What is music cognition?

In the last three years our group has spend quite some energy in promoting music cognition as an interesting field of research in the cognitive sciences. The strategy was simple but effective: simply say ‘yes’ —and think along with— any journalist that contacts you. And, as is often the case in media land, once an idea is out and considered interesting, other media want more of the same. The challenge is, however, to make sure music cognition —its insights and results, its aims and prospects— is represented in an appropriate way, without falling into the trap of being reduced to simple facts that are useful for a popular TV quiz (see video below).

In that sense, I sympathize with initiatives like the ‘Battle of the Universities’ that promote the idea that scientists themselves should to take a lead in presenting their research (instead of ‘complaining’ about the media simplifying it too much :-). However, it is not easy to bring forward the essence of one’s field in an intriguing way.



Outreach —as it is often called— is not the same as ‘going on your knees’ to explain your research to a general audience or making populist interpretations of your field. You are in fact challenged to explain your research and insights in different terms. And that can be very rewarding and even influence to your own thinking. With regard to my own research, I could start talking about the computational modeling of music cognition, and the theoretical, empirical and computational methods that we use, but I’m sure a general audience will quickly loose me. A common trick is to think of a typical example that speaks to everyone’s imagination. I often explain my research in terms of the scientific challenge to make a listening machine. Imagine what that would be like? A machine that can listen and react in a human and musical way. And, of course, it should make the same mistakes! It allows you to explain all kinds of computational modeling notions, what should such a machine know, what should it listen for, and how can we compare and evaluate them? We just heard from our university that MCG has been selected to defend the University of Amsterdam in the Battle of the Universities. A challenge to look forward to ...

What makes a metaphor informative?

Metaphor. When I read that word I always hear the voice of Massimo Troisi (Il Postino) saying ‘Metáfore’. And indeed, like in that movie, metaphor can be mesmerizing and beautiful. However, in music research metaphor has had a debatable role. Metaphors like ‘music is movement’ —it makes you move—, ‘music is a language’, ‘music is distilled emotion’ etc. are often reducing what music is, instead of contributing to a real understanding. While crucial in the arts, metaphor is often all but informative in research.

However, the research of Zohar Eitan (Tel Aviv University) is one of the important exceptions. Instead of taking the ‘music is abstract motion’ metaphor as an explanation of how phenomena in music are constrained —governed by, for instance, the rules of elementary mechanics—, his group designed a nice set of experiments in which participants were asked to imagine a cartoon character while listening to music.

In these listening studies participants had to report when or how the imagined cartoon character was moving in response to the music. Instead of using the physical motion metaphor as an explanation, the association of listeners with physical space and bodily motion was used to reveal how music can influence mental images of motion. Interestingly, it turned out that most musical-spatial analogies are quite asymmetrical. As such providing evidence that, while music and the motion metaphor can influence each other, the latter can not fully capture the actual phenomena.

Eitan, Z., Granot, R.Y. (2006). How Music Moves. Music Perception, 23(3), 221-248. DOI: 10.1525/mp.2006.23.3.221

Is music a luxury?

Could scientific books on music until recently not afford not to include a chapter on the anatomy of the ear, currently most books on music have to include a chapter on the anatomy of the brain. The brain is 'cool, hip and happening'.

Luckily Oliver Sacks' new book Musicophilia: Tales of Music and the Brain does not include this obligatory chapter. Instead, he uses his familiar observational style in revealing his personal and medical interest in people with a particular brain disorder. In this book he focuses on music-related mental phenomena ranging from amusia and absolute pitch to dysharmonia and synesthesia, while also discussing the role of music in Parkinson’s disease, Tourette’s and Williams syndrome. It is a book that fits in a trend of books by scholars that analyze and promote the importance of music in a wider perspective than is normally done by musical experts. Like Steven Mithen’s The Singing Neatherthals and Dan Levitin’s This is your brain on music, this book discusses what is so special about music, while it seriously wonders why some consider it a mere luxury (if not simply cheesecake).

In his book Sacks describes a series of medical cases where a neural deficit reveals something about the workings (or breaking down) of an intrinsic human quality we name ‘music’:

“We humans are a musical species no less than a linguistic one. This takes many different forms. All of us (with very few exceptions) can perceive music, perceive tones, timbre, pitch intervals, melodic contours, harmony, and (perhaps most elementally) rhythm. We integrate all of these and “construct” music in our minds using many different parts of the brain. And to this largely unconscious structural appreciation of music is added an often intense and profound emotional reaction to music. “The inexpressible depth of music,” Schopenhauer wrote, “so easy to understand and yet so inexplicable, is due to the fact that it reproduces all the emotions of our innermost being, but entirely without reality and remote from its pain.... Music expresses only the quintessence of life and of its events, never these themselves.”

The book is a hopeful account of the role music can play in our lives and how our brain is involved in this. It will be released this November in many languages at the same time (I just read a preprint). So this will likely not be the last bit you read about it.

What makes a theory of music surprising?

Quite a while ago, a fellow musicologist referred to me as a ‘positivist’. As I was, at that time, not too familiar with postmodern parlance, I thought of it as a compliment (making an association with the Dutch comedy duo De Positivo’s that were sheer optimistic). It turned out that actually the opposite was meant.

Last weekend in Cologne, being invited to speak at the Gesellschaft für Musikforschung, I was reminded of this remark. For some reason the methods associated with positivism, such as those used in the natural and social sciences, still flag a divide in music research between, for instance, the systematic and historically oriented approaches to music. A divide that seems to be fed by a misunderstanding of much of Popper’s ideas on ‘science’ versus ‘pseudoscience’ (see earlier blog).

While the idea of ‘falsification’ is indeed, as Popper showed in some of his later work, not very useful in historical research —as in archeology, a new found manuscript can easily falsify a long established historical theory— this does not make archeology or historical musicology a 'pseudoscience'. In my opinion, it is not so much the inapplicability of the empirical method (and hence the possibility of falsification) for historically oriented musicology, but the apparent resistance to formulate theories that can be tested, that might characterize the discussion. Is it impossible to make a theory about some aspect of (the history of) music that can be tested (or evaluated), independent of empirical evidence?

I like, particularly in this context, Popper’s idea that a theory can be intrinsically compelling or ‘surprising’, even in the absence of empirical evidence. What is intended here is not ‘surprising’ in the sense that a new fact is found that we did not yet knew about, but a prediction that, while we would expect X —given everything we know—, it actually predicts X is not the case, but rather Y. A prediction that is the consequence of a theory (made up of intuition, empirical observations or otherwise) that is violating our expectations based on what we know. I do not see why both historical and systematic musicology could use that as an additional method.

Did you hear something new this week?

Sharing musical taste, or, even stronger, luring others into your favorite music, is a common activity at school and birthday parties. Nevertheless, the most common way we get to know about new or different music is probably through reviews by experts on concerts and recently released CDs. New in that sense is the role that the Internet starts to play in all this.

In the last few years it became possible to share music and musical taste through Internet communities on an much larger and more personal, one-to-one scale. This is a growing —if not booming— social and cultural phenomenon, that is clearly changing the way communities of music lovers ‘lure’ each other into other musical domains and modes of listening. Next to being facilitated by recent digital recording and distribution techniques, it is mainly the partly unforeseen impact of user-generated content (‘Web 2.0’) that contributes to its success. It seems a development that is worth studying for music, media and music cognition researchers (At the University of Amsterdam a consortium is currently thinking in that direction).

P.S. Did I hear something new last week? Well, yes. Last Monday (in the Bimhuis) I heard for the first time a begana (a 10 string lira), an instrument I saw, around age 8, on the first page of a history of musical instruments book but that I actually never heard. It didn’t sound at all the way I imagined it (listen).

Do you have amusia?

Some people doubt whether they have a sense for musical pitch? However being tone deaf is a relatively rare phenomenon that is studied by neuroscientists (who refer to it as amusia) because it might give us clues about the specificity of music. This week, in exchange for a short blog, a reference to an amusia test made available by the University of Montreal.

Are there four beats or only three?

You are browsing, let us imagine, in a music shop, and come across a box of faded pianola rolls. One of them bears an illegible title, and you unroll the first foot or two, to see if you can recognize the work from the pattern of holes in the paper. Are there four beats in the bar, or only three? Does the piece begin on the tonic, or some other note? Eventually you decide that the only way of finding out is to buy the roll, take it home, and play it on the pianola. Within seconds your ears have told you what your eyes were quite unable to make out — that you are now the proud possessor of a piano arrangement of “Colonel Bogey”.

This is the opening paragraph of an article that was printed in 1979 in the Proceedings of the Royal Society of London. It came out as part of the book Mental Processes: Studies in Cognitive Science by H. Christopher Longuet-Higgins in 1987.

The chapter on music —of which the citation above is part— made a lasting experience on me, and actually made me decide that music cognition is worth dedicating all of one's research to. It made me realize that all those things musicologists and music theorists considered mere axiom’s —such as a meter, an upbeat, or a syncopation— were extremely interesting in themselves, and could be studied using methods from this developing field called ‘cognitive science’.

It is now precisely twenty years ago since H. Christopher Longuet-Higgins’ book was published: An impressive collection of papers with topics ranging from music and language to vision and memory.

It also includes his comments to the Lighthill Report, published in 1973, in which he proposed ‘Cognitive Science’ as a label for what he saw, then, as an emerging interdisciplinary field.

Unfortunately, you have to go to the library to read it: it has been out of print for quite a while.

(See also a text in Dutch —Nieuwsbrief 102— on the same topic.)

A sense for rhythm? (Part 3)

Last Monday, together with four colleagues from the University of Amsterdam, I was asked to speak at the Opening of the Academic Year.

I started with talking about a wonderful study by Erin Hannon and Sandra Trehub (University of Cornell and University of Toronto) called "Metrical Categories in Infancy and Adulthood", and gratefully used some of the sound examples they made available online.

Interestingly, a "one-trial" version of their experiment failed miserably :-) The sign language interpreter picked up the differences immediately and communicated them so clearly to the audience that I had to ask the audience to close their eyes for the other sound examples!

The overall message was, motivated by e.g. the Hannon & Trehub study: Listen as often and as varied as you can, it will improve your sense for rhythm!

Perfect Pitch: You either have it or not?

Last week a paper was published in the Proceedings of the National Academy of Sciences of the USA (PNAS) that generated quite a stir, both in the academic world and in the press. In that paper researchers from the University of California presented the results from an elaborated web-based study (with about 2200 participants) that investigated the ability of Absolute Pitch (AP): being able to name the pitch of a tone without the use of a reference tone. Something some see as a musical gift, others as a burden.

The researchers found a bimodal distribution in pitch-naming ability that was interpreted as “you either have it or not”. Furthermore, they suggested a genetic basis for AP. And that’s were the discussion started ...

While there is some research in the possible genetic basis for AP, related studies (not mentioned in the PNAS paper) have argued, and to a large extend shown, that AP might well be a result of biases due to the task and stimuli used, largely a result of training, and problably more widepread than some think.

For example, Glenn Schellenberg and Sandra Trehub form the University of Toronto found support for a normal, not bimodal, distribution once pitch-naming or reproduction requirements are eliminated (such knowledge about piano keyboards or music notation) and familiar materials (such as soundtracks of tv programs) are used. They argue that good pitch memory is actually widespread.

Oliver Vitouch from the University of Klagenfurt wrote a comment a few years ago, called “Absolute models of absolute pitch are absolutely misleading”, summarizing the state of affairs in AP research, and arguing that it is mainly a result of musical training. Clearly there is little agreement on the claim that AP is a trait.

In addition, I find AP actually not such a special phenomenon. Although we could agree that AP occurs at different levels of preciseness on a continuous scale, in the end we should also agree that Relative Pitch (RP) is far more special. While we might share AP with quite a few animals, RP is far less common, arguably making AP in humans less special.

A 2006 recording of Glenn Gould?

Sony Music recently released a new recording (made in 2006) of Glenn Gould performing the Goldberg Variations. Curious, not? The recording was made using measurements of the old recordings and then regenerating the performance on a computer-controlled grand piano, a modern pianola.

This technology dates from the early nineties, a time when several piano companies (including Yamaha and Bösendorfer) combined MIDI and modern solenoid technology with the older idea of a pianola. Old paper piano rolls with recordings of Rachmaninoff, Bartok, Stravinsky and others were translated to MIDI and could be reproduced ‘live’ on modern instruments like the Yamaha Disklavier. Until now, the only left challenge was to be able to do this for recordings of which no piano-rolls were available.

Besides the technicalties of all this, for most people the real surprise —or perhaps disillusion— might well be the realization that a piano performance can be reduced to the ‘when’, ‘what’ and ‘how fast’ the piano keys are pressed. Three numbers per note can fully capture a piano performance, and it allows for replicating any performance on a grand piano(-la). The moment a pianist hits the key with a certain velocity, the hammer releases, and any gesture that is made after that can be considered merely dramatic: it will have no effect on the sound. This realization puts all theories about the magic of touché in a different perspective.

Nevertheless, while it is relatively easy to make the translation from audio (say a recording from Glenn Gould from 1955) to the what (which notes), and the when (timing) in a MIDI-like representation, the problem is in the ‘reverse engineering’ of key velocity. What was the speed of Gould’s finger presses on the specific piano he used? The Zenph Studios claim to have solved it for at least a few recordings. Only trust your ears :-)