ABSTRACT
This chapter focuses on the output effect of shadowing, that is, how shadowing can promote sentence speech production at second language (L2). Speech production is based fundamentally on the interaction between speakers and listeners. High-speed processing, quasi-simultaneous performance of multiple tasks, interactivity, and locality are fundamental traits originally identified by observation of first language (L2) speech. Slips of the tongue are trivial, often barely noticed speech errors that occur in everyday conversations in a first language. The speech production process at both L1 and L2 has generally been characterized as having three basic components: conceptualization; formulation; and articulation. Once it becomes easy-to-repeat input speech, the situation changes: shadowing training becomes effective in simulating speech production with semantic processing, a multitasking capability that is needed in real-life communication. Lexico-grammatical encoding is the most important process in the acquisition of L2 speech production. How this encoding process is performed greatly influences fluency of speech.
