Participants

The participants (see Table 1) were five families that met the following selection criteria: the family nucleus included the mother, father and at least one child aged between 6 and 36 months; the children represented different developmental stages, and families resided in the same urban area (Barcelona, Spain). Musical background was not a selection criterion. However, in order to better understand the context of each family, parents were later asked about their musical and sociocultural backgrounds during interviews. These families were contacted through the researcher’s personal and professional references, in view of the characteristics of this study and the type of data collection instrument used (see section 2.2). The participants had to make a considerable commitment because each audio recording of a family day was supposed to be a total immersion in their everyday private lives.

Table 1.

Participants

Table 1 Long description

A table with five rows and six columns. The columns are labeled Family, Parents, Child, Siblings, Date of birth, and Age of child at the time of data collection. Row 1: Family 1, Parents Alba/Gabriel, Child Bruna, Siblings 0, Date of birth 18/07/2019, Age of child at the time of data collection 28-34 months. Row 2: Family 2, Parents Ana/Raúl, Child Alex, Siblings Yes, Date of birth 05/03/2020, Age of child at the time of data collection 22-33 months. Row 3: Family 3, Parents Julia/Thomas, Child William, Siblings Yes, Date of birth 23/05/2017, Age of child at the time of data collection 19-26 months. Row 4: Family 4, Parents Karina/Alejandro, Child Charles, Siblings 0, Date of birth 28/06/2020, Age of child at the time of data collection 16-26 months. Row 5: Family 5, Parents Cristina/Salvador, Child Alan, Siblings 0, Date of birth 01/02/2021, Age of child at the time of data collection 9-15 months.

Information on each family’s social and musical background, as reported in the interviews, is included to contextualise the findings. These details enrich interpretation but were not used as inclusion criteria.

• Family 1 (F1): Both parents, university lecturers in early childhood education, value music as a natural, holistic language that plays a central role in family life and cultural transmission. Alba studied piano and solfeggio formally; Gabriel is self-taught in guitar and singing.

• Family 2 (F2): The mother holds a degree in psychology; the father completed secondary education. Ana learned how to play the guitar and theory until age 13; Raúl has no musical training. They use music to navigate challenging parenting moments and enrich daily routines.

• Family 3 (F3): The mother has a PhD in education and degrees in music didactics and musicology; the father studied art history and attended piano classes at a conservatory. They view music as a playful, child-led form of shared experience.

• Family 4 (F4): The mother is a dentist, and the father has a degree in political science, economics and finance. Karina learnt to play percussion instruments at school, and Alejandro had some piano and flute classes when he was a child. They describe music as a way of transmitting calm and a sense of well-being to their child.

• Family 5 (F5): The mother holds a degree in music education and works as a primary school music teacher. The father completed secondary school and took some piano lessons. They view music as an essential family resource with proven benefits.

Data collection instruments

Data analysis

Two procedures were completed simultaneously: examination of the audio recordings and the interviews.

1. (1) Recordings

Firstly, we listen to each one of them (around 420 hours in total) in their entirety. Episodes that featured participation and MI through the voice were identified. MI between the participants was considered if at least one of them employed a vocal discourse that articulated tonal, melodic, and/or rhythmic patterns (Cross & Morely, Reference CROSS, MORLEY, Malloch and Trevarthen2009) to create organised sound (Blacking, Reference BLACKING2006), even if the other party’s vocal discourse was more associated with the communicative expressions of language and the emotions. A spreadsheet with an indexing table was used to log this phenomenon (see Table 2). The timeline of the events in each recording was noted every 30 seconds, and annotations were made indicating the participants and the context.

Table 2.

Part of the indexing table

Table 2 Long description

A table with three columns: Time, Participants in the interaction, and Context. The table has two rows. Row 1: Time, 05:02:00; Participants in the interaction, Mother-son; Context, William's parents have just given him a bath; they are putting on his clothes and combing his hair. Row 2: Time, 05:02:30; Participants in the interaction, Mother-son; Context, Julia starts to sing a lullaby, and William answers her with the same melody.

Following the identification process, each episode (a total of 195 across all recordings) was analysed using the MICAD – Musical Interaction among Children and Adult Descriptors – (see Lerma-Arregocés & Pérez-Moreno, Reference LERMA-ARREGOCÉS and PÉREZ-MORENO2023), specially designed and validated for this study. It was developed from a comprehensive literature review and refined through interaction with the data. The instrument comprises three blocks:

• - Block 1: identification (coding) and duration of each vocal MI episode.

• - Block 2: contextual information (e.g., adult participant, order of participation, setting, situation, probable function).

• - Block 3: musical and communicative features (e.g., type of vocal expression, type of interaction, repertoire, emotional expressions).

This process allowed for both horizontal analysis (in-depth qualitative exploration of each episode) and vertical analysis (global quantification across all recordings).

1. (2) Interviews

The interview transcripts were analysed using a directed content analysis approach (Hernández -Sampieri et al., Reference HERNÁNDEZ-SAMPIERI, FERNÁNDEZ and BAPTISTA2014). Coding was guided by the categories defined in the MICAD, allowing for systematic alignment between parental discourse and observed data. For instance, responses to the question ‘When do you sing to your child/children?’ were associated with the ‘situation’ category (e.g., play time, meals), and answers such as ‘often’, ‘a lot’ or ‘all the time’ were linked to ‘episodes frequency and length’. Table 3 illustrates how selected interview questions corresponded to specific categories from the MICAD instrument.

Table 3.

Relationships between interview content and the categories from the analysis instrument

Table 3 Long description

A table with three columns and five rows. The columns are labeled 'Categories from MICAD instrument', 'Interview question', and 'Indicators (derived from interview)'. The rows are as follows: Row 1: Categories from MICAD instrument, Episode frequency and length; Interview question, When do you sing to your child/children?; Indicators (derived from interview), Frequency and length. Row 2: Categories from MICAD instrument, Order of participation; Interview question, What do you think is the reason (or the reasons) for starting a shared musical activity?; Indicators (derived from interview), The child's immediate needs, the parent's musical interests or the child's. Row 3: Categories from MICAD instrument, Situation; Interview question, When do you sing to your child/children?; Indicators (derived from interview), Moments, scenario, and/or routines. Row 4: Categories from MICAD instrument, Probable Function; Interview question, Why do you sing to your child/children?; Indicators (derived from interview), Causes and/or motivations. Row 5: Categories from MICAD instrument, Repertoire; Interview question, Can you describe the ways in which you sing to your child or with your child?; Indicators (derived from interview), Types of songs, genres from the original cultural heritage or current cultural context.

Episode frequency and length

To understand the families’ level of engagement in vocal MI, we examined the average number and duration of episodes over 5-6 days of continuous audio recordings per family. We also calculated the percentage of daily recorded time (12-16 hours of recorded time) devoted to vocal MI. Table 4 summarises the results:

Table 4.

Frequency of families’ participation in MI

Table 4 Long description

The table has five rows and four columns. The columns are labeled 'Families', 'm = episodes', 'm = length', 'Total time in a day', and 'Participation percentage'. Row 1: Families 1, m = episodes 7.6, m = length 1 minute 21 seconds, Total time in a day 10 minutes, Participation percentage 1.45 percent. Row 2: Families 2, m = episodes 6, m = length 5 minutes 47 seconds, Total time in a day 35 minutes, Participation percentage 4.81 percent. Row 3: Families 3, m = episodes 5.5, m = length 1 minute 34 seconds, Total time in a day 9 minutes, Participation percentage 1.19 percent. Row 4: Families 4, m = episodes 5.8, m = length 1 minute 59 seconds, Total time in a day 11 minutes, Participation percentage 1.61 percent. Row 5: Families 5, m = episodes 8.5, m = length 3 minutes 56 seconds, Total time in a day 33 minutes, Participation percentage 4.63 percent.

Parents were asked: ‘When do you sing to your child/children? [how often and for how long?]’. Their responses reflect frequent and informal use of singing:

• - ‘I sing a lot, with her or without her, not necessarily because Bruna wants me to but because I feel like it. That’s how much we sing every day’ (father, F1).

• - ‘I sing to them a lot, but for no reason in particular, there’s no set time, although it’s almost always at lunchtime and at bedtime’ (mother, F2).

• - ‘Generally, I sing as a way of starting a game with Charles’ (father, F4).

• - ‘At home we sing all the time […] and I almost always use singing as part of a routine’ (mother, F5).

Although parental reports often describe singing as frequent and music as an ever-present element in family life – supporting previous findings (e.g., Custodero & Johnson-Green Reference CUSTODERO and JOHNSON-GREEN2003; de Vries, Reference DE VRIES2009; Evans et al., Reference EVANS, DEAN and BYETT2022; Koops, Reference KOOPS2019) – the recorded data revealed a comparatively brief duration of vocal MI, ranging from 1.45% to 4.63% of the day (M=2,74%, ≈ 16 minutes). This divergence underscores the importance of complementing self-reported data with more naturalistic observation: while interviews shed light on the symbolic and emotional meanings that parents attribute to music, recordings reveal how musical expression emerges in brief yet recurrent forms across the day (this will be explored in the following sections).

Despite their brevity, these MI cannot be dismissed as insignificant. Prior analyses (Lerma-Arregocés, Reference LERMA-ARREGOCÉS2023) show that even short episodes may display rich musical and communicative features – such as anticipation, imitation, synchrony, variation, or melodic and rhythmic development – supporting theoretical perspectives that parents displayed vocal styles depending on the context (e.g., slow and soft for calming, rhythmic and playful for games, Young, Reference YOUNG2023) while infants, in turn, are sensitive to these musical styles, responding with rhythmic or tonal babbling, contour-matching vocalisations, or even initiating turn-taking sequences that resemble proto-duets (Español et al., Reference ESPAÑOL, SHIFRES, MARTÍNEZ, PÉREZ, Español, Martínez and Rodríguez2022; Stadler Elmer, Reference STADLER ELMER2012; Trehub & Gudmundsdottir, Reference TREHUB, GUDMUNDSDOTTIR, Welch, Howard and Nix2015). Thus, short episodes still reveal meaningful musical engagement grounded in expressive exchange.

Order of participation

To examine patterns of participation, we reviewed the 195 episodes to identify who initiated and/or ended each interaction. Table 5 presents both frequency and the percentage distribution of these roles across participants in each family.

Table 5.

Number of times and percentage in which participants initiated and/or ended the MI

Table 5 Long description

A table comparing the number of times and percentage in which participants initiated and ended the MI across five families. The table has five rows and three columns. The columns are labeled Participant, Initiated, and Ended. The rows are labeled F1, F2, F3, F4, and F5, each with sub-rows for Infant, Mother, and Father. Row 1: F1, Infant, 28 (58.3 percent), 25 (43.1 percent). Mother, 13 (27.1 percent), 12 (20.7 percent). Father, 7 (14.6 percent), 21 (36.2 percent). Row 2: F2, Infant, 15 (41.7 percent), 14 (38.9 percent). Mother, 21 (58.3 percent), 22 (61.1 percent). Father, 0 (0 percent), 0 (0 percent). Row 3: F3, Infant, 12 (36.4 percent), 13 (41.9 percent). Mother, 15 (45.5 percent), 12 (38.7 percent). Father, 6 (18.2 percent), 6 (19.4 percent). Row 4: F4, Infant, 14 (46.7 percent), 15 (51.7 percent). Mother, 12 (40 percent), 7 (24.1 percent). Father, 4 (13.3 percent), 7 (24.1 percent). Row 5: F5, Infant, 17 (33.3 percent), 14 (28 percent). Mother, 23 (45.1 percent), 25 (50 percent). Father, 11 (21.6 percent), 11 (22 percent).

Overall, children initiated 43.4% and ended 39.7% of the interactions; mothers initiated 42.4% and ended 38.2%; fathers contributed less frequently (14.1% initiations and 22.1% endings). These data show that mothers and children were more actively engaged in initiating and sustaining MI.

There was often a correlation between who initiated and who ended the interaction. This may be due to: (a) musical closure that naturally concluded the interaction; or (b) a loss of interest by one or both participants.

Parents were asked, ‘What do you think is the reason (or reasons) for starting a shared musical activity?’ Their responses reflect both child- and adult-driven motivations:

• - ‘I think that it emerges from Bruna’s interest or an everyday scenario. For example, a fly comes into the house, and she notices it, so we start singing about the fly’ (mother, F1).

• - ‘Sometimes to calm them down a little but also because they really like music and they ask me for songs’ (mother, F2).

• - ‘As far as I’m concerned, let them begin the activity, it’s rarely me who takes the initiative. The fact is I’m more open to being invited to make music’ (mother, F3).

• - ‘Generally, because I want to get something done, get Charles off to sleep, get him to relax, or when we want to have fun’ (mother, F4).

• - ‘In this case, after I got pregnant, I sang a lot and played the piano all the time, music is something I really live for’. (mother, F5).

In some cases, perceptions and recorded behaviour aligned – e.g., in F1 and F2, both mothers mentioned their children’s interest or their own motivation, which matched the data. In F3, the mother claimed she ‘rarely’ initiated MI, yet she led most often (45.5%). In F4, the mother said she initiated to regulate her son, but the child led slightly more (child: 46.7%, mother: 40%). In F5, alignment was again observed, as the mother expressed strong musical motivation and was the main initiator (45.1%).

The data suggest that MIs are often co-constructed, with both adults and children taking turns in initiating and concluding the exchanges. This points to a dynamic in which musical engagement is not solely driven by adult caregiving goals, but also by children’s own communicative and expressive initiatives. These early signs of mutual responsiveness support the view of MI as a reciprocal practice (e.g., Provenzi et al., Reference PROVENZI, SCOTTO DI MINICO, GIUSTI, GUIDA and MÜLLER2018). In this light, we propose the term Family Musicality (Lerma-Arregocés, Reference LERMA-ARREGOCÉS2023) to refer to the shared musical agency of both children and caregivers, laying the groundwork for deeper exploration in the sections that follow.

Situation

The everyday scenarios in which MI occurred have been categorised for each family. Table 6 presents them in order of the most to the least common detected.

Table 6.

Percentage of vocal MI by scenario and family

Table 6 Long description

The table presents the percentage of vocal MI by scenario and family, categorized from most to least common. It has 17 rows and 5 columns. The columns are labeled F1 percent, F2 percent, F3 percent, F4 percent, and F5 percent. The rows are labeled with different scenarios such as Playtime, Mealtime, Changing nappy or clothes, Emotional regulation, Sleep/nap time, Listening to/sharing music, Moment of entertainment, Bath time, Story time, Going for a walk, Breast-feeding, Waking up, Brushing teeth, Car journey, Conversation, Cooking, Tidying room, Picking up toys, and Unverified. Each cell contains the percentage value for the corresponding scenario and family. Notable trends include high percentages for Mealtime across all families and varying percentages for other scenarios.

Note: [–] indicates no vocal MI episodes in that scenario.

Parents’ responses to ‘When do you sing to your child?’ generally aligned with the most frequent recorded scenarios. For example:

• - F1’s mother mentioned mealtimes – ‘When we’re about to eat’. Recordings confirmed that 43.5% of MI occurred during meals.

• - F2’s mother cited bedtime – ‘Singing is kind of sacred at nap time and at nighttime’, which aligned with 27.8% of MI occurring at that time.

• - F4’s mother mentioned brushing teeth as a moment to sing – ‘I sing to him a lot. I like doing it when I am changing his nappy in the morning, when I pick him up, and sometimes I make up songs to brush his teeth’. Even though it wasn’t the most frequent scenario, MI occurred 13.8% of the time in this situation.

• - F5’s father referred to play – ‘We use singing for routines but also, especially, when we’re playing’, consistent with 16.7% percentage of MI during play.

However, some discrepancies appeared:

• - F4’s mother identified nappy changing as a common situation where shared singing occurred; however, in the recordings, this scenario accounted for only 3.4% of the MI.

• - F5’s father emphasised play, yet emotional regulation was more common in recordings (24.1%).

• - F3’s parents described the presence of MI in emotional regulation scenarios – ‘Certainly, when they needed to be calmed down’ (father, F3); ‘When we spend time together […] at mealtimes, at bath time’ (mother, F3) – yet play (30.3%) and changing clothes (18.2%) were the most frequent contexts.

These contrasts suggest that while parents can recall salient musical episodes, recordings reveal a broader and often unnoticed range of situations in which vocal MI naturally arises. This supports the notion of MI as dynamic, culturally shaped, and intuitively enacted practices (Young, Reference YOUNG2018).

From theoretical perspective, everyday musical exchanges – whether during play, mealtime, or routine care – exemplify what Koops (Reference KOOPS2019) calls Parenting musically: using music to structure caregiving tasks. These findings reinforce that MI is not limited to designated “musical moments” but permeates spontaneous, embodied family life. As Addessi (Reference ADDESSI2009) argues, daily routines offer cyclic, predictable frameworks that foster musical communication. Our data extend these insights, revealing scenarios such as ‘bath time’, ‘story time’, ‘waking up’ or ‘tidying’ that parents often overlook, yet emerged as musically rich.

In sum, vocal MI emerges across a broad range of relational and routine contexts, not as isolated events but as embedded forms of early communication. These findings align with research framing MI as innate, flexible, and culturally grounded practice (Malloch &Trevarthen, Reference MALLOCH and TREVARTHEN2009; Pitt & Welch, Reference PITT, WELCH, Barrett and Welch2023) and highlight the importance of everyday situations not just as background settings, but as structuring elements in how MI unfolds within family life.

Probable function

This finding offers relevant information on how music favours different aspects of the parent-child relationship. When the scenario is played, the most probable function would be to accompany the game and/or entertain, if it affects the child’s emotional regulation: soothing; if it is sleeping time: to get the child to sleep and/or calm him or her down. If the scenario concerns changing nappies/clothes, the probable function would be to get the child’s attention or entertain; if it is about waking up the infant, the function would be greeting. Figure 1 shows, in percentages, the number of times that the categorised functions were identified.

Figure 1.

Probable function of MI.

Figure 1 Long description

Panel Family 5: A horizontal bar graph comparing different functions of musical interactions. The horizontal axis represents the percentage from 0 percent to 80 percent. The vertical axis lists functions: Greeting, Showing affection, Sharing music, Accompanying stories, Accompanying play, Sleeping, Soothing, and Entertaining. Bars are colored blue. Notable values include Entertaining at approximately 50 percent, Soothing at around 30 percent, and Accompanying play at about 20 percent. Panel Family 4: A horizontal bar graph comparing different functions of musical interactions. The horizontal axis represents the percentage from 0 percent to 80 percent. The vertical axis lists functions: Greeting, Seeking attention, Showing affection, Soothing, Accompanying play, and Entertaining. Bars are colored blue. Notable values include Entertaining at approximately 70 percent, Accompanying play at around 30 percent, and Soothing at about 15 percent. Panel Family 3: A horizontal bar graph comparing different functions of musical interactions. The horizontal axis represents the percentage from 0 percent to 80 percent. The vertical axis lists functions: Accompanying entertainment activity, Showing affection, Seeking attention, Sharing music, Soothing, Accompanying play, and Entertaining. Bars are colored blue. Notable values include Entertaining at approximately 50 percent, Accompanying play at around 30 percent, and Soothing at about 20 percent. Panel Family 2: A horizontal bar graph comparing different functions of musical interactions. The horizontal axis represents the percentage from 0 percent to 80 percent. The vertical axis lists functions: Picking up toys, Sleeping, Soothing, Accompanying play, and Entertaining. Bars are colored blue. Notable values include Entertaining at approximately 60 percent, Accompanying play at around 30 percent, and Soothing at about 20 percent. Panel Family 1: A horizontal bar graph comparing different functions of musical interactions. The horizontal axis represents the percentage from 0 percent to 80 percent. The vertical axis lists functions: Soothing, Accompanying story, Sharing music, Accompanying play, and Entertaining. Bars are colored blue. Notable values include Entertaining at approximately 70 percent, Accompanying play at around 20 percent, and Soothing at about 10 percent.

Regarding question – Why do you sing to your child? The answers were:

• - ‘When she was younger, it was mostly associated with some routine, though not so much now that she is older. But for example, when she must wash her hands, then it’s a way of telling her we’re going to eat’ (mother, F1).

• - ‘Music helps me draw him into times of relaxation and sleep. For me it is very important because I use it as a tool to get him to relax and fall asleep’ (mother, F2).

• - ‘[…] we used to sing a lot to amuse William, in the shower for example, which was a horrible moment for him’ (mother, F3).

• - ‘I invent songs to make him laugh or get his attention’ (mother, F4).

• - ‘We sing all the time at home, in the shower, when we’re going to sleep or waking up, while we play’ (mother, F5).

Parents’ perceptions show they were aware of the purpose of singing in their homes. However, results also show that parents found more uses for vocal expression than they reported. Quantitative results show that the most frequent function was entertainment, which occurred in all the participating families, representing 48.7% of the total, followed by play accompaniment, 19.5% of the total. Interestingly, F5 was the only case where soothing (27%) surpassed play accompaniment (13%), possibly due to the child’s age (9–15 months), when MI often supports emotional regulation and bonding (e.g., Bainbridge et al., Reference BAINBRIDGE, BERTOLO, YOUNGERS, ATWOOD, YURDUM, SIMSON, LOPEZ, FENG, MARTIN and MEHR2021; Young & Gillen, Reference YOUNG, GILLEN, Gillen and Cameron2010).

These findings suggest that MI serves multiple overlapping functions, not only to manage routines (as explained in the situation category) but also to nurture affective connection and shared attention. Vocal MI is better understood as a relational and communicative practice shaped by mutual or bidirectional responsiveness, where children are not passive recipients but active participants in musical meaning-making and affective exchange (Español et al., Reference ESPAÑOL, SHIFRES, MARTÍNEZ, PÉREZ, Español, Martínez and Rodríguez2022; Pitt & Welch, Reference PITT, WELCH, Barrett and Welch2023).

Repertoire

This section explores the type of repertoire used during vocal MI episodes, distinguishing between existing and improvised forms, based on observable musical traits in the recordings. A repertoire was classified as existing when recognisable songs or melodies could be identified, and as improvised when the vocal material appeared spontaneous, intuitively generated, or created in-the-moment through melodic or rhythmic invention. Figure 2 shows the distribution of these two categories across families.

Figure 2.

Percentages of types of repertoire.

When asking about the repertoire they used, some answers were:

• - ‘The entire traditional Catalan repertoire for children, also all sorts of music that she (Bruna) has picked up and so she asks for it. Especially music with lyrics, music by the groups El mic and Pot petit, by Jorge Drexler, and “Let it be” by the Beatles’ (father, F1).

• - ‘We sing anything: Iron Maiden, Jurassic Park, Pot petit. They like a bit of musical variety, not just children’s music […] and also songs that my mother and my grandparents sang to me’ (mother, F2).

• - ‘I’m really bad at song lyrics, I don’t remember any of them, so I improvise as much as possible, it’s more like playing a game’. (father, F3).

• - ‘I sing him pop songs that I like, and I change the lyrics intuitively […] the music I sing doesn’t necessarily have to be something I learnt but something that all of a sudden becomes spontaneous’ (mother, F4).

• - ‘Basically, the songs I learnt during my musical training, the typical ones from the traditional children’s repertoire in Catalan’ (mother, F5).

Parents reported using a broad repertoire with identifiable songs – including traditional children’s music, popular tunes, and personally meaningful pieces – which falls within what we classified as existing repertoire. Recordings confirmed that this type of material dominated overall (67.1%), particularly in F1, F2, and F5 (79.6%, 84.6%, and 74.6%, respectively). This supports earlier findings that parents transmit cultural aspects of their society through the repertoire they share with their children (e.g., Campbell, Reference CAMPBELL1996; Trehub & Trainor, Reference TREHUB and TRAINOR1998).

However, while improvisation was rarely mentioned by parents, it emerged in all families’ recordings. Notably, in F4, 71% of vocal MI episodes were improvised, in line with the mother’s description of intuitive and spontaneous singing. In contrast, F3 showed a more balanced distribution, although the father emphasised improvisation more than the recordings suggested.

This discrepancy points out how improvised vocalisations – often playful, situational, and relational – may go unrecognised by adults as meaningful musical acts. Yet, they reflect a key dimension of early MI: responsiveness. As Español et al. (Reference ESPAÑOL, SHIFRES, MARTÍNEZ, PÉREZ, Español, Martínez and Rodríguez2022) argue, children are not passive recipients but ‘skilled participants’ in musical exchanges, actively shaping the communicative process. Thus, MI repertoire should not be seen solely as content to transmit, but as a flexible, co-constructed medium shaped by interactional demands.