<p>For years, understanding the relationship between behaviour and cognition has been a central concern of research conducted in the social sciences. In fields as diverse as anthropology, business, medicine, and education it is widely accepted that the development of practice (as a type of behaviour), depends on a precise understanding of how thought gets carried into action. However, studies investigating the complex interplay between a learner’s cognition (i.e. thoughts, knowledge, beliefs, and feelings about L2 learning) and their behaviour (i.e. language-related activity) are only recently garnering attention. In addition, only few studies have looked at this dynamic process with adult participants beyond the language learning classroom. Framed within the context of naturalistic language learning, this investigation explores the social construction of adult (over 30 years of age) L2 learners’ cognition in an ESOL setting. Specifically it aimed to answer the following research questions: RQ 1. What are the prior language learning experiences of a group of adult migrant learners living in New Zealand? RQ 2. How have these prior language learning experiences influenced the construction and development of their beliefs, assumptions, knowledge (BAK) about language learning? RQ 3. What is their perceived need for English in their current socio-cultural context? RQ 4. How do adult migrant language learners engage in language related activities beyond the classroom? RQ 5. How can this language learning behaviour be reflected in a model of language learner cognition? The study combined a longitudinal, ethnographic approach, with elements of narrative and case study inquiry. Six ‘recently arrived’ (Dunstan, Roz, & Shorland, 2004a) Colombian migrants (five refugees; one immigrant) were asked to talk about and discuss both prior and current experiences learning and using an L2. Through these lengthy in-depth, conversation-like interviews conducted in Spanish (the participants’ L1), told over time, a nuanced picture of the participants’ L2-related cognition emerged. As a result, I was able to more clearly observe the dynamic process in which a language learner’s mental life both impacts and is impacted on by language-related activity throughout their day-to day interactions. The participants are seen engaging in the L2 across a range of settings including at home, the doctor’s office, supermarkets and work. Moreover, in their accounts of this engagement we see change and revision (i.e. development) in their thinking about L2 learning and themselves as language learners, as well as their feelings toward the L2, other L2s and L2 users. A single participant was selected as an exemplary case to examine in detail, and facilitate understanding of this development. A case study approach allowed for a more intricate exploration of how the interplay between thought, emotion, and context impacted on the learner’s approaches to language-related activities. Issues regarding readiness to interact in the L2, intelligibility, language variety, and aversion to the ‘sound of English’ were seen as playing significant roles in the learner’s language development. This analysis resulted in the construction of a framework depicting language learner cognition in action. In terms of implications, this research supports the case for more qualitative research in SLA which centres learners’ perspectives of their L2 related experiences, particularly when so much of what seems to be affecting learning is the learners understanding of themselves and their actions. It also argues that studies in L2 cognition should focus their investigations on the developmental processes involved in the social construction of the mental factors which impact language learning and use. Finally, while belief studies in SLA are expanding the scope of their investigations – by looking to include more emotion and other affective factors, as well as by branching out into self-related constructs such as self-concept and self-efficacy in the foreign language domain – these studies remain limited in their almost microscopic view of learners’ mental lives. The picture of cognition I offer provides a more holistic understanding of this phenomenon which helps account at a macro-level for L2 behaviour. The study also highlights the potential and power of data gathering methods which foreground the participants’ voices and ideas (i.e. in-depth, unstructured interviews told over time) – reminding us that it is important when looking for what drives language learning behaviour to consider what the learners feel and think.</p>