There is an extensive literature in the social and natural sciences that outlines biases and "errors" in the understanding of causal powers. However, there is less work on how (applied) linguists themselves fall into various "traps" when trying to assign cause and consequence to various linguistic phenomena. In this talk, I attempt to explore today's causal reasoning climate in Applied Linguistics, with an eye on key open questions for future research. Drawing on my research on language learning in the third age, I offer a critical review of empirical and theoretical results concerning understanding of the role of 'explanation' – in particular, causal and causal mechanism theories of explanation – as a means of satisfying a psychological need as well as of contributing to explaining prediction, control, and other dimensions of research. How does the quality of an explanation affect our judgments, beliefs and, by extension, our research approaches? Do these features of explanation help us achieve particular epistemic goals? Or do they sometimes lead us astray, leading to errors in our reasoning and decision-making as researchers? I will argue that an improved understanding of causal argument should benefit directly a range of areas such as argumentation, choice of method and statistical testing.
