K. MORI
should they uncover any evidence that seems to support dis-
crimination, they might be accused of being prejudiced them-
selves. Therefore, it was safer for researchers to avoid topics
related to racism altogether, including the LBS issue.
In hindsight, this was my turning point, moving me from be-
ing simply a basic research psychologist to becoming an appli-
ed psychologist. I decided to commit myself to this LBS issue
using whatever expertise I had acquired as a research psycholo-
gist. It would have been very easy to have just speculated or ex-
pounded on the issue without being backed up by reliable re-
search data. However, as a research psychologist, as someone
who wanted the respect of his peers, not just fleeting media
attention, I was aware that I needed experimental evidence to
support any comments I might make about the LBS contro-
versy.
Therefore, I began work on a racism-free version of the book,
which I intended to compare with the original version alleged to
be racist. In order to produce racism-free version, I revised the
original story by changing the protagonist from a black boy to a
black Labrador puppy. I also eliminated the word “Sambo,”
which had a historically pejorative connotation in the US.
However, the basic story remained the same. Therefore, if
readers were to rate both versions as equally interesting, then it
would mean that the alleged racist characteristics had no sig-
nificant effect.
Statistical tests can only show significant differences, but
cannot assure their complete absence. However, the statistical
power analysis developed by Cohen (1988) provided a solution.
If the power of a statistical test is strong enough to reject the
null hypothesis of the equality 95% of the time, its failure to
reject the null hypothesis can mean the correctness of the null
hypothesis with a risk of 5% (Type II error, β = .05). In other
words, if a strong statistical test fails to detect the difference,
then we can consider the difference to be negligible, with the
absence of difference having been proven.
I conducted an experiment to compare the entertainment
value of the two versions of LBS using four-year-old children
and found no difference (Mori, 2005a). I also developed a new
method to assess implicit attitude and proved that those who
had read LBS in childhood showed no difference in their im-
plicit negative attitude against blacks as compared with those
who had had no exposure to the book (Mori, Uchida, & Imada,
2008). Parallel to these experimental investigations, in 1997
under a pseudonym, I published commercially a picture book
based on the non-racist version of LBS featuring my own com-
puter-generated illustrations (Mori, 1997; See Figure 1).
Disappointingly, my experimental studies contributed little
on this issue, probably because they were not promptly ac-
cepted to major Japanese journals. The amusement-comparison
experiment was done in 1990, but the paper based on it was
published fifteen years later in a minor journal (Contact me for
further information on this). However, gratifyingly, the revised
picture book itself enjoyed enormous success. It sold more than
50,000 copies, and I was invited to appear in on a national TV
news program at the time of the book’s publication in 1997. It
provoked controversy yet again, and it seemed have encouraged
other publishers to put several versions of the original LBS back
on the market.
In the course of this research, I came to know the secret rea-
son why the publisher had withdrawn the book in 1988. It had
little or nothing to do with overt racism. Rather, the real reason
the book was withdrawn was a matter of piracy. The publishers
Figure 1.
Chibikuro Sampo: A non-racist version of “The Story of Little Black
Sambo” (Mori, 1997).
had not properly obtained the copyright. At the time of its first
appearance in Japan, in the year 1953, only eight years after
World War II, Japan was not yet “mature” enough to properly
acknowledge the copyrights of foreign books or other intellec-
tual property. Because the book was so popular in Japan, they
continued publishing it even after Japan bounced back to be-
come a leading nation in the world. When it was announced
publicly that LBS was considered to be racist and therefore an
uncomfortable amount of media and academic attention was
suddenly focused on it, this provided a convenient excuse for
the publisher to stop publishing it in 1988 (Mori, 2005b).
Nowadays there are several versions of LBS available in
bookstores all over Japan, including a translation of Banner-
man’s original little story. Whenever I see one, I feel a sense of
pride at being an applied cognitive psychologist who has con-
tributed to this issue.
Acknowledgements
The author would like to express his profound thanks to Re-
becca Ann Marck for her superb work in editing the English
manuscript. She also suggested the main theme of this article.
REFERENCES
Cohen, J. (1988). Statistical power analysis for the behavior al sciences.
New York, NY: Academic Press.
Mori, K. (2005a). A comparison of amusingness for Japanese children
and senior citizens of The Story of Little Black Sambo in the tradi-
tional version and no nracist v ersion. Social Behavior and Personality,
33, 455-466. http://dx.doi.org/10.2224/sbp.2005.33.5.455
Mori, K. (2005b). What have we learned from Chibikuro Sambo? Me-
tropolis, 610, 58.
Mori, K., Uchida, A., & Imada, R. (2008). A Paper-format group per-
formance test for measuring the implicit association of target con-
cepts. Behavior Research Methods, 40, 546-555.
http://dx.doi.org/10.3758/BRM.40.2.546
OPEN ACCESS
10