Hsiang-Yun Chen

Contextualist accounts of “woman,” including Saul (2012), Diaz-Leon (2016), and Ichikawa (2020), aim to capture the variability of the meaning of the term, and do justice to the rights of trans women. I argue that (i) there is an internal tension between a contextualist stance and the commitment to trans-inclusive language, and that (ii) we should recognize and tackle the broader and deeper theoretical and practical difficulties implicit in the semantic debates, rather than collapsing them all into semantics. Moving on, I sketch three strategies to help us advance feminist philosophical endeavors, including how attending to contextual matters can lead us to further reflect on the meta-contextual, such as our role in shaping contexts and whether the working of language is indicative of a larger oppressive social structure.

Artificial intelligence (AI) systems are increasingly adopted to make decisions in domains such as business, education, health care, and criminal justice. However, such algorithmic decision systems can have prevalent biases against marginalized social groups and undermine social justice. Explainable artificial intelligence (XAI) is a recent development aiming to make an AI system’s decision processes less opaque and to expose its problematic biases. This paper argues against technical XAI, according to which the detection and interpretation of algorithmic bias can be handled more or less independently by technical experts who specialize in XAI methods. Drawing on resources from feminist epistemology, we show why technical XAI is mistaken. Specifically, we demonstrate that the proper detection of algorithmic bias requires relevant interpretive resources, which can only be made available, in practice, by actively involving a diverse group of stakeholders. Finally, we suggest how feminist theories can help shape integrated XAI: an inclusive social-epistemic process that facilitates the amelioration of algorithmic bias.

Reluctance to adopt mask-wearing as a preventive measure is widely observed in many Western societies since the beginning of the COVID-19 pandemics. This reluctance toward mask adoption, like any other complex social phenomena, will have multiple causes. Plausible explanations have been identified, including political polarization, skepticism about media reports and the authority of public health agencies, and concerns over liberty, amongst others. In this paper, we propose potential explanations hitherto unnoticed, based on the framework of epistemic injustice. We show how testimonial injustice and hermeneutical injustice may be at work to shape the reluctant mask adoption at both the societal and individual levels. We end by suggesting how overcoming these epistemic injustices can benefit the global community in this challenging situation and in the future.

This chapter addresses the assumed connection between de se attitude ascription and logophoricity in the case of Chinese ziji. It is widely believed that logophors are among the paradigm cases of de se marking, and that long-distance ziji is logophoric. Drawing on a critical examination of a variety of analyses, this chapter argues that long-distance anaphora, de se interpretation, and logophoric marking are overlapping but distinct phenomena. Even if ziji is logophoric, it does not automatically trigger de se requirement. A de se-neutral analysis of ziji is consistent with pragmatic derivations of interpretations that emphasize the self. The findings point to a new approach to long-distance binding, and identify the blocking effect as the key issue for further research.

The phenomenon of intentional identity has bemused philosophical communities since Geach (1967). I argue that the phenomenon is ubiquitous and much more significant than previously acknowledged. The foundations of the problem are implicated in many other well-knownpuzzles, such as Kripke’s (1979) puzzles about beliefs. Thus, the need for a proper analysis is eminently pressing. I specify a template for generalizing intentional identity, identify the challenges involved, and argue that positing a level of representational entity in both philosophy of mind and language(e.g., mental files) is a promising approach to tackling the problem across the board.

Bishop sentences such as “If a bishop meets a bishop, he blesses him” have long been considered problematic for the descriptivist (or E-type) approach of donkey anaphora (e.g. Evans, 1977; Heim, 1990; and Neale, 1990). Elbourne (2005) offers a situational descriptivist analysis that allegedly solves the problem, and furthermore extends its explanatory coverage to bishop sentence with coordinate subjects. However, I throw serious doubts on Elbourne’s analysis. Specifically, I argue that the purported solution is committed to the use of unbound anaphora, and it cannot sustain the claimed empirical adequacy.

I argues that quantificational adverbs are unselective binders over individuals. The Lewisian analysis, however, fails to recognize the ambiguity in some quantificationally modified conditionals. That the Lewisian approach cannot predict some attested reading is known as the “proportion problem.” I propose a solution based on the following ideas: (a) quantificational adverbs bind selectively; (b) a singular indefinite and its anaphoric pronoun may introduce a plural discourse referent, and (c) plural predication is elusive.

In Mental Files, Recanati proposes a non-descriptivist approach to reference in terms of mental files, mental representations that play the role of Fregean mode of presentation. Recanati argues that we refer via mental files and that the reference of a file is determined relationally, rather than satisficationally; files are not to be equated to the information they contain, but typed by their function—to store information gained through certain epistemically rewarding relation to objects in the environment. I offer a critical overview of Recanati’s framework and raise two questions about the nature and workings of files.

What is gender and on what should gender classification be based? Dembroff (2018) has recently claimed that, for reasons of social justice, gender classification should not track extant gender kinds. They further argue for ontological pluralism—the existence of many gender kinds, and recommend that we combat oppression by imitating the gender kinds and classification practices in non-oppressive communities. Contra Dembroff, I argue that the analysis is subject to a number of internal problems, including a misguided self-characterization and a tension between ontological pluralism and imitation. In addition, I sketch a different perspective on the relation between gender kinds and gender classification practices that upholds Dembroff’s intended goal but ameliorates the unintended complications.

Social constructionist analysis of kind terms such as ‘women’ are often criticized as counterintuitive. In response, Haslanger claims that such charges are moot once the distinctions between different types of philosophical analyses and their corresponding concepts are in place. I argue that even with the said distinctions, the Haslangerian definition of ‘women’ is problematic. Drawing on recent discussions on contextualism, metalinguistic negotiation, and the crucial role solidarity plays in politically significant terms, I claim that Haslanger’s replies would lead to consequences contrary to the stated goal of her project. Moreover, I offer a new proposal that takes seriously the aim of ameliorating ‘women.’ My account draws on a dynamic understanding of solidarity based on feminist reconceptualization and makes recourse to Mikkola’s model of trait/norm covariance. Based on the vision that a proper ameliorative analysis in the positive moment should be forward-looking to enhance our coordination in a more just society, my construal elucidates what ‘women’ can and ought to be.