Embodied and situated language processing: Production

This chapter examines language production from an embodied and situated perspective. The chapter begins by outlining the functional architecture of production, from conceptual preparation through articulation, and mapping these stages onto the left perisylvian neural network. It then demonstrates that production is fundamentally a sensorimotor process: coarticulation reveals context-dependent motor planning rather than discrete symbolic assembly, multimodal integration shows that speech coordinates auditory, visual, and somatosensory information, and adaptation studies demonstrate continuous recalibration of sensorimotor mappings based on feedback. Real-time monitoring mechanisms operate through both external auditory feedback and internal perceptual loops, enabling error detection and correction. The tight coupling between production and perception is evidenced by phonetic convergence in dialogue, mirror neuron activation during listening, and shared neural substrates that serve both speaking and comprehending. Moving from words to discourse, the chapter examines how lexical access proceeds through distinct lemma and phonological retrieval stages, whereas sentence production involves incremental grammatical encoding sensitive to accessibility effects. The final section explores how production varies across modalities in sign language, the mechanisms of language selection in bilingual individuals, and the role of cognitive resources in production. Throughout, the chapter emphasizes that production cannot be understood as autonomous symbolic computation but emerges from interactions among cognitive, sensorimotor, and communicative constraints operating in situated contexts. This embodied perspective has implications for the design of virtual reality paradigms for studying production processes.