Music is a universal social phenomenon that has traditionally been experienced in a live context (Nettl and Russell, 1998; Freeman, 2000). The advent of recording technology in the late 19th century heralded a cultural shift in the way that people experienced music, allowing for the convenience of private, in-home consumption (Moreau, 2013). While technology has provided a low-cost, convenient method for music listening, many people continue to attend live concerts, sometimes at great expense in uncomfortable settings (Baxter-Moore and Kitts, 2016; Brown and Knox, 2017). What is it about the experience that motivates listeners to attend live concerts? A survey found that listeners’ strongest musical experiences often took place at live events (Lamont, 2011). Two factors that likely contribute critically to the enjoyment of live concerts are (1) people like the social connexion of experiencing music with other people (Burland and Pitts, 2014; Brown and Knox, 2017) and (2) people like the feeling of being connected to the performers, by being in the same physical space together, with the potential for performers to directly engage the audience (Silverberg et al., 2013; Leante, 2016), and by experiencing a unique live performance as it unfolds over time (Brown and Knox, 2017). Every live performance is idiosyncratic such that events unfold organically and unpredictably, unlike when listening to a recording in which there is no possibility for an audience to directly affect what a performer has already created.

The social effects of experiencing music with other people have been studied to a greater extent than the effects of experiencing a live performance (Freeman, 2000; Egermann et al., 2011; Rennung and Goritz, 2016; Stupacher et al., 2017). Here we examined the effects of live performance while controlling for the social setting. We compared people who listened to a live performance (specifically, a record release party by Canadian rock star Ian Fletcher Thornley’s 2015 solo album Secrets) to people who listened in a group in the same venue without live performers to the album recordings of the same songs from Secrets. Recently, research on audiences of live performances has gained interest (Egermann et al., 2011; Burland and Pitts, 2014; Danielsen and Helseth, 2016; Bradby, 2017; Brown and Knox, 2017), in part because audiences provide an ecologically valid setting for examining group dynamics. Audience experience has been examined with a variety of techniques including real-time subjective responses (McAdams, 2004; Stevens et al., 2009, 2014; Egermann et al., 2013), social networking (Deller, 2011), video analysis (Chan et al., 2013; Silverberg et al., 2013; Stevens et al., 2014; Leante, 2016; Theodorou et al., 2016) and physiological measurement (Fancourt and Williamon, 2016; Bernardi et al., 2017). It is important to understand effects of the concert setting because attendance may increase health: attending a musical performance was found to reduce stress hormones in audience members (Fancourt and Williamon, 2016) and a 10-year longitudinal study suggested that engagement in cultural events, including concerts, may protect against age-related cognitive decline (Fancourt and Steptoe, 2018).

Enjoying music with other listeners may contribute powerfully to the concert experience. Observers of concert audiences judged synchronously moving listeners as experiencing greater rapport and similar psychological states compared to those moving asynchronously (Lakens and Stel, 2011). After adults move in synchrony, even when unaware of their synchronised movements, they remember more about each other, express liking each other more, and show greater levels of trust and cooperation compared to after moving asynchronously (Hove and Risen, 2009; Wiltermuth and Heath, 2009; Valdesolo et al., 2010; Valdesolo and DeSteno, 2011; Launay et al., 2013; Woolhouse et al., 2016). More broadly, periodic movements and physiological rhythms, such as breathing and heart rate, tend to synchronise unconsciously among people in a group (Richardson et al., 2007; van Ulzen et al., 2008; Morris, 2010; Codrons et al., 2014; Miyata et al., 2018).

Entrainment is defined as the ability to synchronise movements with an external auditory stimulus, in this case the timing regularities of music (Phillips-Silver and Keller, 2012). In humans, synchronisation is supported by connections between auditory and motor cortices (Sakai et al., 1999; Janata and Grafton, 2003; Grahn and Brett, 2007; Zatorre et al., 2007; Fujioka et al., 2012) and manifests as oscillatory activity measured in EEG and MEG (Schroeder and Lakatos, 2009; Arnal and Giraud, 2012; Fujioka et al., 2012, 2015; Cravo et al., 2013; Calderone et al., 2014; Cirelli et al., 2014a; Chang et al., 2018a). Interestingly, few non-human species entrain movements to auditory regularities (Merker et al., 2009; Patel et al., 2009; Schachner et al., 2009). The connection between movement synchronisation and social-emotional engagement may have deep evolutionary roots in humans. Infants are not yet able to coordinate their movements to entrain to a musical beat, although they do move faster to music with a faster compared to slower tempo (Zentner and Eerola, 2010). Yet if an infant as young as 14 months is bounced to music synchronously with the movements of another adult, the infant is more likely to help that adult (e.g., to pick up “accidentally” dropped objects needed to complete a task) compared to if an infant is bounced asynchronously with the adult (Cirelli et al., 2014c). Later work revealed that this increased helpfulness extends to friends of the experimenter who bounced with them (Cirelli et al., 2016). In another study, infants who were bounced to music with stuffed animals, choose animals that bounced synchronously with them over animals that bounced asynchronously. These studies indicate that synchronisation of movement with others during music listening is a cue that even infants use in the development of social-emotional bonds and altruistic behaviours (Trainor and Cirelli, 2015; Cirelli et al., 2018).

We examined the effect of live music while controlling for the effects of being with others in an audience. Little research has examined differences between live and recorded performances by manipulating the presence and absence of the performer. Shoda et al. (2016) reported that the heartbeats of audience members at a live performance exhibited greater entrainment with the musical rhythm than those of listeners at a pre-recorded performance. Performer presence was also found to produce greater relaxation in audience members compared to those listening to a recording (Shoda et al., 2016). Contemporary popular performers often play variations of recorded works at live performances (Shoda and Adachi, 2015), suggesting a novelty factor for listeners. Brown and Knox (2017) found that audience members consider this musical novelty as an important motivator for concert attendance. Live concerts also enable audience members to experience an in-person relationship with the performer. Performers can also be influenced by the presence of an audience, and live performances can be acoustically and energetically different than those recorded in the studio (Zajonc, 1965; Yoshie et al., 2016; Bradby, 2017).

We used head movement responses as our main measure of audience experience for several reasons. Moving to the beat during music listening is culturally ubiquitous, with collective movement a hallmark of the contemporary concert experience (Zatorre et al., 2007; Madison et al., 2011; Janata et al., 2012; Davies et al., 2013; Madison and Sioros, 2014; Stupacher et al., 2017). Individuals use a range of movements when listening to music, from foot tapping to head nodding, to whole body movement (Leman and Godøy, 2010). Head movements are particularly relevant as they are a reliable indicator of rhythmic entrainment (Toiviainen et al., 2010; Burger et al., 2013), reveal communication patterns between performers (Chang et al., 2017), reveal directional and emotional communication patterns (Chang et al., unpublished), and even predict who will “match” during speed dating (Chang et al., 2018b). Movement of the head alone—but not legs alone—affects how ambiguous auditorily-presented rhythms are interpreted (Phillips-Silver and Trainor, 2008). This interaction between head movement and auditory perception likely involves the vestibular system located in the inner ear which processes proprioceptive information about head movements (Trainor et al., 2009). Head movements also encode emotional information (Livingstone and Palmer, 2016; Chang et al., unpublished), and may function as a form of non-verbal communication in a noisy environment (Harrigan et al., 2008). Head movements provide information about the nature of an emotion being communicated (Ekman and Friesen, 1967; Witkower and Tracy, 2018). Furthermore, movement smoothness (which increases with movement speed) is greater when communicating joy than a neutral emotion or sadness (Kang and Gross, 2016). Horizontal head movements and forward velocity communicate happiness even without the context provided by facial expression or vocal content (Livingstone and Palmer, 2016). Additionally, movement vigour (average speed) and movement distance have been shown to convey the intensity of emotions (Atkinson et al., 2004). Leow et al. (2015) found that, even when asked to walk at the same tempo, participants walked more vigorously (faster) to more familiar music. One study found that during music listening, greater head speed was correlated with increased spectral flux in low frequencies (associated with greater presence of kick drum and bass guitar) and in high frequencies (associated with hi-hat and cymbals or liveliness of a rhythm), as well as with greater percussiveness, but head speed was not found to be related to tempo (Burger et al., 2013).

In summary, there are many possible factors contributing to movement during music listening including biological imperatives, emotions, and the presence of others. These factors have been studied in highly controlled laboratory settings but have yet to be explored in real-world music listening contexts. In the present study, we were interested in how a live concert affected audience head movements as an index of engagement, specifically, by comparing the movements of concertgoers who experienced a live performance versus a recorded version of the same songs. We were particularly interested in the measure of vigour. Following previous researchers, we operationally defined movement vigour as the average speed of movement over a time interval, regardless of direction (specifically, head distance travelled within a song divided by the total length of the song, giving a value in millimetres per second) (Atkinson et al., 2004; Mazzoni et al., 2007; Zentner and Eerola, 2010). We were also interested in how head movements might be influenced by audience members’ prior admiration for the performers (i.e., their Listener-preference). People are motivated to attend music concerts when they hold a strong preference for the musicians’ work. Musical preferences for genres and artists also play a role in defining social affiliations, particularly during adolescence, where they appear to function as a ‘badge of identity’ within a social group (North and Hargreaves, 1999; Mulder et al., 2010). ‘Fans’ of a particular performer would be expected to enjoy musical performances by that performer, in part because the familiarity gained from repeated exposure to recordings of their music would be expected to increase enjoyment of the performer’s music in general (Schellenberg et al., 2008; van den Bosch et al., 2013). To examine the effect of audience members’ prior preferences for the band, we recruited fans of the performer Ian Fletcher Thornley, along with naïve listeners who expressed no particular preference for the performer. Since the album had not yet been released prior to the concerts, the effects of song familiarity were controlled while examining differences between fans and neutral listeners as neither group had heard the songs prior to the concerts.

In sum, we examined the effects of live versus pre-recorded music and fan status on audience engagement with the music through head movements. Self-reported Fans and Neutral-preference listeners were separately recruited, and randomly assigned to attend one of two concerts. The concerts served as the record release event for Canadian rock star Ian Fletcher Thornley’s 2015 solo album Secrets, featuring new unreleased music. In the Live concert, audience members experienced a live performance by the musicians, while in the Album-playback concert, listeners heard an audio recording of the same songs from the Secrets album. Both concerts were held in the LIVELab, a 106-seat performance hall equipped with a 25-camera optical motion capture system. Head movements of participants were recorded simultaneously throughout each of the two concerts (Supplementary Figure S1). Two aspects of head movement were examined: (1) vigour and (2) entrainment to the beat of the music. We hypothesised that head movements would be faster and better entrained when audiences experienced a live concert compared to a pre-recorded version of the music. We further hypothesised that fans of the performer would exhibit faster movement, and entrain better to the rhythm, compared to neutral listeners.

The question of why people enjoy attending live concerts, when the same music can be experienced more easily and for less money at home, likely involves two aspects: the social sharing of the experience in a group of people; and “live” aspects, including connecting with the artists and experiencing the potential for spontaneity and unpredictability of live music as it unfolds over time, compared to a pre-recorded and unchanging version on a recording that a fan might become familiar with after repeated listening. In our study, we examined primarily the second aspect, comparing listening to a recording of a set of songs from Ian Fletcher Thornley’s 2015 album Secrets to listening to a live performance of those songs, while keeping the social aspect largely the same: both the Live and Album-playback concerts were experienced in the context of an audience in the same LIVELab venue. In the case of this study, audiences were not familiar with recorded songs, but nonetheless may have reacted to the knowledge that the music in the Live condition was unfolding in a unique way that would never be repeated exactly. Necessarily, the visual stimulation differed between the two conditions because of the presence of the live performers. We feel that this is not necessarily a confound—a live performance requires the presence of performers—but future studies might incorporate some visual stimulation that tries to better equate the two conditions, for example, by showing a video of the live performance. We also examined how being a fan of the musical group affected these experiences by comparing self-reported Fans and Neutral-listeners randomly assigned to the Live and Album-playback concert conditions. We focused on head movements, using motion capture to extract the vigour and degree of entrainment of head movements to the beat of music (Toiviainen et al., 2010; Burger et al., 2013).

We found that for both Fans and Neutral-listeners, head movements were more vigorous in the Live than the Album-playback concert, but Concert-status did not affect degree of entrainment to the beat. On the other hand, across both concert conditions, Fans moved their heads more vigorously and with better entrainment to the beat compared to Neutral-listeners. The greater degree of entrainment to the beat in general in Fans likely reflects their greater familiarity with the artist’s musical style. The greater vigour of head movements across groups at the Live compared to Album-playback concert likely represents greater arousal, increased anticipation, and increased connection with the artists and their music during the live concert (Mazzoni et al., 2007; Leow et al., 2014). Amount of musical training varied across audience members, but there were no differences between musicians and non-musicians in either movement vigour or synchronisation to the beat. Similarly, Bernardi et al. (2017) reported that musical training did not affect the degree of synchronisation of autonomic responses to the beat of music experienced in a group setting. Together, these results suggest that entrainment responses in audiences are independent of musical training.

We controlled for song familiarity across Fans and Neutral-listeners by using songs that had not yet been publicly released (the first 7 songs of the concerts). The eighth song, “Blown Wide Open,” on the other hand, was certainly familiar to Fans, and may have been familiar to some Neutral-listeners as its original rendition had achieved double platinum sales in Canada in the late 1990s. Interestingly, when the songs were not familiar, there was no difference in degree of entrainment to the music across the Live and Album-playback concerts. However, for the eighth song that was familiar at least to Fans, head movement entrainment was greater during the Live than Album-playback concert. This suggests that while the vigour of head movements is affected by whether the music is live or pre-recorded regardless of familiarity, familiarity with the music may foster greater entrainment to the beat during live compared to recorded contexts.

Vigour of head movements and degree of entrainment differed across songs. Further, there were interactions for both measures between Songs and Listener-preference, indicating that Fans and Neutral-listeners reacted differently to different songs. This suggests that some songs might excite existing fans differently than naïve listeners, which might inform record company promotion decisions. Concerts are becoming increasingly important for the music industry as the prevalence of piracy results in reduced revenue from album recordings (Frith, 2007; Papies and van Heerde, 2017). Interestingly, the majority of audience members report that cost does not influence their decisions to attend concerts (Brown and Knox, 2017). In general, research on audience development and retention could be important for sustaining the multi-billion dollar music industry (O’Reilly et al., 2014; Papies and van Heerde, 2017).

Music compels us to move, the likely result of connections between auditory and motor areas of the brain (Sakai et al., 1999; Janata and Grafton, 2003; Grahn and Brett, 2007; Zatorre et al., 2007; Grahn and Rowe, 2009; Janata et al., 2012), whose communication during rhythm and beat prediction can be measured in neural oscillations (Fujioka et al., 2012). Certain characteristics of music lead to increased entrainment to music and compulsion of movement, such as beat predictability and rhythmic complexity (Fitch, 2016), the density of events between beats (Madison et al., 2011), moderate levels of syncopation (Witek et al., 2014; Fitch, 2016), and possibly micro-timing deviations (cf. Madison et al., 2011; Davies et al., 2013; Stupacher et al., 2013; Kilchenmann and Senn, 2015). The present study demonstrates that in addition to acoustic characteristics of music, environmental and personal factors influence movement to music as well. Specifically, familiarity with the performer and musical style (Listener-preference) led to increased movement and entrainment, while the live performance (Concert-status) led to a significant increase in movement vigour. Because synchronous movement can lead to prosociality (Hove and Risen, 2009; Wiltermuth and Heath, 2009; Valdesolo et al., 2010; Valdesolo and DeSteno, 2011; Launay et al., 2013; Cirelli et al., 2014b; Trainor and Cirelli, 2015; Rennung and Goritz, 2016; Woolhouse et al., 2016), and because entrainment to music was fostered more by Listener-preference than Concert-status, it is possible that personal factors are more important than environmental factors for generating synchronous movement and subsequent prosociality.

This study adds to the fledgling literature examining music listening in concert settings (Egermann et al., 2011; Shoda and Adachi, 2012, 2015, 2016; Fancourt and Williamon, 2016; Shoda et al., 2016). It provides unique insight into how live music is experienced in ecologically valid conditions, and how that experience is expressed through body movement. Many questions that remain could be addressed in future research in the LIVELab, such as how individual differences in personality affect live concert experiences, how individuals in a concert setting are affected by the movements of those around them, the effects of different musical characteristics (e.g., tempo, instrumentation, presence of improvisation, genre), whether synchronous movements in a concert setting leads to increased prosociality and bonding, and how performers are affected by audiences.

The raw data supporting the conclusions of this manuscript will be made available on https://zenodo.org/ (search for ‘LIVELab) by the authors, without undue reservation, to any qualified researcher.

This study was carried out in accordance with the recommendations of the Canadian Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans (T), with written informed consent from all subjects. All subjects gave written informed consent in accordance with the Declaration of Helsinki. The protocol was approved by the McMaster University Research Ethics Board.

DS involved in data collection and analyses, and the preparation and review of the manuscript. DB involved in the research design, data collection and analyses, and the preparation and review of the manuscript. SL involved in motion data collection, statistical analyses, and review of the manuscript. JB involved in project and research design, recruitment, organisation, and data collection, and review of the manuscript. MW involved in the conception and organisation of the project including artist-Anthem coordination, research design, and review of the manuscript. SM-R involved in recruitment, data collection and review of the manuscript. LT involved in the conception and organisation of the project, research design, review of the statistical analyses, and preparation and review of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.