Learning Additional Languages as Hierarchical Probabilistic Inference: Insights From First Language Processing

Abstract

We present a framework of second and additional language (L2/Ln) acquisition motivated by recent work on socio-indexical knowledge in first language (L1) processing. The distribution of linguistic categories covaries with socio-indexical variables (e.g., talker identity, gender, dialects). We summarize evidence that implicit probabilistic knowledge of this covariance is critical to L1 processing, and propose that L2/Ln learning uses the same type of socio-indexical information to probabilistically infer latent hierarchical structure over previously learned and new languages. This structure guides the acquisition of new languages based on their inferred place within that hierarchy, and is itself continuously revised based on new input from any language. This proposal unifies L1 processing and L2/Ln acquisition as probabilistic inference under uncertainty over socio-indexical structure. It also offers a new perspective on crosslinguistic influences during L2/Ln learning, accommodating gradient and continued transfer (both negative and positive) from previously learned to novel languages, and vice versa.

Individual differences in perceptual adaptability of foreign sound categories

“…Together, these findings point to strikingly similar adaptation processes across L1 and L2 listeners, at least in the context of phonetic category “tuning” in response to distributional changes (cf. Pajak et al, to appear). …”

“…Regardless of the specific model used, the fact that native and non-native listeners demonstrate the same sensitivity and adaptability to changing distributional information in this task suggests that a unified model may be able to account for both L1 and L2 short-term perceptual learning (cf. Pajak et al, to appear); future work should explore how far this similarity extends to L1 vs. L2 learning more generally.…”

Are bilingualism effects on the L1 byproducts of implicit knowledge? Evidence from two experimental tasks

“…The authors argue that their findings pose challenges to explanations of L2 variability that evoke maturational decline in language acquisition capacity, interpreting such findings as evidence that the commonly reported failure of L2 learners to achieve native-likeness across the full range of L2-specific features may actually result from the implicit entrenchment of processing routines from the learners' previous language experience. That seems to be an interesting alternative to the hypothesis of a sudden and biologically determined halt in the brain's capacity to acquire new languages after a given age, as it seems to accommodate both the evidence of age effects on ultimate L2 attainment (LONG, 2013) and the growing body of evidence for human neuroplasticity and continued learning capacity despite increased processing demands over the lifespan (RAMSCAR et al, 2014;PAJAKA et al, 2016).…”

Perception in a Variable but Structured World: The Case of Speech Perception

“…Pajak, Fine, et al (2016) for a discussion of second-language learning as an extreme example from this perspective).…”

“…The ideal adapter framework formalizes this notion by representing situation-specific beliefs about the range (distribution) of generative models that are expected in different types of situations, from very specific (an individual talker) to very general (an entire language).Although it is beyond the scope of this paper to discuss in any detail, this perspective can be applied to second (and third, etc.) language acquisition as well (Pajak, Fine, et al, 2016). This range allows the listener to balance stability and flexibility, relying on prior experience when it is available and relevant, falling back on less informative but more flexible group-level beliefs.…”