PHILOSOPHICAL STUDIES最新文献

Often, a health resource distribution (or, more generally, a health policy) ranks higher than another on one value, say, on promoting total population health; and lower on another, say, on promoting that of the worst off. Then, some opine, there need not be a rational determination as to which of the multiple distributions that partially fulfill both one ought to choose. Sometimes, reason determines only partially, intransitively, or contentiously which of the many “compromises” between these two values is best or most choiceworthy. Norman Daniels, Ruth Chang, Martijn Boot, and Anders Herlitz affirm this opinion, which I shall call “value incommensurability,” “rational underdeterminacy,” or “reasonable disagreement.” To decide between the multiple reasonable compromises on health resource distribution, these philosophers recommend a deliberative democratic process, on two main grounds. First, in such situations, deliberation can produce the determinacy needed for decisionmaking. Second, by treating respectfully and justly even those patients or communities for whom the distributive compromise selected is bad, deliberation shields the legitimacy of that policy. Increasingly, practically-oriented bioethics recommends democratic deliberation even more expansively than these philosophers do—for nearly every decision on health resource distribution and not only when values are incommensurate—on these two grounds and on others. And one could propose a more modest variant on this expansive move as the justification of democratic deliberation. I argue that none of these moves warrants democratic deliberation on health policy.

This commentary on Justin D’Arms and Daniel Jacobson’s Rational Sentimentalism explores two key issues: what exactly is the position D’Arms and Jacobson call ‘rational sentimentalism’, and why exactly do they restrict their theorizing to the normative categories they dub ‘the sentimentalist values’? Along the way, a challenge is developed for D’Arms and Jacobson’s claim that there is no “response-independent” account of the fittingness conditions for emotions such as fear, pride, and amusement.

What is the status of metaphorical meaning? Is it an input to semantic composition or is it derived post-semantically? This question has divided theorists for decades. Griceans argue that metaphorical meaning/content is a kind of implicature that is generated through post-semantic processing. Others, such as the contextualists, argue that metaphorical meaning is an input to semantic composition and thus part of “what is said” by an utterance. I think both sides are right: metaphorical meaning is an input to semantic composition and it is also derived post-semantically. I explain how this is possible by positing that successful metaphor involves coining a new word on the spot; this new metaphorical word is ambiguous with its literal counterpart. I show that an ambiguity theory of metaphor, far from being the obvious non-starter that it has long been treated as, actually offers elegant predictions of a whole suite of otherwise recalcitrant linguistic data.

The dominant practice of AI alignment assumes (1) that preferences are an adequate representation of human values, (2) that human rationality can be understood in terms of maximizing the satisfaction of preferences, and (3) that AI systems should be aligned with the preferences of one or more humans to ensure that they behave safely and in accordance with our values. Whether implicitly followed or explicitly endorsed, these commitments constitute what we term a preferentist approach to AI alignment. In this paper, we characterize and challenge the preferentist approach, describing conceptual and technical alternatives that are ripe for further research. We first survey the limits of rational choice theory as a descriptive model, explaining how preferences fail to capture the thick semantic content of human values, and how utility representations neglect the possible incommensurability of those values. We then critique the normativity of expected utility theory (EUT) for humans and AI, drawing upon arguments showing how rational agents need not comply with EUT, while highlighting how EUT is silent on which preferences are normatively acceptable. Finally, we argue that these limitations motivate a reframing of the targets of AI alignment: Instead of alignment with the preferences of a human user, developer, or humanity-writ-large, AI systems should be aligned with normative standards appropriate to their social roles, such as the role of a general-purpose assistant. Furthermore, these standards should be negotiated and agreed upon by all relevant stakeholders. On this alternative conception of alignment, a multiplicity of AI systems will be able to serve diverse ends, aligned with normative standards that promote mutual benefit and limit harm despite our plural and divergent values.

This paper develops and defends novel accounts of accurate and reasonable doubt. We take a cue from Sosa's telic epistemic normative picture to argue that one’s degree of doubt that p is accurate just in case it matches the level of veritic risk involved in believing that p. In turn, on this account, reasonable doubt is doubt that is generated by a properly functioning cognitive capacity with the function of encoding veritic risk.

There are two leading theories about the ontology of dreams. One holds that dreams involve hallucinations and beliefs. The other holds that dreaming involves sensory and propositional imagining. I highlight two features of dreams which are more easily explained by the imagination theory. One is that certain things seem to be true in our dreams, even though they are not represented sensorily; this is easily explained if dreams involve propositional imagining. The other is that dream narratives can be temporally segmented, involving events which take place across long spans of time; this makes sense if dreams involve sensory imagining, for we often sensorily imagine narratives during wakefulness in the same way. The two considerations are unified by the fact that both highlight forms of content determination characteristic of imagining.

By definition, the implicit commitment of a formal theory (textrm{Th}) consists of sentences that are independent of the axioms of (textrm{Th}), but their acceptance is implicit in the acceptance of (textrm{Th}). In Cieśliński (2017, 2018), the phenomenon of implicit commitments was studied from the epistemological perspective through the lenses of the formal theory of believability. The current paper provides a comprehensive proof-theoretic analysis of this approach and compares it to other main theories of implicit commitments. We argue that the formal results presented in the paper favour the believability theory over its main competitors. However, there is still a fly in the ointment. We argue that in its current formulation, the theory cannot deliver all the goods for which it was defined. In particular, being amenable to a generalised conservativeness argument, it does not support the view that the notion of truth is epistemically light. At the end of the paper, we discuss possible ways out of the problem.

According to internalism about prudential value, the token states of affairs that are basically good for you must be suitably connected, under the proper conditions, to your positive attitudes. It is commonly thought that any theory of welfare that implies internalism is guaranteed to respect the alienation constraint, the doctrine that you cannot be alienated from that which is basically good for you. In this paper, I show that extant formulations of internalism do not have this desirable feature. The central defect of traditional formulations is that they do not respect an important but overlooked truth about alienation: namely, that even if a state of affairs is suitably connected to your positive attitudes, your negative attitudes can nonetheless render you alienated from it. By taking into account the relevance of the negative attitudes, I propose the new internalism—the view that x is basically good for you only if you have a net positive attitude towards it—as a way to advance our thinking about what is required to avoid alienating theories of welfare.

Several influential thought experiments from Harman 1973 purport to show that unpossessed evidence can undermine knowledge. Recently, some epistemologists have appealed to these thought experiments in defense of a logically stronger thesis: unpossessed evidence can defeat justification. But these appeals fail to appreciate that Harman himself thought of his examples as Gettier cases, and so would have rejected this strengthening of his thesis. On the contrary, he would have held that while unpossessed evidence can undermine knowledge, it leaves justification intact. In this paper I seek to undermine the viability of Harman’s position. If this is correct, contemporary epistemology faces a choice: either we reject that unpossessed evidence in Harman-style cases bears on knowledge at all, or else we must allow that it undermines knowledge by defeating justification. The former option must explain why Harman’s thought experiments elicit strong ‘no knowledge’ intuitions; the latter option embraces a minority view about the bearing of social expectations on the assessment of knowledge and justification (= the doctrine of normative defeat).

Do the ethical features of an artwork bear on its aesthetic value? This movie endorses misogyny, that song is a civil rights anthem, the clay constituting this statue was extracted with underpaid labor—are facts like these the proper bases for aesthetic evaluation? I argue that this debate has suffered from a false presupposition: that if the answer is “yes” (for at least some such ethical features), such considerations feature as pro tanto contributions to an artwork’s overall aesthetic value, i.e., as merits or flaws which make something have more or less overall aesthetic value. As the case of ethically laden aesthetic evaluation makes clear, however, good aesthetic judgement is irreducibly multi-dimensional, e.g., “the movie has an engaging soundtrack, tasteful camera work, and takes a misogynistically purient perspective on its female lead.” Such a “non-summative” judgement refuses to reduce those various dimensions of aesthetic value to a single aggregate aesthetic evaluation, like “it’s a 6/10” or “it’s a pretty good movie!” I defend both the modest claim that such non-summative evaluations are not mistaken and the extremist claim that summative (i.e., unidimensional) aesthetic evaluation is defective by considering other domains of normative assessment in which summing seems inappropriate, notably including evaluations of people’s character.