PhD Candidate Fengyue (Lisa) Zhao was recently awarded a Hu Shih Fellowship in Chinese Studies for her work on Mandarin Tone Production in Noise (see abstract below).  

The Hu Shih Fellowship is awarded through Cornell's East Asia Program (EAP) to support graduate students conducting research related to East Asia.

This merit-based fellowship provides Fengyue (Lisa) financial support for one semester without teaching obligations, and allows her to focus full-time on her dissertation research.

The Hu Shih Fellowship in Chinese Studies is named after Cornell Alumnus Hu Shih (Class of 1914).  Hu Shih was a Chinese citizen born in the later years of the Quing Dynasty, and was one of the early East Asian students at Cornell.  Following graduation he want on to become the president of Peking University and served as the Chinese Ambassador to the United States prior to World War II.  He was also active in Chinese linguistics and language issues and advocated using written vernacular Chinese in poetry and literature.  For his contributions in this area, he was nominated for the 1957 Nobel Prize in Literature. 

Abstract:

Mandarin Tone Production in Noise: Effects of Auditory Feedback Availability Across Native, Heritage, and English L2 Speakers

Speech production relies on the dynamic interplay between feedforward motor plans—internally generated motor commands that guide speech movements based on previously learned sensorimotor mappings—and auditory feedback, the process of monitoring one's own speech sounds to detect and correct errors during production.

However, how these two control systems interact in tonal languages and across speakers with different language experiences remains poorly understood. This project investigates how degraded auditory feedback affects Mandarin tone production and how reliance on feedback is shaped by speakers’ language backgrounds, focusing on native, heritage, and second-language learners.

While prior research shows that disrupting auditory feedback affects L2 speech more than native speech, most evidence comes from Western and non-tonal languages and rarely considers heritage speakers, whose early but variable exposure may yield distinct speech motor control strategies.

Through controlled acoustic and articulatory experiments, this study examines Mandarin tone production under normal and noise-masked feedback. By combining these manipulations with detailed language background questionnaires and fluency assessments, the project moves beyond categorical group comparisons to model how individual experience predicts sensitivity to feedback disruption.

The findings will advance theoretical accounts of motor control in tonal production, fill critical gaps in heritage language research, and generate empirically grounded insights for Mandarin pedagogy, particularly in heritage language education contexts.

The Phonetics Lab was well represented at the Annual Meeting on Phonology (AMP) 2026, which was held from October 9-11, 2026 at the University of Maryland, College Park.

Assistant Professor Jennifer Kuo and PhD candidates Yao Zhang and Faolán Boyd presented the following posters:

• Evidence for the Maximum Onset Principle in morpheme segmentation; Dhanya Charan, Jennifer Kuo

• Source-language influence and cross- domain faithfulness in Nuosu Yi loanword phonology;  Yao Zhang, Jennifer Kuo

• Accent cues language-specific phonotactics in Spanish-English bilingual perceptual re-pair;  Faolán Boyd

(Photo caption - left to right - Faolán Boyd, Jennifer Kuo, Yao Zhang) 

On October 3, 2026 the Phonetics Lab spent a beautiful Fall Day birding at Cornell's Sapsucker Woods bird sanctuary.  During our guided tour along the sanctuary trails, we got to traverse varied terrain and viewed (sometimes fleetingly!) a wide variety of native and migratory birds.

After the tour, we spent time at the adjacent Lab of Ornithology building, where we admired bird-related art and shopped at a very cute bird-themed gift shop.  We also visited the Bird Discovery Lab, where we viewed exhibits on bird anatomy and spent way too much time at the interactive Bird Discovery table, where we listened to bird songs while viewing their real-time spectrograms.  

Afterwards, we had a lunch at the excellent Ithaca Ole Ole Korean restaurant and its authentic Korean cuisine. 

PhD Candidate Fengyue (Lisa) Zhao had a memorable summer at the gaming company Roblox, where she worked as a PhD Data Scientist on the Safety team, focusing on text safety and on keeping the platform safe for children playing Roblox games.
 

This experience showed that her PhD training in data processing, visualization, analytical thinking, and storytelling readily carries into industry.

Her work also pushed her to learn new skills such as:

• -handling data at scale
• -scoping ambiguous problems
• -thinking about technology's societal impact
• -working at an industry pace
• -collaborating across departments
• -incorporating feedback
• -experimenting with AI-assisted workflows.

As a linguist, she also enjoyed getting a glimpse into how language is used and evolves in multilingual communities at scale, far beyond the lab.

Fengyue is grateful to Roblox for this opportunity, and to everyone who made it such a meaningful learning experience. She gives a huge thank you to the Text Safety team, the Roblox Data Science community, Bridget Daly and Patrick Fang for their mentorship, Jenna Duque and Charlie Doria from Recruiting for their support throughout the process, and all the lovely friends she made along the way.

Fengyue is excited to carry her new perspectives forward and to see how this influences her research and her career plans.