CEU Electronic Theses and Dissertations, 2025
| Author | Koszeghy, Laszlo |
|---|---|
| Title | Mindless Scrolls: Social Kinds and Power in the Algorithmic Society |
| Summary | This dissertation explains a contemporary shift in the conditions of existence of social reality. Using the conceptual framework of social ontology, I argue that our contemporary social reality depends on our minds – both collective and individual intentionality – in a weaker sense than the traditional picture of social reality holds. Specifically, I establish the following: (i) social kinds constructed through data-driven algorithmic systems and forms of algorithmic power are not dependent on collective intentionality; (ii) algorithmic nudging power is neither dependent nor is enhanced by collective intentionality; and (iii) some forms of algorithmic nudging do not require intentional actions from those over whom it is exercised. As such, this work explores the social ontological implications of the shift towards “algorithmic societies” – how our social reality is becoming “mindles s”. T he dissertation consists of two parts. In Part I (Chapters 1 & 2), I focus on the question of how social kinds, including kinds constructed through data-driven algorithmic systems, are dependent on minds. In Part II (Chapters 3 & 4), I focus on how forms of power enabled by data-driven algorithmic systems relate to our traditional understanding of social power. In Chapter 1, I introduce in detail the traditional picture of social reality, which I base on Searle’s theory of social construction and call Intentionalism. Intentionalism posits that the existence of social kinds depends on their being collectively recognized as existing. The question driving this chapter is whether Intentionalism has general coverage over social reality. After unpacking the most charitable form of Intentionalism by considering criticism the view has received, I conclude that if Intentionalism is weakened on a few key matters, it can retain its general coverage over social reality. In Chapter 2, I argue for two claims. First, socially significant real-time classifications constructed through data-driven algorithmic systems – which I will call Modulatory Algorithmic Kinds (MAKs) – qualify as “social kinds”. This is because, given their multiple shared properties, they can be effectively used for predictions. MAKs, then, qualify as social as they are mind-dependent, representation-dependent, and have large-scale, group-level consequences for social reality. Second, MAKs as social kinds challenge Intentionalism. This is because MAKs are not dependent on collective recognition: the relevant “social glue”, instead, is representation through digital data and algorithmic code. In Chapter 3, explain the forms of algorithmic power enabled by data-driven systems. I first define a minimal notion of social power, clarifying the relationship between social status, intentional action, and social power. I then examine two versions of what I call the Intentionality Constraint: power depending on intentional action and power depending on collective recognition. While I accept the first with caveats, I show that power exercised through data-driven systems challenges the second. Finally, I argue that exercising “nudging” power through data-driven algorithmic systems is a form of social power which neither depends on, nor is enhanced by, collective recognition. In Chapter 4, given how Foucault’s notion of productive power is often contrasted with algorithmic power citing the lack of reflection in the latter, I ask in what sense algorithmic power is not productive. After unpacking productive power and its contrast with intimidation, I specify the sense of “reflection” that those over whom – non-productive – algorithmic power is exercised are lacking. I argue that reflection as the intentional states necessary for intentional action is required for the exercise of both productive power and intimidation, and identify behavioural nudging as the form of algorithmic power which can be exercised without this sense of reflection – thus being non-productive and distinct from intimidation. |
| Supervisor | Kronfeldner, Maria Elisabeth |
| Department | Philosophy PhD |
| Full text | https://www.etd.ceu.edu/2025/koszeghy_laszlo.pdf |
Visit the CEU Library.
© 2007-2025, Central European University