Home for the Curious Musician

Improvisation and the Brain

Photo Credit: Tiago Bortoletto Vaz

We recently found this fantastic piece on how musical improvisation affects the brain on Vicky Williamson’s blog Music Psychology and she graciously allowed us to post it here. Vicky teaches and researches on the relationship between music and the brain at The University of Sheffield. Check out more of her writing here.

Hello Dear Reader,

It is that time of year again. Undergraduates are considering whether or not to continue their higher education following the completion of their undergraduate degree. They are deciding whether to pursue their Masters.

Here at the University of Sheffield we have been busy putting on many Open Days and Evenings, providing information, tours and advice/support for such students.

Recently we held our first evening Open event at the Department of Music, providing a convenient time slot for those with full time jobs or daytime responsibilities.

It was lovely to get away from my computer screen for a couple of hours and spend some time chatting with these (hopefully) soon-to-be postgraduates. So enthusiastic and bursting with ideas. It reminded me of the main reason I originally went into academia; to teach.

Armstrong-in-motionOn this particular Open Evening I met a young student by the name of Robbie. I hope he does not mind me using his name, should he read this blog, and forgives me a possible misspelling. He and I got to chatting about his interest in music psychology and it turned out that Robbie was intrigued by the processes of musical creativity and improvisation.

How do they happen? What is the brain doing when new musical ideas spring to mind? And what might the answers to these questions tell us about the power and possibilities of the human mind ‘at play’?

Well Robbie, this blog is for you.

Shortly after we spoke I found a new article (January 2015) that aimed to summerise the present state of academic understanding regarding brain activity we observe during musical improvisation. The article was written by Roger E. Beaty from the University of North Carolina at Greensboro, United States. It appears in a journal called Neuroscience and Biobehavioral Reviews and is entitled “The neuroscience of musical improvisation”.

Below I paraphrase and summarise the contents of that article. You may find a direct link to the abstract of the article by clicking on this text.

What is improvisation?

Musical improvisation is one of the most demanding tasks a musician can undertake, requiring the real-time generation and production of novel melodic and rhythmic sequences in line with an ongoing musical idea or context. We can suppose that an understanding of how musical improvisation comes about might inform us a great deal about the general processes of human creativity.

One key aspect of the creativity debate is, to what extent does it depend on our ability to ‘let go’ of normal cognitive control? To give in, allow spontaneous thought to burst forth in the hope that we can instantaneously form those ‘rough’ ideas and thoughts into a productive and artistic action.

by Robert Couse-Baker
by Robert Couse-Baker

The most notable theory in this area is that of Pressing, from the 1980s and 1990s. His work followed on from and was informed by the study of expertise, led by Ericsson et al., 1993 and others in psychology.

The Pressing theory states that an expert has undergone a substantial amount of deliberate practice (the oft quoted 10,000 hours minimum) and that this process has automated many of the cognitive and behavioural actions that are demanded in musical performance.

The result of the automating is a partial release from focused attention – in effect, while an expert musician is performing, there is enough ‘spare’ cognitive capacity for other higher order processing, including the generation and on-line evaluation of musical ideas.

Pressing theory claims that improvisation results from the interaction, allowed by this spare cognitive capacity, between long-term domain knowledge and ‘referents’ – ongoing perceptual, cognitive and emotional processes.

Finally, the Pressing theory has an idea about what makes a great improvisation; effective perceptual feedback and speeded error correction. These processes rely on a well tuned system that compares actual output with intended output on a second by second basis, thereby generating future ideas in response to the current instant of music production.

All this depends on well-rehearsed action sequences. In effect, what appears spontaneous and creative to us on the musical stage can often be the result of well-rehearsed patterns and their application at speed. This theory is supported by research into the improvisations of Charlie ‘Bird’ Parker (follow this link for my blog on that paper).

How to study improvisation in the scanner?

If Pressing is right then we should see some telling patterns of brain activity when we scan an individual who is improvising. And perhaps, we will see some overlap with the activations seen during other kinds of creative tasks. Wouldn’t that be neat.

But first we have a problem… how do you improvise in a brain scanner?

I urge you to take a moment and glance at the videos of Charles Limb (he has an excellent TED talk here), a neuroscientists and jazz musician, who took it upon himself to try to answer the above question.

In one of his first studies Limb asked keyboard players to memorise a novel melody before going into the scanner. He then lay them down in the scanner with access to an MRI-compatible keyboard at their fingertips. Not an ideal position but the musicians were able to complete their tasks which were to; a) play their newly memorised melody, b) play a normal scale, and c) improvise.

Limb reported that the task of improvisation was associated with unique activation in a brain network, including the left inferior frontal gyrus (IFG), anterior cingulate cortex (ACC), and the medial prefrontal cortex (MPFC), among several others. What was perhaps most interesting was the comparable widespread deactivation of regions in the frontal lobe of the brain that are associated with conscious monitoring of ongoing activity – perhaps this was the brain ‘released’ from its normal attentional focus and free to ‘play’?

Is the frontal brain just ‘switched off’?

Studies that followed the Limb work found that the story was not that simple – the front conscious monitoring areas of the brain were not all shut off during improvisation. In fact, many of these areas were very active. Bengtsson et al. (2007) argued that this continued activation in frontal areas of the brain makes sense as not everything that might emerge spontaneously in the brain is worth consideration – ideas and thoughts must be monitored, evaluated and selected for action. This executive decision process requires our precious frontal lobes (especially pre-frontal lobes, where motor actions are also driven) to be working, not all switched off.

The default network

The ‘default network’ is a popular new area of study in neuroscience. This pattern of activation in the brain is often reported when a person is lying in a scanner without an overt task to focus on. In these circumstances the mind can switch, at least partially, to internally driven thought processes. You might call it ‘mind wandering’.

This pattern of activation overlaps with that associated with musical improvising.

concentration-vwThe suggestion is that improvising relies on similarly internal self-referential processes, which suspend external conscious monitoring in favour of a partial or almost complete focus on internally generated thought.

Perhaps this is the key to the ultimate demonstration of focused thought in any task; the so-called ‘flow state’. Here concentration is optimal, time melts away, and we are completed absorbed by the task at hand.

However, we must remember that no study to date has argued that improvisation is like mind wandering for the very good reason that improvisation during performance requires that the musician to retain at least a grasp of the musical present, especially if they are part of a group. Without the focus on the present the improvisation may stray away from developing musical ideas and contexts and, ironically, cease to be regarded as optimally creative.

Creativity is not about being utterly random. It is about making the very best outcome from a possible input, one that surprises and delights.

Do you have to be an expert?

As you will recall, Pressing argued that expertise was the foundation of successful improvisation. Without expertise there is no automation, no freeing of precious attention resources, and hence, no room for the mind to play.

Studies have found evidence to support Pressing’s theory in regard to brain patterns. It appears that, when compared to a matched group of non-musicians, expert musicians reliably show the pattern of de-activation with selective focus described above. This pattern is not seen in non-musicians or people in the early stages of training. In particular the de-activation into default network does not seem to happen.

One brain area of note in these studies is the right temporoparietal junction (TPJ). Modulation of activity in this area may provide “task shielding” by filtering or inhibiting information unrelated to the task at hand. It has been reported in tasks of divergent thinking, creative writing, and spontaneous design.

Perhaps therefore, we have at least one candidate for a global brain system that allows experts to embed their mind’s activity within their current task, keeping enough focus on incoming material while shielding their new thoughts from other potentially disruptive attention-related brain processes.

What about playing in a group?

You might be thinking that it makes sense for a different kind of network to be active when a soloist is improvising as to when a group is engaged in collective improvisation. The latter requires more focus on the outside world, yes? You would be right.

improv-in-a-group-vwThere are not many neuroscience studies in the area of collective improvisation – practical constraints being what they are in a scanner – but one recent paper by Donnay et al. (2014) found less activation of the default network in group improvisation situations compared to that reported by Limb in his studies of solo improv.

The stronger demands on performance monitoring in the group situation preclude a fuller ‘shut down’ of external monitoring processing in favour of a tuning into the default network.

But does that mean that group improvisation can never be as ‘good’ as a solo effort? I leave that thought to you, Dear Reader…


Studies of musical improvisation have provided a unique window into the brain at play, and have found several fascinating links to wider studies of creativity across multiple domains.

Several intriguing questions now present themselves – how does the process of brain activation associated with expert improvisation come to pass as someone develops as a musician? Are some people naturally more inclined to such activations, making them naturally ‘better improvisers’? Can we distinguish ‘good’ and ‘not-so-good’ improvisation attempts by reference to the underlying precipitant brain activations? And what , if anything, do these brain studies tell us about how we might teach improvisation to new musicians?

Get the top Flypaper articles delivered straight to your inbox once a week.

Guest Writers

Soundfly welcomes new voices each month to offer unique perspectives, shine a light on unexpected musical worlds, and help our readers find their sound. If you'd like to join us with a guest post, please send articles and inquiries to support(at)!

Get music news and tips delivered to your inbox once a week.