In our second installment of Columbia Linguistics Field Reports, follow our students’ summer travels to Beijing, Greece, and downtown NYC!
Reflections on Princeton in Beijing – Sofia Oltramari and Jacob Sanning (CC ’27)
This past summer, we had the opportunity to participate in the Princeton in Beijing program, an intensive eight-week Mandarin immersion program hosted by Beijing Normal University. Coming into the program as Columbia linguistics students, we were eager to see how complete immersion would change the way we experience language.
From a linguistics perspective, what made this experience invaluable were our real world interactions with native speakers. We got to hear how speakers utilize Mandarin in interaction, beyond the structural features we are taught in classrooms. Prosody is especially interesting in Mandarin as it is a tonal language. It was neat to hear how intonation contours carry pragmatic meaning, the way allophonic reductions occur under prosodic constraints, and how rhythmic lengthening and pausing are used to indicate turns in conversation in Mandarin as opposed to English. Sociolinguistic variation was also very interesting. Much of the Mandarin we heard in Beijing was saturated with erhua, a rhotic suffix attached to final syllables.
What we learned in the classroom and what we heard on the street were two very different dialects, but many Beijingers insist they speak standard Mandarin, not a dialect. In our travels outside Beijing, we also interacted with Jin and Wu dialect speakers — these topolects are even further removed from standard Mandarin! We also noticed some native speakers like to insert English into their conversations. While some of the English used felt as if it was just to identify us as foreigners (the occasional “sorry” or “hello” directed at us), speakers also used quite a few loan words, dropping English nouns into sentences that are otherwise completely in Mandarin. As a student with an interest in semantics and pragmatics, Jacob was particularly interested in hearing how Mandarin’s complex system of sentence final particles work in action — China’s unique linguistic landscape means that particles from local dialects often find their way into speakers’ use of Mandarin!
Princeton in Beijing is known for its rigor. We knew in theory that the program would be intense, but in practice was another thing entirely. Each day started very early with a quiz based on our daily reading. In classes of 4-5 people, we worked with our teachers improving our listening, speaking, and reading skills. Classes ran from 8am to noon, and in the afternoon we had a 50 minute one-on-one class with a teacher. Every day, from Monday through Thursday, we would complete a unit of our textbook, and every Friday we would take an oral and written exam based on that week’s coursework. By far, the most important part of the program was our “language pledge.” The language pledge was our commitment that we signed at the beginning of the program to only speak Mandarin Chinese for the eight week duration of the program, both in and out of the classroom. At first, the language pledge made communication a bit awkward as we struggled to communicate more complex ideas. But over time, these limitations became less noticeable and even made us more creative in our language usage. We learned how to reframe our ideas and embrace circumlocution, forcing us to live in the language we seek to learn.
Both of us are excited to return to Columbia after our summer at Princeton in Beijing. The program truly is one of a kind; it not only helped us develop our Mandarin speaking skills, but also public speaking in general. Taking the plunge with two months of immersion was a daunting experience, but we have no doubt that any student of a foreign language will find that there is no better route to fluency. Students looking to attend Princeton in Beijing next summer, feel free to ask us questions about admissions, financial aid, and details about the program!
Speech Lab Research – Ben Parkhurst (CC ’26)
Since June, I’ve been working as a research assistant at NYU’s Acoustic Phonetics & Perception Lab (APPL). I got this opportunity through a cold-email to the lab director after hearing about NYU’s labs from a classmate, Nina Kapstein (BC ’25).
I worked on the lab’s “f0 Project,” which focuses on investigating the fundamental frequency of gender-diverse American English talkers. The lab is in the Communicative Sciences and Disorders (CSD) department, which helped me connect with speech-language pathologists and researchers to ask about their work. The lab affiliates all have linguistics and CSD backgrounds, which provided valuable insight into the intersection between the two fields. The standard tools for acoustic analysis in phonetics involve at least some coding, which let me apply and build on my two semesters of Python.
My main responsibility was processing spontaneous speech audio files. The end goal was to run the audio files through the Montreal Forced Aligner (MFA), which transcribes audio and shows the duration of each individual phoneme. To prepare files for alignment, I ran them through OpenAI’s Whisper, a speech recognition tool. With one of my supervisors, I wrote a program to convert Whisper’s output into a Praat textgrid. I corrected Whisper’s errors in each textgrid, properly aligning the text to the corresponding section of audio and adding back in any verbal filler or mispronunciations/slip-ups.
Because the MFA uses a list of words and their pronunciations, I had to add several items to that list so the MFA could recognize and align them, such as this rather hilarious mispronunciation of “peanut butter:”
After I finished aligning the files in mid-July, I got to start an independent investigation into the rates of uptalk in the data. When checking the annotations and alignments, I noticed many participants ending declarative sentences with rising intonation—uptalking, despite directing their speech at a computer. I wanted to compare the rates of uptalk in computer-directed speech with a corpus of conversations, where floor-holding motivates uptalk. I used PyToBI on the aligned files to get the intonation pattern of all utterances, then wrote a Praat script to help me visually assess and record each instance of uptalk.
In the fall, I will be returning to the lab to work on their NSF-funded project on how accent and speaker familiarity correlates with speech processing, as well as continuing my uptalk research. The skills I have strengthened and learned in my lab work also helped me to decide my senior thesis topic, which will involve the creation of a pronunciation dictionary and language model for the MFA in Kalmyk.
Summer Research and Family in Greece – Vasily Tselioudis (CC ’26)
This summer, I went to Greece to collect data on how people interpret certain sentence structures in Modern Greek. Modern Greek is a pro-drop, free word-order language with case marking that allows the subject of an embedded clause to be marked in either the nominative or accusative. As a result, in a complex sentence where no noun precedes the matrix verb and the noun following it is marked nominative, that noun can be interpreted either as the subject of the matrix clause or as the subject of the embedded clause. On the other hand, when the noun is marked accusative, it can only be interpreted as the subject of the embedded clause.
Over the course of two months, I sat down with 25 people and had them complete my experiment, which I designed using Cognition.run, an online psychological experiment creator. In this experiment, participants read sentences with the structure I wanted to examine and answered questions to understand what they interpreted from the sentence and how confident they were in their responses. While in Greece, I spent time with my friends and extended family who live there. We went to the beach in Sounio, just an hour away from Athens by car, and the island of Aegina, which is also an hour away from Athens but by boat. I also got to see my sister, whom I hadn’t seen in about 6 months. We spent a lot of time catching up and hanging out by the beach. I am glad that I got to split my time between doing something I am passionate about academically and spending time with my friends and family in a country I adore.
