Network analysis of how music spreads from city to city has yielded evidence that Oslo and Stockholm are the first to discover new bands, and the first to dump them.
"We find empirical support for the claim that a leadership network exists among cities," reads a paper (PDF) describing the research, authored by Conrad Lee and Pádraig Cunningham from University College Dublin. "Some cities are consistently early adopters of new music (and early to snub stale music)."
Lee and Cunningham took data from Last.fm known as " scrobbles", which register when a person plays a song, matching it with their geographical location. It then publishes lists of the most listened-to artists, divided geographically. The way those charts vary over time can show which cities are leaders, and which are followers.
![]()
![]()
The pair took chart information from Last.fm dating back three years for 200 cities around the world, numbering billions of points of data. They cleaned it up to remove noise from cities which, for example, might have only had a few people submitting data. They then adapted a method previously used to detect the leadership networks present in flocks of birds to work out whether a similar leadership network exists among cities, and found that it does.
In Europe, Oslo and Stockholm lead musical taste, and it's interesting to note that London, Birmingham, Brighton, and Bristol have a much stronger follower relationship with Oslo and Stockholm than with each other. Similarly, Cracow and Warsaw in Poland do not follow each other, instead taking influence from German and Scandinavian cities.
The pair also examined relationships within specific genres, finding that although Stockholm is very influential in music as a whole, it's surprisingly uninfluential when it comes to indie music. Oslo, on the other hand, kept its high ranking in that regard, but was beaten to the most-influential crown by Paris, which isn't especially influential when it comes to music as a whole.
In North America, Montreal and Los Angeles proved extremely influential when it comes to indie music, but Atlanta, and Chicago were the most influential cities in terms of hip-hop. Toronto managed to attain second-place rankings in both genres, but fell below Houston, Pittsburgh and the other aforementioned cities for music as a whole.
The researchers also examined two more hypotheses -- firstly that music preferences are closely related to nationality, language and geographic location, and secondly that large cities tend to be ahead of smaller cities. For the first, they found evidence that this might be the case, but with a few interesting exceptions -- most notably that New York and San Francisco are very close, despite being separated by a continent, Canada tends to be more similar to Australia and New Zealand than the United States, and that French-speaking Swiss cities are more likely to be influenced by German cities than French ones.
As for the second hypothesis, the pair found only very weak evidence that this might be the case. "While these correlations between city size and leadership position are positive, most of these relationships are quite weak," the paper reads. "We were surprised that they were not stronger." The exception came in indie music, which had a strong correlation between leadership and city size. "We are not sure why this is the case," say the researchers. "Perhaps this genre is quicker moving or more urban than the others (although presumably hip-hop is also quite an urban genre)."
It's worth noting, of course, that as all the data comes from Last.fm, these results only reflect the demographic that wants to log everything it listens to to a central database on the web. That's going to skew the results a little, but it's not unfair to suggest that this is probably an audience that's at the forefront of music trends, so the conclusions drawn by the report aren't rendered entirely invalid.
If you want to take a look at the full research paper, you can find it over at arXiv.
Updated 15:25 19/04/2012: Paper author Conrad Lee left a comment below to the effect that his work isn't necessarily measuring influence, merely which cities appear to be "ahead of the curve". He's expanded on that a little in a blog post, where he also points out that the model isn't yet shown to be predictive. It's worth a look to see some of the charts from the paper blown up larger, if nothing else.
SOURCE: WIRED
---
Let's talk about our indie snobbish ways and our cities' music scenes and post screencaps of our last.fm charts?
"We find empirical support for the claim that a leadership network exists among cities," reads a paper (PDF) describing the research, authored by Conrad Lee and Pádraig Cunningham from University College Dublin. "Some cities are consistently early adopters of new music (and early to snub stale music)."
Lee and Cunningham took data from Last.fm known as " scrobbles", which register when a person plays a song, matching it with their geographical location. It then publishes lists of the most listened-to artists, divided geographically. The way those charts vary over time can show which cities are leaders, and which are followers.
The pair took chart information from Last.fm dating back three years for 200 cities around the world, numbering billions of points of data. They cleaned it up to remove noise from cities which, for example, might have only had a few people submitting data. They then adapted a method previously used to detect the leadership networks present in flocks of birds to work out whether a similar leadership network exists among cities, and found that it does.
In Europe, Oslo and Stockholm lead musical taste, and it's interesting to note that London, Birmingham, Brighton, and Bristol have a much stronger follower relationship with Oslo and Stockholm than with each other. Similarly, Cracow and Warsaw in Poland do not follow each other, instead taking influence from German and Scandinavian cities.
The pair also examined relationships within specific genres, finding that although Stockholm is very influential in music as a whole, it's surprisingly uninfluential when it comes to indie music. Oslo, on the other hand, kept its high ranking in that regard, but was beaten to the most-influential crown by Paris, which isn't especially influential when it comes to music as a whole.
In North America, Montreal and Los Angeles proved extremely influential when it comes to indie music, but Atlanta, and Chicago were the most influential cities in terms of hip-hop. Toronto managed to attain second-place rankings in both genres, but fell below Houston, Pittsburgh and the other aforementioned cities for music as a whole.
The researchers also examined two more hypotheses -- firstly that music preferences are closely related to nationality, language and geographic location, and secondly that large cities tend to be ahead of smaller cities. For the first, they found evidence that this might be the case, but with a few interesting exceptions -- most notably that New York and San Francisco are very close, despite being separated by a continent, Canada tends to be more similar to Australia and New Zealand than the United States, and that French-speaking Swiss cities are more likely to be influenced by German cities than French ones.
As for the second hypothesis, the pair found only very weak evidence that this might be the case. "While these correlations between city size and leadership position are positive, most of these relationships are quite weak," the paper reads. "We were surprised that they were not stronger." The exception came in indie music, which had a strong correlation between leadership and city size. "We are not sure why this is the case," say the researchers. "Perhaps this genre is quicker moving or more urban than the others (although presumably hip-hop is also quite an urban genre)."
It's worth noting, of course, that as all the data comes from Last.fm, these results only reflect the demographic that wants to log everything it listens to to a central database on the web. That's going to skew the results a little, but it's not unfair to suggest that this is probably an audience that's at the forefront of music trends, so the conclusions drawn by the report aren't rendered entirely invalid.
If you want to take a look at the full research paper, you can find it over at arXiv.
Updated 15:25 19/04/2012: Paper author Conrad Lee left a comment below to the effect that his work isn't necessarily measuring influence, merely which cities appear to be "ahead of the curve". He's expanded on that a little in a blog post, where he also points out that the model isn't yet shown to be predictive. It's worth a look to see some of the charts from the paper blown up larger, if nothing else.
SOURCE: WIRED
---
Let's talk about our indie snobbish ways and our cities' music scenes and post screencaps of our last.fm charts?