Music Style Analysis Could Assist Therapists

July 5, 2019

The inexpressible depth of music, so easy to understand and yet so inexplicable, is due to the fact that it reproduces all the emotions of our innermost being, but entirely without reality and remote from its pain.  Oliver Sacks, Musicophilia: Tales of Music and the Brain

In this quote Sacks is drawing on a long history of using music to diagnose and treat mental health: The ancient Greeks assigned both music and medicine to the god Apollo (who also embodied ideals of masculine beauty). In the twentieth century, music therapy was developed to treat soldiers who had suffered mental and physical trauma in the world wars. How does the act of creating music calm the soul and heal the mind?

New research at the Weizmann Institute of Science may help crack the mystery by quantifying some of the connection between music making and the behavioural effects this activity can arouse.

Some may find the connection between arts therapies and computerized modelling a bit surprising, but for Dr Billie Sandak, both subjects are close to her heart, and she has found a way to use one to provide insight into the other. Sandak is in the group of Professor David Harel of Weizmann’s Computer Science and Applied Mathematics Department.

To construct a computer-based method of investigating a creative process such as playing a musical instrument, Sandak and Harel first had to identify some objective parameters for analysis.

“The ears and eyes of therapists cannot take in every detail of the complex dynamics of the creation process, and hence they may write progress reports that are subjective,” said Sandak.

The method they developed is based on a computer language called Statecharts, created by Harel in the 1980s, which is a visual language for specifying the behaviour of large, multi-component reactive systems. This includes the computerised systems for operating aircraft, but it applies equally to biological systems and social ones. With Statecharts, users can create complex diagrams of all possible events and the transitions between actions, revealing hierarchies and sequences.  Since Statecharts is just like any programming language, the resulting diagrams can then be run, simulating the behaviour of the system. Sandak and Harel had first presented the method and demonstrated how it could be used in art therapy with a modest group of subjects participating in drawing sessions.

In the present study, conducted in collaboration with Drs Avi Gilboa and Shai Cohen of the Music Department at Bar-Ilan University, they used the insights of the earlier research to conduct a broader investigation into a different, related therapy: one that involves playing a musical instrument. The results of this research were recently published in PLoS ONE.

Music therapy most often involves improvisation – whether or not the patient has had any musical training – and the idea is that the unspoken language of music can express emotional states in a way that words do not.

The researchers worked with 108 subjects, men and women, between the ages of 18 and 77. Some had musical training at various levels; others had none at all. All were given the same four tasks: to play improvised pieces on an electronic piano keyboard on four different themes: ‘negative feeling’, ‘ugly’, ‘positive feeling’ and ‘beautiful’.

The music expressing negative emotions and ugliness was, indeed, quantitatively different from that expressing positive emotions and beauty. The negative concepts involved the use of lower notes, more intensity and the playing of several keys at once. The idea of ugliness required the use of more keys even than negative emotions, while beauty involved a softer touch on the keyboard than positive emotions.

There were other differences, as well. Between men and women, for example: Women explored fewer parts of the keyboard, employed a smaller range of notes and intensities than the men, and even pressed fewer keys. This is similar to one of the findings of the earlier research – namely, that women tend to use fewer drawing tools than men.

Age also played a role, as well. Compared to the younger subjects, the older ones played with more keys pressed simultaneously and preferred the black ones – that is, their style exhibited more of the characteristics of negative emotions. For Sandak, this resonates with the earlier research.

“Older people used less color in their drawings and erased more,” she said.

Of course, in music therapy, as in any therapy, there is quite a lot going on around the act of playing music. The researchers now plan to apply their computational method to the more ambitious goal of analysing additional occurrences in the therapy room, including the patient-therapist relationship.

“Ultimately, we want to create a tool that will help the therapist track progress and change through creative therapy in real time,” said Sandak.

“This could be a kind of dashboard enabling the therapist to address issues throughout the session — for example, ‘does the patient avoid high notes and low intensities, and are his/her notes played clearly or in a muddled way?’ ‘Does the patient avoid the colour green in drawing and are his or her lines intense or erased repeatedly?’  Such information, made possible by research results, might be used by therapists to improve treatment.”

Professor David Harel’s research is supported by the estate of Emile Mimran. Professor Harel is the incumbent of the William Sussman Professorial Chair of Mathematics.


More Research

When the Light Particle Saw the Light

A Weizmann Institute study of photons in quantum computing has made a surprising discovery: When photons collide, they create vortices. Vortices are a common physical phenomenon. You find them in…

eastRead More

Mouths Wide Shut: How Cellular Cleaning Out Is Kept in Check

Weizmann Institute of Science Research has discovered that autophagic organelles actively restrict the size of their mouths – but what does that signify? Autophagy literally means ‘self-eating’ and is a…

eastRead More

From Tip to Tail

Researchers from the University of Michigan and the Weizmann Institute of Science have developed the first-ever complete central nervous system on a chip, emulating that of a human embryo, from…

eastRead More

AI Flexes Its Muscles

Life sciences have never been more digital. To learn more about life processes, biologists are collecting massive quantities of data that computer scientists analyse by means of sophisticated computational models…

eastRead More

View All