The Analysis of Speech Sounds Produced by Children Using Praat Software: A Case Study of Late Talkers at Way Kandis Village Tanjung Senang in Bandar Lampung

Language is something that humans have from birth, both verbal and nonverbal. Every child's language development is different. In the early stages, children will use simple words to express what they want to say. Late talker refers to delays the development of spoken language in children, including delays in understanding language, pronouncing words correctly, and the ability to speaking fluently. This study aims to: 1. Explore the phonological processes produced by late talkers; and 2. Explore the dominant sound errors that occur in late talkers. The samples were 15 late talkers aged 3-5 years. The method used qualitative, the data was collected by observation and using the PRAAT software to analyze the pronunciation errors. PRAAT is software written and maintained by Paul Boersma and David Weenink from the University of Amsterdam. The findings show that there are 26 processes, namely lateral substitution, flap substitution, palatal fronting, onset deletion, velarization, cluster reduction, velar fronting, approximats substitution, affricates substitution, devoicing, favorite sound, voicing, nasalization, consonant vowel interaction, nasal simplification, vowel lowring, other harmony patterns (labial targeting coronal, labial targeting dorsal, coronal targeting dorsal), /s/-clusters, harmony targeting coronal, positional variation, syllable deletion, fricativization, final consonant deletion, reduplication, coda cluster simplification, and vowel raising. Late talkers employ velar fronting (41%), lateral substitution (40%), and cluster reduction (19%). The conclusion from this research is that late talkers employ 26 different processes and the process preferred by late talkers is changing velar consonants (k, g, ŋ) into alveolar consonants (t, d, s, z, n, l, r).