Axiom

 Language is a telling clue to

unacknowledged racial attitudes

 Overt racism is declining, but studies show

that unconscious bias remains widespread



 “When the looting starts, the shooting starts,” tweeted

Donald Trump (invoking a slogan from the 1960s), when unrest

broke out after the killing of George Floyd. “Thugs”, the

president wrote, were disrespecting Mr Floyd’s memory. He

almost always use ”Thugs” when some matter of race is at

issue, either the treatment of African-Americans or in

reference to “illegals” and “gang members”, implicitly Latinos.

 Are “looting” and “thugs” a kind of racist code? Many people

detect a “dog-whistle” which, without saying anything

explicitly racist, the president intends listeners to hear.

Mr Trump forcefully disagrees. In 2015 he tweeted: “They now

say using the word ‘thug’ is, like so many other words, not

politically correct (even though Obama uses it). It is racist -

bull!” Plenty of people approve of calling thugs thugs, and

object to being labelled racists for doing so.

 The intentional dog-whistle is not unknown. But people can

allude to race without realising it, too. Explicit racism is

declining in America, as a pile of evidence shows. For

instance, there is a widespread reluctance to admit racist

attitudes that many Americans were once perfectly comfortable

with. On the other hand, unconscious and semi-conscious

prejudice is alive and well. When people use racially linked

language, without overt slurs or other racist tropes,

linguists call it “racialisation”.

 Some ingenious research has teased out the links between

prejudice and language. A classic method for fingering

implicit bias asks subjects to take a quick-fire, button-pushing

test that associates positive or negative words with white and

black faces. In one such study (at implicit.harvard.edu), 58%

of online test-takers are slower to associate positive

language with black faces than with white ones. Just 14%

display the opposite tendency.

 Or consider more deliberate forms of language, formulated by

people who are generally unlikely to consider themselves

racist, journalists. A study by Dana Mastro of the University

of California at Santa Barbara looked at hundreds of news

reports in three big American papers (the New York Times, the

Los Angeles Times and USA Today) about professional athletes

accused of crimes. Such articles were more accusatory of the

alleged perpetrators when they were black. They were more

likely to provide context and humanising detail when the

accused were white.

 Praise can be racialised, too. Crystal Dunn Soubrier, a

black player for America’s women’s football team, wrote

recently that commentators attribute her ability to play in

several positions to athleticism, never to her reading or

study of the game. This seems a persistent pattern in

sportswriting. Kelly Wright, a doctoral student of linguistics

at the University of Michigan, used machine-learning to

predict an athlete’s race based on words that appear in an

article containing their name. Purely relying on the language

of news reports, the model was able to predict that Ronda

Rousey, a wrestler, was white with a 96% certainty, whereas

there was only a 3% probability that Eric Berry, a National

Football League player, was white. So it went with other

athletes; the algorithm’s guesses were almost always loaded

starkly towards black or white.

 When racialisation happens, words do not change their

dictionary definition; rather, they take on associations

with other words. Brains are statistical machines, in which

learning involves a gradual strengthening or weakening of

different synapses. Introduce one concept and related concepts

are “primed”, or made more quickly available to the

consciousness. Just because these associations are hard to

prove conclusively does not make them less real.

 The difference between “racist” and “racialised” helps explain

why Americans often talk fruitlessly past each other when

discussing words such as “thug”. With overt racism waning, it

has become painfully obvious to some people—though not all—that

the submerged part of the iceberg, implicit racial beliefs and

associations, plays a bigger role than was once realised.

Overcoming those is particularly difficult because of their

semi-conscious nature.

 A long struggle has made “racist” one of the worst things

you can be. That counts as a great success, but the corollary

is that the remaining problems are hard to talk about. Some

fear being branded racist just for trying. Seeing the bias

in even innocently intended language is a first step towards

understanding that there is still work to do.

Also check:

Tuesday, July 14, 2020