One of my favorite cartoons as a child was “Speed Racer.” It featured an all-American boy (first name, “Speed,” last name, “Racer”) engaging in that most American of pastimes: driving fast cars. Except that “Speed Racer” wasn’t really American; it was made in Japan, and the original Japanese voices were crudely overdubbed in English. Perhaps I can be excused for not noticing the Japanese origins of the show — I was only 10 years old. Even now, as an adult looking back at those cartoons, the characters do seem awfully American-looking. Or perhaps that’s just my Caucasian bias.
Does everyone see a little bit of themselves in animated cartoon characters? Or do the artists actually draw the characters to look more generic, less racially distinctive? There have been few studies about the perceived race and ethnicity of animated cartoon characters, and none focusing on the unique Japanese anime style.
So Amy Shirong Lu randomly selected 341 main characters out of 3,098 anime films made between 1958 and 2005. Each image was carefully edited to depict only a head-on, facial portrait-style picture. All clothing and background images were edited out, like this:
The character depicted here is Asuka Langley Soryu, from the movie Neon Genesis Evangelion, and of mixed Japanese and German descent. Lu recruited 1,046 people to view a randomly-selected set of 90 of the pictures and judge the characters’ race based on the features depicted in the pictures. The animators’ intended race of each character was judged based on the promotional materials for the film, or watching the movie itself. Still, in 125 of the cases, it was either impossible to determine the character’s race or the character was of mixed ancestry. About half of all the characters were intended to be Asian, while only about 10 percent were Caucasian. Did the viewers responses match the actual race of the characters? Here are the results:
Overall, the respondents didn’t do very well! They correctly identified faces intended as Asian less than half the time, and vastly overstated the number of Caucasians.
But perhaps this was due to the race of the respondents. Lu asked the respondents to identify their own race, and indeed, her sample was strongly biased towards Caucasian respondents. These graphs divide respondents into Asians and Caucasians:
Asians were significantly more likely to say the characters were Asian, and Caucasians were significantly more likely to say characters were Caucasian. So it seems that we are simply more likely to see our own race in anime characters than the race of others. Still, it’s interesting to me that Asians still underestimated the number of intended Asian characters in the cartoons. Lu notes that many critics have accused anime of “ethnic cleansing,” stripping the characters they depict of any ethnic identity. Early anime artists acknowledged a debt to Disney films, and attempted to mimic the Disney style, so perhaps there’s some truth to these accusations.
Lu also suggests that there are many other cues to race besides facial appearance: clothing, behavior, and speech accents also play into the perceived race of a cartoon character. Perhaps if the characters were seen in context, viewers would be better at judging their race.
Lu, A.S. (2009). What Race Do They Represent and Does Mine Have Anything to Do with It? Perceived Racial Categories of Anime Characters Animation, 4 (2), 169-190 DOI: 10.1177/1746847709104647