On Musical Selection in Research: 3. Bottoms Up!

Did you hear that existentialist percussion piece? It was highly cymballic.

We have previously discussed arrhythmic music in reference to ambient and minimalist compositions. Other types of music without a pulse include noise, ambient, and drone music. While these vary drastically in terms of context and performance, and especially volume, they tend to be outlier music and thus rarely fall into standard analysis. However, with the background and neuroscientific analytical vocabulary we’re building, such tracks should lend themselves well to description on a case-by-case basis.

So, we come to a fruitful realm of study, namely how the brain responds to rhythm. Life itself happens rhythmically, and some sort of rhythmic action or resonance can be found within the basic functions of all life on earth. Advanced beat perception, prediction, and locomotive entrainment, is a lively area of research. In other words, we’re studying dance.

In scientific parlance, synchronizing motor processes with auditory cues is called entrainment. When you tap your hand along with a metronome, you have entrained to a repetitive auditory stimulus. The word comes from literally stepping onto a train from a stationary platform. To help remember its meaning, I imagined the train cars as a beat clacking by, then stepping onto the train as a way to anticipate, then synchronize with that steady movement.

When musical entrainment goes too far.

Most animals are terrible dancers. Just terrible. But a short list can do it, and in this presentation by Drs. Aniruddh Patel and Ola Ozernov-Palchik, they postulate that vocal mimicry is the key to whether an organism is capable of the kind of temporal perception/anticipation necessary for accurate rhythmic entrainment. Their research shows a strong connection between aptitude at reading, speaking, and rhythmic discrimination in human children. After all, language itself is temporal, and processing the time-information of speech is inherent to understanding what has been said.

Researchers in the last few years were a little surprised to discover that non-human primates are pretty bad at rhythmic entrainment to music. You’d think it’s related to intelligence, but even primates capable of simple drumming take a long time to learn the skill, and tend to always be a little a late. Rather than anything to do with intelligence, it’s due to a lack of neural motor region coupling to auditory regions, meaning these animals can only judge a beat by determining the interval between pulses, instead of predicting when the next beat is going to land like we do.

Since monkeys are born with all their vocalizations intact, they don’t have the vocal mimicking ability we think is required for rhythmic entrainment. So, can you guess which animals are best at it?

When they said “Everybody,” they really meant it.

The most prominent non-human vocal learners in nature are songbirds, which have quite adorably been proven to be able to spontaneously dance to unfamiliar music at varying tempos. An example is Snowball the Dancing Cockatoo, whose above video deservedly went viral a few years ago. Thanks to this video, Snowball caught the attention of the aforementioned Dr. Patel, thus beginning a lively ongoing discussion as to which animals can and can’t perform this feat, and why or why not.

The list of dancing animals is currently humans, songbirds, elephants, and most recently sea lions. Bonobos are an interesting soft exception, which you can read more about in this article, Beasts That Keep the Beat. There’s also a little more here regarding the relationship between motor anticipation, motor learning, temporal perception, and social engagement, including how these factors relate to beat processing in various species.

While musicality has long been associated with neurologic linguistic processes [1][2][3][4], it is by no means the full story. Music activates very deep brain structures, many of which are not related to language in any way we know. But it turns out auditory mimicry might be an inherent part of rhythmic entrainment and, thus, musical development.

The Auditory Brainstem Response

Hearing a steady rhythm sends matching electric pulses through the brain in an amazing phenomenon called the auditory brainstem response. This describes how auditory stimulus directly corresponds to signals in brainstem activity. The relationship is sensory, i.e. bottom-up processing, and it is also, in my opinion, totally awesome. Look at these paired waveforms below.

Auditory brainstem response
Examples of the auditory brainstem response. More examples.

These are audio stimulus vs. the corresponding EEG readings of a human brainstem. For instance, in the pair labeled “Piano Melody,” notice that the electric impulses in the brainstem line up perfectly with the steady pulse of the piano notes. Such synchronized impulses mark the beginning of a vast connection to many centers of the brain, as laid out in this paper by our trusty friend, Dr. Thaut. Our brains function by sending out pulses of electricity via neuronal networks, so this bottom-up effect from rhythmic stimuli directly alters the landscape of said pulses.

After synchronized pulses to auditory stimuli occur in the brainstem, electric signals travel on to regions such as the spinal cord, the subcortex and cortex, strongly interacting with the motor system. The brainstem regulates most of the basic cyclical body functions, such as cardiac and respiratory processes. It’s also pivotal in maintaining consciousness and regulating the sleep cycle [1].

The Chanda/Levitin paper lists several ways we know the tempo of music affects our physiology. As I’ve pointed out is generally the case, they could find no direct neurochemical relationship to “relaxing music,” but they found a strong correlation between pulse tempo and rhythmic bodily processes.

“These effects are largely mediated by tempo: slow music and musical pauses are associated with a decrease in heart rate, respiration, and blood pressure, and faster music with increases in these parameters. This follows given that brainstem neurons tend to fire synchronously with tempo.”

“The Neurochemistry of Music, Chanda, Levitin, 2013

Heart rate and breathing are directly related to our emotional state at a given time, not only with regards to rapidity but also regularity. Unfortunately, many texts related to this often lack rigor, likely related to sources of said papers offering paid services relating to it. However, some highly regarded researchers – including another researcher hero of mine, Bessel van der Kolk – have observed that PTSD is strongly correlated to heart rate and breathing variability, resilience, and coherence.

For more on the symptoms and treatment of PTSD, I strongly recommend this wonderful book, The Body Keeps the Score by Bessel Van Der Kolk

The Chanda/Levitin paper demonstrates that rhythmic musical stimulus directly regulates vital body functions as a bottom-up electric brainstem response. In other words, cultures have been utilizing percussive beats and group drumming to regulate physiological, neurological, and neurochemical stress and trauma responses for many, many years. When the body starts to go into panic mode, rhythmic stimuli can help prevent a learned traumatic response by, basically, removal of neural resources to cause panic, and by brute-forcing neuroelectric pulses that prevent heartrate and respiration to speed up or get out of sync.

Bilateral Stimulation

Drum circles have a long and well-studied history of helping foster positive therapeutic environments, shown to help with at-risk behaviors, depression, anxiety, and addiction to name a few. Group drumming and/or dancing exists in virtually every culture on Earth, long before the advent of modern psychology, and are often considered explicitly therapeutic or cleansing in nature. While we know communal dancing and hand drumming have therapeutic effects for individuals and communities, we don’t know how or even if that mechanism differs from, say, group discussion, feasting, or other communal events. Rhythmic communal cooperation, specifically, is difficult to differentiate from other forms of group therapy.

So, what’s one way rhythmic stimuli differs from others as a form of therapy? Interestingly, a type of therapy called EMDR uses rhythmically alternating bilateral stimulation via eye movements, knee/thigh tapping, or motorized pulsers to impose a constant mild distraction to each hemisphere of the brain. This is done while the subject goes over troubling memories that usually cause a post-traumatic spiral. There is a growing body of sound evidence showing the efficacy of this method, although how and why it works is still not fully understood [1][2][3]. But as it is a rhythm-based therapy, I thought looking a little closer at the exact methodology might be interesting in our neurological beat processing examination.

EMDR stands for “Eye Movement Desensitization Reprocessing.” The most common method used for it in 2019 is with small motorized pulsers held in each hand. This has been shown to reduce the amount of stress a person feels while recounting traumatic memories. Tapping refers to repetitively tapping on the body, usually the knees or thighs, to achieve a similar goal. I spent some time watching various videos of EMDR and tapping sessions, such as this EMDR one from Dr. Jamie Marich and this resource tapping one from Dr. Laura Parnell.

I paid a lot of attention to the tempi used in these demonstrations, clocking the rates as within a narrow range usually hovering around 85-95 BPM. Dr. Marich completes one cycle (back-and-forth) at about 88-90 BPM, while Dr. Parnell is slightly faster, at about 90-94 BPM. The slowest tempo I could find was about 70 BPM, and none went faster than 100 BPM. So the tempo is always slow, which would make perfect sense when trying to keep someone calm. I can only imagine it would be stressful if someone started hammering on your knees at 160 BPM and asked you to recall upsetting memories.

The crux of this technique and how it functions is called bilateral stimulation. This means continually activating both hemispheres of the brain using a left-right alternating pattern. Eye movements are a very easy way to activate different regions of the brain, which is partly/mostly/largely/probably why they happen as we dream in REM sleep [1][2].

Bilateral tactile sensations have a similar effect to moving the eyes back and forth. If you tap on your left leg with your left hand, and your right leg with your right, alternating at a steady pace, it will also alternate hemisphere activation. There is some evidence that this can effectively reduce the stress response as well. All of which is summarized in this quote from Dr. Robert Strickgold:

“We propose that the repetitive redirecting of attention in EMDR induces a neurobiological state, similar to that of REM sleep, which is optimally configured to support the cortical integration of traumatic memories into general semantic networks.”

EMDR: A putative neurobiological mechanism of action, Strickgold, 2001

Which is super cool. The repetitive distraction induces a dreamlike state, is what he’s saying. Contemporary psychotherapists are developing a verified method of neural engagement that denies the brain the resources it needs to drum up (sorry) fearful or anxious thoughts. To reference the loudness paper again, takes up the neural space that might otherwise promote negative arousal.

Many studies discuss the ability of repetitive drumming – whether listening, dancing, or playing – to induce a trance-like state [1][2][3] similar to hypnosis, all of which falls under temporary, non-pathological dissociation as a form of self-regulation. Which is a hell of a sentence.

“Music emerges as a particularly versatile facilitator of dissociative experience because of its semantic ambiguity, portability, and the variety of ways in which it may mediate perception, so facilitating an altered relationship to self and environment.”

An empirical study of normative dissociation in musical and non-musical everyday life experiences, Herbert, 2011

Now, imagine hand drumming in a circle of friends, colleagues, similarly trained musicians, communally associated peers, or whatever. Hand drumming, by definition, is a self-controlled rhythmic bilateral stimulation in which the player’s hands tactically engage in a steady alternating pattern. This regulates the player’s nervous system by entrainment, which is also synchronous with the surrounding players and audience. One can visually confirm this by observing head bobbing, side-to-side swaying, foot-tapping, etc. in surrounding participants, all of which are also bilateral movements. Collaborative human social behavior is deeply rooted in our brain structures [1][2][3], though studies regarding this social collaboration relating to activities like a drum circle are unfortunately lacking. One can at least guess that this process involves temporal anticipation/prediction/discrimination, as well as motor planning and execution, all confirmed visually, aurally, and socially, which cannot be achieved by similar but arrythmic activities.

Bilateral entrainment, Feynman-style.

In short, psychotherapists are already using hidden versions of rhythms as a way to deal with overwhelming stress and trauma – a contemporary refinement of what we were already doing for millennia with social drumming. I personally wonder if nervous rhythmic actions that produce sound, like idle/anxious knee or desk tapping for example, fits into this picture somewhere.

I would be remiss here if I didn’t mention binaural beats, a fun experiment showing how the brain recreates acoustic beating in the brainstem. Separate sound waves, one in each ear of a pair of headphones, combine as electric brainstem pulses to recreate the type of complex wave one would hear in real world, like this:

Two interacting waveforms resulting in a beat pattern.

The two waves are mathematically pitched so that combining them creates a beating sensation not heard by either ear individually. Many find these extremely soothing as a form of sound therapy, and you can read more about them here and listen to a neat demonstration here.

All this suffices to say that rhythmicity in music arises from a deeply innate neural architecture, the so-called glue that binds together the vast majority of human-organized sound. For the purpose of this series, I spend so much time on it to show that there is a true difference between sound, ambience, or noise vs. auditory stimulus with a pulse.

Types of Rhythm

I’ll now switch track to discuss an infinitesimally small pool of examples showing the various ways rhythm arises within music. Rhythm in a sonic work can be explicitly percussive – i.e. in drum ensembles – such as in an Ewe, samba, or taiko ensemble. You’ll also find it in solo pieces this one for Korean jang-go, as well as this contemporary percussion composition in the raga style, Piru Bole. However, rhythm exists much more than percussion music. Rather, it’s a necessary element in anything with a meter or pulse, whether or not it’s actually percussive in timbre. This makes the effect of rhythm on the brain extremely diverse. Let’s try and characterize some examples using the cake method.

In the above example, we hear a rhythm played via rock drum kit solo. It possesses no pitched or harmonic content. The tempo is very fast, accelerating to around 150 BPM, and contains rapid subdivision and snare rolls. The timbre is obviously percussive, utilizing the full range of the kit, creating a wide spectral texture from the bass drum to the cymbal crashes.

Something rather surprising is that no existing scientific study I can find relates brain activity to drum solo listening. The reason is probably that a drum kit is actually a collection of instruments played simultaneously, making it too difficult to extrapolate meaningful data. Instead, researchers tend to look at rhythmic sounds one at time. This presents the problem of deciding which texture to use when studying rhythm. There’s probably a big difference between how the brain receives the same beat pattern depending on whether it’s middle C on a piano, a sine tone, a metronome click, a bass drum, or a crash cymbal. Indeed, simply the volume of a kick drum has an effect on bodily entrainment. This is is why Garth’s solo up there, as it currently stands, is completely outside of the realm of existing research.

Rhythmic content is fast and complex, accelerando to ~150 BPM. No pitched, melodic, tonal, or harmonic content. Standard drumkit timbres with full spectral range. Quite loud. Structure is through-composed freestyle solo, short in overall length. Live improvisatory rock context.

The first ~1:40 of this track is an arpeggiated melodic line generated with a Roland TB-303. Despite having no sounds that are traditionally percussive, the rapid pace, fast-attack wave shapes, and jumps between low and high pitches creative a highly rhythmic, dancey feel without the need for synthesized drums. Which, of course, makes the composition all the more effective when the drums do come in, then go out, then come in again several times.

Rhythmic content is rapid 4-to-the-floor dance beat. Melody is repetitive, fast, and arpeggiated, containing large leaps, making non- or nearly non-singable. Harmonic content is enharmonic slow, soft pads. All timbres are computer generated/synthesized in the style of acid house techno. Dynamics alternate between quieter sections and loud beat sections, which play out over an unusually long runtime of 16 minutes. Club/raver/electronic dance tradition.

The second movement of Ravel’s sole string quartet composition is a really great example of why music selection is so important when conducting research. The beginning section is highly rhythmic and exciting, featuring constant pizzicato techniques and arpeggiation, just like the previous acid techno track. It also counts as a classical composition and employs so called “relaxing” timbres, namely, strings. When a research paper says it used classical music as a relaxing control, what if this piece snuck its way in? Compare this to Adagio for Strings by Barber. If we don’t know the tempo, timbre, or general feel of each individual piece, both might be characterized as similar in a research environment, although I can guarantee both would activate quite differently in the brain. Indeed, this is a perfect example of why each aspect of a musical selection must be indicated in research, because, generally, this track shares major similarities in some ways with the acid techno track, and other major similarities with Adagio for Strings, but all three pieces would be received quite differently by a listener.

Focusing only on the first movement: Highly rhythmic, achieved via constant pizzicato eight, sixteenth, and 32nd notes. Highly melodic, recurring singable theme. Enharmonic key structure with Euroclassical modulations achieved mostly with arpeggiations between players. Timbres are full-spectrum pizzicato and arco strings as in a traditional quartet. Structure, dynamics, and context are variable as is consistent with the Impressionist Eurotraditional artistic epoch.

Percussive melodic instruments such as the marimba demonstrated above, as well as the xylophone, gamelan, or vibraphone (to name only a few) are very pure combinations of rhythm and pitch. They function by producing a tone with a relatively pure timbre (i.e. closer to a sine wave) and seem to be used about as often as metronome clicks when studying beat processing. I can’t find a study that looks at the difference between pitch and unpitched perception along this axis. Which is weird.

Rhythmic content is fast and steady, highly melodic, enharmonic Eurotraditional key structure, wooden malletophonic timbre and spectral range, relatively consistent medium volume, Baroque tradition.

And so, with Steve Reich’s Clapping Music, we finally venture into the realm of postminimalism and phasing music. We’ll get slightly more into what that means in a second. But I want to focus on this particular piece because it’s another really good example of the difficulty of music selection without being very specific. First of all, this is a highly rhythmic piece with no vocals or pitched content. However, it also very clearly has an human element. The brain may react completely differently when it can tell a percussive sound comes from a human rather than a manufactured instrument, in the same way the brain responds completely differently to vocal sounds compared to otherwise similar sound content.

Instead of breaking it down, let’s instead take a look at how different this piece sounds based on the recording, like this fast-paced ensemble version of Clapping Music. It sounds completely different from this version. Or look at Evelyn Glennie performing it on woodblocks, or these super cool jugglers playing it slowly with bouncy balls, or, my favorite, actress Angie Dickinson performing the piece in the 1967 film Point Blank. They’re all the same piece, but some feature extremely different tempi and timbres, likely provoking strong variations in the neurological listening experience. So, even if a researcher actually mentions the name of the composition (which they don’t always do, at all), unless we know which specific recording of that piece is used, we still can’t rely on any data presented.


We began with Cage’s silent music, then added arrhythmic, meditative harmonies (with some naturalistic textures) to build a minimalist composition. Now, by adding a rhythmic element to minimalist practices, we create the genre (or subcategory) known as postminimalism, also sometimes called phasing music.

In most minimalist music, just like the Pisaro piece or any of the ambient works, slow harmonies progress over a long period of time. In postminimalism, the same process occurs but with an added pulse, and usually short snippets of interlacing melodies that phase in and out. The intro section of Steve Reich‘s Electric Counterpoint is a really perfect example of how this compositional technique arises directly from a true minimalist approach.

Performed beautifully here by Mats Bergström.

The introductory choppy chords eventually move into melodic snippets that fade (or phase) in and out, giving sensation of movement while retaining the slow-moving harmonic structure valued by minimalism. However, by avoiding any longform, easily singable melodies (short, quick notes with large leaps) we avoid true ostinati/thematic content like in other genres.

As a note, if you read descriptions in the links, you’ll notice that the term postminimalism is sometimes lumped in with minimalism, though in contemporary music practice this is considered inaccurate, or perhaps just plain old lazy.

I am very personally interested in the difference between rhythm as accorded by something like Electric Counterpoint vs. a West African Ewe ensemble or Garth’s drum solo up there. Both contain strong rhythmicity, but one comes only from pitched/chordal content and the pulse is very clear, i.e. comprised entirely of eighth notes and eighth rests. The others are unpitched and feature far more complexly interlaced syncopation. Would there be some marriage of stimulus in the Reich piece that we wouldn’t see in the unpitched one? Is the strength of response different by the nature of how rhythmicity is achieved in one or the other? Or the spectral range or waveshape of either? Does the harmonic content neurally allow one to be more repetitive than the other before boredom sets in?

Absolutely none of these questions have been addressed in any research I can find. I think it has a lot to do with the fact that the above questions don’t directly relate to rehabilitation of disease/disorder, and are rather purely theoretical. But still, I can’t imagine having that kind of information would tangentially inform therapeutic musical neuroscience research. Anyway.

Every time we add a new element to our cake, it begins to look more like a cake. Cakeness is subjective, but few would look at Electric Counterpoint and deny that it is a piece of music, which some actually might about the Pisaro piece, and many still do with 4’33”.

Now that the overall point of this exercise is becoming more clear, and just because I love postminimalim, I’ll share some examples of postminimalist recordings below, all of which have similar properties. If you were to characterize each using the elements of music previously described, how would they differ between each work? What would be similarly or identically described?

The Chairman Dances, originally from John Adams‘ opera Nixon in China.
The popular Koyaanisqatsi, a visual tone poem collaboration between composer Philip Glass and filmmaker Godfrey Reggio.
Cruel Sister, by Bang on a Can composer Julia Wolfe.


If lazily described, much of the above music might sound identical to one another – “percussive music,” “rhythmic music,” “fast music,” “electronic music,” or “classical music” as examples. However, I hope it’s growing clear just how variable a listener’s experience of each recording might be. Koyaanisqatsi, for example, is often called minimalist because its tempo is so much slower than the other two examples. But this piece contains more musical elements than the Pisaro. It has a strong rhythmic pulse, a wide-ranging spectral character, and far more timbral information, including the game-changing element of human voices. Even without taking individual bias and tastes into consideration (like if someone happened to know and like the film, for example), the two pieces will objectively have a quite different neurological effect on the listener despite a casual description running the risk of lumping them together as the same.

In the next section, we will discuss pitch, melody, and harmony in greater depth, the latter of which will lead to the discussion of structure in music. As this will lead us squarely into the territory of folk and pop music, instead of a parting image, I leave you with a really chill remix of Reich’s Electric Counterpoint by RJD2 from the album Deadringer. Also feel free to enjoy Jonny Greenwood of Radiohead performing another version of the same piece.

Thanks for reading!

About the author


View all posts

1 Comment

Leave a Reply