record

Thesis Info

LABS ID
00606
Thesis Title
The neuroscience of musical creativity using complexity tools
Author
Dr Shama Sarwat Rahman
2nd Author
3rd Author
Degree
PhD
Year
Oct 2013
Number of Pages
188
University
Imperial College London
Thesis Supervisor
Kim Christensen
Supervisor e-mail
k.christensen AT imperial.ac.uk
Other Supervisor(s)
Henrik J Jensen, Joydeep Bhattacharya
Language(s) of Thesis
English
Department / Discipline
Physics/Neuroscience/Music
Languages Familiar to Author
English
URL where full thesis can be found
hdl.handle.net/10044/1/23280
Keywords
Creativity Neuroscience Complexity Mathematics Brain Patterns Ideation Music Improvisation Interpretation Jazz Classical Piano
Abstract: 200-500 words
This project is heavily experimental and draws on a wide variety of disciplines from musicology and music psychology to cognitive neuroscience and (neuro)philosophy. The objective is to explore and characterise brain activity during the process of creativity and corroborating this with self-assessments from participants and external assessments from professional “judges”. This three-way experimental design bypasses the semantically difficult task of defining and assessing creativity by asking both participants and judges to rate ‘How creative did you think that was?’. Characterising creativity is pertinent to complexity as it is an opportunity to comprehensively investigate a neural and cognitive system from multiple experimental and analytical facets. This thesis explores the anatomical and functional system underlying the creative cognitive state by analysing the concurrent time series recorded from the brain and furthermore, investigates a model in the stages of creativity using a behavioural experiment, in more detail than hitherto done in this domain. Experimentally, the investigation is done in the domain of music and the time series is the recorded Electroencephalogram (EEG) of a pianist’s whilst performing the two creative musical tasks of ‘Interpretation’ and ‘Improvisation’ manipulations of musical extracts. An initial pilot study consisted of 5 participants being shown 30 musical extracts spanning the Classical soundworld across different rhythms, keys and tonalities. The study was then refined to only 20 extracts and modified to include 10 Jazz extracts and 8 participants from a roughly equal spread of Classical and Jazz backgrounds and gender. 5 external assessors had a roughly even spread of expertise in Jazz and Classical music. Source localisation was performed on the experimental EEG data collected using a software called sLORETA that allows a linear inverse mapping of the electrical activity recorded at the scalp surface onto deeper cortical structures as the source of the recorded activity. Broadman Area (BA) 37 which has previously been linked to semantic processing, was robustly related to participants from a Classical background and BA 7 which has previously been linked to altered states of consciousness such as hypnagogia and sleep, was robustly related to participants from a Jazz background whilst Improvising. Analyses exploring the spread, agreement and biases of ratings across the different judges and self-ratings revealed a judge and participant inter-rater reliability at participant level. There was also an equal agreement between judges when rating the different genres Jazz or Classical, across the different tasks of ‘Improvisation’ and ‘Interpretation’, increasing confidence in inter-genre rating reliability for further analyses on the EEG of the extracts themselves. Furthermore, based on the ratings alone, it was possible to partition participants into either Jazz or Classical, which agreed with phenomenological interview information taken from the participants themselves. With the added conditions of extracts that were deemed creative by objective judge assessment, source localisation analyses pinpointed BA 32 as a robust indicator of Creativity within the participants’ brain. It is an area that is particularly well connected and allows an integration of motoric and emotional communication with a maintenance of executive control. Network analysis was performed using the PLV index (Phase Locking Value) between the 64 electrodes, as the strength of the links in an adjacency matrix of a complex network. This revealed the brain network is significantly more efficient and more strongly synchronised and clustered when participants’ are playing Classical extracts compared to Jazz extracts, in the fronto-central region with a clear right hemispheric lateralization. A behavioural study explored the role of distraction in the ‘Incubation’ period for both interpretation and improvisation using a 2-back number exercise occupying working memory, as the distractor. Analysis shows that a distractor has no significant effect on ‘Improvisation’ but significantly impairs ‘Interpretation’ based on the self-assessments by the participants.