The blame dodge
A little while back Sarah Kendzior wrote an interesting article about the effects of Google's choice of languages to develop translators for. Translators have been developed first for the world's more common languages -- English, Spanish, Chinese -- and only later for less-common languages. She discusses the implications this has for deepening the digital divide, bringing more of the internet within the grasp of speakers of the big languages while reinforcing the isolation of people who only speak smaller languages.
In the comments, we quickly see the manifestation of what I call the "blame dodge." Commenters point out, in Google's defense, that their translation development process depends on accessing and processing a large corpus of bilingual text. Such input material is more widely available in languages that are spoken by more, and more tech-savvy, people. So of course Google hasn't developed translators for Piraha or Pitjantjatjara yet! They're not to blame!
The commenters' response is entirely correct and entirely irrelevant. The point of Kendzior's article is not to blame Google, to accuse them of individual wrongdoing. The point of her article is to point out a structural injustice. Google is acting in an entirely rational manner -- indeed, acting in the only way that they can. But because they are acting in a world with preexisting inequalities (in language popularity and internet access -- which are in turn built on other inequalities such as the history of colonialism), their perfectly reasonable actions produce unequal effects.
The blame dodge tries to take discussions of structural injustice and turn them into discussions of blame. Users of the blame dodge can only conceptualize injustice through the framework of individual, blameworthy wrongdoing. So when someone begins talking about a problem, the assumption is that they must be blaming someone for doing something wrong. But nobody's blaming Google. They're just pointing out how social systems perpetuate their inequalities through new channels.
In the comments, we quickly see the manifestation of what I call the "blame dodge." Commenters point out, in Google's defense, that their translation development process depends on accessing and processing a large corpus of bilingual text. Such input material is more widely available in languages that are spoken by more, and more tech-savvy, people. So of course Google hasn't developed translators for Piraha or Pitjantjatjara yet! They're not to blame!
The commenters' response is entirely correct and entirely irrelevant. The point of Kendzior's article is not to blame Google, to accuse them of individual wrongdoing. The point of her article is to point out a structural injustice. Google is acting in an entirely rational manner -- indeed, acting in the only way that they can. But because they are acting in a world with preexisting inequalities (in language popularity and internet access -- which are in turn built on other inequalities such as the history of colonialism), their perfectly reasonable actions produce unequal effects.
The blame dodge tries to take discussions of structural injustice and turn them into discussions of blame. Users of the blame dodge can only conceptualize injustice through the framework of individual, blameworthy wrongdoing. So when someone begins talking about a problem, the assumption is that they must be blaming someone for doing something wrong. But nobody's blaming Google. They're just pointing out how social systems perpetuate their inequalities through new channels.
0 Comments:
Post a Comment
Subscribe to Post Comments [Atom]
<< Home