As individuals, we would like to believe that we are free from biases and that we are somehow immune to acting on the social biases that we have been socialized to since birth. We would like to believe that we can report experiences accurately, recalling events as they truly happened. But, memory is faulty and stereotypes and social biases are pervasive. And, at a level beneath our own control, these biases slip in and influence our memory of events. Eyewitness memory, and the inaccuracy and unreliability of eyewitnesses, is a perfect example of this. But, what about the things we hear? Speech carries a great deal of information; packets of co-varying cues we have been raised to recognize categorically, informing us about a talker's race, accent, emotion, and gender. We have, through our ears, information about events that occur. And, we have in our minds, stereotyped expectations about how various groups of people behave and what various groups of people might say. In this course, we will explore how these two types of information (e.g., the percept of what is actually heard vs. our stereotypes about who is likely to have said what) clash together and influence 'earwitness memory'. We will read and critique journal articles, blogs, and popular science articles, think about the reliability of memory for auditory events, and we will work together to develop three well-designed thought experiments that address questions at the heart of this issue. Along the way, we will learn a bit about the acoustics of speech, social variation in speech, speech perception and spoken word recognition, memory, and experimental design and analysis. Students in this course should be committed to reading the assignments, sharing their ideas about the readings (without concern for 'being right'), and think creatively about ways we can explore the idea of earwitness memory together. While this is a one-quarter course, my goal is to pursue our thought experiments collaboratively, with any interested students in subsequent quarters.
3 units · Letter or Credit/No Credit
As individuals, we would like to believe that we are free from biases and that we are somehow immune to acting on the social biases that we have been socialized to since birth. We would like to believe that we can report experiences accurately, recalling events as they truly happened. But, memory is faulty and stereotypes and social biases are pervasive. And, at a level beneath our own control, these biases slip in and influence our memory of events. Eyewitness memory, and the inaccuracy and unreliability of eyewitnesses, is a perfect example of this. But, what about the things we hear? Speech carries a great deal of information; packets of co-varying cues we have been raised to recognize categorically, informing us about a talker's race, accent, emotion, and gender. We have, through our ears, information about events that occur. And, we have in our minds, stereotyped expectations about how various groups of people behave and what various groups of people might say. In this course, we will explore how these two types of information (e.g., the percept of what is actually heard vs. our stereotypes about who is likely to have said what) clash together and influence 'earwitness memory'. We will read and critique journal articles, blogs, and popular science articles, think about the reliability of memory for auditory events, and we will work together to develop three well-designed thought experiments that address questions at the heart of this issue. Along the way, we will learn a bit about the acoustics of speech, social variation in speech, speech perception and spoken word recognition, memory, and experimental design and analysis. Students in this course should be committed to reading the assignments, sharing their ideas about the readings (without concern for 'being right'), and think creatively about ways we can explore the idea of earwitness memory together. While this is a one-quarter course, my goal is to pursue our thought experiments collaboratively, with any interested students in subsequent quarters.
Offered in Winter 2026 at Stanford University.