Sunday, July 24, 2011

Bullshit Meter

I am totally stealing this from Bee of Backreaction.
It is too awesome to pass up.

A gentleman named Bernd Wurm developed a software, the "Blablameter," which checks text for superfluous/filler/bullshit words (such as those that ads and press releases are full of) as well as awkward constructions. If you run a piece of text through the Blablameter, you will get a result between 0 and 1; the higher the result, the more bullshit. Bee ran abstracts of several papers in her field and got a range or results, from very light to very heavy on bullshit.

The German version of the Blablameter,
http://www.blablameter.de, has interesting Questions and Answers (Fragen und Antworten), where it explains what the software does (e.g. can't check for accuracy of the content, just certain linguistic features).

If you just want to run a piece of your scientific writing through it, here's the English link.

http://www.blablameter.com/

I ran the abstract of a paper I am going to submit in the next few days, and the result is



Understandably, I am feeling very smug. :)

10 comments:

Alex said...

Um, I think I need to revise some abstracts.

prodigal academic said...

Heh. My last abstract got 0.31.

Cath@VWXYNot? said...

BRILLIANT!

Gears said...

Oh no!!!! just check my three papers that are currently under review.

0.26
0.34
0.54

WTF!!!?!!!

Hermitage said...

*blows finger guns* 0.18. Not bad for a PhD n00b, hehehe.

Aisling said...

Hum. Anything of significance in my field seems to have been written by a "PR expert" or at best "shows indications of bullshit English". I wonder how this tool works exactly.

GMP said...

Aisling, I am not sure how the tool works, but don't take it too seriously. I think it's calibrated towards journalism -- good newspaper articles are between 0.1 and 0.3. I am sure there are differences in style from field to field though. But mostly consider it good fun.

I read the low- (0.08) and high-scoring (0.66) abstracts from Bee's website; while the high scoring one isn't bullshit, it is a pretty poorly written abstract. On the other hand, the very low-scoring one didn't dazzle me with awesomeness either, but it is significantly better than the 0.66 one.

Everyone -- thanks for the comments and thanks for playing! Hermi, way to go!

Miss Outlier said...

Reminds me very much of Asimov's Foundation: "There is a branch of human knowledge known as symbolic logic, which can be used to prune away all sorts of clogging deadwood that clutters up human language. I applied it to the protection treaty that we signed and something like ninety percent of it boils right out of the analysis as being meaningless; what we end up with can be described as a recognition that our “ally” cannot protect us”.

Science fiction does have an uncanny way of coming true... :)

Bee said...

And how funny, since I've long given up checking trackbacks, that I coincidentally (via ScientistMother's Breakout Story) would come across your blog today :-) And, indeed, good score! Best,

B.

Bee said...

Actually, for what I understand (there is unfortunately very little information on the website) the tool is aimed at advertisement and targets besides common empty phrases the use of nominal style (Google this, you'll figure it out). Check the comments on my blog for an example that scored 1.2!