Peter Keller: On Becoming More Skilled Ensemble Musicians with Insights from Cognitive Science

It was a late spring day in the 5th grade, and the school I attended was having its annual let’s-get-the-kids-outside-and-make-them-compete-at-random-made-up-sporting-events extravaganza. There was tug-of-war, water balloon tosses, running laps while balancing an egg on a spoon, potato sack races, etc.

Anyhow, my friend Shane and I were paired up to compete in the three-legged race (in case you’ve never seen this, here’s what it looks like ). We tied our adjacent legs together, lined up with our classmates at the edge of the field, put our arms around each others’ shoulders, and decided which leg we’d start with. We also agreed to say “one, two, one, two” out loud together to help us keep our legs coordinated.

When we heard the signal to go, we said “one” together, stepped forward in sync, and gradually built up speed, continuing to count out loud. It wasn’t long before we were legit running, where all four (or three?) of our legs were completely off the ground between strides.


When we got to the finish line and looked back, we discovered that the race wasn’t even close! None of the other teams had even gotten to the halfway point yet. Some teams were on the ground, some were dragging their partners across the field, it was the hilarious mess that you’d expect from a bunch of 5th graders (exactly like these kids ).

Nearly 40 years later (😳), I still remember how in-sync we were that day, and how effortlessly we ran across the field. And though I haven’t talked to Shane in years, I’m betting he still remembers that day too. 


Indeed, whether it’s playing sports, dating, or collaborating in an ensemble with other musicians, we’ve all experienced moments of “chemistry” (or the complete lack thereof) with certain people.

Where we just feel like we’re more in sync on some fundamental level. Like, our natural inclinations and instincts seem to align so well, that being together or playing together not only feels easy, but is kind of a thrill in and of itself.

Have you ever wondered why this is? Or more specifically, what the underlying neural or behavioral bases for this kind of chemistry might be? And whether these are skills we can learn and develop? And if so, how?

Meet Peter Keller

Peter Keller is currently Professor of Neuroscience in the Center for Music in the Brain and the Department of Clinical Medicine at Aarhus University in Denmark, and Professor of Neuroscience of Music in the MARCS Institute for Brain, Behaviour and Development at Western Sydney University in Australia.

His work today centers around trying to better understanding the cognitive and motor processes that enable musicians to play effectively together in ensembles. But once upon a time, Peter was a conservatory student majoring in trombone, musicology, and composition. An experience which, as you’ll see from today’s chat, led to many of the questions that he explores in his research today.

In today’s chat, we’ll explore…

  • 2:42 – Peter’s journey from trombone to psychology, and a particular aural skills class, in which the teacher’s unique methods may have influenced some of Peter’s research interests today.
  • 11:43 – What is “anticipatory auditory imagery,” and what effect does this have on both our experience of the performance, and what the listener hears as well? (This is something legendary tuba player Arnold Jacobs often spoke about in his teaching, which trumpet player Kristian Steenstrup describes in his podcast episode at 2:42 here.)
  • 14:22 – How does anticipatory auditory imagery work when it’s not just you, but you’re playing with other musicians too? (Pianist Vivian Hornik Weilerstein talks about this exact thing in her podcast episode too – just in different words. You can listen to that at 41:02 here.)
  • 17:31 – And how do musicians adapt or adjust to spontaneous things in performance?
  • 22:14 – Are there any ways to test, measure, or predict how effectively one may be able to play with others?
  • 26:56 – Why is striving for “perfect” ensemble – i.e. 0 milliseconds of asynchrony between musicians – actually undesirable? (This reminded me of Vivian’s explanation of “lined-up ensemble” vs. “emotional ensemble” at 27:56 here)
  • 28:35 – Is there a research-based rationale for teachers’ asking students to conduct in lessons or move physically while mentally audiating a piece they’re working on to help them internalize a stronger sense of rhythm or pulse? (This is something that violinist Catherine Cho speaks more about, at 10:02 of her podcast episode here!)
  • 30:27 – Why is dance music most effective when it’s loud?
  • 34:06 – Can people play effectively together even if there are individual discrepancies in their philosophies about effective ensemble playing?
  • 37:52 – What could account for different members of the same ensemble having very different impressions of the same performance, where some thought it went well, and others thought it didn’t go well? (Cellist Merry Peckham speaks about this in her podcast episode, at 35:50 here!)
  • 39:30 – Are there certain characteristics that might make some individuals predisposed to be better at certain instruments than others?
  • 41:58 – If there are specific ensemble skills that seem to be essential for effective ensemble playing, do we currently have ways of measuring or “testing” for these skills, so we can identify who these effective ensemble players might be?
  • 43:50 – What implications could this research on ensemble skills have on orchestral auditions? As in, is it possible that some ensembles may be getting enough data about a candidate’s instrumental skills, but not enough data about their ensemble skills, to make decisions about which musicians would mesh well within that ensemble?

One quick thing before you listen!

Peter references his Ensemble Skills Framework at about a third of the way into the episode, but really starts alluding to it much earlier than that. I think the visual model really helps to make many of these concepts more concrete and organize them in our brains, so take a quick look below and just keep it in mind as you listen to the episode:

Peter Keller
From Keller, P. E., Novembre, G., & Hove, M. J. (2014). Rhythm in joint action: psychological and neurophysiological mechanisms for real-time interpersonal coordination. Philosophical Transactions of the Royal Society B: Biological Sciences, 369(1658), 20130394.

Subscribe to the weekly podcast via iTunes | Spotify | Google Podcasts | Stitcher

Someone asked me the other day, what genre or type of movies that I liked best. And I hadn’t really thought about that question up until that point. But as I’ve thought about it more, I think I’ve always really enjoyed the origin story type movies like how did Peter Parker become Spider Man? Or how did the X-Men come to be and so forth. What I like is you get to see how these little tiny things that don’t seem very important, at the time, end up foreshadowing what’s to come. And so before we dive into your research, and what it is that you do now, I was wondering if it might be okay to start by asking you to share a little bit of your background because you do have this blended, unusual background of music and psychology. And I’m curious to learn more how that might have come to be and how that led to your current work and particular research interests.

Sure. In terms of origin stories, unfortunately, there are no superpowers to divulge. But my early interests were in music as a performer, and composer. And in fact, I used to, don’t tell my teachers this, but I used to wag school, take days off school, to stay home and write music if I was inclined to do so occasionally. So I was very, very enthusiastic, motivated about that. And also performing starting off playing in brass bands, playing a baritone, and then euphonium, eventually realizing that for a career as a brass player, it made sense to pick up the trombone to have more opportunities to play in orchestras and actually get paid to do so. So I started studying the trombone and attended a tertiary level Conservatory, and then a couple of universities before settling to do a performance major at the Canberra School of Music working with a very excellent trombone teacher who’s there only very briefly, Michael Mulcahy. He eventually got a posting in the Chicago Symphony Orchestra where he continues to play. So I was very serious about trombone. When he left, I realized that I was interested in other things as well. And in fact, psychology.

So one of the interesting experiences I had at the Canberra School of Music, was the aural skills course, was of course, there were component the usual components, like melodic dictation, and all these sorts of things. But the teacher there had method that was quite unique at the time, it was very active method. We would pile into his office in groups of four or five musicians with our instruments, and do a lot of performance exercises, not as well as perception, but production exercises, focusing not just on pitch, but really on performing complex rhythms. And importantly, together as a group, so learning how to coordinate complex rhythms together poly rhythms, but even just complex forms of musical texture. And that was something that I hadn’t experienced before, it was something that had been a challenge as an ensemble performer. And, of course, at the time, I didn’t realize the implications of it fully, I kind of just put it back of mind. And in the meantime, started studying psychology, I moved back to Sydney from Canberra. And once Michael Mulcahy left, I decided to put aside also the performing and go back to composition and music theory and finish my degree in those disciplines, at the same time taking up psychology. That was more just as a hunch, and almost as a plan B in case the music career didn’t take off. Psychology seemed like an interesting prospect, and one that had some sort of security.

But of course, once I started studying, I realized that there was more to that, and developed a passion for it. And even more to the point, when studying music, so in music theory, my thesis was on the Czech composer Leos Janacek, and amongst his many contributions, he’s most well known for his operas. And what was special about those is that the way he treated the setting of the text, the language. He did so in a way that was supposed to enhance the emotional truthfulness or veracity, that he was trying to communicate the emotions of his characters as reliably and accurately as possible. And to do so he developed what he called the, of course, not in the same language, but it translates to the speech melody theory. And what he would do would be going around the streets, notating snippets of people’s speech in different emotional states, you know, somebody could be agitated, they could be happy, joyful, they could be morbidly sad. He would walk around and surreptitiously notate, I guess in a notebook, the contours of their speech much maybe like Messiaen around with birdsong and so on but Janaceck was interested in emotional prebiotic qualities of speech and not so much to then steal them and use them in his operas when setting the text, but to train himself to then imagine the text of a libretto and compose a melody that was faithful to the contours of somebody making that utterance in a particular emotional state in a real life context. And interesting thing about it, he developed this theory while taking a pause from composing one of his operas, Jenůfa. And my thesis was based on looking at the first act and comparing with the second third act, in terms of how the qualities that you would expect of a speech, like, what the melody should look like to differ between those acts. And there was some evidence for this, and we did some statistics to quantify what was going on.

But the turning point for me was then thinking, okay, that’s all great, we’ve counted every note in the score and annotated every contour. And really, it was quite a big job. So in theory, the second and third act should be more emotionally forceful than the others. There was an issue there to the dramatic potency changes, but the statistics was able to deal with that and account for that. But what the real question then was, well, actually does it work? And to do that, I realized that one has to do perceptual experiments, to play the music to people, ask them about their emotional responses, or even better objectively do some physiological or other recordings to gauge those. So I decided that’s when I had to take psychology seriously to learn the methods to answer those sorts of questions, which was essentially musicological, but they were really about how humans perceive, how their brains make sense of an experience, the music, and one needs to really jump into that to learn the methods rigorously to apply them in a way that is scientific. So that was the kind of a switching point for me when I really switched from music to psychology. But still, the object of study was music and the way people process music. It wasn’t at that stage a question of, I want to understand the brain and perception in general and I’m going to use music as a tool to do that, which you know, a lot of work going on at the time was more in that vein. Since then, of course, now I’m operating in both ways. And I have one foot in both camps, because I’ve learned to appreciate the complex beauty of the human brain. And I think it’s an interesting object to study its own right. But importantly, it’s only interesting insofar for me at least, how its role in, of course, allowing us to behave in the way we do, and especially in social contexts. And the prime example of that, I think, is the way humans interact through music. And that’s the topic of most of my research at present.

So I’m kind of curious about this aural skills course, because I feel like it might transition into, I think the first paper that I came across of yours was the 2012 paper on the role of anticipatory auditory imagery. I’m not sure if you coined that phrase, but it’s the first time I’ve seen that phrase, anywhere. And essentially, if I understand the paper correctly, or the review correctly, it was how engaging in mental imagery during performance was not something that had been studied very much. But it does absolutely seem to have an effect on the performance itself, and what the audience hears and so forth. Because, you know, thinking back to when I was a kid, I think I just let my mind wander wherever it felt like going at the time, and it could be related to the audience, or food or TV show that I’d watched before and didn’t get to see the end of, and I didn’t really know what to do with my thoughts and performance. It just kind of did what it did. And I mean, is that partly related to and I know you’re experiencing describing or skills is more about kind of playing together and ensemble and that sort of thing. But I don’t know what the question is. But I’m very curious about this, this aural skills class, and maybe how that relates to this idea of anticipatory auditory imagery. And maybe you can tell us a little bit more about that.

Sure, I can say something about both those cases, how those concepts or that concept came about, and another related one, because, in fact, it’s a good point. Most of the research topics that I’ve ended up focusing on arose through insights that came kind of through struggle, things that I was struggling with, as a performer, of course, things that came naturally, if anything did their back of mind, then they don’t rise to the level of consciousness. And of course, one could think about things theoretically, become an armchair philosopher and think about problems and discuss them. But I was kind of, although my research is not typically seen as applied in terms of solving a real world problem, it is applied in the sense that it arose through thinking about things that I was struggling with, or finding problematic, and most of them are resolvable. But the interesting thing is the way to resolve it and the ways in which the resolution might differ for different people, depending on their experiences, capacities, and other predilections.

So in terms of how the concept of anticipatory auditory imagery came about, now, that’s an interesting division that I saw in musical training and pedagogy at the time that most of it was focused on training performance skills as a soloist because they I was preparing you for auditions. Funnily enough auditions to play in ensembles with others. But the training for playing in ensembles, though there was such training, but it wasn’t as systematic. You know, each particular lecture might have a particular idea about ensemble performance and talk about it in a different way, which is an interesting fact in itself that they could all play together well with different perspectives. But I remember the ensemble class, doing sectional work, practicing excerpts together, it was kind of something we did occasionally, it wasn’t something where there was a particular professor in charge of that that was part of the curriculum, maybe things have changed, I haven’t really followed up with formal ways. And it does vary with institutions. But at the time that I was studying that wasn’t common, whereas solo performance goes through is more codified, there was more consensus about how to teach it with some variation, of course. So the anticipatory auditory imagery came more from the idea of striving to produce an ideal sound as a performer, myself, and this is something that the trombone teacher was very big on talking about, maybe it’s part of the Chicago School and the Chicago conceptualization. I think it’s more broad than that, of course, but it was something that was very much front of mind. So the idea was that yes, it’s no point really picking up the horn and playing a sound if you don’t actually have a clear idea of what that sound should be. So that was really the genesis of that idea whether the term whether that originated in my research, I don’t know, I’m cautious about making any claims about that. Because one often finds then that if you read back in the literature, you’ll find that the ancient Greeks were talking about it or something I don’t know. But certainly, the concept wasn’t being discussed in scientific terms quite as much with music.

But apart from an interesting side note, that William James, one of the fathers of experimental psychology, in his book on the Principles of Psychology, does talk in his chapter on free will about the singer imagining that the singer does not need to think about, you know, the muscles involved in producing a particular sound, she just needs to imagine the ideal sound, and it will happen, of course, with the appropriate training and so on. So actually, William James, in his so called idiomotor principle, hit the nail on the head. And that’s exactly what anticipatory auditory imagery is. It specifies its role in producing one’s own performance. But then, as you mentioned, this oral skills course, where we were playing together, kind of, I don’t know how explicitly, I realized that at the time, that it did bring awareness that one has to and we know that people are taught to listen to other performers. But listening, in a sense, can be thought of, if it’s not explained in more detail, as something that’s passive and reactive. We wait for something to happen, it happens, we pay attention to it, then and listen. The idea is that the most efficient way is to actually anticipate to listen to something before it actually occurs. And that requires mental imagery. So that’s the idea of anticipatory auditory imagery, using it to plan your own actions to imagine your ideal sound. And then you’ll be more likely to produce it than if you don’t have that vivid image. And then also to predict co-performers’ sounds, of course, we have limited capacities to do that. And there are interesting questions about how many different things we can attend to or how many different parts we can imagine, I think one would expect that perhaps individuals such as conductors who deal with who are able to hear a high score have special capacities in that regard. And I think there’s some evidence for that.

And the other aspect that came to the fore in the aural skills course was, and I’ve touched on this just a moment ago, when I when I referred to a tension, that really another big challenge was dividing a tension between what you’re doing as an individual performer, and what your co performers are doing. And this will vary with the level of complexity of what you’re producing. And in our course, we were mainly focusing on rhythmic complexity of the relationship between parts. And that actually led to another maybe I do have a penchant for coining odd terms, because my PhD thesis was on concept which are referred to as prioritized integrative attending. And that was simply the case that when you are in a complex situation, a lot of people doing things and you’re doing it together, collectively performing, you need to, in a sense, divide your attention between what you’re doing and what others are doing, but also monitor the overall output if the idea is to play in tune and in time, obviously. So however, the whole thing as a group will not work if each individual doesn’t play their part correctly, or in the way that’s agreed upon. So in that sense, attention is prioritized. One’s own part is highest priority. Of course, with a lot of practice, various aspects of it become automated, but usually those are the technical aspects and the expressive aspects one still would like to pay attention to especially if one wants to introduce spontaneous sounding variations to make the performance exciting to make it sound improvised, even if it’s not. So the technical aspects might be automated, but the expressive aspects can benefit from continued attention. And in a group, however, this involves, of course, paying attention to others as well and to the overall output. So that’s something that I found another challenging aspects, that’s not something that came naturally to me at least I’m not sure what other people’s experiences are. So those were two separate concepts that came out of that early music training.

The third concept that I’ve become interested in and these three collectively form a core of the theoretical framework I work with, which I call the ensemble skills framework is what I call adaptation. And that’s the reactive component, the attentional component is spreading one’s monitoring capacities across different sources of information. The anticipatory auditory imagery, or just anticipation capacity is using one’s own sensory motor system to anticipate to plan one’s own actions and predict what others are doing. But of course, one is reacting to things that aren’t planned, for example, that arise spontaneously or even God forbid errors that occur interpersonal timing errors, or tempo errors, any sort of an improvised music, just unforeseen unexpected events that somebody introduces, that are thrilling but need to be dealt with. They can’t be through so thrilling that everyone puts down the instruments and just revel in the in the thrill of it, the show goes on. So reactive processes come in and I’m mainly interested in timing of more recently focused on other aspects. Tembre blend, for example, in choral singers is another topic, that’s very interesting. But most of the work I’m interested is in timing, because if you’re not in time, and not in tune, whatever else you do is maybe not going to be as effective. So timing is very, very fundamental. It’s very obvious if it’s out. And so reactive error correction mechanisms for temporal adaptation.

So these three concepts have genesis in separate experiences that I had, but actually turned out to be very much related. And at the moment, what I’m interested in is how the brain deals with them, how they together work together to determine the quality of coordination. And also how they’re influenced then by by other things that are undeniably important, higher cognitive functions, knowledge, so familiarity with the parts of co-performers, but also their idiosyncratic way of playing. Because in fact, the tension or competition between different performance goals and intentions is what makes listening to an ensemble performance, like watching a thrilling drama perhaps where there’s tension and resolution, not just at the level of, for example, tonal harmony, or cadence resolving and so on, but actually at the way that people are interacting at the 10s of millisecond timescale as we do when we’re walking down the street and see interactions that heated or exciting in some way. So those factors are all couched in a social context. So then another thing that came to the fore, eventually, is that these processes of attention and anticipation and adaptation are in some way related to our personalities, we have predispositions for reacting to others in certain ways to certain degrees, some people may find it easier than others to anticipate what others are doing to pay attention to others, to divide attention. And this is not something about just linked to concepts such as fluid intelligence, and so on. This is something more fundamental and more social, in a way. And for that reason, the ensemble scores framework that I’ve been working with, I kind of think of as generalizing, in fact, using the same sensory perceptual motor, emotional machinery and capacities that we have in the brain for navigating our social lives more generally. So I call this the idea of music as a microcosm of social interaction. And it’s probably one could even go so far as to say that, to understand it, it’s safest to assume that that’s what humans evolved to do. That was a major factor driving evolution. That’s something a contention that’s difficult, if not impossible to prove, and people will argue about it for a long time. But I think it’s a useful assumption to make as a basis for guiding a research program.

I really liked the ensemble skills framework or this theoretical model that you put together, because I think, typically we think of good ensemble players. It’s kind of a vague, abstract concept. And it’s hard to know exactly what are the skills and I don’t know if we’re at the point yet where we can assess or train these specific skills or how effectively that can be done. But I do like how it makes things more concrete, and I’ll try to post a picture of it somewhere but the three core factors being anticipation, attention and adaptations, you described and, and then of course with these other factors, if I understand it correctly, like understanding your co-performers, your ensemble members’ preferences or playing styles and how they tend to produce things, who’s in charge of leading in different sections, and so forth. And, and then I’d really love to kind of ask you about the social factors or the personality factors, because I think I don’t know if it was yesterday, or some other lecture where I heard you talk about this empathy study and tempo prediction, which I just found really quite fascinating. But this general kind of human skill of empathy would be related to whether you’re able to play more effectively with people or not.

Yeah, that, in fact, surprised me also, in terms of the robustness of the relationship, we’ve seen it time and time again. And the assessment tool for assessing empathy that we use has four different sub scales, and there’s only one of those that seems to be related. This tool has four sub scales relate to two cognitive dimensions of empathy, and two more emotional or affective dimensions. And amongst the cognitive ones, perspective taking is one of them. The other one kind of oddly is termed fantasy, but that refers to how you kind of identify with characters when you’re reading a novel how much you kind of firm. So not just see somebody else perspective, but see it really identify with a character in a work of fiction, and the others, the effective components are personal distress and empathic concern. So you can guess from the title is what they refer to. And it’s only perspective taking, that seems to be correlated with the ability to predict timing of another performer, or even a computer generated sequences varying in tempo. Although it turns out that perhaps the reason for that relationship is actually not so surprising. And that’s because they both seem to rely or involve similar neural substrate. So parts of the brain’s motor system or system that is used to control actions, so not just not just to move not just primary motor areas, but a more distributed system in the brain that controls goal directed actions to produce certain effects, certain sensory effects. And this system can be used, we use it when we’re imagining performing an instrument. And we’re not actually overtly moving necessarily, or certainly not with the instrument. But if we’re experienced, we can imagine all the component processes of doing that. And if we scan the brain, while individuals are doing that, the level of activity looks very similar to when actually performing. This depends on experience, the more experienced one has, connections are formed, and one detects that to a stronger degree. So it’s the same system. And when we’re interacting with others, and taking their perspective, we’re in fact, simulating their actions. And that’s one of the hypotheses about how we have this ability, how we have this capacity, we’re able to simulate another’s action. So when we observe somebody doing something, we covertly use our own motor system to simulate that. And that allows us to better understand their actions and what their goals might be, and to predict the future time costs of their actions. And that’s what’s crucial in coordinating with them.

So the idea is that, yes, we use particular brain network when we’re empathizing with people in daily life. And we use that same network when we’re imagining or internally simulating musical actions of others. And interesting questions arise, the effects of whether we play the same instrument or not, okay, as a trombone is how well can I simulate a pianist or violinist actions, certainly not down to the level of fine motor skills, and even demonstrating fingering of a violin with the wrong hand. And it makes no difference to me whether I use the left or right hand, because I don’t have this direct experience. But however, our bodies apart from that are the same apart from those skills. And when we perform as you know, a lot of information is carried through the visual modality or the music is typically considered as an auditory art form, especially since the advent of recording audio recordings, but really live performance and the visual component we know, especially if that’s been denied from us is really compelling, and add something so visual information is important. And so therefore, I can covertly simulate the body motion of another performer, even if I can’t play their instrument and simulate the details of the fine movements involved in in triggering natural sounds. And that seems to be enough because those body movements we know from other research, yoked are kind of lawfully related to phrase structure. Okay, they might not be the body is not moving at the level of, you know, 16th notes in at a fast tempo, but it’s moving at the level of phrasing. And that’s important for coordination. And coordination occurs at multiple hierarchical levels. We want to coordinate the notes that are going on, but of course, we also want to coordinate things at a higher level at the level of phrasing and that’s really where group musicality can come out, having coordination at multiple levels simultaneously.

Another interesting aspect that comes out is that what is perfect coordination. And often the assumption in research is when you measure people doing a task coordinating, they’re playing on MIDI pianos or they’re tapping a key or drumming together. The assumption is that the task is to achieve zero asynchrony, zero milliseconds asynchrony, zero 1000s of a second between performers. But of course, if you generate music like that, it’s not necessarily the best sounding performance. Some variation is important. It tells us something it creates that tension. It also allows certain instruments to emerge from the texture and be perceptually salient. So sounds that occur slightly earlier. It’s a phenomenon referred to as melody lead, it’s something that is very finely calibrated, it’s in the order of 20 milliseconds or 20 thousandths of a second. Other style of music relies on systematic differences between on sets of instruments. For example, in groove based music, there’s a concept referred to as the Why beat sound a lot of funk type music where the bass and bass drum play with a consistent lag of 30 milliseconds around so a bit larger than what you’d expect with melody lead. And then at certain structural points in phrases, instrumentalists will, it’ll sound best when they’re playing sometimes up to 90, 100 milliseconds out of time with one another. This depends on the particular instruments, for example, the rise time stringed instruments, wind instruments, where the sound onset takes a while to ramp up while being very short amount of time. But I’m comparing this to a percussion instrument, even a piano, of course, drums and other percussion as well, where perhaps the tolerance for synchrony in the optimal levels can vary.

I’m curious about something you said a little bit ago about body movement or body sway. And I had a teacher Don Weilerstein, who would have me conducting lessons on occasion to try to internalize physically the pulse or the direction of the phrase and Catherine Cho, a violinist at Juilliard takes it maybe even one further where she used to audiate in her head, the concerto that she was playing, or working on, while she would walk around her yard or neighborhood to try to more holy with her body internalize a sense of pulse. And I wonder if those are related at all to what you found in your research. But even you know, this idea of I’m sure, every instrumentalist has been guilty of throwing their pianist off at points because they made it seem like they were going to end up at a certain place at a certain time. But then, you know, you rush or you drag or something funky happens and times not actually communicated effectively in the way it’s going to be. And, and I wonder if there’s something that ties all these things together, perhaps that’s related to what you found in your research?

Yes, there is very, very much so. So what do you refer to making the pulse overt and walking is a great example, walking by imagining because it regulates the timing. Because if you don’t, you when walking, you know, you’ll fall over or won’t be able to avoid obstacles, effectively, posture, and the way one moves is important. It regulates timing. It’s interesting, the mechanisms by which that might occur. There’s some evidence that it might be because it stimulates the vestibular system. So the system that controls posture and balance, and it’s tightly linked and plays a important role in rhythm perception work going back decades by a collaborator of mine, Neil Todd has demonstrated this, and he’s continuing this really fascinating work on the role of the vestibular system, in rhythm, perception and production. So it’s not just an auditory sense. But vestibular.

This is one of the potential explanations for why dance music is most effective when played loud, because the vestibular threshold is around. It varies in individuals, but usually between 90-95 decibels. And you know, you’ll find people complaining about loud music in dance clubs, when people were still going to dance clubs, which can, it’s usually to be effective and enjoyable exceeds that level. And the idea is that it really gets you moving, because it’s stimulating to the vestibular system, which gives you the sense that you’re moving and empowers you to actually move into enjoy it, it’s pleasurable, it’s linked to the reward system in the brain, as well. So just to go back to your, the point that you raised, I think movement needs to be regulated, navigate through the environment, when you’re conducting, you actually have this nice, effective redundancy, because you’re not just imagining the sound, but you’re making a movement, a movement that is fairly natural to make, and therefore you have visual input, proprioceptive input, you’re actually enhancing the sensory information that you have about the timing of what you’re imagining or planning to produce. So making it overt actually constrains it. And the word constraint, I mean, in a good way, because one of the problems with human action control is or one of the things that traditionally thought to be a problem is the immense degrees of freedom. We have the ways our bodies constructed. We can do things in so many different ways. And that’s an advantage of course, it has been an advantage in evolution for You can also think of it computationally as a disadvantage, because there’s no time, despite the brain being very rapid and powerful as a computing device, it’s unable to compute joint angles of all the joints in the body required to produce a particular action. So if you use a motor program that is, falls naturally out of the way the body is constructed and conducting movements fit within this description, you can then use that as a scaffold to upon which to hang all sorts of complex musical thoughts

And this is not maybe 100% related to that, but it seems that there’s an awful lot of useful things that we could be thinking about when we’re performing. And it depends on the situation and so forth. I liked how earlier you separated how a lot of the technical and mechanical movements associated with playing your instrument, hopefully, by the time you’re on stage, largely operate out of conscious awareness, you don’t have to think about those things. And of course, it goes into the whole research area on choking and so forth. But I also liked how you made it very clear that you know, there’s this other part of your performance that very much should be in your conscious awareness while performing that’s expressive part. And so I wonder if say we were all psychic, and could read our favorite musicians minds while they were playing, I’m wondering if you could take a guess as to like, what sorts of things would we be able to read in their minds as they’re, whether it’s in a quartet or whether it’s in a full orchestra performance? Like what sorts of things should we be aspiring to be focused on when it comes to thinking about expressive moments?

Well, I think the reality and the ideal probably differ quite a bit. And we might even be shocked at what performers are really thinking about in some cases. And I say this, having interviewed some who I consider to be elite, ensemble performers, I haven’t yet published all of this work. There’s snippets of it in some works that I’ve written where I’ve included quotes from, from members of Chicago Symphony, and even a lot more chamber musicians, including the Labèque sisters, and nobody said anything really shocking, I was a little bit joking about that. But you know, there was a lot of variation in the way people approach the task of performing as ensemble musicians, even at the level of the goals, the things that they think are most important. In fact, I’ve had situations where I’ve interviewed members of the same ensemble, and the interview process has opened up situations that they weren’t even aware of where they have different ideas, and as a result, have many arguments about certain things.

So and even if not arguing one interesting thing that sticks to my mind, a string quartet that I interviewed, a very high level string quartet, one member said, It’s absolutely vital to think of the string quartet as a single instrument with 16 strings. You’ve probably heard people say this, and it does make sense in terms of emphasizing the cohesiveness working as a single unit, single organism, if you like. And yet another member of the same quartet independently, so interviewed on separate days without talking about it had the other view that they did, what I think is exciting is that is four separate instruments. And of course, they’re working together, but didn’t seem to see much value in thinking of it or necessity, thinking of it as a single instrument with 16 strings. So rejected that idea. So that was interesting. And this particular quartet has been at the top of the international scene for more than 15 or even 20 years by now. So that sort of thing is kind of interesting.

And that’s why I really think I link it to social interaction, ensemble performance that is, because it doesn’t have to be all this cult-like conformity and single mindedness, the conflicts, many conflicts that can arise can be what makes it exciting. Conflict might not be the right word, because it has all sorts of connotations. But I think it is a minut level, they are exciting conflicts that have to be resolved. And that’s part of we know from research on musical meaning and emotional communication, that what’s exciting and what’s rewarding and pleasurable in listening to music is when there’s a certain degree of tension or expectancy is violated to some degree. But then the resolution is important at some point, or at least, aesthetically, it’s certainly valuable. To challenge in understanding this if one wants to, for example, teach it or break it down in a certain way or explain to people, the people’s tastes in music might vary. When I say tasting music, that’s obvious, but I actually mean what they like in terms of how an ensemble is performing. And that’s because I think it is related to one’s own way of interacting with others. So when we go back to the ensemble scores framework, how adaptive I am, how my capacities for anticipating, my capacities related to attention will in some way, be linked to what I appreciate when I’m observing somebody else interacting. So it might want to be in some cases similar to what I’m doing, some cases different. I was talking about, I remember now that talking about the degree of conflict, and the degree of conflict and resolution that we appreciate, there might be some cultural constraints on that what’s fashionable or acceptable in a particular time point in history or a particular place. But there are also individual differences in what we appreciate that way. So concepts such as something being aesthetically pleasing is very much linked to psychology, individual psychology, but I think not just at a conceptual level of meaning, and so on, but really down to a sensorimotor level. I think it can be boiled down, I’m not advocating some sort of reductionist approach where that’s all we need to focus on, but I think it needs to be considered to have a full understanding of not just how we perform music, how we teach it, but also how we appreciate it. And ultimately, why we do it as humans.

And I don’t know if this is related, but I was talking to Merry Peckham who’s one of the founding members of the Cavani quartet here in the States. And she talked about how over the course of several decades of performing together, she could recall maybe just a handful of times where all four of the members of the quartet agreed that a performance had gone well. Where usually like two or three would have thought that it went well but then one thought it wasn’t or some of them thought it went poorly and that made me wonder if if there’s something related to this?

Yeah, I mean, I don’t know exactly what the situation was that you’re describing. But I’ve certainly heard performance say similar things. I think it once again, underscores how this is related to things going on outside the musical sphere, because people’s moods might influence how they perceive things. And it also depends on what the expectations where, for example, some people may be more inclined to think that it’s a good thing, if the performance deviates in certain ways from the rehearsals, but other performers may find that unsettling and prefer that co-performers stick to what was agreed upon in rehearsals, and certainly the dress rehearsal, say, whereas others might be more, we could tie this to, for example, risk taking. Some people may appreciate risk taking more than others. So this could influence how people feel things go, also what they’re paying attention to. Maybe also the roles that they have, there may be particular programs where one performer feels that they are a little bit more out on a limb, for whatever reason, so that could affect how well they think something goes.

I think it’s an interesting question. from an educational perspective. You know, there’s, I think, quite some years ago, there was interest in whether certain individuals are predisposed to play certain instruments better like, is there a certain personality type better suited to the flute? And is that different from somebody better suited to play the trombone or the drum kit or whatever? And, you know, that’s an interesting question, it gets into the idea of stereotypes as well, which of course, can be misapplied to negative effect. The way I think about it, when I’m linking up concepts such as personality, empathy, we discussed before with anticipation capacities, I don’t think of these as hard constraints, I think of it as useful to know that our ability to predict tempo changes seems to use co-opt brain networks that we also use when you’re enjoying interacting with others to understand what they’re doing and why they’re doing it. So we have this link. So it means perhaps, that we know that somebody if they’re playing a role in an ensemble, where, for example, they’re if they’re designated as a follower, the accompanying, they will find it easier to predict the co-performer, if they are high in this capacity in empathy. If they’re not, it doesn’t mean that they should not play their role, it just means that there’s a targeted training regime in the education system. It should help boost their performance and that skill, whereas their ability to adapt and react might be actually relatively high compared to others, so they won’t have to work as hard on that. So I think we all have to work hard to achieve excellence, but it probably pays to apply that work in a targeted way that addresses things that we may have difficulty with. And we will all have difficulty with different things. And this is linked to our way, our brains are anatomically structured through experience. And also, of course, the way things developed from genetic blueprints, but really experience shapes. This is undeniable, this is one one big thing that we’ve learned, but also the way our psychological makeup is shaped by the interactions that we have through life. And these are not hard constraints. They are just weak predispositions that it would be useful to take into account when shaping one’s training,

Which is kind of a nice way bring it back to I think what you started talking about In your origin story, which is your experience in your educational process led you to identify these questions that you had. And so I have this question that I have sort of a secret hidden agenda for asking, and I’ll tell you what it is in a second, but I don’t want to like influence your answer by telling you what the context is. And so if these ensembles skills are important for one’s ability to perform effectively in ensembles of varying sizes, any suggestions or ways that one could measure these ensemble skills somewhat effectively?

Yeah, for timing, we have actually a battery of tasks, it’s maybe sounds nicer to call it a suite of tasks. But in psychology, we call them task batteries. A collection of tasks that can be applied, ideally, takes about an hour, but I think, a short version in about 20 minutes and assesses based on a mathematical modeling technique. And we ask people to tap along with different pacing sequences that are supposed to capture certain timing demands that would arise in an ensemble situation. And then we can analyze the timing of the responses, and get estimates of how much somebody is adapting how much they’re anticipating, and how they’re integrating their own actions with those of the external partner, which is in this case, is a computer driven auditory sequence. And this gives us estimates of people’s ensemble skills. Now, as I said, the important aspect is this is just about timing. And that’s the most basic requirement. And there’s many other things on top of that, but I think it’s a start, and I think that could be maybe extended. It is a great interest of mine to actually try to apply this in educational contexts for both assessing pre and post training. I would not advocate using this as a selection tool for entry or anything like that. It’s really more just to identify where people might need to devote a bit more extra effort to develop their skills as an ensemble musician.

Right, yeah. And so it’s good that you mentioned that at the end, because my question was coming out of two parts, one, you know, the pedagogical implications of the work that you do in terms of how can we identify young musicians who might benefit from more of this kind of experience, or that kind of training and so forth. But the other was, I’m wondering, this somehow reminds me of the NFL, the National Football League in the United States, and this combine that they have where they measure everyone’s agility and vertical leap, and they have this really long, complicated paper and pen kind of assessment to find out personality factors, and so forth. And it’s always very unclear as to how much of this is actually predictive of who’s going to be successful at the professional level. So my question about selection was in orchestral auditions are set up differently in different countries. But I wonder if the way that they’re set up now, I mean, our orchestra is maybe not getting enough data about a particular candidate’s fit or ensemble playing abilities to be able to select members who would really mesh well, within that ensemble.

I’m not an expert on the practices of all the orchestras and that, of course, I’ve talked to some musicians about the audition process, and having prepared for that sort of thing myself many years ago, have have some idea. In my understanding, it varies there, it depends on the orchestras and their resources and how, what their programs are like, you know. I’ve heard things range from, you know, having a very brief audition and that’s it. There’s a trial period, usually, but whether that trial period is really how long that trial period is, and what it entails, really varies. And even the audition process, whether it involves a section audition, where you perform just with the section, so I was a trombone player, as I mentioned, so with the lower brass section, or if it’s just a solo audition that varies, then the, as I mentioned, the length of the trial, the weight carried by the, for example, the musical director versus the actual section members and the rest of the orchestra, whether to offer continuing appointment. So it varies so much. And there’s probably not enough data because a lot of this information, of course, is confidential. So it’s not the case that one can collect the data and see the relationship between people scores in a solo audition and how long they actually last, whether it’s a full career or not, and what the reasons were for not lasting if unfortunately, that wasn’t the case. It’s a tricky issue because of confidentiality and privacy concerns. I think it would be, of course, a lot of people would want to know the answer to that question. But the prevailing issue would be that I think it really varies, which country you’re in and what level orchestra it is, how busy they are, how quickly they need to appoint somebody, and in fact the role whether it’s a role where there’s just one instrument versus a sectional arrangement.

And I didn’t mean to put you on the spot in terms of affecting in any way, the future direction of how orchestral auditions are run. But I’m just curious if you had your research hat on and we’re thinking, oh, you know, if I’m in charge of this orchestra, and I really want to get people together who are a good fit for one another, like, what quirky unusual thing that isn’t being looked at right now would I want to try to inject into the process to see if I could find people who would mesh well together?

Well, I would, I think it wouldn’t hurt put it that way to assess ensemble skills. And in fact, this is something that I’d be very interested to do across orchestras of all standing, across different sections, because I’m expecting what’s important is complementarity. You don’t want everybody, there’s no such thing as this is the ideal combination of reactive adaptive capacity, because you actually need a situation where, where some individuals are good reacting, and some are good at anticipating. And perhaps and this is a hypothesis to be tested that might depend on instrument and role within a section with a section leader or not. But it’s just that it’s a hypothesis. It could be the case that it’s not as important as other things and explains a very small amount of variance in what’s going on. So it’s really negligible and one shouldn’t worry about it, invest much in it or make decisions based on it. We really don’t know. My hunch would be that it is certainly worth investigating. Because I think based on my own experiences, it would have, I think been helpful to me, at least in understanding what’s going on. It maybe would have allayed some frustrations and that sort of thing.

Transcribed by


  • Here’s a link to Peter’s 2012 paper on anticipatory auditory imagery that I alluded to at the beginning of the episode. 

More Peter?

If you’d like to explore more of Peter’s research, below are a few handy links. 

You can check out his dissertation or the Janacek thesis that he mentioned in today’s episode, or even listen to a few of his musical compositions at his website:

Or follow his Twitter feed at:

Or reach out to him via:

Header photo credit: Wissenschaftskolleg

Ack! After Countless Hours of Practice...
Why Are Performances Still So Hit or Miss?

For most of my life, I assumed that it was because I wasn’t practicing enough. And that eventually, if I performed enough, the nerves would just go away and everything would take care of itself.

But in the same way that “practice, practice, practice” wasn’t the answer, “perform, perform, perform” wasn’t the answer either. In fact, simply performing more, without the tools to facilitate more positive performance experiences, just led to more negative performance experiences!

Eventually, I discovered that elite athletes are successful in shrinking this gap between practice and performance, because their training looks fundamentally different. In that it includes specialized mental and physical practice strategies that are oriented around the retrieval of skills under pressure.

It was a very different approach to practice, that not only made performing a more positive experience, but practicing a more enjoyable experience too (which I certainly didn’t expect!).

If you’ve been wanting to perform more consistently and get more out of your daily practice, I’d love to share these research-based skills and strategies that can help you beat nerves and play more like yourself when it counts.

Click below to learn more about Beyond Practicing, and start enjoying more satisfying practice days that also transfer to the stage.

BOGO pricing is now in effect! (through 11:59pm Sunday)

Sign up anytime now through Sunday (Dec. 4) at 11:59pm Pacific, and you’ll receive a second bonus Beyond Practicing account – at no additional cost – that you can gift to a friend, colleague, family member, student, or teacher (i.e. a practice buddy to explore the course with 😁).

Click the red button below to learn more about the course and get the holiday buy-one-get-one-free offer.


Leave a Reply

Your email address will not be published. Required fields are marked *


Join 48,000+ musicians!

Get the latest research-based tips to level up in the practice room and on stage, from one week to the next.

You'll also receive other insider resources like the weekly newsletter and a special 6-day series on essential research-based practice strategies that will help you get more out of your daily practice and perform more optimally on stage. (You can unsubscribe anytime.)

Download a

PDF version

Enter your email below to download this article as a PDF

Click the link below to convert this article to a PDF and download to your device.

Download a

PDF version

All set!

Discover your mental strengths and weaknesses

If performances have been frustratingly inconsistent, try the 3-min Mental Skills Audit. It won't tell you what Harry Potter character you are, but it will point you in the direction of some new practice methods that could help you level up in the practice room and on stage.

Hello. Add your message here. Learn more
Beat nerves with a friend, with BOGO pricing on the Beyond Practicing course Join Today