PreprintPDF Available

Open Source Investigations and the Technology-Driven Knowledge Controversy in Human Rights Fact-Finding

Authors:
Preprints and early-stage research may not have been peer reviewed yet.
Preprint

Open Source Investigations and the Technology-Driven Knowledge Controversy in Human Rights Fact-Finding

Abstract

In this chapter, we first describe the settled practices of human rights fact-finding that open source investigations have disrupted. Although the authority to shape these practices is centralized largely with Western human rights institutions populated by professional experts, the more decentralized underpinnings of open source investigation – namely, the use of information produced by civilian witnesses and through diverse networks – have an equally long history. We go on to detail how the rise of new technologies in human rights fact-finding has allowed for the participation of new actors in the form of civilian witnesses and analysts and necessitated the participation of others in the form of technologists and machine processes. These new actors bring with them not only new data and new methods, but also new norms about what human rights knowledge should be. The clash of these new elements with established practices produces a knowledge controversy in which much is possible and much is at stake. In the subsequent chapter section, we take a closer look at what is at stake through examining the power relations within human rights fact-finding revealed and disturbed by this knowledge controversy. Namely, we look at the power to shape human rights methodology, because methodology rules in and rules out particular types of human rights information with respect to evidence. It thus rules in and rules out particular types of corresponding subjects and witnesses of violations with respect to access to human rights mechanisms that can help them, in turn, speak truth to power. Ultimately, we are concerned with the impact of these power relations on pluralism, or the variety and volume of voices that can speak and be heard, both in terms of shaping the practices of human rights fact-finding, and in terms of access to human rights mechanisms that help subjects and witnesses speak truth to power.
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
1
Draft of chapter accepted for publication in: Koenig A, Dubberley S and Murray D, Digital
Witness: Using Open Source Information for Human Rights Investigation, Documentation
and Accountability (Oxford University Press, Forthcoming 2019)
Open Source Investigations and the Technology-Driven Knowledge
Controversy in Human Rights Fact-Finding
1
by Ella McPherson, Isabel Guenette Thornton, and Matt Mahmoudi
It’s 2018. There’s been yet another attack in the Syrian conflict - this time in Douma,
Eastern Ghouta, where a bombing has left the marketplace devastated. A team of student
investigators have been trying to identify the position of a local hospital after they located a
video on Twitter showing dozens of heavily injured, hospitalized individuals. These students
belong to the Digital Verification Corps (DVC), Amnesty International's global project to train
the next generation of human rights investigators at several university campuses worldwide.
DVC students discover and verify open source information to help Amnesty researchers find
evidence of human rights abuses. The video in question has been tweeted by a number of
different users, each time accompanied by written text claiming that it documents the
victims of the marketplace bombing. Should the students be able to geolocate the hospital,
they will be better able to assess the scale of the attack through an indication of the
minimum number of victims – information they can pass on to Amnesty’s investigators, who
are trying to establish the facts of the event. The video gives little away; shot from inside
the hospital, few geographical clues are revealed. “What’s that?” says one of the students,
pointing at two or three frames in the video that reveal a child in polka-dot trousers being
carried into the hospital. Scrolling back to another video the team had geolocated earlier
that day, the students compare this child with a similar-looking child in similar-looking
polka-dot trousers being carried away from the scene of the marketplace attack. The
student investigators excitedly agree: They’ve found a match that helps them establish the
location of the hospital.
This scenario — student investigators helping Amnesty’s research by analyzing
civilian witness video posted publicly online — was unimaginable only a few years ago.
Investigations based on the proliferation of open source information like the Twitter video
mentioned above have transformed the established practices of human rights fact-finding.
As we explain below, this transformation is in terms of who is involved (amateurs and
technologists as well as professional human rights fact-finders), the data under scrutiny
(including social media content and publicly-available databases such as Google Earth Pro),
the methods used (such as cross-referencing the metadata of open source civilian witness
content), and the norms about knowledge production that participants bring to the table
(an emphasis on quantitative versus qualitative, for example).
1
We gratefully acknowledge the funding for this chapter, European Commission Horizon 2020 (H2020)
Industrial Leadership (IL), grant 687967, ChainReact: Making Supplier Networks Transparent, Understandable
and Responsive
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
2
One way to understand this sort of transformation is as a type of knowledge
controversy. A knowledge controversy can occur when previously settled and taken-for-
granted practices of knowledge production (like the human rights evidence produced
through the practice of fact-finding) are unsettled and questioned because of the
introduction of a novel element in the form of new participants, data, methods and/or
norms.
2
The rise of open source investigations is part of a knowledge controversy in human
rights fact-finding. This particular knowledge controversy is driven by the adoption of new
technologies in the production and evaluation of human rights information for evidence in
advocacy and courts.
In this chapter, we first describe the settled practices of human rights fact-finding
that open source investigations have disrupted. Although the authority to shape these
practices is centralized largely with Western human rights institutions populated by
professional experts, the more decentralized underpinnings of open source investigation –
namely, the use of information produced by civilian witnesses and through diverse networks
– have an equally long history. We go on to detail how the rise of new technologies in
human rights fact-finding has allowed for the participation of new actors in the form of
civilian witnesses and analysts and necessitated the participation of others in the form of
technologists and machine processes. These new actors bring with them not only new data
and new methods, but also new norms about what human rights knowledge should be. The
clash of these new elements with established practices produces a knowledge controversy
in which much is possible and much is at stake. In the subsequent chapter section, we take
a closer look at what is at stake through examining the power relations within human rights
fact-finding revealed and disturbed by this knowledge controversy. Namely, we look at the
power to shape human rights methodology, because methodology rules in and rules out
particular types of human rights information with respect to evidence. It thus rules in and
rules out particular types of corresponding subjects and witnesses of violations with respect
to access to human rights mechanisms that can help them, in turn, speak truth to power.
Ultimately, we are concerned with the impact of these power relations on pluralism, or the
variety and volume of voices that can speak and be heard, both in terms of shaping the
practices of human rights fact-finding, and in terms of access to human rights mechanisms
that help subjects and witnesses speak truth to power.
Established human rights practices disrupted by the knowledge controversy
The appearance of a knowledge controversy marks the shattering of a prior
consensus about an established practice of knowledge production and thus the type of
knowledge produced by that practice.
3
In this section, we provide an overview of the
settled and established practices for producing human rights knowledge disrupted by the
adoption of new technologies. These practices, circumscribed by a particular set of actors
and a particular set of methods, emerged from a set of institutions who have come to
dominate human rights fact-finding in the international arena.
4
At first glance, open source
2
Sarah J Whatmore, ‘Mapping Knowledge Controversies : Science, Democracy and the Redistribution of
Expertise’ [2009] Progress in Human Geography 1; Andrew Barry, ‘Political Situations: Knowledge
Controversies in Transnational Governance’ (2012) 6 Critical Policy Studies 324.
3
Whatmore (n 1); Barry (n 1).
4
Diane Orentlicher, ‘International Norms in Human Rights Fact-Finding’ in Philip Alston and Sarah Knuckey
(eds), The Transformation of Human Rights Fact-Finding (Oxford University Press 2016); Dustin N Sharp,
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
3
investigations seem a radical departure from these established, orthodox practices, but a
longer gaze sees this new development as an extension of two other long-established
practices in human rights fact-finding: the use of civilian witness information for evidence as
well as collaboration among a diverse array of networked individuals and institutions.
The dominant institutional framework drawing on human rights fact-finding has
consolidated over time, growing increasingly bureaucratic and elite. Early fact-finding
involved select diplomats and legal experts conducting field visits and reporting their
findings to intergovernmental organizations like the United Nations.
5
Following high level
concerns critiquing human rights fact-finding for – in the words of a 1964 United Nations
Special Committee debate – its “scantiness and unclearness,” a number of attempts were
made to standardize fact-finding methodology.
6
Under the influence of prominent human
rights NGOs, fact-finding evolved to focus on witness interviewing, often with or through
known sources with established credibility.
7
Upon returning from the field, these
investigators wrote up their reports, which were made public in order to shame clearly-
identified states and other actors into complying with human rights norms. Fact-finding in
these international NGOs continues to be conducted largely by elite, often Ivy League-
educated investigators who usually have been trained in law; even those from non-Western
nations have often received their educations in the West.
8
They maintain close relationships
with other elite members of the political and press classes in order to more effectively lobby
them behind closed doors.
9
At intergovernmental bodies, similar investigations are
undertaken by sets of experts belonging to each of the ten UN human rights treaty bodies.
These include, for example, fact-finding missions and investigations carried out by the UN
Office of the High Commissioner of Human Rights on behalf of the UN Human Rights
Council, the General Assembly, and the Security Council. The formal and bureaucratic
nature of these institutions developed in part as a counterweight to the formal and
bureaucratic targets of their reports: states.
In both intergovernmental and international NGO institutional contexts, an
insistence on the prescriptive and legalistic as experts seek to construct what appear to be
objective and undeniable facts has become the dominant modus operandi of human rights
institutions.
10
This insistence, as Philip Alston describes it, “risked producing a somewhat
formulaic and relatively inflexible style and format.”
11
That said, these orthodox fact-finding
practices, undertaken by dominant institutions, are not necessarily taken up by other actors
working in human rights and more broadly in emancipatory projects.
12
For example, post-
war anti-colonialists avoided the use of human rights discourse, despite decolonisation
‘Human Rights Fact-Finding and the Reproduction of Hierarchies’ in Philip Alston and Sarah Knuckey (eds), The
Transformation of Human Rights Fact-Finding (Oxford University Press 2016).
5
Philip Alston, ‘Introduction: Third Generation Human Rights Fact-Finding’, Proceedings of the Annual Meeting
(American Society of International Law) (2013).
6
BG Ramcharan, ‘Introduction’, International Law and Fact-Finding in the Field of Human Rights (Martinus
Nijhoff Publishers 1982) 1; Philip Alston and Sarah Knuckey, ‘The Transformation of Human Rights Fact-
Finding: Challenges and Opportunities’ in Philip Alston and Sarah Knuckey (eds), The Transformation of Human
Rights Fact-Finding (Oxford University Press 2016).
7
Alston (n 4).
8
Obiora Okafor, ‘International Human Rights Fact-Finding Praxis: A TWAIL Perspective’ in Philip Alston and
Sarah Knuckey (eds), The Transformation of Human Rights Fact-Finding (Oxford University Press 2016).
9
Sharp (n 3).
10
ibid.
11
Alston (n 4) 61.
12
Alston and Knuckey (n 5); Orentlicher (n 3).
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
4
emerging during and in the immediate aftermath of the establishment of the Universal
Declaration of Human Rights in 1948. Instead, self-determination was the operative
language, working towards a collective vision of liberation, rather than universal individual
rights. This was, in part, because anti-colonialists remembered all too clearly that Western
concepts of emancipation were the backbone of colonialism’s justification.
13
Still, the
dominant discourses, practices, and institutions of human rights tend to eclipse these
alternative emancipatory ideas and movements, with the consequence not only of diverting
attention and funds, but also of depreciating alternative actors, methods, and norms of
knowledge production.
14
The multiple actors, multiple methods and diverse data of open source human rights
investigations may seem at odds with these established practices. Even at the core of
bureaucratic and standardized human rights fact-finding, however, we see that open source
investigation is as much a continuation of as a break from these established practices, which
have always featured the use of civilian witness information and networked collaboration.
For example, in one of the earliest accounts of the mobilization of civilian witness
information in advocacy and accountability, the British human rights campaigner Emily
Hobhouse arduously documented the Second South African War of 1899-1902. This war
was between Great Britain and the two Boer republics over the expansion of British forces
in South Africa and control over the Transvaal gold mines. About her attempts to document
the conditions of refugee camps run by the British that had deteriorated into concentration
camps, Hobhouse said, “It is hardly possible to draw up an ordinary conventional report.”
Instead, she relied on a combination of fieldnotes, letters of correspondence, and
statements by Boer women and children, along with her photography.
15
As camera use became more common among the general public, spontaneous acts
of civilian witnessing increased. A searing example is the 31-year-old plumber George
Holliday’s capture on his camcorder of the Los Angeles Police Department brutally beating
Rodney King after stopping him for a traffic violation on March 3rd 1991. Although the
responsible officer was eventually acquitted, the electrifying images seen around the world
made the power of citizen media to shed light on human rights struggles palpable. They
presaged the new forms of data pulled into human rights fact-finding through open source
investigations, brought forth especially by advancements in camera-phones and social
media platforms.
Today, civilian witness data can be solicited and captured relatively securely and
digitally through reporting mechanisms such as digital forms tailored to particular situations
of human rights violations, like those created by The Whistle, or through relatively secure,
highly adopted messaging services, such as WhatsApp, Signal, and Telegram. As platforms
such as Facebook, Twitter, Instagram, and YouTube encourage “intimate storytelling” and
“voluntary self-disclosure,”
16
open source investigators may also discover civilian witness
information through using deep searches on these platforms. Like Emily Hobhouse, open
13
Samuel Moyn, The Last Utopia: Human Rights in History (Belknap Press 2012).
14
David Kennedy, ‘International Human Rights Movement: Part of the Problem?’ (2002) 15 Harvard Human
Rights Journal 101; Günter Frankenberg, ‘Human Rights and the Belief in a Just World’ (2014) 12 International
Journal of Constitutional Law 35.
15
Guardian Research Department, ‘19 June 1901: The South African Concentration Camps’ The Guardian (19
May 2011) <https://www.theguardian.com/theguardian/from-the-archive-blog/2011/may/19/guardian190-
south-africa-concentration-camps> accessed 31 December 2018.
16
Cristina Miguel, ‘Visual Intimacy on Social Media: From Selfies to the Co-Construction of Intimacies Through
Shared Pictures’ (2016) 2 Social Media + Society 1.
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
5
source investigators draw on a multiplicity of data sources, and today have access to a
plethora of new types of data and methods. Satellite imagery, for example, has advanced to
an unprecedented extent, with remote-sensing satellites now able to capture images in up
to 30 centimeters resolution, meaning that each image pixel captured by satellites is now
representative of 30 square centimeters on the ground. This is enough to capture
everything from infrastructure and missile sites down to troop units and vehicles. In another
example, corporate information publicly available online enables researchers to analyze
complex networks that may hide abusive practices such as modern slavery. Equally, data
scraping techniques allow even the most inexperienced of investigators to download large
chunks of data from across multiple sites in a matter of moments. Common to all of the
new trends in technology-assisted investigations is that they make data from hard-to-reach
places more readily available.
Because of the variety and scale of data involved, open source investigations often
require collaborations among a diverse network of actors, another practice with a long
tradition in human rights fact-finding. For decades, transnational advocacy networks
incorporating human rights investigators – but also including journalists, church leaders,
grassroots activists, and politicians – have worked together on the basis of shared values.
17
Given that a significant portion of the work of transnational advocacy networks is oriented
around communicating between institutions and across geographic locations, the global
architecture of the internet has further enabled the spread and effectiveness of their work.
A variety of technologists, human rights practitioners, architects, academics, and activists
are increasingly coming together to form networks and tools for supporting and improving
evidentiary and advocacy techniques, as evident in projects produced, for example, by
Forensic Architecture or Bellingcat. Private actors from the technology sector provide the
platforms and tools for data collection, analysis, and output, either indirectly or directly, for
the purposes of these investigations. Civilian analysts, crowds of amateurs who receive
training in order to help with labor-intensive analysis tasks, are another new addition to
open source investigation networks. For example, the Amnesty Decoders project on Raqqa
relies on digital volunteers to look through past and current satellite imagery to track
buildings over time in order to help identify periods and sites of airstrikes.
Amnesty International’s DVC is a pertinent illustration not only of the use of open
source investigation methods that have emerged across transnational networks, but also of
the many layers of individuals who have to interpret and process relevant data into some
form of knowledge about the particular event. A researcher in the vicinity of a human rights
related event, for example, might be alerted to its occurrence by a witness, or on rare
occasions, through personally witnessing the event. In some cases, the researcher is unable
to either reach the area in question or to cover enough relevant ground to fully scope the
event. Instead, she contacts Amnesty to request support from the Digital Verification Corps.
Having been briefed on the often limited information known about the event, the Corps -
which at the time of writing has a presence in the universities of Pretoria, Berkeley, Toronto,
Essex, Hong Kong, and Cambridge - proceeds to put together a multidisciplinary team of
investigators to conduct discovery surrounding said event. The discovery process includes
deep searches on Twitter, Facebook, YouTube, and other media fora that may have
portrayed the event within the given territory and timeframe. Investigators collect and
17
Margaret E Keck and Kathryn Sikkink, ‘Transnational Advocacy Networks in International and Regional
Politics’ (1999) 51 International Social Science Journal 89.
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
6
archive these videos and images, and proceed, collaboratively (at times across campuses),
to process the information using techniques such as reverse image searches to see if the
content has appeared online previously; satellite imagery to scan for landmarks, signs,
buildings, roads and landscapes to discover the geo-coordinates of the event; and weather
data to corroborate time and place. Investigators might then look for further videos that
portray the same event posted by different accounts or from different angles. The
researchers ask, can we find additional pieces of media that corroborate what we initially
discovered? Have investigators from other campuses found something different or reached
different conclusions? Finally, in conversation with the DVC manager housed at Amnesty
International, the Corps authors a report that establishes the probable veracity of the event
and documents the verification process. The report is then either sent to the researcher on
the ground to aid in further fact-finding and/or used by Amnesty International, in
combination with its in-house researcher’s observations, to write a press statement or
advocate for some action by various stakeholders. Though it has roots in established
traditions, human rights fact-finding in such instances is still a significant departure from
human rights professionals using conventional methods; the new actors involved bring new
understandings of knowledge, as we explore next.
New actors in human rights fact-finding and the struggle for interpretive authority
Today, data about human rights has become increasingly accessible and is no longer
solely the provenance of traditional human rights actors and experts. Advances in digital
communication have provided various platforms for the collective development of new
techniques to gather, contextualize, and verify data. These changes allow new actors
unassociated with traditional human rights organizations, like technologists, volunteer
digital verifiers, civilian witnesses, and even algorithms, to participate in the location,
interpretation, verification, and promotion of human rights information and offer traditional
actors new methods to apply to their work. Thus, both the “who” and the “howof
traditional expertise are shifting: Who, then, are human rights experts today, if human
rights information is no longer restricted to experts with relatively exclusive access to sites
of struggle? How is information properly contextualized and verified outside of traditional
paradigms of known authorship and chains of custody - particularly when there’s so much
more information, and so much of it is anonymous or from unknown sources?
Understanding the ways in which the “who” and the “how” of expertise are changing should
ultimately lead us to the “why: Why is it that these experts who are using certain methods
are endowed with authority, and what norms, values, and power dynamics does this
authority uphold?
One of the consequences of a knowledge controversy is the questioning of
traditional expertise and established authority figures.
18
A critical aspect of knowledge
production, contested during a knowledge controversy, is interpretive authority: The
authority to build information, which is inherently limited, into a coherent account or story
and to ascribe it meaning. Location-specific experts, for example, provide context for and
explain the significance of information that may not seem meaningful to the layperson or
even other human rights professionals without relevant geographic and socio-political
background knowledge of the event. With open-source investigations, data may be made
available to all —a video posted on YouTube or images tweeted out, for example — but the
18
Whatmore (n 1); Barry (n 1).
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
7
interpretation of this information is often challenging. Even putting aside instances of
malicious fakery, original posters of information may provide vague, misleading or no
context in which to understand their posts; similarly, in advocacy, disparate instances must
be tied into a bigger story that reflects their shared context. Multiple possible
interpretations often exist when data is meagre, or when posters provide conflicting
explanations of events. Therefore, this interpretative function is key in the transformation
of human rights information into evidence for advocacy and court. It is up to experts to
gather additional information to fill in the blanks or to verify or challenge existing
explanations. Spaces for negotiation are crucial in this interpretative work, where
stakeholders with different types of expertise and knowledge collaborate to produce truth-
claims, persuading others of their views, contesting alternate framings, and acknowledging
potential ambiguity in the interpretation.
Traditionally, interpretive authority has been the province of human rights
professionals, who may visit or work in situ where human rights violations are taking place
and develop networks of local informants; the information from these networks is then
transformed into evidence by the human rights expert according to the accepted methods
of her organization and her own expertise. This provenance is unsettled during the
knowledge controversy by new actors who bring new understandings of interpretive
authority, which we consider in turn below. Relatively new human actors in human rights
investigations include civilians, who enter the sector either as spontaneous civilian
witnesses who share their digital documentation of events around them, or as analysts
tapped to deal with the deluge of digital data, as in Amnesty’s Digital Decoders project or
the DVC. Technologists are also newcomers, and they volunteer or are invited by
traditional experts to assist with developing methods for analyzing and managing the deluge
of digital data relevant to human rights research. In order to incorporate new civilian
analysts, human rights professionals must teach them a relatively standardized, relatively
straightforward set of methods. As a result, human rights professionals are collaborating
with technologists in developing tools to assist both civilians and experts in verification and
analysis, helping to clarify, systematize, and speed up the investigative process. As we
return to below, increasingly-automated tools, because they can make information analysis
decisions autonomously, can almost themselves be considered new actors with
technologists acting behind them. In addition to being invited to collaborate directly with
human rights groups, technologists are becoming increasingly powerful new actors in these
investigations in their own right, as large private sector companies that deal in digital
information must grapple with data about human rights abuses that appear on their
platforms. For example, YouTube, Facebook and Twitter have all developed policies about
how to treat content that violates the company’s community standards but may provide
important documentation of a human rights violation, and have created teams dedicated to
investigating and addressing those violations. Startups from the technology-for-good sector
seeking to address human rights aims have also proliferated.
Challenges around interpretive authority are inevitable when claims to authority
depend on different backgrounds. Human rights professionals have authority based on
their experience, proximity to witnesses, understanding of socio-political contexts,
methodological expertise, and deep institutional knowledge, including familiarity with
advocacy values and practices. In contrast, civilian witnesses have the authority of
authenticity and deep cultural, social and historical knowledge. Technologists invent tools
and have the necessary technical expertise to analyze larger bodies of data in new ways.
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
8
Given that it is unlikely that any one actor will have access to all types of expertise, these
different experts must collaborate. During these collaborations, it may be difficult and to
translate methods, implicit knowledge, and underlying systems of values across boundaries.
As mentioned above, civilian witnesses’ use of digital technologies to document
potential situations of human rights violations represents a welcome step-change in the
amount of human rights information available. This information does not always translate
so easily into evidence, however. In contrast with the scientifically-based, truth-claims
epistemology central to dominant human rights institutions’ fact-finding methodology –
where things have to be exactly what they say they are, as proven through the triangulation
of data and methods – civilian witnesses may bring differing understandings of interpretive
authority to their notions of how to produce human rights knowledge. For example, a
witness may share a report on social media but, lacking photographs of this particular event
(or wary of posting images of victims), they may choose instead to post a proxy photograph
of a similar event. For these witnesses, this epistemology of illustration facilitates a
meaningful testimonial and truth telling that highlights their subjective experience: What
happened was (a lot like) this. Human rights practitioners may believe this witness and
want to support her; social media shares by human rights organizations can endow civilian
witnesses with legitimacy, facilitate their visibility and, through that visibility, provide some
protection against reprisals. At the same time, however, human rights practitioners are
vulnerable to accusations of misinformation if they (re-)post or support testimonial that
includes material, such as proxy images, that isn’t consistent with a professional truth-
claims epistemology. This can result in tension between would-be allies, as witnesses seek
to uphold their report’s legitimacy as authentic and original, and their interpretive authority
in making reports in the way that is meaningful to them, despite the fact that it then cannot
be situated within interpretive and epistemological norms of human rights organizations.
Excitement about the potential benefits of technologists engaging in human rights
discovery and reporting processes has been well documented, and indeed it is difficult to
imagine how human rights reporting in the age of digital communication can succeed
without technical expertise. However, this expertise also comes laden with techno-culture
norms that reframe interpretive authority, sometimes in ways that are misaligned with
implicit goals and values of the human rights sector. For example, a tension exists between
the techno-capitalist goal of efficiency and pluralist goals of negotiation and ambiguity.
19
Technologists working on behalf of human rights organizations to develop digital human
rights reporting apps often advocate for the accumulation of a greater amount of data,
specifically the kind of evidence that is easier to interpret using technical means (statistics;
shorter stories with identifying details; certain types of photographs), despite the
aforementioned dominant practice of long-form testimonials developed between witnesses
and human rights practitioners.
This continued emphasis from technologists on more data that can be technically
quantified (often with the idea that machine learning can separate the “good data” wheat
from the “bad data” chaff – a tricky proposition when dealing with involved testimonials)
offers a different interpretive paradigm and standard for fact-finding than in-depth
testimonials informed by personal relationships and bolstered by the longevity of the area-
knowledge of a human rights professional. This new paradigm, sometimes described
19
Luis Suarez-Villa, Globalization and Technocapitalism: The Political Economy of Corporate Power and
Technological Domination by Luis Suarez-Villa (Routledge 2012).
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
9
pejoratively by practitioners as “quantity over quality,” represents an uncomfortable
incursion of technologist conventions and values into this traditional human rights territory
– even as some practitioners hope that it can provide information from sources that they
might not receive otherwise.
20
A focus on data rather than narrative is also seen in the
quantitative turn within human rights led by figures such as Patrick Ball, a statistician who
analyzes large scale human rights abuses and to who has often provided expert testimony
on war crimes. However, an emphasis on the quantitative creates new problems; per Sally
Engle Merry, the “seduction” of quantitative data as able to provide concrete, seemingly
objective truths often obscures the power relations and assumptions inherent in the
development of quantitative systems of measurement.
21
The latter is particularly of concern in the rise of machine learning and automated
decision-making in human rights information analysis. Anxieties about digital tools in effect
overstepping interpretive authority – that is, being used to achieve seemingly objective
conclusions, that, nonetheless, are based on potentially flawed, biased or limited
assumptions – suggests that these tools may be considered new actors in their own right,
with varying degrees of interpretive authority. These new machine “actors” represent both
processes and analytical “judgements.” For example, programs may analyze the shadows in
an outdoor image to suggest the time of day and year; analyze the metadata attached to
images and video; judge if social media posts contain human rights-relevant content using
discovery algorithms; or suggest to civilian analysts working on a set of data which methods
to try in what order. In this last instance, such programs suggest the use of methods both
based on human judgement and machine judgement, with varying emphases. While the
mechanized paradigm is not new, having clear antecedents in Fordian models of efficiency
and mass production which similarly envisioned the human and machine working together
as a mechanized unit, it is newly expressed in human rights within the context of advances
in both machine judgement and instantly mediated digital communication. This interaction
between old paradigms and new advances, in which new methods help us reflect on old
structures, is typical of a knowledge controversy.
These machine actors are developed by human actors (technologists, sometimes
working in collaboration with human rights professionals), and therefore inevitably reflect
the biases and limits of their creators. However, a techno-romantic view of machines as
omniscient – and machine judgement as transparent and uncontaminated by human bias –
has often buried this authorship. Norms like the fetishization of empiricism (prizing the use
of quantitative methods, outputs and formats regardless of their fit for the research
question or context), concerns with objectivity, anxieties about the fallibility of experts in
the context of the “post-truth” phenomenon, and the anticipation that institutions such as
courts require data that is gathered, verified and framed in an empiricized way, have
reinforced this perception that machine judgement is relatively pristine. Implicitly, this
perception implies that human judgement is comparatively contaminated. This assumption
and the resulting obfuscation of potential biases in machine judgement already has
consequences for vulnerable people, as Rebecca Wexler reveals in her work on the use of
algorithms to inform legal judgement within the criminal justice system, and Virginia
20
Isabel Guenette Thornton, Ella McPherson and Matthew Mahmoudi, ‘No Tech, Low Tech, Slow Tech: Human
Rights Practitioners’ Resistance to ICT4D’ [Forthcoming] International Journal of Human Rights.
21
Sally Engle Merry, The Seductions of Quantification: Measuring Human Rights, Gender Violence and Sex
Trafficking (The University of Chicago Press 2016).
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
10
Eubanks discusses in the context of the flawed algorithmic indicators that are used to
separate children from impoverished parents in Allegheny County, Pennsylvania.
22
Each interpretive paradigm or epistemology views the balance of authority in truth-
claims differently – whether, for example, authority should reside in statistical analysis, big
data patterns, qualitative accounts by experts, or testimonies direct from witnesses. As new
actors bring different paradigms and epistemologies of interpretive authority, this diversity
of perspectives increases the risk of errors from actors who might misunderstand
information produced across boundaries. This may occur, for example, as a result of
uncritically replicating information that is seemingly direct from witnesses or is the
automatic product of analytical tools produced without an informed consideration of
context. Actors may repost material that has not been properly verified, perhaps because
the tweet in question has already been taken up widely by the press or other social media
users, exacerbated by fast-moving visibility methods like hashtag activism. Actors may
reproduce statistics based on definitions or categories that they would not agree with were
those categories made transparent. They may mishandle the chain of custody when
gathering information, particularly digital information, such that the data is no longer
admissible in courts. With machine processes, analysis is limited by the tools and methods
offered, which may falsely suggest that a series of positive results on each test is enough to
produce confidence, when tests may have varying relevance to the questions at hand.
23
This
can be contrasted with traditional human-centered processes that emphasize the
understanding that methods must remain flexible, and sometimes be created anew, to be
an appropriate fit for the context.
Furthermore, new civilian witness and technologist actors may underestimate the
scope of interpretive work within analysis and verification and, indeed, may be skeptical of
seemingly intuitive processes, used by experienced human rights practitioners, that they are
unable to easily replicate or understand. Human rights professionals construct, in the words
of Stephen Hopgood, “an appearance of objectivity out of subjectivity” to define what
constitutes a human rights violation – which, per Richard Wilson, requires “a suppression of
the authorial voice and the deployment of a language purged of all tropes, metaphors, and
figurative elements,” as a way to empiricize and simplify complex interpretive processes.
24
The underlying interpretive work – such as resolving the question of how to deal with
ambiguous or incomplete data, inferring broader meanings of social and political
movements from individual stories, categorizing complex experiences, and attempting to
convey suffering – is thus easily obscured and devalued. The inherent uncertainty of such
interpretive work, which exists for all methods, may be further obfuscated by the use of
digital analysis tools that seem to give an empirical result. In the next chapter section, we
go on to consider the implications of the rise of these new actors, with their contested
understandings of interpretive authority, on the core human rights norm of pluralism.
22
Rebecca Wexler, ‘Life, Liberty, and Trade Secrets: Intellectual Property in the Criminal Justice System’ (2018)
70 Stanford Law Review 1343; Virginia Eubanks, Automating Inequality: How High-Tech Tools Profile, Police,
and Punish the Poor (St Martin’s Press 2017).
23
Orentlicher (n 3).
24
Stephen Hopgood, Keepers of the Flame: Understanding Amnesty International (Cornell University Press
2006) 5; Richard A Wilson, ‘Representing Human Rights Violations: Social Contexts and Subjectivities’ in
Richard A Wilson (ed), Human Rights, Culture & Context: Anthropological Perspectives (Pluto Press 1997) 149,
cited in Hopgood (2006) 5.
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
11
The knowledge controversy around open source human rights investigations and its
implications for pluralism
Knowledge controversies are unsettling, but this very state of unsettlement is
productive in terms of analyzing power relations and working to make them more equal.
25
The appearance of a knowledge controversy marks the shattering of prior consensus about
a topic of knowledge and the methods for creating that knowledge – a consensus that
naturalizes the knowledge and makes it seems as if it just is rather than is constructed by
social actors.
26
The clash of actors, methods, data, and norms in a knowledge controversy
creates an opportunity to be reflexive about the current practices of knowledge production
and how they came about – as well as what we think knowledge production should be. Part
of the latter involves identifying the norms that we value in knowledge production, and
assessing how directions in the knowledge controversy measure up to those norms. This
may be an uncomfortable process, as it requires us to momentarily step away from our
stake in the knowledge controversy, about which we may feel passionately, in part because
it is wrapped up in our own power positioning.
The norm of knowledge production that concerns us here is one that is central to
our academic disciplines as well as to our work on The Whistle, an academic start-up
focused on supporting the reporting and verification of digital human rights information. It
is also a core norm for human rights fact-finding, given this practice’s concern with speaking
truth to power and giving voice to the voiceless. The norm in question is pluralism, which,
as mentioned above, is the variety and volume of voices that can speak and be heard, both
in terms of shaping the practices of human rights fact-finding, and in terms of access to
human rights mechanisms that help subjects and witnesses speak truth to power. Pluralism
of knowledge production is supported in at least two ways: the creation and maintenance of
spaces for opportunity and the creation and maintenance of spaces for negotiation. Spaces
for opportunity provide chances to participate in knowledge production. Spaces for
negotiation provide chances to negotiate that participation into the ultimate decision of
what knowledge is produced, how, by whom, and why. This distinction is important, as it
tells us something about not only the quantity of pluralism (spaces for opportunity), but also
its quality (spaces for negotiation). It has parallels to a long-standing debates in
development and citizenship studies and practices around what exactly participation should
look like, with greater quality of participation seen as having greater benefits in terms of
equalizing power relations.
27
Our overview of the settled practices of human rights fact-finding disrupted by the
current knowledge controversy indicates that these settled practices provided spaces for
opportunity and spaces for negotiation, albeit limited. Human rights practitioners have long
included civilian witness accounts, translating them into the standardized reports necessary
for influence at the international, institutional level. The traditional, orthodox practice of
face-to-face interviews allows not only spaces for opportunity for civilian witnesses to
25
Sarah Franklin, Biological Relatives: IVF, Stem Cells, and the Future of Kinship (Duke University Press 2013);
Sarah Harding, ‘Feminism, Science, and the Anti-Enlightenment Critiques’ in Linda J Nicholson (ed),
Feminism/Postmodernism (1990).
26
Barry (n 1); Brian Martin and Evelleen Richards, ‘Scientific Knowledge, Controversy, and Public Decision-
Making’ in Sheila Jasanoff and others (eds), Handbook of Science and Technology Studies (Sage 1995);
Whatmore (n 1).
27
Sherry R Arnstein, ‘A Ladder of Citizen Participation’ (1969) 35 Journal of the American Planning Association
216.
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
12
convey their knowledge to fact-finders, but also spaces for negotiation, as this knowledge
transfer happens through a conversation that allows its participants explain their respective
versions of interpretive authority and their respective epistemologies so as to arrive at a
mutual understanding of the events under discussion. For example, a civilian witness who
remembers the violation largely through the emotional as well as physical trauma they
experienced may meet a fact-finder who is interested in an account that emphasizes specific
facts, such as the place, date, and time of the violation. In cases like this, the interlocutors
can, through negotiation, build bridges between differing visions of what human rights
knowledge is and how to construct it. In particular, through listening to the civilian witness,
the fact-finder can honor her account while also translating it for consumption by human
rights institutions and possible use as evidence admissible in court.
28
The resource-intensive method of face-to-face interviews has, however, always
meant a limit on spaces for opportunity for witnesses. These limits stem not only from the
cap on the number of civilian witnesses who can cross paths with fact-finders during their
research, but also from methodological limitations: Those who have experienced violations
that are less documentable using the traditional methodology of shaming based on witness
testimonies (for example, violations without survivors or violations involving social,
economic, and cultural rights) have generally had greater difficulty in accessing the
machinery of human rights process.
29
At the broader scale of institutional politics, we see
limitations on spaces for negotiation arising from the power dynamics among the different
sets of actors involved. Though civilian witnesses and collaborative networks are important
to developing human rights information into evidence, orthodox human rights institutions –
because of their resources, their credibility, and their proximity to the corridors of power –
usually have the dominant interpretive authority. This power imbalance ultimately
circumscribes the potential space for negotiation in the production of human rights
knowledge. It is also a historical precedent that is illuminatory for the dynamics of the
current knowledge controversy.
First of all, despite predictions that information and communication technologies
would allow the pursuit of accountability to be conducted extra-institutionally through peer
networks of citizens, societies show no signs yet of entering a post-institution age. The
multi-layered human rights architecture that is advocated by bodies and agencies of the
United Nations (UN) is still very much at the center of how human rights practice is
executed in the international context. There are, for instance, still significant constraints to
what can be considered a legitimate method for evidence capture, with a preference for
established sources with whom institutions have built trust – a fact that is in tension with
the adoption of open source methods, a prime feature of which is a greater diversity of
voices (the majority of whom will not have the credibility of long-term sources and
informants). International NGOs, and the UN bodies in particular, are here to stay and have
a great degree of state-power behind them in shaping dominant epistemologies and norms;
traditional ways of “doing” human rights have, in other words, become somewhat
cemented at this level. This makes it difficult to truly open up space for negotiation on
human rights methodologies. Participants in open source investigations are thus
potentially subject to a dilemma vis-à-vis human rights pluralism. On the one hand, they
28
Ella McPherson, ‘Technologies for Human Rights Witnessing: Humans, Machines and Ethics’ (Working
Paper).
29
Kenneth Roth, ‘Defending Economic, Social and Cultural Rights: Practical Issues Faced by an International
Human Rights Organization’ (2004) 26 Human Rights Quarterly 63.
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
13
must plug in to the existing institutional framework to be effective in delivering citizens
access to the mechanisms of human rights accountability; on the other hand, to do so, they
may not have the space to negotiate alternative interpretive frameworks and norms into
the production of human rights knowledge, but rather have to shoehorn this diversity into
the standardized, dominant methods of these institutions.
Secondly, these institutional dynamics have set a precedent for how struggles over
spaces for opportunity and spaces for negotiation might play out. Human rights
professionals are well aware of power dynamics, including the inequalities around pluralism,
involved in orthodox human rights practices;
30
this explains some of the excitement that has
emerged about the spaces for opportunity afforded by new communication technologies.
31
These technologies have been invaluable for the spontaneous civilian witness who, upon
experiencing a human rights atrocity or crime, can take to any number of platforms to seek
uptake of their information among publics and professionals – a particularly attractive (if
risky) possibility in the absence of a domestic rule of law system to which victims can appeal
for remedy. Though inaccessible in some parts of the world (e.g. due to cost or
authoritarian censorship) or differentially accessible within a community (e.g. due to gender
or age norms around technology ownership), a camera and access to the internet would
seem enough to shed light on an unseen event, regardless of geographic context or the
nature of the atrocity.
32
It is not, however, just digital divide issues in terms of access –
which can map onto traditional issues of fact-finder access to dangerous or remote regions
– that limit potential new spaces for opportunity. Divides in terms of technical and
information literacy are a concern as well.
33
Abuses that are easier to capture and
corroborate visually, such as those occurring in public places, may be more easily identified
and analyzed than harms like sexual violence, which are more likely to happen in private
and are difficult to document with current tools like smartphones. In another limitation to
spaces for opportunity, human rights organizations’ long-standing central concern with
credibility means this is still a metric by which potential civilian witnesses are assessed;
exclusionary power dynamics persist around this, where greater credibility can be
associated with greater social capital, such as the number and type of followers.
34
Algorithmic privilege, namely a user’s relatively prominent positioning in other users’ social
media timelines due to the opaque workings of timeline algorithms, and content
moderation, or the removal of user content from social media platforms due to its
perceived violation of community standards, are other ways in which new spaces for
opportunity are limited.
Still, it seems spaces for opportunity are growing with new technologies not only in
terms of civilian witness information but also in terms of the evaluation of that information
for evidence. New networks are coming together across diverse professions to improve
digital analysis techniques, including through the development of new opportunities for
30
Alston and Knuckey (n 5).
31
Molly Land, ‘Democratizing Human Rights Fact-Finding’ in Philip Alston and Sarah Knuckey (eds), The
Transformation of Human Rights Fact-Finding (Oxford University Press 2016).
32
Christoph Koettl, ‘Citizen Media Research and Verification: An Analytical Framework for Human Rights
Practitioners’ (University of Cambridge Centre of Governance and Human Rights 2016).
33
Ella McPherson, ‘Digital Human Rights Reporting by Civilian Witnesses: Surmounting the Verification Barrier’
in Rebecca Ann Lind (ed), Produsing Theory in a Digital World 2.0: The Intersection of Audiences and
Production in Contemporary Theory, vol 2 (Peter Lang Publishing 2015).
34
Ella McPherson, ‘Advocacy Organizations’ Evaluation of Social Media Information for NGO Journalism: The
Evidence and Engagement Models’ (2015) 59 American Behavioral Scientist 124.
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
14
civilian analysts. In these scenarios, expertise is established through using accepted,
rigorous (and often reproducible) methods: this is akin to what Diane Orentlicher calls
“accountability-through-methodology” rather than the authority of experts as such.
35
Put
another way, the “how” of human rights investigations leads to the “who” and the “why” of
accepted expertise, instead of the other way around. While a focus on methods —
particularly reproducible and empiricized methods that can, in theory, be deployed by
anyone to verify or challenge human rights stories — can be seen on the one hand as
creating new spaces for opportunity, it can be seen on the other as undermining the value
of other forms of human judgement, and thereby reducing pluralizing spaces for negotiation
and ambiguity.
A focus on methods in the current knowledge controversy means a focus on
technologies, and these technologies and their associated technologists bring new norms of
knowledge construction into the mix – or their increasing prominence makes their
associated norms more dominant. A significant proportion of the new actors and methods
introduced with open source investigations hail from the technology sector. Think, for
example, of the use of Facebook, Twitter and WhatsApp as human rights information
transmission mediums, or of the development of human rights-specific applications by
programmers. The programmers building these technologies, whether mainstream or
working in a human rights niche, typically receive their formative training in a cultural
context increasingly shaped by Silicon Valley values. A cornerstone of these values is the
norm of efficiency, with associated knowledge values of quantification and objectivity that
allow for more efficient analysis. Prizing efficiency means prizing the ability to do the same
with less resources or to do more with the same amount of resources. The latter is
especially important in the informational sector, given the oft-cited statistic that 90 percent
of the world’s data was produced in the past two years.
36
In other words, prizing efficiency
prioritizes a technology solution to a technology problem – that of big data.
Like every other information-based profession, the human rights sector faces a big
data problem. Often, open source investigations involve the discovery of a vast amount of
data, much more than was previously possible, which is both the exciting and challenging
potential of these new forms of information collection – particularly given the time and
expertise needed to verify new forms of digital data.
37
Given the hours involved, the
interest among many actors in the human rights world and beyond in the development and
adoption of new technologies for making the collection and analysis of this data more
efficient is understandable. This emphasis on efficiency, however, not only squeezes out
space for negotiation but also may make negotiation a less desirable norm of knowledge
production.
Knowledge production achieved through spaces for negotiation flourishes within
human relationships, over time, with effort and exchange – elements at odds with the norm
of efficiency. The efficiency of technologies often derives from their replacement of human
endeavour with machine work, which often reduces time constraints in part through
35
(n 3) 509.
36
Bernard Marr, ‘How Much Data Do We Create Every Day? The Mind-Blowing Stats Everyone Should Read’
(Forbes, 21 May 2018) <https://www.forbes.com/sites/bernardmarr/2018/05/21/how-much-data-do-we-
create-every-day-the-mind-blowing-stats-everyone-should-read/> accessed 3 September 2018.
37
McPherson, ‘Digital Human Rights Reporting by Civilian Witnesses: Surmounting the Verification Barrier’ (n
32).
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
15
eliminating interpersonal contact.
38
Think, for example, of a civilian witness reporting a
human rights violation through a web-based form rather than face-to-face with a fact-
finder. A different instance is the rise of satellite imagery to document human rights
violations. This has created new opportunities for inclusion in human rights cases, but such
a data source places a great distance between those affected on the ground and the human
rights fact-finders picking up their cases. In another example, ICTs allow the Digital
Verification Corps to work collaboratively across time and place, yet this separation does
not always lend itself well to in-depth conversation. Negotiation is a human-to-human
process; it is difficult to negotiate with a machine – despite advances in artificial intelligence
– or even sometimes to negotiate with another human through a machine.
If the rise of technologists and their tools in this knowledge controversy casts doubt
on human judgement and non-mechanized processes (as contaminated, and also as
inefficient compared to machine processes), this introduces consequences for interpretive
authority. Human actors may increasingly be encouraged to act as consistent and efficient
machines through following protocols that resemble algorithms, while machines are
allowed to act as humans by making evaluative suggestions or even decisions about
information.
39
In terms of the former, programs supporting civilian analysts to assist with
verification often provide a series of methods and tools following an ‘if-then’ structured
series of tasks – an algorithmic paradigm. Examples of the latter include programs in
computer vision, where machines automatically identify objects in pictures, or algorithmic
identification engines, where machines make judgements about what material might be
valuable in human rights investigations such as through identification of key words.
40
The
interpretive authority of machine processes is ascendant; human actors’ space for
negotiation is compressed by the new pseudo-mechanized processes involved in human
rights fact-finding as well as by algorithms’ invisibilization of interpretive moments and bias
that otherwise might be identified and interrogated. Even if unintentional, this obfuscation
of the falsity of machine processes’ implied certainty is always political, as removing space
for ambiguity and negotiation removes interpretive agency and connotes that only a single
epistemology is valid. As a result, the norm of negotiation in knowledge production is
depreciated.
The reduced spaces for negotiation resulting from the rise of technology tools and
norms has a further consequence of introducing new inequalities and exacerbating existing
ones. In other words, some actors have access to more space than others. For example,
regarding negotiation around the shape and transparency of tool design, the elite
populations working at Western NGOs are more likely to have connections to technologists
than those in the Global South.
41
This inequality maps onto the existing dominance of the
West over the South in terms of the direction of human rights knowledge.
42
It also means
that the design priorities in technology for human rights will more likely come from Western
perceptions that may or may not fit with Southern realities – again, a wider, long-standing
38
McPherson, ‘Technologies for Human Rights Witnessing: Humans, Machines and Ethics’ (n 27).
39
ibid.
40
Jay D Aronson, ‘Computer Vision and Machine Learning for Human Rights Video Analysis: Case Studies,
Possibilities, Concerns, and Limitations’ (2018) 43 Law & Social Inquiry 1188; Alan Blackwell and others,
‘Computer Says “Dont Know" - Interacting Visually with Incomplete AI Models’ (2018)
<https://digital.lib.washington.edu/researchworks/bitstream/handle/1773/42857/DTSHPS18-Proceedings-
final%20v2.pdf?sequence=8&isAllowed=y> accessed 2 January 2019.
41
Alston and Knuckey (n 5).
42
Okafor (n 7).
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
16
problematic for human rights ideas and methods.
43
In another example, while actors
external to the dominant human rights institutions may have increasing opportunities for
their information to be seen by these institutions, they may have less space to negotiate the
interpretation of this information, as this visibility occurs mediated by machines rather than
by humans. The consequence is that, as under the established practice of human rights
fact-finding, information might come from the “bottom,” but, as Dustin Sharp puts it,
‘Solutions generally come from the “top.”
44
In sum, the rise of open source human rights
investigations may mean that, though we see more actors engaged in human rights
knowledge production, we hear them less.
Conclusion
In this chapter, and in the spirit of the knowledge controversy approach, we have
raised more questions than answers and more possibilities than conclusions about the
changes wrought by and around the rise of open source human rights investigation.
45
We
pose these questions in the aim of creating spaces for reflection among practitioners and
scholars, rather than state answers that might shut out alternative perspectives. As we
have explained, human rights fact-finding is in the midst of a knowledge controversy
spurred by the rise of new technologies and manifesting in expanding practices, like open
source investigation, involving new actors, who bring new norms. Knowledge controversies
are marked by a departure from the taken-for-granted, established order, which in turn
provides insight into the normative, practical, and power-related dimensions of the previous
status quo. They feature potential misalignments in norms and epistemologies between
actors, and contested expectations around methodological practices – conflicts visible in
public and professional anxieties about changing paradigms of expertise, interpretive
authority and power in the production of human rights knowledge. We have focused here
on the implications of this knowledge controversy and its contests for the pluralism of
human rights knowledge production. We go beyond concerns with the quantity of
pluralism, associated with spaces for opportunity for participation, to the underexamined
quality of pluralism manifested in spaces for negotiation over the methodologies used to
establish the knowledge, and the associated interpretive authority necessary to do so. This
more nuanced understanding of pluralism ensures that we push analytically far beyond the
mistaken assumption that the wider availability of technologies for human rights is in and of
itself considered a measure of greater pluralism.
46
Open source human rights investigation surfaced in a field dominated by
institutional players, but also featuring looser networks, including civilian witnesses. This
new practice is an evolution of the latter, but must also fit into the orthodox institutional
framework to produce evidence that will be taken seriously by these still-dominant
mechanisms of accountability and justice. As such, despite the significant new, exciting
capacities and expanded spaces for opportunity for new actors to participate in the human
rights knowledge production that new technologies have afforded, these are arriving in a
context where the precedent has been to provide a narrower space for negotiation over the
43
Sharp (n 3).
44
ibid 76.
45
Whatmore (n 1).
46
Tamy Guberek and Romesh Silva, ‘Human Rights and Technology: Mapping the Landscape to Support
Grantmaking’ (Partners for Human Rights Information, Methodology and Analysis 2014)
<https://www.fordfoundation.org/media/2541/prima-hr-tech-report.pdf> accessed 3 January 2018.
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
17
interpretation of that knowledge. This precedent has dovetailed with a troubling trend,
which is that spaces for negotiation may be narrowing in the relatively new practice of open
source investigation as well. Norms that have arrived with new technologist and technology
actors – such as efficiency, quantification, and objectivity – can clash with the norm of
negotiation. Not only are new, efficiency-oriented practices edging out spaces for
negotiation, including reflection on information interpretation, but they are also
depreciating negotiation as a norm of knowledge production – despite its benefits for
pluralism.
At the time of writing, we are still in the midst of the knowledge controversy in
human rights fact-finding. It remains to be seen how competitions and collaborations will
develop. Will heterogeneous norms, goals, and methods resolve over time, leading to
flexible, hybrid practices? Will they solidify into rigid practices dominated by particular
actors? Or will they continue to provoke anxieties and struggles around contested
interpretive authority?
Along with the anxieties of knowledge controversies come the benefits of the
openness they create. This openness, however, is short-lived. It only lasts as long as the
knowledge controversy. The actors involved often rush to settle the controversies that arise
because norms and practices of knowledge production are unstable and evolving, and work
based on these norms and practices can be discomfiting, slow and difficult to complete.
Returning to a taken-for-granted and standardized state allows us to work more efficiently,
but it comes at the expense of a critical awareness of why and how we are producing
knowledge that would otherwise allow us to question and adjust our norms and practices.
Just as the opening of a knowledge controversy is an important moment for the state of
knowledge production, so is its closure. This is the moment when knowledge and methods
settle and become widely accepted anew. One version in the controversy has won. Its
proponents have earned the power to define how knowledge is produced and how much
space for opportunity and for negotiation is built into the system.
47
As a result, we encourage participants not to rush to resolve the current knowledge
controversy, but rather to dwell in the openness it creates.
48
This is the moment to reflect
on existing and desired norms and practices of knowledge production, to retain or change
them, and to evaluate new norms and practices against them. It is a time to ensure that
flexibility exists to tailor the traditional and the new to best fit each fact-finding situation. It
is now that participants should continue to build on the reflexive turn in human rights fact-
finding to think not only about implications of changes for spaces for opportunity and
negotiation and how to protect and grow these spaces, but also about relative inequalities
of access to these spaces among different populations. Questioning along the way, even as
you read this book, how interpretive authority and expertise are developed through norms,
practices and methods, why these methods are emphasized over others, and who are
empowered as experts as a result will draw attention to the power dynamics and potential
inequalities inherent in this knowledge controversy.
The idea of a knowledge controversy may be at once familiar and unsettling for
human rights practitioners. Given the contested nature of human rights reports,
controversy is the murky air practitioners breathe. That said, human rights professionals are
usually trying to clear the air, to settle the dust and allow the facts to emerge. Where they
47
Martin and Richards (n 25).
48
Whatmore (n 1).
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
18
see violations, they employ methods to make their evidence as incontrovertible as possible.
Part of this process entails the meta-method of publicly communicating the rigor of their
methods in their reports and on their websites. So while human rights practitioners might
be comfortable with controversies on the level of evidence about what happened in specific
instances of human rights violations, more disconcerting is a knowledge controversy in how
we arrive at human rights evidence overall. Still, even though practitioners may breathe a
sigh of relief when this technologically-afforded knowledge controversy closes and methods
naturalize again, we urge them to retain some of its openness in their production of
knowledge.
Settling a knowledge controversy prematurely not only risks sedimenting power
dynamics within human rights investigations, it may also pose risks to human rights
investigations as nefarious actors may co-opt new tools, methods and forms of interpretive
authority in ways contrary to pluralism. We are reminded by the increase of digital fakery
scandals that the positive developments leveraging digital technologies for open source
investigations may also be overshadowed and further complicated by new techniques
adopted by nefarious state and non-state actors, and developed against the backdrop of
leaps in artificial intelligence and machine learning. Though digital verifiers such as the
Atlantic Council's Digital Forensic Research Lab and Bellingcat have fairly settled techniques
for exposing disinformation and digital fakery, deep fakes – digital scenarios, including the
video and audio-based imitation of individuals, generated via artificial intelligence – pose
perhaps one of the greatest information challenges on our horizon, as Scott Edwards
describes in chapter X of this book. Furthermore, as open source investigation develops
against the backdrop of the fake news era, it has become ever more common for opponents
to attempt to discredit human rights fact-findersmethods and findings, and such
discrediting discourses may take hold among broader publics. The socio-technical changes
introduced in information-sharing in the late digital age must therefore be approached with
caution; there is much to be excited about in open source investigations and human rights,
but also reason to tread carefully to avoid falling into noxious challenges to human rights
documentation. With spontaneous witnesses of human rights-related incidents already
considered suspect by their lack of established credentials, aforementioned developments
risk further jeopardizing marginalized voices.
As the risks for human rights investigations are so high, open source human rights
investigation is, in a sense, the canary in the new information coalmine. The perspective of
human rights practitioners is incredibly valuable for other knowledge professions such as
journalism and academia, as, being at the frontier, they can provide thought leadership in
terms of how to navigate this terrain as equitably and inclusively as possible. This can be
showcased in an ethical approach to open source investigation that supports spaces for
opportunity and negotiation – thinking all the while about how to settle the knowledge
controversy with the utmost consideration of pluralism and power relations, while also
retaining its spirit of reflexivity and flexibility.
Bibliography
Alston P, ‘Introduction: Third Generation Human Rights Fact-Finding’, Proceedings of the
Annual Meeting (American Society of International Law) (2013)
Alston P and Knuckey S, ‘The Transformation of Human Rights Fact-Finding: Challenges and
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
19
Opportunities’ in Philip Alston and Sarah Knuckey (eds), The Transformation of Human
Rights Fact-Finding (Oxford University Press 2016)
Arnstein SR, ‘A Ladder of Citizen Participation’ (1969) 35 Journal of the American Planning
Association 216
Aronson JD, ‘Computer Vision and Machine Learning for Human Rights Video Analysis: Case
Studies, Possibilities, Concerns, and Limitations’ (2018) 43 Law & Social Inquiry 1188
Barry A, ‘Political Situations: Knowledge Controversies in Transnational Governance’ (2012)
6 Critical Policy Studies 324
Blackwell A and others, ‘Computer Says “Don”t Know" - Interacting Visually with Incomplete
AI Models’ (2018)
<https://digital.lib.washington.edu/researchworks/bitstream/handle/1773/42857/DTSHPS1
8-Proceedings-final%20v2.pdf?sequence=8&isAllowed=y> accessed 2 January 2019
Engle Merry S, The Seductions of Quantification: Measuring Human Rights, Gender Violence
and Sex Trafficking (The University of Chicago Press 2016)
Eubanks V, Automating Inequality: How High-Tech Tools Profile, Police, and Punish the Poor
(St Martin’s Press 2017)
Frankenberg G, ‘Human Rights and the Belief in a Just World’ (2014) 12 International Journal
of Constitutional Law 35
Franklin S, Biological Relatives: IVF, Stem Cells, and the Future of Kinship (Duke University
Press 2013)
Guardian Research Department, ‘19 June 1901: The South African Concentration Camps’
The Guardian (19 May 2011) <https://www.theguardian.com/theguardian/from-the-
archive-blog/2011/may/19/guardian190-south-africa-concentration-camps> accessed 31
December 2018
Guberek T and Silva R, ‘Human Rights and Technology: Mapping the Landscape to Support
Grantmaking’ (Partners for Human Rights Information, Methodology and Analysis 2014)
<https://www.fordfoundation.org/media/2541/prima-hr-tech-report.pdf> accessed 3
January 2018
Guenette Thornton I, McPherson E and Mahmoudi M, ‘No Tech, Low Tech, Slow Tech:
Human Rights Practitioners’ Resistance to ICT4D’ [Forthcoming] International Journal of
Human Rights
Harding S, ‘Feminism, Science, and the Anti-Enlightenment Critiques’ in Linda J Nicholson
(ed), Feminism/Postmodernism (1990)
Hopgood S, Keepers of the Flame: Understanding Amnesty International (Cornell University
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
20
Press 2006)
Keck ME and Sikkink K, ‘Transnational Advocacy Networks in International and Regional
Politics’ (1999) 51 International Social Science Journal 89
Kennedy D, ‘International Human Rights Movement: Part of the Problem?’ (2002) 15
Harvard Human Rights Journal 101
Koettl C, ‘Citizen Media Research and Verification: An Analytical Framework for Human
Rights Practitioners’ (University of Cambridge Centre of Governance and Human Rights
2016)
Land M, ‘Democratizing Human Rights Fact-Finding’ in Philip Alston and Sarah Knuckey
(eds), The Transformation of Human Rights Fact-Finding (Oxford University Press 2016)
Marr B, ‘How Much Data Do We Create Every Day? The Mind-Blowing Stats Everyone
Should Read’ (Forbes, 21 May 2018)
<https://www.forbes.com/sites/bernardmarr/2018/05/21/how-much-data-do-we-create-
every-day-the-mind-blowing-stats-everyone-should-read/> accessed 3 September 2018
Martin B and Richards E, ‘Scientific Knowledge, Controversy, and Public Decision-Making’ in
Sheila Jasanoff and others (eds), Handbook of Science and Technology Studies (Sage 1995)
McPherson E, ‘Technologies for Human Rights Witnessing: Humans, Machines and Ethics’
(Working Paper)
McPherson E, ‘Advocacy Organizations’ Evaluation of Social Media Information for NGO
Journalism: The Evidence and Engagement Models’ (2015) 59 American Behavioral Scientist
124
McPherson E, ‘Digital Human Rights Reporting by Civilian Witnesses: Surmounting the
Verification Barrier’ in Rebecca Ann Lind (ed), Produsing Theory in a Digital World 2.0: The
Intersection of Audiences and Production in Contemporary Theory, vol 2 (Peter Lang
Publishing 2015)
Miguel C, ‘Visual Intimacy on Social Media: From Selfies to the Co-Construction of Intimacies
Through Shared Pictures’ (2016) 2 Social Media + Society 1
Moyn S, The Last Utopia: Human Rights in History (Belknap Press 2012)
Okafor O, ‘International Human Rights Fact-Finding Praxis: A TWAIL Perspective’ in Philip
Alston and Sarah Knuckey (eds), The Transformation of Human Rights Fact-Finding (Oxford
University Press 2016)
Orentlicher D, ‘International Norms in Human Rights Fact-Finding’ in Philip Alston and Sarah
Knuckey (eds), The Transformation of Human Rights Fact-Finding (Oxford University Press
2016)
McPherson, Guenette Thornton & Mahmoudi, “Open Source Investigations[Draft]
21
Ramcharan BG, ‘Introduction’, International Law and Fact-Finding in the Field of Human
Rights (Martinus Nijhoff Publishers 1982)
Roth K, ‘Defending Economic, Social and Cultural Rights: Practical Issues Faced by an
International Human Rights Organization’ (2004) 26 Human Rights Quarterly 63
Sharp DN, ‘Human Rights Fact-Finding and the Reproduction of Hierarchies’ in Philip Alston
and Sarah Knuckey (eds), The Transformation of Human Rights Fact-Finding (Oxford
University Press 2016)
Suarez-Villa L, Globalization and Technocapitalism: The Political Economy of Corporate
Power and Technological Domination by Luis Suarez-Villa (Routledge 2012)
Wexler R, ‘Life, Liberty, and Trade Secrets: Intellectual Property in the Criminal Justice
System’ (2018) 70 Stanford Law Review 1343
Whatmore SJ, ‘Mapping Knowledge Controversies : Science, Democracy and the
Redistribution of Expertise’ [2009] Progress in Human Geography 1
Wilson RA, ‘Representing Human Rights Violations: Social Contexts and Subjectivities’ in
Richard A Wilson (ed), Human Rights, Culture & Context: Anthropological Perspectives (Pluto
Press 1997)
ResearchGate has not been able to resolve any citations for this publication.
Chapter
Full-text available
This is the author accepted manuscript. The final version is available in "Produsing Theory in a Digital World 2.0: The Intersection of Audiences and Production in Contemporary Theory. Volume 2"
Article
Full-text available
This article is based on a study that analyzes the use of pictures to build and convey intimacy through social media interactions across two platforms of different genres: Badoo (dating/hook-up platform) and Facebook. The study explores what kinds of pictures participants (aged 25–49 years) consider to be intimate and whether they disclose those kinds of images through these platforms. My discussion moves from the photographs (often portraits and selfies) that participants choose to disclose in order to (re)present themselves on the network to the topic of negotiating the publication of pictures with other users. Participants mainly identified as intimate sexy pictures and images showing sexual orientation or relationships status. In the context of Badoo, double sexual standards were found to be a key issue in the disclosure of sexy pictures. For Facebook, associated intimacy and censoring of images, with revelation of sexual orientation, publication of relationship status, and sharing of emotions with wider or unintended audiences, were the main topics discussed. Using in-depth interviews and user profile analysis, the aim of this article is to understand the roles of images within public intimacy practices in different types of relationships across two distinct social media platforms.
Article
Full-text available
This article contributes to the emergent literature on the use of social media at advocacy organizations. Much of this existing literature focuses on these organizations’ production of social media information; this article, however, explores the complementary and relatively unexamined consumption of social media information that can form part of advocacy work. By drawing parallels between journalism and advocacy, the article develops two theoretical models of how advocacy organizations evaluate social media information as part of this consumption. These models differ according to the information values at their cores and according to how these values are evaluated in practice; correspondingly, the models interact differently with social media’s affordances. The key information value for the evidence model is the veracity of the information’s metadata, and this is largely evaluated through a time-intensive verification process requiring corroboration and drawing on human expertise. In contrast, the key information value pertaining to the engagement model is participation, which is evaluated by measuring the volume of participants in the information’s production and transmission. The affordances of social media are often hindrances for the evidence model, because they can make metadata more difficult to verify. In contrast, the engagement model capitalizes on social media affordances, because these affordances facilitate participation as well as the evaluation of participation volume using digital analytics. In addition to shedding light on approaches to social media information evaluation at advocacy organizations, this article urges researchers and practitioners to be alive to related barriers to pluralism as they study and use these approaches.
Chapter
Full-text available
Martin and Richards identify four distinct approaches: positivist, group politics, constructivist, and social structural. After describing each of the four and briefly applying them to controversies over fluoridation and vitamin C and cancer, the authors compare them across different areas (epistemology, focus of analysis, conceptual tools, closure, partisanship of the analyst, and decision-making). Martin and Richards note that most analysts don't fall entirely within any of these 'ideal types' and that certain analyses are more appropriate for certain situations than others. In the final section, they exploit this tension by providing two ways that the four approaches could be integrated for a more nuanced and comprehensive analysis than any one approach could individually accomplish. [Nitin and Chad, STS 901 - Fall 2006]
Article
Citizen video and other publicly available footage can provide evidence of human rights violations and war crimes. The ubiquity of visual data, however, may overwhelm those faced with preserving and analyzing it. This article examines how machine learning and computer vision can be used to make sense of large volumes of video in advocacy and accountability contexts. These technologies can enhance the efficiency and effectiveness of human rights advocacy and accountability efforts, but only if human rights organizations can access the technologies themselves and learn how to use them to promote human rights. As such, computer scientists and software developers working with the human rights community must understand the context in which their products are used and act in solidarity with practitioners. By working together, practitioners and scientists can level the playing field between the human rights community and the entities that perpetrate, tolerate, or seek to cover up violations.
Article
Globalization and Technocapitalism considers the global reach of a new capitalist era, exploring the nature of 'technocapitalism' as grounded in new forms of accumulation, commodification, and corporate organization. As technological creativity, corporate research, and talent flows become more important than ever, this book explores the manner in which globalization acquires new contextual features that will become central to the macro-social dynamics of the twenty-first century. It thus sheds light on the resultant growth in global inequalities and more intrusive forms of global domination that are grounded in emerging sectors, such as nanotechnology, biotechnology and its diverse fields, such as genomics, synthetic bioengineering, bioinformatics and biopharmacology, and related advances in computing and telecommunications. A rigorous examination of developments in contemporary capitalism as driven by the forces of globalization, Globalization and Technocapitalism will be of interest to scholars working in the fields of social and political theory, international political economy, political philosophy, science and technology studies and globalization.
Article
As it is practiced by the world’s biggest international human rights NGOs (INGOs), fact-finding has become an elite activity, carried out, for the most part, by a class of professionalized 'experts.' Over the last several decades, their work has catapulted organizations like Human Rights Watch and Amnesty International to positions of prominence and influence in myriad global political and policy processes. The work of these individuals and organizations is projected to be (and indeed is sincerely imagined to be) neutral and apolitical, the product of technocrats concerned only with exposing abuses by powers small and large. While no one can doubt that INGO-led fact-finding has been a force for much good, raising the level of global human rights awareness, the collection and dissemination of human rights facts and knowledge have also been intimately bound of up with politics, power, and the reproduction of hierarchies, making the technocratic view of human rights fact-finding highly problematic. In this chapter, I argue that there is a particular need to think carefully and critically about the role of human rights fact-finding in generating institutional legitimacy and power, in privileging certain questions of social justice over others, and in potentially serving to narrow the terrain through which broader projects of social change might take place. Understanding fact-finding not as a technocratic exercise, but as part of a set of complex institutional and global power dynamics with distributional consequences may suggest the need to democratize the field, including the collection and dissemination of human rights facts. Though it has fallen out of favor with some of the biggest NGOs, 'capacity building' is one potential model that could be used to try to diversify and pluralize the production of human rights knowledge, making human rights fact-finding more of a true global project rather than the domain of a relatively restricted set of elite institutions. Yet capacity building projects carry their own problematic dimensions, implying, among other things, a one-way transmission of expertise from elite to ‘local’ organizations, and may serve to propagate the very hierarchies and elite strategies for change typified some global NGOs. Thus, simplistic recipes of 'more local, less global' are not sufficient to address some of the more problematic aspects of fact-finding practice. Ultimately, I argue, there is a need for NGOs, both local and international, to diversify and democratize not only in terms of composition, but in terms of fundamental advocacy paradigms. Fact-finding and the projection of professional, technocratic expertise has served as an impressive platform for high-level lobbying and pressure politics via the mobilization of shame, but is ultimately no substitute for developing a genuine human rights constituency akin to other social justice movements throughout history. While fact-finding may still serve as an engine of growth and legitimacy for some INGOs in the 21st century, it needs to be leveraged to support a richer palette of advocacy tactics if the human rights 'movement' is to move outside of the elite circles to which it has largely heretofore been restricted.