Production Training
Visual Feedback
Visualizing differences in sound categories found in an L2 can help learners adapt to native pronunciations. Learners’ L2 speech production was recorded and analyzed according to frequency. Consonants and vowels have particular acoustic properties, corresponding to what is known as formants, which are concentrations “of acoustic energy around a particular frequency in the speech wave” (Wood, 2005). These can be represented visually in graphs, which was provided in one training method (Kartushina et al., 2015), where visual representations of the learners’ speech production are displayed next to the “ideal” version of that sound, so that learners can see how their production differs from the native pronunciation, based on objective acoustic feedback. Kartushina et al. (2015) posited that feedback would be much more effective in helping learners acquire the correct pronunciation, compared to mere repetition and exposure to L2 sounds.
Automatic Speech Recognition
Automatic Speech Recognition (ASR) technology has also been introduced into second language learning, particularly within the field of Computer-Assisted Language Learning (CALL). Available software can analyze the speech production patterns of a second language learner, and identify if they have produced the word or phrase correctly. If the software is unable to identify the response as one of the possible correct options, the learning system then encourages the learner to make a second attempt at pronouncing the word or phrase (van Doremalen et al., 2016). With the advent of smartphones, games utilizing the in-phone microphone have been designed using ASR technology, to help second language learners practice their pronunciation in simulated dialogue.
Perceptual Training
Word Identification Practice
Repetitive practice is also often used to train speech perception in a second language. In one particular study (Huensch & Tremblay, 2015), adult native Korean speakers learning English as their second language underwent eight 20-minute daily sessions of online training. They were then exposed to words in isolation and words in carrier phrases, and, given two choices, had to identify which word was in English. Immediate feedback was provided on whether their answer was correct or incorrect, and then they heard the word one more time. This is similar to drills carried out in second language classrooms, where teachers will play audio excerpts to second language learners and ask them to identify what they hear.
High Variability Phonetic Training
Recent advances in technology have also made it possible to tackle one of the problems with learning to perceive speech in a second language: the lack of invariance. Thomson (2012) argues that High Variability Phonetic Training (HVPT) is a way to truly harness the power of computers to aid L2 learners, since it provides accurate feedback to L2 learners and is not something that can be replicated in the classroom. It uses a wide range of input that was derived from recordings of the speech of different native speakers of the target language to represent one particular word. This shows learners that the same word can be pronounced in an acceptable variety of ways. Through this method, learners can be exposed to more sounds from the target language than if they were only exposed to their second language teacher, and can learn a fuller range of acceptability for second language speech sounds.