
Felt emotional ambivalence, bittersweet emotional experience, and other cases of apparent coexistence between affectively discordant emotions raise a much-debated empirical question: do affectively discordant feelings actually coexist (Coexistence hypothesis), or do they just seem to us to do so, due to their rapid alternation (Vacillation hypothesis)? This paper seeks to clarify what is at stake in addressing this question, without thereby endorsing any of the standard answers to it as the debate is usually framed; instead, it argues that coexistence and vacillation are best treated as distinct, context‑dependent mechanisms rather than rival global accounts of mixed affect. The analysis I propose reveals that the denial of Coexistence does not directly depend on commitment to ‘bipolar’ models of affective valence as is normally accepted, since the polar opposition of pleasure and unpleasure does not by itself exclude Coexistence. Rather, this denial stems from commitment to a particular ontology of emotions, namely, the view that occurrent emotions are partly constituted from a simple, unitary core affect state that, at any given instant, must bear either positive or negative valence. Having disentangled the Vacillation hypothesis from debates about the structure of valence, the paper further argues that recent psychology literature has misinterpreted Kahneman’s ‘vacillation hypothesis’, overgeneralizing it beyond its original theoretical context. Kahneman originally proposed Vacillation for specific cases involving commensurable evaluative comparisons (e.g., framing effects in economic decision-making), not as a general account of mixed emotional affect. The paper thus distinguishes between two mechanisms: vacillation, which applies when competing evaluations operate on a shared metric, and coexistence, which applies to cases in which emotion episodes are focused on distinct, incommensurable value-aspects of emotional situations.
This article develops the concept of manipulative injustice to identify a normative vulnerability in legal argument schemes. Dialectical models of legal argumentation do not themselves prescribe the allocation of burdens found in particular legal doctrines, nor are they incapable of representing strategic or bad-faith speech. However, even dialectical legal argument schemes may be vulnerable to asymmetrical burdens of proof with inadequate mechanisms for addressing pretext, evasion, and bad faith. The article develops this claim through the treatment of pretext and bad faith, focusing on the McDonnell Douglas burden-shifting framework and the “Pretext Plus” approach associated with St. Mary’s Honor Center v. Hicks. This comparison reveals an important distinction between the dialectical defeat of an explanation and the legal sufficiency of a claim. I argue that manipulative injustice can arise when asymmetrical burdens of proof converge with inadequate inferential mechanisms for evaluating objective and circumstantial evidence of concealed motive. The article identifies three dimensions for evaluating such schemes: evidence conditions, specifying what evidence supports an inference of pretext or bad faith; defeat conditions, specifying when such evidence defeats a proffered explanation; and consequence conditions, specifying what legal significance follows from that defeat. Manipulative injustice thus provides a diagnostic framework for evaluating when legally instantiated argument schemes become vulnerable to strategic manipulation.
The standard account of epistemic injustice locates its causes in individuals, in their prejudices, their failures of attention, and their unwillingness to revise. This paper argues that this account is incomplete, perhaps fundamentally so. Polarization, systematic exclusion, and the erosion of public trust are not simply aggregates of individual vice. They are structurally distinct: stable patterns of obstruction that emerge from how communities argue, not merely from how individuals think. I call these patterns collective epistemic vices, or CEVs. The analysis draws on Cassam’s (2019) obstructivist framework, according to which what makes something epistemically vicious is its tendency to block knowledge acquisition, transmission, or revision. The paper extends this framework to the collective level. Three vices are examined. Collective Non Acceptance disqualifies speakers or positions before their reasons can be assessed. Collective Contempt moralizes disagreement by reading epistemic difference as evidence of bad faith. Collective Hostility encodes these exclusions into platform architectures and institutional incentives. Each vice produces recognizable forms of epistemic injustice, including testimonial, hermeneutical, and zetetic injustice, but through mechanisms that accounts centered on individuals cannot fully capture. The deeper problem, which occupies much of this paper, is that vices and structures are not straightforwardly related as cause and effect. They sustain each other. Epistemic repair, if it is possible, requires coordinated intervention at both levels.
This paper is divided into two main parts. In the first part, I defend a non-descriptivist account of imperatives, based on Paul Portner’s idea that imperatives serve to manage speakers’ to-do lists. I show that, once we recognize the multiplicity of operations that can be performed on a to-do list, the account can explain weak uses of imperatives without postulating additional structure on the to-do list or another list alongside it. In the second part of the paper, I present a logical framework which integrates the dynamic forces associated with declaratives, imperatives, and rejections thereof. I use this framework to inferentially explain the meaning of deontic modals such as must in terms of imperatives. The resulting inferential expressivist account has the resources to explain performative uses of must and hitherto unaccounted for data about their occurrence pattern.
This paper addresses a central phenomenological puzzle: how to reconcile the traditional view of compassion as an instance of suffering with contemporary neuroscientific findings characterizing it as being hedonically positive. Building on the attitudinalist framework, I argue that this conflict arises from a failure to see emotions as unfolding episodes. By thinking of emotions as felt readinesses to act, we can understand compassion’s valence not as a fixed hedonic tone, but as a dynamic trajectory within a structured affective space. The argument proceeds in three stages. First, I establish that positive and negative emotions are defined by their structural orientation, welcoming or rejecting value. Second, I apply this to compassion, arguing it is fundamentally a mode of taking displeasure in another’s suffering. Finally, I show that the “positivity” noted by scientists is the result of the benevolent action tendency being allowed to run its course, transforming initial distress into the expansive experience of relief and connection.