In English, reductions might occur when there is shared knowledge among the utters, or when the word is repeated. Previous literatures have investigated this reduction phenomenon following the priming of a repeated word or a homophone. Possible explanations were provided based solely on auditory and phonological cues, without anything to do with semantics, while Hamlaoui et al. (2018) suggests that prominence shift is due to a shift in focus marking from one word to another, particularly in two sentences consisting of the same antecedent appearing in the same syntactic position. It led me to investigate the roles that focus marking play when a prominence shift happens within sentences.
In English, the prosodic marking of the focus is realized through an enhanced prominence (reflected in higher pitch, intensity, and longer duration), as the already-given information is realized through a reduced prominence (reflected in lower pitch, intensity, and shorter duration).
Take this sentence as an example: on the left, the ball is bigger than the doe, and on the right, the HOUSE is bigger than the doe. Usually, speakers will enhance the prominence of the word 鈥渉ouse鈥� and reduce the prominence of the second 鈥渄oe.鈥� Thus, we could say that the prominence is shifted from 鈥渄oe鈥� to 鈥渉ouse.鈥�
This project aims to answer to questions:
- What kinds of repetition produce a prominence shift in a word?
- Does the syntactic position of the repeating word influence prominence shift?
To answer these two research questions in short, we found that both homophones and identical words would produce a prominence shift. Also, syntactic position indeed affects the production of a prominence shift. Target words in the same syntactic position showcase more salient prominence shift.
Two production experiments were conducted and audio data was collected from the participants. Data processing was a key process in this study. Audio data was transcribed to Textgrid files. There are many ways to process textgrid files, including using local Praat or Python scripts. Still, their usability is relatively low for someone unfamiliar with the two scripting languages. Thus, the project also aims to make linguistics data collection and processing more easily by improving the existed spreadsheet experimenter developed by my supervisor, Professor Michael Wager, and by developing new scripts that can be deployed online.
I was taking a laboratory linguistics class with my supervisor, and I found myself interested in prosody and the experimenter he used. ARIA appealed to me then. I thought that it would be a great opportunity if I could continue working on a prosody project and improve the functionality of the existed scripts. This is not my first internship. My last internship was at a head tech company. Compared to traditional industry internships, the ARIA program provided me with a safer learning and working environment - I had more freedom in organizing my work at my own pace. More importantly, I can stay in touch with fellow students and many researchers.
The first learning objective for me is to learn the existing experimenter's structure, Prosobeast Annotation Tool, and Praat scripts. The experimenter includes many functions I was unfamiliar with, such as reading files and fetching audio data. Thus, it is essential for me to know how everything works. My second learning objective is to build a data-annotating pipeline and test it with the collected data. The annotated data then needs to be plotted and analyzed using R.
The biggest highlight of the summer was when the annotation script successfully ran. I learned many things that weren鈥檛 taught in class. In my previous computer science classes, I was only taught to solve problems and take tests and rarely found the opportunity to develop practical tools. For this project, I was able to learn to use new Python libraries and new programming languages. I was so excited to see my code runs for the time after days of debugging and re-configuring. It was always rewarding to apply what I've learned to build real-life solutions.
It was always hard to find where to start at the beginning. The biggest challenge for me seemed to be transforming a conceptual plan into implementable codes. Thanks to my supervisor and Professor Branislav Gerazov, they helped curtail the plan. They provided me with a clear coding framework, making it possible for me to build the script from its most straightforward function to the harder ones. Remote working is another challenge for me as it is a double-edged sword. On the one hand, it gave me the freedom to work flexibly; on the other hand, I had to deal with many distractions. To avoid distractions, I locked my cell phone during the day and tried to keep a daily to-do list to ensure that I made progress during the day.
ARIA has sparked my interest in many areas. I look forward to being involved in more projects as I really enjoyed the process of learning new knowledge and skills. The website itself is not deployed to the public net, and I feel committed to making it available for the public. I'm willing to learn and contribute more to academia. I also hope to pursue graduate study in the future, hopefully within the field of theoretical linguistics. Also, this project allowed me to gain some experience in webpage development, including webpage design and data visualization. With the skillset I've gained, I wish I could dive more into the interface of linguistics and technology.
Lastly, I would like to thank the donor of the Undergraduate Experiential Learning Opportunities Support Fund for providing me with this opportunity to explore the research project that I was interested in. It is genuinely a privilege for me to be funded. Also, I'm extremely grateful for my supervisor, Professor Michael Wagner, for his generosity and support along this rewarding learning process.
