Alan Edward Williams

Composer/ cyfansoddwr


Leave a comment

Why the “pop music revolution” research doesn’t tell you what you think it does

 

There have been a number of attempt recently at applying an evolutionary biology approach to the study of popular music, and on the face of it, this makes a kind of sense. The thinking is that because we can speak of heritage and inheritance in music, then a similar process of evolutionary change occurs in music as happens in the natural world.

A recent such study, published by the Royal Society, no less, claims that it the first such study to use quantitative data analysis, and that therefore the results of the study are “scientific”. This is a big claim – while in central Europe it may be usual to speak of Wissenschaft in relation to music, the English language tradition is to claim entirely different methodologies for the scientific and cultural realms.

The authors of this study claim their aim is to provide a “rigorous test[s] of clear hypotheses based on quantitative data and statistics”. While I’m not qualified to comment on the statistical analysis provided in the study, it seems to me that even without testing the data analysis methods, there are some fundamental flaws in the argument which undermine the basic assumptions of the study.

Using spectrum analysis, the team analysed a huge number (over 17,000) of top selling songs according to several musical parameters, and for now, I’m only going to deal with one of these: harmony. The example shown in the article deals with a tiny snippet of Queen’s Bohemian Rhapsody. The example shows the 10th -20th second of the song extract, and the chords (shown in the “supporting information”) for seconds 14-19 of this extract. It’s not clear where exactly this is in the song, as they don’t say which part of the song the 30” comes from. The spectral analysis suggests a Gm-E flat major chord change. The trouble is, there never is such a change in the song. There are plenty of Gm7 – Cm7 changes, as part of a VI-II-V7-I cycle of fifths pattern in B flat major, as you might expect in a song which uses functional key-based harmony interspersed with enharmonic note-in common modulations. As anyone who reads chord symbols knows, the top three notes in closed position of Cm7 do indeed make E flat major triad (E flat, G and B flat). It’s often a moot point in classical music and in jazz as to whether a chord is “really” chord II7b or chord IV6. Either way, it doesn’t really matter in functional harmonic analysis, since they are both approach chords to the dominant.

However, the study is based in part on a statistical analysis of the prevalence of particular chord shapes – major, minor, minor7 and dominant 7 shapes. Indeed, one of the methods by which one genre is distinguished from another is by the prevalence of one or more of these shapes. BUT if the technology used can’t reliably distinguish one of the basic categories of chord itself (as we’ve seen, it’s often moot whether you hear a chord as belonging to one category or the other) then there’s no reliable conclusion that can be drawn from the precarious tower of culturally ill-informed inference they pile on top of it.

Moreover, what is proposed is a kind of neutral listening – an ideal listener in fact. The machines used (even though as I’ve shown, they are far from perfect in their analysis) are gheld up as the ideal listener. We know that this is problematic. Listening to music is a performative action, i.e. it is an active process of selection and pattern recognition. What is recognized and valued by the listener is culturally determined. The structure of punk – as measured by the study’s methodology – would be almost indistinguishable from rock: a fact recognized by many commentators. But what was different about it was the energy, the type of energy with which the instruments were played, the accents in which the vocalists sang. In the terms of the study, these differences would be negligible. But in the 1970’s they might have been enough to get your head kicked in. Musicologists of popular music have long recognized the inadequacy of traditional categorizations in pop music such as harmony and melody, since these aspects ignore key features of the music, described by Sheila Whitely amongst others in the title of here book The Space Between the Notes.

The fact that a cursory reading of the article reveal both factual inaccuracy and cultural naiveté suggest that the authors of the article had do better to start with an undergraduate course in critical theory and harmonic analysis.

Advertisements