Language use in an Old Order Amish community in Kansas

Old Order Amish are a religious group with three languages in its linguistic repertoire: Pennsylvania German (PG), American English (AE), and Amish High German (AHG). A considerable amount of research examined PG-speaking communities, analyzing the causes of language change (whether it is caused by language contact or internal processes), the factors determining language choice in situations like family or work, and the spread of linguistic innovations between speech islands. However, few studies examine the language alternation within speech situations, language use in the worship service, or language use at the level of individual utterances (discourse level). Furthermore, rural communities are underrepresented in research on PG and few studies exist on areas with a low density of PG-speakers.;The present study addresses these research deficits by describing and analyzing an Old Order Amish speech community in Anderson County, Kansas. The speech community is geographically distant from other PG-speakers. Data has been collected through interviews, translation tasks, and participant observation. The present study analyzes four major areas of the Anderson County speech community: First, the study describes the social structure as well as cultural and religious norms of the community. These factors influence language use, linguistic change, and communicative contacts to PG and AE-speakers outside of the speech community. The present study sets out to test with ethnographic methods how many contacts exist to other PG-speakers in geographically distant speech communities. Second, a detailed analysis of the linguistic structure of Anderson County PG will be provided, employing comparative linguistic methods, with focus on language change and contact to AE and other varieties of PG. It will be examined whether changes in Anderson County PG are caused by internal processes or language contact. Third, two theoretical models of language choice, the domain model and the network model, are tested with the data from the Anderson County speech community. Based on these data, limitations of domain and network models are demonstrated. Finally, the sociolinguistic structure of the worship service, its theological and social functions, and the language use in this setting are analyzed with ethnography of speaking and discourse analysis methods. Focus will be on the sermons. The data from Anderson County reveal a communicative problem in the sermons, the "preacher's dilemma": the preachers quote and interpret the scriptures which are in AHG. However, preachers and other congregation members have only limited AHG-proficiency and, thus, do not easily understand all words or phrases used. Switching to AE is restricted by the sociolinguistic norms and PG does not provide the necessary lexical equivalents of the words in questions. The preachers manage this dilemma by employing the communication strategies metalinguistic remarks and self-translations.;The analysis of the Anderson County speech community shows that the community has complex contacts to other PG-speaking communities and undergoes a change in the employment structure that causes an increase in AE-contacts. The linguistic data show little AE-influence beyond the lexicon, but ambivalent results regarding the influence from other varieties of PG. The tested models of language choice prove to be suitable for some settings, but not applicable for complex and highly regulated speech situations like the worship service. In the sermons, the role of the preacher and the theological function of the sermons supersede other factors of language choice including sociolinguistic norms. The use of codeswitching-based communication strategies in the sermons shows that violations of sociolinguistic norms are accepted if they serve the main function of the sermons and are kept to the necessary minimum. The communication strategies are necessary repair mechanisms for communicative problems. Thus, both problem and solution are connected to the specific structure of multilingualism in the speech community. Despite the regulated ceremonial setting and in contrast to implication from past research, the sermons prove to be dynamic speech events in which all languages of the repertoire fulfill important functions.;The dissertation addresses research deficits in four different areas that have little been addressed in research so far: First, a profile of language structure and language use in a isolated, rural PG-speech community is presented. Second, hypotheses on the sources of language change in PG are tested. Third, the language use in the worship service is described in detail and, fourth, language use on the discourse level is analyzed.