Caroline Palmer, PhD

Professor, Department of Psychology
McGill University
(514) 398-6128

My research program combines two related issues in cognitive psychology: how people remember long sequences typical of speech and music, and how they produce those sequences. Many theories of memory for speech, written language, pictures, and other human endeavors focus on the problem of serial order: knowing what comes next in a sequence. What most theories do not address is the time course of retrieval: when particular sequential information is available, and for how long. My research focuses on the time course of serial order in music performance, one of the most complex of human skills. Not only must item information be recalled correctly in music, but temporal information (when and for how long events should occur) must also be recalled correctly. My work has established that performers’ memory for musical sequences can be extremely accurate (smaller than 3% error rate), despite the complexity, length, and temporal requirements of music (seeFinney & Palmer, 2003).

One theme of my research addresses the range or scope of planning in sequence production; the range of planning, similar to memory span measures that gauge short-term memory capacity, refers to the span of sequence items that are accessible at a given time during performance (see Palmer, 2005). We have developed a formal theoretical framework for serial recall in music performance that makes time-dependent predictions of the range of planning during production (see Palmer & Pfordresher, 2003), as well as an account of the relationship between speed and accuracy (see Pfordresher, Palmer, & Jungers, 2007). We also address how the scope of planning changes during different stages of skill acquisition (see Palmer & Drake, 1997). Child pianists’ performances indicated skilled-related increases in monitoring one’s own behavior, in anticipating upcoming events, and in generalizing beyond specific motor movements. Comparisons with adult performers indicated these cognitive capacities change most during the first 5 years of skill acquisition, whereas domain-specific knowledge increases across all stages of skill acquisition (see Palmer & Meyer, 2000).

A second theme addresses the motor actions that underlie skilled performance, and properties of goal-directed movement that are specific to individuals. Using motion capture techniques, we record pianists’ finger movements. Dynamic (time-dependent) properties of motion specific to individuals and fingers inform about how personal identity may be rooted in voluntary, goal-directed actions (see Dalla Bella & Palmer, 2006). Motion during sequence production also relects the biomechanical constraints that arise from musicians’ finger and hand movements as they tap on a table (see Loehr & Palmer, 2007), or perform on an instrument (see Palmer, Carter, Koopmans & Loehr, 2007).

A third theme of our research extends our memory findings to understanding how people perceive stable, categorical events in a continuously changing world. Listeners tend to perceive musical sequences as temporally regular; people without any musical training can clap along to continuously fluctuating music with little effort. An ability to perceive temporal regularity is remarkable, given our findings that actual music performances are temporally irregular. We model with dynamical systems approaches how listeners are informed by the temporal fluctuations in music performance, by capitalizing on the systematic nature of the variability (see Large & Palmer, 2002). We have also shown that synchronization of tapping with music is aided by listeners’ sensitivity to phase differences in the onsets of produced and perceived beats (see Loehr, Palmer & Large, 2007). This work indicates that temporal structure is fundamental to understanding how people perceive meaningful units in a continuously varying auditory world, and it offers a primary resource for aiding people’s memory for and learning of auditory sequences