The spread of a social media phenomenon like a meme or video is significantly complex - more than even a modern virus like Ebola.
It is difficult to pinpoint the earliest instance of the usage of the word ‘viral’ for anything other than the spread of a disease. The Oxford dictionary dates it back to 1989 – for describing the rapid spread of information as ‘viral’. The term has, however, become commonplace only recently, as we have entered the information age. Now, everything from memes and videos, to advertising campaigns and tweets ‘go viral.’ But aside from the frequent casual use of the term, an interesting question arises – does ‘viral’ media spread in the same manner as a virus, or does it have some radically different way of propagating?
The study of virulent diseases might have some answers.
Viruses spread in a characteristic manner. Starting, like an earthquake, from a single epicentre and they spread like a wave. A simple analogy is the propagation of a ripple in still water. Virulent disease pandemics across generations are consistent with this pattern, be it 14th-century bubonic plague in Europe or the SARS outbreak less than 15 years ago. This isn’t very obvious. Development of various forms of travel has made the world a smaller place. Thus, viruses that would earlier be contained close to their point of origin, now break geographical barriers and spread across continents. How does this seemingly faster spread still manifest in the familiar wavelike pattern? It turns out, that if the outbreak is ‘normalized’ by the speed of travel, scientists recover the wavelike spread of typical and much earlier virus events. This exercise of ‘normalizing’ is widely used in science. If we note down the approximate weight of the brain of different mammals on a piece of paper, a pattern is not readily apparent. However, the ratio of brain weight to the animal weight is almost a constant! Here, normalising the data by respective weight of the mammal in question unearthed the inherent pattern. Can an effective technique of normalization help find a pattern in the way social media phenomenon spreads?
The spread of a social media phenomenon like a meme or video is significantly complex – more than even a modern virus like Ebola. Unlike conventional disease spread, for a video to go viral, there doesn’t need to be person to person contact or even geographical proximity. Rather these events depend on the social networks that tie particular regions together. Social networks add a unique degree of randomness to this phenomenon because the ‘strength’ of connections between places is not directly dependent on easily measurable factors like distance. Yet, recently a group of scientists paper put up on arxiv) found out that, amidst all this chaos, social media phenomenon exhibits a much too familiar pattern – wavelike, just like your run of the mill virus. They analyzed the spread of one particular social media event which caught the fancy of the entire planet – Psy’s whimsically weird Gangnam Style dance frenzy.
The researchers collated data from across the world by looking at tweets tagged with terms like ‘Gangnam’. By studying these tweets, they charted the course and time of arrival of the video in a particular region. At first, no pattern emerged. But then, the researchers found a way to normalize all the data. And, voila – a wavelike spread was recovered. As it turns out, the spread of such content strongly depends on the ‘socio-political’ ties between regions. That’s understandable. A person who is part of a group or clique is more likely to communicate and share information with someone within his group – regardless of which country the person resides. This strength in a social connection is encapsulated in a new factor, an ‘effective’ distance if you may say so. And the Gangnam dance video data, normalized by an ‘effective distance’, shows the same spread characteristics of an ancient disease.
These results are incredibly useful if not altogether surprising. Even though viral media content was expected to flow in a manner similar to actual viruses in the physical world, that precise connection wasn’t established until now because of how convoluted social networks are. This work is the first definitive and quantitative proof that an innocuous piece of social media content and a virulent disease, both become an epidemic in the same fashion. It answers the question we all asked back in 2012 – How did a video featuring an unknown figure doing a relatively unknown dance style (k-pop) become the most watched YouTube video of all time.
While Psy’s video originated in South Korea, the epicentre of it going viral is probably the Philippines, according to the researchers. The Philippines has a larger population of English speaking people and many Filipinos have settled in developed countries. At first, the video could have caught the fancy of a group of Filipinos. They then would have shared it with some of their relatives abroad, from where it got a launchpad. This work can be critical in helping companies advertise smartly. Targeting a small region with deep social ties to many other places could help snowball a small campaign into a nationwide craze. India presents some unique test cases because of its diverse set of cultures. For example, a product suitable for low-income rural households is difficult to advertise for. But, what if the company set up a small campaign designed around second-tier towns with migrant workers in plentiful. Strong social ties can easily be exploited for viral transfer of information. On the flipside, these studies could also become a tool for the anti-social, but tech-savvy members of our population. Strategies can be designed for the optimal spread of hate videos and fake news. Ever wondered how a video primed to incite hatred, probably cooked up in a cave somewhere, even makes its way to your mobile screen? Maybe, you are the last in a line of a specially crafted, virulent and wavelike epidemic.
Studies, such as the one described in this piece are interesting because they deal with scenarios that are increasingly relevant to us. They are also important because they attempt to establish connections between two very different worlds – the virtual world of information and the physical world of diseases. There is no denying that technology built on this research can be exploited for wrong means, but if science can destruct, it can also construct. Maybe further research can help design powerful countermeasures to prevent epidemics from occurring – virtual or not.