Content area
This dissertation investigates the linguistic and ideological conditioning of ergative case variation in Basque, focusing on two distinct bilingual communities: Getxo (Basque-Spanish) in the Basque Country, and Boise, Idaho (Basque-English), home to a diasporic heritage speaker population. Ergative nominal marking (-k) is among the most ideologically charged grammatical features in contemporary Basque, often functioning as an index of linguistic legitimacy. While previous studies have centered on child speech (difficulty (Austin, 2013; Duguine & Köpke, 2018; Ezeizabarrena, 1996; Ezeizabarrena, 2011; Meisel & Ezeizabarrena, 1996; Zawiszewski 2006) and native speech and documented its social salience(Rodríguez-Ordóñez, 2015, 2022; Gondra et al., 2024), this study extends the analysis to non-canonical speakers in research and incorporates a qualitative, ideologically focused approach linking ergative production to speaker beliefs and self-perceptions. It examines how linguistic, social, and ideological variables shape ergative variation, and how different speaker profiles—particularly early L2 and heritage speakers, often excluded from traditional sociolinguistic analysis—navigate ideologies of authenticity and correctness.
Framed within the new speaker paradigm (O’Rourke et al., 2015; Ortega et al., 2015), the dissertation challenges essentialist notions of nativeness still active in the community—especially those based on family transmission or dialect use. Building on critiques of the native/non-native binary (Birkeland et al., 2024; Flores & Rosa, 2023; Lynch, 2008), it argues that early L2 and heritage speakers—though often ideologically positioned as less legitimate—are central to the contemporary Basque-speaking landscape.
To explore these issues, this study investigates how ergative marking varies across written controlled tasks and natural speech in both communities and across speaker profiles and examines the language ideologies that shape the salience and interpretation of this feature. Methodologically, it adopts a mixed-methods design combining (i) quantitative analysis of two structured tasks (acceptability judgment and elicitation tasks) and spontaneous interview speech, and (ii) qualitative discourse analysis of speaker ideologies, beliefs, and self-perceptions.
This dissertation makes two central contributions. First, it demonstrates that nativeness cannot be captured through fixed binaries (e.g., native/non-native; new/traditional; heritage/monolingual native), as acquisition type, self-perceived nativeness, and speaker identity influence both linguistic performance and ideological positioning in overlapping yet non-equivalent ways. Some early L2 outperformed heritage speakers and L1 speakers; some heritage speakers outperformed both early L2 and L1 speakers too, while some L1 speakers did not even identify as traditional or native. Hence, the use of ergativity may reflect their diverse experiences as bilinguals. These findings show that ergative variation is not simply a result of incomplete acquisition or attrition. Rather, it reflects how bilingual speakers—regardless of profile—navigate two structurally opposed case systems: ergative-absolutive (Basque) and nominative-accusative (Spanish and English), leading to systematic patterns such as -k omission or reanalysis of subject roles.
Second, the study reveals how the Mother-Tongue Ideology (Skutnabb-Kangas, 1989) and Standard Language Ideology (Milroy & Milroy, 1999) mediate beliefs about ergativity. In Getxo, -k is tightly linked to correctness and authenticity, enforced through schooling and peer surveillance. In Boise, these ideologies are weaker, and ergative omission is less policed. Drawing on Irvine and Gal’s (2000) semiotic model, the dissertation argues that ergative marking becomes iconized as a symbol of legitimacy, fractally mapped onto speaker categories, and used to erase proficient yet non-conforming speakers—such as proficient early L2 and illiterate heritage speakers. This process of iconization and erasure is particularly pronounced in Getxo. In Boise, its effects are more limited, primarily affecting speakers who are in regular contact with the motherland's ideologies through the education system or personal acquaintances. In sum, this work shows how grammatical variation is not only structured by linguistic constraints but also shaped by ideological regimes of authenticity and belonging in both revitalized and diasporic contexts.