On Color-Blindness and Observation Method

Thoughts, US Culture
four diverse women

The author (second right) and international students on a trip to Washington, DC

There are by now several compelling arguments why being “color-blind” does not address the problem of racism but merely masks it. A recent study in Social Psychological and Personality Science, detailed in New York Magazine, found that children were self-censoring to avoid referring to people by race. Middle-school children were asked to identify a photo of a person out of a group of 40 by asking yes-or-no questions. It was discovered that:

Overall, the odds of children mentioning race were 4 times lower than mentioning gender […], despite the fact that both questions were equally useful for completing the task. It was also clear that children noticed race: Over 97% reported noticing that the photos varied by race in the posttask measures.

The results were consistent across different (self-reported) racial backgrounds. The children who avoided mentioning race despite noticing it explained their choice as not wanting to be rude or racist. The study concludes that the stigma on mentioning race hurts minority children as race is central to their identities.

While this is a powerful argument to stop avoiding race in conversation and pretending to live in a postracial society, one should be careful not to interpret the findings as an encouragement for referring to people by their (perceived) background — something the authors never explicitly state in the article. Whatever benefits a person may derive from a strong racial identity, that identification needs to be determined by the person themselves and not assigned to them by the external observer.

Identifying Race by Observation Is Inaccurate

The first, and obvious, problem with the observation method is that its results are inconsistent between different observers and may not reflect the person’s self-reported background. A 2002 study of racial classification by observers, conducted at the University of Michigan, found that “[w]hite men and Asian men consistently demonstrate a simplistic understanding of race. They take relatively little time to classify others, are the least likely to identify people as multiracial, and the most likely to identify people as black.”


Portrait of the author (R) as a young student, by a person I met in Providence, RI, in 2008 and would like to credit if I could just remember his name!

People’s perceptions of others’ backgrounds seem to be informed by their own life experiences and exposure to diversity. So, depending where I’ve been in the world, I’ve been perceived to be Armenian by Russians; Iranian, (Southeast Asian) Indian, or Latina by Americans, and Romanian by Germans. While I have the utmost solidarity with minorities and oppressed groups everywhere, I don’t actually belong to any of the groups mentioned.

One of the motivations for collecting race information by various organizations is to identify and address racial disparities. However, having an outsider identify a person’s background tends to be inaccurate for people who identify as Hispanic or multiracial. Moreover, in the realm of healthcare, broad categories like “Hispanic” or “Asian” fail to capture or explain the vast differences in health outcomes within the group. Finally, and most disturbingly, educational and wealth outcomes for immigrants of color are best predicted by how assimilated the person is into US society than by their skin color or actual parentage. Recent and “unassimilated” West Indian, Haitian, and Latino immigrants often do better than their assimilated counterparts — “[a]ssimilation means learning the racial order of the United States, and for people of color it means joining the lower ranks of that racial order.” From the standpoint of having accurate data to inform institutional decisions, the best practice is to let the person self-identify.

Does Not Reflect Complexities of Lineage

The second problem with observer identification of race is that it assumes clear-cut lineages and identities. It ignores the history of Afro-Latin@ individuals who might just be labeled “black” in the US and denies the blackness of light-skinned individuals who identify as black. In addition, assigning a race to individuals from the outside does not let those who so choose honor both/all sides of their family and describe themselves in terms that best reflect that identification.

Denies The Person’s Autonomy

The most nefarious consequence of the observation method is that it essentializes race and ascribes it to the person as something objective, immutable, and determined at birth. No one would deny that every human is descended from a group of people who lived in certain (sometimes different) areas and belonged to certain culture(s). At the same time, the notion that people of one (perceived) race somehow have more in common genetically than any two random people does not hold up. In its statement on race, the American Anthropological Association points out that “there is greater variation within ‘racial’ groups than between them.”

If we agree that race is a social construct, does this mean that a person can pick and change their race at will? This is a question I will not attempt to give a definitive answer to. Rachel Dolezal, the former NAACP chapter president who claimed to be black, suffered a wave of backlash once it came out she did not actually have any (recent) African ancestry. My understanding of the criticism leveled at her is that she sought to reap the benefits of membership in a traditionally oppressed group without suffering the marginalization that other individuals in that group experience, which has more to do with cultural appropriation than with the innateness of race.


The author (right) and classmates on a boat ride

Self-serving motives aside, people have chosen and changed their race based on societal forces and personal preferences. A researcher looking into the Russian census in 2000 reported that “in the region of Krasnodar, where the Armenian population experiences persecution, many were reluctant to identify themselves as Armenians on a government census questionnaire.” Similarly, many will recall people of mixed African-European(-Native) heritage passing for white in order to escape persecution. Not all fluidity in racial identification stems from discrimination. Millions of Americans changed their answers about their race between the 2000 and 2010 census, perhaps to better reflect their backgrounds. In addition, racial divisions are time- and region-specific, as evidenced, for example, by groups of Polish, Jewish, and Italian immigrants to the US who came to be seen as white over time.

To tell a person “you are what you look like to me” is to deny them their agency and family history they choose to cherish. If perceptions by others take precedence over self-identification, then every Sikh person attacked for being perceived as Arab/Muslim (because Mulsim is a look, ya know?) must immediately change their self-identification. Rather than shaming a person who doesn’t but you think should identify as black or Jewish (yes, they were viewed as an ethnic group in the former USSR) and telling them how they would still be targeted by oppressive regimes despite their self-identification, we should encourage anti-racist activism both among the marginalized group, their allies of other backgrounds, and society at large.

Now What?

So have we come full circle to being “color-blind” and not referring to people by their race? That is not at all my argument. As race is an important aspect of identification and cohesion for many people, I am not advocating making it unmentionable. Instead, we should try to:

  1. Avoid classifying people based on our often limited and inaccurate knowledge of their background
  2. Ask the person what they would like to be called and respect their choice if that’s not something they would like to disclose
  3. Respect their self-identification and not try to dispute it (with the Rachel Dolezal caveat above)

I have tried to incorporate relevant research into this piece. However, no post can cover all facets of an issue as complex as this, so I welcome additions, corrections, and references to angles I have not considered.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s