The language of the World Bank

Text linguistics, concerned with analysing discourses, is currently teaming up with computational text analysis. It allows to plow through huge piles of texts systematically. Research at Stanford Literary Lab is one example of such an application. The New York Times covers one of the most recent studies they conducted:

A computer analysis of more than 65 years of the [World Bank’s] annual reports found a sharp decline in factual precision, replaced by what the researchers call management discourse, a bureaucratic gobbledygook whose meaning is hard to decipher.

For those who’d like to read more about it, the New York Time article unfolds the subject in a comprehensible way. You may also want to have a look at the research paper that provides striking graphs revealing the increase of catchphrases like management as well as other strategies to construct a good feeling without saying much concrete, for example connection of governance with positively connoted adjectives.

I heard Franco Moretti of the LitLab talk about his research about two months ago. When he was asked how he sees the possibilities that these new data alalysis tools provide, interestingly his answer didn’t seem very convinced; What it basically does, he said, is „cutting knowledge and pleasure“, because (close) reading is replaced with programming (he calles it „distant reading“). It remains to be seen how many new possibilities it will open up.

(Article via Input Output)

Schreibe einen Kommentar