Exaggeration in Moderation and Simple Signal

Rachel with a computer screen showing an example of her research.

Funding Source




Assistant Professor of Linguistics

Project Summary

Every day we make several unconscious decisions about how we communicate. My summer research focused on investigating what governs these decisions. Consider the following example, where a participant in a conversation exaggerates:
A: “I was sick all day yesterday!”

B: “Really? All day? That sounds awful.”

If A was really only sick for a few hours, it is unlikely that B would feel deceived. Why would B accept A’s exaggeration, even though it contains false information? It may even be that, because the exaggeration highlights the interest of the information being communicated, B prefers this statement to one containing no exaggeration. Information in general is valuable, but information becomes more valuable the more exciting and interesting it is. So what motivates speakers like A to share their valuable resource in the first place? Being listened to and seen as a source of interesting information increases one’s social standing, and exaggeration can make information seem more interesting. There is a limit, however, as people who exaggerate too much gain a reputation for being misleading and unreliable. If this proposed explanation is correct, then we should expect that moderate exaggerators will experience an advantage both over those who do not exaggerate at all and over those who exaggerate all the time. I designed an experiment to test this idea where participants attend a virtual farmers’ market and try to purchase the best apples available from three different farmers, making their decision based on the farmers’ statements alone. One farmer never exaggerates about the quality of his apples, one farmer exaggerates sometimes, and the last farmer exaggerates every time. If moderate exaggerators have a social advantage, then participants should buy more apples from the second farmer, all else being equal.

In parallel to designing my own experiment, I worked on an ongoing project between my mentor Dr. Roberts and Dr. Greg Mills of the University of Groningen called Simple Signal. In Simple Signal, we are attempting to discover what kinds of behavior people identify as human communication. Pairs of participants send and receive simple on-off signals for 15 minutes. Afterwards, they are asked to judge whether they were receiving signals from a human, or (depending on condition) “random signals” or “signals from a computer.” Components of the signals can be adjusted, including turn-taking, signal length, and signal frequency. The participants’ judgments in response to these adjustments offer insight what make up the fundamental components human interaction.
The mentorship through CURF has allowed me to hone several important academic skills, such as refining a large concept into a single explorable question, examining the weaknesses of an experiment design, and even learning how to program in a new language. As I plan on attending graduate school, these skills are very useful to me and will play an important role in my future research endeavors.