My research interests

Broad: Phonology, Phonetics, Typology, Computation

  • Phonology and its interfaces, especially phonetics and morphology
  • Phonological typology
  • Computational linguistics, especially computational phonology
  • Laboratory Phonology
  • Statistics methods
  • Historical Chinese phonology


  • Consonant harmony, especially long-distance harmony related to laryngeal features
  • Constraint-based models, e.g. Maximum Entropy Model, Noisy Hamonic Grammar, and Stochastic OT
  • Tone, especially tone sandhi
  • Formal language theory
  • Prosody-semantics interface

Current Projects

Explaining tone sandhi in Linchuan

In this project, the part of disyllabic tone sandhi has been finished in my recent paper (see Papers). In this article, I combine the analysis of tone inventory and tone sandhi (I would like to give Prof. Adam Albright credit for this idea). The ordinary idea was to analyze tone inventory by referring to Dispersion Theory (Flemming 2006) and P-map (Steriade 2001) . In the progress of writing the paper, I found a way to rule out marked tones in tone inventoris step by step, and combine this analysis with tone sandhi. In my paper, I discussed the computation of the constraint MATCH-SLOPE, which might be interesting to the researchers of tone and tonal alternation.

The next steps of the project are:

  1. Analyzing tone sandhi in trisyllabic sequences. I will do a fieldwork in the January of 2018;
  2. The typology of tone inventories in Chinese languages (see below);
  3. Examining the computation of the constraint of slope correspondence MATCH-SLOPE in my paper;
  4. Modeling the learning of tone inventory and tone sandhi in the framework of MaxEnt and Noisy Harmonic Grammar;
  5. Testing contour tone targets, e.g. slope in Linchuan.

The typology of tone inventories in Chinese languages

In my paper on tone sandhi (Dai 2017), I proposed a typology of tone inventories in Chinese languages by using the data from Coblin (2011), Coblin (2015), and Bao (1999). The database is relatively small (about 43 Chinese languages), and my following work is to collect the data of the tone inventories in Chinese languages.

A map of the tone inventories of the world’s languages (Maddieson 2013; extracted from WALS). See the intensive red dots in Chinese languages.


A Grammar of Gan (Chinese)

This is a project on language documentation. Gan Chinese is one of the most unexplored Chinese language/dialect. In Linchuan, one of Gan Chinese, the Linchuan ‘language’ has almost degraded to an ‘accent’ of Mandarin. My project starts at Linchuan, my hometown, from phonology to morphosyntax. I expect to write a grammar for this language in ten years.

The VOT value of Jakarta stops

This is the final project of LING 106. I am working with a speaker of Jakarta Malay (< Indonesian) Ludwig Adisiswoyo who is my classmate and friend at UCSB.

The diachrony and synchrony of the gender system in Tsez

This is a term paper I am writing for LING 181, under the oversight of Prof. Bernard Comrie.

Statistics in Linguistics

I am interested in statistics in linguistics research. This interest comes from Statistics in Linguistics, a course at UCSB taught by Prof. Stefan Gries. I uploaded the code I edited to convert Praat .text data to .csv columns to my Github (Link to this code; Update on Oct. 13. 2018: after I posted my code about three months, I found a similar code from Aaron Braver’s webpage [Link]).

Screenshot 2017-12-13 15.30.44
Cluster analysis I did in the assignment of LING104 taught by Prof. Gries