1 Introduction

We live in a world undergoing digital transformation. This is a process quite novel in the course of human history, where “digital” is understood not ‘merely in terms of ones and zeros but rather as the capacity to process data’ and the ‘automation of data processing’ (Hui, 2016, p. 35; p. 54). Given what we know about previous periods, it has been relatively easy for us to see how the development of digital technologies has opened new possibilities for human life. In our creative industries it might be said that ‘we are living through a digital renaissance’ of popular culture and society (Waldfogel, 2018, p. 253). And despite disagreements on measurement, the perception remains that ‘the growth of the digital economy is unprecedented and has been a major contributor to recent economic growth’ (Moulton, 2000, p. 34). Some have even taken it as far as to suggest that human flourishing in the advanced digital age means we should ‘retreat to “virtual” worlds that are created and sustained by the infrastructure that we have built’ (Danaher, 2019, p. 3). In short, the future is digital, and it offers us enormous opportunities.

Yet the digital world is not without serious risks. Digital can exacerbate existing modern technological issues of ‘efficiency, instrumentalization, domination of humans, tyranny, alienation, and the end of mankind’ (Van de Poel, 2020, p. 502), and present new, unexpected crises. Amongst theorists of technology, there is now a well-established literature on modern technological pessimism, including the work of Martin Heidegger, Jacques Ellul, and Lewis Mumford. Critical approaches stress the presence of ‘technological hegemony’ (Feenberg, 1992, p. 309) within overlapping areas such as gender (Wajcman, 1991), race (Benjamin, 2019), and human rights (Jørgensen, 2019). Hegemony and the technological domination of humans therefore remain pressing concerns, and opportunities afforded by digital should be judged with the risks it presents to human flourishing.

This paper attempts to bring theory from classical political philosophy, sociology, and media theory together to analyse one such conceptual risk posed by the digital world: digital despotism. Sabine Pfeiffer describes digital despotism in largely economic terms as the ‘digitally augmented hollowing-out not only of privacy but also of labor rights and the democratic potential of industrial relations’. Building on Michael Burawoy’s Marxist social theory of hegemonic despotism, Pfeiffer argues the digital kind ‘combines bureaucratic control (computers), physical control (machinery, assembly lines) and bodily control (wearables like data glasses and smart textiles to control and track individual behavior)’ and emerges from Industrie 4.0. Digital despotism, so understood, is a globally-structured economic regime of brutal control, silent submission, industrial deregulation, and individual atomization (Pfeiffer, 2017, pp. 36–38).

I argue that we should not leave the idea of digital despotism at this aggregate formulation. Digital despotism could be much enhanced by putting it in a critical conversation with its historic roots in the despotic master–slave relation through themes theorized by Aristotle. As Agamben says, ‘slavery is to ancient humanity what [modern] technology is to modern humanity: both, as bare life, watch over the threshold that allows access to the truly human condition’ (2015, p. 78). With this in mind, we can group Aristotle’s themes along two related dimensions within the context of the Politics’ discussion of masters, slaves, and slavishness in general. Ontological dispossession implies that humans are i) property (ktēma ēi anthropos ōn, Pol. 1.1254a17) and so lose ownership capacity; ii) bodies to use (chrēsis tou sōmatos, Pol. 1.1254b19); and ii) dependent parts of their master (meros ti tou despotou, Pol. 1.1255b12). Cognitive dispossession implies iv) a loss of the deliberative faculty (ouk echei to bouleutikon, Pol. 1.1260a13); and v) a psychological excess or deficiency in their force of character (plērē thumou or athuma, Pol. 7.1327b23–29). The end result of these dimensional dispossessions is a vulnerability to forms of despotic political rule (polis doulōn kai despotōn, Pol. 4. 1295b21–22).

This paper is therefore structured as follows: I will first show how the despotic master–slave relation still sits at the core of contemporary global Marxist historicist, republican, and/or postcolonial accounts that focus on digital slavery and/or domination. Secondly, by engaging common themes in Aristotle’s political philosophy identified above, I argue that digital despotism can work to dispossess humans ontologically as owners into an increasingly disempowered dual category of digital user/used. Likewise and thirdly, I argue that cognitive deliberative and psychological digital conditions, embedded in the design of digital architectures and use of algorithms for example, can bypass or frustrate our decision-making capacity. And finally, I accordingly categorize some of the means of resisting digital despotism, within its overall context of rendering people politically vulnerable to forms of despotic rule like authoritarianism. As we examine what the Politics has to say about masters, slaves, and slavishness, what emerges is not an inevitable, permanent, or simply aggregate condition of structural domination in production and/or consumption, but a dangerous syn-drome – literally a “running together” of things – which operates along ontological and cognitive dimensions.

By conceptualizing digital despotism through Aristotle’s account of the despotic rule of masters over slaves, this thematic approach aims to follow (Aytac, 2022, p. 14) in offering a ‘diagnostic argument identifying problematic power relations’ in the digital world that focuses on a form of (digital) domination at the extreme. In highlighting salient dimensions of the master–slave relation that we ought to avoid, we are better able to categorize thematic risks and potential responses to human autonomy in this world. These dimensions are presented as distinct parts of a single syndrome, and so I offer an integrated approach that stresses the interrelation between various despotic conditions in the digital world that we might otherwise see and want to treat only separately. Moreover, by situating the master–slave relation in Aristotle’s broader political philosophy we gain a theoretical account of how individual-level conditions can translate into political effects on the order of regime type – this is something missing from Pfeiffer’s conception of digital despotism. Relatedly and finally, grounding a concept of digital despotism in classical political philosophy allows us to develop a deeper understanding of our legitimate grievances regarding digital technologies, reminding us why we should care about their deployment, and adding new dimensions to traditional accounts that focus on digital slavery and domination – while avoiding minimization of the coercive, material violence involved in what Forsdyke terms the ‘experiences of slavery’ (2021, p. 102, passim) itself. Much as Aristotle’s ideas on katharsis and/or rhetorikē (and pathos, ethos, logos), for example, have informed and improved our understanding of media, his ideas on slavery and despotism can therefore theoretically ground contemporary concerns about digital technologies and be used to offer a new vantage.

It is nevertheless important at the outset to clarify the methodological limitations involved in this approach. While there has been an exciting recognition in recent scholarship that ‘exploring classical ideas can provide pertinent insights into digital age issues’ (Clements, 2022, p. 769) – see for example Shannon Vallor’s technomoral virtue ethics approach (2016); Charles Ess’ pluralistic global information ethics grounded in Aristotelian pros hen [“towards one”] equivocals (2006); Aristotelian phronēsis and AI (Eisikovits & Feldman, 2021); and Aristotelian friendship and online penpalship (Kristjánsson, 2021) – reading contemporary categories into an ancient text should be done with prudence. The structure of slavery in classical Athens, and the political economy of digital platforms are very different contexts, and thematic analysis can sometimes simplify relations at the expense of complex contextualization. While a degree of presentism is inescapable when interpreting historical texts, the aim of this paper is limited to making despotism a common principle of analysis in the classical Greek master–slave relation, and the contemporary digital world. Ismard’s (2017) work on ancient slavery is instructive here: ‘the goal is not to establish any kind of system into which each society would be sorted, but to implement a principle of specification in order to situate and identify an issue in the context of each of the societies being studied’ (p. 25), such a method is ‘constructivist and reflexive, developing its own categories without turning them into objects of analysis that then define the universal character of a category in themselves’ (p. 26).

Relatedly, the goal of this paper is not to offer a new or coherent reconstruction of Aristotle’s theory of natural slavery in the Politics, which has alternatively been condemned as a ‘battered shipwreck of a theory’ (Garnsey, 1996, p. 107) or an argumentatively coherent ‘defensible defence’ (Simpson, 2006) or ‘defensible piece of Aristotelian philosophy’ that might even be construed as ‘at least potentially a critical theory’ (Schofield, 1990, p. 16; p. 11). Instead of offering definitive answers to the difficult and controversial questions around disentangling Aristotle’s descriptive comments on actual Greek slavery from a philosophy of natural slaves, reading despotism as a syndrome through the Politics’ discussion of masters, slaves, and slavishshness converges on what the vast majority of Aristotle scholars agree anyway: that Aristotle is concerned with ‘the essential character of the master–slave relation’, and with distinguishing political rule from despotic rule (Schofield, 1990, p. 21). Despotism as it appears in Aristotle is therefore the site of analysis in this paper, and not any one single factor that would make his theory of natural slavery causally (in)coherent.

2 (Digital) Despotism, Mastery, and Slavery

Sabine Pfeiffer’s account of digital despotism is not alone in using the semantic field of despotism, mastery, and slavery to describe phenomena in the digital world. Relying on Italian operaismo theory developed in the 1960s, Alessandro Delfanti has argued that digital capitalism relies on an augmented form of despotism and ‘machinic dispossession’ to control labor (Delfanti, 2019). Christian Fuchs examines the ICT supply chain as a whole to suggest that the sector very often rests on slave work, which can therefore be called ‘digital slavery’ (Fuchs, 2014, pp. 155–181). Jack Qiu’s Goodbye iSlave unites both production and consumption dimensions of digital media to argue moreover that the free labor involved in user generated content is a form of ‘manufactured iSlavery’ (Qiu, 2016, p. 106).

Qui accepts a historicist definition of slavery that is ‘often fluid, contingent upon contexts and norms that are subject to change’ (Qiu, 2016, p. 106). Joining Qui are some postcolonial scholars whose general commitment to ‘co-construct[ing] decolonial sociotechnical solutions’ (Cruz, 2021, p. 1847) is often premised on concerns that the digital world risks of ‘new forms of slavery’ (Marwala, 2020, p. 220), ‘invisible and virtual…slave badges, slave collars, slave chains and slave cuffs’ (Nhemachena & Mawere, 2020, p. 133), and ‘a superior form of slavery…the sum of all colonialities, aka summa colonialities (Benyera, 2021, p. 77). Countering this conceptual expansion somewhat, Fuchs notes that we should not too readily apply slavery ‘to every form of exploitation’. The Facebook user, for example, is not a slave – especially when compared with directly coerced Congolese miners in the ICT supply chain. Fuchs nevertheless contends that ‘all labour and class relations have certain dimensions of slavery’ (Fuchs, 2017, p. 692).

On the other hand, Paulin Ismard has noted this trend, particularly amongst Marxist and global historicist approaches, to displace slavery into an overarching category of involuntary, unequal relations of dependence and domination, where ‘its rise is repositioned as central to capitalist modernity’s expansion’ (Ismard, 2017, pp. 16–17). Republican accounts of freedom and domination have also stressed the master–slave relation – Philip Pettit, for example, summarizes ‘[the] republican tradition [as] unanimous in casting freedom as the opposite of slavery, and in seeing exposure to the arbitrary will of another, or living at the mercy of another, as the great evil’ (1997, pp. 32–33). More recently, scholars have extended republican ideas into the digital world with related concepts of ‘digital domination’ (Aytac, 2022; see also Taylor, 2021) and ‘digital freedom’ (Oldenbourg, 2022). There is little doubt that this work has been both important and fruitful. But perhaps there is something yet to gain from going deeper into the historic extremity of the despotic master–slave relation that underlies, yet is not fully co-extensive with contemporary notions of domination.

Indeed, we may generate useful insights into the nature of the digital despotism by examining it with its theoretical forbear and extreme limit in Aristotle’s Politics, which discusses the master (despotēs), slave (doulos), and slavish (ton doulon or andrapodōdēs) types. Aristotle provides one of the richest theorizations of despotism and slavery, as one of the most influential and notorious theorists on this topic in Western thought. Scholars have already noted that Aristotle’s ideas on slavery could be characterized as an ‘extremity’ (Trott, 2013, p. 178), emerging from Greek antiquity as a (mis)appropriated ‘exemplar [to Antebellum pro-slavery writers] of what an “unmitigated” slave society can accomplish’ (Monoson, 2011, p. 272). As such, if we assume that today we still value the heights of human flourishing (eudaimonia) for all that Aristotle’s Politics generally reserved only for free and cultivated Athenian male adults, it is worth inspecting the despotic lower depths of those radically excluded from this life. By doing so, we might ‘reveal some unnoticed aspect, some unexpected angle or concealed property’ and gain the ‘freedom and pleasure of unravelling and reassembling the constituent elements of intellectual operations (Detienne, 2008, xiv–xv).

3 Ontological Dispossession: The New Digital User/Used

How might the digital world affect our capacity to be owners and users of tools? To answer this, we said that it is instructive to evaluate risks in the digital world alongside an extreme limit case: slavery. Ideas on slavery in Aristotle’s Politics give us perhaps the most extreme version of this limit case: humans who are slaves by nature are i) articles of property, ii) bodies used as tools to use other tools, and iii) belongings of others. The Politics gives the essential characteristics of Aristotle’s theoretical natural slaves as follows: a human being who nevertheless does not belong to themselves (mē autou) but rather belongs to another (allou), where belonging to another means being a human article of property (ktēma ēi anthropos ōn). An article of property itself is a tool (organon) for action that is separable from the owner (Pol. 1.1254a15–18.). Slaves are therefore considered tools who do not belong to themselves, suffering from a severe loss of their independent humanity. They inhabit despotic conditions of human dispossession and human impairment, which can serve as warnings for what we ought to guard against reproducing in the digital world. Aristotle’s text suggests that cultivated men who act slavishly and learn the work (erga) of slaves actually void the distinction between master and slave (Pol. 3.1277a33–1277b7). It is certainly not the loss of the capacity to be a master that is or should be relevant for us. But what should concern us is any loss of human autonomy that renders us ontologically similar to the politically vulnerable category indicated by slavishness.

3.1 Loss of Ownership Capacity (ktēma ēi anthropos ōn)

Firstly, Aristotle’s Politics gives us a theoretical account of a slave by nature as one who is human property themselves (ktēma ēi anthropos ōn, Pol. 1.1254a17). This does not mean that the despotic relation theorized by Aristotle necessarily implies that all slaves in the Politics could not own their own property or use tools – all subordinated workers are tools that wield many tools (Pol. 1.1253b34), for example. But this extreme version of slavery to be found in the text assumes humans are articles of property that do not belong to themselves. Aristotle offers this conception to distinguish despotic rule from other types of rule, as ‘seeing the slave instrumentally, as part of the master's property, was necessary to make the distinction between different forms of power’ (Vlassopoulos, 2011, p. 122). As a result, Aristotle’s natural slaves have no real capacity to own their own property, or indeed the tools they used, in any meaningful sense that could be considered free from despotic control. In reality, we know that generally Athenians’ private household (oikos) slaves could not rely on the law to enforce rights to owning property – ‘they were incapable of owning property at all’ (Harrison, 1968, p. 236), although there certainly were exceptions including relatively wealthy public slaves (dēmosioi) owned by the state (In. Tim 1.54). That Athenian law denied protection to the private household slave’s ability to own property therefore coincides on this point of despotic control with the kind of dispossessed human Aristotle wants to theorize in his natural slave.

Exclusion from ownership, particularly as it relates to the tools we use, is an indicator of structural despotism. This exclusion was de jure as well as de facto in the case of Athens. Today we might argue that the global proliferation of intellectual property law regimes over the last century suggests that a similar de jure exclusion operates in the digital world – intellectual property ownership is legally protected, after all. However, de facto exclusion is another story. As Joshua Fairfield has written, the digital world threatens individuals with the ‘death of property’ in its traditional form. Digital property ownership is being built on a despotic ‘feudal model’ that relies on ‘command and control’ and is subject to the ‘iniquities of power distribution in a digital society’ (Fairfield, 2017, p. 13; p. 21). De jure protection of intellectual property can moreover be subverted for de facto exclusion from ownership according to these power iniquities. As Zuboff (2019) has shown, it is not difficult to see that this is partly the result of ‘unprecedented concentrations of knowledge and power’ (p. 180) in the digital world that has yielded incredibly powerful corporations like Amazon, Microsoft, Meta, Alphabet, and Apple.

But another, interrelated reason for these iniquities is the ongoing accretion of technological dependencies that has come to determine the development of new technology and the use of tools. The case of an individual worker – whether Ancient Athenian or contemporary American – owning a hammer is far removed, in terms of successive ownership structures, from a contemporary digital user who ‘must rely on the owners above them’ (Fairfield, 2017, p. 21). These users are reliant on a conglomerated matrix of energy, hardware, network infrastructure, software, application, and content proprietors to be able to exercise any residual form of ownership that might be left to them. Thus, while infrastructural dependency is by no means unique to the digital age, increased layers of technical mediation dilute ownership to a greater extent. An app to control a toaster requires many more levels of technical dependency than an analogue toaster does. Simply put, we are much more likely to own our hand tools than our digital tools.

3.2 Use of the Body (chrēsis tou sōmatos)

The generalized loss of individual ownership associated with the rise of the digital world is already contained in the everyday designation of digital user. While we may not enjoy full ownership over our digital experience, whether work or consumption, we nevertheless retain some level of ability to use these tools. Use (chrēsis) is still an important indicator of control over something else. As Agamben notes in his profound treatment of this subject, ‘The expression “the use of the body” (he tou somatos chresis) is found at the beginning of Aristotle’s Politics (1254b18), at the point where it is a question of defining the nature of the slave’ (2015, p. 3). Aristotle goes on to state that the position of the ruled is akin to a flute-maker, while the position of the ruler is akin to the flute-player who uses (chrōmenos) what the former makes (Pol. 3.1277b28–30). The scope of use matters. In the Politics, this scope differs from what is afforded to free, cultivated Athenians, because slaves were also, in turn, actually used by the former.

Secondly, then, we see that the Politics suggests that slaves differ from free, cultivated Athenians in the scope of their user experience. Aristotle states that the slave’s function (ergon) is the use (chrēsis) of their body (tou sōmatos) primarily (Pol. 1.1254b18–20), though Athenian slaves did in practice occupy a range of complex roles requiring intellectual effort (Ismard, 2015, pp. 63–94). Conversely, the Politics repeatedly stresses that masters and free, cultivated Athenians should only learn how to use (chrēsthai) slaves and subordinates. They should not make it their focus to learn the content of slaves’ menial (diakonikas) tasks, like cooking and handiwork, that slaves in turn use tools to perform. Otherwise, they risk becoming slavish themselves. Aristotle even suggests this as the reason why cultivated Athenians use an intermediary – a steward (epitropos), who takes on the direct task of managing slaves so the former can engage in politics and philosophy (Pol. 1.1255b21–38; 3.1277a33–1277b7).

What is crucial in these passages above is the Politics’ reference to knowledge (epistēmē, epistasthai) that cultivated Athenians ought not to learn in their education. It is skill and education that determines the restrictions on use – slaves in the Politics should only be educated to the extent that they can be bodies to use as tools for the use of other tools, while Athenian freemen can use other people to take care of necessary affairs and cultivate themselves. Though retaining the ability to use something certainly implies some level of control over it, skill and education therefore determine the possibilities of use open to the user. This does not mean that Athenian freeman need to be skilled themselves in everything they require to live well. Having some idea about what the work and skills of others entails, so that one can use them to perform these tasks, is all that is sufficient for the Athenian freeman when he takes on the role of master: the master must know how to command the things that the slave must know how to do (Pol 1.1255b34–36).

The kind of household management Aristotle’s Politics contemplates in this interpretation is no less relevant to the use of digital tools. Aristotle separates two very different kinds of user according to their education and tasks: a higher skill use of others, and a lower skill use of tools. Higher skill users are able to shape the direction of tasks and their outputs. Lower skill users are circumscribed to performing according to the function of the tool and what is ultimately required of them. Just like their classical counterparts, digital users are more likely to have a high degree of control over their use of tools, even if they don’t own them, when they have acquired superior skills. A small minority of software engineers, for example, are not only able to interact with a given online platform at strategic levels, but can also actively create and manage their own digital tools of engagement. In contrast, someone who has only learnt to browse the internet or read an email is far more circumscribed in how they can use digital tools. The skills and usage gap, a core component of what has been called and popularized as ‘the digital divide’ (Van Dijk, 2020), defines the contours of what the digital world means for a majority of digital users when ownership has been even further hollowed out by greater economic consolidation and accreted infrastructural dependencies, and the scope of usership is unequally distributed. When this happens, users are at risk of becoming a class of digital used whose experience with these technologies is ultimately precarious and dependent on forces belonging outside of their control.

3.3 Part of the Owner (meros ti tou despotou)

Finally, then, we should ask what it means not to have an independent sense of belonging. Not only does the Politics suggest that slaves are property, but they are also unable to belong to themselves. They are ‘of another’ (allou, Pol. 1.1254a15–16). Belonging to someone else, as a possession, obviously indicates a despotic relationship. But the Politics goes further than this – natural slaves are in fact an actual part of their masters (meros ti tou despotou), an animate part of the body (meros ti tou sōmatos) that is separated (Pol. 1.1255b12–13). It’s difficult to fathom a state of being more completely dispossessed of human independence than this. To be a part of the master’s body is to function as an organ as much as a tool, which is why it is telling that the Greek term for both is organon. Both master and slave are locked in a complex dyadic relation of dependency (Pol. 1.11252a25–1252b1) which almost entirely favors the interests of the former.

Belonging to another means that one has lost an important element of independence and control. This extreme version of slavery contained in the Politics sees the slave as a wholly dependent body part of another. Such an idea seems alien to our contemporary world notwithstanding the existence of modern slavery and coercive forms of labor and servitude. But the example of the slave with a non-interfering master has been a central point of contention between contemporary liberal and republican accounts of freedom, as well as those who have attempted to find a middle-ground in freedom as independence (List & Valentini, 2016). Independence is still an enormous challenge in a digital world where humans have the ‘technical ability to turn our behavior and social activities into data points that can be collected and analysed’ (Hintz, 2020, p. 536), and where our presence and actions are increasingly dependent on the will of those who best control platforms and their underlying infrastructures. For example, Aytac notes that the ‘digital public sphere is one indispensable mechanism in which citizens participate in public life and engage in contestatory practices…[but] [i]nstead of being unconditional and independent, the existence and functioning of these channels of influence still depend on the goodwill of corporate executive boards’ (2022, p. 10). A caveat worth mentioning at this point is that ‘digital vs. analogue’ or even online vs. offline can be understood to be continuous ‘modes of presentation’ to epistemic agents (Floridi, 2008, p. 151). Nevertheless, if digital expressions of who we are, our digital selves, form exploitable parts of a system that can be bought, sold, monitored, edited, and extinguished, this digital self does not ultimately belong to us in a meaningful way and is of another (allou).

The risk of a significant loss of independence in the digital world, approaching the complete loss faced by the category of human that Aristotle’s Politics describes as natural slaves, is especially a cause for concern because of how the former directly undermines independence in our offline existence. Data points, or parts, of our digital selves such as inter alia location, health, preferences, relationships, tell real information about our offline lives. This subordination in the digital world can therefore flow through to our offline existence as well. As Bernard Stiegler (2016, p. 194) has suggested, ‘automated forms of social control’ have the tendency of disaggregating and ‘dis-integrating’ us, so that the “I” in individual is lost, leaving us mere ‘ndividuals’. Despotism in the digital world can therefore operate along a sinister dimension of dependency that is often acutely subtle and diffuse. In sum, a loss of ownership in the digital world can therefore be accompanied by a dispossession of human independence.

When humans have lost the ownership of their digital tools, face increasingly circumscribed use of what is left, and are dispossessed of their basic independence, we find a worrying structural approximation to the conditions of abject slavishness. In this situation, digital users become tools in themselves and can be used as such by a variety of different groups for a variety of different ends. Of course, the despotic position that people can be tools, which the Politics takes as regards not only slaves, but all subordinate craft workers in general, is by no means unique or exclusive to him or his period. That workers in particular can become nothing more than used tools or organs is also observed by Marx regarding capitalistic production processes (1867 [1976], p. 414 [548]), where factory work goes one step further by subordinating the worker to machines so that ‘the machine makes use of him’ instead of the other way around. Beyond the factory and into the digital world we see that individuals are no less at risk of the same dispossession.

4 Cognitive Dispossession: Deliberative and Psychological

Natural slaves in the Politics are not only essentially dispossessed as independent humans, but they also face cognitive restrictions and impairments that cultivated – and morally excellent – Athenian freemen do not. In this way the freedom of inquiry and decision-making is intimately tied to the possibilities of social and political freedom. Natural slaves are said i) not to wholly have the ability to deliberate (holōs ouk echei to bouleutikon) and make decisions (Pol. 1.1260a13), and slavish types are moreover said to ii) suffer from either an excess or deficiency in their force of character ((plērē thumou or athuma) or spiritedness (Pol. 7.1327b23–29). In short, deliberative and psychological factors also frustrate the ability of slavish types to act independently according to practical reason. By applying these themes of cognitive dispossession to the digital world, we can identify areas of risk in its design, function, and experience.

4.1 Loss of Deliberative Faculty (ouk echei to bouleutikon)

Firstly, the text of the Politics claims that slaves by nature are unable to properly deliberate. This is a serious incapacity for a number of reasons. For one, this incapacity strikes at moral excellence, since it means that these types are unable to make choices (prohairesis) about practical means to ends. According to Aristotle, choice is a deliberate desire (bouleutikē orexis) of things within our power (Eth. Nic. 3.1113a10–13). It is a prerequisite for moral excellence, since the latter is itself an active disposition (hexis) for making the right choices (prohairetikē) at a relative mean (Eth. Nic. 2.1106b36–1107a1). Without a capacity to deliberate, a natural slave may possibly identify practical goals but cannot figure out the conduct required to achieve them. Actions that might be deemed virtuous or good, like taking action in the heat of the moment to save someone from harm, are for these slaves less a function of deliberate cognition than passionate happenstance. As a result, Aristotle’s virtue of phronēsis, a form of ‘reflective judgement’ that is by definition capable of self-correction (Ess, 2020, p. 565) and a ‘metacognitive capacity to make complex moral decisions’ (Kristjánsson, 2022, p. 173), is made virtually impossible. This frustrates an ongoing recognition that phronēsis is critical for civics in a digital age (Clements, 2023).

While scholars have tended to focus on phronēsis and moral excellence, the slave’s more basic incapacity to make deliberative decisions (ouk echei to bouleutikon) can have equally severe consequences for political life. To lack proper deliberative capacity also means an inability to come to the decisions required for governing the political community. While ‘prohairetic activity…distinguishes slaves from nonslaves but secures no absolute boundaries and offers no permanent foundation’ (Frank, 2005, pp. 37–38), it is partly through this permeable boundary of decision-making that the Politics disenfranchises slaves from citizenship and political life. We know that slaves in Athens, still ‘central to the apparatus of the state’, were largely excluded from political participation and ‘did not have political or legal rights’ even though the Greeks’ ‘strenuous efforts’ to enforce a clear distinction between slave and free ‘was constantly destabilized’ (Forsdyke, 2021, p. 198). Public slaves often carried out the administrative tasks of the state, for example (Ismard, 2015, passim). Coming to decisions about how to run the state coincides in importance for both theory in the Politics and actual practice: the language of bouleutikon is actually contained the Athenian political institution, the boulē, which served as a council for governance of the daily affairs of the city.

Undermining our capacity to deliberate and make decisions, especially about our own moral and political lives, is therefore a dangerous dimension of cognitive dispossession. This can occur both overtly and covertly through the design and implementation of choice architectures in the digital world. These choice architectures exist because we do recognize that very often people do not make optimal decisions that either maximize their own payoffs, the payoffs to society at large, or ideally both. Sometimes these architectures rely on more or less overt “nudging” in the right direction. Nudging decision-making towards optimal outcomes through a choice architecture that does not inherently restrict options seems like an inoffensive and moreover useful exercise of governance. But problem arises when we take it too far – instead of a choice architecture that embraces a kind of libertarian paternalism, as behavioralists have termed it (Thaler & Sunstein, 2008), we find architectures that mirror the kind of despotic paternalist relation of the master towards the slave in the Politics.

One obvious way, therefore, that digital world choice architectures can be used to dispossess people of their ability to deliberate and make decisions is if they are overtly designed to do so. Architectures like these are not really about choice at all, but rather embody values (Klenk, 2021) inducing behavioral command and control, often under the guise of reducing human error and/or increasing productive efficiency. For example, using warehouse algorithms to dictate the minutiae of how workers should do their jobs smacks of an exploitative relationship that serves a private firm’s interests over workers, while also undermining human deliberation and decision-making. In algorithms such as those deployed by Amazon, human decision-making is precisely frustrated by the centralization of knowledge about the overall workflow and order process outside of individual workers: ‘this form of algorithmic management is based upon the standardization and taskification of the labor process made possible by this incorporation of knowledge in the machine. In turn, this allows Amazon to strictly control masses of workers who can quickly be put to work in the warehouse’ (Delfanti, 2019, p. 47). Ifeoma Ajunwa (2023, p. 3) states the problem most acutely: ‘these new technologies perform automated decision-making with machine learning algorithms, often ignoring the gestalt of the worker in favor of numbers on the screen’.

But architectures can also covertly result in an effective cognitive dispossession of deliberative decision-making. Choice (or lack thereof) architecture is itself determined by the space of operation. If the space to manoeuvre is restricted, there is only so much that users can do. Ajunwa (2023, p. 98) points to the example of automated hiring platforms as an example of “platform authoritarianism” where ‘a lack of choice as to whether to use a platform is coupled with a demand to engage with that platform’ and is ‘effectuated by design choices for hiring platform user interfaces that constrain user choice regarding what information must be entered into the system’. Aytac (2022, p. 11) explicitly argues that ‘social media companies’ algorithmic control amounts to quasi-public domination because it arbitrarily interferes with citizens’ choices about modes of discursive engagement’. In short, ‘a small group of corporate actors have quasi-monopolistic control over the terms and conditions of citizens’ online speech and contestatory practices’ (p. 10).

Similarly, while the digital world contains shiny new possibilities and experiential opportunities over which to exercise decision-making, the interactive sum total of inter alia proprietary systems, security protocol, and predictive and persuasive algorithms profoundly delimit both the choices we can make, and our independence in doing so. Danaher (2019, p. 119) has called these latter concerns possible violations of the ‘optionality and independence’ conditions of autonomy, though separates out and discounts their effect on (practical) rationality. To frame the problem more acutely – it is not simply an issue that we might yield decision-making to algorithms that allow us to simplify our choices when deciding what to watch on Netflix. It is also a failure to recognize that it is the same operative digital world and deliberative capacity at risk which we also employ for hugely consequential decisions in human life, like voting. As Oldenbourg puts it, social media companies wield immense power because they ‘are able to manipulate the autonomous decision-making processes of their users’ (2022, p. 1; see also Sahebi and Formosa, 2022, p. 70). Stiegler (2016, p. 194) is at once more general yet even more explicit: ‘[d]igital automata have succeeded in bypassing the deliberative functions of the mind’.

4.2 Imbalanced Force of Character (plērē thumou or athuma)

Secondly, we saw that another element of making choices, aside from deliberation, is affective in nature. As Marcus puts it, ‘[affective] state of mind, contemporary mood, may mediate how judgments are made’ (2000, p. 232). It bears repeating that for Aristotle, choice is a deliberate desire (bouleutikē orexis) of things within our power. Orexis, or desire, speaks to this affective element within us that should be directed by reason. If our capacity to deliberate is undermined, so is our ability to make good choices. But Aristotle also identifies another type of desire in addition to a deliberative kind aimed at a plurality of goods (boulēsis) and one based on pleasure (epithumia), namely a desire that has to do with our force of character or spiritedness: thumos (De an. 3.432b4–7). Giles Pearson (2012, p. 139; p. 131) draws on Aristotle’s comments (especially in Eth. Nic. 4.1125b–11256a) to advance a persuasive, if controversial, account of thumos in Aristotle as a ‘retaliatory desire’ that makes it virtually synonymous with anger (orgē): ‘a desire for revenge in response to a perceived slight’ – although admits that in certain ‘atypical’ cases such as the Politics a broader meaning seems apparent. For our purposes what is important is that Aristotle associates both an excess and deficiency of thumos with a lack of autonomy.

On one hand, the text of the Politics claims that Asians (i.e., Persians) and some Greek ethnic groups (relative to each other) have souls endowed with thought (dianoētika) and technical skill (technika) but are athuma and so remain enslaved (douleuonta) and ruled (Pol. 7.1327b26–37). Without a strong force of character, or a desire to retaliate to slights, the implication here is that these groups are unable to muster psychological resistance against despotism. Jill Frank (2005, p. 31) describes this situation as follows, ‘Aristotle instead calls Asians natural slaves on the basis of what he sees as their lethargy [my emphasis], which is to say, their apparent tendency to forget how to act on their own initiative, or “inactivity”’. Though the Politics specifies these groups as relatively intelligent and technologically sophisticated, this is deemed insufficient to guarantee the political freedom possible for cultivated Athenians. Leaving aside the nevertheless important question of whether Aristotle held protoracist and ethnicist views (Lockwood, 2021), a useful insight here is that in the Politics, technological sophistication and intelligence do not automatically translate into freedom from despotism – some force of character or retaliatory desire should also be present in people.

In fact, a modern concern is that technological sophistication and systems can actually dissipate our psychological force of character. In the industrial factory, Marx’s comments on capital and automated machinery reference man’s natural barrier and resistance (Widerstand) that must be overcome for capitalist productivity, going so far as to suggest that the turn towards exploiting the labor of women and children is in part because of their more docile (fügsamere) nature (1867 [1976], p. 392 [527]). The idea of technology dissipating or wearing down human psychological resistance is by no means confined to industrial work, however. Pointing to Amazon’s $1bn bet on the videogame streaming platform, Twitch, Tung-Hui Hu (2022, p. 43; vii–viii) notes that this is ‘a sign that so-called passive users are simply the next marketplace for digital capitalism’s expansion’. Hu’s novel theorization of ‘digital lethargy’ as a state of ‘of being passive, or wanting to disassociate and be anyone but yourself, or avoiding decisions [my emphasis]’ is one that actually forecloses the respite offered by true passivity – digital users are ‘always “on” as far as technology is concerned, even if you think you’ve logged off’. In short, Hu’s work supports the idea that digital world creates mountain ranges of added complexity and a raft of new hierarchical relations that can erode and dispossess our psychological force of character.

While Hu’s picture of digital lethargy actually attempts to transvalue this form of passivity through ambivalence (we come to ‘[endure] a condition rather than [refuse] it…[sidestepping] the liberal-democratic narrative in which a subject “finds one’s voice” and begins to participate in political change’ [xxvii]), it nevertheless recognizes passivity and lethargy as problems in the history of Western political thought. But if we consider a lack of thumos as a problem, there also lies a danger at the opposite extreme of thumos which is usually more acutely visible to us, especially on social media platforms. The Politics offers a relative characterization of Europeans and certain Greek ethnic groups as full of spiritedness or force of character (plērē thumou) but deficient in intelligence (dianoias) and technical skill (technēs). As a result, they lack political governance (apoliteuta) though they remain somewhat free (Pol. 7.1327b23–26). It would be a mistake, however, to think that this freedom is equivalent to the freedom that is secured through good political governance. Again, what is important for our purposes here is that the Politics diagnoses an acute problem with excessive spiritedness or force of character (thumos).

Too much force of character or spiritedness, especially in the inflection of retaliatory anger that Pearson (2012) suggests, can be deleterious for practical reason and general cognition. There is no shortage of literature and empirical data on the role that anger, aggression, and hatred play in the operation and success of social media platforms. Siegel (2020, p. 63), for example, relies on social media hate speech research by ElSherief et al. (2018) on Twitter to suggest that ‘accounts that instigate hate speech tend to be new, very active, and express lower emotional awareness and higher anger and immoderation in the content of their tweets’. Ganesh (2020) explicitly uses the language of ‘white thumos’ in developing a framework to analyse the racialized flow of radical right sentiment across social media. The kind of unbridled excessive thumos associated with anger, aggression, and hatred is precisely the type that clouds judgement and can lead to increased political polarization and instability. For Aristotle, it runs the risk that humans degrade themselves into preferring the life of animals, a life of utter slavishness (pantelōs andrapodōdeis, Eth. Nic. 1.1095b14–22) subservient to the passions and not reason – a situation from which the Politics saves the natural slave while nevertheless reiterating a functional closeness (Pol. 1.1254b20–26).

To conclude, if we find ourselves unable to, or impaired in our capacity to deliberate and make decisions on one hand, while having an excess or deficiency with respect to our force of character or spiritedness on the other, we find ourselves at risk of the cognitive dispossession Aristotle’s Politics once associated with slavishness. Of course, these two components of cognition are interrelated since a failure to deliberate can trigger affective consequences and vice versa. It would therefore appear relatively easy to find ourselves overwhelmed, given how much needs to be balanced. In discussing Aristotle’s moral psychology, Giulia Sissa (2018) notes the careful balance of conditions that must occur so that we can behave decisively like ‘a Homeric king, a political leader, and a military commander’ (p. 167 read with Eth. Nic. 3.1113a3–10). The problem comes in, however, not when we find ourselves sporadically unbalanced, but when this imbalance becomes a persistent feature of our lives: we become ‘habituated to slavish behavior’ (Frank, 2005, p. 32). From restrictive digital choice architectures to decision-making algorithms, and from digital lethargy and passivity to social media anger, aggression, hate and polarization, we can therefore all too easily find ourselves in this habituated position where our general ability to make (good) decisions – including moral and/or political ones – is compromised by the digital technological milieu we inhabit.

5 Digital Despotism: Addressing Vulnerability to Authoritarianism

What is the social and political significance, then, of finding ourselves ontologically and cognitively dispossessed? One of Aristotle’s major political concerns with a people becoming too slavish (lian andrapodōdes) is because, inter alia, it means their ability to judge as a collective is severely compromised (Pol. 3.1282a15–16). Moreover, more slavish types (doulikōteroi) will end up enduring despotic rule (despotikēn archēn) because they cannot foment the requisite resentment (duscherainontes) that can serve as the basis for revolution (metabolē) against despotism (Pol. 3.1285a18–23 read with 5.1306b3–6). Despotic rule, based on the rule of master over slave, is an extractive one that only ever serves the interests of the ruled accidentally. So too with tyranny, which the Politics analogizes and compares closely with the relation of master over slave (Eth. Nic. 8.1160b7–9; 1161a31–1161b11). The end result of tyranny is a city of masters and slaves (polis doulōn kai despotōn) for a people (Pol. 4.1295b21–25).

If we accept the idea that the digital world risks a syndrome of human dispossession along ontological and cognitive dimensions, then the political implications of this dispossession should concern us as well – namely a modern, digitally-enabled vulnerability to tyrannical or authoritarian control. In fact, the starting point has usually been the observed political outcome rather than the risks and conditions of possibility. ‘Networked authoritarianism’ (Mackinnon, 2011, p. 33) is described as the result of an already ‘authoritarian regime embrac[ing] and adjust[ing] to the inevitable changes brought by digital communications’. More recent focus has been on an overarching term ‘digital authoritarianism’ which refers to ‘the use of digital information technology by authoritarian regimes to surveil, repress, and manipulate domestic and foreign populations’ (Polyakova & Meserole, 2019, p. 1). What these descriptions suggest is that digitally authoritarian political outcomes are top-down and determined by established authoritarian regimes rather than technological conditions of possibility and syndromic risk factors that may affect any regime type. Digital despotism suggests instead that states and political actors across the board can exploit a common set of digital risk factors to generate more or less tyrannical or authoritarian outcomes Table 1. Even in democratic states, the presence of uncontrolled digital platforms can thwart ‘citizens’ individualized, unconditional, and efficacious influence on the political processes [as] a necessary condition for popular democratic control’ (Aytac, 2022, p. 13).

Table 1 Thematic Framework for Digital Despotism 
Full size table

It is helpful, then, to summarize what this syndrome looks like so far, in terms of the thematic framework including corresponding digital risks:

The syndromic view of digital despotism has the benefit of integrating various and often otherwise isolated concerns scholars and practitioners have had about the effects of digital technologies within a classical theory of politics.

Another implication of seeing digital despotism as a syndrome is that we are better able to recognize that interventions to address it should not be isolated. Instead, interventions should involve a multipronged targeting of the components that risk a vulnerability to tyrannical or authoritarian control. Why? Let us take, as an example, the risk of ownership capacity loss. To mitigate against ontological dispossession and the transformation of humans into a large class of disempowered digital users, framing digital despotism thusly suggests that we might consider ways to reclaim ownership of digital tools and platforms. This can be done in a collective manner by increasing public ownership and accountability of digital technology companies. However, as Susskind (2018) points out, the last century has taught us that ‘sweeping state ownership can lapse into authoritarian dictatorship’ (p. 329).

Other possible interventions, therefore, might stress the ability of private individuals and/or non-state actor communities to have a more direct level of ownership and control over digital technologies. Ideas gaining increased traction recently include cooperative community ownership, which involves ‘shared ownership and governance among people who depend on an enterprise, shared profits, and coordination among enterprises rather than competition’ (Schneider, 2016, pp. 15–16); and predistribution, which involves ‘a more equal distribution of economic power and rewards even before government collects taxes or pays out benefits’ (Hacker, 2011, p. 35) and can be used to allocate fractional shares, dividends, and corporate voting rights in technologies and technology companies prior to government taxation. Importantly, these interventions are not necessarily mutually exclusive with increased public, collective ownership of digital technology, but rather enhance it along the individual level. And yet, ownership without digital skills development will still leave people vulnerable because it is ‘those who do have access, capacities, skills, and resources that allow them to take ownership’ (Postmes, 2007, p. 173). Addressing a loss of ownership, therefore, also means addressing skills inequalities.

If independence has been eroded, we also come to a problem with the free ability of private citizens to exercise direct ownership. If, as we identified earlier, individuals have been digitally “disintegrated” to such an extent that they form exploitable parts that can be manipulated, then this loss of independence jeopardizes both skills development and ownership. Likewise, if the elements of cognitive dispossession are not also addressed, then any one of them might begin to weigh heavily enough on individual autonomy that they create the conditions for the emergence of other digital syndromic risk factors. A multipronged approach to addressing digital despotism recognizes that it is not enough to address one risk factor alone, and that political vulnerability to authoritarianism locates itself across multiple areas of risk in human life. In concrete terms, therefore, interventions aimed at ensuring forms of digital ownership must go hand-in-hand with increasing skills (Clements, 2020, p. 576), enhancing digital independence (Fairfield, 2017, p. 78), stimulating deliberative decision-making (Clements, 2020, p. 581; Ess, 2009, passim), and encouraging psychological balance or ‘affective intelligence’ (Marcus et al., 2000, p. 2) as well as psychological integrity to avoid digital technologies’ destabilizing effects (Burr & Floridi, 2020, p. 15) – all with respect to the digital world.

The previous discussion assumes, however, that addressing digital despotism is the preserve of intentional technocratic policymaking. This neglects the sphere of real individual and collective resistance to despotism. Scholars have noted that slaves in Athens could offer resistance through various forms of self-interested cooperation, passive acquiescence and marginal non-cooperation, and/or active resistance (Garnsey, 1996, p. 9). For example, while Athenian slaves could not legally own their own property in general (with the exception of public slaves), they could and did address their dispossession by sometimes seizing their master’s possessions. The Greek orator and politician, Demosthenes, relates the case of the slave Moschion, who was found to be stealing money. from his master without the latter’s knowledge, keeping it for himself (di’heautou) until he was apparently found out (Dem. 48. 12–21). Moschion’s story is an unhappy one – the money was repossessed, and he was tortured twice. He may have even been his master’s ‘slave manager and business agent’ himself (Forsdyke, 2021, p. 112). But failure and the undoubted complexity of slave relations does not negate acts of resistance: ‘from the perspective of slaves, such acts were not theft but justified reappropriation of the fruits of their own labor’ as an ‘important technique of covert resistance’ (Forsdyke, 2021, p. 205).

What might a digital corollary to the stealing slave look like? The obvious answer here is piracy as a form of user resistance. If we continue this line of thinking for the other risk factors that we have identified then our framework offers, additionally, a way to categorize familiar resistances that have emerged through media theory and praxis. Firstly, in becoming digital users/used people can and do counter ownership loss with piracy (Strangelove, 2005) and/or the development of free & open-source tools (Coleman, 2004). Secondly, in being restricted in our digital experience according to a widening skills inequality, countering with self- and community-led digital (re)skilling (Mossberger, Tolbert, & LaCombe, 2021) is an option. Thirdly, to a loss of independence it is possible to strengthen collective and individual data sovereignty efforts like citizen lobbying for increased oversight and/or deliberately providing obfuscatory data about ourselves to platforms (Kitchin & Fraser, 2020; Kitchin, 2021, pp. 224–226). Fourthly, to a loss of deliberative decision-making brought about by choice architectures and algorithms there remains options of resisting more or less intensely through hacking/hacktivism (Jordan & Taylor, 2004) or repair: incremental and internal algorithmic corrections (Velkova & Kaun, 2019). Finally, to an imbalanced force of character associated with problems of aggression or passivity people alternatively look to digital detoxes (Syvertsen, 2022) or (social) media activism (Meikle, 2002), for example. These forms of resistance are by no means exhaustive, or exclusive to one dimension of dispossession. And while any one or more of these forms might invariably cross a threshold of social, legal, and/or political acceptability – and it is not the aim here to engage in a normative evaluation on this basis – it is their possibility that destabilizes the formation of digital despotism.

6 Conclusion

I have suggested that classical political and technological thought still has something to say about contemporary concerns in the digital world, and have offered a thematic framework of despotism through which the former can inform the latter. I suggest that it is through one of the most infamous theorists of slavery – Aristotle– that we might better understand and hopefully avoid a syndromic loss of our humanity by the modern digital world. Indeed, recent thinkers in global Marxist historicist, republican, and/or postcolonial traditions have precisely turned to the language of slavery when describing real and/or potential conditions of human existence brought about by digital technology.

Digital despotism, already theorized at the aggregate level as an economic and technological regime of exploitation that has largely been applied to industrial relations, captures something at work in the appeal to the language of slavery while trying to avoid minimization of the material, coercive violence involved in the experiences of slavery itself. By supplementing this aggregate economic conception of digital despotism with one that a emphasizes dimensions of individual i) ontological, and ii) cognitive dispossession, we can better conceptualize the thematic risk factors that create the technological conditions for contemporary vulnerabilities to authoritarianism and tyranny. By reflecting on the extremity of Aristotle’s ideas about slavery, we can see how we might risk ontological dispossession in the digital age when we find ourselves converted into the dual category of digital users/used; we are grossly unequal with respect to digital skills; and we lack an independent digital sense of self. We risk cognitive dispossession when choice architectures and algorithms increasingly restrict our ability to make deliberative choices; and digital technologies and platforms affect us to the extent that we find ourselves habitually in excess or deficiency with respect to our thumotic force of character – either too passive, or too aggressive. Consequently, our ability to make decisions and form good judgments is compromised: ‘future humans will not continue to hone their practical judgment and phronesis, practicing the foundations of their agency’ and instead slip into, ‘moral patiency’ (Eisikovits & Feldman, 2021, p. 196) and political vulnerability.

By seeing digital despotism as a syndrome of human dispossession we are able to recognize risk factors that are clearly identifiable, nevertheless interrelated, and which require a multipronged intervention. Doing so also broadens and deepens the scope of existing accounts of domination, slavery, and despotism as applied to the digital world –including Marxist economic, neo-republican, and/or postcolonial discourses – by providing bridges between different areas of focus that nevertheless stem from a longer historical, common concern with despotic control in human life. Moreover, we are better able to situate and categorize individual and collective forms of contemporary resistance such as inter alia piracy, free & open-source toolmaking, self- and community-led digital (re)skilling, citizen lobbying and the use of obfuscatory techniques, hacking/hacktivism, algorithmic repair, and digital detox or (social) media activism, as responses to the dispossessory tendency of digital technologies. At bottom, grounding digital despotism in Aristotle’s philosophy warns us that if we do not guard our individual autonomy in the face of new digital technologies, we face the genuine prospect of despotic political governance.