Personality disorders (PDs) are some of the most prevalent and high-risk mental health conditions, and yet remain poorly understood. Today, the development of new technologies means that there are advanced tools that can be used to improve our understanding and treatment of PD. One promising tool – indeed, the focus of this thesis – is computational language analysis. By looking at patterns in how people with personality pathology use words, it is possible to gain access into their constellation of thinking, feelings, and behaviours. To date, however, there has been little research at the intersection of verbal behaviour and personality pathology. Accordingly, the central goal of this thesis is to demonstrate how PD can be better understood through the analysis of natural language. This thesis presents three research articles, comprising four empirical studies, that each leverage computational language analysis to better understand personality pathology. Each paper focuses on a distinct core feature of PD, while incorporating language analysis methods: Paper 1 (Study 1) focuses on interpersonal dysfunction; Paper 2 (Studies 2 and 3) focuses on emotion dysregulation; and Paper 3 (Study 4) focuses on behavioural dysregulation (i.e., engagement in suicidality and deliberate self-harm). Findings from this research have generated better understanding of fundamental features of PD, including insight into characterising dimensions of social dysfunction (Paper 1), maladaptive emotion processes that may contribute to emotion dysregulation (Paper 2), and psychosocial dynamics relating to suicidality and deliberate self-harm (Paper 3) in PD. Such theoretical knowledge subsequently has important implications for clinical practice, particularly regarding the potential to inform psychological therapy. More broadly, this research highlights how language can provide implicit and unobtrusive insight into the personality and psychological processes that underlie personality pathology at a large-scale, using an individualised, naturalistic approach.
