Nobody’s perfect – Error monitoring in the performing brain

 

Gabriela Bury, December 13th 2016

Every time I attempt to order my favourite coffee at Costa, my tongue decides to fail me in every possible way. I often mispronounce my order so bad I end up walking out of the coffee shop carrying a boring Americano instead of my beloved Mocha, too polite to rectify my mistake. As I suffer my coffee tragedy in silence, I can’t help but wonder – why? What went wrong after I carefully planned out what I was going to say?image002

Performing everyday actions like walking, speaking or making basic decisions seems automatic to us. In reality, our brains need to process a great amount of information for us to execute such precise motor and cognitive tasks. We first formulate theintention to perform a certain task, before our brain plans the action and executes it. The brain is then involved in constant monitoring through auditory, tactile and proprioceptive feedback, making sure that the executed behaviour is consistent with what we intended.

Our brains are continuously involved in this complicated loop of processing, sometimes monitoring many loops at the same time, and it is only normal for something to go wrong every now and then. We stumble over our words, our feet get tangled, we perform the wrong action. This feeds back into the loop, signaling to the brain that we’ve made an error.

The ability to monitor our actions and errors is crucial for goal-directed, adaptive behavior in a changing environment. Error-monitoring has been extensively studied in simple choice or reaction-time tasks, which can be easily manipulated to induce mistakes. Mismatch in monitoring can occur when the participant made an actual error, or when the feedback he received on his action was manipulated in a way that did not match his expectations. EEG recordings showed that errors led to more negative activity in the dorsal Anterior Cingulate Cortex (dACC) of the brain 50 ms after error onset (Error-related negativity, ERN), and feedback errors led to more negative activity in the brain 250 ms after unexpected feedback onset (Feedback ERN; Simon, 2009 ; Nieuwenhuis et al., 2004).

image006

Source : Simon, 2009 and Nieuwenhuis et al., 2004

Although simple choice and time-reaction tasks help study action- and error-monitoring, in real life this process involves many components such as memory retrieval, sensory-motor association, or precise control of motor actions. Something this complex can hardly be represented accurately by repetitively pressing a computer key in a lab, and the study of more complex movement is necessary for a fuller understanding of what happens when we make mistakes. This, unsurprisingly, is when music comes into play.

Musicians are indeed ideal candidates for the study of action monitoring. A professional pianist does not only display impressive motor skills; he is also a memory expert able to retrieve very long pieces of melody, and capable of constant monitoring of his performance through auditory feedback. I am no professional pianist, but I like to fool myself into thinking that my 14 years of formal musical training should enable me to play perfectly a piece I know by heart. More often than not, I end up throwing my book of Chopin’s waltzes on the floor and storming off, knowing perfectly well – as every good neuroscientist should – that my lack of work and motivation are not the ones to blame for my repetitive failures. No, the real culprit is, of course, my brain.

In a study by Maidhof and colleagues (2010), auditory feedback was manipulated so as not to match the expectations of professional pianists who were either playing a melody, or simply listening to one. The participants therefore had the perception of making a mistake, as the feedback differed from the expected melody. The same feedback error-related negativity observed in simple tasks was identified in EEG recordings of both groups of participants, pointing towards similar mechanisms underlying the processing of expectancy violations in more complex tasks such as music.

More interesting yet is what happens in the brain when an actual mistake is made during a performance. Both Herrojo Ruiz and Maidhof and colleagues (2009) analysed piano performances and isolated specific types of errors. All notes were played at the same high tempo, however a delay was observed in the onset of wrong notes, and even the keypress velocity decreased when a wrong note was played. Pianists did not just slow down when mistakes were made; they slowed down before they even had the time to press the wrong key and hear their mistake. This happened even when pianists were prevented from receiving auditory feedback and did not hear their mistake. EEG recordings confirmed this by showing a pre-error negative activity increase 100 ms before the erroneous key was played

image008

Source : Maidhof et al., 2009

A further study by the same authors (2013) used motion capture to record the hand movements of pianists while they were performing, blindfolded. This not only gave rise to funky little piano-playing skeleton hand images, but also supplied behavioural information about what happens before error onset (see https://vid.me/4nS2 to see the skeleton hands in action). This provided data about tactile feedback – the feedback the pianist’s brain gets from touching the right or wrong key. This feedback is obtained as the key is touched, not pressed, and is therefore processed before a key is played. The interval between the onset of tactile feedback (key touch) and the key being played (key press) was greater when an error was committed than when the right key was pressed. The analysis of the combined motion capture and EEG data might suggest that tactile feedback could play a part in error detection, but its role still needs further investigation. All that is sure is that pianists slowed down in response to a mistake they did not yet fully commit.

image012

The pianist’s hand before

image013

The pianist’s hand after

How can the brain possibly predict an error before it happens? Some, like Maidhof and his colleagues, say that it is probably due to the brain’s predictive control processes that compare the predicted outcome of an action with the action goal before its actual realisation. Some, like myself, prefer to think that musicians’ brains are basically crystal balls – and will definitely flaunt those superpowers at the next party they attend. All will hopefully agree that studying the neuroscience of music, however cool and artsy it may sound, can actually help us understand a lot more about our brains’ incredible capacities.

This blog was written following Clemens Maidhof’s presentation to Goldsmiths’ Music, Mind and Brain students on 27/10 2016 as part of the ‘Invited Speaker’ series.

For more details on the Music, Mind and Brain MSc, please visit: http://www.gold.ac.uk/pg/msc-music-mind-brain/.

References

Dikman, Z. V., & Allen, J. J. (2000). Error monitoring during reward and avoidance learning in high-and low-socialized individuals. Psychophysiology37(01), 43-54.

Lutz, K., Puorger, R., Cheetham, M., & Jancke, L. (2013). Development of ERN together with an internal model of audio-motor associations. Frontiers in human neuroscience7.

Maidhof, C., Rieger, M., Prinz, W., & Koelsch, S. (2009). Nobody is perfect: ERP effects prior to performance errors in musicians indicate fast monitoring processes. PLoS One4(4), e5032.

Maidhof, C., Vavatzanidis, N., Prinz, W., Rieger, M., & Koelsch, S. (2010). Processing expectancy violations during music performance and perception: an ERP study. Journal of Cognitive Neuroscience22(10), 2401-2413.

Maidhof, C., Pitkäniemi, A., & Tervaniemi, M. (2013). Predictive error detection in pianists: a combined ERP and motion capture study. Frontiers in human neuroscience7, 587.

Nieuwenhuis, S., Yeung, N., Holroyd, C. B., Schurger, A., & Cohen, J. D. (2004). Sensitivity of electrophysiological activity from medial frontal cortex to utilitarian and performance feedback. Cerebral Cortex14(7), 741-747.

Ruiz, M. H., Jabusch, H. C., & Altenmüller, E. (2009). Detecting wrong notes in advance: neuronal correlates of error monitoring in pianists. Cerebral Cortex19(11), 2625-2639.

Advertisements
This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s