UC Davis

Diachronic linguistics has long been a field of competing and contradictory claims, most recently with source-based theories and lineage-specific generalizations opposed to explanations for change based on more universal principles, and with various single-factor theories often staking out uncompromising positions that demand discarding research done in other areas of the field. While this is not new, and bold positions have been held at all points in the past of historical linguistics and other diachronic disciplines within linguistics, diachronic linguistics has a unique benefit compared to other subdisciplines of linguistics. It sits at the nexus of many, or even most fields of linguistics. Linguistic typology, historical linguistics, sociolinguistics, and numerous other fields all contribute to this discipline. The problem to date has been how best to capitalize on this advantage. It should be possible to bring the research of all these interconnected disciplines together to speak to diachronic issues rather than leave estranged the findings of each from the others. Historical linguistics has traditionally operated under an assumption that the languages of today operate under the same rules and limitations as the languages of the past. Known as the Uniformitarian Principle, this premise has been relatively uncontroversial in historical linguistics and linguistic typology but has been underutilized in the field. I will argue here that if we systematically exploit the consequences of the Uniformitarian Principle to their fullest, it presents the answer to the problem laid out above. The Uniformitarian Principle can be leveraged to allow synchronic and diachronic research to speak to the same questions. With this in mind, the full heft of synchronic linguistics can apply across diachronic studies — research into efficiency or typology for example can naturally be expected to have findings applicable to understanding how languages change. Just as importantly, the Uniformitarian Principle implies that the findings made within the whole scope of linguistics are by their very nature generalizable to language change. In this dissertation, I will lay out how, by fully leveraging the Uniformitarian Principle, a comprehensive model of language change can be constructed, where synchronic and diachronic data are all treated equally. I will offer a critique of the broad state of the field of diachronic linguistics, in the manner of a philosophy of science analysis. In doing so, I formulate a novel typology of linguistic research as it applies to diachronic analysis and demonstrate how this new understanding can be leveraged to break down conflicts in the field, and in doing so develop a new multi-factor model for understanding language change as a complex system reacting to differential pressures. Where pressures for change align or are unopposed, we should be able to predict such changes in the system in order to alleviate the relevant pressures and bring the system into alignment with them. Where pressures conflict with each other, we predict synchronic variation in general and either no change from the current language state or change into an alternative state within the range of synchronic variants predicted by the opposing forces. This more complex modeling approach allows us to avoid overly reducing the scope of the problem and to acknowledge that the individual strands of research already conducted on language and language change are not necessarily wrong, but rather waiting to be joined with one another and combined into a bigger picture analysis. This advances diachronic analysis beyond the era of single-factor modeling and allows a more cohesive manner of integrating findings from the many disciplines of linguistics.