MNDLB5 Whitepaper #1: Music and Human Performance: A Unified Approach

Daniel Casanno, Professor of Politics at Fairleigh Dickinson University and Research Lead at MNDLB5

Neel Lee Chauhan, Non-Executive Director at MNDLB5 and Governor at Toronto Metropolitan University

Download PDF version here: https://drive.google.com/file/d/1Q6SB_pDvf96FWwET1h0PtDTMXKav9edp/view?usp=sharing

Research Summary:

This white paper explores how music influences human performance across cognitive, athletic, and social domains by strategically inhibiting distractions. Depending on task complexity, music can enhance focus, reduce fatigue perception, and improve motor coordination. However, its effects vary based on musical characteristics and individual differences, highlighting the importance of personalization in optimizing performance.

The MIND (Musical Inhibitions of Negative Distraction) effect describes how music can take up cognitive processing space, reducing awareness of internal fatigue signals. This occurs because the nervous system has limited bandwidth for processing information, and music can partially block discomfort signals, enhancing human performance.

Key Findings:

1. Cognitive Performance: Music can improve concentration in low-engagement tasks by blocking internal distractions but can hinder more complex cognitive tasks, particularly when it includes lyrics or demands high cognitive processing.

2. Physical Performance: Music enhances endurance and reduces fatigue perception in moderate-intensity exercise by occupying cognitive bandwidth and synchronizing with movement. However, its effects diminish in high-intensity exertion, where physiological signals become too strong to override.

3. Individual Differences: The impact of music depends on personal preferences and emotional connections. Music that resonates with the listener is more effective in enhancing both cognitive and physical performance, explaining inconsistencies in prior research and underscoring the need for customized interventions. Music designed to maximize performance must therefore be both emotionally resonant, and have characteristics appropriate to the activity at hand: a combination that may be impossible without some degree of personalization.

   
   

The limited capacity of our cognitive systems (as in Kahneman 1973) means that we can only attend to a small portion of the torrent of sensory stimuli bombarding us at any time. This constantly roving attention was likely critical to the survival of our distant ancestors, who needed to maintain constant vigilance in the face of predators, but is less helpful in the modern world, where maintaining focus on a task – be it cognitive or physical – is often desirable. While it might seem that exposing an individual to increased stimuli would necessarily lead to distraction, and therefore to decreased task performance, the limited capacity of our cognitive systems means that the strategic use of stimuli like music can actually optimize performance, by inhibiting undesirable stimuli, and improving both performance and the subjective experience of the task.

Depending on the task at hand, the degree to which music can crowd out other stimuli, and the nature of the stimuli that it is crowding out are likely to vary. For instance, past research has shown that performance on an attention-based task is enhanced by music that requires limited cognitive processing, while performance in athletic tasks is best enhanced by music that requires substantial cognitive processing.

Importantly, while the inherent objective qualities of the music (such a tempo and lyrical complexity) being used matter, so too do the subjective qualities of the music, especially the emotional attachments that the individual has to it. The capacity of a particular track to inhibit undesirable stimuli also depends on the interaction between these features and the individual: what works well for one person might not work nearly as well for another. As such, any system that attempts to use strategic musical inhibition of stimuli to enhance performance must be mass customized  in a way that is not possible with existing technologies and systems.

Strategic inhibition of stimuli

The use of music to strategically inhibit stimuli is grounded in two concepts that have been well established in cognitive psychology: limited capacity and automatic processing of sensory stimuli. Modern research on the limited capacity of cognitive systems goes back to Kahneman’s (1973) work on perception (though work as early as James in 1890 talk about the limited capacity for controlled, as opposed to automatic, processing of information). Essentially, he argues that there is a substantial mismatch between the amount of sensory information received by the brain and the brain’s processing capacity such that we can only attend to a small portion of the incoming information. Generally, our brains are well adapted to attend to the portions of the world around us that are most important to our survival: if you are walking down a trail, and see a snake, you’re likely to immediately attend to it, regardless of whatever else is on your mind (as in Isbell 2006). But limited capacity means that attending to one set of stimuli means not attending to others. In the example of the cocktail party effect (Cherry 1953), for instance, individuals have been shown to have the capacity to selectively attend to one stream of speech, while ignoring others. All of the conversations are being received by the sensory centers of the brain, but we are very capable of picking out one conversation to follow, selectively ignoring the others.

The metaphor of bandwidth has long been used to describe the limited capacity of the nervous system to process and transmit information (Karageorghis and Priest 2012). As such, exposure to music can interfere with the processing of other signals by taking up processing space, for instance, drowning out the experience of fatigue during light and moderate intensity workouts. In this case, internal stimuli – those being generated from within the body – are sending signals about fatigue or soreness to the brain. But if musical stimuli are occupying a sufficient portion of the brain’s processing capacity, those internal stimuli will be partially blocked, meaning that the individual will simply be unaware of the full extent of their fatigue. We refer to this as Musical Inhibitions of Negative Distraction (MIND) or the performance enablement effect of musical stimuli.

In addition to limited cognitive capacity, the use of music to strategically inhibit stimuli is based on the concept of automatic processing of sensory information. Decades of work in cognitive psychology, with modern work linked back to Schneider and Shiffrin (1977), argue for two distinct ways in which the brain processes information: slow, serial, conscious processing, which relies on the small amount of information that can be held in working memory (and so is often aided by technological enhancements, like written notes), and fast, parallel automatic processing. These types of processing are largely handled by different regions of the brain (as in Schneider and Chein 2003). When stimuli call upon the individual to suppress automatic processes (as in the Stroop test, 1935, which asks individuals to differentiate between the text of words and the colors they’re printed in), they can do so only with great cognitive effort. 

For our purposes, what’s important about these separate processes is the “automatic” nature of the second information processing system. Music has been shown to lead to bi-lateral activation of brain areas, throughout the cortex, neocortex, and paleo-cerebellum (Peretz and Zatorre 2003; Tramo 2001; Platel et al 1997). Importantly, some elements of music – especially rhythms – seem to be processed by deeper areas of the brain that are generally outside of conscious awareness. In fMRI studies, rhythmic music has been shown to activate areas related to motor control, movement and balance (Levitin and Tirovolas 2009). Our brains respond to music, processing it automatically, even when we’re not trying to attend to it. As a result, music is necessarily competing with other stimuli for the limited cognitive capacity available to the individual, and the fact that it is working on deeper, more evolutionarily ancient areas of the brain mean that it may even have an advantage over other stimuli (much like the snake distracting a hiker).

Put together, this means that music has the capacity to crowd out other stimuli that might get in the way of performance across various tasks. In studies of areas as diverse as cognitive, attention-based tasks, athletics and social dancing, researchers have documented  the degree to which music can enhance performance. The argument being made here is that all of these effects can be understood as arising from the same bases, and much of the inconsistency in the reported effects of music on performance across these task areas can be attributed to a combination of individual level differences, differences in the music being used, and the interaction of the two.

Often, the ways in which music crowds out other stimuli leads to worse performance on a task, as in the common example of driving. In a simple driving task, music leads to better task performance; when the driving task became more complicated, music led to worse performance (Wang et al 2015). Individuals seem to understand this, at least unconsciously (another recurring theme in this area of research), hence the propensity of individuals to turn the radio off when they need to concentrate on the road. Whether music tends to enhance or decrease performance on a task depends on a combination of the task, the music, and the individual. Understanding how music can crowd out other stimuli, and the conditions under which this enhances performance, means that we can build musical stimuli that will reliably do so across task areas.

Music and Cognitive/Attention-Based Tasks

The most well-developed literature around the effects of music on task performance comes from studies looking at cognitive tasks, especially those which require attention and focus. In the laboratory, studies in this area often rely on vigilance tasks, in which (typically) participants are asked to look at a screen, and report when something happens. These tasks are often easy and relatively dull: the challenge is maintaining concentration for long enough to complete the task appropriately. Such tasks don’t have perfect ecological validity – outside of the TSA, few people are tasked with watching something and trying to stay alert while looking for a relatively rare outcome – but they’re generally used as a simplified version of a task that many people find themselves facing: looking at a screen, and trying to maintain concentration on the task in front of them. Studies that have made use of more externally valid tasks – for instance, having participants work on a project of their own (as in Haruvi et al 2021), find similar results to lab-based studies, but there are some inconsistencies, perhaps due to task complexity. These sorts of issues with external validity often extend to population validity, with much of the work on the effects of music in this area using samples of psychiatric patients (such as Courtright et al 1990; see Garrido, Davidson and Odell-Miller 2013).

In studies of the effects of music on attention and concentration to these sorts of tasks, researchers have generally relied on an arousal-based model, largely based on the Yerkes Dodson (1908) Law.  The Yerkes Dodson Law holds that performance on attention-based tasks is optimized when overall arousal levels are moderate: if arousal levels are too low, people become bored; if they’re too high, distracted. Music can help to increase the level of arousal, and therefore get people closer to the optimal levels of arousal, and therefore, performance – but only when the task itself has relatively low arousal (that is, doesn’t require a great deal of cognitive effort). When the task being attended to requires a great deal of attention already, any additional arousal caused by music would be likely to reduce task performance.

Exhibit 1: Task Focus Model

In the work on this area, researchers describe the resulting horseshoe shaped model of performance, in which both low and high levels of arousal lead to decreased performance, and performance is optimized at moderate levels. This work also leads to the important distinction between two types of distractions: internal and external (Unsworth and Robison 2016; Lenartowicz, Simpson and Cohen 2013). Internal distractions arise from the individual: thinking about other tasks that have to be completed (checking a social media feed, or thinking about the shopping), or attending to bodily sensations (hunger, itches and such). External distractions arise from stimuli that the individual is not trying to attend to: a baby crying, or a phone ringing. Either one of these distractions reduces performance on cognitive tasks, but the horseshoe model of performance is misleading in that these distractions have very different remedies. In one case, performance is enhanced by increasing stimuli; in the other, by reducing it. 

On an anatomical basis, the part of the brain that regulates attention is thought to be located in the posterior parietal cortex (Schneider and Chein 2003; Baizer, Desimone and Ungerleider 1993), as evidenced by studies showing that damage to this area impedes the ability of individuals to identify and attend to important stimuli (Mesulam 2000). Not surprisingly, problems in this area have been linked to disorders in attentive capacity, such as ADHD (Salehinejad et al 2020).

On the level of brain wave activity, music tempo impacts beta wave activity (Hurless et al 2013), and beta waves (13-30 mHz) correspond with conscious processing, alertness and attention to a task (Lim, Yeo and Yoon 2019; Ismail et al 2016).

Music seems to have the most beneficial effect in tasks that are relatively dull and are compromised when the mind wanders (as in Davies, Lang and Shackleton 1973). The cognitive story of what’s going on here is relatively simple. While the brain has limited cognitive capacity, that capacity is still substantial, and a dull task simply doesn’t use all of the “bandwidth” available. That absence is quickly filled by other processes that would typically burble along below conscious awareness (internal distractions), which compete with the desired task for attention, reducing performance, or interrupting a flow state of concentration (a phenomenon studied since Csikszentmihalyi’s work in the 1970s; see Marty-Dugas and Smilek 2019 for a review).

The automatic processing of music means that music crowds out internal distractions, preventing them from entering conscious awareness. This does not mean that these considerations are not present in the individual, but rather that limited cognitive capacity means that they don’t rise to the level of conscious awareness, and disturb concentration. This helps to explain the dramatic differences in research on the effects of background music versus non-melodic background noise. Since there is no reason for noise – especially artificially generated white noise – to receive privileged processing in the brain, it simply adds an external distraction on top of any internal distractions that may be present. Studies on the effects of such noise show almost universally that it reduces performance in cognitive tasks across a variety of areas (Banbury and Berry 2005; Evans and Maxwell 1997; Boman 2004), including memory (Boman, Enmarker and Hygge 2005) and cognitive task performance (Furnham and Strbac 2002; Smith et al 2003).

Exhibit 2: Conscious Awareness Model

In contrast, when tasks are relatively engaging (like reading comprehension, or remembering something that’s been said), music tends to interfere with the process (Kiss and Linnell 2024). Automatic processing of music means that the brain will process musical input – perhaps even favoring it – even when the cognitive capacity needed to do so takes away from the cognitive capacity necessary for the task being attended to. In such a case, the music becomes an external distraction, going beyond crowding out internal distractions and begins to crowd out the attended task, reducing performance.

This two-sided effect of music – increasing performance in some instances, but not others – is evident in the existing research, which shows strongly mixed effects of music and similar stimuli. So far, there are two moving parts in this process: the complexity of the task and the cognitive capacity needed to process the music. But there is also a third: the strength of the internal distractions. As these distractions become more urgent – remembering that you have to do something important, or getting very hungry, or thirsty – the music can no longer crowd them out: they break through into conscious awareness.

Characteristics of Music and Performance

Assuming we can hold constant the strength of the internal distractions, the extent to which music enhances performance in cognitive tasks therefore depends on the balance between the amount of processing required to carry out the task being attended to, and the amount required to process the music. Performance on a simple task might be enhanced most by music that requires relatively effortful cognitive processing, while more complex tasks might benefit from music that requires less processing. At some level of complexity, any processing of music is likely to detract from task performance. 

There has been a great deal of research looking at the relationship between characteristics of music and their role in enhancing concentration. While music has been characterized by eight separate dimensions (Levitin 1999: contour, loudness, meter, pitch, rhythm, tempo, timbre and spatial location), only a few of these have been the subject of study for their impact on attention. Even this sort of dimensionalization is a simplification, as individual elements of a song – like mood – are often modelled as multi-dimensional (i.e. Bhat et al 2014).

For instance, pop and hip hop tend to lead to less concentration relative to other genres, likely because of the higher lyrical complexity and unexpected changes in inherent characteristics like tempo commonly found in these genres (Haruvi et al 2021). Park, Kwak and Han (2020 had a relatively cognitively taxing task (filling out a questionnaire) and found that classical/jazz music helped participants maintain attention better than the other kinds of music. Music without lyrics such as house music, progressive house, techno, dubstep and are likely to require less processing capacity than music with lyrics, just as music with rapid changes in tempo or chord structure are likely to require more processing. 

Teasing out the effects of language within music from other characteristics is tricky, but researchers have figured out ways to isolate it as a variable. In a 2022 study, Brouwer et al tested participants in a speech recognition task, in which they had to type sentences that they heard while listening to three different versions of two Katy Perry songs. One version was the original in English (a language understood by the participants, though the sentence in the task was in the participant’s native language, Dutch); the second was a karaoke version without any lyrics, and the third was a version with nonsense lyrics, originally produced for the video game The Sims (so, instead of “Last Friday Night,” participants heard “Lass Frooby Noo”). As might be expected in a speech recognition task, participants did better in the karaoke version, where there were no lyrics to interfere with speech recognition. But they also did better when listening to nonsense lyrics than when they were exposed to English lyrics, apparently because processing sensical words required increased cognitive processing, even when they were attending to a different task. In this case, the difficulty of the attended to task remained constant – but the processing required by the music ramped up to an extent that it reduced performance on the main task. 

Similar results pertain to the effect of lyrics and lyrical complexity on concentration and attention. Shih, Huang and Chiang (2012) find that lyrics play a substantial role in whether music tends to increase or decrease levels of attention: when lyrics are present, attention to task among works drops relative to a baseline of no music (or similar music without lyrics). While this may seem odd, it makes sense within the model put forward: songs with lyrics requiring more processing, and the intervention of different areas of the brain than the processing of music without lyrics (Furnham and Allass 1999).

To complicate matters, individuals seem to differ in how much cognitive effort they use when processing music. Work on music students  (Ohnishi et al 2001; Zatorre et al 1998) suggests that individuals with musical training process classical music more deeply – engaging language centers of the brain, for instance – than individuals without such training. 

But these responses to inherent characteristics of music don’t tell the entire story, as they interact with the traits of the individual (Kiss and Linnell 2024). Characteristics of music can be broadly categorized as inherent (referred to as “intrinsic” in Sloboda and Juslin 2001, and “internal” in Karageorghis, Terry, and Lane 1999) and subjective (“extrinsic” in Sloboda and Juslin 2001; “external” in Karageorghis, Terry, and Lane 1999; other terms have been used for both of these categories as well). Inherent characteristics are based on the features of the music, such as tempo, melody and complexity; essentially the eight characteristics references above. These are experienced in approximately the same way by all listeners, can be easily compared across tracks, are what recommendation engines use to categorize music. Subjective characteristics arise from the interaction of the track and the listener. Bishop (2010) points to the popularity of the theme from the movie Rocky as an example of a track having emotional resonance, independent of any inherent qualities of the track, an experience familiar to anyone who has had “a song” shared with a spouse or loved one. The duality of these response categories is mirrored by the parts of the brain activated by exposure to music, with the limbic system responding to inherent characteristics like tempo, and more evolutionarily recent parts of the brain, like the neocortex, being responsible for subjective responses (which require integration with memory and emotional centers). 

So, music that has emotional resonance for the individual, or lyrics that are meaningful to them, are likely to lead to require greater cognitive processing than hypothetical music with identical inherent traits but no such connections. In the existing research, the importance of these emotional connections to music are evident in the well-established effects that arise from having participants in research studies choose their own music, rather than having music randomly assigned to them. Responses to music are shaped not just by the nature of the task, but also by both the inherent qualities of the music and the subjective qualities of how it is perceived by the listener (as well as the interaction between the two). This interaction helps to explain the commonly reported effects of music chosen by the participants in studies as having a very different effect than music assigned to them (as in Cassidy and Macdonald 2009; Huang and Shih 2011), even when the inherent qualities of the music do not systematically vary (see also Park, Kwak and Han 2020). Lynar et al (2017) find that the subjective qualities of the music are much more important than inherent qualities in helping participants maintain concentration. 

The fact that participants in these studies are being assigned to music, rather than selecting it, has led  to lots of concerns about the ecological validity of results in this area (Kiss and Linnell 2024). In the real world, people aren’t randomly assigned to music that they wouldn’t listen to normally (much less to white noise), nor are they made to carry out dull tasks with an experimenter looking over their shoulder.

Effects of the preferences of listeners may help to explain the significant disparities in findings on the topic (Huang and Shi 2011). As de la Mora Velasco and Hirumi (2020) report in their review of the literature on music and learning, identified effects have been wildly inconsistent across studies, even ones that use the same music as a stimulus. Meta-analyses have generally found that there is no effect of music on concentration, but that’s largely driven by positive and negative effects cancelling each other out (Kämpfe et al 2011). Focusing on the amount of cognitive effort needed to process the music – which involves both the inherent and the subjective qualities of the music, including emotional resonance that may attach to some tracks – helps to explain this disparity. Even when researchers are controlling for what music is being played, they cannot control for differences in how the subjective qualities of the music that lead to differential processing loads on individuals.

Some of the most intriguing evidence for how internal states of the individual impact the processing of music come from Darrow et al (2004). Their study looked at attention and performance in cognitive tasks among students, divided between music majors and non-majors. They found that instrumental music had dramatically different effects on music majors than it did for non-majors, increasing both their focus and performance on a cognitive task. The idea seems to be that students who were studying music were processing the instrumental music more deeply, so it was more effective at blocking distractions than the same music when listened to by students without musical training. 

It might be thought that calming music would have the capacity to reduce arousal, and therefore improve performance in relatively difficult concentration-based activities. If this were the case, it would provide a potent counter example to the causal model presented here, as music would require processing, but would somehow reduce the overall level of arousal. But while there is some argument that that music can be used to reduce arousal in exercise activities (as in Kiss and Linnell 2023), especially in post-exercise recovery, there does not seem to be any research showing that calming music can reduce arousal in a concentration context (at least relative to silence). Calming music might reduce anxiety or other negative states that can interfere with cognitive tasks – but they do so by crowding out internal distractions, not by reducing overall arousal levels.

Research has also pointed to the importance of volume of music on concentration tasks. Turner et al (1996) found that music played at a low volume worsened cognitive performance, as did music played at a high volume. Such findings fit in well with the proposed model: music played at a high volume would be more distracting, and therefore take away cognitive resources needed for the task being attended to. Music played at a very low volume requires additional processing for the listener to fully understand it, and would have much the same effect (though through different means).

Closely related to cognitive and attentional tasks are creative tasks. While promoting creativity would be a worthwhile endeavor, there is little research on how music impacts it (see Xiao et al 2023 for an exception). Part of this is a measurement issue: the phrase “standardized measure of creativity” is almost an oxymoron, and the construct validity of commonly used measures like the Alternative Uses Task (AUT) isn’t clear. Another issue is that the cognitive roots of creativity aren’t entirely clear. Is it about connections coming up, like distractions, when the mind is occupied elsewhere? If so, music might crowd out creativity. Is it about novel connections being made between concepts being attended to? If, so music might help creativity, unless creativity is such a cognitively intense exercise that music is taking up required processing capacity. Without answers to these basic questions about creativity (and, perhaps, better ways to measure it), the likely effects of music on creativity even with the model proposed here are unclear.

Objective versus Subjective Measures of Attention

In addition to measuring attention to a task, and task productivity, researchers have also made use of subjective measures of attention. In some cases, this is necessitated by diversity in the sorts of tasks undertaken by individuals. In order to maximize the ecological validity of a study, participants should be left to do the sorts of tasks that they would carry out normally (rather than visual vigilance tasks, or shortened games of Tetris (as Haruvi et al 2021). However, this raises a measurement problem: if everyone is doing different tasks, or they’re carrying out concentration-based tasks that don’t have clear “scores” associated with them (like writing a paper, or coding a program), objective measures of productivity or attention may not be available.

Subjective measures may also be useful for concepts that are difficult to measure externally, like “flow state,” focus, or generalized productivity. While some studies make use of EEGs and similar devices to measure physical correlates of such concepts (as in Lim, Yeo and Yoon 2019; Haruvi et al 2021), other measures may provide valid results with less respondent burden.

In such cases, researchers have often turned to subjective measures of attention, focus and productivity. In these measures, rather than directly looking at how well an individual performed on a task, researchers look at self-reports of attention, concentration or productivity (as in Homan, Drody and Smilek 2023; Kiss and Linnell 2021; Rowell and Flick 2019). Interestingly, studies that have looked at the effects of background music on cognitive tasks have generally found stronger effects than those that look at objective measures (see, for instance, Rowell and Flick 2019). The effect here seems to be that music makes tasks more enjoyable – especially when participants are allowed to choose their own music – even when it does not improve performance, and even when the task is difficult enough that the music is likely to be making the task more difficult (as in Homan, Drody and Smilek 2023) by crowding out processing capacity needed for the task being attended to (similar effects have been found for the effects of music on exercise). 

At first glance, this may seem like a combination of a measurement problem (something like a placebo effect – effects being attributed to music that aren’t there) and a non-finding (if the music isn’t helping attention or productivity, why should we care?). But the subjective experience of the task is inherently important. Even when listening to the right kind of music doesn’t make individuals more productive or enable greater concentration, increasing subjective measures of enjoyment or productivity reflect greater satisfaction and happiness with the task, making the individual more likely to continue on with it. Task perseverance, too, is an important outcome, and if music can increase that, it’s valuable regardless of other effects.

Music and Exercise/Athletic Tasks

In studies of the effects of music on exercise, researchers commonly refer to Rejeski’s (1985) parallel processing hypothesis. This holds that since processing space – often conceptualized like the RAM of a computer chip – is limited, music can increase athletic performance for light to moderate levels of exercise by essentially drowning out physiological symptoms of fatigue, improving performance. As the exercise becomes more intense, these physiological signals become stronger, more urgent, overwhelming any the effect of the music. While the existing literature does not explicitly make this connection, this is very much an instantiation of the same effects seen in the earlier discussion of the effects of music on cognitive activities. The key difference between the types of tasks is the role that music would play in enhancing performance. In cognitive tasks, music is expected to enhance performance by leveraging the preferred processing of music to crowd out unwanted stimuli that might otherwise interfere with the task being attended to. In exercise-based tasks, music can be used to enhance performance by distracting from signals that the individual does not want to attend to. 

This means that there is a substantial difference in the qualities of the music that are expected to enhance performance between the types of tasks. In cognitive tasks, performance is enhanced when the amount of cognitive processing required by the music is enough to crowd out internal distractions, but not so much as to take capacity away from the task being attended to. In exercise tasks, performance is enhanced when the music requires so much processing that it crowds out other signals. In both cases, the goal is to interfere with internal distractions, but the qualities required to carry out this task differ widely.

The parallel processing model (Rejeski 1985) holds that music increases performance in exercise tasks by distracting the individual from signs of fatigue, essentially distracting them into being able to push themselves harder (Terry et al 2020; Hutchinson et al 2018), though the effect is be expected to be stronger at lower levels of fatigue.

There is little disagreement that music can enhance athletic performance, though the research carries a number of caveats. Listening to music during repetitive exercises increases efficiency and energy output (as in Yamashita et al 2006), but it doesn’t seem nearly as effective during high intensity exercises (Karageorghis and Priest 2012). The effects of the music appear to be strongly mediated by the participant’s enjoyment of the music: liking a track almost seems like a prerequisite for it having an effect (as in Nakamura et al 2010).

Such caveats and seeming inconsistency of findings become easier to understand within the model described previously. When exercising, the body sends signals about fatigue, pain or other internal states that may lead the individual to cease exercising, reduce their pace, or otherwise lead to suboptimal performance. The role of music, then, is to crowd out these signals, much as it crowds out internal distractions in order to enhance performance on cognitive tasks. However, music is not expected to be as effective in doing so during exercise tasks as it is during cognitive tasks, as the internal distractions are processed in many of the same deep parts of the brain as music is. Music was able to effectively interfere with internal distractions to cognitive activities because processing of music is privileged over the signals it was competing with, but it is not privileged over the physiological signals that are the equivalent internal distractions during exercise.

Studies have shown that during low intensity exercise, music significantly cuts perceived exertion during the task, and seem to cut biological markers of fatigue. These effects are generally improved when participants are allowed to pick their own music (Karageorghis and Priest 2012b). Crust and Clough (2006) argue the individual level traits such as personality type almost certainly interact with the effects of music on performance. People process information differently; why should their responses to music be the same? The potential importance of individual level variation is evident from the findings presented in Karageorghis and Priest (2012b): the effect of music on the individual is strongly moderated by the method by which the music is selected. 

Our model helps to explain such findings, why liking a track, a track having emotional resonance, or other subjective qualities matters so much in studies of exercise. The more processing a track requires, the more effective it will be at blocking out internal distractions, and tracks that have emotional resonance, meaningful lyrics, or otherwise interact with pre-existing states of the individual will require more processing in different areas of the brain. In a treadmill task, participants listening to motivational music – music which had emotional resonance related to performance, and would therefore require more processing, from different areas of the brains – displayed greater endurance than those listening to a neutral track, and both conditions outperformed silence (Karageorghis et al. 2009).

At low or even moderate levels of exercise, music, especially when it has such interactive subjective properties, is able to crowd out some of the internal signals that might otherwise lead the individual to slow down or stop their exercise. As the level of effort increases, however, these internal signals become too strong to be blocked out by music, and the effect of music on performance fades. 

Researchers have looks at various dimensions of music for impacts on exercise performance, including genre, intensity and volume (i.e. Kreutz et al 2018; Waterhouse, Hudson and Edwards 2010). Though, as Crust and Clough (2006) find, these elements cannot be considered completely in isolation, and interact with each other within a track. For instance, a beat works differently when paired with certain melodic structure than it does alone. Priest and Karageorghis (2008), for instance, find that participants look for motivational music followed by a purposive drumbeat: such tracks would engage both the deep parts of the brain that process motion and tempo, as well as the parts of the brain processing the message and emotional context of the music. As such, it would maximize the degree of distraction from the internal signals that might otherwise inhibit performance. Similar to the music chosen by individuals during cognitive tasks, there seems to be some intrinsic understanding of what qualities best drive performance for a particular individual during a particular task that line up well with the model proposed here.

For instance, Rasteiro et al (2020) find that listening to preferred music increases the amount of time some participants can run in a treadmill task, without changing the physiological indicators that normally govern how hard someone can exercise, and when they feel that they have to stop (blood lactate levels and such).

North and Hargreaves (1996) argue that individuals have some awareness of the differential requirements of various exercise tasks, choosing music that engenders the optimal level of physiological arousal for the task at hand. That is, when people choose exercise tracks, they are, on some level, choosing tracks that will effectively crowd out signals that would reduce their performance. This, too, could be an important reason why preferred tracks have a much bigger effect on outcomes than tracks that are assigned to participants in exercise trials: the tracks that they’re choosing are those that are likely to enhance their performance. For instance, Crust (2004) finds evidence that playing music significantly increased endurance in a treadmill walking task compared to white noise, but while participants preferred familiar (over unfamiliar) music while walking, there was no sign that they performed better when listening to their own preferred music. 

In that study and other, and as with cognitive activities, there is a gap between the effects of music on measurable outcomes and the effects of music on how an individual feels. Even though music has not generally been shown to increase endurance or other measures of performance in high intensity exercise activities, it does increase reported enjoyment of exercise. Even without any impact on physiological markers of exercise, it still can have psycho-physical effects changing the subjective experience of that exercise. Participants who experience less of the negative impacts of exercise are likely to exercise more in a single session, and exercise more often (Annesi 2001; Karageorghis, Terry, & Lane, 1999This is an important outcome, as increased enjoyment of exercise is likely to lead individuals to exercise more.

During high intensity workouts, fatigue signals are strong enough that they cannot be silenced by music, but even then, the music alters the subjective experience of the exercise (Karageorghis et al., 2009). In studies like Elliott, Carr and Orme (2005;  Elliot, Carr and Savage 2004), participants were therefore able to push themselves harder, but without reporting a subjective experience of greater tiredness. During high intensity exercise, music doesn’t seem to impact performance, but it does impact the way in which participants experience exercise and fatigue: they don’t run farther or faster, but they feel better about it (Hardy and Rejeski, 1989; Shaulov and Lufi 2009).

In exercise tasks, tempo has been singled out as an especially important inherent quality of music: individuals seem to respond directly to the tempo of background music, speeding up in response to a faster tempo, and slowing down in response to a slower one (Waterhouse, Hudson, and Edwards 2010; Van Dyck et al 2015; Szabo, Small and Leigh 1999). in tasks like running or walking on a treadmill, the body has been shown to match the tempo of the music being played. One hundred twenty beats per minute (bpm) has been proposed as the “natural” rhythm of the human body, based on the spontaneous speed at which people tap their fingers or walk (Hirasaki et al 1999). 

This would support an idea that exercise efficiency might be maximized by tracks that play at around 120 bpm, but Schneider et al (2010) argue that exercise on a treadmill at least is optimized at around 160 bpm (in line with the preferred music choices of individuals, though there is significant individual level variation, as in Smoll and Schultz 1982).

When an individual hears a beat, the regions of the brain that control movement are immediately and uncontrollably engaged (Fujioka et al., 2012; Grahn & Brett, 2007), with neuronal oscillations aligning with the beat (Calderone et al., 2014; Chang et al., 2016, 2019; Schneider, Askew, Abel, and Struder 2010), as is evident in 3 Hz delta activity. It is generally thought that music in this case is hijacking relatively primitive portions of the brain which set up a repeated task (like running) at a particular rhythm, requiring very little intervention to keep it going, minimizing cognitive load (and potentially increasing efficiency of energy use by allowing the body to prepare for movements more effectively, as in (Roerdink, 2008). Physiologically, this is linked with lower blood lactate levels and oxygen consumption (Terry et al 2012).  

Even this relatively simple relationship between activity and music – moving in time with a beat – is at least partially subjective in nature, with Kornyasheva et al (2010) showing differential activation depending on the preferences of the individual. Beats that are preferred by the individual lead to stronger activation in the ventral premotor cortex, which they interpret as the process of the brain “tuning in” to a liked track.

Such work would imply that beat/tempo is likely to be processed very differently than other elements of a track, and may have a substantial effect on how exercise tasks are carried out (as in Terry et al 2012), even when the track doesn’t impact how well the tasks are carried out. From a research perspective, these results provide more evidence for the preferential processing of music in deep parts of the brain that underlies the model being proposed.

Music has also been shown to play a role in post-exercise cool downs. Jones, Tiller and Karageorghis (2017) show that certain kinds of music – especially high tempo music with strong emotional resonances, can help athletes recover from intense activity more quickly than other types of music, or neutral stimuli. This seems to work because the athletes are able to synch physiological indicators to the music, and it seems likely that music with emotional resonances to the individual would help to crowd out internal distractions that might otherwise interfere with recovery. In some sense, recovery periods from intense exercise are similar to moderate exercise, pointing to a role for music in facilitating recovery (see also Karageorghis et al 2021).

In addition to individual level differences, there are some indication of group-level interaction effects in how music impacts exercise. Some studies find significant differences between men and women in the effects of music on exercise performance, such as Karageorghis et al (2010) who observed sex differences in the effect of motivational music. Etaugh and  Michals (1975) another difference between men and women on the effect of music, but Ghazel et al (2021) suggest that some of the differences between men and women in measures, especially subjective ones, could be linked with menstrual cycles.

There’s a reason why individuals commonly listen to music during exercise, and make use of very different music than they would listen to during cognitive tasks. At low and moderate levels of exercise, music can effectively crowd out physiological signals that would tend to reduce performance, and even at high levels of exertion, music tends to improve subjective measures of performance and reported enjoyment of the activity. The research also makes clear, however, the importance of subjective and interactive qualities of the music, with much stronger effects when individuals are allowed to pick their own tracks, or at least the genre, rather than having it assigned to them. While this leads to serious validity issues for researchers who are trying to isolate the effects of various dimensions of the music, it also means that there is no magic bullet, no single combination of inherent elements that is likely to improve exercise performance across individuals. The music that best enhances exercise performance likely varies by both the individual, the task, and their state of mind.

Music and Athletic Tasks

Less studied is the effect of exercise on athletic (rather than exercise) tasks, and the work that has been done generally points to weaker effects of music on athletic tasks than exercise tasks (Terry et al 2020). Athletic tasks are an interesting case, as they combine elements of cognitive tasks like concentration and attention with physical activity. As such, maximally distracting music – the kinds that enhance performance during light to moderate exercise tasks – would not be useful, as they would take away cognitive resources needed for concentration on the athletic task at hand. But music might still be useful in crowding out internal signals of fatigue, or, perhaps, intrusive thoughts, or over-thinking of tasks that have already been well learned in training. For instance, Jebabli et al (2023) find that listening to (preferred) music improves performance in a jump shot task among trained basketball players. Put simply, athletes know how to do these well-rehearsed tasks, and the goal of the music is to strategically inhibit internal distractions that could reduce performance in them.  If this is the goal of strategic inhibition during athletic tasks, the music that would best enhance performance is likely more similar to that used for cognitive than for exercise tasks.

An example of what this looks like in the research comes from Bishop (2010). Bishop describes an experiment in which a dance track was manipulated so as to have three different tempos (99 bpm, 129 bpm and 161 bpm), and two different volumes (55 and 75 dBA), with white noise and silence as control conditions. Participants (all tennis players)  were then asked to carry out a computer simulation of a tennis-based task (returning a serve) and measured on how quickly they were able to respond correctly to the stimulus.  Interestingly, while volume (an intrinsic characteristic of the tracks) was related to response time, reported enjoyment of the music (a subjective feature) had a stronger effect. This seems likely driven by the operationalization of the tennis task in this study, which was much more similar to the tasks used in studies of attention and concentration than true athletic tasks which require both concentration and sustained physical activity. Other experimental tasks with greater ecological validity (like table tennis returns in Petri, Schmidt and Witte 2020) have shown similar, limited, effects.

Other work on athletics and music have looked at the potential effects of ramp-up music: music that athletes listen to before competing in an event (such music has also been studied for its effect on exercise performance, as in Ballmann et al 2021). Collins et al (2023) finds that music is the most commonly used pre-event priming method among elite athletes (followed by motivational self-talk). The physiological effects of exposure to music tend to persist for several minutes after the musical stimulus has ceased (Bishop, Ross and Karageorghis 2008). Applying this directly to athletics, Meglic et al (2021) find that the use of music during a warm-up activity tends to increase some markers of performance among collegiate athletes. The music that athletes listen to before an event (especially for activities that do not facilitate listening during competition) can have an effect on performance during the event. This is likely due to the role of that music in crowding out internal distractions (like anxiety) that may arise before an athletic performance.

A recurring issue for the external validity of all of these studies is that, as Ballmann (2021) points out, the effects of music on athletic performance are highly conditional on individual preferences. From the perspective of the model proposed here, this makes sense: much of the effect of music in crowding out distractions arises from subjective qualities of the music that activate different parts of the brain than the inherent qualities. But while the effects of music are likely to be individual in nature, the music rarely is: music is played throughout an arena or field during a match or practice. Even if the music being played improves performance among some listeners, it’s unlikely to do so for everyone. As athletes are unlikely to wear headphones while playing, the ramp-up and cool-down effects of music on the individual may well be the most relevant to performance. As studies have shown, music that requires a great deal of processing capacity, because of emotional resonances, lyrical complexity and other inherent features is likely to serve this function best.

Music and Dancing

While researchers have had difficulty carrying out research on athletic activities in a way that has ecological validity to activities in the real world, these challenges are nothing compared to the challenges of measuring social dancing activities, especially those in a club context. Still, the model proposed here gives us a clear indication of what qualities of music are expected to maximize performance (however we want to define it) and enjoyment of the dancing, club (i.e. Pacha in Ibiza,Spain etc) or music festival experience (i.e. Tomorrowland in Antwerp, Belgium etc.).

In a club or music festival setting (as opposed to dances with formalized movements, such as a waltz), the goal of the music is to (1) create and facilitate dancing/movement, (2) keep individuals dancing/moving for as long as possible, and (3) to create a sense of social cohesion among the individuals dancing. On this basis, we can define performance in terms of creating/facilitating movement on an individual level, and longevity in that movement. Put another way, the right music should make more people dance, and keep them on the dance floor longer.

As with athletic performance, there are two strains of internal distractions that music should strategically inhibit in order to maximize performance. The first comes from internal distractions like self-consciousness or embarrassment, which may stop individuals from coming to the dance floor. Such internal distractions are likely best crowded out by highly distracting music, especially music with strong emotional resonances. The second comes from fatigue or related signals, just as in exercise tasks. While music cannot completely crowd out such signals, it can reduce signals that arise from light to moderate exertion, and can change the subjective experience of heavy exertion.

Senn et al (2023) find that the reported urge to move in response to music (“feeling the groove” in Stupacher et al 2022, Etani et al 2023) is driven by feelings of energy (energetic arousal) from the music, their positive feelings towards the music, and the regularity of the tempo. This fits in well with the model that’s been established: strong, regular beats (though the relationship with complexity of tempo is complicated; see Witek et al 2014), as discussed previously, are thought to activate deep areas of the brain related to movement, and subjective and interactive qualities of the music lead to greater interference with internal signals that might lead someone to stay on the sidelines. From a brain area activation perspective, music with such qualities differentially activates parts of the brain involved in the reward system and supplementary motor areas (Matthews et al 2020); other studies have shown music with such qualities leads to differential effects on the level of neurons (Cameron et al 2019).

The role of music in facilitating dance and movement is strongly influenced by its inherent rhythmic and harmonic characteristics. Music with a pronounced, rhythmic bassline and an energetic tempo tends to be more effective at inducing movement, which is why electronic dance music (EDM) is often cited as particularly movement-inducing (Burger & Toiviainen, 2020). While rhythmic stability is crucial, an element of dynamic variation—such as syncopation, bass drops, and build-ups—also plays a key role in engaging listeners and encouraging physical response (Stupacher et al., 2022; Turrell et al., 2021). These elements align with the proposed model, as they heighten arousal and help to override internal distractions. Music with these movement-inducing qualities has been shown to elicit physical responses across different populations, with Kragness et al. (2023) finding increased movement even among infants.

However, these inherent qualities interact with subjective responses to the music (Bechtold, Curry and Witek 2024). As would be expected, dance music has a greater effect on people who have more experience with it, as it is likely to have greater emotional resonance for them, and therefore activate more areas of the brain. In an effort to test the effects of dance music on information processing, Fukuie et al (2022) gave Stroop tests to individuals before and after listening to the music. Individuals with greater familiarity with the music performed better on the Stroop test after listening to high groove music, indicating a greater level of distraction, and that inhibitory functions were suppressed. Dotov, Bosnyak and Trainor (2021) argue that the effects of music on movement are likely mediated by social cues: seeing others respond to music is likely to make us behave similarly. This, too, points to the importance of strategic inhibition arising from music: only when they inhibit internal distractions can individuals give themselves over to the music.

Researchers have also looked at the subjective experiences of dance and movement in response to music: not surprisingly, enjoyment of dance creates a number of positive emotions in the individual (Vander Elst, Vuust, Kringelbach 2021). As with exercise, this implies that even when music cannot provide the distraction necessary to change performance measures (as in the case of heavy exertion), it can still change how people feel about having participated in it.

While there is less research on dancing and club activities than for other tasks described here, the research that does exist – especially that which looks at the components and effects of “groove” – make the case that it fits nicely in with the research already described. Music that has certain inherent qualities (much the same qualities that make music work well during exercise activities) facilitates dancing by maximally distracting from internal distractions and activating deeper portions of the brain. However, these inherent qualities interact with the subjective qualities of the music, especially emotional resonance, to both further distract the individual, and create greater enjoyment of the music and the dance experience. Moreover, EDM music seems to generally match the criteria associated with both groove and distraction, with a combination of strong beats and some degree of unpredictability.

Music and Human Performance: A Research Agenda

The research described here suggest that music can enhance human performance in a variety of areas, and while the way in which music does so varies, there is a single underlying model of strategic inhibition of internal stimuli that unites all of the use cases. Certain cases require low levels of distraction to allow for attendance to a task; others require maximal distraction in order to crowd out as many distractions as possible (leaving little cognitive processing available for other tasks). 

The research has also hinted that individuals are aware of this relationship, choosing tracks that work best for the task they are participating in, especially those that have emotional resonance, or other subjective qualities. The interactive nature of these qualities of music – the inherent and the subjective – may help to explain many of the divergent findings in the field over the years. Simply put, the controls used, such as giving all participants the same music, may be experienced very differently by the participants, biasing the results of the study towards null effects. Individuals are good at picking music that blocks out some stimuli in order to concentrate better, or has strong beats/drops and emotional resonance to facilitate exercise tasks. Throughout all of the work on the topic, research has also shown that liking the music matters a great deal: even if a track has all of the inherent characteristics that are useful for a task, it won’t work as well if the individual doesn’t enjoy it.

But what happens if the music that someone enjoys doesn’t have the subjective qualities needed to optimize their performance in the task? If an individual likes a track, they don’t have the option to increase the tempo a bit, or reduce the lyrical complexity, or otherwise personalize  it in order to enhance their performance. They have the option of listening to a track that might be too distracting – or not distracting enough – or trying to find a different one to listen to.

But if individuals were able to personalize or mass customize  the traits of a track – especially one that they liked – we expect that we would see both increased performance relative to the baseline track, and greater enjoyment of the task (which is important, as it leads to greater perseverance in the task going forward). We believe that EDM is an especially good starting point for this sort of work, as the structure of the music lends itself well to Artificial Intelligence driven mass-customization of core  music characteristics (Tempo, Drop, Rhythm & Beat, Lyrical Complexity, Timbre & Loudness, Melodic & Harmonic Structure, Genre Influence).

The closest thing to this approach is work from Haruvi et al (2021), in which participants were attached to portable EEG headsets that monitored focus levels while listening to either silence, pop music playlists from Spotify or Apple labelled “focus” or “flow,” or what they refer to as “engineered soundscapes.” These soundscapes – sold by companies like Endel, Melodia and others – are combinations of sounds and noises that supposedly increase cognitive focus.

Participants then did an hour’s worth of cognitive tasks: 30 minutes of an activity chosen by the participant, 3 minutes of math problems, 3 minutes of the Alternative Uses Task, and two levels of Tetris. After each task, participants recorded their subjective level of focus, enjoyment and stress. But, even here, individuals were not able to alter the characteristics of the music.

More basic research across tasks is needed, especially in the areas of athletic performance and the facilitation of dancing, but replication of existing results on concentration, attention and more externally-valid cognitive tasks would also be valuable. Any research going forward should also make use of standardized measures, like the FAIR (Frankfurter aufmerksamkeits-inventar) concentration test and the Brunel Music Rating Inventory (Karageorghis, Terry, and Lane 1999).

The goal of this work should be to identify ways in which individuals can control their own musical inputs in order to maximize performance, while also enjoying the music they’re listening to. Music is not, and cannot, be just about maximizing performance. It must also be something that individuals enjoy. Allowing individuals to mass-customize the music that they are already using to help them perform their best allows them to do both. The psychologist William James (1890) wrote that “the more details of our daily life we can hand over to the effortless custody of automation, the more our higher powers of mind will be set free.” To this, we would add: the more we can silence, at least temporarily, the parts of our minds that limit our capacities, the more we may be able to accomplish.

Footnotes:

[1] Individuals with some degree of musical training tend to process it in different regions of the brain, especially in areas more related to language (Ohnishi et al 2001), which seems to correspond to the naming of elements in the music (Zatorre et al 1998).

[2] Similar effects are seen in the distinction between low impact and high impact exercise in the next section.

[3] One exception with this is a mismatch between the language in which lyrics are sung and the fluency of the listener – complex lyrics in one language may not be experienced as such by someone who does not speak that language, or may require greater processing for someone who has less than full fluency in the language.

[4] See Rowell and Flick (2019) for a rare study focusing on ecological validity, looking at the effect of preferred soundscapes on writing production among first-year writing students.

[5] Karageorghis, Terry, and Lane (1999) list four dimensions of motivational qualities of music. Rhythm response, musicality (pitch, harmony), cultural impact, and associations. The last two are subjective.

[6] Karageorghis and Priest (2012) similarly divide the effects of music on performance into four categories: ergogenic (changes to measurable exercise outputs like speed or endurance) and psychological (the emotional  experience of the exercise task for the participant), psychophysical (the perception of fatigue or physical effort, generally measured through the Borg Rating of Perceived Exertion (RPE) scale, and psychophysiological (measurable effects of mental states, like heart rate or blood pressure). [7] Other research has shown the extent to which the process of synchronization with the beat can be relatively complex (Toiviainen et al., 2010), moving on every other beat, and so on.

[8] Note that our definition here of athletic tasks is limited to tasks that require concentration and responses to changes in the environment as well as sustained physical activity. Tennis, hockey, basketball would all be considered athletic tasks; running, swimming or other tasks that rely mostly on repetitive motions. Such tasks would be more akin to intense exercise tasks.