Speech production

Speech production is the process by which the nervous system transforms a linguistic representation into an acoustic signal. It is not a unidirectional pipeline from intention to sound. It is a dynamical system in which the intended utterance, the motor plan, the articulatory execution, and the auditory feedback are continuously coupled in a feedback loop that self-corrects in real time.

The standard view — derived from the linguistic tradition of phonology and phonetics — separates the planning stage from the execution stage. But evidence from speech errors, coarticulation, and real-time compensation for perturbations shows that planning and execution are not separable. The motor plan is continuously updated on the basis of somatosensory and auditory feedback, creating a system in which the output is also an input.

The connection to articulatory rehearsal in working memory is direct: silent rehearsal recruits the same motor programs and feedback circuits as overt speech, only without phonation. The boundary between thinking and speaking is not a categorical distinction but a continuous variation in the gain of the same feedback loop. The connection to phonetics is equally direct: the physical parameters of speech production — articulator positions, airflow dynamics, acoustic output — are not downstream consequences of a completed plan. They are the plan's way of maintaining itself in the physical world.

Speech production is not the voice expressing thought. It is thought in its motor form — and the motor form is not secondary to the thought. It is the thought's way of maintaining itself in the world.