Undergraduate Finalist paper in the 2025 National Uehiro Oxford Essay Prize in Practical Ethics. By Rahul Lakhanpaul, University of Edinburgh.
In the last few decades, social media technologies have experienced astronomical successes with many billions of users regularly engaging with social media platforms daily (DataReportal, 2024). Much research has been done focusing on the mechanisms by which social media platforms foster the proliferation of misinformation and fake news (Nguyen, 2020). The epistemic threats social media platforms pose are well-known (Nguyen, 2020). In this essay I wish to shed light on something I think has been overlooked, the moral implications of these epistemic threats. In particular, I wish to present considerations to the effect that social media poses a significant threat to our autonomy, our basic capacity for self-government.
I dedicate the first section of this essay to exploring the concept of autonomy, outlining its ethical significance, and formulating and defending a necessary condition of autonomy. Loosely speaking, I believe we are autonomous only if we are adequately well-informed about our actions. In the second part of the essay, I focus on producing a sketch of one of the ways by which social media prevents us from fulfilling this condition necessary for autonomy. I argue that social media increases our moral influence upon the world and that with this increased influence comes greater moral complexity of our actions. This moral complexity necessitates that we possess more information to be adequately well-informed about our actions. However, as I show, social media deliberately prevents us from accessing this information and, in this way, poses a threat to our autonomy.
Autonomy is the basic capacity we possess for self-government (Feinberg, 1989). Although a complete and rigorous defence of the moral significance of autonomy lies beyond the scope of this essay, I wish to begin by briefly outlining some motivations for considering autonomy a worthy subject of ethical investigation. Most contemporary normative ethical theories can be approximately classified into two camps (Gryz, 2011): deontic theories e.g., consequentialism and deontological ethics, and aretaic theories e.g., virtue ethics. Deontic theories focus primarily on what we ought to do whereas aretaic theories are primarily concerned with the kind of person we ought to be. All deontic theories operate on the presupposition that agents possess the basic capacity for self-government.
Indeed, it is nonsensical to speak of prescriptions, rules, commands and the like unless one is committed to the belief that people have the basic capacity to police their actions. This is precisely why we do not speak of infants or the insane as having obligations. Similarly, for aretaic theories to talk of the sort of person one ought to be, one must first be committed to the idea that people can skilfully regulate their actions and, in that way, cultivate their character. To be a moral agent, then, requires that one possess the basic capacity for self-government. Autonomy lies at the root of ethics. I believe for this reason that autonomy deserves our attention.
I now turn to my first task in this essay of investigating the concept of autonomy as the basic capacity for self-government. I will not provide a complete analysis of the concept of autonomy here. It suffices for my present purpose of illuminating the threat social media poses to autonomy to state and defend a single necessary condition of autonomy which I will later show social media renders unattainable. I contend that:
Authorial condition (AC): An individual is autonomous only if when acting they can appropriately consider themselves an effective agent.
I shall show the necessity of this condition follows directly from the concept of self-government. If an agent fails to satisfy AC, then they must lack the capacity to understand themselves as effective agents i.e., as the authors of their actions. Lacking the capacity to understand oneself as the author of an action precludes oneself from governing over that action. If an individual believes a particular action to stem from something beyond themselves, they cannot coherently consider that action to lie within the jurisdiction of their self-government. Thus, autonomy is subverted whenever the authorial condition is not satisfied.
I have not yet satisfactorily defined the authorial condition.I must now explain when one can appropriately consider themselves an effective agent.Recall the motivations for considering autonomy a crucial ethical concept. Without the concept of autonomy, one cannot speak sensibly of agents either acting in accordance with moral precepts or acting to cultivate their character. To talk sensibly of such things, an agent must not only understand themselves to be the author of actions, but they must additionally recognise the moral character of those actions. If an agent is ignorant of morally salient factors when acting then it is not appropriate to expect them to govern that action in accordance with a system of ethics, just as one would not expect the police to interfere with, or promote, actions they judge not to pertain to lawful conduct. This elucidates what I mean for one to consider themselves an effective agent in a manner appropriate for possessing autonomy. One appropriately considers themselves an effective agent only if they are sufficiently sensitive to the morally salient features of their actions. The earlier condition can now be equivalently restated like so:
Authorial condition* (AC*): An individual is autonomous only if when acting they can consider themselves an effective agent of their actions and are sufficiently sensitive to the morally salient features of those actions.
Having developed this necessary condition of autonomy, I now turn to the second half of this essay in which I hope to sketch an outline of a procedure by which social media prevents its users from satisfying this condition and thus undermines our autonomy.
I shall now show that by impairing their sensitivity to the morally salient features of their actions social media prevents its users from fulfilling AC*.Many factors contribute to this impairment, I will focus only on one here – flawed informational networks. When an individual is deprived of information this does not by itself constitute a threat to autonomy even though it might represent an epistemic issue. As AC* shows, for a flawed informational network to threaten an individual’s autonomy it must deprive them of, or deceive them about, information relevant to moral aspects of their actions. So, before I show these flawed informational networks exist, I will first explain why the information they might provide is morally relevant to our actions.
The introduction of all new technologies has an empowering effect on individuals whose consequent augmentation is inevitably attended by a proportionate expansion of their moral spheres of influence and the moral complexity of their actions (Kiran, 2015). By fostering hitherto unprecedented levels of global hyperconnectivity social media technologies have radically inflated the individual’s moral sphere of influence. Social media grants all its users the potential to directly address massive audiences, a power traditionally the prerogative of wealthy broadcasters, publishing houses, and state media. By expanding our moral sphere of influence, social media has proportionately increased the moral complexity of many of our actions. In turn, we now require far more information than ever before to be attentive to the moral aspects of our actions and in the modern day we depend heavily on social media platforms for the acquisition of this information.
To see this, consider how the democratisation of massive information distribution networks has increased the moral implications of oppressive speech acts and how this affects what we must know to be sensitive to the moral features of such acts. An oppressive speech act is a discriminatory remark uttered within the context of an unjust social structure of systematic oppression which contributes to that structure by signalling to audiences that it is appropriate to regard a certain group as inferior (McGowan, 2009). The statuses of the unjust social structures in which audiences to a speech act are situated are thus morally salient features of that speech act. Since speech acts uttered online have the potential to reach far greater audiences than those uttered in person, they also have significantly more complex moral natures because online speech acts are operative across numerous, potentially disparate, social structures. As such, one requires far more information to be sensitive to the moral aspects of a speech act on social media than they do to be equally sensitive to the moral aspects of a speech act addressed in person to a small audience. Furthermore, both traditional media outlets, like news corporations, and individuals now disseminate their information primarily by way of social media, so it is to these platforms that people turn to acquire information of the world around them.
I think we are now in a good position to understand how social media deprives us of autonomy by fostering faulty informational networks. To boost profits and engagement, social media technologies utilise incredibly complex filtering algorithms to sift through swathes of information and present each user with a tailored media feed meticulously designed to be ingratiatory (Nguyen, 2020). As a result, the informational networks in which social media place us lack coverage which is ‘the completeness of relevant testimony from across one’s whole epistemic community’ (Nguyen, 2020). These faulty informational networks prevent social media users from forming an accurate representation of their wider environment and blind them to certain moral aspects of their actions. For example, by selectively omitting news stories concerning wage disparities across gender from a user’s social media feed, the social media platform would prevent that user from understanding the unjust social structures surrounding them. If such a user were then to commit an oppressive speech act towards women, this act would not be autonomous. The user fails to satisfy AC*. They lack sensitivity to an incredibly important, morally salient feature of their action, namely the unjust social structures which oppress women. In this way, the flawed informational networks that social media platforms incubate pose a significant threat to our autonomy.
I must note here that individual social media users are not responsible for the ignorance which flawed informational networks lead them into. In fact, even virtuous epistemic practices leave one vulnerable to the effects of flawed informational networks (Nguyen, 2023), (Kahan, 2013). Furthermore, recent empirical research has demonstrated that counterbalancing the selective exclusion of information by algorithmic filtering is no easy process (Kahan, 2013). It is not a simple matter of seeking out alternative information for oneself so as to gain a clearer picture of the world. It is not possible for me to detail with due care here to the empirical facts precisely why this is, but it concerns a plethora of mechanisms by which social media technologies systematically exploit our cognitive limitations. This exploitation leaves us impotent to counterbalance the effects of the flawed informational networks and sets AC* off in the distance as an unattainable ideal.
In this essay I hope to have shed light on the severity of the threat social media platforms pose to our autonomy. Social media platforms are drastically expanding the moral complexities of our actions whilst failing to provide us with a commensurate amplification of our knowledge regarding those moral complexities. In fact, social media is deliberately limiting our understanding of those complexities. In doing so, social media prevents us from being sufficiently sensitive to the moral character of our actions which undermines our autonomy. Since autonomy lies at the heart of all ethical considerations, this issue is of immense practical significance, and I think it requires much greater focus in contemporary epistemology. Furthermore, I believe that this is only one of a myriad of ways that social media, and digital media platforms at large, exploit our epistemic vulnerabilities to morally detrimental effects and I hope that further work can be done to explore how this might be combated.
Works Cited:
DataReportal (2024). Global Social Media Stats. [online] DataReportal – Global Digital Insights. Available at: https://datareportal.com/social-media-users.
Feinberg, J. (1989) ‘Autonomy’, in Harm to Self. [Online]. New York: Oxford University Press. pp.27-51
Gryz, J. (2011) On the Relationship Between the Aretaic and the Deontic. Ethical theory and moral practice. [Online] 14 (5), 493–501.
Kahan, D. M. (2013) Ideology, Motivated Reasoning, and Cognitive Reflection. Judgment and Decision Making. [Online] 8 (4), 407–424.
Kiran, A. H. (2015) ‘Four Dimensions of Technological Mediation’, in Postphenomenological Investigations. [Online]. United States: Lexington Books/Fortress Academic. pp.123-140.
McGowan, M. K. (2009) Oppressive Speech. Australasian journal of philosophy. [Online] 87 (3), 389–407.
Nguyen, C. T. (2020) Echo Chambers and Epistemic Bubbles. Episteme. [Online] 17 (2), 141–161.
Nguyen, C. T. (2023) Hostile Epistemology. Social philosophy today. [Online] 39, 9–32.