My first paper in the Journal of Research in Music Education describes an interview study in which artist-level jazz improvisers described their thinking while looking at notation and listening to audio from a just completed improvisation. Read article
I later used the same methodology to interview developing improvisers. Interesting, the more experience they had, the more their thinking resembled that of the advanced improvisers. Read article
I'm currently analyzing data from a new study using a similar methodology with improvisers from different cultures
After the improvisers in my first study mentioned melodic patterns repeatedly, I decided to investigate this using a corpus of Charlie Parker solos. I indeed found that Parker uses a LOT of patterns. Read article
I'm currently working with researchers on several new projects using jazz corpora. See the amazing "dig that lick" project based on the Weimar database
Knowing that tonal jazz solos contain lots of repeated patterns, now I wondered if it would be possible to teach a computer to improvise in different styles simply by feeding it solos. Using a mathematical technique called Markov chains, I worked with researches in mathematics and computer science to design a very simply computer program that can improvise in different styles. Interestingly, the program does not know any stylistic rules. It simply reuses patterns from a given input corpus. Read article
Having shown that patterns indeed appear in tonal jazz solos, I wanted to know if I could modulate improvisers' use of patterns by diverting their attention. The answer is yes. When having to do an unrelated counting task while improvising, advanced jazz pianists used more patterns! Read article
In this study published in the Journal of Research in Music Education, I show that music instruction with a focus on improvisation may have cognitive benefits. Read article.
For the second study, we used advanced improvisers who either sang or imagined memorized or improvised music while in an MRI scanner. Using functional Magnetic Resonance Imaging (fMRI) we were able to identify brain areas that were more active during improvisation. Interestingly, the connectivity between those areas was less mirroring the results from our EEG study. Read article
What happens in the brain when you improvise? To investigate this, I worked with researches in neurophysics using electroencephalography (EEG). For the initial study, we asked music students to play simply melodies either from memory or improvised. Interestingly, in line with other research, we saw less connectivity between brain areas during improvisation compared to memorized. Read article
Diffusion Tensor Imaging (DTI)
Here were used DTI to study structural differences between jazz musicians and non-musicians.