids, or transcultural identities be observed?
Show Summary Details
Open access

Gray zones: The fluidity of Wisconsin German language and identification

Samantha Litty, Christine Evans and Joseph Salmons


Gray zones: The fluidity of Wisconsin German language and identification*

Abstract: Vor mehr als 150 Jahren wurden viele deutschsprachige Siedlungen in Wisconsin etabliert, in denen ein breites Spektrum von Dialekten, regionalen Umgangssprachen und auch standardnahen Varietäten gesprochen und geschrieben wurde. Im Laufe der Zeit wurden neue Generationen zwei- bzw. dreisprachig (Englisch, Deutsch, oft Plattdeutsch) und dann langsam monolingual englischsprachig. Damit verschwanden alte regionale Marker, oder zumindest verloren sie an Bedeutung. Anhand von literarischen Texten aus dem späten 19. Jahrhundert und Aufnahmen, die sich über eine Zeitspanne von den 1940er Jahren bis ins Jahr 2013 erstrecken, finden wir parallel zu diesen rein linguistischen Entwicklungen eine Dekonstruktion von alten sozialen und soziolinguistischen Gruppierungen und Identitäten. Die ältesten Daten deuten darauf hin, dass Sprecher mehrere deutlich unterschiedliche Varietäten des Deutschen beherrscht oder mindestens erkannt haben. Im 21. Jahrhundert sprechen unsere Consultants „nur Deutsch“, und sie verwenden eine Mischung von Dialektmerkmalen, die früher getrennt gehalten wurden. Identitäten und Identitätsgruppen sind verhandelbar und im ständigen Fluss, und wir argumentieren, dass Generationen von Deutschsprechern in Wisconsin sich auf relativ niedriger Ebene mit „Deutschtum“ und verwandten Regionalgruppierungen identifiziert haben. Alte Grenzen wurden dekonstruiert, und diese Entwicklungen beeinflussen eine neue Wisconsin Identität, die noch im Aufbau steht.

Schlagworte: Sprachkontakt, Dialektkontakt, Sprache und Identität, Sprachwandel

Keywords: language contact, dialect contact, language and identity, language change ← 183 | 184 →


While immigrant language speech communities are often represented in terms of social and linguistic continuity, we argue here that over the course of more than a century and a half in Wisconsin, speakers of various kinds of ‘German’ have constantly and often dramatically reshaped both their speech and how they identify themselves. We focus on the intersection of these social and structural dynamics over time. Considering how German speakers in Wisconsin have and have not seen themselves over time allows us to capture likely motivations for some highly unusual linguistic developments.

The original speakers of German in 19th century Wisconsin brought repertoires of linguistic knowledge and practice that identified them regionally and socially in rich and complex ways. Like in German-speaking Europe, speech readily identified speakers as northern or southwestern and distinguished them as more or less educated, or at least more or less invested in Standard Language Ideology (Lippi Green 2011). Today’s speakers continue to produce an array of known features, but in unexpected combinations and without the earlier social marking. Socially, we see these changes as a striking example of Brubaker’s (2006) notion of ‘groupness’ and we use that to inform the structural linguistic changes in these communities.

Structurally, from early times, Northern or Rhineland speakers in Wisconsin pronounced gemacht as jemacht — realizing Standard German /g/ as [j] in onsets (g-spirantization), while southwestern speakers produced fest as fescht, where /s/ is realized as [ʃ] before coda consonants, especially after /r/ (shibilization). Northerners and standard speakers had front rounded ‘umlaut’ vowels in schön and grün, while southern, less standard speakers unrounded to scheen and griin, /ø:/ and /y:/ versus /e:/ and /i:/ ([front vowel] unrounding). At the same time, more standard speakers used the genitive case, essentially absent from the dialects, while less standard speakers had no standard-like genitive and little or no command of the standard dative/accusative distinction, giving variation between die Mutter der Frau vs. die Frau ihre Mutter (among other constructions) and mit dem Vater vs. mit den/der Vater. In contemporary Wisconsin German varieties, we see instances of these very distinct regional and standard features being mixed that in a European context not only could not be found in the speech of a single person, but which are such strong regional markers that they are puzzling when first encountered.

Socially, discussion of German-American ‘identity’ typically focused as defining moments on the conflicts of the World War I and even World War II eras, on Nativism before that and on issues of ‘ethnic revival’ since the 1970s. In recent decades, the understanding of ‘identity’ in German-American communities has ← 184 | 185 → moved from a static to a constructivist one and the understanding of German-American varieties has moved from a static one to one anchored in notions of koinéization and contact-connected change. Here, we develop this general line of analysis, with direct connections to linguistic structure. We support this with evidence from eastern Wisconsin (especially Dodge County) from the 19th century to the present, as well as with indications of its social meaning. Specifically, for four datasets from different time periods, we look at structural features in terms of regional and standard character as introduced in the first paragraph above and examine what we can glean about the social interpretation of those features by speakers and listeners.

In this chapter, we first sketch Brubaker’s model and expand it to structural linguistic issues (§ 1) and give background on the community, data sources and methods (§ 2). We then present evidence on linguistic features that Wisconsin speakers have had over the last several generations and on how they see themselves sociolinguistically. We begin with representations of German speech and identification in a 19th century play set in a fictitious Wisconsin city (§ 3). We follow with evidence from recordings made in the 1940s (§ 4), recordings made in the 1960s (§ 5) and finally contemporary recordings (§ 6). This arc of evidence shows steady patterns of change. Socially, this involves change from a focus on European regional identity to a more generalized ‘German’ identity to a specifically Wisconsin identity. Linguistically, this is paralleled by a loss of German dialects in these communities, shift to a far less regional but highly variable German and eventually to a regionally distinct variety of English. In the conclusion, (§ 7), we argue in particular for the multifaceted nature of socially constructed identities at each stage, one interconnected with language and dialect. Taking language and dialect as a rough correlate of identity, we focus here directly on linguistic variation.

1.  Brubaker’s ‘Groupness’ and the connection to language structure

As already suggested, looking at how German speakers in Wisconsin have, and have not, seen themselves over time allows us to capture likely motivations for some highly unusual linguistic developments. Brubaker’s (2006) Ethnicity without Groups offers us a tool for understanding the social processes at hand, one that fits well with the linguistic processes we see in the history of German in Wisconsin. Brubaker calls for researchers to abandon previous notions of ‘identity’ and instead to take a fully dynamic approach to the topic. He argues that the constructivist position has become “complacent and clichéd” (2006, p. 3) and he pushes it farther, thinking in terms ← 185 | 186 → of unbounded, dynamic groups and “group-making” itself as a “project” (p. 13). He argues against using the notion of “identity” as an “analytical concept” (p. 61).1

Criticizing “the tendency to take bounded groups as fundamental units of analysis (and basic constituents of the social world)” (p. 2), he focuses on ‘groupness’, “a contextually fluctuating conceptual variable” (p. 11). While it seems widely accepted that both individuals and groups forge social groups, Brubaker stresses that memberships in groups not only changes, but the intensity of groupness can wax and wane.

Following Brubaker, our basic argument is that time and again in the history of German-speaking Wisconsin, group membership has been adjusted as individuals over their lifetimes and as generation after generation have negotiated a place socially and linguistically in an emerging Wisconsin society. Social scientists have long stressed that organizations and identities are not handed to us but rather that they are constructed and negotiated. For example, people, can cease to consider themselves ‘Pomeranian’ and begin to consider themselves ‘German’. But more novel and interesting, to us at least, is Brubaker’s insight that such memberships can be low-level or intense. While ‘being German’ (or ‘Norwegian’, ‘Polish’, ‘Czech’, etc.) in Wisconsin doesn’t have much concrete impact on people’s daily lives, they do often readily identify themselves as belonging to one group or the other, showing that this ancestry-based distinction has meaning for them. This groupness may not be particularly visible to non-group members, which leads us to argue that for our Wisconsin German speakers, groupness is usually the former, i.e. low-level, punctuated with moments of greater intensity.

Language plays a role in this as we accommodate to or differentiate ourselves from those around us. Changes in social affiliations are mirrored in and contribute to changes in linguistic behavior. People who came to the United States seeing themselves as ‘Pomeranian’, for example, became to some extent ‘German’ and ‘Wisconsinites’, and we can correlate these identifications with distinct linguistic patterns. One key, reflected in our title, is that these are never binary categories, but malleable and highly variable patterns, a set of gray zones.

We see these ever-changing features of structural linguistic variation and social identities most dramatically in instances where highly salient patterns from very different regions or from sharply different parts of the standard to non-standard continuum come together in the speech of individuals, where the social meaning of the structural features of language have clearly taken on new meanings and ← 186 | 187 → associations. We explore such examples from German and English here in the context of what is known about the speakers.

We argue that Germanness and its relatives (Low Germanness, notably) have at least almost always been distinctly low level for Wisconsinites. Previous research on diasporic German has assumed, often tacitly, a retention of dialects from the Old World and, with it, a group identity which our evidence suggests is largely fictional. For instance, Eichhoff states (1985, p. 234, emphasis added) that:

The dialects still extant in rural areas [of Wisconsin] preserve practically all the distinctive marks of the mother dialects in the German areas from which they originated. With the exception of the borrowing of lexical items from the English language, little if any development occurred.

These issues of course bear on identity as well as language and we present evidence below that paints a very different picture, one where much development has occurred both with regard to regional and social variation. In a certain sense, the dialects may not have changed so much as been lost in a shift to a general colloquial American German, showing dynamic development over time.

The mix of features we present here requires a different kind of account from those current in the literature on language and dialect contact. The Wisconsin German varieties we look at are not, for instance, typical of koinéization or new dialect formation where there are often ‘compromise’ features, widely shared features and features that are not socially or regionally marked. The mix is not typical of reallocation where existing features take on new meaning. It is also not indicative of a grammar gone wild, as some claim for terminal generation speakers. Instead we see that highly marked or salient regional and social features appear to spread. These features have lost their earlier social and regional significance, which we suspect may be traceable to a “loss of normativity” (drawing on a comment by Peter Rosenberg at the conference).

The constant, dramatic change in linguistic behavior from generation to generation supports Fuller’s notion that ever-changing identity correlates with ever-changing linguistic behavior: “Language maintenance is not necessary for a distinct social group identity to persist, but the reverse is true: without a distinct group identity, a minority language will not be maintained” (2008, p. 15). With that context in mind, we turn now to the community and the data.

2.  Background on community and data

We draw our data here from an area between Madison and Milwaukee, Wisconsin, namely Dodge County and an immediately adjoining area to the south in ← 187 | 188 → Jefferson County. Overall, this is one of the heaviest areas of German settlement in Wisconsin and, more importantly for present purposes, has perhaps the best and longest record of documentation of German varieties in Wisconsin, including dialects and standard-like varieties.

Over the last century and a half, eastern Wisconsin has seen a steady stream of varieties of German brought to the region by waves of immigrants, and those varieties have further developed here. In Dodge County, large-scale German immigration began in the 1840s and continued into the early 20th century. In many cases, immigrants and even their children and sometimes their grandchildren did not acquire English (Wilkerson / Salmons 2008, 2012, as well as Frey 2013 on areas just to the north), with 20–30% of the population in some communities still reporting being German monolingual in the 1910 U.S. Census, well over a half century after the major time of immigration. Such people often attended German-language schools and churches, participated in other German-speaking institutions and read local German-language newspapers and other publications.

While over a third of the state’s population once spoke German of some sort, those varieties have steadily been lost during a shift to English. In many families, the first-generation immigrants were Low German-speaking, often specifically speakers of Pomeranian (East Low German), with varying knowledge of Standard German. Alongside Standard German often (but not only) learned in schools, heard in churches, read in newspaper, etc., many used a colloquial variety of German that we will call ‘High German’ here, characterized by a standard-like degree of the Second Consonant Shift, for instance, and readily intelligible to speakers of contemporary European German. They and later generations continued to use those languages as they acquired and began to use English. Today’s generations are largely English monolingual. In short, over the course of 150 years, many families have gone from bilingual (Low German-High German) to trilingual (Low German-High German-English) to bilingual ([High] German-English) to monolingual (English). Along with this shift, the regional diversity people once understood as being part of German slips away. People even until today hear some range of regional variation, but that’s really where it ends. The last generation of learners didn’t acquire active command of most relevant regional and social variation. This change is being followed by the recent and still ongoing emergence of a distinct Wisconsin English (e.g. Salmons / Purnell 2010). As the languages used change, so too do the group identities of the speakers.

The earliest known recordings of German in the area were made in the late 1940s by Lester W.J. ‘Smoky’ Seifert, with speakers born as early as the American Civil War (1861–1865). A second set of recordings was made by Jürgen Eichhoff ← 188 | 189 → in the late 1960s, with speakers typically born around the turn of the 20th century. The most recent recordings were made by graduate students at the University of Wisconsin-Madison in the spring of 2013. The real-time depth of our records is thus approaching 70 years, from 1947 until 2013, with speakers born over a yet longer time span, from the American Civil War through 1947. The oldest speakers were born into communities where German speakers were still immigrating, where, as just noted, institutional use of the language was widespread and strong, and where German monolingualism was still common. The youngest, in contrast, are the last Wisconsin-born speakers of German in their communities and grew up as the institutional use of the language was fading in most domains – though church services in German remained widespread into the 1970s and 1980s – and where even use of German in private domains had receded.

The core of our analysis draws on those audio recordings. We first present social evidence, then structural evidence. Our data come from three sets of recordings, all stored in the Max Kade Institute Sound Archives at the University of Wisconsin – Madison. We are able to examine the historical development of these communities thanks to the decades of recordings and we have access to a wide range of sociolinguistic and structural linguistic information thanks to those interviews.

Each dataset was produced under very different circumstances and for very different purposes. Seifert was documenting the German varieties of Wisconsin, recording both High German and dialect from many speakers, while Eichhoff was very specifically working to document Low German varieties. Contemporary work is much closer in spirit to Seifert’s, with the aim of meeting Mattheier’s challenge (1993, p. 50):

Eine Sprachinsel muss in ihrer Geschichtlichkeit ebenso erforscht werden, wie hinsichtlich ihre gegenwärtigen soziolinguistischen und linguistischen Strukturen, also nach Struktur und Status des Varietätenspektrums. (A language island must be investigated in its historicity just as in regard to its current sociolinguistic and linguistic structures, thus according to structure and status of the variety spectrum.)

Additionally, the interviewers differ. While he was a professor, Seifert was also a native speaker of Wisconsin German who knew many of the people interviewed and had grown up like most of them on a farm in bilingual Wisconsin. In the recordings he speaks a Wisconsin variety of English, a kind of High German that is surely familiar to his interlocutors and, where appropriate, a dialect of German widely used at the time in Dodge County. Eichhoff, also a professor, grew up in Northern Germany and speaks very standard, even professorial, German on the tapes, along with very fluent West Low German and fluent but accented English. ← 189 | 190 → The most recent interviews have been conducted by graduate students who are L2 speakers of German and who cannot generally speak Low German.

The technology and consultants’ familiarity with the technology vary between the collections as well. Seifert made phonograph records using a SoundScriber machine with a microphone attached to the device, while Eichhoff produced reel-to-reel recordings. The newest recordings are made both directly onto computer hard drives and with portable digital recorders. It is unlikely that Seifert’s speakers had ever seen a SoundScriber before the session and less likely that any had been recorded. This surely presented serious issues with the Observer’s Paradox. In Eichhoff’s day, we suspect that the technology was more familiar, even if people were not used to being recorded. Today, people are not only familiar with computers but also with recordings and the broadcasting of voice, e.g. voicemail and Skype, though even some of the most recent consultants do not have internet access at home and/or do not regularly use a computer.

At the same time, the interview conditions have changed in ways that will have likely influenced the behavior of the speakers. In the following analysis, it is important to keep in mind that socially sensitive variables were probably deployed by speakers differently when speaking to someone from their own community who spoke a German similar to theirs versus someone from Germany or someone speaking High German as an L2.

Our first challenge, especially with the historical recordings by Seifert and Eichhoff, is to establish what we can about which varieties of German (and English) the interviewees spoke and how well. To do that, we rely on direct and indirect evidence in the recordings, including the production of the speaker, the apparent comprehension of the speaker and reports of linguistic knowledge and usage. But let us begin with data from the time before sound recordings.

3.  19th century German Wisconsin literature

We can start tracing social and structural features of Wisconsin German using literary sources from the 19th century and early 20th century. Local German plays, short stories and novels show the same variables under discussion in our recordings. Alfred Ira – a pen name for the minister and newspaper publisher Alfred Friedrich Grimm – wrote an extensive body of work that often includes (e.g. 1911) entire passages and dialogues written in three languages, Standard German, English and Pomeranian Low German, so that a reader needs to have reasonable command of all three to follow the story.

But this literature can reveal more about the knowledge expected from an audience in terms of regional and social variation, as we illustrate with structural ← 190 | 191 → linguistic features and language attitudes found in a play published in 1892 in Milwaukee by Julius Gugler (1848–1919), For Mayor Gottfried Buehler (despite the title, written in German). Gugler came to Milwaukee from Stuttgart at age 6. He writes Standard German but represents the speech of a diverse set of German-American characters. These include speakers of clearly Standard German, southwestern-accented colloquial German and what Gugler calls “ein gemischtes Deutsch” (‘a mixed German’), German-American speech involving code switching. Gugler uses these different styles and dialects to depict characters of different social groups. As shown below, descriptions of the characters are not only about the personality of the characters, but also their speech patterns. This characterization implies something about the people who would have attended performances of this play or read it: Because of the specificity of each of these descriptions, it is clear that these varieties would have meant something to the audience and that this meta-linguistic discussion would have been meaningful to actors. We show this with three examples from the play. First, consider Gugler’s description of the speech of a housemaid, Dörthe:2

Dörthe ist Repräsentantin der großen Rasse von Dienstmädchen, die ein gemischtes Deutsch, wie ich es angewandt, sprechen. Das Pommerisch-Mecklenburgische ist dabei vorherrschend. Man vermeide das Städtisch-Berlinische indessen soviel als möglich. (Dörthe is a representative of the great class of housemaids, who speak a mixed German, as I have applied it. Pomeranian-Mecklenburgish is predominant. City-Berlin dialect should be avoided as much as possible.)

We infer that, when this play was written in 1892, the German-speaking, theater-going public in Milwaukee was not only familiar with the Low German, Standard High German and English presented in the play, but also that the structural patterns associated with dialects such as Pomeranian-Mecklenburgish and City-Berlinish signified something to the audience. To appreciate the play, then, the audience had to understand the social and regional values of these speech patterns. Take Dörthe’s first soliloquy early in the play:

“Gott, was ist das heut vor ein excitement in dies Haus! Kein Augenblick nich Ruhe! (wirft sich auf das Sopha) Das up stairs und down stairs laufen! Und Miß Rosie thut mich onfahren, als ob ich heute nich meine Arbeit dhun thäte wie alle Dage! (gähnt und reckt sich) Na, bei die bin ich nich so, obgleich man sich in dissen Land doch Nichts nicht jefallen zu lassen braucht und mich die Miß Maier in der Dritten Straße schon oft einen halben Dahler mehr die Woche versprochen hat, wenn ich zu sie kommen wollte!” (“God, what excitement there is in this house today! Not a moment of rest! (throws herself on the sofa) ← 191 | 192 → Running up and down the stairs! And Miss Rosie is on my case, as if I weren’t doing my work today like every day! (yawns and stretches) Nah, I am not like that with her, although in this country one doesn’t need to put up with anything and Miss Maier on Third Street has often promised me half a dollar more per week, if I wanted to come to her!”)

When we look at Dörthe’s speech, there are several markers to take note of. Most clearly, there are loan words and translations from English. Double negatives which appear are not found in Standard German, but are used dialectically and found in recordings from contemporary speakers in the area. Similarly, the use of tun as an auxiliary is common in regional colloquial speech, but not accepted in Standard German. Finally, the ge- prefix of Standard German appears as je-, which is an expected form in northern dialects. We conclude that these linguistic features evoked the expected social and regional associations for audience members, and thus contributed to the characterization of Dörthe. A second and quite distinct situation is that of a Low German speaker, whose speech is described this way:

Der Ticketpeddler Muß starken plattdeutschen Anklang haben. Zuviel mag indessen dem Verständniß seitens der Zuhörer nachtheilig sein. Wenn mein Versuch im Plattdeutschen von einem Schauspieler, dessen Muttersprache Niederdeutsch ist, verbessert werden kann oder soll, so hüte er sich nur davor, die Sache allzu echt zu machen. Der allgemeine Ton wohlgetroffen ist dem zu genauen Markiren des Einzelnen in der Sprache vorzuziehen. (The Ticketpeddler must have a strong Low German accent. However, overdoing it may be disadvantageous to the listener’s comprehension. If my attempt at Low German can or should be improved by an actor who is a native speaker of Low German, he should beware not to make the speech too real. The successful general tone is preferable to the exact characterizations of the language.)

In fact, the language he uses is heavily Low German, but includes significant elements from High German, like at the end of this passage:

M—ja! Det mag’t wohl sind, und ersten as ick noch jrün west, hefft ick’t ooch so anseht, aberscht min Nochbar Jochen Snut söt tau mi: Dat’s Allens was annerscht in Amerika; da möt man smart sind, hat hei segt! Wer d’ meiste betohlt, de hat’s; ob einer nu die Stimmen vor die Wahl kauft oder sie nachher kaufen dhut …(M—yes! That may well be the case, and at first when I was still green, I also viewed it like that, but my neighbor Jochen Snut said to me: That’s all different in America; there one needs to be smart, he said! Whoever pays the most, gets it; whether one buys the votes before the election or buys them after…)

The Low German here shows northeastern features, typical of Pomeranian, where the Ticketpeddler is said to be from, such as g-spirantization in complex syllable onsets, like in jrün for Standard German grün, more commonly realized in Low German as grön or gräun (Herrmann-Winter 1999). But it differs in details, such as the third person plural form of ‘to be’, given here twice as sind where the expected form would be sünd, [zYnt] (Russ 1989, p. 116, Lindow et al. 1998, p. 96). ← 192 | 193 →

The speech of the Peddler shows the use of aberscht (Low German aberst), as well as annerscht (typically anners in Low German), both indicative of the shibilization found in southern and southwestern Germany. The key here is that the Peddler, a native speaker of Low German who otherwise shows no apparent southern or southwestern features, uses it. The area in Germany where this feature is common is nowhere close to Pomerania, as can be seen on the relevant Sprachatlas map ( or a simplified map (, by clicking ‘fest/fescht’). One might suspect that Gugler simply missed his Low German target, though it is remarkable that this shibilization of -rst > -rscht is one of the very features we find in Eichhoff’s recordings with Low German speakers in Wisconsin and which we still find today.

We can illustrate Gugler’s representation of regionally-colored colloquial German (southwestern, in this case) and very standard German by this exchange between the young lawyer Gustav Dorn and mayoral candidate Gottfried Buehler:

BUEHLER: Deiner Ansicht nach bin ich also ein solch verwerflicher Candidat, daß nur der Umstand, daß wir verschwägert sind, Dich abgehalte hat, mich öffentlich zu blamire? (setzt sich wieder)

DORN: Da gehst Du nun zu weit, lieber Gottfried. Im Gegentheil, ich versuchte Dich vor einer Blamage zu retten. Ich schätze Deine Ehrbarkeit, und liebe Dich um Deiner ausgezeichneten Eigenschaften willen, aber um einer städtischen Verwaltung vorzustehen, dazu gehören hervorragende administrative Fähigkeiten, welche zu entwickeln Dein früheres Geschäft Dir wenig Gelegenheit gewährt hat.


DORN: Der Hauptbeweggrund meiner Opposition aber lag in der Überzeugung, daß Dich lediglich der Ehrgeiz und die Langeweile treibt und nach einem Gespräch mit Marie bin ich zu dem Entschluß gekommen, ich thue ihr und vielleicht am Ende auch Dir einen Gefallen wenn ich handelte wie ich that, um die Familie vor einer immerhin möglichen Blamage zu behüten.

BUEHLER: In your opinion, then, I am such an objectionable candidate that only the fact that we are related by marriage prevented you from publically disgracing me?3 (seats himself again)

DORN: With that you go too far, dear Gottfried. On the contrary, I tried to save you from disgrace. I value your respectability and love you for your excellent character, but to preside over a municipal administration, that requires outstanding administrative abilities, which your earlier business allowed you little opportunity to develop. ← 193 | 194 →


DORN: The main reason for my opposition, however, lay in the belief that you were merely driven by ambition and boredom, and following a conversation with Marie I came to the conclusion that I would be doing her, and perhaps in the end you as well, a favor, if I acted as I did, in order to protect the family from still possible embarrassment.)

This conversation sounds very stiff and outdated to the contemporary ear, using complex syntax and structures, including the genitive case, which are not unprecedented, but are certainly not typical of conversational speech. Features of general colloquial American German, such as tun as an auxiliary, are notably absent, as are any clear dialectal features for Dorn. Dorn’s highly Standard German would have been understood by the audience, and listeners would have recognized this speech as distinct from the non-Standard varieties of Dörthe and the Ticketpeddler. The conscious use and juxtaposition of Standard German, Low German and other German varieties, not to mention English, in literary works like Gugler’s play highlight not only the widespread multilingualism of German speakers in 19th- and early 20th-century Wisconsin, but also the fact that particular features held clear regional and social associations.

4.  Evidence from 1940s recordings

In the 1940s, Seifert recorded German speakers in eastern Wisconsin, including at least 11 speakers from Dodge and neighboring areas of Jefferson County. These early recordings show basically no code-switching, and speakers appear to have full control of the two or more languages they speak. Some lexical borrowing occurs between varieties in all directions, mostly from English into Low German and German and occasionally from German into Low German. (See below for an example of this basic pattern from the later Eichhoff recordings.)

These recordings often give remarkable indications of the linguistic repertoires of the speakers. About half of the interviews recorded by Seifert are in his own native Oderbruch dialect (East Low German) with four labeled ‘High German’ and the others as ‘Pomeranian.’ The basic finding from listening to the recordings is that there is a remarkably clear distinction between standard High German with colloquial features and straightforward dialect. Beyond lexical borrowings and some regional character, mostly from prosodic patterns, the High German sounds impressionistically different from contemporary European German in ways that can be interpreted as reflecting contact with American English. For instance, American English-like approximate /r/ occurs for some speakers in coda positions. While similar allophones exist in some varieties of German (like some Saxon dialects), those dialects are not found in this area to our knowledge. ← 194 | 195 →

One simple example suffices here to show both regional and social aspects of variation in speech, from a Seifert interview recorded in 1947 with a male from Waterloo (Jefferson County, west of Watertown and just south of the Dodge County line). He was born in Wisconsin in 1864 and reports in the relevant U.S. Census data that he was a harness maker, likely retired at the time of the recording. Most of the recording consists of English-to-German translations and the speaker uses a colloquial German with an array of diverse regional features and many standard-like features. Consider this prompt and response, first transcribed in ‘eye dialect’ and then in Standard German orthography:

Seifert:  ‘The hens lay better in the spring than in the fall.’

RS:  die Hihner, die legen besser im Frihjahr wie im Herbscht

Die Hühner legen besser im Frühjahr als im Herbst.

The sentence shows patterns found throughout the speaker’s interview, including the standard-like use of dative case twice (im) and the non-standard and non-northern unrounding of front rounded vowels ([i:] for [y:] twice here, but unrounding of all front rounded vowels in the recording). He uses wie for the comparative particle rather than standard (and typically northern) als, and shows shibilization in Herbst. At the same time, in other parts of the recording the speaker spirantizes final /g/, e.g. [haitsəda:x] ‘these days’.4

We do not have this speaker’s linguistic biography – he appears in the 1900 to 1940 Census records and each time is shown as born in Wisconsin with German-born parents – and we do not know whether he spoke one or more dialects as well, but we know from the basic demographic history of the area that he lived in an area of heavy Low German influence where High German was widely known. Still, for a speaker born during the American Civil War, we find a set of distinctly non-northern features alongside some northern features and both distinctly standard and non-standard features.

5.  Evidence from the 1960s recordings

Eichhoff interviewed 7 speakers in this area, to our knowledge, 6 males and only one female. All of the speakers are interviewed in Low German. All speak Pomeranian, very fluently, with Eichhoff speaking English or Standard German to provide prompts for sentence translations and North Saxon Low German dur ← 195 | 196 → ing free conversation. Because his goal is to record Low German, speakers use overwhelmingly Low German on the recordings, but indirect evidence shows that speakers had good comprehension of contemporary Standard German and American English.

To the first point, Eichhoff gives the speakers complicated sentences in High German and they all respond as easily as they do to his English as to his Low German prompts. In particular, he records most of the famous Wenker Sätze used in early German dialectology, including these:

  Im Winter fliegen die trockenen Blätter in der Luft herum.

‘In the winter, the dry leaves fly around in the air.’

  Es hört gleich auf zu schneien, dann wird das Wetter wieder besser.

‘It’s going to stop snowing in a minute and then the weather will get better again.

  Tu Kohlen in den Ofen, damit die Milch bald an zu kochen fängt.

‘Put coals in the stove, so that the milk will start to boil soon.’

  Der gute alte Mann ist mit dem Pferde durch das Eis gebrochen und in das kalte Wasser gefallen.

‘The good old man broke through the ice with his horse and fell into the cold water.

In fact, only one speaker seems to have trouble with Eichhoff’s Low German – as just noted Eichhoff is speaking North Low Saxon, while these people are speaking Pomeranian, varieties that are potentially not straightforwardly mutually intelligible – and another speaker seems to have a little trouble with his English. Yet no one seems to blink at his extremely standard Northern German. Speakers almost always do a linguistic biography at the end and they generally make explicit that they know German and Low German (also often called Platt). They all produced most numbers in German, not in Platt (with one speaker as an exception), and one asserts in answer to a direct question about what language he’s using that they are in Platt not German. The consultants talk about using German and Platt, making clear that Deutsch does not include Low German for them. And they have High German loans in their Platt that are striking to the ear. One expressly uses a High German word and flags it as such: “de ‘Rahm’ seggt man, nich? Op Dütsch” (You say ‘the cream’, right? In German). This suggests that he could not recall the Platt and gave a German form instead, in fact a cognate with the Low German Rohm, though German has both Rahm and Sahne for this (with regional variation).

In addition to High German-to-Low German translations, Eichhoff does English-to-Low German translations, using sentences from a questionnaire Seifert had created. One speaker begins doing these in High German and self-corrects. Then, less than 30 seconds later, he starts doing them in High German and Eichhoff pushes him back into Low German. ← 196 | 197 →

Eichhoff:      ‘Light the fire.’

Speaker:      O…mach ma’s Feuer an.

O…mach mal’s Feuer an.

Eichhoff:      ‘The wood is in the stove.’

Speaker:      Das Holz is im Oven.

Das Holz ist im Ofen.

Eichhoff:      Nu Pla’dütsch. ‘Light the fire.’

Now in Low German.

Speaker:      [Pause] Mach ma mach mach ma’s Füer aa.

Mach mal das Feuer an.

Eichhoff:      ‘The wood is in the stove.’

Speaker:      Dat Holt is im Oven.

Das Holz ist im Ofen.

We conclude that these speakers understand complex sentences in Standard German easily and evidence suggests that they have decent command of it, based on production (one uses dative) and their self-reports. This is consistent with evidence from three Wisconsin communities in Schwartzkopff (1987: 148) that most speakers controlled the local dialect and standard-like German (see also Fuller 2008, p. 14).

6.  Evidence from contemporary recordings

The first two authors of this paper – in part together with another fieldworker, Clinton Ford – recorded interviews with 28 speakers in this area, 15 of which are used here, 6 males and 9 females. The majority of speakers have ancestral ties to Pomerania, but consultants from one family mention ancestors from German-speaking Hungary and a few others do not know what parts of Europe their German-speaking ancestors emigrated from. That said, Sewell (2014) clearly shows, after several generations in Wisconsin, family histories become quite complex, with many individuals having ancestry from various regions, countries and language areas.

During the interviews, consultants completed several tasks, including free conversation. Consultants were asked to give information related to their ancestry as well as language usage throughout their lives so that we could create linguistic profiles. They completed a picture guided narration task using Frog, where are you? by Mercer Mayer and a translation task where they were asked to translate English sentences into the kind of German that was most natural to them. Consultants had varied responses as to where their families came from, how long they have been in Wisconsin and when they used German and with whom. Something many consultants had in common is that after childhood, they report far fewer opportunities for speaking German than previous generations had likely had, including, for example, ← 197 | 198 → marriages where one partner did not know German, and German was not taught to their children.

Contemporary speakers can comprehend the fairly standard European German of the graduate student interviewers, who sometimes needed to accommodate sound changes (notably [front vowel] unrounding) and word selection. Many speakers have contact with European German speakers and make regular references to the ‘correctness’ of European German. For them, Standard German plays a role in their assessment of their own speech. Through interactions with European relatives or acquaintances, for example, contemporary speakers of German in Wisconsin learn that ‘gleichen’ in the sense of ‘to like’ is not found in European German, but sometimes speakers assume that their own speech patterns differ more from European German than they actually do. As an example, one consultant observed:

B.  We would say ‘Guck mal’, but I know that’s not German.

The expression in question is, in fact, widespread in European German, albeit colloquial, and so illustrates how speakers of Wisconsin Heritage German recognize their German as different from that spoken in Europe. While these consultants are aware of some lexical and structural differences between their German and Standard German, there is typically little to no reference to regional dialects. Others suggest a classification of the German spoken in Wisconsin as a mix of German and English, in one instance calling it a ‘slanguage’. Those who mention growing up with Low German input from Platt-speaking parents do not themselves produce more than a few words of Platt or phrases that they remember their parents/grandparents saying. For these speakers Low German is ‘rough’ or a less correct language, but they do not report it as having a clear relation to a particular region in Germany.

Today, Wisconsin German-English bilinguals speak varieties that show many features of colloquial German with regional coloring. If we use the classic defining characteristic of German dialects, the Second Sound Shift, their speech is in most cases standard-like, unlike the speech of virtually any of their ancestors. Beyond that, their language reflects varying mixtures of dialect features and features of Standard German as it is spoken in Germany today. Certain characteristics of Standard German morphology, such as -t in third person singular verb forms and ge- prefixes, are largely present, while other areas show pronounced variation. The syntactic and phonological patterns of the speakers show greater variation. A majority of the speakers produce both preterit (simple past) and periphrastic perfect forms, while only one (the youngest female, with Standard German training in school) produces inflected relative pronouns. The comparative wie is used by twice as many speakers as als, the latter consistent with Northern German usage and ← 198 | 199 → cognate with the Low German form as. The widespread use of tun as an auxiliary is strikingly non-standard – a feature long fought in standard language education – but one found in most of the regions from where their families originally came.

For the most recent recordings, we compiled a list of features that we expected to find based on self-reported ancestral origin. We also list other features that struck us as unexpected or especially distinct. After the feature list was compiled, we listened to the recordings to determine what features each consultant produces. The table below lists these with examples. Some are found in several regions while others are specific to a single region in German-speaking Europe. A few are non-standard, but regionally widespread. Bolded are features we expected to find based on region as well as a standard vs. non-standard distinction.

Table 1  Expected distribution of selected regional and non-standard features


← 199 | 200 →


The following patterns emerged from a case study of a mother and her two daughters, whose family originally came from Northeastern Germany. Linda, the mother, 101 years old at the time of interview, exhibits expected features as well as unexpected features common in Northern and Northwestern Germany. Susan, 69, exhibits expected features as well as unexpected features common in Northern, Northwestern, Southwestern and Southern Germany. Brenda, 66, exhibits expected features as well as unexpected features common in Northern, Northwestern and Southwestern Germany. While the mother and daughters each exhibit varying individual features, the following common set in Table 2 was remarkable to us.

Table 2  Categorization of selected linguistic features in contemporary Wisconsin data


The coupling of dialect and standard traits throughout the data and, more significantly, within individual speakers indicates that the original social and pragmatic meanings associated with the traits have faded in this context. Furthermore, the ← 200 | 201 → features present reflect a mixture of regional dialects that would not come into contact in Europe.

In addition to the original 15 German-speaking consultants from whom we draw linguistic evidence, several speakers from the area were interviewed who do not speak German, but grew up in a majority German-speaking community or come from families in which parents and grandparents spoke German. Often it is these consultants who emphasize traditions of the past and who verbalize their attachment to being ‘German’. For consultants who grew up with German in the home, being ‘German’ was just something that happened. In these cases, the speakers place less importance on maintaining overtly ‘German’ traditions and rituals, and their ‘Germanness’ stems from the fact that they happened to have German parents and learn the language at home. These comments suggest that the consultants maintain only a low-level identification with German. At the same time, during interviews most speakers show great interest in their heritage, modern Germany and ‘correct’ German. Through the linguistic behavior and personal observations of these speakers, we see that identification with the categories of ‘German’, ‘Low German’ and ‘American’ has shown and continues to show remarkable flexibility in these communities.

7.  Conclusions

In the earliest (and least reliable) dataset, 19th century literary representations, we see a relatively clean separation of forms associated with particular varieties: Characters represented by Gugler as being from Pomerania speak, from the available evidence, unremarkable Pomeranian, save for the curious presence of shibilization. Gugler represents speakers of colloquial southwestern German, presumably his own native variety, and it too is unremarkable. The distinctively German-American characters have lexical borrowings, code-switching, and influence from English very close to what contemporary speakers have, like the ubiquitous discourse markers such as well. The most standard German passages would count as stiff and archaic in Europe today.

A broad pattern of evidence strongly suggests that in the late 19th century many speakers had native-like command of multiple, very distinct kinds of ‘German’. We have presented evidence from literature here, but this is thoroughly consistent with evidence from letters and various metalinguistic comments in newspapers and elsewhere. Whatever the speakers’ active command, the variants were understood and had social value. For example, Low German features used in High German signaled something about a speaker’s regional background in Europe and educational/social background. And the use of English within ← 201 | 202 → German discourse signaled something about education and social status. Use of heavily formal standard German, more or less per definitio, signals something both about educational/social status and language attitudes.

Seifert recorded numerous speakers using different varieties, while Eichhoff’s speakers use Low German but show good comprehension of and indications of production skills in High German. In our fieldwork, no speakers have been able to complete the picture narration in more than one variety, though Lucht was able to record some speakers doing the narration in Low German (2007).6 As the range of dialects and languages has contracted, the stylistic register has also contracted, including integration of regional features from dramatically different parts of Europe into the speech of most individuals. This is not, we argue, typical sociolinguistic reallocation, and while it shows characteristics of partial koiné formation, it differs in important ways.

We have seen an almost complete shift from Low German varieties, especially Pomeranian and other East Low German dialects/languages, to a variety that has the defining characteristics of High German, e.g. in terms of the Second Consonant Shift. Such contemporary speakers show traces of their Platt-speaking heritage and at the same time features that are emphatically not of East Low German provenance. For instance, these speakers often use the distinctively northern pattern of realizing /g/ as /j/, so that gegangen is pronounced jejangen. These forms preserve the features of most of their ancestors. But those same speakers often produce strikingly southern patterns like pronouncing German wir as mir/mer and using the western and especially southwestern shibilization pattern -rst- > -rscht. Once regionally identifiable, these features are now simply ‘German’.

While the inventory of dialectal features we have investigated here remains amazingly consistent across more than 120 years, the situation has not been static. From Gugler’s literary representations through Seifert’s and Eichhoff’s audio recordings from the 20th century and current fieldwork, we find speakers producing onset /g/ as /g/ or /j/, producing final /g/ as /x/, producing third person singular present verbs with or without final -t, using tun as an auxiliary, and so on. What has changed is how and by whom they are used.

Our evidence shows the dynamic nature of linguistic patterns and social ‘identification’. This constant reconfiguring of ‘identification’ fits Brubaker’s notion of ← 202 | 203 → groupness as flexible and low-level. The dynamic nature of what languages/dialects people spoke or speak parallels this. Earlier markers of particular regional ‘identity’ or connections with ‘Standard German’ have been bleached of social meaning. Gugler’s regional and social stereotypes are no longer recognized. The range of command of dialects/languages found by Seifert and Eichhoff is gone. Our consultants speak ‘German’, not ‘German’ and ‘Pomeranian’ in Dodge County. The use of previously salient regional features like je-, -rscht, or front vowel unrounding are what now exists as ‘German’ in these communities. Remarkably, receptive capability remains strong, so that we can speak Standard German with them and many of them speak with European Germans.

We conclude that there has been steady change over time and now a kind of deconstruction of social and linguistic borders is taking place in Wisconsin. In correspondence about this paper, Rosenberg makes the important comment that vital language islands are complete language systems with wide ranges of styles and registers, and compromises or losses thereof are a sign of the retreat of the language island (personal communication). Though we leave it to future research to more fully address register compression the Wisconsin German setting, the above data show a steady dissolution of former distinctions as Low German and southern/southwestern features become ‘just German’ and Standard and dramatically non-standard features become unmarked, as well. This is highly consistent with Fuller’s view of identification as motivation for maintaining language rather than vice versa. Most of the last German speakers demonstrate only low-level identification with ‘Wisconsin German,’ and non-speakers now talk much more about a ‘German identity’ than speakers do. A new Wisconsin identity is recent and its formation is still ongoing. There may be construction of new boundaries, but it is regional rather than immigrant and further patterns will be seen in English, not German.


Brubaker, Rogers: Ethnicity without Groups. Harvard University Press: Cambridge 2006.

Frey, Benjamin: Toward a General Theory of Language Shift: A Case Study in Wisconsin German and North Carolina Cherokee. (doctoral thesis) University of Wisconsin – Madison: Madison 2013.

Fuller, Janet M.: “Language and Identity in the German Diaspora”. In Schulze, Mathias et al. (eds.): German Diasporic Experiences. Wilfried Laurier: Waterloo, Ontario 2008, pp. 3–19. ← 203 | 204 →

Goss, Emily L. / Salmons, Joseph C.: “The Evolution of Bilingual Discourse Marking: Modal particles and English markers in 19th century German-American dialects”. International Journal of Bilingualism 4, 2000, pp. 469–494.

Gugler, Julius: Dramatisches. Selbstverlag: Milwaukee 1892.

Herrmann-Winter, Renate: Plattdeutsch-hochdeutsches Wörterbuch. Hinstorff: Rostock 1999.

Ira, Alfred: Aus der alten Kaffeemühle. Antigo Publishing: Antigo 1911.

Jacob, Alexandra: Niederdeutsch im mittleren Westen der USA: Auswanderungsgeschichte, Sprache, Assimilation. Verlag für Regionalgeschichte: Bielefeld 2002.

Lindow, Wolfgang et al.: Niederdeutsche Grammatik. Verlag Schuster: Leer 1998.

Lippi Green, Rosina: English with an Accent: Language, Ideology and Discrimination in the United States. Second edition. Routledge: London 2011.

Litty, Samantha: Time for /t/: VOT variation in three southern Wisconsin counties. (manuscript).

Lucht, Felecia: Language Variation in a German-American Community: A diachronic study of the spectrum of language use in Lebanon, Wisconsin. (doctoral thesis) University of Wisconsin – Madison: Madison 2007.

Mattheier, Klaus J.: “Sprachinselsoziolinguistik: Beobachtungen und Überlegungen an deutschsprachigen Sprachinseln“. In Salmons, Joseph (ed.): The German Language in America, 1683–1991. Max Kade Institute: Madison 1993, pp. 38–61.

Milroy, James: “Social network and prestige arguments in sociolinguistics”. In Bolton, K. / Kwok, H.: Sociolinguistics today: international perspectives. Routledge: London 1992, pp. 146–162.

Nützel, Daniel / Salmons, Joseph: “Structural Stability and Change in Language Contact: Evidence from American German”. Language and Linguistics Compass 5, 2011, pp. 705–717.

Rose, Mary Aleene: Language, Place and Identity in Later Life. (doctoral thesis) Stanford University: Palo Alto 2006.

Rosenberg, Peter: Der Berliner Dialekt und seine Folgen für die Schüler. Geschichte und Gegenwart der Stadtsprache Berlins sowie eine empirische Untersuchung zu den Schulproblemen dialektsprechender Schüler. (dissertation). Niemeyer: Tübingen 1986.

Salmons, Joseph / Purnell, Thomas: “Language Contact and the development of American English”. In Hickey, Raymond (ed.): The Handbook of Language Contact. Blackwell: Oxford 2010, pp. 454–477. ← 204 | 205 →

Schirmunski, Viktor M.: Deutsche Mundartforschung: Vergleichende Laut- und Formenlehre der deutschen Mundarten. Akademie Verlag: Berlin 1962. New edition by Naiditsch, Larissa with Wiesinger, Peter (eds.), translated by Fleischer, Wolfgang. Peter Lang: Frankfurt 2010.

Schwartzkopff, Christa: Deutsch als Muttersprache in den Vereinigten Staaten. Teil I: German Americans: Die sprachliche Assimilation der Deutschen in Wisconsin. Steiner: Stuttgart 1987.

Sewell, Alyson: A socio-historical study of heritage languages: Accounting for variation among Wisconsin German varieties. Second International Conference on Heritage/Community Languages. UCLA. March, 7–8, 2014.

Wilkerson, Miranda / Salmons, Joseph: “‘Good old immigrants of yesteryear’ who didn’t learn English: Germans in Wisconsin”. American Speech 83(3), 2008, pp. 259–283.

Wilkerson, Miranda / Salmons, Joseph: “Linguistic Marginalities: Becoming American without Learning English”. Journal of Transnational American Studies 4(2), 2012, acgcc_jtas_7115, from ← 205 | 206 → ← 206 | 207 →

*  The conference from which this paper grows gave us opportunities to sharpen our understanding of the setting we work with and we are deeply grateful to the organizers and participants for that. In addition to them, we thank the audiences at the Viadrina conference for discussions on this topic and comments on earlier versions of this manuscript, and Peter Rosenberg for extremely helpful feedback on a full draft of the paper. We are also grateful to Marit Ann Barkve, Ben Frey, Monica Macaulay, and Alyson Sewell for their constructive suggestions on a prior draft. An earlier version of the paper was presented at the 4th Workshop on Immigrant Languages in America (WILA), Reykjavík, Iceland in 2013, and feedback from that audience was crucial to the later development of the topic.

1  We see this terminological/conceptual critique as badly needed within linguistics and closely related to Milroy’s (1992) dismantling of the notion of ‘prestige’ in sociolinguistics.

2  Throughout, we retain Gugler’s orthography.

3  In German-American varieties, blamieren often means ‘to blame’ but we read it in the present context as more likely still having the European German meaning.

4  In word-final position (and often in syllable codas), German famously has ‘final devoicing’ or fortition, so that Tag ‘day’ ends with [k] for many speakers. Northern and some central varieties spirantize /g/, which together with fortition yields [x].

5  According to Rosenberg (1986 and personal communication), unrounding was not unknown in the North through the beginning of the 20th century. The forthcoming Norddeutscher Sprachatlas shows unrounding in the Berlin-Brandenburg regiolect as well as some varieties of Pomeranian. Thus, front vowel unrounding could have been a native feature of some Low German varieties brought to Wisconsin, though it is not common in northern varieties today. Likewise, shibilization is reported occasionally in some words, like Durscht and Wurscht.

6  While Low German has receded or perhaps disappeared in this area and some others, e.g. to the north in Sheboygan County, it is still spoken especially in central Wisconsin. See, for instance, Jacob 2002.