THIRTEEN

An Orderly Assemblage of Biases:

Troubling the Monocultural Stack

Jason Edward Lewis

The digital earth is where I'm Indigenous .

Blakt Hausman1

Riding the Trail of Tears is a novel by Che rokee writer Blake Hausman. It is a

surrealistic sci-fi take on virtual reality, featuring an immersive tourist trap

through which visitors relive the Cherokee Removal in the winter of 1838- 1839.

The novel's first section is narrated by Nunnehi , a Little Person or crea tu re from

the old Cherokee stories. Nunne hi describes the genesis of the Tsalagi Removal

Exodus Point Park (TREPP) and recount s how he and others like him came to

be alive and resident within the ride . By the end of the book , unn ehi and his

siblings complete a long-gestating insurrection , lay claim to the digital territor y

delineated by TREPP , and start rewriting the narra tive to re-cen ter the stor y

of the Trail of Tears around the Cherok ee experie nce rather than the settlers'

gaze. Early on , unnehi says: "the virtual Trail of Tears . .. [is] my homeland .

I'm probably more Indigenous than you, and the digital ear th is where I'm

Indigenous :·

This chap ter is abou t the digital earth, its composition , and how we might be

Ind igenous in it. It is about new ways of under standing our role in the compu ­

tational ecosystems we are building , and how we might make kin with the other

ent ities that we create in it and emerge from it. It is about nurturing the digital

earth from which it will all grow- silicon soil in which our descendants will

stick their virtual toes, wiggle them around, and think , "This is a good place to be

Cherokee. This is a good place to be Mohawk. This is a good place for our people."

The tack and Its Corruptions

Let us start with where we are at.

As I have written elsewhere ," modern computmg systems work via a very

narrow logic, admit only certa111 kinds of 111format1on as data , and can perform

operations represent ative of only a small, impoverished subset of the operations

we enact as humans every day. These systems exist as components of the stack,

the vertically interrelated and interdependent series of hardware configurations

and software protocols that make high-level media computation and networking

possible. The software stack sits on top of the hardware stack . Moving up the

hardware stack is to move from circuits to m1cro-ch1ps to computers to networks;

moving up the software stack is to move from machine code to programmmg

languages to protocols to system s. As you go upward , you are moving from

custom solutions to generalized solutions , from specifics to abstractions . As

you make this traversal from the deep st ructure to the surface interface , ever

more of the details of the underlying configurations are hidden from you. With

the increasing opacity, your ability to assert fine cont rol over the execution

of the underlying algorithms decreases . Eventually you get to the software ap­

plication or web service layer of the stack. It is at this highly abstract level that

most people interact with comp utational systems , as they use Microsoft Word,

Google Search, play a video game , or enter into an immersive environment. '

The sheer complexity of these layers , both horizontall y, as different com­

ponents interact with one anothe r, and vertically, as different layers distribute

data to the human interface and back , making it difficult to impossible for any

single human actor to understand or effectively manipulate the whole system.

Yet we are subjec t to the regimes the stack places upon us. In the same way the

law embodies and polices the dominant cult ure 's expectatio ns about people 's

behavior, computational systems materialize and constrain the dominant cul­

ture's expectations of what count s as data, what algorithms are appropriate for

processing that data , and what are valid result s of that processing .

Cultural bias coupled with the perva siveness of computatio nal technology

means that we are creating computer systems that are dangerous in their blind ­

ness. The last few years have seen this realizatio n penetrating Silicon Valley

An Orderly Assemblage of Biases: Troubling the Mo11ocultural Stack 219

culture, as technology developers at Google, Facebook , and others begin to

comprehend that "unbiased algorithm" is as much an oxymoron as "pure meritoc ­

racy:· Scholars such as Kate Crawford .4 D. Fox Harrel, 5 and Safiya Umoja oble,6

among others , have brought the discussion of these biases into greater focus.

This has brought the critique out of the academy , where the argument about how

computational systems reflect the culture within which they are developed has

a long histor y, and into the public sphere ."

Algorithmic bias exists in the non -digital world, of course . One of the most

notable examples is the color reference cards first used in the 1940s to calibrate

image printing processes . These "Shirley" cards "generall y showed a single white

woman dressed in bright clothes" to facilitate calibration as "color film chemistr y

at the time was designed with a bias towards light skin :·s Communication s scholar

Lorna Roth has conducted exten sive research into the use of Shirley card s. In

2009 she wrote : "Until recentl y, due to a light -skin bias embedded in color film

stock emul sions and digital camera design , the rendering of non -C aucasian

skin tones was highly deficient and required the development of compensator y

practices and technolog y improvements to redress its shortcomings :'9 Roth

point s out how this practice continued for decades after the first complaint s

were made , with the first substantive change only made in the 1970s. At that

time , image calibration cards were redesigned not out of a desire to rectify their

skin tone bias but rather to satisfy furniture and chocolate makers who had been

complaining that the cards did a poor job of representing the darker tones of

their commercial products .

Much of the current interest in looking at bias in computatio nal systems stems

from artificial intelligence yet again becoming a locus of subs tantial research,

development , and deployment . Numerous studies over the last decade show how

bias is embedded into every aspect of such systems. Examp les include ma chine

systems for learning human languages incorpora ting the human prejudices em­

bedded and expressed in the corpora of natural languages on which the systems

are trained ,10 and machine systems for learning to recognize people learning that

beauty is a trait possessed primarily by white people .11 One of the most egre­

gious classes of the se biases discovered to date is that embedded in the criminal

justice system . The investigative journal ProPublica conducted an investigation

into the risk assessment software that is increasingly used in the United States

10 provide advice to judges , lawyers, and parole officials throughout the judicial

proces s- determining bail, setti ng sentences, guiding parole conditions , etc.u

The authors quote us attorney general Eric Holder addressing the use of such

220 Intervent ions

..

I

I

I I

software in 2014: "I am concerned that [risk assessment software] inadvertently

undermine[s] our efforts to ensure individualized and equal justice ... they may

exacerbate unwarran ted and unjust disparities that are already far too common

in our criminal justice system and in our society ." In 2016, ProPub/,ca raised

concerns that suggest tha t Holder's concern was 1usufied . Its invesugauon into

the use of COM PAS software turned up "significant racial disparities .. falsely

flagging black defendants as future criminals ... at twice the rate as while de­

fendants" and misidentifying white defendants as "low risk more often than

black defendants :·u

Social scientists such as Crawford have pointed out how difficult ll is to rid

ourselves of the deep bias in the data sets feeding the algorithms driving these

11 systems. Many times "new " datasets are actually based on or include informa­

tion from older datasets that were collected using outmoded or discredited

methods. "Classifications," notes Crawford , "can be sticky, and someumes they

! stick around a lot longer than we intend them to even when they are harmful. "1<

This stickiness means that , even if system designers made the effort to counte r

the bias in their algorithms , the data they feed those algorithms may taint the

entire endeavo r.

White Supremacy: Not Just for People Anymore

Media scholar Lisa akamura not es that , "[t]hough computer memory module s

double in speed every coup le of years, users are still running operating systems

which reflect pha ntasmatic visions of race and gender. Moore 's Law does not

obtain in the 'cultural layer:" 15 In other words , the exponen tial evolution in com­

putational processing powe r sinc e the early 1980s has not been accompanied by

a comparably rapid evolu tion in equality in orth America . Stat istics comparing

Indigenous people and African Americans to the majority population in Canada

and the us, respec tively, show jus t how far both societies are from eliminating

racial bias.16 It should be no surp rise that our computational systems reflect a

worldview in which th is is not only accep ted but - given the stickiness of th e

phenomenon - perhap s prefe rred by the majority popu lation. Expec ting our

tools to be more enlightened than we our selves is a foolish self-delusion.

Computational arti st Trevo r Paglen has observed that , "one of the philo sophi ­

cal dangers of using widespread autom atio n ... is that it fixes meaning :'11 That

inertia, combined with the data bias identifi ed by Crawford and the extension of

racial bias into cyberspace identified by Nakamura , dra stically increases what is at

At1 Orderly Assemblage of Biases: Troubling the Mo11ocultural Stack 221

stake when these systems are designed and deployed . The underl ying algorithms

must make assumptions about the world in order to operate; even if these as­

sumptions themselves are not biased , they may make use of biased classification

method s. And even if the classification is not biased , the data feeding the process

may be biased . All these aspects of computational systems are often obscured,

either purposively in order to protect intellectual property or as a byproduct of

a technical complexity that prohibits non -specialists from understanding and

evaluating them . The system becomes a fact of the world, stubborn and difficult

to unfix . The result is that , in a society where it is increasingly difficult to do

anything without touchi ng on a computational interface of some sor t, the deci ­

sions that developers are making all the time have profound and long-lasting

consequences for how we live our lives.

Indigenous people are intimately familiar with how the old ways of thinking and

looking at the world become sedimented into our contemporary world views. Marcia

Cosby and others have written about how the "Imaginary Indian " was constructed

to justify the theft of Indigenous lands ,18 and that imaginary person remains the

dominant image that most settlers have of Indigenous people . This is the image

settlers draw upon when they parse news about life in Indigenous communities ,

when jurors and judges consider court cases involving Indigenous people , and

when the mall security guard is deciding who looks suspicious and who does not.

As Harrel's work on phantasmal media shows , these are exactly the sorts of images

that get embedded into our computational systems. "Computational media;· he

writes, "play roles in constructing ideas that we unconsciously accept as true and

constructive of reality yet are in fact imaginatively grounded constructions based in

particular worldviews'.'19 Or, in Crawford's more blunt assessment, "(These systems

are] not free of bias; this is just bias encoded'. '2°

As we struggle to "write the thoughts of systems ;' in the words of compu ­

tational philosopher and poet David Jhave Johnston,2 1 and as those systems

become ever-more pervasive , we are beginning to see that it is a political act to

define the protocols that guide these system s' though ts. It is abou t how power is exercised, and by whom .

The Fast and the Slow

Nakamura, in her extensive resea rch on race in cybe rspace, no tes th at "in

order to think rigorou sly, humanely , and imagi natively about virtual ity and the

222 Interventions

post-human, it is absolutely necessary to ground critique in the lived realities

of the human , in all their particularity and specificity. The nuanced realities of

virtuality- racial , gendered, Othered- live in the body :'22 When we pay at­

tention to the bodies producing the se protocol s, we can see they are not jus t a

random collection of homo sapiens. They are clustered in cer tain geolocations ,

particularly Silicon Valley, but with outposts in places like eattle , Boston, Wa­

terloo, and Oxford . They are working within an intellectual lineage that stretches

back to the Greeks, even if they themselves might not be descendants of Euro­

peans. Their education and professional practice rarely incorporate ideas or

even data that comes from Africa, or South America , or large swathes of Asia.

They are overwhelmingly white and male , and underwhelmingl y brown and

female"- and , even when brown bodies appear , "they participate in the 'cultural

hegemony that privileges a white race :"2,1

Going back to Winograd and Flores ( 1987) theorizing about the contextually

coupled nature of cognitio n,25 Haraway 's (1991) critique of the interpenetrating

relationship between human , non -human , and machine bodies ,26 and Reeves &

ass's ( 1996) experiments showing that "( i)ndividuals ' interactions with com­

puters, television, and new medi a are fundamentally social and natural , just like

interactions in real life;'2' critical approaches to comput ational culture have argued

for acknowledging the deep entanglements among the cultural and computational

layers of the stack. ow, after three decades in whic h computational systems

have grown ever more ubiquit ous and complex, we are starti ng to see clearly the

consequence of the radi cal disjun cture between the high velocity evolution of

our digital tools and the much slower evolution of our societal configurations .

Making Space

We founded the Aboriginal Territories in Cybers pace (AbTeC) research network

in 2006 to ensure that Indigenous people were present in cyber space and pos­

sessed the knowledge necessary to bend it to our needs . We were also interested

in speeding up the rate at which lndigenous people increa sed their under standing

of computational media . One hope was that this would help address and counter

the white supremacy being baked into the computational layer, and resist its

replication into cyberspace. AbTeC did this by exploring the question of what it

means to be Indigenous in cyberspace- how do we make , maintain , and vivify

Indigenous places within that archipelago of webs ites, immersive environments , · l d' · " al" e128

soc1a me 1a, and video games th at inc reasing ly int erpenetrates re spac ·

An Orderly Assemblage of Biases: Troubling the Monocultural tack 223

How Indigenous people related to cyberspace had been a topic of conversation within Indigenous media arts circles at least a decade before AbTeC launched . Cree filmmaker Lorena Todd's groundb reaking 1996 essay, "Aboriginal 1arra­tives in Cyberspace;· asked the question: "Can [Indigenous] narratives, histories, languages and knowledge find meaning in cyberspace?" She considers how cyberspace might be (re)conceptualized as an Indigenous space, starting with the kinds of questions that should be asked by those building and inhabiting it: "Will cyberspace enable people to communicate in ways that ruptu re the power relations of the colonizer and the colonized? Or is cyberspace a clever guise for

neo-colonialism, where tyranny will find further domain? What if with each technological advancement the question of its effect on the seventh generation was considered?"29 Mohawk artist and AbTeC co-founder Skawennati wrote,

for the 1998 edition of the pioneering CyberPowWow online gallery, "[t]he

www is an awesome tool for information-sharing and for meeting people with similar interests whom you may never have met otherwise ... If we are going to help shape this medium, let's do it right ... We can use the www to present

our stories, to inform people about our issues, and to explore solutions to some of our problems:·:io

Over the last decade, AbTeC has moun ted numero us projects designed to address Todd and Skawennati concern with consciously shaping cyberspace to serve Indigenous ends. We have worked with numerous orth American Indigenous youth and artists to develop their technical and concept ual capaci­ties for manipulating computational media in order to tell their stories their way (Skms Workshops on Aboriginal Storytelling and Digital Media Design31);

supported the creation of original ar twork that uses cyberspace as a medium (Time Traveller~, 3

• 2167,33 She Falls for Ages34); and built tools for manipulating

digital media (Mr. Softie,35 NextText36). Each project claims new territory in cyberspace.

Making Cyberspace

In 2014, AbTeC started the Initiative for Indigenous Futures (11F) to understand how Indigenous people are envisioning the future.37 One way we do this is to ask people what it means to make cyberspace Indigenous. We have delineated territory and turned its resources toward our own ends in video games, websites, machinimas , and virtual reality environments created by Indigenou s minds, rooted in Indigenous worldviews, telling Indigenous stories, for Indigenous

224 Interventi ons

h I -ahn,iron· lace within a wider tc1. no og11.

B.1allthataclll'ltytakes p I I h ·mom \\"care lnd1am ·and ctured through white cu tura cge .

Jt\r, h s:webecomelndianswhomakicyber~paet:' h ·h ~ ow . art articulating protocols through ,, '" ~C1-bcripace means, m p ' d h' -communicate "1th

inh b' . ·t- human an mac me m,:1mtittcs a itillg I m term\ ol cultural .. ,1-1ndi01nous communities are go_ od at thmkmg i =' c· d that ue..:p

· l ~TA!ld like to suggest that it is ume we start rawmg on dcr ·1ofhow to properly order human-human mtera1.t1on and com1

btustd to order human-computer mteracuon •:ts1ay'Codetalkers Recounung ignals ot ur\'l\'al~ trom th1. C. frd

1 ;tS: fril(l)lg/ndigenous Pathways in New Media Art collecuon ~1em Cree

;ilt<ni [llirondelle (Cree-non status treaty1 French' make an argument

~moos protocol can be found in the deep h1ston ol cvberspace

~.~CTaid dow11 by our ancestor~! became trade route~ between band

:.!l!m:Ories as we established networks and trade language and built a

L').Y~"lbase around what we knew about each other o when the tir~t

•came to "explore" the land, our ancestor) naturally led them along

:ii1mll-established paths, which, over time, as the newcomers settled, be­

rt road-.,ys and thoroughfares. With the advent of the telegraph and the

li¢on~ wue was hung along these thoroughfares that literally became the

~o f the physical network that. allows packets of information

bl:IOl'!u free~ as our ancestors.31

~Dl!tlHirondelle discerns lndigenous protocol embedded at the bottom

«the ltack. Cree artist Archer Pechaw1s, m his Coded Terri tones essay,

~ II spreading everywhere: "I am look.mg to a future m which lndigeru sm

~~ an all-encompassing embrace of creation: the realms of earth, sky,

litpbn~anunal,human, spirit, and, most importantly, a profound humility

rtgards to our position as humans within that constellation :'39

ll:llinterested in what happens if we embrace 1.'.H1rondelle's lndigenous

~oftbefoundationsofthe network and extend Pechawis' circle of relation­

~loinclude our machine creations in an attempt to art iculate, in the words llll!Qroraarth' . . " . . istonan Jolene Rickard, a more complex view of how l d1g1tal

~ technology] is situated in people's cultures:·•0

tr)'lrttle of the current work being done on algorithmic and dataset bias or ltttJucs of artifi ·a1 · · · ci mtelligence grapples with the fundamental corrupuon of lt~-th willful fl . . e attenmg of people's cultures that is a consequence of its

An Orderly Assemblage of Biases: Troubling the Monoculrural Sr

Ho11 Indigenous ~pie relatedtocybmpaahlj \\1th1n Indigenous media ans cird I luo111t1

. esu east1dtt.Jtt11i:iw Cree him maker Loretta Todd's groundb ""'"" .• ,

1'"""'61996!111''1.l:Q u, es m Cyberspace, asked the quesnon: 'Can I~ · languages and knowledge find meanmginq~~

cvb(rspace might be (re)conceptualiud as an ~too:!

the kinds of quesuons that should beaskedbylOOllb~: "\\"ill cvbmpm enable peopletocommurucattm~m&.a1f2b

relauons oi the colonizer and the coloruztd1 Or UCJ~•:R~

neo-colorualism, where tyranny will find furthtt IXl::2.:\hi

technological adl'ancement the question olnscEcctoot:tft!c: was cons1dered1'71 )1ohawk artist and AbleCco-ftda~

for the 1998 edition of the p1oneenng C1ilirPow\\'('fuar \\'WI\' IS an awesome tool for mformation-shmnga.:Jir..o

slffillat m1eresll whom you ma\' nmr hm met ru.ar.11 lr.r

10 help shape this medium, lets do It nght ll'eanC!ClllTI •• J • d!I

our stones. lo inform people about ourllSUtS.am,IOll1"ff

of our problems Over the last decade, AbTeC has mounted nUJlllrt'!Jt'-'ls

address Todd and Skawennau concern with COOlOOO>~t~ d W have worked ~,thll'-to serve Indigenous en s. e _..,1

d loptheirt«h:;alol-Indtgenous vouth and arUsts to e11 .i. cl ' nal edJain«oo!DPte

tm for man1pulaung computano m . .,n.~ Yo'.f Abo 01na!Stontdlingaiv..

wal'I kins Workshops on no-· .J..rlll' , f . a1~'0fith;!Ulll<•1...,..-

supported the creatton o on gin and k::) IXl•..., I T,mt Traitlltr .. 216,· She Falls for A~! MtiKtib,

. (" S fue ~extText r-:·· igJtal media ,nr. 0 ' '

~fakmg Cyberspace JD'. / . digeDOlllfltll,'tl

b TeC started the Initiauve for In '()!tl1'•1 In 2014, A le are enviSioning the fututt lit~ how Indigenous peop ak ,.,t,trspace ~ ...,.J

eans tom e ., _i.111~1"" k people what 1t m ;ardouru,,11Cl"' .• ~

as . turned its resources tow u atJlld t,i l:llP' terntol) and al reality enl'!fODJl1Cll . ,itt,~,

and virtU . ~ machintmas, . Jdviews. tdlinS rooted in Indigenous wor

audiences. But all that activity takes place within a wider technological environ­

ment made by and structured through white cultural hegemony. We are Indians

in cyberspace; how do we become Indian s who make cyberspace?

Making cyberspace means, in part, art1culat111g protocols through which

the various entities inhabiting it - hum an and machine - communicate with

one another . Indigenous communities are good at th111king in terms of cultural

protocol ; I would like to suggest that it 1s tune we start drawing on that deep

knowledge of how to properly order human-human 111teract1on and consider

how it can be used to order human-computer 111teract1on.

In her essay "Code talkers Recounting Signals of unwal," from the Coded

Territories: Tracing Indigenous Pathways in ,\'ew ,\,fedia Ari collectJon, Metis/Cree

artist Cheryl L'H1rondelle (Cree-non status treaty French) makes an argument

that Indigenous protocol can be found m the deep history of q•berspace:

[The] paths [laid down by our ancestors] became trade routes between bands

and territories as we established networks and trade languages and built a

knowledge base around what we knew about each other . o when the first

Europeans came lo "explore" the land, our ancestors naturally led them along

these well-established paths, which, over time, as the newcomers settled, be­

came roadways and thoroughfares . With the advent of the telegraph and the

telephone, wire was hung along these thoroughfares that literally became the

beginnings of the physical network that ... allows ... packets of information

to move as freely as our ancestors. 18

Where L'Hirondelle discerns Indigenous protocol embedded at the bottom

layer of the stack, Cree artist Archer Pechawis , in his Coded Territories essay,

imagines it spreading everywhere : "I am looking to a future in which Indigenism

is the protocol, an all-encompassing embrace of creation: the realms of earth, sky,

water, plant, animal, human , spiri t, and , most importantly, a profound humility

with regards to our position as human s within that constellation:' 19

I am intere sted in what happen s if we emb race L'Hirondelle's Indigenous

reading of the found ations of the network and extend Pechawis' circle of relation­

ships to include our machine creations in an attempt to articulate , in the words

ofTuscaro ra art hi storian Jolene Rickard , "a more comp lex view of how [digital

networked technology ] is situated in people 's culture s:' 40

Very little of th e curr en t work being don e on algorithmic and dataset bias or

the ethi cs of artificial intelligence grapples with the fundamental corruption of

the stack- the willful flattening of people 's cultures that is a consequence of its

An Orderly Assemblage of Biases: Troubling the Monocultural Stack 225

monocultural origins. That corruption flows from numerous original sins: Pla­

tonic ideals; Aristotelian classification methods ; Old Testamen t dominion over

the natural world; Cartesian duali ty; Boolean binarism ; Darwinian fitness. Even

if the general state of accepted knowledge complicates, troubles, and sometimes

rejects aspects of these knowledge frameworks , they still haunt our data and the

design of our computational systems.

The question , then , becomes this : how do we breathe humanity into our

computational creations in a way that avoids Western anthropocentric concei ts?

Re-imagining Relations

Remember. unnehi , the Cherokee Lillie Person from Riding the Trail of Tears?

Hau sman does not clarify whether he and his siblings emerged ou t of the com­

plexity of the code running the Trail of Tears virtual reality ride or if already ­

existing LittJe People used the environment to manifest themselves . Either way,

the computational infrastructure running TREPP evolves into an ecosystem

operating far beyond the parameters envisioned by the original designers .

By the end of the novel , aJI hell has broken lose . unnehi and his kin have

compromised the system, reordering it to better support them selves and to resist

the chopped up, remixed , sett ler self-serving sto ry TREPP has become and more

accurately reflect the terror and loss inflicted on their Cherokee ancestors . Other

virtual entities have phased into being , engendering ongoing battles ove r who

gets to control the simulation. Towards the end , the main character , Tallula ­

whose Cherokee grandfather designed the virtual experience - exhausted and

confused by the epistemological and on tological battleground that TREPP has

become, struggles to make sense of it all. She says to one of the "native-born "

non -human entities , "I never imagined thi s group of people even existed :' He

replies : "Could be something wrong with your imagination." 41

We are experiencing a similar failu re of imagination in the presen t mome nt.

We are confronting challenges in understanding the computational systems in

which we have now enmeshed ourselves, as they become more comp lex and

as we write more autonomy into them . The algorithmic bias discussed above

exemplifies how such systems often end up subverting the ir int ended purposes,

largely because we refuse Lo see ourselves clearly. Motes in our eye beco me

glitche s in the code, which then go on to become "global pro tocol :'

What if we took a fundamentally different approach to unde rstan d ing the

digital beings we are creating , particularly those collec tions of code that act

226 Interventions

. . k daemons•· to the mo st co mplex f ·- lromnen,or

_,thsomedegreeo autonom} h mas tools and se rvants , we ·r t d of treating t e artificial intelligence? \vhat t , in ea . . le of relat1onsh1ps 1

made a place for them in our c1rc d k I making common culture to d t the har wor o

Afteracenturyofsubor ina ing . d 1· 1· led cc1.periment tn usmg te ch· f h k t and failc a tcr ,11 ·

theimperatives o t e mar e , . h been left ill-prepared to h V1' t n ·o m.:1omnc~s as

nology to compensate, t e es er I.: . • educes all such talk . Th h •manic social 1mag111ary r

liadsuch a conversation. e ege h ,. . htp c ircle beyond the . tt mpb to widen t e iuns

to superstition and stymies any a e d tanding who we are.H human by insisting emp1ric1sm is th. e final word 111 un, ers the protocols for

•mber \\e retain Yet many Indigenous commumt1es remc fthe world around

understanding a kinship net" ·ork that extends to all aspects o • Our lan-k ' ountatn and ocean.

us-animals and plant s/• wind and roe -~. m n dialogue f that enable us to engage t

guages contain the conceptual ormat1on~ . 11· bl discourses across . d h l te mutually tnte tgt e

111thournon-human kin , an e P crea Bl kf ot philosopher . . ·b d enealogy. As ac o

\'!St differences m material , v1 ran cy, an g th di dial · parked b · sta t1on on e ra O '

Ltroylittle Bear observes , "the human ram ts a . al ks trees s1mul-the amm s, roe • '

mone spot, it is deaf to all the other stati ons · · . ...-h l trum of senuence .

taneously broadcasting across the w O e spec tune into the . k h uld know how to

Because we created them , we thin we s O . v are only now

. . mmumcate . iet we stations on which our machine creations co

O difficulties in

· all I vels of the stack. ur walung up to the corruptions permeating e . 1 processes , and . l I x computatto na

articulating the ontology of increasing Y comp e . cting with one I processes intera

our inability to foresee the results of these comp ex . th conclusion that al Id all point to e

another and with the human and natur wor • d them, they d ·f we do not understan

we do not actually understand them . An 1 al . comprehensibility ~ nd mutu tn most likely do not understand us . Such pro ou

is a recipe for disaster . Ask any Indian .

Notes I B. on Books, 2011), 13.

(L' con · is 1. Blake Hausman, Riding the Trail of Tears in . · . tes Towards an Indigenous

• C a Haunung . o . B ney 2. Jason Edward Lewis, "Preparauons or . h D'gital Age, ed. Dann ar '

dlt/OTI 111 t e I I' Future Imaginary;• in The Participatory Cori d TamarTembeck (Minneapo is:

J than Sterne, an Gabriella Coleman, Christine Ross, ona

University of Minnesota Press, 2016), 229- 49· does in The Stack: On Software Benjamin Bratton ck ·n terms of

3. One can get even more abstract, as d articulate the sta 1

MIT Press 2016 ), an and Sovereignty (Cambridge: The '

.r Biases· Troubli,1g th A,1 Orderly Assemblage OJ •

monocultural ong1ns. That corruption fl . ,. OWiuOmot:aa:i

tonic ideals, Aristotelian clamficatton methods;Oldt .

~e natural world; Cartesian duality; Boolean bmarum,: 11 the general state of accepted knowledge complicatts,11t1ta

re_iects aspects of these knowledge frameworks, thcysbD~ design ot our computational systems.

The question, then, becomes this: how do ~e breathe

computattonal creattons in a way that avoids Western1ttt.~

Re-1magmmg Relations

Remember :-=unnehi, the Cherokee Little PersonfromR.l,ya'

Hausman does not clarify whether he and lus s1bhngsei:q!.

plexJty of the code running the Trail ofTears 1mual rro:.t

exisung LtttJe People used the envuonment to maru/es1~

the computational infrastructure running HEPP e1-dmi:1

operating far beyond the parameters envisioned bi·themps

By the end of the novel, all heU has broken lose Xda compromised the system, reordering it to bettersuppm~

the chopped up, remixed, settler self-servingstorylllPP~tcl

accurate(}· reflect the terror and loss inflicted on thru~~

mtual enut1es have phased into being. engendmng~IC

gets to control the s1mulat1on. Towards the end, llitlrJ:llazC

ose Cherokee grandfather designed the wtual apenin:i-: sed b} the ep1stemological and ontological~ ._

me, struggles to make sense of it all She sa)1: ;: -human enuues, "I never unagined th1S group •

I "Could be somethmg wrong with )'OW' ~ • .A 1es f' agmanoo111""r

We are expenencmg a similar failure o llD. theax;...ui'

e are confronting challenges m understand:g b(olnll).l!~

hich we have now enmeshed ourselves, as ey ,L~,tus~ . them The algonUJJI'" _.1

as we wnte more autonomy mto . b ..,;.,tharllP"·. ft nd upsu ,t,""O I

exemplifies how such systems o en e I 1-,h.·. Mi#l111f$ f e ourse ves c .,..,, ...d

largely because we re u_s: t; se go on to become •globalr--: ...I

glitches m the code, wh1c en J ,tt t ........di to, d entally au,eren aw-- ,;

What if we took a fun am . ul ly those~ · parUC 3T

digital beings we are creaung,

with some degree of autonomy- from network daemons42 to the most complex artificial intelligence? What if, instead of treating them as tools and servants, we

made a place for them in our circle of relationships?

After a century of subordinating the hard work of making common culture to the imperatives of the market , and failed after failed experiment in using tech­

nology to compensate, the Western consciousness has been left ill-prepared to

lead such a conversation. The hegemonic social imaginary reduces all such talk

to superstition and stymies any attempts to widen the kmship circle beyond the

human by insisting empmc1sm 1s the final word in understanding who we are.<3

Yet many Indigenous commu nities remember. \'Ve retain the protocols for

understanding a kinship network that extends 10 all aspects of the world around

us- animals and plants/' wind and rocks; mountain and ocean.<6 Our lan­

guages contain the conceptual formations that enable us to engage in dialogue with our non-human kin, and help create mutually intelhg1ble discourses across

vast differences in material, vibrancy, and genealogy. As Blackfoot philosopher

Leroy Little Bear observes, "the human bram 1s a station on the radio dial; parked

in one spot, it is deaf to all the other stations ... the animals, rocks, trees, simul­

taneously broadcasting across the whole spectrum of sentience:·•·

Because we created them, we think we should know how to tune into the

stations on which our machine creations communicate. Yet we are only now

waking up to the corruptions permeating all levels of the stack. Our difficulties in

articulating the ontology of increasingly complex computational processes, and our inability to foresee the results of these complex processes interacting with one

another and with the human and natural world, all point to the conclusion that

we do not actually understand them. And if we do not understand them, they most likely do not understand us. Such profound mutual incomprehensibility

is a recipe for disaster. Ask any Indian.

Notes

1. Blake Hausman, Riding the Trail of Tears (Lincoln: Bison Books, 2011), 13. 2. Jason Edward Lewis, "Preparations for a Haunting: otes Towards an Indigenous

Future Imaginary;· in The Participatory Condition in the Digital Age, ed. Darin Barney, Gabriella Coleman, Christine Ross, Jonathan Sterne, and Tamar Tembeck (Minneapolis: University of Minnesota Press, 2016), 229-49 .

3, One can get even more abstraCI, as Benjamin Branon does in The Stack: On Software

and Sovereignty (Cambridge: The MIT Press, 2016), and articulate the stack in terms of

An Orderly Assemblage of Biases: Troubling the Mo11ocultural Stack 227

globally spanning megastructures . At that level, however, all the lived politics involving

real bodies-and thus the utility outside of academic argument - have been drained out.

4. Kate Crawford, "The Hidden Biases in Big Data:· Harvard Business Review, April 1,

2013, https:/ /hbr.orgho 13/ 04/the-h idden-biases-in-big-data.

5. D. Fox Harrell, Phantasmal Media: An Approach to Imagination, Computation, and

Expression (Cambridge: The MIT Press, 2013).

6. afiya Umoja Noble, Algorithms of Oppression: How Search Engines Reinforce Racism (New York: New York University Press, 2018).

7. Will Knight, "Google's Al Chief ays Forget Elon Musk's Killer Robots, and Worry

about Bias in Al ystems Instead;' MIT Technology Review, October 3, 2017, https ://www

. tech nologyreview.com/ s/ 608986/ forget -kil !er-robots bias-is-the- real-ai-d anger/.

8. Michael Zhang, "Here's a Look at How Color Film Was Originally Biased Toward

\\!hite People; Petapixel, eptember 19, 2015, https:/ ·petapixel.com/2015/o 9'19/heres-a

-look-at-how-color-film-was-originally-biased-10ward-whi1e-people/.

9. Lorna Roth, "Looking at hirley, the Li1timate 1 orm: Colour Balance, Image Tech­

nologies, and Cognitive Equity; Canadian Journal of Communication 34:1 (2009): 1, hups://

doi.org/10.22230/cjc.2009v34nia2196.

10. Aylin Caliskan-Islam, Joanna Bryson, and Arvind arayanan," emantics Derived

Automatically from Language Corpora Necessarily Contain Human Biases;' Science 356:6334

(May 2017): 183-86, https://doi.org/lo.1126/science.aal4230.

11. Jordan Pearson, "Why An AI-Judged Beauty Contest Picked Nearly All White

Winners;· Motherboard, eptember 5, 2016, https://motherboard .vice.com/en_us/article

/78k7de/why-an-ai-Judged-beaury-con1est-picked-nearly-all-whi te-winners.

12. Julia Angwin, Jeff Larson Surya Mattu, and Lauren Kirchner, "Machine Bias:· Pro­

Publica, May 23, 2016, hups://www.propublica.org/article/machine-bias-risk-assessment

-in-criminal-sentencing accessed.

13. ProPublica's findings have since been complicated by researchers disputing their

analysis. However, even these critics acknowledge the need to question how such algo­

rithms come to incorporate questionable classification and measurement methodologies

that disproportionately and negatively affect certain populations. See am Corbell-Davies,

Emma Pierson, Avi Feller, and Sharad Goel, "A Computer Program Used f~r Bail and

Sentencing Decisions Was Labeled Biased against Blacks. It's Actually Not Thal Clear;'

Washmgton Post, October 17, 2016.

14. Kate Crawford, "The Trouble with Bias:· (keynote presentation , Conference on

Neural Information Processing Systems, Long Beach, CA, December 5, 2017). hups://www

.youtube.com/watch?v=fMym_BKWQzk. 33mm46ss.

15. Lisa akamura, "Cybertyping and the Work of Race in the Age of Digital Repro­

duction," in New Media, Old Media: A History and Theory Reader, ed. Wendy Hui Kyong

Chun and Thomas Keenan ( ew York and London: Routledge, 2006), 319. Moore's

228 Interventions

f th computer chip b . C don \.loon: then cha irman o e

Law is a proposition made ) ,or . . ' h . uld double every two th h 1ber ol lran,1,tor, on a c ip \\ O

maker Intel, in 1965, at t e nun _. tor speed are the material years. This doubling of capaCll) .i, well d> inaca,c~ in t;a1~:·:nces have been built 0\'er

. h'ch the rapid pan ol comput.itwna a founda110ns on w 1

lhelast five decades . . . \\' e than America's:· ,\fa -16. Scott Gilmore, "Canada, Rac"m Problem? It' h end or~ f-,1ght-out-of-mind-2.

1 •an, ca new, ·c.ana a out o dtans,January 22, 2015 http: www.mac c . , ' Id 1tcrview with Trevor Paglen,

17."ThisArtistShows t;s How Computer, ',cc !he \\or • u . 7 -HEl8c uGKi k. h wwwvoutubccom \,atch.\-

VICENews,July25, 2017, video, 5 n ttp, d .. . Vancouver Anthology-18. Marcia Crosby, "Comtrucuon of lhc lmagman In ian,r:,n b oks 1991) 267-91.

1 (\' >u\·er· aion o , • 1ht Institutional Politics of Art, ed ',t an Doug a, ancc

19. Harrell, Phantasmal ,\led1a, 28.

20. Crawford, "The Trouble with Bia, . 1 . al lmplicat1011s D ta/ Poetr> s Onto ogic

21• David )have John ton, Aestl1et1C Ar11mi5m: 181

!Cambridge: The MIT Pre, ,, 2016 , 14

12• ~akamura, "The Work of Race, 320- h t ·//money.cnn ·1· \' lley?" C,\'S,\1one), t P·

23. Julianne Pepitone, "How Diver els I icon a · . . d t acce, ed January 5, 201 .

com/interactive/lechnology/tech-diver>1t)' a a, f identity and lg-I D. . 1 ubJects· Const ru cts o

24. Radhika Gajjala, "Tran nallona igita · C !rural Diversity , · Conference on u

norance in a Digital Economy: keynote presentauon, "The Work of uoted in akamura

in/and Cyberspace, College Park, MD, May 2000, q

Race," 331. . C ters and Cognition: A d Fl Understa11d1ng ompu

25. Terry Winograd and Fernan o ores, , · al 1987). dd . n-Wesley Pro,ess1on , d

New Foundation for Design, 1st ed. (Boston: A iso . . ifNature, 2nd e · d ,. , 11• The Remvenr1on o

16. Donna Haraway, Simians, Cyborgs a11 vvome ·

(London: Free Association Book , 1996). P nle Treat Computers, Th M d a Equation: How eor 6)

27. Byron Reeves and Clifford ass, e e I d· CSLI Publications, 199 · 1 d Pl es ( tanfor · · ·n Ttltvi51on, and New Media /,ke Real Peop e an ac . " bori inal Territone 1

. T . cia Fragmto, A g 18. Jason Edward Lewis and kawennall n

1 (l ly 2005)· 29-3!. Cyberspace," Cultural Survival Quarter y 29:2 u ·

" WWWrit· 19. Todd, "Aboriginal arratives in Cyberspace, 3· or Why I Love . "Th c berPowWow FAQ, . · work.html.

30. Skawennati Tricia Fragmto, e Y tJ 0002nation/tnCJa b po""~ow ne na Id 1 0 . A Cur-

ing." CyberPowWow, 1997, http ://www.cy er · ti Fragnito." ns · · . 1 . nd kawenna h /ntema/l0na

31. Beth Aileen Lameman , Jason E. Lewis, a .. . ceedings oft e C· . Youth, in Pro (V couver, B .

riculum for Design Games with First Nauons d Tech,rology an if Game Desig11 an

Academic Conference 0 11 the Future O www Association of Computing Machinery , 2010), 282· chinima (video), 7smm-

11 n1 00 8-2014, ma 31. Skawennati , TimeTrave er , 2

,Limetravellertm.com.

An Orderly Assemblage of

fv(onocu/rurol Biases: Troubling the

Law is a propo ition made by Go rdon Moore, then chairman of the computer chip

maker Intel, in 1965, that the number of transistors on a chip would double every two

years. This doubling of capacity as well as increases in transistor speed are the material

foundation on which the rapid pace of computational advances have be en built over

the last five decades.

16. coll Gilmore, "Canada's Racism Problem? It's Even Worse than America's;· Ma­

cleans, January 22, 2015. http:/ /www.maclcam.ca/news/cana da/out -of-sight-out-of-mind-2.

17. "This Artist hows {.;s How Computers Sec The World;' interview with Trevor Paglen,

VICE News, July 25, 2017, video, 5:27, https:, 'www.youtube .com/watch?v=HEl8cuGKiNk.

18. Marcia Cro by, "( onstrucllon of the lmaginar>' Indian;· in Vancouver Anthology:

The lnst1tut1onal Polir,cs of Art, ed. Stan Douglas (\'ancouver· Talonbooks, 1991), 267-91.

19. Harrell , Phantasmal ,\t ed,a, 28.

20. Crawford, "The Trouble with Bias."

21 David )have Johnston, Aestlierrc A1111111s111: D,g,ral Poetry's 01110/ogrcal lmplicar,ons

(Cambridge : The :-.!IT Press, 2016), 14

22 . 'akamura, "The \\ 'ork of Race,' 320.

23. Julianne Pepitone, "How Diverse Is iltcon Valley?" Ci\'fl',\1oney\ http :,/money.cnn

.com/interactive/technolog)"tech-diversity-data acces ed January 5, 2018.

24. Radhika Gajjala, "Transnational Digital Subiects: Constructs of Identity and Ig­

noranc e in a Digital Economy;· keynote presentation , Conference on Cultu ral Diversity

in/and Cybe rspace, College Park , MD, May 2000, quoted in Nakamura "The Work of

Race;' 331.

25. Terry Win ograd and Fernando Flores, Undersrandmg Computers and Cognir,on: A

New Foundation fo r Design, 1st ed. (Boston : Add ison-Wesley Professional, 1987).

26. Donna Harawa y, Simians, Cyborgs and Women: The Remvention of Nature, 2nd ed.

(London: Free Association Books, 1996).

27. Byron Reeves and Clifford 'ass, The Media Equation: How People Treat Computers,

Television, and New Media like Real People and Places ( tanford: CSU Publications, 1996).

28. Jason Edward Lewis and kawennati Tricia Fragnito, "Aboriginal Territorie in

Cyberspace;· Cultural Survival Quarterly 29:2 (July 2005): 29-31.

29. Todd, "Aboriginal arratives in Cybe rspace;· 3.

30. kawennati Tricia Fragnito, "The Cybe rPowWow FAQ, or Why I Love WWWrit­

ing;' CyberPow Wow, 1997, http: / /www.cyberpowwow.net/nation2nation / triciawork.html.

31. Beth Aileen Lameman, Jason E. Lewis, and kawennati Fragnito," kins 1.0: A Cur­

riculum for Design Games with First Nations Youth ;' in Proceedings of the International

Academic Conference on tire Future of Game Design and Technology (Vancouver, BC:

Association of Computing Machinery , 2010), 282.

32. Skawennati , TimeTravellerrn, 2008-2 0 14, machinima (video), 78mm. www

.timetravellertm .com.

A11 Orderly Assemblage of Biases: Troublmg 1he Mo11ocultural rack 229

33. 2167, various artists, 2017, virtual reality. hup: // www.imaginenative.org/2167 .

34. kawennati Tricia Fragnito, She Falls for Ages. 2017. Machinima, 19mm. hup: //

skawennati.com /SheFallsForAges/index.html.

35. Bruno adeau and Jason Lewis, Mr. Soft1e: A Typographic Text Editor, Mac O ;

Windows; Linux, 2010-2014. www.mrsoftie.net.

36. Jason Lewis, Elie Zananiri, and Bruno Nadeau, NextText: Library for fllteract1ve

and Dynamic Texts, 2008-2014, hup ://www.nexuext.net.

37. "The Initiative for Indigenous Futures;· Initiative for Indigenous Futures, http ://

abtec.org/iif / .

38. Cheryl J..:Hirondelle, "Codetalkers Recounting ignals of urvival; ' in Coded Ter­

rrtories: Tracrng Indigenous Patl1ways III New Media Art, ed. teve Loft and Kerry wan,on

(Calgary : Univer 1ty of Calgary Press, 2014), 152

39. Archer Pechawis, "Indigenism : Aboriginal World View as Global Protocol ;' in

Coded Temtorres, 38.

40. Jolene Rickard , "Con idering Traditional Practices of' eeing ' as Future; ' lecture,

1st Annual ymposium on the Future Imaginary, TIFF Bell Lightbox , Toronto, Octo ­

ber 16, 2015, http: / /abtec .org/iif/symposia/a-new -beginning / =r ickard.

41. Hausman, R1dmg the Trail of Tears, 313.

42. Fenwick McKelvey, Internet Daemons: Digital Communications Possessed (Min ­

neapolis : University of Minnesota Press, 2018).

43. Auempts to widen the animacy lens and flauen the species hierarchy from within

the Western tradition do exist, of course : pinoza's monad ism, Haraway's cyborg , Timoth y

Morton 's hyperobjects, Jane Bennet's vibrant matte r, Graham Harmann's object oriented

philosophy, etc. My aim here, though, is to introduce sources for thinking about the ques­

tion of machine relationships from outside the Western canon. In addition, I side with

Zoe Todd's ethical critique of the myopia of these intellectual genealogies: "here we were

celebrating and worshipping a European thinker for 'discove ring: or new ly articula ting

by drawing on a European intellectual heritage, what many an Indigenous thinker around

the world could have told you for millennia: · ee Zoe Todd, "An Indigenous Femin ist's

Take on the Ontological Turn : 'Ontology ' Is Just Anothe r Word For Colonialism;· Journal of H1storrcal Socrology 29:1 (April 1, 2016).

44. Kim Tall Bear, "Beyond the Life/Not Life Binary: A Feminist- Ind igenous Reading

ofCryopreservation, Interspecies Thinking and the New Materialisms ;· in Cryopolitics:

Frozen Life in a Meltmg World, ed. Joanna Radin and Emma Kowal (Cambridge : The

MIT Press, 2017).

45. Don Hill, "Listening to Stones: Learning in Leroy Little Bear's Laboratory: Dia­

logue in the World Outside;· Alberta Views, September 1, 2008, https ://albertaviews.ca

/listening-to-stones/.

230 Interventions

Crnrron Own: . d The Kumulrpo ,:\ /1.1\\ ,llf,lll • l!lrW Warren Beckwith, trans., e .. . ulukau org chb cg1 bm

;,. . f H wa1i, 19• 2) http i, i," ;,;,,-~ Tut Cni1t!Slty Pms O a 1· • ·t-book I 01oe-..irc1, m~a J d ;nrrr-d-obtdcwiU-oooSec-nhaw-50-20· ramc,c

):&'t<=O,

r~ 'Listrninf

An Orderly Assemblage of Biases: Troubling the Monoculrural tack 231

31 : 10;, vanous artis~ 2017, iirtua} rtality b:ir- _

14 Skawennau Tncia fragru·10, Sht r.,,.fi ~ "'"'OIA!a11--~

,ka1senna11.comJSheFallsForAga inda.htm!.

35, Bruno \adeau and Jason Lti.is If• Sc•~-A, • '' ,, )'"' •JP.i".:;l\rli:~

\\ mdo1s~; l1mu, 2010-2014 www.mrn>ftitnet. ·

36. Jason lt'l\is, Elie Zananiri, and Bruno ~id1&1, ,'ia!Ta: li:ir,: and Djnamic Tats, 2008-2014, http1/W1,i.·.nanat.r4

37. "The Imuatil'e for Indigenous Futum,' lrutwntbirq:m1.,,

ahtec.org, iifl.

3 Chel)i I.:HuondeUe, "Codelillt~ ~ ? IIX-'ll'l u

ntontS: Tracing Jndrgtnous Pathwa;1 rn,\'rw,lftJ:JArt,ec.!ilti!xc~

Calgil}~ t.:ruwnitr of Calgary Pres, 2014 , 1µ

39 Archer Pecha1s,s, "Indigenism: Aboriglnll ll'adil liulltl

Ccdtd Ttmtorres, 38.

40. lolene Rickard. "Cons1denngTradJuom!Pn..11a1ik!1·-

1)I Annual ·~mposium on the Future lmaginuy. TIFF!ld1¢'1,'

ber 16, 2015, http: ahtec.orgiiifisymposia 1-nei,·-beg.t.;:i :'1ir:

41. Hamman. R1dmg tht Trailo/Ttars.31).

42, Fenwick McKelvey, lnttmtl Datmons: Drgiu/ ~-- Ill

neapolis. Lnivers1ty of .Minnesota Pre;s. ioiB) 4J. Anempts to widen the animacJ lens and !hnmthts;asls:r!'li

the Western tradition do exis~ of oount:Splllozai~~1:t:!

' rib I JDJltt!, G~ ~ Morton's hyperobiects, Jane Bennets f ran .w..

th gh.. iointrOducrS!'t.'tlliriU'!

PhiJo,ophy. etc. "-1Y aim here, ou 15 , , . ·d tbtll'estc111CZ:SJ

uon of machr.ne relationships from outsi e ltldkd~~ ZoeToddseth.icalcntiqueofthe®'Opiaofthest , ,~d

tJunkafor~ celebraung and worshipping a European ID~~

el]ectual benugt, ,--balllll-"' · ~ by Jra~1ng on a European uu . • 1.oe rc.Jd '.I:~

the world could have told t'OU for millenrua. ~\i\tdliJ~ I

Take on the Ontological Tum: 'OntolOg)·' lsJU!I

Is . / . 9'I (April I, 2016). of Historica oc10 og; z . . ' Lifi 8in11):A~- J

44 Kim Tall Bear, "Beyond the Lifef.iot ande ···~t'li;~ •· · ToJnking 111< ' ~

of Cryoprt)ervalion, Interspecies Radin and fJnml ~ Froztn Lift rn a Melting World. ed. Joanna .

. !,lfl'jd MIT Press, 2017) . LCJrllinginlPO)"lP

D Hill "Listening to Stones. tdl]ha ~ ~ll;l 4 5, on ' . • AJ/,trla Vim, St!'

logue in the World Ou151dC.

/listening-to-stones·

46. Martha Warren Beckwith, trans., ed., The Kumulipo: A Hawaiian Creation Chant

(Honolulu: The University Pres, of Hawaii, 1972). http://www.ulukau.org/eiib/cgi-bin

/library?e=d-obeckwi12-ooo ec 11haw-50-20-framese1-book-1-01oescapewin&a=d&d

=Do&toc=o. 47. Hill, "Listening."

An Orderly Assemblage of Biases: Troubling the Monocultural Stack 231