Day 3 – The power of musical imagery

Day 3 of the Edinburgh conference began with brilliant sunshine and a light breeze. I awoke early to put up my poster, and those of other members of my research group in the Playfair library, an amazing long reading room. I spent a few minutes perfecting the art of sticking pins into the presentation boards using only one hand before going anywhere near the precious posters! (My wrist was still in a brace at this point following a sprain). Eventually however, all the posters were up and looking lovely. I was eagerly anticipating my poster presentation, talking to whoever desired to learn more about our work on visuo-spatial processing in amusia.

Once this task was complete, our group convened at the Chocolate Soup cafe and had another lovely breakfast sitting in a sun drenched window alcove. It felt like a morning meal in the Mediterranean rather than Edinburgh! I had some work to catch up on before heading to one of the most interesting symposium for me; that called ‘Mind and brain in musical imagery’.

The symposium was convened by Andrea Halpern and Robert Zatorre, who have been working together in this area now for some time. The focus of the talks was on our ability to re-experience music in the mind. This retrieval is often under one’s control, and can be useful for musical practise or simply personal entertainment. The authors argued that at least for well learned music, musical imagery captures several nuanced aspects of the actual music with remarkable fidelity.

1)      Andrea Halpern – Dynamic aspects of musical imagery

Andrea reviewed a number of the recent studies from her lab that have demonstrated that the basic qualities of music, including pitch and tempo, can be represented in auditory imagery; although it should be noted that all the work she presented had been conducted on individuals with at least 8 years of music lessons.

First, she presented a fascinating ‘real-time’ paradigm where she tracked  emotional response to both real and imagined music, by having people continuously update (2 clicks per second) ratings of valence (positive or negative) and arousal (agitated to calm). She showed the patterns of both over time, for rating taken during both imagined and real music: they overlapped almost identically, showing that people’s imagery of an evolving piece must have been very similar to the experience of hearing the music.

Second, she carried out an ‘anticipatory imagery’ paradigm in an fMRI scanner. People were played familiar pairings of songs and they looked at the activation between the first song ending and the second song beginning. This is the equivalent of looking at your brain activity when you listen to a favourite album, and as a song ends you know very well the track that is coming next. In this gap, even though there was no music, they found activation in the motor areas, cerebellum, basal ganglia and putamen; all areas traditionally associated with hearing real music. What’s more, the level of activation correlated highly with self-ratings of the vividness of the imagery.

2)      Peter Keller – Mental imagery in musical performance

Peter is interested in the numerous anecdotal reports from music performers, that auditory and motor imagery are a vital part of their practise routine. He sought to understand the brain dynamics behind this process and to understand if and why musical imagery, which is a relatively demanding cognitive process, really benefitted musicians’ preparations.

He showed a number of studies looking at both online and anticipatory imagery. He showed that imagery enables, through action planning, movement execution that is characterized by efficiency, temporal precision, and biomechanical economy. He also had some interesting notions that by working with mental imagery of an upcoming performance group musicians can actually improve their predictions about others’ action timing. In the end he concluded that effortful mental imagery exercises are justified when artistic perfection is the goal of performance.

3)     Petr Janata – Acuity of mental representations of pitch

Petr showed some very interesting work looking at how individual ability to represent pitch mentally can vary widely, and also that this ability is likely supported by multiple memory systems that can be boosted with musical training. His tasks were a fairly standard two-tone discrimination format, based on the Deutsch paradigms of the 1970s, and a scale mis-tuning task, where you mistune the last tone of a scale and then remove the few tones that immediately precede that wrong note. Overall, people do well on both tasks but acuity improves with musical training.

He also presented some ERP data with the same paradigms where he divided people into good and bad performers using a median split. He showed that the good performers demonstrated a stronger P3a component in their ERPs, which he interpreted as an indication of their ability to better detect deviant pitches. Good performers also had a smaller N100 component, which he suggested reflected their improved ability to instantiate accurate mental pitch images.

4)      Robert Zatorre – Beyond auditory cortex: working with musical thoughts

Robert is, of course, most well known for his neuroimaging studies of musical imagery  so I was looking forward to hearing his latest findings. At the start of his talk he played a little game with the audience asking them to imagine ‘Auld Lang Syne’  in their head. Everyone nodded that they could do this well. Then he asked us to imagine it played on a xylophone; again a good audience response. Then he asked us to image the hip hop version! This was a good illustration of the power and flexibility of musical imagery, but for me it had an unfortunate side effect – I had my dodgy hip hop version  in my head all the way through lunch!

Robert showed two groups of new studies, the first looking at imagery during a transposed motif reversal task (i.e. can you spot a version of a tune that has been reversed once it is shifted a key) and recognition of temporally reversed melodies (i.e. where the tune is the same but backwards).

Using imaging during these types of tasks he has found a network of activity in the brain outside the traditional auditory cortex, which he postulates is important for imagery. He particularly focused on the intraparietal sulcus region. He suggested that this region had a key role to play in effecting transformations of sensory inputs from one reference frame to another, thereby allowing us to play with sounds in our heads. This has fascinating implications for areas like composition, and showing how new musical ideas might emerge from existing musical knowledge.