Fictional characters present a problem for semantic theorists. One approach to this problem has been to maintain realism regarding fictional characters, that is to claim that fictional characters exist. In this way names originating from fiction have designata. On this approach the problem of negative existentials is more pressing than it might otherwise be since an explanation must be given as to why we judge them true when the names occurring within them designate existing objects. So, realists must explain the (...) intuitive truth of such statements. Some realists have appealed to pragmatics to explain this, but have not developed these positions fully. What follows is an original account of negative existentials based on the pragmatic process of modulation. Modulation affects the meaning of ‘exists’ such that its extension is merely those things that exist physically. It is then argued that the modulation approach provides a more natural account of the intuitive truth of negative existentials involving fictional characters than an account based on conversational implicatures. Finally, the modulation account is defended against objections presented against similar accounts. (shrink)

Realists about fictional entities often distinguish the properties that a fictional character has and the properties a character holds. Roughly, this is the distinction between the properties that a character really possesses and the properties it fictionally possess. But despite the popularity of this distinction in realist circles, it gives rise to a number of subtle issues about which fictional realists can and do disagree. In this paper, we aim to clarify these issues and defend three related theses. One: that (...) to say that Hermione holds the property of being female is just to say that the singular proposition ⟨Hermione, being female⟩ is true according to Harry Potter. Two: that ordinary objects can hold properties insofar as propositions like ⟨Napoleon, being French⟩ can be true according to fictions like War and Peace. Three: that the distinction between having and holding should not be thought to play any semantic role within the context of fictional realism. (shrink)

This essay explores the use of platonist and nominalist concepts, derived from the philosophy of mathematics and metaphysics, as a means of elucidating the debate on spacetime ontology and the spatial structures endorsed by scientific realists. Although the disputes associated with platonism and nominalism often mirror the complexities involved with substantivalism and relationism, it will be argued that a more refined three-part distinction among platonist/nominalist categories can nonetheless provide unique insights into the core assumptions that underlie spatial ontologies, but it (...) also assists in critiquing alternative uses of nominalism, platonism, and both ontic and epistemic structural realism. (shrink)

According to Timothy Williamson, we should accept the simplest and most powerful second-order modal logic, and as a result accept an ontology of "bare possibilia". This general method for extracting ontology from logic is salutary, but its application in this case depends on a questionable assumption: that modality is a fundamental feature of the world.

This paper is about intentional identity, the phenomenon of intentional attitudes having a common focus. I present an argument against an approach to explaining intentional identity, defended by Nathan Salmon, Terence Parsons and others, that involves positing exotic objects. For example, those who adopt this sort of view say that when two astronomers had beliefs about Vulcan, their attitudes had a common focus because there is an exotic object that both of their beliefs were about. I argue that countenancing these (...) exotic objects does not help us explain intentional identity. (shrink)

This paper concerns Kripke’s puzzle about belief. I have two goals in this paper. The first is to argue that two leading approaches to Kripke’s puzzle, those of Lewis and Chalmers, are inadequate as they stand. Both approaches require the world to supply an object that the relevant intentional attitudes pick out. The problem is that there are cases which, I argue, exhibit the very same puzzling phenomenon in which the world does not supply an object in the required way. (...) The second goal is to draw out a more general lesson about Kripke’s puzzle. I argue that Kripke’s puzzle should be understood as intimately related to a phenomenon known as ‘intentional identity’, and that an adequate account of Kripke’s puzzle should be extensible to cases in which the relevant attitudes are empty. (shrink)

We want to say both that Sherlock Holmes does not exist, and that he is a fictional character. But how can we say these things without committing ourselves to the existence of Sherlock Holmes? Here I develop and defend a non-commital paraphrase of quantification over fictional characters, modeled on the non-commital paraphrase Kit Fine provides for quantification over possibilia. I also develop and defend the view that names for fictional characters are weakly non-referring, in Nathan Salmon’s sense, and so provide (...) us with a non-commital means to express singular propositions. The resulting position allows us to reap the benefits of Fictional Realism without paying the associated ontological cost. (shrink)

The aim of this paper is to explore the possibility that, at least, some metaphysical debates are ‘metalinguistic negotiations’. I will take the dispute between the dominant approaches of realism and the anti-realism ones about the ontological status of scientific models as a case-study. I will argue that such a debate may be better understood as a disagreement, at bottom normatively, motivated, insofar as a normative and non-factual question may be involved in it: how the relevant piece of language ought (...) to be used. Even though I will generally assess the prospects for a broadly deflationist approach, I shall outline a sense in which the dispute can be recast as ‘minimally substantive’. (shrink)

The problem of intersubjective identification arises from the difficulties of explaining how our thoughts and discourse about fictional characters can be directed towards the same (or different) characters given the assumption that there are no fictional entities. In this paper I aim to offer a solution in terms of participation in a practice of thinking and talking about the same thing, which is inspired by Sainsbury's name-using practices. I will critically discuss a similar idea that was put forward by Friend (...) in terms of participation in what Perry calls a notion-network. I will then argue in favor of Sainsbury's baptism-based approach against Perry's information-based approach and I will answer some recent objections that Friend put forward against the former. (shrink)

Against standard descriptivist and referentialist semantics for fictional reports, I will defend a view according to which fictional names do not refer yet they can be distinguished from one another in virtue of their different name-using practices. The logical structures of sentences containing fictional names inherit these distinctions. Different interpretations follow.

The review praises the philosophical quality, but is less enthusiastic about the scholarship and historical accuracy.

This paper addresses the title question and provides an argument for the conclusion that so-called phenomenal intentionality, in both its relational and non-relational construals, cannot be identified with intentionality meant as the property for a mental state to be about something. A main premise of the argument presented in support of that conclusion is that a necessary requirement for a property to be identified with intentionality is that it satisfy the features taken to be definitory of it, namely: the possible (...) non-existence of the intentional object and aspectuality. By taking this premise on board, I attempt to show that phenomenal intentionality cannot be identified with intentionality because, appearances notwithstanding, it ultimately satisfies neither of the two above mentioned features. (shrink)

There are two major semantic theories of proper names: Semantic Descriptivism and Direct Reference. According to Semantic Descriptivism, the semantic content of a proper name N for a speaker S is identical to the semantic content of a definite description “the F” that the speaker associates with the name. According to Direct Reference, the semantic content of a proper name is identical to its referent. Semantic Descriptivism suffers from a number of drawbacks first pointed out by Donnellan (1970) and Kripke (...) (1972). Direct Reference faces difficulties of its own, most importantly the problem of empty names. The most promising Directly Referential solution to this problem is the Unfilled Proposition view, according to which utterances of sentences containing empty names semantically express unfilled propositions. But this view faces the problem of accounting for the intuition that negative existentials involving empty names are true. The most promising way of dealing with this problem within Unfilled Proposition theory is to suppose (i) that utterances of sentences may be used to pragmatically convey propositions they do not semantically express, and (ii) that the proposition pragmatically conveyed by a speaker S's utterance of a sentence containing an empty name N (where “the F” is a definite description S associates with N) is identical to the proposition semantically expressed by an utterance of the sentence obtained by replacing N with “the F”. Call this view “Pragmatic Descriptivism”. With respect to the problem of negative existentials, Pragmatic Descriptivists can insist that, although an utterance of “Santa does not exist” is literally neither true nor false, our taking it to be true may be explained as the result of our having confused the unfilled proposition it semantically expresses with the clearly true descriptive proposition it pragmatically conveys. Despite its theoretical virtues, Pragmatic Descriptivism has recently come under fire. Everett (2003), in particular, has advanced four different lines of criticism, to which Adams and Dietrich (2004) have responded in some detail. In this article, I have two main aims. The first is to argue that Adams and Dietrich's replies to Everett's criticisms (with one exception) are ineffective. I conclude that there is no acceptable strategy for solving the problem of empty names within Direct Reference theory. The second is to argue that there is a promising alternative to Semantic Descriptivism and Direct Reference that requires us to fill unfilled propositions with names, thereby solving the problem of empty names. (shrink)

The aim of this paper is an ontological clarification of the concept of artefact. The following questions are addressed: 1. Do artefacts constitute an ontological category of objects in its own right, and if so, how could this category be characterized? 2. How do artefacts come into existence? 3. What kind of artefacts are there, and in which relations do they stand to each other? It is argued that artefacts are characterized essentially through their genesis and that they owe their (...) existence to mental acts of a particular kind, which I call “acts of creation”. It is maintained that there are not only concrete, material artefacts, but also abstract artefacts, that is, artefacts that cannot be perceived through the senses and are not located in space. Furthermore, it shall be claimed that in a sense those abstract artefacts are even primary to the concrete, material artefacts. (shrink)

This paper explores the prospects of combining two views. The first view is metaphysical rationalism : all things have an explanation. The second view is metaphysical essentialism: there are real essences. The exploration is motivated by a conflict between the views. Metaphysical essentialism posits facts about essences. Metaphysical rationalism demands explanations for all facts. But facts about essences appear to resist explanation. I consider two solutions to the conflict. Exemption solutions attempt to exempt facts about essences from the demand for (...) explanation. Explanation solutions attempt to explain facts about essences. I argue that exemption solutions are less promising than explanation solutions. I then consider how explanation solutions might be developed. I suggest that a “generative” approach is most promising. I tentatively conclude that the prospects for combining metaphysical rationalism and metaphysical essentialism turn on the viability of a generative approach. This sets the agenda for defending the combination as well as the more general project of explaining essences. (shrink)

In this paper, I will defend a new compositional semantics for single-domain free logic. This semantics makes use of a distinction between the semantic value of a singular term and its semantic referent. The semantic value of a singular term is conceived of as a set that either contains the semantic referent or no element at all. The semantic referent is the object that the term designates. Before I will introduce this new semantics for single-domain predicate and an S5-type modal (...) logic in detail, I will present five related problems of compositionality of the standard semantics for these logical systems. After that I will show in detail how my new proposed alternative semantics can be used to solve nearly all of the outlined five problems. (shrink)

Sainsbury 2005 and 2009 offers a theory of empty names that purports to account for the content and truth-value of all utterances involving them. The goal is to do this while offering a homogenous semantic treatment: both empty and nonempty names make the same kind of contribution to truth-values. The account is based on a new theory of reference that purports to be an alternative among nondescriptivist accounts. According to the new theory, there is reference even without referents. In this (...) paper I argue that the theory does not offer a homogeneous semantics for names and that, thus, it fails to offer a more satisfactory alternative to the theories already available. I conclude by briefly describing a way in which a theory could in fact offer such homogeneous semantics for names via a cognitive theory of empty names. (shrink)

In this paper, I will show that typical formal semantic reconstructions of the rigidity of proper names neglect the important aspect that the rigidity of names is determined by our ordinary use of a name relative to the actual world. This fact was clearly pointed out by Kripke, but overlooked by the subsequent discussion concerning this topic. Based on this diagnosis, I will distinguish three different actualized notions of rigidity. Firstly, I will introduce two different new varieties of known versions (...) of rigidity; namely, actualized persistent and actualized obstinate rigidity. Secondly, I will make use of the tools provided by free logic and introduce a new and overlooked version of rigidity, which I will call actually restricted rigidity. Against this background, I will argue that we have different options to formally model the rigidity of proper names in natural languages. Which option we choose, mainly depends on our philosophical background assumptions. For someone who believes in non-existent objects and the possibility of naming such objects, it is the best choice to hold that proper names are actually obstinately rigid designators. For someone who rejects non-existent objects and the possibility of naming such objects, actually restricted rigidity is a better choice than actualized persistent rigidity. (shrink)

Since Kaplan : 81–98, 1979) first provided a logic for context-sensitive expressions, it has been thought that the only way to construct a logic for indexicals is to restrict it to arguments which take place in a single context— that is, instantaneous arguments, uttered by a single speaker, in a single place, etc. In this paper, I propose a logic which does away with these restrictions, and thus places arguments where they belong, in real world conversations. The central innovation is (...) that validity depends not just on the sentences in the argument, but also on certain abstract relations between contexts. This enrichment of the notion of logical form leads to some seemingly counter-intuitive results: a sequence of sentences may make up a valid argument in one sequence of contexts, and an invalid one in another such sequence. I argue that this is an unavoidable result of context sensitivity in general, and of the nature of indexicals in particular, and that reflection on such examples will lead us to a better understanding of the idea of applying logic to context sensitive expressions, and thus to natural language in general. (shrink)

In this paper, I propose a new way to distinguish between indexicals, like “I” and “today”, and demonstratives, like “she” and “this”. The main test case is the second person singular pronoun “you”. The tradition would generally count it as a demonstrative, because the speaker’s intentions play a role in providing it with a semantic value. I present cross-linguistic data and explanations offered of the data in typology and semantics to show that “you” belongs on the indexical side, and argue (...) that they can be generalized to a novel criterion for distinguishing between indexicals and demonstratives. The central theoretical claim is that the semantic values of indexicals are objects which play certain utterance-related roles, which are fixed independently of the words being used in the utterance. For instance, the speaker plays the speaker role whether or not she uses the word “I”, and the addressee plays that role whether or not the speaker uses the word “you”. Demonstratives, on the other hand, pick out objects that play no such role, and are instead helped by the speaker’s word-specific intentions. (shrink)

The most popular and influential strategies used against semantic externalism and the causal theory of reference are critically examined. It is argued that upon closer scrutiny, none of them emerges as truly convincing.

The canonical version of possible worlds semantics for story prefixes is due to David Lewis. This paper reassesses Lewis's theory and draws attention to some novel problems for his account.

In this paper, I defend a view of names from fiction compatible with the Millian theory of proper names. Unlike other attempts at providing a Millian analysis of names from fiction, my approach gives semantic recognition to our pre‐theoretic intuitions without postulating metaphysically dubious entities. The intuitively correct treatment of typical examples, including true negative existential statements, is obtained by appealing only to independently motivated results in the semantics for natural languages.

In this paper, I defend a view of names from fiction compatible with the Millian theory of proper names. Unlike other attempts at providing a Millian analysis of names from fiction, my approach gives semantic recognition to our pre‐theoretic intuitions without postulating metaphysically dubious entities. The intuitively correct treatment of typical examples, including true negative existential statements, is obtained by appealing only to independently motivated results in the semantics for natural languages.

This paper investigates an objection often raised against metaethical error theory. The challenge runs as follows. Metaethical error theory says that all substantive ethical sentences are false. But if a sentence p is false, then given a standard semantics for “not,” ¬p must be true, and vice versa. On the face of it, one can’t hold that p and ¬p are both false. After presenting a more refined version of the challenge (in the form of a set of initially plausible (...) and yet jointly inconsistent principles), the paper examines a common way out of the puzzle, finds it unsatisfactory, and offers some alternative escape routes that, it is submitted, fare better than the standard one. (shrink)

As our data will show, negative existential sentences containing socalled empty names evoke the same strong semantic intuitions in ordinary speakers and philosophers alike.Santa Claus does not exist.Superman does not exist.Clark Kent does not exist.Uttering the sentences in (1) seems to say something truth-evaluable, to say something true, and to say something different for each sentence. A semantic theory ought to explain these semantic intuitions.The intuitions elicited by (1) are in apparent conflict with the Millian view of proper names. According (...) to Millianism, the meaning (or 'semantic value') of a proper name is just its referent. But empty names, such as 'Santa Claus' and 'Superman', appear to lack a .. (shrink)

The mere fact that ontological debates are possible requires us to address the question, what is it to claim that a certain entity or kind of entity exists—in other words, what do we do when we make an existence-claim? I develop and defend one candidate answer to this question, namely that to make an existence-claim with regard to Fs is to claim that we can refer to Fs. I show how this theory can fulfil the most important explanatory desiderata for (...) a theory of existence; I also defend it against the charges of illegitimate ‘semantic ascent’ and of making existence counterfactually dependent on human linguistic ability. (shrink)

The overgeneration argument attempts to show that accepting second-order validity as a sound formal counterpart of logical truth has the unacceptable consequence that the Continuum Hypothesis is either a logical truth or a logical falsehood. The argument was presented and vigorously defended in John Etchemendy’s The Concept of Logical Consequence and it has many proponents to this day. Yet it is nothing but a seductive fallacy. I demonstrate this by considering five versions of the argument; as I show, each is (...) either unsound or lacks a troubling conclusion. (shrink)

According to two-dimensional semantics, primary intension and secondary intension can play the role of reflecting the cognitive aspect of an expression like Fregean sense does. The aim of this paper is to argue that this role is likely a myth. To argue for this, I attempt to show that cognitive aspects of fictional names cannot be explained within the framework of two-dimensional semantics. To be more specific, I consider four ontological theories about fictional characters that two-dimensional semanticists might be tempted (...) to choose. And then, I argue that none of them can be combined with two-dimensional semantics in a satisfactory way. (shrink)

ABSTRACTI propose a different account of fictional objects from the ones already present in the literature. According to my account, fictional objects are culturally created abstract objects dependent for their existence on the pretence attitude adopted by a group of people towards a single fictional content. My work is divided into three parts: in the first one, I present how fictional objects come into existence according to my proposal; in the second part, I illustrate how the existence of fictional objects (...) so conceived may be ontically indeterminate; in the last part, I consider what happens when vague existence and indeterminate identity are claimed within fictional texts. (shrink)

In this essay, I appeal to the mental file approach in order to give an anti-realist semantic analysis of statements containing fictional names. I claim that fictive and parafictive uses of them express conceptual, though not general, propositions constituted by mental files, anchored in the conceptual world of the corresponding fictional story. Moreover, by positing a referential shift determined by the presence of a simulative referential intention characteristic of those uses, it is possible to take them to be true with (...) respect to those conceptual worlds. As for metafictive uses, since they are grounded on mixed intentions, they are considered to express hybrid propositions, partly conceptual and partly referential. (shrink)

This paper introduces a persistence puzzle involving two sports clubs with a somewhat intertwined history. As one might assume, the implications of the puzzle go far beyond a mere plea for a precise metaphysical analysis of certain perplexing quandaries regarding sports clubs and represents a challenge for our everyday understanding of social groups. To overcome the supposed impediment to the puzzle, as a starting point, I will accept the assumption that these entities are both social groups and abstract artifacts and (...) will further argue that they are _temporal abstract objects_. Since temporal objects are in time, it will be determined that the cause of the conundrum is the seeming inability to correctly account for the I-relation between the temporal parts, and I will elaborate on why such an explanation is required. Ultimately, I’ll introduce the notion of _I-shifting_, which will explain how temporal abstract objects persist through time and consequently solve the puzzle. (shrink)

This article focuses on Jonathan Berg’s Theory of Direct Belief as presented in his 2012 book Direct Belief. An Essay on the Semantics, Pragmatics, and Metaphysics of Belief. After regimenting Berg’s key theses and discussing the sources of their general unpopularity, I proceed to reconstruct Berg’s book-length argument for his conclusions. I here make explicit that Berg relies on a range of strong meta-semantic principles and assumptions. I conclude that even if Berg has brought considerable methodological rigor to the on-going (...) debate over the semantics of natural language attitude ascriptions, and has proposed an elegant and consistent theory, he has not offered compelling reasons to accept his preferred methodological constraints in light of the difficulties, which those constraints impose upon attitude ascription semantics. (shrink)

Impossible fictions are valuable evidence both for a theory of fiction and for theories of meaning, mind and epistemology. This article focuses on what we can learn about fiction from reflecting on impossible fictions. First, different kinds of impossible fiction are considered, and the question of how much fiction is impossible is addressed. What impossible fiction contributes to our understanding of "truth in fiction" and the logic of fiction will be examined. Finally, our understanding of unreliable narrators and unreliable narration (...) in fiction needs to accommodate stories that, on the face of it, cannot possibly be true. (shrink)

Creationism about fictional entities requires a principle connecting what fictions say exist with which fictional entities really exist. The most natural way of spelling out such a principle yields inconsistent verdicts about how many fictional entities are generated by certain inconsistent fictions. Avoiding inconsistency without compromising the attractions of creationism will not be easy.

One of the central debates in the philosophy of language is that between defenders of the causal-historical and descriptivist theories of reference. Most philosophers involved in the debate support one or the other of the theories. Building on recent experimental work in semantics, we argue that there is a sense in which both theories are correct. In particular, we defend the view that natural kind terms can sometimes take on a causal-historical reading and at other times take on a descriptivist (...) reading. The meaning will shift depending on the conversational setting. The theoretical view has roots in work by Kitcher. We present some original experiments that support the thesis. (shrink)

Grelling’s Paradox is the paradox which results from considering whether heterologicality, the word-property which a designator has when and only when the designator does not bear the word-property it designates, is had by ‘ ȁ8heterologicality’. Although there has been some philosophical debate over its solution, Grelling’s Paradox is nearly uniformly treated as a variant of either the Liar Paradox or Russell’s Paradox, a paradox which does not present any philosophical challenges not already presented by the two better known paradoxes. The (...) aims of this paper are, first, to offer a precise formulation of Grelling’s Paradox which is clearly distinguished from both the Liar Paradox and Russell’s Paradox; second, to offer a solution to Grelling’s Paradox which both resolves the paradoxical reasoning and accounts for unproblematic predications of heterologicality; and, third, to argue that there are two lessons to be drawn from Grelling’s Paradox which have not yet been drawn from the Liar or Russell’s Paradox. The first lesson is that it is possible for the semantic content of a predicate to be sensitive to the semantic context; i.e., it is possible for a predicate to be an indexical expression. The second lesson is that the semantic content of an indexical predicate, though unproblematic for many cases, can nevertheless be problematic in some cases. (shrink)

In this paper, I argue that sentences that contain ‘omission’ tokens that appear to function as singular terms are meaningful while maintaining the view that omissions are nothing at all or mere absences. I take omissions to be fictional entities and claim that the way in which sentences about fictional characters are true parallels the way in which sentences about omissions are true. I develop a pragmatic account of fictional reference and argue that my fictionalist account of omissions implies a (...) plausible account of the metaphysics of omissions. (shrink)

Names are rigid designators, but what kind of rigidity do they exhibit? Both “obstinately” and “persistently” rigid designators pick out O at every world at which they pick out anything at all. They differ in that obstinately rigid designators also pick out O at worlds at which O fails to exist; persistently rigid designators have no extension whatsoever at worlds at which O fails to exist. The question whether names are obstinate or persistent arises in two contexts: in arguments against (...) rigidified descriptivism (on the grounds that names and rigidified descriptions exhibit different kinds of rigidity) and in considerations about what it means for an expression to be “directly referential.” This paper contends that names are persistent, not obstinate. The position thus clarifies the relationship between Millian theories of names and rigid designation, in addition to undermining attempts to rehabilitate the modal argument against rigidified descriptivism. (shrink)

According to Millianism, the semantic content of a proper name is its semantic referent. Many names, however, lack semantic referent; hence, so-called ‘empty’ names. Empty names raise various problems for Millianism. T.C. Ryckman, Fred Adams, Garry Fuller, Robert Stecker, Kenneth Taylor, and Nicole Wyatt, among others, have defended Millianism against these problems by appeal to pragmatics . I introduce Millianism and the problems raised by empty names for the view, then examine Pragmatic Millianism , its strength, its varieties, and why (...) the previous arguments against PM do not succeed. I then provide my argument against PM: the view oversimplifies the complex phenomenon of association between names and descriptions. I discuss an objection to my argument and rebut that. Finally, I try to draw an outline of a positive view. (shrink)

Neo-Russellianism, which incorporates both Millianism (with regard to proper names) and the thesis of singular Russellian propositions, has widely been defended after the publication of Kripke's Naming and Necessity. The view, however, encounters various problems regarding empty names, names that do not have semantic referents. Nathan Salmon and Scott Soames have defended neo-Russellianism against such problems in a novel way; to account for various intuitions of competent and rational speakers regarding utterances of sentences containing empty names, Salmon and Soames appeal (...) neither to entities similar to Fregean senses, e.g. propositional guises or modes of presentation, nor to Gricean implicatures. In this paper, however, I argue that their view slips into neo-Meinongianism; it is committed to nonexistent objects, assigns various properties to them, and allows quantifiers range over such entities. This, I conclude, makes Salmon and Soames' view less appealing, if not implausible. (shrink)

It is usually taken for granted, in discussions about fiction, that real things or events can occur as referents of fictional names . In this paper, I take issue with this view, and provide several arguments to the effect that it is better to take the names in fiction to refer to fictional surrogates of real objects. Doing so allows us to solve a series of problems that arise on the reference-continuity view. I also show that the arguments philosophers usually (...) rely on in order to ban surrogates are not as serious as they have been taken to be. In the first part of the paper, I describe the two conflicting views. In the second part, I discuss several specific arguments in favor of surrogates. In the third part I take up the kind of reasons ordinarily offered against them. (shrink)

One of the main issues concerning different versions of content externalism is whether or not they are compatible with the privileged access thesis. According to the so-called ‘illusion version’ of externalism, in reference failure cases (such as cases in which an empty proper name is involved) the subject suffers an illusion of entertaining a thought. In this paper, I shall concentrate on a recent argument offered by Jessica Brown, which she calls the “illusion argument”, to the effect that the illusion (...) version of externalism undermines the privileged access thesis (Brown, 2004). After criticizing Brown's argument, I shall try to reconstruct the illusion argument in a more defensible and straightforward way. I will exploit, in my argument, solutions proposed by Goldman and Alston for the so-called ‘generality problem’ (Goldman, 1986; Alston, 1995). Moreover, I shall offer a stronger formulation of the global reliability condition for knowledge, upon which my reconstruction of the illusion argument is based. (shrink)