WHEN TWITTER first arrived five years ago, it took a lot of flak. While the service was touted by an advance guard of social media boosters, effusing about the new way to connect, the feeling for many was: 140 characters, what’s the point?
But after the Iranian revolts, the Arab Spring and the London riots – all events where Twitter was a part of the story – the little bird is not so easily dismissed. You may or may not choose to participate, but Twitter has shown that it’s an effective way to organize, a way to follow the news, and, at times, a force that makes the news.
To these acknowledged distinctions, it is time to add another: Twitter has become an established tool for scientists.
Twitter’s staccato ephemera would seem the very opposite of seriousness. How essential is it to know that someone failed to celebrate the birthday of Lil Wayne? And will that ironic observation about the weather be worth revisiting in a year?
But the of-the-moment fizziness is precisely what makes Twitter a valuable source. It provides a peek inside the collective mind, its moods and opinions, all in time-stamped text blocks, ready-packaged for computer analysis. Scientists have microscopes for the very small, telescopes for the very distant, and now they have Twitter to peer into the zeitgeist.
Last week, for example, a pair of sociologists at Cornell University released a fascinating portrait of the planet’s shifting mood and what drives it. Their approach was typical of the new social science, though massive in its scale. They trawled through some 509 million tweets, gathered between February 2008 and January 2010, using linguistic software to score their positivity based on word choices.
This gave them a mood timeline for 2.4 million individuals from 84 countries, and allowed them to make some interesting observations.
The day, they found, has an emotional arc. Our happiness has a peak in the morning, before the typical workday begins, and then fades as the day progresses, only to climb again late in the day.
The daily downer, you would think, is work. Certainly, the sub-optimal moments of working life are a popular topic on Twitter. But the same pattern is found on the weekend – the morning peak (delayed by sleeping in) followed by a fall and a slow climb at day’s end. Beneath the complex emotions of any given day, it appears, there is a tide, driven by sleep and our basic biology.
The report, which was published in Science by Cornell sociologists Scott Golder and Michael Macy, also provided insight into the question of why there is more depression during the winter. It’s not exactly the lack of light in winter, they suggest, because the length of the day did not affect average moods. Instead, positive mood tracks how quickly the length of the day is changing, falling as we approach the winter solstice and the days shorten, and rising as daytime extends.
The Cornell team is part of a growing endeavor. Scientists have shown that Twitter can be used as a rough (and inexpensive) opinion poll, and that it can be used to successfully predict opening weekend box offices.
Last year, there was much eye rolling when the Library of Congress announced that it would make Twitter part of its collection. Imagine, 140-character bursts on Britney Spears, overheard at Starbucks, and what happened last night housed in same hallowed institution as the Lincoln Bible and Washington’s diaries. Yet if there has been a consistent gap in our histories, it has been to overlook the common experience. In recent decades, historians have worked hard to augment the chronicles of wars and politics with social history and the details of daily life. You cannot understand a time or a place if you do not understand how people actually experienced it.
The early Twitter studies, while perhaps not profound in their conclusions, show that tweets have their place in history – and science.