Gossip comprises huge chunk of corporate emails: Study

Washington -  The average corporate email user sends 112 emails every day and one out of every seven of those messages can be called "gossip," says a new study.

Eric Gilbert, assistant professor at Georgia Tech's School of Interactive Computing, US, examined hundreds of thousands of emails from the former Enron Corporation.

Gilbert found that - by the definition of "gossip" as messages that contain information about a person or persons not among the recipients - nearly 15 percent of the emails qualify as office scuttlebutt.

What's more, Gilbert, an expert in social computing, found that gossip is prevalent at all levels of the corporate hierarchy, though lower levels gossip the most. "Gossip gets a bad rap," he added, according to a Georgia Tech statement.

Still, another finding was that "negative" gossip, characterised through a natural language text processing analysis, was in fact 2.7 times more prevalent than positive gossip, though a significant portion of the messages were "sentiment-neutral."

The findings, according to Gilbert and doctoral student Tanushree Mitra, represent an important test of anthropological theories about gossip in what can reasonably be called the world's most popular electronic social medium: email.

"A recent survey of that literature summarised gossip as having four main purposes: information, entertainment, intimacy and influence. We found evidence of all those categories in the Enron emails, relating to both business and personal relationships."

"Enron certainly had what could be called a 'cowboy culture,' but I suspect the way they behaved internally to each other did not differ significantly from most other US corporations," Gilbert said.

Indeed, the Enron corpus - some 600,000 messages purchased following the company's bankruptcy and now made freely available for study - represents the world's largest publicly accessible body of naturally occurring emails.

It has provided grist for numerous scientific and technical advances.

For example, Gilbert said, email spam filters took a huge leap forward in efficiency in 2005 due largely to advancements made from analysing the Enron corpus.

Gilbert's and Mitra's findings were presented at the sixth International AAAI Conference on Weblogs and Social Media (ICWSM '12) at Trinity College in Dublin, Ireland.