Viewing language as a complex cultural evolutionary system has greatly advanced our understanding of some fundamental questions concerning its biological bases, cognitive implementation, universal tendencies and the spatio-temporal patterning of linguistic diversity. Genetic biasing represents one important but understudied factor, influencing the diachronic trajectory of language change by affecting language transmission across generations (Dediu & Ladd, 2007; Ladd et al., 2008; Dediu, 2011b). This project will advance our understanding of the nature, mechanisms and consequences of such biases affecting language and, in particular, speech. The project builds on my previous ground-breaking work, by focusing on three main directions: (1) computational modelling of genetically biased language transmission (Dediu, 2008, 2009), (2) statistical and evolutionary analyses of typological and genetic diversities (Dediu, 2011b; Dediu & Ladd, 2007), and (3) the collection of data on vocal tract variation and its effects on speech. Direction (1) will help understand the types of biases and the social, linguistic and demographic conditions that lead to genetically biased language transmission. Moreover, it will highlight the "traces" these processes leave, allowing their identification in real data. Direction (2) will apply advanced statistical and evolutionary techniques to identifying aspects of language and the human genome that might be involved in biased cultural transmission. Direction (3) will involve primary data collection of vocal tract parameters from key populations and individuals, experimental articulatory phonetic techniques, and computer models of the vocal tract used to link patterns of vocal tract and speech variation. We will also conduct an extensive review of the various literatures (phonetics, dentistry, medical genetics, anthropology...) concerning vocal tract variation, its genetic bases and effects on speech. These findings will open new directions for exploring language evolution and change as well as linguistic universals and variation. The project will also develop new quantitative methods for studying language.