Bloggers raise alarm over sleep of reason

The standfirst of "You can't tell me anything" ( October) says that "the gleefully bull-headed ignorance shown by politicians, bloggers and others" suggests that "scientific evidence and scholarly analysis may soon count for nothing". But nothing in the article supports the claim that bloggers are on the side of ignorance. What I have seen over the past 10 years is a steady increase in mainstream journalism spreading stupid claims about academic research while academic bloggers battle to set the record straight.

Let me cite just one current example, from linguistics. Journalists all over the world have been repeating the claim made by Ralph Fiennes in front of the British Film Institute to the effect that "a world of truncated sentences, sound bites and Twitter" is degrading our culture and causing us to lose our ability to use longer words the way Shakespeare did. Now, that's a factual claim, but not one journalist asked whether it was true. A lone blogger, Mark Liberman at Language Log, raised that question and addressed it with a small experiment. He took a Shakespeare play, some P.G. Wodehouse stories and a collection of tweets, and used off-the-shelf software to check the word lengths and compare them. The result: Fiennes believes the diametrical opposite of the truth. Tweets gathered by a University of Pennsylvania student newspaper show greater average word length than Wodehouse, who shows greater average word length than Shakespeare.

In that case, the blogger had to do the research himself, and publish it. But I could cite many such cases of bloggers battling misconceptions fostered by careless mainstream reporting about empirical matters. "You can't tell me anything" argues that evidence and analysis are needed when evaluating claims, and that holds just as much for unsupported slurs against bloggers as for anything else. Studying academic group blogs, especially in the sciences, rapidly reveals that they are an important bastion against mainstream press practices such as over-hyping research findings, or neglecting them, or simply not asking whether they exist.

Geoffrey K. Pullum, Professor of general linguistics, University of Edinburgh