Matti Eklund

Forthcoming in Darren Bradley (ed.), Carnap and Contemporary Philosophy. This paper is centered on Carnap’s views on rationality. More specifically, much of the focus is on a puzzle regarding Carnap’s view on rationality that Florian Steinberger has recently discussed. Not only is Steinberger’s discussion of significant intrinsic interest: his discussion also raises general questions about Carnap interpretation. As I have discussed in earlier work, there are two very different ways of interpreting Carnap’s talk of “frameworks” – and, relatedly, different ways of interpreting Carnap’s principle of tolerance. Carnap can be interpreted as either a relativist or as what I call a language pluralist. Steinberger’s puzzle arises given the relativist interpretation; I believe the language pluralist interpretation is correct. Most of the discussion will concern the correct interpretation of Carnap, and what this means for Steinberger’s puzzle. While I will not here mount a full defense of the language pluralist interpretation, I will pause to discuss Vera Flocke’s recent criticism of it. Towards the end, I will describe a puzzle regarding rationality different from Steinberger’s. The puzzle that I describe does arise already for the language pluralist

Here is the liar paradox. We have a sentence, (L), which somehow says of itself that it is false. Suppose (L) is true. Then things are as (L) says they are. (For it would appear to be a mere platitude that if a sentence is true, then things are as the sentence says they are.) (L) says that (L) is false. So, (L) is false. Since the supposition that (L) is true leads to contradiction, we can assert that (L) is false. But since this is just what (L) says, (L) is then true. (For it would appear to be a mere platitude that if things are as a given sentence says they are, the sentence is true.) So (L) is true. So (L) is both true and false. Contradiction.

In part of The Fragmentation of Being, Kris McDaniel discusses the possibility that we—persons—are not fully real, and the normative upshot of this. The broader metaphysical context is a view on which different things have different degrees of being and what is discussed is the possibility that persons do not have the maximal degree of being. McDaniel thinks that this has a problematic normative upshot: we would not matter. I do not agree. Here I go through some reasons for thinking that the possible metaphysical view discussed does not have the normative upshot that McDaniel thinks it has.

My theme here will be vagueness. But first recall Quine’s arguments for the indeterminacy of translation and the inscrutability of reference. (I will presume these arguments to be familiar.) If Quine is right, then there are radically different acceptable assignments of semantic values to the expressions of any language: different assignments of semantic values that for all that is determined by whatever it is that determines semantic value are all acceptable, and all equally good. Quine even argued that the indeterminacy is so radical that some sentences are true under some acceptable assignments but false under others.1 Still, Quine does not allow intermediate truth-values or truthvalue gaps. (As I will put it, avoiding the disjunctive formulation: does not allow that there are sentences which are neuter.) Quine holds on to classical logic and bivalence and requires each acceptable assignment to be classical and bivalent.2.

In contemporary debates about ontology, one prominent skeptical view emphasizes the existence of different possible languages for doing ontology. Eli Hirsch, in recent years the most prominent proponent of a view like this, has defended the claim that “many familiar questions about the ontology of physical objects are merely verbal. Nothing is substantively at stake in these questions beyond the correct use of language” and the claim that “quantifier expressions can have different meaning in different languages”.1 Ted Sider, while critical of the type of view Hirsch defends, has in many places prominently singled it out for critical discussion. In his (2001), he associates with Carnap the view that “different frameworks employ different semantic rules for the quantifiers”, and says that a theorist in Carnap’s tradition would say that the stuff-ontologist and the thing-ontologist have different “frameworks” and that “[W]ithin these frameworks there are answers to what there is, but any question about which framework is the right framework is metaphysical in the pejorative sense of being a pseudo-question”.2 In his (forthcoming), Sider discusses at length the view that disputants in ontological disputes use the same sentences with different meaning and hence in fact each speak the truth, and what he describes as the attendant view that “there are multiple candidate meanings for the quantifiers”.3 In his (2005), Cian Dorr critically discusses “the idea that there are many different possible languages which differ systematically in the truth-values they assign to general ontological claims”.4..

Oystein Linnebo's book Thin Objects is partly devoted to defending the view that some objects are “thin” in that their existence does not impose any substantive demands on the world. In this paper, I discuss the concern that the defense relies on there being entities that serve as the referents of predicates. Linnebo thus seems to assume the thinness of those entities. In the course of my discussion, I also discuss what Linnebo says about the role of criteria of identity in his discussion of reference and existence.

Supervaluationism, traditionally conceived, is the conjunction of three theses: Vagueness in a language gives rise to there being a multitude of acceptable assignments of semantic values to some expressions of the language, These assignments correspond to possible completions of the meanings of vague expressions, Truth is truth under all acceptable assignments, and falsity is falsity under all acceptable assignments. Supervaluationism has three chief virtues. It preserves classical logic. It provides an account of what vagueness is . And it extends nicely to the vagueness of singular terms. I argue, however, that the existence of vagueifiers ‐ expressions like‘roughly’– pose problems for thesis of Supervaluationism, and that the supervaluationist can get around these problems only at a serious cost. Toward the end of the paper, I present an alternative analysis of vagueness, which preserves the main virtues of Supervaluationism, but which identifies vagueness as semantic Overdetermination, rather than as semantic underdetermination

The basic elements of this framework were elaborated upon and defended at some length in Wright’s Truth and Objectivity —henceforth, T&O—which was a marvelous book. The present volume of essays, almost all previously published and including some that predate T&O, continues to discuss the same themes, and is, in virtue of the significance of the ideas discussed and the high level of the discussion, likewise a very important work. The collection is divided into five parts: Realism Reconfigured, Replies to Critics, Ethics, Truth, and Response-Dependence and Cognitive Command. The oldest essay in the collection is “Realism, Anti-Realism, Irrealism, Quasi-Realism,” in section 1, Wright’s 1987 Gareth Evans Memorial Lecture in Oxford. This essay, as well as “Moral Values, Projection, and Secondary Qualities,” from 1988, predates T&O itself. The only previously unpublished material is the latter part of “Minimalism and Deflationism,” in section 4, an essay edited from material from different sources. Each section of the book begins with an introduction, but the introductions contain only brief summaries of the papers.

According to a certain pluralist view in philosophy of mathematics, there are as many mathematical objects as there can coherently be. Recently, Justin Clarke-Doane has explored what consequences the analogous view on normative properties would have. What if there is a normative pluriverse? Here I address this same question. The challenge is best seen as a challenge to an important form of normative realism. I criticize the way Clarke-Doane presents the challenge. An improved challenge is presented, and the role of pluralism in this challenge is assessed.

Jaakko Hintikka has argued that ordinary first-order logic should be replaced byindependence-friendly first-order logic, where essentially branching quantificationcan be represented. One recurring criticism of Hintikka has been that Hintikka'ssupposedly new logic is equivalent to a system of second-order logic, and henceis neither novel nor first-order. A standard reply to this criticism by Hintikka andhis defenders has been to show that given game-theoretic semantics, Hintikka'sbranching quantifiers receive the exact same treatment as the regular first-orderones. We develop a different reply, based around considerations concerning thenature of logic. In particular, we argue that Hintikka's logic is the logic that bestrepresents the language fragment standard first-order logic is meantto represent. Therefore it earns its keep, and is also properly regarded as first-order.

Supervaluationism, traditionally conceived, is the conjunction of three theses: Vagueness in a language gives rise to there being a multitude of acceptable assignments of semantic values to some expressions of the language, These assignments correspond to possible completions of the meanings of vague expressions, Truth is truth under all acceptable assignments, and falsity is falsity under all acceptable assignments. Supervaluationism has three chief virtues. It preserves classical logic. It provides an account of what vagueness is. And it extends nicely to the vagueness of singular terms. I argue, however, that the existence of vagueifiers ‐ expressions like‘roughly’– pose problems for thesis of Supervaluationism, and that the supervaluationist can get around these problems only at a serious cost. Toward the end of the paper, I present an alternative analysis of vagueness, which preserves the main virtues of Supervaluationism, but which identifies vagueness as semantic Overdetermination, rather than as semantic underdetermination.

There is a vexed question in the literature on Marx of whether Marx was somehow anti-morality or if on the contrary he was instead defending a particular, perhaps rather radical, conception of morality. This question will be my starting point. But I will have nothing to contribute to the scholarly question of what Marx’s view was. Rather my aim will be this. Whatever in the end is the correct interpretation of Marx, it is undeniable that there are passages in Marx which if taken literally and straightforwardly give voice to an anti-morality stance. One reason not to read these passages straightforwardly is that it is unclear what it even can mean to be anti-morality, as opposed to merely being against one or other popular conception of morality. I will however see what sense can be made out of anti-moralism; and a positive suggestion will be presented. (My claim will however not be that what I suggest is what Marx had in mind.) The discussion of Marx and anti-moralism will serve as a springboard for bringing up certain other issues. I will discuss thick concepts, and how to resolve a particular problem concerning how best to understand the nature of thick concepts. Generally. I will discuss the semantics of moral terms. I will also present a new type of moral antirealism.