This summer I worked as a research intern at the Language and Cognition lab at Stanford University, where my job was not only data collection, participant recruitment and data analysis, but I also contributed my ideas to the study design and thought about how our study would fit in the bigger picture of research in the particular area. During the internship, I met fellow interns who are extremely brilliant psychologists, linguists and computer scientists hailing from distinct areas of the country and had excellent discussions with them at our weekly seminars. I was able to learn about curiously interesting ideas and perspectives as well as develop friendship with my peers who have similar goals and dreams in the future—to understand language as a cognitive system and more broadly, how the human mind processes and manipulates languages that are inherently symbolic to achieve concrete communicational and social goals.
My project this summer is mainly exploring children’s understanding of pragmatic inferences, in other words, their ability to “read between the lines”. Specifically, we are trying to understand whether children are able to process implicatures given a certain context and if not, what could improve their performance. Pragmatic inferences are ubiquitous in our daily life. For example, if A says : “Is John good-looking?” B answers: “Well, he has a good personality”. In this case, B is not just commenting on the personality of John, but also implying his answer to A’s question—which is very likely a negative one. So a pragmatic inference is when the meaning of a sentence goes beyond its literal sense. We, as speakers, follow certain rules, as summarized by Paul Grice. One of these rules is that one’s speech has to be informative but economic. As adults, we are proficient in dealing with these inferences, whereas children are found to be not as good. However, they are shown to be somewhat sensitive to which sentences are informative and which ones are not. So my mentors designed this experiment to assess children’s real-time processing of implicatures, where children are shown two plates—one has a banana, and the other has a banana and a carrot. If a sentence says: “Elmo’s plate has a banana”, children are expected to look at the plate that only has a banana more than the other one if they understand the speaker’s intentions, like the adults would. And it turns out that children younger than 4 are still not inferring the speaker’s intention. So we added a prosodic cue, which is a contrastive stress on the word “banana”, so the sentence becomes “Elmo’s plate has a BANANA”. We hypothesized that children should be able to pick up the speech cue to help them disambiguate the referent. As of now we only have preliminary data for the study, and they look very promising. I’m hoping to collect more data by the end of this year and see whether our hypothesis can be validated.
In terms of research methodology, the program is very computationally oriented so the interns are taught various extremely useful skills including programming languages like R, Javascript and HTML, experience with version control and code-sharing, and building mathematical models to simulate cognitive processes of language. The interdisciplinary nature and connection with cutting-edge technology really opened my eyes to a whole new world of conducting research and gave me fresh perspectives in the field of psychology.
If anyone is interested in the area of psycholinguistics, I strongly suggest they apply for a research position in the lab and explore what they can offer. For psychology research in general, I passionately encourage my fellow Wabash students to search for resources online in their area of interest and exploit them as best as they can for unique opportunities that could benefit their future career.
-Charles Wu ’15
