Implicit learning of complex auditory temporal structures with even and uneven meters

Abstract

Complex auditory sequences (e.g. speech and music) unfold in time. With exposure, listeners can extract regularities in these sequences and develop expectations about the identity (ordinal dimension) and the timing (temporal dimension) of upcoming events. When upheld, these expectations permit faster, and more accurate responses to events, compared to when expectations of the ordinal and/or temporal dimensions are violated (Brandon, Terry, Stevens, & Tillmann, 2012; Buchner & Steffens, 2001; Nissen & Bullemer, 1987; Shin & Ivry, 2002). In everyday life, expectations can be acquired without an intention to do so, and without attention being directed to the regularities in the sequence (Perruchet, 2008; Shanks, 2005). That is, sequential regularities can be learned implicitly. Implicit learning (IL) research has primarily investigated learning of ordinal dimensions of visual sequences. The research presented in this thesis investigated IL of complex temporal structures in auditory sequences with rhythmic features typical of music. Musical rhythms of many cultures have a hierarchical beat structure: meter. Meter is the perception of cyclic patterns of strong and weak beats. The more salient the strong beats, and the more regular and frequent their distribution across time, the stronger the meter is perceived to be. In most Western tonal music, strong beats are evenly spaced in time (even meters). However, in music from the Balkan region for example, strong beats can be unevenly spaced in time (uneven meters). Through a lifetime's exposure, listeners develop expectations of meter characteristic of their musical environment. In the experiments reported in this thesis, rhythms with even and uneven meters are used to examine the flexibility of Western listeners to implicitly learn temporal structures. In the current research project, the development of temporal expectations was examined using an auditory serial reaction time task (SRTT). Participants with no or minimal formal musical training identified as quickly and accurately as possible sequences of auditory events (e.g. pseudo-random ordering of syllables). Unbeknownst to participants, the temporal presentation of events followed a repeating series of inter-onset intervals (IOIs). It was hypothesised that, as temporal expectations were acquired over exposure blocks, reaction time (RT) to identify the syllables would decrease. It was also hypothesised that, as temporal expectations were violated at the introduction of a new rhythm at a test block, RT would increase. The results of six experiments provide evidence of IL of temporal structures, or auditory rhythms, when the ordinal structure was unpredictable. Learning occurred not just of the grouping structure but also of the timing between groups of events. Furthermore, listeners of Western tonal music implicitly learned rhythms with culturally familiar and culturally less familiar meters. Together, these findings demonstrate the capacity of listeners to develop temporal expectations of musical rhythms that either uphold or violate long-term, culturally acquired expectations. The findings also highlight the efficacy of implicit learning as a means of developing temporal expectations in a single exposure session. Demonstrations of IL of temporal structures have important implications for settings where an explicit instruction to learn a rhythm may impeded learning (e.g. music education, motor skills rehabilitation, speech therapy).

Date of Award	2013
Original language	English