<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.0 20120330//EN" "JATS-archivearticle1.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink">
  <front>
    <journal-meta />
    <article-meta>
      <title-group>
        <article-title>Towards a Conception of Sociotechnical Pathology</article-title>
      </title-group>
      <contrib-group>
        <aff id="aff0">
          <label>0</label>
          <institution>Tampere University</institution>
          ,
          <addr-line>Tampere</addr-line>
          ,
          <country country="FI">Finland</country>
        </aff>
      </contrib-group>
      <pub-date>
        <year>2021</year>
      </pub-date>
      <fpage>0000</fpage>
      <lpage>0001</lpage>
      <abstract>
        <p>The concept of social pathology has long belonged to the toolkit of social scientists, and several critical social philosophers have found it indispensable for linking social ontology to social criticism. While different conceptions of social pathology, as well as their applicability as diagnostic tools for social wrongs, have been debated, a common area of neglect becomes apparent when we consider pathological states of social wholes, such as societies, as not only socially but technically constituted. As a first step towards filling this gap, this paper introduces the concept of sociotechnical pathology. Drawing on existing work on social pathology, it discusses four different general conceptions of sociotechnical pathology as a diagnostic tool for analyzing socially and technically constituted social wrongs. The paper contributes to philosophical inquiry at the cross-section of critical social philosophy and philosophy of technology by paving way for substantive conceptions of sociotechnical pathology.</p>
      </abstract>
      <kwd-group>
        <kwd>sociotechnical pathology</kwd>
        <kwd>social pathology</kwd>
        <kwd>sociotechnical systems</kwd>
        <kwd>critical social philosophy</kwd>
        <kwd>social wrongs</kwd>
        <kwd>social ontology</kwd>
      </kwd-group>
    </article-meta>
  </front>
  <body>
    <sec id="sec-1">
      <title>Introduction</title>
      <p>
        Comparisons between the health of organisms and that of societies, be it in the form of
poetic metaphors or scientific concepts, have long persisted in vocabularies for
analyzing social issues, for better or worse. The concept of social pathology – although
debated in many respects – has been seen as indispensable for the project of social
philosophy in that it seems particularly apt for analyzing phenomena, processes, and
practices such as reification, ideology, alienation, invisibilization, social inequality,
exploitation, domination, and oppression
        <xref ref-type="bibr" rid="ref17 ref18 ref6 ref9">(see Zurn, 2011; Honneth, 2014; Laitinen &amp;
Särkelä, 2019)</xref>
        . The indispensability of the concept lies in the notion that, in analyzing
and criticizing specific wrongs of social life and of social entities (e.g., societies), lenses
of interpersonal morality and political legitimacy can fail to give us a grasp of wrongs
distinctively of a social kind
        <xref ref-type="bibr" rid="ref17 ref9">(Laitinen &amp; Särkelä, 2019; 2020)</xref>
        .
      </p>
      <p>Debates in social philosophy concern the proper conception of social pathology, and
a specific issue pertains to the biological and medical connotations of the term (see</p>
      <p>Copyright © 2021 for this paper by its authors.</p>
      <p>Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0)
Honneth, 2014; Laitinen &amp; Särkelä, 2019). In this paper, a common area of neglect
across specific substantive conceptions is highlighted: the technical constitution of
pathological conditions of social reality, which provides the backdrop for, and
intertwines with, their social constitution. From databases and automated
decision-making systems to networked digital infrastructures, technologies
(increasingly) enable and mediate practices which can be understood as pathological in
the sense described above. This highlights the need to consider the intertwining social
and technical causes of pathological conditions qua social wrongs. To this end, this
paper initiates a discussion on pathological conditions of social reality grounded in
sociotechnical arrangements and argues for the concept of sociotechnical pathology.
The general claim, that the concept of social pathology fails to account for the
technological, is stated in Section 2, although it is supported throughout the paper. The
paper proceeds to consider four ways of conceptualizing sociotechnical pathologies,
drawing on work in critical social philosophy and other fields such as critical algorithm
studies and AI ethics. Two normativist conceptions are discussed in Section 3, and two
naturalist conceptions in Section 4. I will not defend a specific substantive conception
of sociotechnical pathology, although for the sake of transparency I disclose that my
sympathies lie with the fourth conception. The final section provides a summary of the
paper.
2</p>
    </sec>
    <sec id="sec-2">
      <title>Philosophy, pathology, and the sociotechnical fabric</title>
      <p>
        The concept of social pathology is motivated by the seeming narrowness of moral and
political philosophy as lenses for analyzing what can be characterized as ‘social
wrongs’ – wrongful states, diseases, or disorders of social life, of social entities, or of
social reality as a whole
        <xref ref-type="bibr" rid="ref17 ref9">(Laitinen &amp; Särkelä, 2019; 2020)</xref>
        . Common examples of social
wrongs would include alienation, invisibilization, anomie, misrecognition, and
reification
        <xref ref-type="bibr" rid="ref18">(see ibid.; Zurn, 2011)</xref>
        . Given the distinct nature of these social wrongs, “the
concept of pathology seems a handy way of distinguishing the project of ‘critical social
philosophy’ from the projects of ‘political philosophy’ or ‘moral philosophy’” in that
“questions of moral rightness and political legitimacy appear one-sided and do not
grasp the specific disorder[s] addressed”
        <xref ref-type="bibr" rid="ref13 ref17 ref9">(Laitinen &amp; Särkelä, 2019, 87; Neuhouser,
2012)</xref>
        . To describe a social pathology as a failure to adhere to moral principles, as
injustice, as a failure of democracy, or as political illegitimacy, is to under-describe it,
even if such failures capture certain salient aspects of that pathology
        <xref ref-type="bibr" rid="ref13 ref17 ref4 ref9">(see Neuhouser,
2012; Harris, 2019; Laitinen &amp; Särkelä, 2019)</xref>
        . Put simply, moral and political
firstorder standards may fail to specify what is distinctively wrong in the case of social
wrongs.
      </p>
      <p>
        Notably, the biological and medical terminology of social pathology has raised
concern in discussions in the social sciences and social philosophy. Some take the term
‘pathology’ to naturalize social phenomena and to constitute a counterproductive way
of describing and criticizing social issues. Others find the notion of pathology as
indispensable for diagnosing social wrongs
        <xref ref-type="bibr" rid="ref17 ref6 ref9">(see discussions in Honneth, 2014; Laitinen
&amp; Särkelä, 2019)</xref>
        . This issue is discussed in this paper only to the extent that is
necessary for highlighting ontological commitments of specific conceptions of social
pathology.
      </p>
      <p>
        The concept of social pathology is generally meant to aid in the analysis of social
wrongs as distinct from other wrongs, and thereby implies a deviation from common
use of the concept of ‘wrongness’ as interchangeable with ‘wrongness’ in the sense of
interpersonal violations of morality or as injustice. Indeed, it is here taken that injustice,
for example, is a wrong-making feature alongside others, including other political
wrongs, and (non-)interpersonal violations of moral principles1. Hence, actions and
states-of-affairs can be, in this account, understood as wrong in different senses – wrong
qua morally wrong, wrong qua unjust, and so on. This demarcates the area of inquiry
and critique understood as critical social philosophy, distinguishing it from political
and moral philosophy: whereas, say, political ethics is concerned with political rights
and wrongs, critical social philosophy concerns itself with social rights and wrongs
        <xref ref-type="bibr" rid="ref17 ref9">(see
Laitinen &amp; Särkelä, 2019; 2020)</xref>
        .
      </p>
      <p>Given the general consensus regarding ethical and political issues related to
technology, it seems clear that sociotechnical analysis is necessary for understanding
the interrelated social and technical factors underlying many of the relevant wrongs.
However, to motivate the concept of sociotechnical pathology, I need to argue that this
is the case with social wrongs as well.
2.1</p>
      <sec id="sec-2-1">
        <title>Motivating the concept of sociotechnical pathology</title>
        <p>
          What motivates the concept of sociotechnical pathology as a diagnostic tool of specific
subset of wrongs? The reasoning here goes as follows: If social reality can suffer from
pathological conditions (i.e., if there can be social wrongs) and if technical objects are
at least minimally constitutive of social reality, then it seems social wrongs can be at
least minimally technically caused and/or constituted. In other words, if social
pathology refers to there being something wrong with the social fabric
          <xref ref-type="bibr" rid="ref10 ref12 ref2">(Laitinen &amp;
Särkelä, 2020, section 3)</xref>
          , sociotechnical pathology could be understood as something
being wrong with the ‘sociotechnical fabric’ – or what I refer to as a ‘sociotechnical
arrangements’ in the plural form. Accordingly, it seems a diagnostic tool for pathologies
of the social must, regardless of the substantive conception of sociotechnical pathology,
cover its technical causes (e.g., the functions and outputs of technical systems, or
‘performance’ by technical artefacts) and conditions of possibility (e.g., technological
resources, conditions, affordances, infrastructures, and networks, which constitute the
backdrop for the construction of social reality).
        </p>
        <p>
          This reasoning incorporates two core theses commonly maintained by sociotechnical
systems theorists: (1) both social and technical factors are constitutive of systems’
functioning and success, and that (2) interactions between these factors can involve
linear and non-linear causal relationships. This leaves open the possibility that
pathological conditions of social reality are undesigned and unexpected: social and
1 Roughly, interpersonal violations of moral principles occur when a person wrongs another,
while what I call non-interpersonal violations comprise wrongs against non-persons, such as
animals.
technical factors can malfunction, perhaps simultaneously, leading a sociotechnical
system (or “organism”) to a pathological state. Conversely, interventions on one or the
other, or both, can work to “cure” pathological states (or not). In this sense, with the
term ‘minimal constitution’ used above I simply refer to the myriad of ways
(networked) technical objects can causally contribute to the existence of pathological
conditions, directly and indirectly. For example, technical objects and material and
digital infrastructure can (and do) constitute the backdrop of social reality and activity.
In this sense, there is always a sociotechnical arrangement underlying social reality.
Technical objects – e.g., algorithmic decision-making systems – also increasingly
perform functions that actively maintain and transform social reality, for better or
worse. Hence, technology can be at least one cause in the causal chain leading to social
wrongs. For example, alienation, invisibilization, and ideologies such as colonialism
and capitalism, among other ills of the social typically diagnosed as pathologies
          <xref ref-type="bibr" rid="ref18">(see
Zurn, 2011)</xref>
          , take novel forms as they are increasingly mediated by such systems in the
hands of amalgams of public and private power
          <xref ref-type="bibr" rid="ref1 ref12 ref14 ref17">(see Eubanks, 2018; Noble, 2018;
Zuboff, 2019; Mohamed, Png &amp; Isaac, 2020)</xref>
          . Sociotechnical analysis, then, is not only
necessary for understanding the backdrop of social reality; it is necessary for in-depth
analysis and critique of sociotechnical arrangements and causes underlying
pathologies. As automated decision-making systems also exhibit what may be called
‘functional autonomy’ in that they can causally act on the world without being directly
manually operated, social wrongs can be exacerbated or amplified due to the
continuous and ubiquituous nature of their operation, as well as a lack of both
technical and organizational transparency. In these senses, understanding of pathological
conditions can be left too ‘thin’ if not complemented by sociotechnical analysis.
        </p>
        <p>Technology is primarily understood here as a backdrop of social activity or as
performing functions relevant to social reality and its state. There is, however, room in
this view for stronger arguments to the end that technical objects would themselves be
(quasi-)social (quasi-)agents in some relevant sense2, and thereby function as subjects
of pathological experience capable of being wronged. This would effectively
strengthen their ontological status as part of agential, social relations constitutive of
what have been traditionally understood as social entities in (critical) social ontology.
As my aim is to only pave way for detailed substantive conceptions of sociotechnical
pathology, I commit here only to a minimal view where technical objects can function
as (partial) metaphysical grounds of a social ontology.</p>
        <p>Now, a social pathologist – even if convinced by the relevance of technological
mediation for discussions of social pathology in some extra-theoretical sense, or with
respect to moral and political considerations – might regard the sociotechnical framing
as redundant or irrelevant to the diagnosis of pathologies. Indeed, could we not merely
ascribe instrumental status to technology? To do so, I maintain, would be to neglect
that the concept of the ‘sociotechnical’ is precisely meant to describe the
interdependence, intertwined-ness, and reciprocal construction of practices and
contexts irreducible to the social or the technical alone: social reality is (and has been)
a sociotechnical reality in the strict sense.</p>
        <sec id="sec-2-1-1">
          <title>2 For relevant discussions see Coeckelbergh (2012) and Gunkel (2012).</title>
          <p>
            This section has provided general motivation for the concept of sociotechnical
pathology. For present purposes, I shall hope the reader is convinced of the necessity
of sociotechnical analysis for understanding pathological conditions of social reality.
To pave way for substantive accounts in this respect, I will next sketch four different
conceptions of sociotechnical pathology following a four-fold classification for
conceptions of social pathology proposed in
            <xref ref-type="bibr" rid="ref9">Laitinen and Särkelä (2019)</xref>
            3. The first two
conceptions are called normativist conceptions as they ascribe the status ‘pathological’
to a subset of social wrongs or failures (Section 3). The other two conceptions are
dubbed naturalist; here, social pathology is the diagnosis of something as distinctively
a social wrong (Section 4). In other words, the first two conceptions maintain that social
failures are diagnosable as pathologies if they are wrong, while the second two use the
concept of pathology to identify social wrongs (Ibid., 87). I shall briefly review these
conceptions of social pathology, considering the implications of the notion of
sociotechnicality with respect to each conception. I also point to some possible pitfalls
and shortcomings along the way. I will not defend any specific conception, although
for the sake of transparency I note that my sympathies lie with the fourth conception.
3
          </p>
        </sec>
      </sec>
      <sec id="sec-2-2">
        <title>Normativist conceptions of sociotechnical pathology</title>
        <p>Normativist conceptions of social and sociotechnical pathology ascribe the status
‘pathological’ to a subset of social wrongs. The first conception discussed here is
antitheoretical as it maintains no common conceptual structure underlying social wrongs
can be found; it only uses ‘pathology’ as a label for socially criticizable issues.
Proponents of the second conception will disagree, holding that there is a unified structure
to be found in sociotechnical pathologies.
3.1</p>
      </sec>
      <sec id="sec-2-3">
        <title>An anti-theoretical conception of sociotechnical pathology</title>
        <p>
          According to the first (anti-theoretical and anti-naturalist) normativist view of social
pathology, normativist view of social pathology, the term 'social pathology' simply
refers to things that are in some sense social and wrong
          <xref ref-type="bibr" rid="ref17 ref9">(Laitinen &amp; Särkelä, 2019,
83–84)</xref>
          . Social pathology, in this view, is an umbrella term that captures family
resemblant practices, ideologies, and (pluralities of) dynamics, for instance,
which inhibit human flourishing, or violate conditions necessary for leading a
good life
          <xref ref-type="bibr" rid="ref17 ref4 ref9">(Laitinen &amp; Särkelä, 2019; see also Harris, 2019)</xref>
          . As the aforementioned
need not share in an inner logic or structure, this conception can be understood as
anti-theoretical
          <xref ref-type="bibr" rid="ref17 ref9">(Laitinen &amp; Särkelä, 2019)</xref>
          . As a diagnostic and a tool for social
critique, the conception of social pathology is nonetheless here “thicker” than traditional
liberal critique and, as such, cannot avoid from committing to at least a minimally
normative ethical view
          <xref ref-type="bibr" rid="ref4 ref5">(Honneth, 2007; Harris, 2019)</xref>
          . Proponents of this
conception might, however, see no reason beyond smooth communication for using the
3 For categorizations of conceptions of social pathology along different lines, see
          <xref ref-type="bibr" rid="ref4">Harris
(2019)</xref>
          .
concept of ‘pathology’ because the concept may carry “naturalizing, biologizing,
universalizing, medicalizing, organistic, vitalistic, and uncritical overtones”
          <xref ref-type="bibr" rid="ref17 ref9">(Laitinen
&amp; Särkelä, 2019, 84)</xref>
          .
        </p>
        <p>As this first conception provides no theoretical account of social pathology, not
much can be said about translating the concept into the language of sociotechnical
systems aside the general comments in Section 2 concerning the sociotechnical
constitutions and causes of social wrongs. Sociotechnical pathologies, according to this
anti-theoretical account, do not share in features or structure amenable to analysis aside
their wrongness, however. In this sense, sociotechnical analysis can only offer negative
characterizations of social wrongs, and retrospectively inform their sociotechnical
“etiology” once they are identified. For proponents of non-essentialist views of social
wrongs, the conception can yet be of use, perhaps.
3.2</p>
      </sec>
      <sec id="sec-2-4">
        <title>Dynamic and disorder: Structural conceptions of sociotechnical pathology</title>
        <p>
          A second set of normativist conceptions of social pathology is unified by the thesis that
social pathologies share a structure amenable to theoretical analysis. This theoretical
structure can be analyzed in anti-naturalist terms, however, without the need for
naturalistic biological, or medical terminology some consider problematic
          <xref ref-type="bibr" rid="ref17 ref9">(see
discussion in Laitinen &amp; Särkelä, 2019)</xref>
          . The shared structure underlying social
pathologies can be that of a ‘second-order disorder’
          <xref ref-type="bibr" rid="ref18">(Zurn, 2011)</xref>
          , or a negative
self-perpetuating dynamic
          <xref ref-type="bibr" rid="ref13 ref4">(Neuhouser, 2012; see also Harris, 2019)</xref>
          , for example.
        </p>
        <p>
          In Christopher Zurn’s view, pathologies, including “ideological recognition,
maldistribution, invisibilisation, rationality distortions, reification and institutionalised
self-realisation”, share a conceptual structure in that they “all operate by means of
second-order disorders”
          <xref ref-type="bibr" rid="ref18">(Zurn, 2011, 345)</xref>
          . With second-order disorders Zurn refers to
“constitutive disconnects between first-order contents and second-order reflexive
comprehension of those contents, where those disconnects are pervasive and socially
caused” (2011, 345–346). Such a disconnect may hold between the deliberate
invisibilization of certain groups, and the group’s members’ experiences and
comprehension of this wronging, for example. An alternative view is that social
pathologies take the form of an almost unstoppable, self-perpetuating dynamic which
exacerbates initial bad circumstances, possibly even without social agents
acknowledging this dynamic
          <xref ref-type="bibr" rid="ref13">(Neuhouser, 2012)</xref>
          . Issues such as colonialism, capitalism,
and global warming would seem to lend themselves to social pathology diagnoses as
either disorders or (pluralities of) vicious dynamics, depending on the view.
        </p>
        <p>
          Could the presently discussed normativist conceptions of social pathology provide a
theoretical basis for conceptualizing sociotechnical pathologies? To start, a diverse
literature on data- and algorithmically driven forms of capitalism and colonialism,
exploitation, oppression, and dispossession
          <xref ref-type="bibr" rid="ref12 ref14 ref17">(Noble, 2018; Mohamed, Png &amp; Isaac,
2020; Zuboff, 2019)</xref>
          suggests various forms of technologically mediated social
wronging are irreducible to violations of principles of interpersonal morality or lack of
political legitimacy alone. Not only do they operate also at distinctively social areas of
life – at the level of the social fabric itself – but they feed on structural inequalities,
asymmetric relations of power, resource- and data-hungry computation, and thus have
effects that traverse and transform spheres of (social) life. These patterns of social
wronging often involve layered opacity of which the technology is partially
constitutive: “black box” algorithms are used by “black box” companies and public
organizations (or amalgams thereof), and this layered lack of transparency effectively
denies subjects informational resources necessary for reflexive comprehension.
Individuals may not know why, how, and when their data is collected or, whether and
how an algorithm configures decisions regarding their fate, and whether the
aforementioned are justified
          <xref ref-type="bibr" rid="ref1">(see, e.g., Eubanks, 2018)</xref>
          . The conception of
sociotechnical pathology as a second-order disorder seems to capture something
important, accordingly: technology is not a mere instrument constitutive of the
first-order conditions of sociotechnical reality – sociotechnical arrangements can also
prevent subjects’ comprehension of them, and effectively obstruct social critique.
        </p>
        <p>
          Likewise, the self-perpetuating negative dynamic discussed by
          <xref ref-type="bibr" rid="ref13">Neuhouser (2012)</xref>
          seems apt for describing systemic failures of sociotechnical arrangements and practices.
Consider, say, feedback loops in algorithmic decision-making
          <xref ref-type="bibr" rid="ref11">(see, e.g., Lum &amp; Isaac,
2016)</xref>
          : Algorithmic systems are first fed structurally biased data, after which they enact
on that data, further disadvantaging the disadvantaged. The resulting data is captured
and fed back into the systems as ‘ground truth’ for future decisions. The loop continues
in virtue of the self-reinforcing nature of the sociotechnical arrangement, including the
data pipeline. To the extent that this is a common pattern – and, indeed, it seems to be
– the notion that sociotechnical pathologies share such a self-perpetuating structure
more generally has some initial credibility.
        </p>
        <p>I should here emphasize that the technological aspect of the discussed social wrongs
is indispensable as an object of analysis in both cases, be it as second-order disorders –
i.e., disconnections between sociotechnical reality and subjects’ comprehension of it –
or ‘negative spiral’ dynamics. By acknowledging the fact that the technology is partly
constitutive of the pathological disconnect or dynamic, and by incorporating the
technical layer into our conception of pathology, we fill the gap that is unbridgeable
within the framing of social pathology alone.</p>
        <p>
          Now, there are certain problems with the conception of sociotechnical pathology qua
second-order disorder or dynamic. Regarding pathologies as second-order disorders,
Laitinen and Särkelä aptly note that “the fault need not lie in the disconnect between
reality and reflection, but in the social reality itself” (2019, 85). Indeed, both
invisibilization and brute force can be reflexively recognized by individuals
experiencing such wronging, for example, although the contrary may also be the case
in cases of “learnt self-invisibilization” (ibid.). Hence, this conception seems to neglect
pathological first-order wrongs where no disconnect occurs. This critique arguably
applies to the analogous conception of sociotechnical pathology: while social wrongs
in the first-order – e.g., algorithmic oppression
          <xref ref-type="bibr" rid="ref14">(Noble, 2018)</xref>
          – can function opaquely
due to the “black box” nature of both modern machine learning algorithms and
companies using them, and thereby create disconnections between reality and
reflection, this is not necessarily the case. We need only look at contemporary forms of
technological resistance to verify this claim. For example, “protective optimization
technologies”, which fool machine learning systems and meddle with their predictions,
are used to counter the pervasive optimization logics built into decision-making
systems, and to disarm technological surveillance
          <xref ref-type="bibr" rid="ref8">(Kulynych et al., 2020)</xref>
          . There seems
to be little reason to suppose that pathologies, such as oppressive sociotechnical
practices and arrangements, could not be reflexively experienced as wronging. As
Laitinen and Särkelä note (2019, 96), the conception of pathology as a second-order
disorder seems to identify “reflexive” as opposed to “social” pathologies.
        </p>
        <p>
          Regarding the dynamic conception of sociotechnical pathology, it would seem that
first-order wrongs as sociotechnical phenomena can exhibit structural features other
than unstoppable self-perpetuating dynamics. Conceptually speaking, structural
similarities can likely be found underlying cases where algorithmic systems reproduce
social inequality, but it is not clear that this applies to sociotechnically enabled or
induced alienation and invisibilization, for example. It is also possible that there are
diverse structures for wrongs under a single category, such as those mentioned
previously, which means that a normativist conception might benefit from a more
complex and nuanced account of pathological structures underlying sociotechnical
phenomena
          <xref ref-type="bibr" rid="ref17 ref9">(cf. Laitinen &amp; Särkelä, 2019, 86–87)</xref>
          . Alternatively, one might adopt a
naturalist approach, where the question regarding unified theoretical structure does not
similarly arise. These will be considered next.
4
        </p>
      </sec>
    </sec>
    <sec id="sec-3">
      <title>Naturalist conceptions of sociotechnical pathology</title>
      <p>
        Proponents of naturalist views use the concept of social pathology in a contentful
manner in order to identify what is wrong with social reality
        <xref ref-type="bibr" rid="ref17 ref9">(Laitinen &amp; Särkelä, 2019,
87)</xref>
        . In the organistic view, social pathology is conceptualized as a disease in a social
entity, which, in this view, shares a structure similar to that of a biological organism.
Another conception, the so-called processual conception, conceives of pathologies as
stagnation and degeneration in natural social lifeprocesses. I will argue that, as
technology can be understood as an inorganic complement or ‘prosthetic’ harnessed for
the reproductive aims of social wholes, both conceptions provide a suitable conceptual
scheme for substantive conceptions of sociotechnical pathology, although the latter
conception may be preferable.
4.1
      </p>
      <sec id="sec-3-1">
        <title>The organistic conception of sociotechnical pathology</title>
        <p>
          The first conception embraces the naturalistic terminology of pathology: a social
pathology is an ‘illness’ in a social organism – specifically, a deviation from the
“reproductive values and ends of society”
          <xref ref-type="bibr" rid="ref17 ref9">(Laitinen &amp; Särkelä, 2019, 88)</xref>
          . This
‘organicist’ view is attributable, most notably, to Durkeim and his analysis of anomie
([1895] 2013) but it has gained new life in the hands of
          <xref ref-type="bibr" rid="ref6">Honneth (2014)</xref>
          who
rehabilitated the idea of society as an organism and social pathology as a ‘disease of
society’. Importantly, social pathology as illness or disease is not a phenomenon
diagnosed at the level of individuals, or even the “organs” of the society-as-organism,
but the organism as a whole. In virtue of this framing, the organicist view allows for
diagnosing pathologies not as family resemblant wrongs (with the cost of an
antitheoretical approach to social pathology) or as wrongs sharing a second-order
property (with the cost of misdiagnosing first-order social wrongs as non-pathological),
but as ones interfering with the reproductive function of the social whole (insofar as
that whole takes an organistic form). Thus, whereas normativist conceptions are
uncomfortable with the naturalistic terminology, and would indeed dispose of it, “[t]he
critical force of the concept” in the organistic view lies precisely “in the supposition
that societal reproduction can fail by analogy to the way in which the self-maintenance
of a living organism is disrupted when it falls ill”
          <xref ref-type="bibr" rid="ref17 ref9">(Laitinen &amp; Särkelä, 2019, 89)</xref>
          .
Notably, within the organicist view “[m]orality and politics can conflict with other
institutional spheres of the societal whole”, and so they too can “fail at their
reproductive tasks, in which case they will be diagnosed as requiring revision” (ibid.).
        </p>
        <p>How to accommodate sociotechnical diagnosis of pathology with the organistic
view? One option is to understand technology as a ‘prosthetic’ part of the social whole4.
Here, technologies are functional replacements of (or complements to) the organism’s
functional subcomponents, its “organs”. Consider, say, Big Data datasets and
automated decision-making systems as technical prosthetics of social organisms.
Datasets could be understood from a Derridian technology-as-prosthesis viewpoint as
augmenting a social organism’s memory in that they perform and complement the
archiving processes necessary for the reproduction of the social whole. The
implementation of automated decision-making systems in the public sector is often
guided the vision that institutional functions are optimized, made more effective,
consistent, and objective by “tapping into” this prosthetic memory. In this sense,
datasets as technology not only function as archives and memories, but they enable the
optimized reproduction of the self-maintaining social organism through interaction
with other technologies-as-organs, such as automated optimization systems.</p>
        <p>Now, within this view, a pathological condition can be understood as sociotechnical
insofar as disruption of the social organism’s self-maintaining process is attributable to
technical factors (in addition to social ones). The technical object as prosthetic can be,
on the one hand, directly rejected by the social whole akin to how transplants or
artificial limbs can irritate the body and be rejected by it. This would be the case when
a specific technology or technological condition is reflexively taken as unfit to serve
the ends of the social whole. For example, the use of carbon-intensive technology
running on fossil fuels can be understood as pathological in this sense. There is a large
consensus on addressing the climate crisis, and various ideological dynamics, such as
capitalist and colonialist logics qua social patterns, contribute to (and reproduce) the
present pathological condition of inaction. Importantly, however, from the
sociotechnical point of view, technology itself is one causal reason of the organism’s
disease by effectively and actively undermining conditions of its self-maintenance in a
causal-functional sense.</p>
        <p>
          The notion of the interrelatedness and reciprocal construction of the social and the
technical is essential here, as ideological narratives, for example, can distort
second-order reflexivity regarding pathological conditions. Under pathological
4 An alternative view is a relational one, where the ontological status of technical artefacts is less
neatly separable from questions of epistemology. Relational views have been di
          <xref ref-type="bibr" rid="ref15">scussed in
Coeckelbergh (2017</xref>
          ) and
          <xref ref-type="bibr" rid="ref3">Gunkel (2012)</xref>
          , for example.
conditions, “technosolutionist” narratives can create vicious cycles of alienation akin
to what was proposed by
          <xref ref-type="bibr" rid="ref13">Neuhouser (2012)</xref>
          . Mark Coeckelbergh highlights this issue
vividly by raising the concern of AI technology possibly becoming an “alienation
machine”, that is, “an instrument to leave the Earth and deny our vulnerable, bodily,
earthly, and dependent existential condition” (2020, 196). This is an example of a more
indirect pattern of disruption where technical affordances (or lack thereof) configure
second-order reflection. Note that the pathology is not in these cases “attributed to the
many suffering individuals, not even to the malfunctioning institutional ‘organs’ but
only to the social organism”
          <xref ref-type="bibr" rid="ref17 ref9">(Laitinen &amp; Särkelä, 2019, 89)</xref>
          . Given the sociotechnical
bases of the pathological condition, however, the “interface” between the social and
prosthetic organs of the organism cannot be omitted from diagnosis. Threads of the
social fabric, language (e.g., narrative), technology, and social activity mutually shape
each other (Coeckelbergh, 2017), and the sociotechnical fabric as a whole, respectively.
        </p>
        <p>
          As was stated above, morality and politics can also fail their reproductive functions
within the organistic view. Accordingly, it may be that ethical and political standards
for technology development and use need to be revised to safeguard the organism’s
technologically mediated self-maintaining process. Concrete examples abound, as
several calls have been made for approaches to ethical governance of technology better
suited to criticize existing power asymmetries in contemporary societies
          <xref ref-type="bibr" rid="ref7">(Kalluri,
2020)</xref>
          , more reflective of social inequalities and histories of oppression
          <xref ref-type="bibr" rid="ref2">(Fazelpour &amp;
Lipton, 2020)</xref>
          , and which involve public and deliberative forms of democratic
decision-making
          <xref ref-type="bibr" rid="ref16">(Wong, 2020)</xref>
          . Within the organistic view, then, a pathological
condition where the operative first-order normative standards are considered incapable
of serving the reproduction of social values and ends can also be indirectly technically
constituted, as it were. In such cases, the irritation located at the interface of the
prosthetics and the organism is indicative of a disconnect between the ends of the social
whole, on the one hand, and the standards applied in assessing the fit between social
and technical components of the organism, on the other.
        </p>
        <p>
          Now, the organicist conception is vulnerable to certain criticisms which I do not
consider here5. However, one significant disadvantage of the organicist view I wish to
highlight here is “that the organism analogy squeezes society into such a static shape
that radical social critique becomes impossible”
          <xref ref-type="bibr" rid="ref17 ref9">(Laitinen &amp; Särkelä, 2019, 91)</xref>
          . In other
words, as the reproductive ends of social organisms are taken as given, critique of those
ends would itself be understood as pathological. This implies that sociotechnical
arrangements and conditions of possibility cannot be criticized within this view insofar
as they serve the values and ends of the social whole, whatever they may be. This is
problematic, firstly, because the need for critique of societal ends informed by
sociotechnical considerations is partly what motivates the concept of sociotechnical
pathology to begin with (cf. ibid.). Secondly, this seems to also, in an indirect manner,
limit the space for criticism of first-order normative content – namely, the ethical and
political standards applicable to evaluation of sociotechnical arrangements. As the
normative force of critique of first-order contents will likely be derived partly from
5
          <xref ref-type="bibr" rid="ref9">Laitinen and Särkelä (2019)</xref>
          discuss, for example, problems with ascribing similar idealized
structures to organisms and the societies.
conceptions of societal ends
          <xref ref-type="bibr" rid="ref5">(see Honneth, 2007)</xref>
          , the “naturalization” of the operative
ends of a social whole can shield first-order normative standards from critique insofar
as those standards are aligned with the naturalized ends.
        </p>
        <p>A viable conception of sociotechnical pathology should retain the diagnostic power
of the concept afforded by the naturalist frame without committing one to a view with
no room for transformative social critique. The last conception discussed here promises
to do just that.
4.2</p>
      </sec>
      <sec id="sec-3-2">
        <title>Sociotechnical pathology as stagnation and degeneration of social life</title>
        <p>Another naturalist conception of social pathology builds mutability into the very
concept of social reality by understanding pathological states as things being wrong
with social lifeprocesses. Laitinen and Särkelä, who attribute the ontological view
underlying this conception to John Dewey, summarize its ontological commitments as
follows: “Whereas organicism rests on the conception of the organism as an ideally
organized self-maintaining substance […] the idea of a distinctively social life, by
contrast, is committed to the idea of a lifeprocess operating above and beneath the living
body and conceives life as irreducible to the organism” (2019, 94). Social life,
according to this conception, can develop into a pathological state by losing its
“transformative growth” and stagnating into “a merely organic process”, or by failing
to reproduce its form, degenerating “into mere inorganic processuality” (Ibid.). For the
social pathologist, the diagnosis thus consists not in locating a systemic disorder or
malfunction; the pathology is found in the absence of transformative growth natural to
social lifeprocesses or in the degeneration of that life into “mere inorganic
processuality” (Ibid.). In contrast to the organicist view, mere maintenance of form is
not sufficient for social life, which is by definition growing, although it is necessary to
the extent that form is required for reproduction of that life above a mere organic genus
of living: “social processes need to disintegrate in order to integrate” and critique, in
this view, “become[s] a medium of life”, respectively (ibid., 95). Importantly, then, as
this conception builds critique into the notion of social life, it retains both the necessity
and transformative power of social critique as a core task of critical social philosophy
without adopting the restrictive ontology of the organistic view.</p>
        <p>
          The processual view of social reality seems a viable scheme for conceptualizing
sociotechnical pathology if we understand technology as prosthetic to the social
lifeprocess, respectively. A sociotechnical arrangement, according to this conception,
is pathological if it either (i) contributes to the stagnation of social life or (ii)
degenerates social life into mere processuality (as a genus of life). The advantage of
this view is that it seems to capture many (if not all) previously discussed social wrongs
in the production of which technical conditions and (partial) causes play a vital role.
For example, it would seem to capture the social wrongs of algorithmic colonialism and
surveillance capitalism in a manner faithful to the (partially) technical constitution of
such wrongs with the social fabric: they are wrong qua pathological in that they stagnate
social life, and the technical mediation of these ideological practices amplifies that
wrongness by making them ubiquitous and opaque, thus escaping reflection by the
social whole. Furthermore, we can also diagnose sociotechnical practices and
arrangements as pathological when they do not necessarily exacerbate but “merely”
reproduce failures of social life. The reproduction of social inequality and structural
oppression with and through automated systems
          <xref ref-type="bibr" rid="ref1 ref14">(Eubanks, 2018; Noble, 2018)</xref>
          , for
example, lends itself to a diagnosis of pathology (in addition to involving apparent
moral and political wrongs). In these cases, social life is pathological because it
effectively stagnates its natural transformative growth through technology.
Sociotechnical arrangements that generate and sustain ‘feedback loops’ (see above), for
example, reproduce the very technological conditions actively preventing
transformative growth: as data that capture social inequalities are fed back into the
prosthetic memory of social life in a looped manner, pathological conditions of social
life are further cemented. Note that this is analogical to the issue considered with the
organistic view: social life is pathological because it is taking the baseline consideration
for what the social fabric ought to be like6 as given, constructing the (in)organic bases
of social life according to ends that are closed and pre-determined, as opposed to open
to critical scrutiny.
        </p>
        <p>
          Importantly, however, the ontological commitments of the processual conception do
not preclude the possibility of social critique. Technology can be understood as
prosthetically complementary to the reproduction of values and ends of social life
without prescribing it these ends in a “naturalistic” manner. That is, the presented ends
of social life can be criticized, although this can only be done from within social life
itself; and indeed, critique is part of that social life
          <xref ref-type="bibr" rid="ref17 ref9">(see Laitinen &amp; Särkelä, 2019, 94–
96)</xref>
          . Hence, this view accommodates for both the reproductive and transformative aims
of technology development and use without closing out the possibility for social
criticism when actual sociotechnical arrangements fail to deliver on those promises.
Because critique is necessary for social lifeprocesses under any given sociotechnical
arrangement, this view can retain an optimism regarding technology without dissolving
into ‘technological solutionism’ or ‘technochauvinism’ (Broussard, 2018), without
allowing technology to become an ‘alienation machine’ (Coeckelbergh, 2020, 196), and
without rendering critical social philosophy toothless at the face of pathologies. Note
that it also allows for understanding sociotechnical practices of resistance as critique
without diagnosing them as necessarily pathological. Within the organistic view, the
use of protective optimization technologies in the name of political resistance, for
example, would be considered pathological if it stands in opposition to the reproductive
aims of the social whole. In contrast, here, the exercise of resistance with and through
technology can effectively expose pathological conditions of social life by making
technological systems’ logics of optimization and utility maximization transparent and
thereby socially contestable
          <xref ref-type="bibr" rid="ref8">(Kulynych et al., 2020)</xref>
          . Thus, the processual view of
sociotechnical pathology can provide the critical social philosopher a more nuanced
view of social critique as it is conducted in technologically mediated ways.
        </p>
        <p>
          Social critique could be itself pathological, however, by leading to the death of
reproductive ends worth preserving or, conversely, by blocking transformative practice
by preserving societal values
          <xref ref-type="bibr" rid="ref17 ref9">(Laitinen &amp; Särkelä, 2019, 95)</xref>
          . Insofar as the practice of
social critique itself is understood as self-transformative and fallibilistic concerning its
        </p>
        <sec id="sec-3-2-1">
          <title>6 For a discussion on ought to be -norms, see Laitinen &amp; Särkelä (2020).</title>
          <p>
            own method, this i
            <xref ref-type="bibr" rid="ref15">s not an issue (see Särkelä, 2017</xref>
            ). The present frame, however,
highlights that critique also operates from within the sociotechnical arrangements that
constitute its object. To remain truly self-transformative and fallibilistic, the pathologist
needs to also account for how the technical mediates the relationship between
“metacritical” and critical dimensions of sociotechnical diagnosis, respectively.
5
          </p>
        </sec>
      </sec>
    </sec>
    <sec id="sec-4">
      <title>Conclusions</title>
      <p>This paper has discussed the sociotechnical constitution and causes of social wrongs
qua pathological conditions of social reality. The primary aim has been to introduce and
motivate the concept of sociotechnical pathology, which bridges the gap between
critical social philosophy and philosophy of technology by framing sociotechnical
systems and arrangements as indispensable parts of social (or sociotechnical) ontology,
and thereby the analysis and critique of pathological conditions of, or “wrongs-with”,
the social fabric. The concept is to aid critical social philosophy in the analysis of social
wrongs, such as algorithmic oppression, algorithmic colonialism, surveillance
capitalism, among other social ills, such as alienation and loss of meaning. Four
conceptions of sociotechnical pathology were discussed:



</p>
      <p>The anti-theoretical, normativist conception: ‘sociotechnical pathology’ refers to
diverse sociotechnically constituted social wrongs with no unifying theoretical
structure.</p>
      <p>The structural, normativist conception: ‘sociotechnical pathology’ refers to
sociotechnically constituted social wrongs which have a common theoretical
structure amenable to analysis.</p>
      <p>The organistic, naturalist conception: a sociotechnical practice or arrangement is
pathological when it obstructs or prevents the reproduction of the values or ends of
a social whole.</p>
      <p>The processual, naturalist conception: a sociotechnical practice or arrangement is
pathological when it stagnates or degenerates the lifeprocess of a social whole.</p>
      <p>The general outlines offered here hopefully aid others on their journey towards
constructing substantive conceptions of sociotechnical pathology, at least by pointing
to some possible pitfalls along the road. In any case, the general argument for the
necessity of sociotechnical considerations in analyzing social wrongs can be regarded
as independent from this mapping exercise. There are surely other ways of
conceptualizing sociotechnical pathologies, but any conception of sociotechnical
pathology will plausibly concede some version of the general argument put forth here.
Broussard, M. (2018). Artificial unintelligence: How computers misunderstand the world. MIT</p>
      <p>Press.</p>
      <p>Coeckelbergh, M. (2012). Growing moral relations: Critique of moral status ascription. Palgrave</p>
      <p>Macmillan.</p>
      <p>Coeckelbergh, M. (2017). Using words and things: Language and philosophy of technology.</p>
      <p>Taylor &amp; Francis.</p>
      <p>Coeckelbergh, M. (2020). AI ethics. MIT Press.</p>
      <p>Durkheim, É. ([1895] 2013) The Division of Labour in Society. Free Press.</p>
    </sec>
  </body>
  <back>
    <ref-list>
      <ref id="ref1">
        <mixed-citation>
          <string-name>
            <surname>Eubanks</surname>
            ,
            <given-names>V.</given-names>
          </string-name>
          (
          <year>2018</year>
          ).
          <article-title>Automating inequality: How high-tech tools profile, police, and punish the poor</article-title>
          .
          <source>St. Martin's Press.</source>
        </mixed-citation>
      </ref>
      <ref id="ref2">
        <mixed-citation>
          <string-name>
            <surname>Fazelpour</surname>
            ,
            <given-names>S.</given-names>
          </string-name>
          , &amp;
          <string-name>
            <surname>Lipton</surname>
            ,
            <given-names>Z. C.</given-names>
          </string-name>
          (
          <year>2020</year>
          ).
          <article-title>Algorithmic fairness from a non-ideal perspective</article-title>
          .
          <source>Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society</source>
          , pp.
          <fpage>57</fpage>
          -
          <lpage>63</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref3">
        <mixed-citation>
          <string-name>
            <surname>Gunkel</surname>
            ,
            <given-names>D. J.</given-names>
          </string-name>
          (
          <year>2012</year>
          ).
          <article-title>The machine question: Critical perspectives on AI, robots, and ethics</article-title>
          . MIT Press.
        </mixed-citation>
      </ref>
      <ref id="ref4">
        <mixed-citation>
          <string-name>
            <surname>Harris</surname>
            ,
            <given-names>N.</given-names>
          </string-name>
          (
          <year>2019</year>
          ).
          <article-title>Recovering the critical potential of social pathology diagnosis</article-title>
          .
          <source>European Journal of Social Theory</source>
          ,
          <volume>22</volume>
          (
          <issue>1</issue>
          ), pp.
          <fpage>45</fpage>
          -
          <lpage>62</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref5">
        <mixed-citation>
          <string-name>
            <surname>Honneth</surname>
            ,
            <given-names>A.</given-names>
          </string-name>
          (
          <year>2007</year>
          ).
          <article-title>Pathologies of the social: the past and present of social philosophy</article-title>
          .
          <source>Disrespect: The Normative Foundations of Critical Theory</source>
          ,
          <string-name>
            <surname>Honneth</surname>
          </string-name>
          , A (ed.). Cambridge: Polity Press.
        </mixed-citation>
      </ref>
      <ref id="ref6">
        <mixed-citation>
          <string-name>
            <surname>Honneth</surname>
            ,
            <given-names>A.</given-names>
          </string-name>
          (
          <year>2014</year>
          ).
          <article-title>The diseases of society: Approaching a nearly impossible concept</article-title>
          , Särkelä,
          <string-name>
            <surname>A.</surname>
          </string-name>
          <source>(trans.)</source>
          .
          <source>Social Research: An International Quarterly</source>
          ,
          <volume>81</volume>
          (
          <issue>3</issue>
          ), pp.
          <fpage>683</fpage>
          -
          <lpage>703</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref7">
        <mixed-citation>
          <string-name>
            <surname>Kalluri</surname>
            ,
            <given-names>P.</given-names>
          </string-name>
          (
          <year>2020</year>
          ).
          <article-title>Don't ask if artificial intelligence is good or fair, ask how it shifts power</article-title>
          .
          <source>Nature</source>
          ,
          <volume>583</volume>
          (
          <issue>7815</issue>
          ), pp.
          <fpage>169</fpage>
          -
          <lpage>169</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref8">
        <mixed-citation>
          <string-name>
            <surname>Kulynych</surname>
            ,
            <given-names>B.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Overdorf</surname>
            ,
            <given-names>R.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Troncoso</surname>
            ,
            <given-names>C.</given-names>
          </string-name>
          , &amp;
          <string-name>
            <surname>Gürses</surname>
            ,
            <given-names>S.</given-names>
          </string-name>
          (
          <year>2020</year>
          ).
          <article-title>POTs: protective optimization technologies</article-title>
          .
          <source>Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency</source>
          , pp.
          <fpage>177</fpage>
          -
          <lpage>188</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref9">
        <mixed-citation>
          <string-name>
            <surname>Laitinen</surname>
            ,
            <given-names>A.</given-names>
          </string-name>
          , &amp;
          <string-name>
            <surname>Särkelä</surname>
            ,
            <given-names>A.</given-names>
          </string-name>
          (
          <year>2019</year>
          ).
          <article-title>Four conceptions of social pathology</article-title>
          .
          <source>European journal of social theory</source>
          ,
          <volume>22</volume>
          (
          <issue>1</issue>
          ), pp.
          <fpage>80</fpage>
          -
          <lpage>102</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref10">
        <mixed-citation>
          <string-name>
            <surname>Laitinen</surname>
            ,
            <given-names>A.</given-names>
          </string-name>
          , &amp;
          <string-name>
            <surname>Särkelä</surname>
            ,
            <given-names>A.</given-names>
          </string-name>
          (
          <year>2020</year>
          ).
          <article-title>Social wrongs</article-title>
          .
          <source>Critical Review of International Social and Political Philosophy</source>
          , pp.
          <fpage>1</fpage>
          -
          <lpage>25</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref11">
        <mixed-citation>
          <string-name>
            <surname>Lum</surname>
            ,
            <given-names>K.</given-names>
          </string-name>
          , &amp;
          <string-name>
            <surname>Isaac</surname>
            ,
            <given-names>W.</given-names>
          </string-name>
          (
          <year>2016</year>
          ).
          <article-title>To predict and serve?</article-title>
          .
          <source>Significance</source>
          ,
          <volume>13</volume>
          (
          <issue>5</issue>
          ), pp.
          <fpage>14</fpage>
          -
          <lpage>19</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref12">
        <mixed-citation>
          <string-name>
            <surname>Mohamed</surname>
            ,
            <given-names>S.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Png</surname>
            ,
            <given-names>M. T.</given-names>
          </string-name>
          , &amp;
          <string-name>
            <surname>Isaac</surname>
            ,
            <given-names>W.</given-names>
          </string-name>
          (
          <year>2020</year>
          ).
          <article-title>Decolonial AI: Decolonial theory as sociotechnical foresight in artificial intelligence</article-title>
          .
          <source>Philosophy &amp; Technology</source>
          ,
          <volume>33</volume>
          (
          <issue>4</issue>
          ), pp.
          <fpage>659</fpage>
          -
          <lpage>684</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref13">
        <mixed-citation>
          <string-name>
            <surname>Neuhouser</surname>
            ,
            <given-names>F.</given-names>
          </string-name>
          (
          <year>2012</year>
          ).
          <article-title>Rousseau und die Idee einer pathologischen Gesellschaft</article-title>
          .
          <source>Politische Vierteljahresschrift</source>
          <volume>53</volume>
          (
          <issue>4</issue>
          ), pp.
          <fpage>628</fpage>
          -
          <lpage>745</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref14">
        <mixed-citation>
          <string-name>
            <surname>Noble</surname>
            ,
            <given-names>S. U.</given-names>
          </string-name>
          (
          <year>2018</year>
          ).
          <article-title>Algorithms of oppression: How search engines reinforce racism</article-title>
          . NYU Press.
        </mixed-citation>
      </ref>
      <ref id="ref15">
        <mixed-citation>
          <string-name>
            <surname>Särkelä</surname>
            ,
            <given-names>A.</given-names>
          </string-name>
          (
          <year>2017</year>
          ).
          <article-title>Immanent critique as self-transformative practice: Hegel, Dewey, and contemporary critical theory</article-title>
          .
          <source>The Journal of Speculative Philosophy</source>
          ,
          <volume>31</volume>
          (
          <issue>2</issue>
          ), pp.
          <fpage>218230</fpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref16">
        <mixed-citation>
          <string-name>
            <surname>Wong</surname>
            ,
            <given-names>P. H.</given-names>
          </string-name>
          (
          <year>2020</year>
          ).
          <article-title>Democratizing algorithmic fairness</article-title>
          .
          <source>Philosophy &amp; Technology</source>
          ,
          <volume>33</volume>
          (
          <issue>2</issue>
          ), pp.
          <fpage>225</fpage>
          -
          <lpage>244</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref17">
        <mixed-citation>
          <string-name>
            <surname>Zuboff</surname>
            ,
            <given-names>S.</given-names>
          </string-name>
          (
          <year>2019</year>
          ).
          <article-title>The Age of Surveillance Capitalism: The Fight for the Future at the New Frontier of Power</article-title>
          . Profile Books.
        </mixed-citation>
      </ref>
      <ref id="ref18">
        <mixed-citation>
          <string-name>
            <surname>Zurn</surname>
            ,
            <given-names>C.</given-names>
          </string-name>
          (
          <year>2011</year>
          ).
          <article-title>Social pathologies as second-order disorders. Axel Honneth: Critical Essays: with a reply by Axel Honneth</article-title>
          , Petherbridge, D (ed.). Brill.
        </mixed-citation>
      </ref>
    </ref-list>
  </back>
</article>