Matthew Willis

While voluntary epistemic isolation is often regarded as epistemically detrimental, particularly in discussions of echo chambers and epistemic bubbles, this perspective overlooks the potential benefits isolation can offer to broader epistemic communities. This paper presents a social defense of strategic epistemic isolation, arguing that voluntary epistemic isolation by some agents can, under specific conditions, support collective epistemic progress. Drawing from landmark findings in network epistemology and agent-based modeling, which show that strategically limiting connectivity between agents can prevent premature consensus and support the discovery of superior solutions, I show that epistemic isolation can sometimes serve, rather than hinder, collective epistemic progress. Utilizing more recent models (Jann & Schottmüller, 2025; Heydari Fard, 2025), I further argue that epistemic isolation can be necessary for collective epistemic progress when networks of epistemic agents must contend with rampant misinformation or conformity pressure. This socially-oriented approach challenges the strict view that all voluntary epistemic isolation is unjustified and provides a stronger basis for responding to typical objections concerning bootstrapped corroboration. I conclude by outlining practical conditions for epistemically justified isolation, particularly for groups experiencing ongoing epistemic injustice.

Democratic discourse is increasingly strained by social risk and deep ideological division, leaving citizens reluctant to engage across disagreement even when such engagement is essential for collective epistemic progress. In this paper, we argue that large language models (LLMs) can function as tools for navigating difficult disagreements by creating low-risk epistemic spaces, or environments where agents can test reasons, rehearse arguments, and explore opposing perspectives without incurring the interpersonal costs that often derail human-to-human exchange. We distinguish between two forms of difficult disagreement – socially fraught and deep – and show how LLMs may mitigate the epistemic hazards of each. In socially fraught contexts, LLMs can reduce reputational and emotional vulnerability while redistributing epistemic labor that otherwise disproportionately falls on the marginalized. In cases of deep disagreement, LLMs can flag divergent frameworks, scaffold meta-cognitive awareness, and mediate the transfer of information without triggering individually rational incentives to misinform. While engaging with LLMs will never be a proper substitute for democratic dialogue, we contend that carefully designed LLMs can lower the barriers to more productive engagement across difference under non-ideal conditions.

This paper challenges the viability of pragmatism as a response to skepticism, with particular focus on the epistemic regress problem. Pragmatist epistemology is characterized by three commitments: radical fallibilism, optimism about inquiry, and a method of post hoc belief evaluation. These tenets jointly aim to resist skepticism without lapsing into dogmatism, but I argue they are not mutually sustainable when subject to scrutiny. The post hoc method of generating warrant requires stable evaluative standards to avoid arbitrariness, but these standards themselves become vulnerable to regress if fallibilism is applied consistently. Furthermore, even if such criteria are granted, determining whether a belief satisfies them invites a new target for regress. I call this the evaluation problem, and I argue that it undermines the pragmatist’s anti-skeptical ambitions. Ultimately, I suggest that embracing a form of Pyrrhonian meta-epistemic skepticism may be more consistent with the pragmatist’s fallibilist commitments.

The traditional epistemic regress problem preys upon the thought that knowledge requires something more than correct judgment. That is, if we must earn or succeed at something in order to lay claim to knowledge, the skeptic can employ the epistemic regress problem to attack our ability to meet those success conditions. Despite the fact that many contemporary epistemologists do maintain some success conditions for knowledge, epistemology largely carries on unphased by the epistemic regress problem. In this paper, I will argue that the traditional epistemic regress problem is insufficient in its generality and that, for this reason, many who should otherwise engage with the problem are correct to ignore it. After elaborating on this problem, I will propose a more effective version of the epistemic regress problem based on a new, minimal epistemic standard that creates epistemic success conditions applicable to all normative epistemologies.

The epistemic regress problem targets our ability to provide reasons for our beliefs. If we need reasons for our beliefs, then we may also need to provide reasons for those reasons, and so on into regress. Because the epistemic regress problem is often cast as an attack on our ability to achieve justification, it is often thought that epistemic positions which do not rely on notions like justification escape without difficulty. The first goal of this dissertation is to establish the generality of the epistemic regress problem, beyond all technicalities regarding the nature of justification. To do this, I propose a new minimal epistemic standard, that we should hold no bald assertions to be epistemically acceptable. I then use this epistemic standard to construct two new forms of the epistemic regress problem. The first version of the problem is reminiscent of the classical problem in that it attacks our ability to locate any such reason. The second version of the problem attacks our ability to unite any such reason to a particular claim that it is supposed to support. The second goal of this dissertation is to argue that pragmatism fails as a solution to the epistemic regress problem. The pragmatist seeks to avoid the regress by provisionally accepting their beliefs to be evaluated later. In so doing, the pragmatist seeks to have reasonable beliefs, but only after having accepted them provisionally. After eliminating other theoretical alternatives that the pragmatist may use to bolster their position, I argue that the pragmatic solution is not successful because (1) it is committed to some foundational beliefs that are necessary to facilitate the evaluation of their provisional beliefs which fall prey to the regress, and (2) the employment of these criteria of evaluation encounters the regress in its second form. I conclude by advocating for an understanding of Pyrrhonian skepticism under which we are not required to eschew our everyday beliefs that is nevertheless commensurate with the epistemic regress.