Marianne Gullberg, Lund University & Chloë R. Marshall, University College London

The debate surrounding the limits of the adult capacity for language learning persists despite the advancements in the last decades of usage-based approaches to learning and statistical learning paradigms which all offer valuable insights. Many of these studies rely on artificial language input and/or on extensive training paradigms, obscuring the true potential of the adult learning mechanisms during initial encounters with new natural languages and without training. A further question that remains unexplored is whether learning mechanisms operate similarly or differently depending on language modality (spoken vs. sign language). 

In this talk, we present an overview of findings from a series of so-called first exposure studies examining how adults deal with input in an unknown language after just a few minutes of contact and without training. In all studies, we exposed adults with different first languages to a natural, contextualised, continuous but controlled new language in order to examine the adult capacity for word recognition, meaning assignment, and phonotactic generalisation. We investigated these capacities in spoken languages conveyed in the auditory modality and accompanied by gestures in the visual modality; and in signed languages which exist only in the visual modality. We discuss both modality-general and modality-specific findings. These studies reveal a powerful adult capacity for processing linguistic input regardless of modality, and for language learning. They also demonstrate domains of limitations. Overall, these studies highlight the robustness of the human cognitive system for learning language and underscore the potential for effective language learning in the wild.