Linguistic phylogenies are commonly inferred from abstract cognate classifications that encode relationships among lexemes. Although widespread, this practice has well-recognized limitations: it discards the phylogenetic signal contained in segmental word forms; restricts the range of evolutionary questions that can be addressed; and treats cognacy judgments, which are hypotheses, as observed data. We introduce a comparative framework that addresses these limitations by modeling the evolution of aligned cognate word forms directly. Our approach adapts the TKF91 model of molecular evolution, originally developed to account for insertion and deletion events in DNA sequences, to the domain of linguistic data. By operating on segmental strings rather than abstract character codings, the framework enables phylogenetic inference from observable word forms and supports quantitative investigation of sound change. We demonstrate its utility through analyses that illuminate patterns of segmental stability and the evolution of phonological inventories.

Divergence-time estimation is one of the most important endeavors in historical linguistics. Its importance is matched only by its difficulty. As Bayesian methods of divergence-time estimation have become more common over the past two decades, a number of critical issues have come to the fore, including model sensitivity, the dependence of root-age estimates on uncertain interior-node ages, and the relationship between ancient languages and their modern counterparts. This study addresses these issues in an investigation of a particularly fraught case within Indo-European, the diversification of Latin into the Romance languages. The results of this study support a gradualist account of their formation that most likely begins after 300 CE. They also bolster the view that Classical Latin is a sampled ancestor of the Romance languages (i.e., it lies along the branch leading to the Romance languages).

Grammaticalization is characterized by robust directional asymmetries (e.g., Kuteva et al. 2019). For instance, body-part nominals develop into spatial adpositions, minimizers develop into negation markers, and subject pronouns become agreement markers. Changes in the opposite direction are either rare or unattested (Garrett 2012:52). Such robust cross-linguistic asymmetries have led some scholars to reify grammaticalization trajectories as universal mechanistic forces (Heath 1998:729). One consequence of such a view is that the ambient morphosyntax of a language has little or even no relevance for grammaticalization. This paper uses Bayesian phylogenetic methods to demonstrate the critical role that pre-existing morphosyntax can play in grammaticalization. The empirical basis for this claim is the grammaticalization of definite and indefinite articles in the history of Indo-European: indefinite articles developed at a faster rate among languages in which a definite article had already emerged compared to those lacking a definite article. The two changes are thus correlated. The results of this case study suggest that there is much more to be learned about when and why grammaticalization occurs by investigating its relationship to the pre-existing linguistic system (cf. Reinöhl and Himmelmann 2017:381).