This chapter addresses pronunciation in second language (L2) learning, which ranges from phoneme-level pronunciation to conversation training. First, the definition of phonemes and their relationship with articulation are explained. Vowels and consonants are classified according to different dimensions. The concept of distinctive features is also described. These provide a basis to model and identify phoneme-level pronunciation errors. Suprasegmental features such as stress and rhythm are also addressed. Next, speech analysis methods are described. While formant analysis is effective for diagnosing the pronunciation of vowels, articulatory attribute detection is explored for comprehensive analysis of all phonemes. The chapter then introduces automatic speech recognition (ASR) technology to detect pronunciation errors. Settings of minimal pairs of words, prompted text, and free input can be designed. ASR models are also used for pronunciation grading. The goodness of pronunciation (GOP) score is computed for each phoneme and aggregated over all phonemes in the utterance. Nonnative speech modeling is crucial for effective L2 pronunciation learning.

This chapter examines the concept of L2 speaking by detailing several technologies that can be used to support the development of oral production in a foreign language. Relevant theoretical and historical concepts are first discussed to give readers a foundation to understand the factors that influence the L2 speaking process. The next sections delve into emerging technologies that show promise in supporting speaking development. The chapter concludes with future directions related to L2 speaking teaching and learning.

This chapter explains significant speech and translation technologies for healthcare professionals. We first examine the progress of automatic speech recognition (ASR) and text-to-speech (TTS). Turning to machine translation (MT), we briefly cover fixed-phrase-based translation systems (“phraselators”), with consideration of their advantages and disadvantages. The major types of full (wide-ranging, relatively unrestricted) MT – symbolic, statistical, and neural – are then explained in some detail. As an optional bonus, we provide an extended explanation of transformer-based neural translation. We postpone for a separate chapter discussion of practical applications in healthcare contexts of speech and translation technologies.