The Automatic Selective Perception (ASP) model posits that listeners make use of selective perceptual routines (SPRs) that are fast and efficient for recovering lexical meaning. These SPRs serve as filters to accentuate relevant cues and minimize irrelevant information. Years of experience with the first language (L1) lead to fairly automatic L1 SPRs; consequently, few attentional resources are needed in processing L1 speech. In contrast, L2 SPRs are less automatic. Under difficult task or stimulus conditions, listeners fall back on more automatic processes, specifically L1 SPRs. And L2 speech perception suffers where there is a mismatch between the L1 and the L2 phonetics because L1 SPRs may not extract the important cues needed for identifying L2 phonemes. This chapter will present behavioral and neurophysiology evidence that supports the ASP model, but which also indicates the need for some modification. We offer suggestions for future directions in extending this model.

Spoken word recognition is an automatic and smooth everyday process for most of us in our first language (L1), but it can be challenging in a second language (L2). Bilinguals’ recognition of spoken L2 words is characterized by L1 interference in how words are phonologically encoded in the mental lexicon, and how they are activated during comprehension. This chapter provides an overview of phonological processing during spoken word recognition in bilinguals, describing how phonological knowledge in L1 impacts the processing of native and non-native speech for various phonological dimensions. The chapter then surveys major experimental findings in L2 phonological perception and lexical access processes, highlighting the connection between the two, and showing that the phonolexical representations created by L2 learners are L1 influenced. This survey is contextualized by an outline of the various “forces” that further shape processing (e.g. orthography, vocabulary size, or lexical factors). Finally, the chapter outlines how L2 phonological processing develops over time, and how learners succeed at optimizing their processing and creating more target-like phonolexical representations.