Stark on Keiser, 'Pennsylvania German in the American Midwest'

Steven Hartman Keiser
Joel Stark

Steven Hartman Keiser. Pennsylvania German in the American Midwest. Durham: Duke University Press, 2012. 197 pp. $20.00 (paper), ISBN 978-0-8223-6769-7.

Reviewed by Joel Stark (UW-Madison) Published on H-TGS (February, 2014) Commissioned by Josh Brown

"A Language That Has Outgrown Its Name"

Pennsylvania German (PG) is interesting among minority immigrant languages in America in that it continues to thrive to this day as the first language for hundreds of thousands of speakers across the United States long after immigration from abroad has ceased. The accepted view is that PG formed in colonial Pennsylvania through the leveling of Palatine German dialects. These processes resulted in a relatively homogeneous, independent variety of German before the turn of the nineteenth century. But that is not to say that PG has been without linguistic variation. Indeed, scholars have studied lexical and phonological differences across varieties of PG in southeastern Pennsylvania since the early twentieth century. Only recently has the focus shifted to the new frontier of PG linguistics: the American Midwest.

In essence, Pennsylvania German in the American Midwest by Steven Hartman Keiser is a compendium of more than ten years of original research documenting dialect divergence and sociolinguistic variation in PG. As is tradition in sociolinguistic studies of PG, ethnographic research plays an important role in explaining the relationships between linguistic and social variables. In this way, Keiser advances the field of PG linguistics, carefully describing the social practices and geographic distribution of the ethnoreligious groups among whom PG is actively used and passed on to children. Without exception, these groups include the Old Order Amish and Old Order Mennonites.

Pennsylvania German in the Midwest serves primarily to explain phonological, morphological, and lexical variation in present-day PG. To that end, the author shows how the divergent features are conditioned both geographically and socially. We learn that "a new Midwestern dialect of Pennsylvania German has emerged and with it a new sense of regional identity for its approximately 160,000 speakers." In fact, more people speak PG across the Midwest than in the original enclaves of southeastern Pennsylvania, making PG truly "a language that has outgrown its name" (p. 1). Additionally, by considering patterns of social interaction between Anabaptist groups, the book provides an outlook for possible future dialect divergence (or convergence) among PG speakers in new Amish settlements in the Midwest. Finally, the author argues that PG is truly an American language due to its orientation toward American (as opposed to European) developments; its long history and westward expansion on American soil; and, among others, its connection to patterns of social variation that allow speakers to identify dialectal regionalisms, despite complete mutual intelligibility between dialects.

Pennsylvania German in the Midwest is relevant, first and foremost, as a resource for linguists interested in modern PG dialectology, though a broader audience of linguists interested in German as a minority language will also find the book of great use. In that vein, the book handles data relevant to human geography of Palatine Germans in their migrations to and within America, both preceding and following the American colonial period. The subsequent focus is on the westward migration patterns by Amish communities as they continued during the nineteenth century to the present, a phenomenon that helped spur the dialect divergence that comprises the core subject of the book. The introductory and concluding chapters synthesize several complex patterns of migration and social interaction, as well as their linguistic implications, that will be relevant for non-linguists interested in Anabaptist ethnography and variable (speech) behavior.

Chapter 1 traces the migration of Amish communities from their origins in sixteenth-century Switzerland to the Palatinate, then further to the American Midwest, providing a critical chronology for the development of PG. Although the (pre)history of PG includes many ethno-religious groups, the focus on the Amish during the precursory stages of the language occurs at no detriment to the central arguments in latter chapters, as the Amish are by far today's most representative group of the language. Furthermore, Keiser notes that many of the less conservative Anabaptist groups included in the study that speak PG, such as the Beachy Amish or Horning Mennonites, have origins in more conservative Old Order groups, and so focusing on the Amish is a reasonable approach that covers the most ground in the development of the language.

The Radical Reformation beginning in 1520 provided unfavorable circumstances for religious minorities, and after many decades of persecution, Swiss Anabaptists moved from Canton Bern in Switzerland to the German Palatinate in the 1670s. Many families stayed in the Palatinate and neighboring Hesse for a number of years before migrating to America; this is a pattern that continued into the nineteenth century. It is believed, Keiser states, that the linguistic norms of the Palatinate region were adopted by Amish migrants, among others, prior to moving to the New World.

The American Revolution was a watershed period for the development of PG. Virtually cutoff from European immigration, the children of colonial speakers of German varieties in southeastern Pennsylvania further leveled out dialectal differences, forming a relatively homogenous variety of German. It is during this period, shortly before 1800, Keiser notes, that we may speak of a distinct PG variety. Following the American Revolution, the largest Amish settlement was at Northkill in Berks County, Pennsylvania. As families grew bigger and land became more expensive in the nineteenth century, entire families moved westward with a pivotal stopping point in Somerset County, Pennsylvania (much like the Palatine region in centuries prior). Migrant families from eastern Pennsylvania, as well as European Amish immigrants in the nineteenth century, stayed in Somerset County for an extended period (often a generation or more), before moving into Ohio and beyond. These patterns are central in explaining the relative homogeneity of PG across America: as Swiss Amish moved northward to the Palatinate, they acquired local Palatine features; and as nineteenth-century European Amish immigrated to Pennsylvania, they acquired the local variety of Somerset County before heading west. Throughout the nineteenth century, subsequent migration patterns of the Amish within the Midwest are described as "portable communities" (p. 6). Frequent moves resulted in related, yet distinct dialects of PG: Midwestern Pennsylvania German (MPG) and Pennsylvania Pennsylvania German (PPG).

Chapter 2 synthesizes three main areas of previous dialectological research on PG: bilingualism and language contact as a source of language change and dialect differentiation in PG; sociolinguistics (especially sectarian versus nonsectarian); and regional variation (almost exclusively within Pennsylvania). Examining evidence first identified by Mark L. Louden, Pennsylvania German in the Midwest focuses on divergent linguistic features, mostly phonological, that distinguish MPG from PPG.[1] These divergent phonological features include allophonic realizations of the diphthong /aɪ/, /r/, and /l/. "It is the emergence, spread, and social evaluation of these features that are examined in this book" (p. 47).

In chapter 3, the author describes his methodology and, in detail, patterns of interaction between communities in the Midwest: the two main communities are Holmes County, Ohio, and Kalona, Iowa, since they are the oldest and largest settlements in the Midwest, and serve as the primary communities of study in the book. Migration patterns and data on intercommunity contact show that connections between midwestern Amish and Pennsylvania Amish are virtually nonexistent, contributing to noticeable distinctions both culturally and linguistically.

Methodologically, the author collected data both as a participant observer in Holmes County and by conducting four separate studies that he discusses in chapters 4, 5, and 6. For the main study, he conducted sociolinguistic interviews and translation tasks, eliciting tokens from seventy participants in Kalona, seventy participants in Holmes County, and nine participants in the new settlements of Grant County, Wisconsin, where MPG and PPG are in a situation of potential contact. For comparative purposes, data were also collected from thirty-nine participants in Lancaster, Montgomery, and Bucks counties in Pennsylvania.

The dependent variables, as mentioned previously, are the phonemes /aɪ/, /r/, and /l/. Possible variants of the phoneme /aɪ/ include monophthongs [æ:] or [ɛ:], but it may also be realized conservatively as [aɪ] depending on the age and dialect of the speaker. One reason variation in this phoneme is so salient is that /aɪ/ occurs in many high-frequency words, including the name of the language, Deitsch (German). The variable /r/ has three allophones in MPG that depend on syllable position: "vocalic [ɐ] in final position, approximant [ɹ] in simple onset position, and apical tapped [r] in complex onsets and intervocalically" (p. 66). The tapped apical [r] in complex onsets and intervocalically is of interest in this study, as this deviates from PPG, which has adopted the English approximant [ɹ]. In MPG, /l/ is realized as a clear [l], whereas in PPG it is realized as dark (or vocalized) [ɫ].

Independent variables include phonetic environment, manner and place of articulation, syllable position, participant's home community, age, gender, style, religious affiliation, and occupation. In the following chapters, statistical models are employed to tease out correlations between the independent and dependent variables, allowing the author to predict which social groups will produce specific phonetic variants.

Chapter 4 is the core of the book, presenting and analyzing data on the multiple, socially significant realizations of the phoneme /aɪ/ between speakers of MPG and PPG of various ages; home region (Kalona, Holmes County, Grant County, and Pennsylvania); and religious affiliation (Old Order, Beachy, etc.). Results from over 2,100 tokens show that MPG speakers produce a variety of progressive forms, but primarily the monophthongs [æ:] and [ɛ:]. Together, [aɪ], [æ:], and [ɛ:] comprise 82 percent of the data for this part of the study (p. 85). Other possible tokens in MPG include diphthongs with a low vowel nucleus: [a] with an upglide; [æ] and [ɛ] with an inglide; and in "advanced" cases, monophthongal /e/, though these instances occur less often. Where MPG speakers do not monophthongize /aɪ/ is when the following phonetic environment contains an unstressed vowel, such as in words like scheier (barn). Speakers of PPG produce, nearly categorically, a conservative [aɪ] diphthong, except before /r/ and /l/. This is known as the Lancaster rule, as noted by Loudon, a condition corroborated by Keiser's data.[2]

Using a statistical model of logistic regression, one conclusion drawn in chapter 4 is that "place" is not a significant variable as a predictor for monophthongal tokens in MPG. The author demonstrates that "we can confidently talk about a single sound change progressing along similar paths in both of these speech islands," Kalona and Holmes County (p. 91). Conversely, "age" and "gender" are predictive variables. Older speakers of MPG are more likely to produce conservative (diphthongal) forms of /aɪ/, whereas women are more likely than men to produce progressive (monophthongal) forms.

Because the monophthongal forms of /aɪ/ in MPG overlap in the vowel space with /e:/, Keiser discusses the potential for vowel merger between these phonemes. He finds that at least one speaker in Holmes County displays tokens indicative of near merger. Furthermore, in a perception study, a selection of Holmes County speakers "were unable to consistently distinguish these phonemes in the speech of at least two Holmes County speakers" (p. 108). While Kalona displays no evidence of phoneme merger, Holmes County may be leading the way if speakers continue to innovate in this direction.

For comparative purposes, the study continues in Grant County, Wisconsin, where two new communities of young families have recently settled. The two communities differ in language spoken (PPG versus MPG) and religious practice. The results are consistent with previous findings, indicating variable phonological realizations that the speakers are aware of. Although these groups do not presently come into close enough contact to induce linguistic convergence, the close proximity of PPG and MPG may be an area for future linguistic change within the Midwest.

Chapter 5 further portrays dialect divergence between MPG and PPG, specifically variation in liquid consonants /r/ and /l/. Data show that MPG speakers have adopted the English approximant [ɹ] in simple onsets, but have variable realizations of /r/ (alternating with tapped [r]) in complex onsets. Conversely, data show that PPG has "almost complete convergence with the phonology of the English of southeastern Pennsylvania" (p. 136). In sum, MPG has conservative forms of liquid consonants, whereas PPG has more "progressive" (English) forms.

Chapter 6 handles lexical variation between MPG and PPG, which is a salient marker of dialect divergence. The patterns of lexical variation are complex. For instance, MPG patterns are closest to current Amish usage in Lancaster, however several points of divergence are among the most noticeable. For example, what midwesterners call an emer (bucket), speakers of PPG in Lancaster call a kiwwel. This is as noticeable among speakers of PG, Keiser states, as are variations in regionalisms in American English, such as pop or soda. Precise categorical differences cannot be drawn out of the lexical data as easily as the phonological data due to the fact that lexical variation within PPG is quite drastic. For instance, the most lexical variation occurs among the socioreligious divisions within Lancaster County. The Old Order Mennonites in Lancaster County are found to be more conservative with respect to native PG vocabulary when compared to their Old Order Amish neighbors.

Other topics covered in chapter 6 are the frequency of loanword occurrence and rates of phonological incorporation of loanwords, which are not indicative of dialect divergence between MPG and PPG. Non-feminine morphology in feminine possessive constructions (mei Schwester sei Haus) "is now nearly categorical among Old Order Amish in Holmes County" (p. 156), but never occurs among Lancaster County Amish and Mennonites. Finally, the merger of the dative and accusative case can be observed in MPG.

Chapter 7 summarizes the various studies (chapters 4 through 6) that comprise the heart of the book, and provides an analysis of how various identities of the ethno-relgious communities that speak PG might influence the future of PG. To be sure, Pennsylvania German in the Midwest is a collection of exceptionally up-to-date and forward-looking studies of a language with deep roots in American soil. Its speakers are aware of dialectal variations, primarily in the phonology and lexicon, that are diagnostic of the area where a speaker is from. Similar to the carrying phrase Pahk ya cah in Hahvahd Yahd in American English, the "PG speakers even have their own stock carrying phrases used to illustrate the differences between [dialects]: for /r/, Ich hab mei Ohre verfrore am fahre 'I froze my ears driving'; and for /ai/, drei veisi Weibsleit 'three white women'" (p. 160). These caricatures help illustrate the salient differences between regional dialects, but show the homogeneity and potential for mutual intelligibility between dialects as well.

As the language grows in the Midwest and beyond, so too does the opportunity for language contact between MPG and PPG communities, as well as between PG speakers and speakers of non-English languages in the United States. For the title of this review, I have adopted Keiser's statement that "PG is a language that has outgrown its name" (p. 1). Indeed, the future of PG linguistics, as Keiser has done here, must address the rapid growth and subsequent dialectal and sociolinguistic developments of the language outside of Pennsylvania--in the Midwest and beyond.


[1]. Mark L. Louden, "Patterns of Sociolinguistic Variation in Pennsylvania German," in The German Language in America, 1683-1991, ed. Joseph C. Salmons (Madison: Max Kade Institute for German-American Studies, University of Wisconsin-Madison, 1993), 284-306; and Mark L. Louden, "Linguistic Structure and Sociolinguistic Identity in Pennsylvania German Society," in Languages and Lives: Essays in Honor of Werner Enninger, ed. James R. Dow and Michèle Wolff (New York: Peter Lang, 1997), 79-91.

[2]. Louden, "Linguistic Structure and Sociolinguistic Identity, " 81.

Printable Version:

Citation: Joel Stark. Review of Keiser, Steven Hartman, Pennsylvania German in the American Midwest. H-TGS, H-Net Reviews. February, 2014. URL:

This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.