Miscellaneous - May 2015

  by Benjamin Howarth

published: 22 / 7 / 2015

Article

So it turns out that the Beatles did not spark a musical revolution after all - at least not if you read the Daily Telegraph. Earlier this month, the paper's science editor Sarah Knapton reported: "A new study suggests that the Beatles may have been given too much credit for the musical revolution which started in the US in 1964." The study was conducted by academics from two London universities (Imperial College and Queen Mary) which had used an innovative computer program to spot distinctive patterns in chord progressions, beats, lyrics and vocals from 17,000 songs from the US pop charts between 1960 and 2010. One of the authors, Professor Armand Leroi of Imperial College, told the Telegraph "The music historians all talk about how the Beatles came to America and changed everything but it's entirely coincidental... They didn't make a revolution or spark a revolution, they joined one." This seems like an unusual statement, because the study itself concluded that the 1964 'British Invasion' actually was one of only three significant musical revolutions during the period in question. By questioning the impact of the Beatles, Leroi in fact appears to be undermining the headline conclusion of his own study. Indeed, when the BBC interviewed another of the authors on its flagship 'Today' programme, Dr Matthias Mauch appeared to say the exact opposite from his colleague, arguing that the study shows 1964 as the year the British Invasion bands helped a radical, rocky sound take over from the 'dominant seventh' chords common in the jazz and blues records of earlier decades. So, what's going on here? Why is one of the authors telling the BBC one thing and another telling the Daily Telegraph something else? Luckily (unlike the majority of academic studies of this kind) the findings have been published on an 'open access' basis, meaning ordinary members of the public can read them in full (http://rsos.royalsocietypublishing.org/content/2/5/150081) What we find when reading the original study is that the researchers did, as the BBC say, identify 1964 as the first of three major periods of revolution. But, they also found that the first two revolutions represented a much less significant change than the third. So - and this may come as a shock to regular readers of Mojo and Classic Rock - a scientific analysis of the characteristics of music does indeed (as the Daily Telegraph reports) conclude that the revolution ushered in by the Beatles was much less dramatic than the shift in popular taste that occurred when hip hop conquered the US charts. And then we come to a big 'But'. The dates suggested by the researchers for the next two revolutions are highly questionable. 1983 is identified as the year the 'disco' sound caused a revolution. In fact, it appears that the researchers are actually just recognising 1983 as the year of the drum machine – disco itself (defined, as we all know, far less by its drum patterns than by its basslines) had peaked several years earlier. Then 1991 is defined as the year rap broke through. This may be technically true, in the sense that 1991 was the year rap began to become the dominant sound on the US charts, but it was certainly not the year the sound established itself. For me, two albums represent the point at which rap broke out of the sidelines and became a mainstream concern – Public Enemy's 'It Takes A Nation of Millions...' and NWA's 'Straight Outta Compton'. Both came out in 1988. If we must pick an arbitrary date to signify the rap 'revolution', 1988 seems as good a bet as 1991. But the researchers don't agree with me. They set their stall out early, writing that conventional music history, "though rich in vivid musical lore and aesthetic judgements, lack what scientists want: rigorous tests of clear hypotheses based on quantitative data and statistics." Music history, they say, compares poorly with other cultural disciplines, such as linguistics, where a mass of data has been collected to produce the kind of forensic examinations that wouldn't shame an evolutionary biologist. But, now, with the vast archive of popular music having been digitised, that data problem has been resolved – and the Imperial/Queen Mary team have set about conducting the kind of quantitative analysis linguists have relied on for decades. Alas, immediately you start to see some potential limitations in the findings. Before the researchers began analysing their data, they already grouped each of the 17,094 songs into thirteen distinct stylistic groups. To be fair to the researchers, they have been pretty careful with this, using tags from Last FM to cross reference different genres. But, the researchers will inevitably have been forced to group era-defining classics alongside pale imitations - the unique riff found in '(I Can't Get No) Satisfaction' is in no way diminished by having a similar tonal structure to another, less memorable riff. What does it matter if it has the same chords and timbre as another, inferior riff? The researchers did not, in fact, even analyse the whole song, but a 30-second sample (they do not, as far as I can tell, justify how they decided which 30 seconds to sample). A 30-second sample hardly does justice to the six minutes of Dylan's 'Like A Rolling Stone' – or indeed any song shunning the typical verse, chorus, verse format. And, so, we come to that confusing conclusion about the Beatles. The Beatles and their friends/rivals the Rolling Stones, we are told, "exaggerated existing trends in the Hot 100 towards increased use of major chords and decreased use of 'bright' speech and increased guitar-driven aggression and decreased use of mellow vocals." This means that "their evolutionary trajectories were all established before 1964, implying that, whereas the British may have contributed to this revolution, they could not have been entirely responsible for it... their songs resemble the rest of the Hot 100: for these musical attributes, they were merely on-trend." All well and good. But it misses something rather significant, which (eventually) the academics have no choice but to acknowledge - their "extraordinary popularity". No serious music critic would argue that the Beatles invented a new genre of music on their early albums – but they were the first self-contained pop group, writing their own songs and setting their own agenda, and they were the first not to fall out of fashion alongside the style to which they were associated. Ultimately, the academics involved have gone to a lot of trouble to tell us that the Beatles and the Stones were selling back to America a new version of what it already had. We knew that – but we also know a lot more about the Beatles and the Stones and we know that the technical structures underpinning their music make up only a tiny part of their story. That is not to say there is nothing of interest in the academics' work – and you might say that we should be more charitable about such an exhaustive exercise. Indeed, had the authors shown a little more modesty, they might have been entitled to it. Instead, they boldly state, "Those who wish to make claims about how and when popular music changed can no longer appeal to anecdote, connoisseurship and theory unadorned by data." To which, I say, if 'data' means feeding 30 second snippets of songs into a computer and classifying the whole of popular music into thirteen overly broad genres, I think I'll stick with anecdote and connoisseurship. The authors' intention was to temper the highly subjective discourse around popular music with a little hard scientific method. And who can blame them for that? Unfortunately, they undermined themselves from the start by trying to draw too much from too little data. In choosing to call their study "The evolution of popular music" the authors were asking for trouble. What they have actually done is measure the tonal range of the US singles chart. By using the emotive language of "evolution" and "revolution", the authors undermine their own conclusions. Their conclusion that 1991 was the year rap music truly broke through and saturated the US singles chart is interesting – but, as I've already said, the researchers are asking us to abandon common sense in favour of total trust in 'data'. To persuade me, they need to go into more detail – what was so different about the rap music being released in 1991 to the rap music being released in 1988? Have the authors, in fact, not identified a revolution at all but merely the record companies cashing in on the trailblazing success of earlier years by releasing more and more records in this style? They argue that the 'evolutionary trajectories' that were already in place before the Beatles touched down in the US are enough to suggest the Beatles' significance has been overstated. They may be correct (though I would counter by saying there are many other factors they've failed to consider). But they then make no attempt to explain why rap's 1991 'revolution' does not come with the same caveats - despite Public Enemy, NWA, Run-DMC and many others all being well established as chart acts long before 1991. Have the researchers, in fact, stumbled upon a major trend no one else had noticed – and then, managed to miss it themselves? Perhaps it was not popular taste as a whole that was revolutionised in 1991 – after all, that was the year that Nirvana and Pearl Jam became megastars by drawing heavily on the music of the 1970s – but hip-hop. Was 1991 – not recognised by many historians as an especially pivotal year for the genre - the year hip-hop through off the shackles and entirely divorced itself from the wider traditions of pop – making more of its output recognisably distinctive from other genres? Or, was the reverse true, and this the year hitherto conventional pop stars learned to imitate the sound of 'real' hip-hop artists, thus meaning the charts came to be dominated with what had previously been a niche sound? The answer will be buried somewhere in our academics' data, but disappointingly, they haven't found it. Instead, for all the talk of revolutions, we get three broad but unremarkable basic findings. Firstly, that the Beatles and the Stones - and their inferior British Invasion comrades - sounded much like the American artists they were influenced by (which we knew, but we also know that there was something else happening that made them so much more popular). Secondly, that drum machines were very popular by 1983 (again, not news to me or you). And thirdly, that rap music began to be the dominant sound on the US charts from 1991 onwards (again, we knew this, but the real questions are 'why?' and 'was this because rap changed or consumers changed?'). In fact, there IS a potential goldmine buried in the report – the finding that, even after these three revolutions, the conventional view that the pop charts became more homogenised over time is inaccurate. The researchers say that the charts were just as diverse in 2010 as they had been in earlier decades. This is a finding that we've not been able to conclusively prove before now, and one that the data set seems actually suited to demonstrating. What a pity, then, that the authors of this study seem more interested in getting cheap headlines in the Daily Telegraph than actually telling us something interesting about pop music. Like too many academics, they've got themselves bogged down in the limitations of another field (in this case, music history and criticism) and forgotten to be so forensic about the limitations of their own. Had they been less determined to slay sacred cows, and then dug a little deeper, they might have told us something we didn't already know.

Also In Condemned to Rock 'n' Roll