Getting a Read on World Happiness Through the Words of Bloggers

AA: I'm Avi Arditti with Rosanne Skirble, and this week on WORDMASTER: using bloggers as a measure of the world's happiness. How does that add up? Just ask Chris Danforth, an applied mathematician at the University of Vermont.

Chris Danforth
CHRIS DANFORTH: "A colleague of mine, Peter Dodds at the University of Vermont, we started thinking that now would be a great time to start looking at the Internet as a source of information about how people influence each other. So traditionally it's very difficult to measure how one person influences another in a controlled environment. And so we got in touch with the folks at, which is this beautiful Web site put together by Jonathan Harris and Sep Kamvar, who are computer scientists and artists.

"And what they've been doing for the last five years or so is, they wrote a program to crawl over all the blogs around the world that they had access to, and search for sentences that have the words 'I feel' or 'I am feeling in them' along with any demographic information available about the person who wrote the sentence. So typically that's age and sex and latitude and longitude. And over the last four years they were able to get roughly ten million human feelings from over two million individuals."

AA: "These were all English language blogs? Or other languages."

CHRIS DANFORTH: "Yeah, that's right. These are all English language blogs."

RS: "And what did you do with that information? Why were you interested in their site?"

CHRIS DANFORTH: "Well, we thought it would be a great opportunity to try and develop what we're calling a hedonometer or a real-time measure of happiness around the globe. So we took all those sentences and we searched in those sentences for particular words that are known to be substantive.

"We have emotional happy-sad scores for roughly a thousand words from a psychology study at the University of Florida."

AA: "Which has a great name, the Affective Norms for English Words, ANEW, study."

CHRIS DANFORTH: "Yeah, yeah, it was great! So they took, you know, a typical psych experiment. They took intro college freshmen in a Psych 101 class and they showed them some of these substantive words and asked them on a scale of one to nine, sad to happy, how they felt when they read that word."
RS: "Give us some of these words. I'm sure our audience would be interested to know what are the happy words, what are the ones that scored high?"

CHRIS DANFORTH: "Sure, so words like triumphant and paradise and love, those scored all the way up near nine. Words like luxury, glory, those words score in the seven to eight range. There are some words right in the middle like street, paper, engine, those words don't mean particularly -- they don't have a ... "

AA: "They're neutral words."

CHRIS DANFORTH: "Yeah, they're fairly neutral words. So then words like disgusted and hostage and trauma, those are words that end up near the bottom, and suicide, rape, funeral. These words are very sad. So there's a selection of one thousand thirty-four of these words that we then, every time we see those words in a blog, we can attach the score to it."

AA: "So you've got these ten million sentences all beginning with the words 'I feel' or 'I am feeling' gleaned from various English language blogs worldwide, and what did you discover?"

CHRIS DANFORTH: "Well, we found some things that we certainly weren't expecting. First of all, we found that happiness has been actually growing in blogs over the last four years at a fairly steady clip. I guess what I would have anticipated was that it would sort of oscillate around the average value and there would be happy days and sad days, but I didn't expect it to be growing.
"There were certain days on which happiness was spiking, days like Valentine's Day and Christmas. Turns out that the election of Barack Obama on November fourth, two thousand eight, that was the happiest day by a long shot that we looked at. And we have a way of measuring which words are responsible from the deviation from the typical day, and for that day it was the word proud. So there were a number of people who were saying 'I feel proud' or something like that on that day, enough to make it the happiest day in our study.
"And there were some other days that were particularly sad, there were dips. So the fifth anniversary of 9/11 was a fairly big dip, and the death of Michael Jackson last month was one of the bigger dips as well."

AA: "And yet you say -- so you've noted that happiness has been growing over the last four years, and yet we had [gone] into the depths of a recession and a financial crisis and the wars and all, and yet even so, English language bloggers were showing themselves to be happier than from four years ago?"

CHRIS DANFORTH: "Yeah, there were dips last year in the fall due to the economy, we think, but as a trend it's still going up."

AA: Next week, more with Chris Danforth from the University of Vermont. And that's WORDMASTER for now. Archives are at With Rosanne Skirble, I'm Avi Arditti.