Complexity of Social Media Messages is Rising—But Why?

Social media is everywhere, and like any new technology, a new generation of parents worry about how being on social media platforms affects children. In particular, people seem to worry about social media making us stupid, or causing our language to deteriorate, as we learn to communicate in shorter bursts, and with emoticons and gifs rather than wordy passages.

However, surprising new research provides a counterintuitive response to these naysayers. Ivan Smirnov of Moscow's National Research University Higher School of Economics studied the complexity of social media messages sent between 2008 and 2016, and analyzed the way they changed over time. The messages were sent and received by almost a million VK users (VK is like Facebook in Russia) in St. Petersburg. The data set included more than one billion words.

Complexity on the rise

Smirnov noted the ages and education of the people studied, and then went on to ask himself: how did the complexity of the messages change over time? He decided to use average word length as a proxy for complexity—an imperfect proxy, but not a meaningless one. He found that people started to send more complex messages as they got older and on through their 20s, and then plateaued in their 30s. The complexity of their messages started growing again starting in the early 40s.

However, one particularly interesting finding was that message complexity was constantly increasing, in a way that couldn't be attributed to aging alone. This might have been due in part to technological advances; it's a lot easier to text on a modern smartphone than it was on an old cellphone that demanded you tap the “2” button three times to get the letter “c”, for example, and so modern text messages are almost certainly more complex. Better interfaces prompt more complexity.

Still, an overall increase in complexity this notable at all ages was a dramatic finding. For example, 15-year-olds in 2016—the least complex age group—create more complex posts than any group did in 2008. Is this due to better interfaces alone? Are we simply that more adapted to expressing ourselves this way?

Are we getting smarter, or interfacing more effectively?

Smirnov believes this change is an observation of the Flynn effect, or more specifically, a digital Flynn effect. The Flynn effect is the sustained, substantial increase in IQ that has been noted and measured in people around the world since the 1930s. The effect is so substantial that what was “average” IQ in 1930 is less than 80 by today's standards. Most people believe this is down to improvements in education and nutrition, more familiarity with tests, smaller families, and other social factors.

In any case, whether we're measuring it just like we did in the 1930s or not doesn't matter so much. What matters more is that between 2008 and 2016 little complexity seems to be lost—and some may have been gained. Maybe being open to technologies and learning to interface with them more effectively can improve performance across the board.