Vici Grant for Yiya Chen’s research on melody in Speech
Unlike Dutch and English, most languages employ melody not only to add nuance to sentences but also to differentiate between words. How languages do this remains largely unknown. Linguist Yiya Chen will start a research project to investigate the linguistic, psychological and neural processes underlying the parallel communication of word identity and sentence meaning through melody. We asked Yiya about her drive to do this research.
Why do you want to do this specific research?
‘It is a natural extension of what I have been doing so far. The proposed topic builds upon my research interests in the production and processing of speech melody which speakers of the majority of the world languages employ to not only add nuance to sentence-level meanings but also distinguish word meanings. Speech melody is essential for effective speech communication and social interactions.
The knowledge and experiences gained from my previous projects (in particular VIDI and ERC-starting grant), as well as the questions and knowledge gaps that I became aware of during those research experiences directly motivated the current project. I’d also like to mention that discussing my research ideas with colleagues at Leiden and elsewhere has also been motivating and encouraging.’
What will your research bring to society and other researchers in the future?
‘The proposed project is theory oriented. Speech melody is a pervasive feature of linguistic communication and in most languages where melody is also used to distinguish word meanings, our understanding of this topic has remained a fundamental knowledge gap in our understanding of how human languages work.
Understanding melody in speech does have underpinnings in human behavior and is of societal relevance.
- Speech corpora including lexical tone typically are based on impressionistic observations with tones transcribed in symbolic labels that abstract away from the details of pitch patterns. These acoustic details, however, can be important sources of information for our understanding of the way tones are produced and perceived, and consequently evolve.
- Data and knowledge gained in our project can potentially provide insights on how to improve ‘naturalness’ in synthesized speech and how to more efficiently recognize human interlocutors’ linguistic, social, and emotional information, which are essential for human-like computer interaction systems.
- Proper understanding of melody in speech communication can also be of interest to, e.g., second language learners.
- Carrying out the research project in China and West Africa will boost research interests and facilitate scientific trainings at local communities and universities in both regions, where experimental research is still seriously under-developed, despite recent progress.'
How much money will you receive and what will you spend it on?
'I will receive 1.5 million euros. Most of part of it will be spent on allowing myself some time and hiring young talents (2 PhD students and 2 Postdocs) to work as a team on five subprojects which address the three research questions (listed in the summary). The rest of the funding will be spent on running experiments and disseminating research results.'