The hunt for greater understanding of how content spreads throughout a network is a long and arduous one. As the viral nature of these phenomenon suggests, many take their cue from biology. Such is the case with a recent study conducted by the University of Michigan and Facebook.
They analysed 460 million shares of over 1,000 text status style memes (y’know, the kind that ask you to like or share if you agree – yeah, those annoying things). The study was done between 2009 and 2011, so in order for people to share the meme, they had to literally copy and paste the text into their own status box and ping it out.
One such meme was “No one should die because they cannot afford health care and no one should go broke because they get sick. If you agree please post this as your status for the rest of the day.”
The message itself was copied around 470,000 times, with one variant of it adding thinks that to the beginning (60,000 copies) and another adding we are only as strong as the weakest among us to the middle. All of these messages resonated most strongly with people who described themselves as liberals.
When they began analysing the meme itself, they found that 89% of the copies were exact, with the remaining 11% containing a mutation (variation) of some kind. Some of these variations did nothing to change the message, whilst some sought to appeal to new audiences (ie no one should die because the government is involved with health care). A further grouping of mutations took the meme off down less serious routes, such as No one should be without a beer because they can’t afford one.
The interesting part of the study of course is whether the spreading of this content conformed to any kind of pattern. The researchers modeled the evolution of the memes using the Yule process, which is used to describe the similarity of genes within a population.
“It surprised me that the model fit really well,” said Lada Adamic, an associate professor in the U-M School of Information and the Department of Electrical Engineering and Computer Science, and a member of Facebook’s data science team. “We can’t claim that memes evolve exactly like genes, but there are a bunch of very convincing parallels.”
One such parallel is with how genetic information is passed from one gene to another. In lateral gene transfer this information is passed laterally between different organisms rather than the more traditional path down between generations of the same organism. Bacteria are a good example of an organism that transfers genetic information in this way. Adamic believes that memes often go about their work in a similar fashion, with many similar facets of a meme passing from one to another. For instance, many memes contain annoying phrases such as 95% of you won’t repost this, but the 5% who are <insert desirable trait> will.
“Some of the variants would have the snippet and others wouldn’t, and the ones that had it were more popular,” Adamic said. “They were little boosters.”
Another attribute common in gene and meme reproduction is that mutations tend to occur more often at the beginning or end of a sequence. And in both processes, shorter sequences of similar information tended to spread more broadly. Those are just a few of the similarities detailed in the study. The researchers say the work opens a window into how people interact and persuade one another.
“In many ways memes shape the culture and communities through which they move,” said Eytan Adar, an assistant professor in the U-M School of Information and the Department of Electrical Engineering and Computer Science. “The thing we really wanted to understand was how the cultures shaped the memes. This work has implications for understanding how people communicate and influence each other, and let’s us measure people’s beliefs, desires, values, and even what entertains and amuses them.”
So, interesting stuff. Suffice to say, the research could have limited appeal due to its focus on quite a narrow range of memes, with the sharing mechanism used now largely superseded. Whether the findings from the study are therefore as transferable as aspects of the memes themselves is open to debate.Original post