Sensing the beat

Article: Grahn, J. A., Henry, M. J., & McAuley, J. D. (in press). FMRI investigation of cross-modal interactions in rhythm perception: Audition primes vision, but not vice versa. NeuroImage.

For those who have read some of my blogs about ICMPC 11 you may already know the name Jessica Grahn. Jessica is a very talented researcher based at the MRC at Cambridge University, who does a cracking job of presenting as well. She does a lot of work on rhythm perception and production (including the work presented at ICMPC 11). This is not really my field, but I always enjoy reading about her studies. A new paper authored by her has just appeared in the journal Neuroimage, and I have spent my early morning delving through the contents. Here, as usual, is my little potted summary.

Introduction

Understanding how our brains use information from the environment to keep time is one of the holy grails of music psychology. We seem so intuitively driven to sense and respond to a beat that many have argued, quite persuasively, that beat perception must be a key component in the evolution of musical ability. In this recent paper Jessica and her team focus on how the brain uses and combines temporal information from different modalities (i.e. visual, auditory) in order to guide behaviour.  More specifically they are interested in the idea that the auditory system demonstrates a ‘specialization (and priority)’ for processing temporal information (including that of sensing a periodic “beat”) over other sensory systems.

Method

For method, they chose to compare beat perception in the auditory and visual modalities using a behavioural paradigm that has been recently developed by Jessica and her co-authors Molly Henry and Devin McAuley .

Participants are presented with sequences of discrete elements (tones or flashes of light) in a range that is maximally reliable for inducing beat perception in humans (600ms). After hearing a reference sequence of 4 elements (two elements followed by a break, then the other two elements) participants must judge whether a subsequent sequence is speeded up or slowed down by comparison (see ‘control trials’ in Figure 1 below). The speeding up or slowing down was achieved by increasing or decreasing the time interval before the final element. In ‘test’ trials, the implied beat of 600ms is made ambiguous by inserting a rogue element 300ms in-between the first and second elements of the first group. Basically this means that the final element of sequence in this case always sounds a little ‘early’ or ‘late’ with respect to the expected beat.

The final crucial element to the experiment was the trial order: visual followed by auditory (VA) or auditory followed by visual (AV). The authors were interested in whether the order in which they completed the trials affected beat perception, as well as whether different brain regions were associated with beat perception in different sensory modalities. Participants did practice trials to get familiar with the procedure, and then they were put into an fMRI scanner.

Results

Behaviour: Participants were clearly able to detect beat changes in both modalities. Individuals showed greater sensitivity to beat perception in the auditory modality, as measured by the authors’ own measure of ‘beat sensitivity’. The details of this measure are rather complex but basically it refers to how much participants are able to extract the implied 600ms beat in test sequences and how much they rely on the rogue 300ms element (therefore showing less sensitivity to the implied beat). Another interesting finding was that sensitivity to a visual beat was improved after listening to an auditory sequence (AV condition); this was not explained by an improvement in simple temporal discrimination – it was unique to beat sensitivity.

Brain: The authors found mostly expected, but also one or two unexpected areas of specialist activation depending on whether stimuli were visual or auditory. Also a network of overlap across modalities that was consistent with previous research (including my favourites; BA 44, SMA, premotor cortex and superior temporal gyri – they often appear in language/music overlap studies). Interestingly the order of stimuli presentation also affected brain activations: Activity in the putamen (part of the striatum within the basal ganglia) was always noted during auditory condition, but in the visual conditions only the AV participants showed activation in this region that was correlated with beat sensitivity.

Conclusions

The present evidence supports the idea that beat perception is superior in auditory sequences compared to when the exact same stimuli are shown in the visual modality.  With regards to the cross-modality findings, the authors propose that exposure to auditory rhythms primes an internal representation of a beat, which can then be exploited by the visual system. This increase in visual beat sensitivity following auditory rhythms was associated with an increase in activity within the basal ganglia, a result which supports this brain area as being crucial to the internal beat generation in humans.