On the Possibilities of a Political Theory of Algorithms

Date01 February 2021
Published date01 February 2021
Subject MatterArticles
/tmp/tmp-18SHV7tWvVavc8/input 959853PTXXXX10.1177/0090591720959853Political TheoryPanagia
Political Theory
2021, Vol. 49(1) 109 –133
On the Possibilities of
© The Author(s) 2020
Article reuse guidelines:
a Political Theory of
DOI: 10.1177/0090591720959853
Davide Panagia1
This essay asks how we might articulate a political theory of algorithms.
To do so, I propose a political ontology of the algorithm dispositif that
elaborates how algorithms arrange the movement of energies in space and
time, and how they do so automatically. This force of arrangement is what I
refer to as the dispositional power of algorithms that I identify as a political
physics of vital processes. The essay is divided into three sections. The first
provides readers of Political Theory with a discussion of three notable works
in the field of critical algorithm studies relevant to a political theory of
algorithms. The subsequent sections of the essay elaborate an understanding
of the political ontology of the algorithm dispositif by focusing (first) on the
difference that a virtual ontology introduces to our political reflections and
(second) on the cybernetic operation of negative feedback that I identify
as foundational to understanding an algorithm’s political physics of vital
processes. I conclude that any political theory engagement with technical
media can’t simply rest on an epistemic analysis of the normative effects
of media but must also pursue an investigation into a medium’s modes of
existence in the world.
political algorithms, entropy, cybernetics, science and technology studies,
1UCLA, Los Angeles, CA, USA
Corresponding Author:
Davide Panagia, UCLA, 4289 Bunche Hall, Los Angeles, CA 90095, USA.
Email: davidepanagia@ucla.edu

Political Theory 49(1)
In recent years, the technical capacities of algorithms have become an object
of critical investigation throughout the humanities, the social sciences, and
the natural sciences. Scholars in diverse disciplinary fields (ranging from
computer engineering, mathematics, the law, information studies, cognitive
neurobiology, cinema studies, etc.) are asking about the uses, abuses, and
effects of algorithms in everyday life. There is no doubt that algorithmic gov-
ernance is an actually existing regime of rule in contemporary democratic
life.1 From sentencing algorithms in criminal trials, to predictive policing
algorithms, to sorting algorithms that generate targeted campaign advertis-
ing, to biometric algorithms that monitor border crossings, and even the algo-
rithms that manage traffic signaling, sewerage systems, and other essential
flows, all these versions of algorithmic governance participate in democratic
politics and they do so at a microscopic and macroscopic level.
The purpose of this essay is to contribute to this growing body of litera-
ture, but from a different angle. In these pages I propose a political ontology
of the algorithm dispositif. Rather than advancing epistemic norms for algo-
rithmic reform (as one often finds in critical algorithm studies), my focus is
on the political ontology of this technical medium. Such a focus will allow us
to better understand the medium’s forms of participation and how its techni-
cal operations introduce transformations into the world. Specifically, I seek to
understand how algorithms arrange the movement of energies in space and
time, and how they do so automatically. This force of movement and arrange-
ment is what I refer to as the dispositional power of algorithms that I identify
as a political physics of vital processes.2
My investigation takes inspiration from, and shares a family resemblance
with, Aristotle’s analysis of action in Poetics (and simultaneously his analysis
of action in Politics and Nicomachean Ethics, as well as his analysis of
movement in Physics, Book II). As in Aristotle’s inquiry, I ask how a medium
operates and what forms of association does it introduce. For Aristotle the
answer lies in a study of the actions of the plot, or muthos (emplotment).3 He
asks, what does a tragedy do? And his answer is that it represents action.4
Aristotle’s analysis was (and is) radical and would have seemed counterintui-
tive to a Classical Greek readership; ancient poetry was not associated with
textual representation but with song and musicality. So to insist (and defend)
the idea that the task of poetic criticism is literary and that one’s judgments are
to be oriented toward a work’s purpose, or its telos as a unity, remains a foun-
dational ontological insight about the mimetic nature of a medium.5 I depart
from Aristotle in that I do not defend the idea that a technical medium requires
a teleological account of purpose, nor do I rely on Aristotle’s hylomorphic
account of the unity of form and matter. Instead I follow Gilbert Simondon

and Georges Canguilhem’s pluralist ontologies of technical objects in order to
better understand the dispositional powers of algorithms.6
An investigation into the political ontology of a technical medium is impera-
tive from a political, a critical, and an ethical perspective. Only by understanding
the multifaceted complexities of a medium’s modes of existence can we then
articulate a critical political vocabulary adequate to its forms of political partici-
pation. To invoke James Tully’s helpful formulation (offered in a different,
though related, context), if algorithms generate a “questionable regime” of “prac-
tices of governance” then an inquiry into their operational ontology is relevant to
our political understandings of that questionable regime.7 My argument is that in
order to appreciate algorithms as political media we must understand them as
automated systems for the perpetual arrangement and rearrangement of relations.
An algorithm’s agentialism, in other words, regards the recursive dispositional
ordering of bodies and energies in space and time, hence my preference for the
term “algorithm dispositif” that I explain and elaborate throughout.8
But why do we need to understand the political ontology of a medium?
My answer is at once simple and straightforward: if we are to make claims
about a medium’s political effects, then an investigation into the modes of
existence of technical objects, the relations these generate, and the transfor-
mations in perceptibilities, temporalities, and forms of association they intro-
duce is at once urgent and necessary. This essay is thus a continuation and
extension of my research in the aesthetics of politics that attempts to under-
stand how technical media take part in the agential life of a polity.9 In this
respect, I am in full agreement with Peter Polack’s analysis of algorithmic
governance that substitutes a focus on algorithm bias with an analysis of
algorithmic domains. As Polack notes, the critical approach of “algorithmic
reformism elides [the principle of algorithmic design] by proposing new
algorithmic solutions to satisfy the same design problems, neglecting to
acknowledge how design problems implicate consequences irrespective of
their particular algorithmic solutions.”10 My addendum to Polack’s argument
is to show how a political ontology of the algorithm dispositif helps us
acknowledge the political stakes of the design problems he analyzes.
Like Polack, Louise Amoore has forwarded a similar line of argument. In
Cloud Ethics she shows how an algorithm’s mode of functioning doesn’t track
well with the ambitions of epistemic reform in critical algorithm studies: “the
output of an algorithm is never simply true or false but is more precisely an
effect of the partial relations among entities,” she explains, and then continues,
“A pattern of false positives from a biometric algorithm, for example, can never
be simply false because the threshold is immanently adjustable.”11 Relying
on Polack’s and Amoore’s research, my point is this: a political theory of

Political Theory 49(1)
algorithms should not only attend to the epistemic and normative dilemmas that
emerge from algorithmic governance but also to the ways in which algorithmic
systems generate “new limits and thresholds of what it means to be human.”12 In
this regard, an already existing human solution to the problems of algorithmic
governance—like the reformist solution that relies on an individualist ideal of
agency, freedom, and responsibility—may not be enough because even if epis-
temic reforms are implemented, we are still dealing with the political participa-
tion of an automated system of everyday governance rooted in a mediatic
ontology that is virtual and founded on negative feedback systems of control.
Amoore, Polack, and others argue that algorithms are a different sort of
medium than traditional communication technology and, thus, that we need
to think differently about what political judgment means vis-à-vis a world of
algorithmic governance. In a world where human judgment is expected to be
either true or false, right or wrong, algorithms ground decision-making dif-
ferently basing it on distributed cognition, probability, correlation, and infi-
nite recursion and recombination of outcomes. The analogy Amoore shares
with her readers is at once helpful in understanding this distinction, but also
challenging; the system of order that has been introduced by algorithms is not
a Porphyrian one “that categorizes...

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT