Using Music to Develop a Multisensory Communicative Environment for People With Late-Stage Dementia

Amy Clare, DClinPsych; Paul M. Camic, PhD; Sebastian J. Crutch, PhD; Julian West, BA; Emma Harding, Bsc; Emilie Brotherhood, Msc


Gerontologist. 2020;60(6):1115-1125. 

In This Article


The goal of grounded theory is to identify a core or overarching category that brings together key aspects of the analysis (Griffiths & Smith, 2016). For this study, music making was seen as a way for people to communicate in later stage dementias. The essential overarching aspect of the music-making groups appeared anchored in multiple communicative actions that occurred between musicians and residents, musicians and staff, and residents and staff. These were not linear forms of communicating with a clear beginning and end but were often a network of complex and subtle interactions discovered through close observation. The process that operated to form the core category involving these multiple types of communication is a multisensory environment created by the music-making groups. The multisensory nature of the environment provided a unique and supportive communicative atmosphere.

A Multisensory Communicative Environment Through Music for People With Late-stage Dementia

The emerging grounded theory is composed of three interacting components: multiple communicative actions, social interactional components, and agency (Figure 1). Multiple communicative actions act as the core category because they included most of the data (Glaser, 1998). The careful, sensitive noticing of and responding to the communicative actions of participants by the musicians set the foundation for the social interactional components of communication (mirroring, turn taking, and humor) and agency (e.g., choosing to sing or be silent) to take place. A unique multisensory environment that was fundamental to the group supported the development and maintenance of the three components.

Structure of the Grounded Theory

Multiple Communicative Actions. The data consisted of all observable behavior of the participants across the music sessions. The data were coded and categorized into 10 communicative actions: laughing, body positioning, playing an instrument, looking, pointing, dance-like movement, talking, facial movement, smiling, and singing. A description of the communicative actions and sample codes are shown in Table 3. Each action involved a level of physicality in that they involved movement of all or certain parts of the body and contained either nonverbal or verbal components. They formed the core category of communicative actions because, as theorized by Argyle (1988), any verbal or nonverbal behavior that has the potential to influence another individual, intentionally or not, can be seen as communicative. Within the groups, it seemed that each of these actions had the ability to influence the behavior of the musicians or staff and therefore moved from being solely a physical action to a communicative action.

Noticing and Responding. If a communicative action by a resident was noticed it could be responded to by a musician, other resident, or staff member. A communicative action being responded to often led to occurrences of other communicative actions, for example, laughter from a resident leading to laughter from a carer and eye contact. There were times when the communicative actions were not noticed, for example, if the musicians were focusing on a different resident; it was observed that at these times the residents tended to cease the communication attempt or turn their attention to objects around them such as their instrument or their clothes. For example, an interaction with Anne in session seven was observed as follows:

Staff member turned away, [Anne] sitting stiller, looking at baton, looking towards staff member, smiling at her [staff member still turned away] looking down.

Noticing and responding to both verbal and nonverbal methods of communication, were essential because they allowed for the second component of the theory to develop. Yet it is also relevant to point out there are some situations where a response is not expected or required and its absence does not necessarily reduce further communication.

Social Interactional Components

Mirroring. The noticing and responding to the communicative actions allowed for mirroring and turn taking to occur with the musicians modeling these processes within their collaborative music making (Table 4). The musicians used mirroring as part of their practice be it through body positioning, vocalizing, or through the use of their instruments.

Jim, session five: given wood block by person sat next to him, he starts to tap it and musician starts to play the drum mirroring the rhythm. Jane, session seven: musician holding an instrument with Jane and she starts making mouth movements, possibly vocalising, musician mirrors facial expression and mouth movements back to Jane with vocalisations matching her attempts to make sounds

Turn Taking. Turn taking in this study related to a nonverbal or verbal back and forth exchange between two people. Mirroring entered into this process as a way of initiating and prolonging an interaction. Turn taking was enabled by the processes of noticing and responding to the multiple types of communication. The musicians would often use their instruments to start a turn taking exchange in response to the resident playing an instrument.

Tom, Session seven: Tom hits xylophone - each bar in order. Musician responds by mirroring notes and rhythm of this with oboe, Tom then hits the bars again and the musician responds, turn taking exchange continues. Harry, session seven: music stops, tapping fingers in rhythm on drum in silence whilst smiling and sticks tongue out at musician. Takes it in turn with musician to imitate each other's drumming. Smiling and eye contact with musician. Lifts up hands when had enough as in 'I'm beaten' and smiles – everyone laughs.

Both mirroring and turn taking processes require a negotiation by both communication partners and a careful noticing, inviting, or stepping back in response to subtle cues.

Humor. Humor was seen as a social interaction because it required more than one individual and relied upon another person noticing and responding to its use. It involved a complex cognitive process: gaining other people's attention, deciding what to say or do, and carrying this out with the intention of making others laugh. This humor was evident in both verbal and nonverbal behaviors and often had a playful aspect to it. Harry used humor on several occasions, utilizing both verbal and nonverbal means, a particular incident was during the welcome song of session five and the following occurred:

Staff member sang his name – Harry said 'who me' whilst pointing at himself, people laughed, he smiled and laughed and jokily looked inside his jacket for someone else and then looks at musician.

Agency. According to Boyle (2014), agency is "the ability to initiate social action or at least influence own personal circumstances" and can be indicated by both behavior and emotion. Agency (Table 5) was evident throughout the group through the residents having the opportunity to choose how to participate but also through the opportunity to influence the music being created. The musicians were essential to the development of agency (see negative case) through their noticing and responding to the communicative actions of the residents, providing opportunity for social interaction through mirroring, and turn taking as well as inviting the residents to participate in and influence the music being made. As a result, the residents could choose how they interacted with the group and the individuals within it. Participating in the group did not rely on a person playing an instrument and several of the residents chose to interact in other ways as shown by the range of communicative actions observed. For example, Mark spent time with his eyes shut, however, small changes to his facial musculature and his eyes flitting between being open and closed indicated he was not asleep, as shown in session one:

Eyes shut whilst instruments are handed out. Lively music – brisk with oboe – eyes open and mouth less slack. Watching musician across the room introducing an instrument to a resident. Percussion comes in. Eyes shut but mouth tight.

Some of the residents showed agency via how they wanted to interact by choosing to hand back or put down an instrument, for example:

Harry, session three: handed an instrument by musician, they are talking about it and he is turning it over then he hands it back to musician. She returns it to him and points to show another musician demonstrating it. He hands it back to her.

At other times, residents were able to show their agency in the groups by being given the opportunity to direct the music by the musicians (e.g., responding to their communicative actions or inviting them to create a sound by demonstrating how to play an instrument, verbal encouragement or mirroring). This was shown by the examples of turn taking and also in the following observations:

Tom, session five: Tom starts hitting a wooden block with a mallet, the musician joins in gently tapping the drum and mirroring the rhythm Tom is playing. The flautist joins in following this rhythm. Other residents start playing instruments in time or tapping in time. Tom stops and the music stops. There is talking and then Tom is invited to start playing again, he does so, and the musicians follow his rhythm again.

Multisensory Environment

An overarching multisensory environment involving communicative actions, social interaction and agency, was a fundamental aspect of the intervention. It is multisensory in the sense that it contains visual, auditory, physical, and tactile components. The instruments could be seen, heard and touched; everyone was seated in a way that allowed them to see, hear and potentially physically interact each other (e.g., see facial expressions, body movements, see and hear laughter, and spoken words). The multisensory environment reinforced that the group accepted and responded to a variety of different means of communication that relied upon the use of different senses. It also allowed social interactions to occur, as well as the opportunity for agency, using a variety of communicative actions. It gave the residents the opportunity to make use of and interact in ways that appealed to a variety of senses.

Within this multisensory environment, there were three factors that influenced the three main components of the theory. These emerged as theoretical memos during the data analysis following observations that the communicative actions for all the participants tended to change in response to the same factors: time, one-to-one interaction and the characteristics of the music.

Time. Time related to experiencing the multisensory nature of the group on a weekly basis, particularly those aspects of the group that were repeated or similar every week: the music at the beginning as the group when settling in, the welcome song and the music at the end. The familiar melodies used may have orientated the residents to the purpose of the group or acted as a reminder of what happens and how they can participate leading to a change in observed communicative actions over time. This change indicated that some of the residents had remembered parts of the routine aspects of the group as seen by singing along to the words or using more exaggerated movements in each successive session. For example, Jane during the welcome songs in session one where she sat with eyes shut, mouth open and sitting back and then in session seven where she was sitting back, eyes wide open, mouthing along to the chorus whilst looking at the musician. Another participant, Harry, sat during the first session with his hand over his mouth watching the musicians. As each session passed, he uncovered his mouth a little more until by the seventh session he was singing along with his hands in his lap.

One-to-one Interaction. This aspect related to the residents having time where they were receiving direct attention from another individual (e.g., a carer or a musician). The one-to-one interactions in the group were multisensory in nature involving physical closeness, eye contact, clearly visible facial expression, playing of instruments, touch, and spoken words. These interactions seemed to lead to changes in communicative actions for all of the residents.

Anne, session seven: music has stopped and talking, Anne has been looking down at instruments for a while, torso still, foot tapping. Musician looking at and moving to be in front of Anne. Anne looks up. Given a drum, Anne says 'yes' to musician asking, 'shall we play together?'. Anne starts hitting the drum quickly and loudly, smiling, muscles tense. 'Too loud?' she says frowning, 'I don't know?'; musician plays the djembe drum following her rhythm – she sits back a bit and keeps hitting it, now smiling. Harp joins in, she starts playing quicker. Musician sings her name next to her, she suddenly looks up and smiles at her.

There were many instances of these changes but there were also some occasions where a decrease in communicative actions was observed when the person giving the attention turned away:

Jane, session five: Musician has turned to face Jane, sat next to her, sings her name and holds her hand. She turns to face musician, makes eye contact and smiles. Musician turns away and Jane shuts her eyes and smile fades.

Characteristics of the Music. These are related to the tempo, character, and type of instrument used to create the music at a set moment. Tempo refers to the music's speed whereas character refers to its nature, for example, whether the music was gentle, brisk, smooth flowing, or lively or a combination of these. The type of instrument (Table 2) refers specifically to which instrument was playing the predominant line or which instrument seemed to lead to changes in communicative actions when it joined in the music. A larger drum, known as a djembe drum, in particular seemed to lead to a change in communicative action for most of the residents.

The character of the music making was multisensory in nature; instruments could be felt and handled and the music could be heard but also seen through body movements and facial expressions of other people in the room. When the tempo, character, and type of instrument interacted in certain ways, it seemed to produce more changes in communicative action than at other times, specifically when the tempo was faster and the character was lively at the same time when the djembe drum being played.

Mark, session five: eyes drooping, minimal music in the room. Passing instrument back to member of staff. Music gentle. Watching flautist. Increase in tempo and music becomes lively – does not take offered instrument but smiles and laughs. Djembe drum next to him starts being hit. Some nodding, starts to drum on legs in time as music picks up speed. Given shaker, starts tapping it on head in time. Jane, session five: Jane is sat with her eyes closed, music is building through different instruments joining in, tempo increases, lively and brisk, flute playing predominant line, djembe drum joins along with other percussion in the room (other resident playing a tambourine), Jane opens her eyes and looks at musicians, keeps her eyes open, looks at resident playing tambourine.

A Negative Case

A session involving the same participants but listening to recorded music was analyzed following the method described above. This played an important part in the theory's development because the observations during this session did not seem to fit the developing grounded theory. In relation to the multiple communicative actions, it was notable that there were fewer observations for most of the subcategories except for "talking" which increased. For example, one resident (Anne) spent most of the session talking aloud, she did not appear to be talking to anyone in particular as indicated by her lack of eye contact to anybody in the room. When another person responded to her verbally she continued to talk aloud without looking at them. In addition, there were fewer observations of mirroring, turn taking or humor. In contrast to the live music, some residents tried to stand up as though to leave the room, which may have been a way of showing agency by choosing to leave or demonstrating disinterest. With these observations in mind, this study proposes that the live music carried a unique quality by involving a multisensory communication environment.