People more drunk at weekends, researchers discover

by

A parody from 2000

It’s open season on Wikipedia these days. The project’s culture of hatred for experts and expertise has become the subject of widespread ridicule. Nick Carr christened it “the cult of the amateur”.

But what has professional academia done for us lately? Here’s a study from the University of Amsterdam to ponder.

New Scientist reports that researchers for Professor Maarten de Rijke at the Informatics Institute have been recording words used by bloggers, in an attempt to find interesting or unusual patterns. What revelations did the team’s MoodViews software unearth?

The team discovered that the LiveJournal label “drunk” becomes increasingly popular each weekend. And around Valentine’s Day, “there is spike in the numbers of bloggers who use the labels ‘loved’ or ‘flirty’, but also an increase in the number who report feeling ‘lonely’.”

It gets better.

The team also noticed that on the weekend of the publication of the most recent Harry Potter book, bloggers used “words like ‘Harry’, ‘Potter’, ‘shop’ and ‘book’,” PhD student Gilad Mishne reveals.

This work really should put the Nobel Prize Committee on Red Alert. Alongside the existing scientific prizes for Chemistry, Physics and Physiology and Medicine, the Laureate Committee should design a new category for the “Bleeding Obvious”, or the “Dying Ridiculous”.

More seriously, let’s look at what this episode teaches us.

Two things are immediately obvious: Mishne’s study was considered worthy of academic funding, and it was considered worthy of an article in a popular science magazine.

The study doesn’t tell us anything we didn’t know before: unless you’re surprised by the revelation that people get more drunk at weekends, or people talk about Harry Potter books more when a new Harry Potter book goes on sale. The study is really considered funding-worthy and newsworthy because of what’s unsaid – the implication that the aggregation of internet chatter will reveal some new epistemological truth.

Alas, as marketers have already discovered to their cost, the deeply unrepresentative composition of internet enthusiasts means that no such conclusions can be drawn. An agency that market-tested a campaign with bloggers and received an enthusiastic response, came to grief when it flopped with the real public. One only need look at the rise and fall of the Howard Dean campaign to see how steep these delusions are, and how quickly they disintegrate on contact with the wider world. Many more similar and expensive mistakes will be repeated in the years to come.

This belief, that there’s a “collective conscious” out there we haven’t spotted yet because we haven’t been looking hard enough (or haven’t had fast enough computers), underpins a lot of the sillier technology evangelism today. It should be considered right alongside attempts to teach religious cosmologies on a schools’ scientific curricula as part of the same phenomenon. Who can blame the creationists, or supporters of “Intelligent Design”, for trying to push their agenda into schools, when you can earn a living doing much the same thing at a grown-up University?

But it also highlights a profound crisis in academia, which isn’t simply limited to science.

The traditional model of scientific enquiry is to make empirical observations on the world, and then find a mechanistic explanation that’s reproducible and in some way new. (We oversimply crudely – Kuhn’s Structure of Scientific Revolutions for example finds many worthwhile contradictions with this view – but bear with us). By contrast, research like this finds nothing we don’t already know, which ought to disqualify it from the label research. Unless one is suffering from an acute case of Asperger’s, every conclusion here can be explained by psychology, sociology or economics – at a level so basic that a child has no difficulty understanding. So this isn’t so much an example of science, but an example of pseudo-science trying to elbow aside other disciplines.

Or maybe there’s a much simpler underlying explanation.

Mishne’s previous publication, the delightfully titled Leave a Reply: An Analysis of Weblog Comments was conducted with an internet marketing company employee. Mishne has interned at the company, BuzzMetrics, himself. Perhaps the answer is that corporate sponsorship is now so pervasive in academia that what results in place of scientific research are these worthless little tracts of technology puffery. And it’s so subtle, that it’s only when we come across a real clunker like this, that we notice.

0 responses to “People more drunk at weekends, researchers discover”