vrijdag 2 mei 2014

The tragic case of distant reading


Franco Moretti teaches literary at Stanford and I feel conned in spending 15 Euro on his 2005 book 'Graphs, Maps, Trees, Abstract models for literary history.' Moretti is also a Marxist of some kind and as a working man I feel doubly ripped of: at my minimum wages that represents 2 hours of work.

For some reason in literature, in academia, scholars are expected to write with the convolution of the Mississippi river and the jargon of a turbine-engineer. It is good because it will always prefer to make you read the original rather than its interpretations. Mr Moretti is no exception: his prose is 'magical' in the personal and extrapsychological spectrum, or spectra, (the model is not the map and the network is an inverted social doxa of personalized metastatis vis a vis Hegel [in his early period]). Like, obscure, self-referencing and self-aggrandizing.

The outset of Mr Moretti project I can completely follow and admire: millions of book have been published, it is impossible to read them all, perhaps, from a distance, using text mining and statistics, something of value can be learned.

What happens next however is of a quality that I find staggering. The absolute bottom is reached on page 25 where a geograph shows 'US film comedies as a percentage of top five box office hits 86-95'. In some places, Central Europe and Israel, this can be as high as 30%, elsewhere, Serbia and Taiwan, it is 0%. Why there is the need to suddenly mention film I do not know BUT if an academic in the humanities can do no better than explain why comedies "are hard to export" with the observation that "contemporary comedies make large use of jokes, which are often lost in translation" I can only petition any serious body giving to man a forum to discontinue all their connections with him. Not only is it bad science: the statement needs proof with linguistic analysis (would Hungarian really be better suited to translate US jokes than Serbo-Croatian?). It is incomprehensible that a professor in literature can't come to the idea that humor is, at least in part, culturally defined.

The book continues without ever picking up steam, some graphs are nice, most seem pointless, most visualizations fail to convince me. There is much talk about form and genre and evolutionary metaphors but it is all cliquish more about reference than real understanding. If you want to do quantitative science, provide the data and the software, tell the world exactly how you came to your results, make it falsifiable. Mr Moretti begins with the right idea and then manages to ruin it completely with his own continuing presence, he seems concerned more with staking his claim than with doing a proper job.

I refrain from using the world charlatan but I am not sure why.

Geen opmerkingen:

Een reactie posten