Lexical access in speech production

authors

  • Kerr Emilia
  • Ivanova Bissera
  • Strijkers Kristof

document type

COUV

abstract

The speed and ease with which we produce words has puzzled researchers for decades. Uttering a single word comprises a great number of mental operations like conceptual selection (‘choosing’ the concept we are about to name), lexical retrieval (selecting the correct, and grammatically specified lemma for that concept), phonological encoding (retrieving the phonological form of the word), articulatory preparation and, lastly, producing the correct sequence of sounds that represent the intended word. All of this is processed by our brain in a few hundreds of milliseconds and is largely error-free. In other words, even though language production is an immensely complex psychomotor skill, we nonetheless manage to achieve it (apparently) quite effortlessly. Hence, our brain must be particularly efficient and successful in organizing the representations and dynamics underpinning our ability to speak. Understanding the nature of that organization has been a key research endeavor in the field of language production. In this chapter we will offer an overview of some of these potential architectures for the retrieval of the mental representation of words, better known as lexical access. To access a word is to retrieve it from the mental lexicon, a vast lexical storage that for an average adult language user comprises, according to different estimates, from 1000 to about 100,000 words (e.g., Levelt, 1989). In essence, for spoken word production, lexical access means coupling conceptual representations and their phonological forms. Below we will give an overview of four different types of word production models, which serve as a guide to highlight the different possible cognitive and neurobiological architectures that can support lexical access. First, we will review traditional serial and interactive theories (Part I and Part II), and then we will move onto some more recent models, namely dual-stream (Part III) and parallel (Part IV) models of lexical access in speech production.

more information