Exploring the effects of phonetic overlap and background noise on incremental processing in children

Jun 13, 2020 | NURJ x EXPO 2020 Poster

Sophia Liu

AdviserTina Grieco-Calub

Sophia is a current junior from Birmingham, Alabama majoring in neuroscience and minoring in global health studies. Sophia came into college uncertain about undergraduate research but discovered the field of communication sciences and disorders from my freshman seminar “Language and Childhood.” As an immigrant who grew up speaking both Mandarin and English, she is interested in how children with different experiences acquire language and use it to communicate with the world. As a result, Sophia joined the Hearing and Language Lab (HLL) to learn about how our experience with sound affects our ability to understand and learn language. Over the past two years, Sophia worked on studies about how children learn new words and how children with and without hearing loss process phonologically competing words. In the summer of 2019, Sophia was given a summer grant to conduct my own independent project on children’s ability to process rhyming words.


In everyday conversation, individuals actively process speech in order to comprehend and respond in real-time. As a word unfolds, listeners activate possible lexical candidates and actively determine the target word as they receive more information, a process referred to as incremental processing. This process requires knowledge of one’s native spoken language and the ability to recognize individual phonemes. While this process allows for rapid word recognition, it can also result in phonological competition among multiple candidate words that have phonetic overlap. This has been shown for words that share the same initial phonemes (e.g., candle and candy). However, other types of phonetic overlap, such as rhyme (e.g., candle and sandal), are not well explored. The present study implemented an eye-tracking paradigm in school-age children to quantify incremental language processing of words that rhyme with or share offset phonetic information with other words. The experimental method enabled us to measure the extent to which children looked at target images (e.g., candle) as well as images that served as rhyme competitors (e.g., sandal), shared offset competitors (e.g., poodle), and unrelated distractors (e.g, diaper) as the target speech unfolded. Children completed the task in quiet as well as in the presence of broadband background noise, which disrupts the acoustic representation of the speech input. The data suggest that phonetic overlap does impact incremental processing in quiet. However, the background noise condition slows processing speed to a certain extent for each of the three competitor image condition.