Content uploaded by Ron Eglash
Author content
All content in this area was uploaded by Ron Eglash on Apr 03, 2020
Content may be subject to copyright.
Race-positive Design: A Generative
Approach to Decolonizing Computing
Abstract
Removing racial bias from algorithms or social
process is necessary, but alone it is insufficient.
The “bias” framework tends to treat race as
unwanted noise; best when suppressed or
eliminated. This attitude extends to classrooms,
where an attempt to be “colorblind” leads to what
Pollock calls “colormute”; fearful of even
mentioning race. Just as feminists developed
“sex-positive feminism” in the 1970s, we now
need race-positive design. Thinking about race as
positive presence—as cultural capital; histories of
resistance; bindings between lands and peoples—
can be a generative force in computing
development. Here we detail the application and
assessment of African fractals, Native American
bio-computation; urban artisanal cyborgs and
other hybrid forms in which race-positive
technology design can make important
contributions. These include community-based CS
education; computational support for sustainable
architecture; unalienated labor in human-machine
collaboration, and other forms of generative
justice.
First Author
Ron Eglash
School of Information
University of Michigan
105 S. State St.
Ann Arbor, MI 48109-1285
eglash@umich.edu
Second Author
Audrey Bennett
School of Art and Design
University of Michigan
agbennett@umich.edu
Third Author
Michael Lachney
College of Education
Michigan State University
East Lansing, MI 48824
lachneym@msu.edu
Fourth Author
William Babbitt
Science and Technology
Studies
Rensselaer Polytechnic
Institute
Troy, NY 12180
Permission to make digital or hard copies of part or all of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. Copyrights
for third-party components of this work must be honored. For all other
uses, contact the owner/author(s).
CHI 2020 Extended Abstracts, April 25–30, 2020, Honolulu, HI, USA.
© 2020 Copyright is held by the owner/author(s).
ACM ISBN 978-1-4503-6819-3/20/04.
DOI: https://doi.org/10.1145/3334480.XXXXXXX
*update the above block & DOI per your rightsreview confirmation (provided after acceptance)
Keywords
Decolonization; generative justice; education;
ethnocomputing.
CSS Concepts
• Human-centered computing~Collaborative and
Social Computing
Introduction
Imagine a feminist movement that only focused on the
bad things men do. Artists like Frida Kahlo, Judy
Chicago, and Lina Iris Viktor would be irrelevant, as
would computer scientists like Phoebe Sengers,
innovators like Leah Buechley, and entrepreneurs like
Limor Fried. These feminists have all worked in
domains in which they put out visionary designs,
alternative technologies, even new ways of doing
business. In short, they were not ignoring critique but
realized that feminism also needed to be generative.
Perhaps the clearest of such statements occurred when,
in reaction to a sweeping declaration against any sexual
practices that might be suspected of being patriarchal
(including some much beloved in the queer
community), activist-scholars such as Gayle Rubin and
Mireille Miller-Young invented the term sex-positive
feminism. We contend that such an intervention is
needed now for race-positive design.
Removing racial bias from algorithms or social
processes is necessary, but alone it is insufficient. The
U.S. now holds 22% of all prisoners on earth; it is, as
Angela Davis puts it, a prison-industrial complex.
Removing racial bias from sentencing would barely put
a dent in it. The "bias" framework tends to treat race as
unwanted noise, a bug to be eliminated, thus
preserving the system containing the bug. This bug-
elimination view extends to classrooms, where an
attempt to be "colorblind" leads to what Pollock calls
"colormute," fearful of even mentioning race. Consider
the contrast between "feminist"--a positive statement
of female agency--and "anti-racist," which names only
what one is against. "Critical Race Theory" similarly
places the emphasis on critique. Even our terminology
fails to articulate a race-positive vision.
There is a wonderful history of race-positive design
movements. The 1960s slogan "Black is Beautiful," a
simple but powerful example, gave rise to the Black
Arts Movement; AfriCOBRA and other Chicago south
side organizations (which included George Lewis's
pioneering work on African-influenced computer
music); Columbia University's "Urban Center" focused
on design justice; and so on. Wired senior editor Jason
Parham, drawing mainly from McIlwain's Black
Software, writes that in the 1990s the proliferation of
African American online hubs--NetNoir Online, Melanet,
GoAfro, Universal Black Pages (as well as other online
ethnic communities such as LatinoLink and
CyberPowWow) provided a momentary glimpse of what
grassroots platform ownership looked like before
"corporate gentrifiers like Google and Facebook moving
in and taking over." [16]
Thus, just as sex-positive feminism did not ignore
critiques of patriarchy, race-positive design does not
ignore critiques of racism. Rather, it enables thinking
about race as a positive presence—as cultural capital,
histories of resistance, bindings between lands and
peoples—and the means by which it can be a
generative force in technologies for just and sustainable
futures.
Figure 1: Race-positive designs
influenced by African fractals:
Animated gif by Francois
Beaurain and Medina Dugger;
Ethiopian architecture by Xavier
Vilalta; SciFi novel by Nnedi
Okorafor.
For example, figure 1 shows some outcomes of the
African fractals project. By documenting the ways that
African traditions used recursive scaling in built
environments, hairstyles, metal sculpture, textiles,
divination codes, and other domains, we were able to
make these “heritage algorithms” available for
contemporary projects in STEM education, artisanal
production, architecture, arts, urban agriculture,
science fiction novels, and other design related
disciplines. But before we dive into that discussion, we
need to ask a more fundamental question: what
constitutes race for the framework of race-positive
design?
Race is recursive: rethinking the relation of
race, genetics, and society
Much of the writing in the racial bias genre is
accompanied by the caveat that race does not "really"
exist; it is merely a social construction. It is
understandable how this confusing stance came about.
Social construction has been used effectively to show
that categories such as "female hysteria," "health
benefits of tobacco," and other once-real phenomena
were merely illusionary social inventions. [12] Racist
biological theory posited that genetic differences for
Black, Native American, and Latinx populations
determined innate intelligence; therefore, efforts to
create equity in income or academic success were
hopelessly contradicted by nature. By attacking race
itself as a social construction, one could dismiss this
genetic determinism. But it does so by dismissing
genetics, creating a forbidden firewall between
biological and social domains.
A race-positive perspective, in contrast, need not assert
that our current understanding of racial genetics is
correct; but can still engage the utility and flexibility of
race and gene relationships. For example, Hubbard and
Wald compare screening for Tay-Sachs disease in the
Jewish population, which was highly successful, with
screening for sickle cell trait in the African American
population, which had many negative effects: by the
mid-1970s almost all of the major airlines grounded or
fired employees who were sickle-cell carriers (even
those having only one copy of the gene, thus no
anemia). [13] The U.S. Air Force Academy did the
same, despite the fact that the claims for oxygen
deprivation under physical stress had little scientific
support. [2, 3] That a group enjoying greater social
privilege can take advantage of a race-gene
relationship, while an oppressed population cannot,
suggests that we need to demand improving race-gene
social justice, not dismiss it as a social illusion.
In studying the "reveal" YouTube videos of African
American citizens looking at their DNA ancestry tests,
social scientist Alondra Nelson reports:
I've spoken with African Americans who have
tried four or five different genetic genealogy
companies because they weren't satisfied with
the results. They received different results each
time and kept going until they got a result they
were happy with. [21]
The idea of a flexible "self-fashioning" of racial identity
--a process that combines genetic information and
personal decisions--may seem counter-intuitive. [17]
But it is increasingly common for endeavors such as
medical diagnosis, where we take official
recommendations regarding genetics, bloodwork, etc.
and make our own best guess for diet, treatment, and
so on. If we can democratize health care, why not
democratize race? New research in epigenetics shows
that our DNA expression, even across several
generations, is affected by DNA methylation from diet
and other environmental factors. My genetically
influenced hair, skin etc. might inspire changes in diet
and other behaviors, which creates epigenetic effects
modulating gene expression. Why not reconceive race
as part of this recursive molecular ecology? We need to
move away from the colonial view of race as a means
of controlling populations and lands. Exploring its
recursive composition across many time scales--a
network of factors including genetic, dietary and social
effects--we can positively engage the real, rather than
endlessly run from an illusion. [4]
Decolonizing Computing: a generative
approach
One of the frameworks that allow a more positive role
for racial identity is that of decolonization. Tuck and
Yang's "Decolonization is not a metaphor" essay
illuminated the importance of starting from the view of
actual lands and peoples, and the same is true if we are
seeking to decolonize computing. [20] We use the
metaphor of the internet cloud, but server farms
require massive amounts of physical energy, as well as
cooling. Facebook, Apple, and Microsoft all have server
farms located by Columbia river basin hydroelectric
dams, which provide both electricity and cooling. Native
communities destroyed or damaged by these "eminent
domain" dams include Nez Perce, Umatilla, Warm
Springs, Lummi, Slockish, Colville, Nespelem, Sanpoil,
Sinixt, Palus, Wenatchi, Entiat, Methow, Sinkaietk, and
Yakama to name but a few. Colonization is not a distant
historical event; it is an on-going criminal act, and
computing is increasingly taking the role of the mob
boss.
The reason why so many native nations have been
affected is, in part, the enormous span of the Columbia
river basin: 64 dams spread across seven states and
parts of Canada. But it is also because Indigenous
cultures deliberately focused on water systems to
create a generative economy. In a generative economy,
the value that is generated by humans and non-
humans is not extracted. Rather it is circulated and
maintained in unalienated forms. Archaeological and
ethnographic evidence shows that deliberate
enhancement of aquatic habitats, transplantation,
predator management and other factors sustained
through a complex of spiritual and social dynamics
created ecosystems of profound productivity: human
populations rose, but food species populations did not
decline. [14, 19] Computational models are one way to
support decolonization; for example, Lansing's
simulations of traditional rice irrigation in Bali were
used to defeat forced adoption of pesticides and
artificial fertilizers, and gain U.N. status as a world
heritage site. [15] But broader application can be
achieved if we adopt the generative framework as a
generalizable model for contemporary, information-
driven systems, from industrial automation to online
communities; even including AI. [10]
Industrial processes create destructive effects across
the three major domains of life: they alienate ecological
value by extracting agricultural products and returning
chemicals rather than compost. They alienate labor
value by paying workers a fraction of the value they
create, and removing creativity and agency from
Figure 2: From traditional
algorithm in hair to students’
algorithm on the screen
work practices. And they alienate semiotic value by
colonizing our social networks, privatizing the public
sphere. Designing for a transition to a generative
economy requires computational innovation in all three
domains, replacing one-way extraction with generative
cycles, as well as networking between the domains,
restoring the links between growing, making and
socializing. [10]
A key strategy for race-positive design is to start with
systems in which there is already at least some survival
of unalienated value generation--traditional
agroecology; artisanal labor; grassroots media--and
develop technologies that can empower and network
those into larger systems. We tend to think of
"artisanal" with some skepticism, as corporations and
wealth-encrusted specialty shops have appropriated the
term. But let us consider the artisanal aspects of Black
hair salons and barbershops. They have been
wellsprings of unalienated labor and semiotic value
since their origin, linking African hair genetics and
traditions of combing, braiding, dreadlocking, and
shaving with varieties of contemporary cosmetology
innovation that cross race and class lines with
astonishing power.
Figure 3 shows an example of the heritage algorithms
embedded in black cosmetology practices. From
interviews with braiders, we gained an understanding
of their computational thinking (iterations of geometric
transformations). These interviews were a crucial step
in documenting internal Black algorithmic traditions,
rather than imposing European understandings
externally. From there, we developed a blocks-based
scripting interface, which allows students to simulate
existing cornrow designs, and create new patterns on
their own. This interface is the software component of a
larger website (csdt.org) that provides cultural
background, lesson plans, and other teaching support,
called "Culturally Situated Design Tools" (CSDTs).
CSDT blocks such as "translate by % width" are not
part of Western math fundamentals; rather, they
portray how practitioners of the tradition conceive it. It
creates a minor challenge for teachers, who have to be
willing to show students how the concept of calculating
percentages, and the concept of geometric translation,
can be put together. But it is worth the trouble: studies
show statistically significant improvement in STEM skills
and interest in comparison to baseline measures using
CSDTs for underrepresented students. [6, 9] Likewise,
statistically significant improvement has been shown in
controlled studies with other CSDTs. [7, 1] Once
students have developed their own unique creative
design, they can physically render this pattern with
digital fabrication into jewelry, textiles, and other
forms. Figure 3 shows student projects as 3D printed
mannequin heads, which were then passed to braiding
shops. shops. As adult cosmetologists became more
involved, the race-positive design concept
spontaneously emerged in their descriptions of the
experience. For example:
I'm excited about what I'm learning today
because I realize it wasn't all about slavery, it
wasn't all about, you understand, the negative of
me having this skin tone. It's more so about that
pride and us understanding that we are a people
that should be celebrated in a sense.
They also made STEM contributions, suggesting
additional research areas: these became digital
innovations in pH sensing, hair damage testing, and so
on (figure 3). Later work explored potential
Figure 3: Generative STEM cycles
value back to community sources
collaborations with digital systems for local food
growers, creating ingredients for organic hair products.
All of which is not meant to suggest any particular
solution, rather it is an illustration of how bottom-up
development, starting with beloved culture-based
traditions and practices, can be enhanced by
computational innovation and networked to form a
growing ecosystem of unalienated value circulation.
While the above example focused on African American
culture, similar efforts have been made for Latinx,
Native American, and African practices. [10, 5] Some of
our experiments with these generative computing
efforts have looked at white racial identity as well.
White males in their teen years are especially
vulnerable to the seductions of far-right supremacist
groups. One factor appears to be certain strategies of
the political left, in cases that emphasize making white
youth feel more ashamed of their privilege and
ancestors' crimes. [11] As a race-positive counter-
strategy, CSDTs have included examples such as the
Appalachian quilting CSDT. Here students can see how
the "radical rose" (a symbol of the abolitionist
movement) was a popular motif in pre-civil war quilting
because of solidarity from poor whites, overturning
preconceptions of what constitutes white rural heritage.
Another strategy involves CSDTs using Celtic interlace
designs. Celtic tribes once spread across Europe from
Ireland to Turkey; they were essentially Europe's
Indigenous culture. The only difference is that the
colonizers--the Roman Empire; and so on--were
internal; Europe essentially colonized itself. Our use of
heritage algorithms in simulations of contemporary
Celtic patterns for leatherwork, textiles etc. is not
limited to white youth: it turns out that a Celtic design
imaginary, from Latinx Dungeons and Dragons
aficionados to Black Game of Thrones cosplayers, exists
across a very multi-ethnic fan base. Figure 4 shows a
design created by an African American student who
began simulating cornrow braiding, replaced the plaits
with circles, and laser-etched a pendent he called
"Viking horns." Such cultural hybridity is a common
outcome for the generation that taught us how our old
categories of race, sex, and class can become a fluid
palette for portraits of more just and livable worlds.
Conclusion
One of the most quoted lines from sex-positive
feminism was from Wolf: "orgasm is the body's natural
call to feminist politics." Race-positive design can also
insist on the radical potential of delight. It is a call to
playful celebration of old cultural roots and new hybrids
that take effective aim against both individual and
structural racism. And it is a powerful guard against the
tendency of racism to force us into adopting a fearful,
paranoid stance, turning colorblind into colormute.
References
[1] Babbitt, W., Lachney, M., Bulley, E., and Eglash,
R. 2015. Adinkra mathematics. A study of
ethnocomputing in Ghana. REMIE Multidisciplinary
Journal of Educational Research 5, 2: 110.
[2] Bowman, J. E. 1977. Genetic Screening Programs
and Public Policy. Phylon (1960-) 38, 2: 117.
[3] Duster, T. 2003. Backdoor to eugenics. Routledge,
New York, London.
[4] Eglash, R. 2013. Toward a Cybernetics of Race.
Determinism and Plasticity in Ideological and
Biological Systems. In Anthropology of race.
Genes, biology, and culture, J. Hartigan, Ed.
Figure 4: from cornrow
simulation to Viking horns
School for Advanced Research advanced seminar
series. SAR Press, Santa Fe, 57–82.
[5] Eglash, R. 2018. A Generative Perspective on
Engineering. Why the Destructive Force of
Artifacts Is Immune to Politics. In Engineering a
Better Future, E. Subrahmaian, T. Odumosu and J.
Tsao, Eds. Spinger, Cham., 75–88.
DOI=10.1007/978-3-319-91134-2_9.
[6] Eglash, R., Bennett, A., O Donnell, C., Jennings,
S., and Cintorino, M. 2006. Culturally situated
design tools. Ethnocomputing from field site to
classroom. American Anthropologist 108, 2: 347–
362.
[7] Eglash, R., Krishnamoorthy, M., Sanchez, J., and
Woodbridge, A. 2011. Fractal simulations of
African design in pre-college computing education.
ACM Transactions on Computing Education (TOCE)
11, 3, 17.
[8] Eglash, R. 1999. African fractals. Modern
computing and indigenous design. Rutgers
University Press, New Brunswick, NJ.
[9] Eglash, R. and Bennett, A. 2009. Teaching with
hidden capital. Agency in children's computational
explorations of cornrow hairstyles. Children Youth
and Environments 19, 1: 58–73.
[10] Eglash, R., Robert, L., Bennett, A., Robinson, K.
P., Lachney, M., and Babbitt, W. 2019.
Automation for the artisanal economy. Enhancing
the economic and environmental sustainability of
crafting professions with human–machine
collaboration. AI & Soc 29, 3: 236.
[11] Gray, B. J. 2018. The politics of shame. https://
www.currentaffairs.org/2018/03/the-politics-of-
shame.
[12] Hacking, I. 1999. The social construction of what?
Harvard Univ. Press, Cambridge, MA.
[13] Hubbard, R. and Wald, E. 1999. Exploding the
gene myth. How genetic information is produced
and manipulated by scientists, physicians,
employers, insurance companies, educators, and
law enforcers. Beacon Press, Boston, MA.
[14] Lansing, J. S., Lansing, P. S., and Erazo, J. S.
1998. The Value of a River. Journal of Political
Ecology 5, 1: 1–22.
[15] Lansing, J. S. and Cox, M. P. 2019. Islands of
order. A guide to complexity modeling for the
social sciences. Princeton University Press,
Princeton, NJ.
[16] McIlwain, C. D. 2020. Black Software. The
Internet and Racial Justice, from the AfroNet to
Black Lives Matter. Oxford University Press, New
York, NY.
[17] Odumosu, T. and Eglash, R. 2010. Oprah, 419,
and DNA. Warning! Identity Under Construction.
In Diasporas in the New Media Age. Identity,
Politics, and Community, A. Alonso and P. J.
Oiarzabal, Eds. University of Nevada Press, Reno,
NV, 85–109.
[18] Sutton, S. E. 2017. When Ivory Towers Were
Black. A Story about Race in America's Cities and
Universities. Fordham University Press.
[19] Thornton, T. F. and Deur, D. 2015. Introduction to
the Special Section on Marine Cultivation among
Indigenous Peoples of the Northwest Coast. Hum
Ecol 43, 2: 187.
[20] Tuck, E. and Yang, K. W. 2012. Decolonization is
not a metaphor. Decolonization: Indegeneity,
education & society 1, 1: 1-40.
[21] Younge, G. 2006. New Roots. https://
www.theguardian.com/science/2006/feb/17/
comment.columnist