#TheDress. #PizzaRat. #Icebucketchallenge.
If you're a person who spends much time at all on the Internet, those hashtags ring a few bells: The gold-and-white (or was it blue-and-black?) dress that drove Americans mad in February of 2015, the valiant little rat dragging a slice of pizza down the subway stairs in New York in September, and the enormously viral amyotrophic lateral sclerosis fundraiser of summer 2014.
None of these things would seem to have much in common, except that people shared, commented on and remixed them with abandon. But researchers are taking a closer look at memes as a new form of social glue — and one that can reveal a great deal about how we connect with one another and spread information.
"I think that the appeal of these creations is that when you create a meme, you're simultaneously expressing yourself, but also connecting to others," said Limor Shifman, a lecturer at Hebrew University of Jerusalem who researches memes.
Some memes, like the Ice Bucket Challenge or The Dress, connect large swaths of people. Others form a sort of private language for online subcultures. On the message board 4chan, for example, only an outsider would fail to recognize "Caturday" as the practice of posting cat pictures to the board on Saturdays. Memes, Shifman said, create the borders of communities.
Originally, the term "meme" was popularized by evolutionary biologist Richard Dawkins, who used it to refer to bits of information that travel from person to person. In today's Internet-speak, memes more often refer specifically to certain images, hashtags or videos that can be endlessly shared, built upon and repurposed. Shifman differentiates memes from viral content because of this repurposing. People may share viral videos and articles, but memes are pieces of content that people modify and make their own.
Shifman's work has found that successful memes have common features. They often involve ordinary people (like little David, a kid who slurred his speech after a dental procedure and become famous as "David after dentist). They're often repetitive (think the music video for "Gangham Style," or Rebecca Black's "Friday"). They're also simple to imitate, making it easy for people to bring their own community's interests into the meme.
In 2008, for example, parts of the Internet went wild reposting images of actor Ryan Gosling with phrases starting "Hey girl…" The results ranged form the straight flirtatious ("Hey girl, feel my sweater. Know what it's made of? Boyfriend material) to the opaquely nerdy and academic ("Hey girl. Are you saying standpoint theory could help solidify my argument about the Doctor Who/female companionship relationship dynamic as fundamentally problematic with keen feminist undertones?). Part of the beauty of the Ryan Gosling meme was that anyone could customize it to their own tastes.
Many memes are funny, but humor isn't an essential element of meme-ness, Shifman said. "We are the 99 percent," the political slogan of the Occupy Wall Street movement, is a meme. So is "Black Lives Matter," a hashtag and activist group addressing police brutality and bias in the justice system against African-Americans. "It Gets Better," a video campaign trying to prevent gay teen suicide, was another meme that people built upon and made their own.
"When you broaden this definition of memes, you can also think about their social significance in terms of political action, in terms of activism," Shifman said. "I think it has quite a lot of power because it's bridging the personal and the political. On the one hand, each person expresses his or her own ideas, and on the other hand, because the meme has this shared template, it's also political."
Memes also help smooth over the challenges of communicating on the Web. Conveying emotion through text can be hard; image macros and GIFs provide a shorthand with more flexibility than simple emoticons. Dislike what someone has to say on a message board? Reply to their comment with a picture of Star Trek's Captain Picard facepalming. Expecting a blowup of epic proportions in the comments section of an article on a hot-button topic? There's a GIF of Michael Jackson eating popcorn for that, which came from the 1982 music video for "Thriller".
Sometimes, GIFs can get very personal. Most borrow the facial expressions of celebrities, or even animals, to get a point across. But on October 29, Republican presidential candidate Chris Christie actually reaction-GIFed himself. In response to a tweet from the New York Times sending him a link to the subscription page so that he can read an article about himself, Christie posted a GIF of himself with a reaction that's difficult to put into words:
Memes can also reveal surprising insights about human preferences. Though the rule of thumb in news has long been, "If it bleeds, it leads," memes suggest that people actually prefer positivity to negativity. A study just published in the journal PeerJ on September 30 found that of nearly 20,000 tweets from September 2014, 42 percent were neutral, 36 percent were positive and a mere 21 percent were negative. Negative information spread quicker than positive information, the researchers found. The first retweet on a negative tweet typically comes in half the time of the first retweet on a positive tweet. But positive tweets end up reaching more people in their own slow-and-steady way.
"Individuals online clearly tend to prefer positive tweets, which are favorited as much as five times more than negative or neutral ones; the same holds true for the amount of retweets collected by positive posts, which is up to 2.5 times more than negative or neutral ones," the researchers wrote.
Negative tweets might engender an immediate, strong reaction, they speculated, but a subconscious bias toward optimism (known as the "Pollyanna effect") leads to broader engagement with good news.
This kind of research, part of the NSF-funded Truthy Project at Indiana University, aims to predict what kind of content gets shared and how.
"What's really interesting for us is the structure, especially if there are reoccurring patterns and generic laws that can be taken from the data," said Giovanni Ciampaglia, a computer scientist at the Indiana University Network Science Institute.
Researchers on the Truthy Project have developed algorithms that can predict whether a meme will go viral up to 2 months before it does, based on factors like the position in the network of early adopters and the diversity of the community. They've also found tell-tale signs in the way information spreads that can reveal whether the spread is organic or if someone is trying to manipulate it somehow. Astroturfing campaigns, Ciampaglia said, might try to create multiple profiles to make it look as though there's broad, grassroots support for a certain policy. But the abnormal spread of information through these fake communities can reveal the deception.
Ultimately, understanding the spread of memes has a host of practical applications, Ciampaglia said, from learning how to spread information more effectively to figuring out how false information spreads. In that sense, memes aren't much different from normal human conversations — except that they can inflate quickly and bind very large groups of people, all the while leaving clear footprints in their path.
"You can actually track the development of new ideas and concepts," Ciampaglia said.