This paper defends the view that the Faculty of Language is compositional, i.e., that it computes the meaning of complex expressions from the meanings of their immediate constituents and their structure. I fargue that compositionality and other competing constraints on the way in which the Faculty of Language computes the meanings of complex expressions should be understood as hypotheses about innate constraints of the Faculty of Language. I then argue that, unlike compositionality, most of the currently available non-compositional constraints predict (...) incorrect patterns of early linguistic development. This supports the view that the Faculty of Language is com- positional. More generally, this paper presents a way of framing the compositionality debate (by focusing on its implications for language acquisition) that can lead to its even- tual resolution, so it will hopefully also interest theorists who disagree with its main conclusion. (shrink)

The topic of this dissertation is the nature of linguistic competence, the capacity to understand and produce sentences of natural language. I defend the empiricist account of linguistic competence embedded in the connectionist cognitive science. This strand of cognitive science has been opposed to the traditional symbolic cognitive science, coupled with transformational-generative grammar, which was committed to nativism due to the view that human cognition, including language capacity, should be construed in terms of symbolic representations and hardwired rules. (...) Similarly, linguistic competence in this framework was regarded as being innate, rule-governed, domain-specific, and fundamentally different from performance, i.e., idiosyncrasies and factors governing linguistic behavior. I analyze state-of-the-art connectionist, deep learning models of natural language processing, most notably large language models, to see what they can tell us about linguistic competence. Deep learning is a statistical technique for the classification of patterns through which artificial intelligence researchers train artificial neural networks containing multiple layers that crunch a gargantuan amount of textual and/or visual data. I argue that these models suggest that linguistic competence should be construed as stochastic, pattern-based, and stemming from domain-general mechanisms. Moreover, I distinguish syntactic from semantic competence, and I show for each the ramifications of the endorsement of a connectionist research program as opposed to the traditional symbolic cognitive science and transformational-generative grammar. I provide a unifying front, consisting of usage-based theories, a construction grammar approach, and an embodied approach to cognition to show that the more multimodal and diverse models are in terms of architectural features and training data, the stronger the case is for the connectionist linguistic competence. I also propose to discard the competence vs. performance distinction as theoretically inferior so that a novel and integrative account of linguistic competence originating in connectionism and empiricism that I propose and defend in the dissertation could be put forward in scientific and philosophical literature. (shrink)

According to Amie Thomasson's Modal Normativism (MN), knowledge of metaphysical modality is to be explained in terms of a speaker’s mastery of semantic rules, as opposed to one’s epistemic grasp of independent modal facts. In this chapter, I outline (MN)'s account of modal knowledge (§1) and argue that more than semantic mastery is needed for knowledge of metaphysical modality. Specifically (§2), in reasoning aimed at gaining such knowledge, a competent speaker needs to further deploy essentialist principles and information. (...) In response, normativists might contend that a competent speaker will only need to appeal to specific independence counterfactuals, on analogy with quasi-realism about morality. These conditionals fix the meaning of our terms at the actual world, independently of the particular context in which a statement is evaluated. However, I show that this strategy causes several problems for the account (§3). While those problems might perhaps be avoided by endorsing a certain picture of modal metaphysics (Modal Monism), such a picture involves notorious issues that normativists will have to address (§4). It is thus doubtful that (MN) can explain knowledge of metaphysical modality. Still, it may explain some modal knowledge without committing to Modal Monism. As I show (§5), semantic mastery may suffice for gaining knowledge of logical-conceptual modality or analyticity. (shrink)

It's widely accepted that social facts about an individual's linguistic community can affect both the reference of her words and the concepts those words express. Theorists sympathetic to the internalist tradition have sought to accommodate these social dependence phenomena without altering their core theoretical commitments by positing deferential reference-fixing criteria. In this paper, we sketch a different explanation of social dependence phenomena, according to which all concepts are individuated in part by causal-historical relations linking token elements of thought.

The epistemic modal auxiliaries must and might are vehicles for expressing the force with which a proposition follows from some body of evidence or information. Standard approaches model these operators using quantificational modal logic, but probabilistic approaches are becoming increasingly influential. According to a traditional view, must is a maximally strong epistemic operator and might is a bare possibility one. A competing account—popular amongst proponents of a probabilisitic turn—says that, given a body of evidence, must \ entails that \\) is (...) high but non-maximal and might \ that \\) is significantly greater than 0. Drawing on several observations concerning the behavior of must, might and similar epistemic operators in evidential contexts, deductive inferences, downplaying and retractions scenarios, and expressions of epistemic tension, I argue that those two influential accounts have systematic descriptive shortcomings. To better make sense of their complex behavior, I propose instead a broadly Kratzerian account according to which must \ entails that \ = 1\) and might \ that \ > 0\), given a body of evidence and a set of normality assumptions about the world. From this perspective, must and might are vehicles for expressing a common mode of reasoning whereby we draw inferences from specific bits of evidence against a rich set of background assumptions—some of which we represent as defeasible—which capture our general expectations about the world. I will show that the predictions of this Kratzerian account can be substantially refined once it is combined with a specific yet independently motivated ‘grammatical’ approach to the computation of scalar implicatures. Finally, I discuss some implications of these results for more general discussions concerning the empirical and theoretical motivation to adopt a probabilisitic semantic framework. (shrink)

In this paper I show that we have strong empirical and theoretical reasons to treat the verbs we use in our semantic theorizing—particularly ‘refers to ’, ‘applies to ’, and ‘is true of ’—as intensional transitive verbs. Stating our semantic theories with intensional vocabulary allows us to partially reconcile two competing approaches to the nature and subject-matter of semantics: the Chomskian approach, on which semantics is non-relational, internalistic, and concerns the psychology of language users, and the Lewisian approach, (...) on which semantics is fully relational, specifies truth-conditions, and has metaphysical implications. ITVs have two readings: an intensional, de dicto reading, and a relational, de re reading. A semantic theory stated with the de dicto readings of our semantic verbs captures the core insights of the Chomskian approach to semantics, in part because it allows us to assign extremely fine-grained semantic values to expressions, even when those expressions are empty. On the other hand, the de re reading yields a theory that is fully relational, and issues in truth-conditions. The resulting theories are related—and compatible—in that they are expressed by two different readings of the very same semantic vocabulary, and plausibly, the distinction between these two readings is one of scope. (shrink)

In philosophy of language, a distinction has been proposed between two aspects of lexical competence, i.e. referential and inferential competence (Marconi 1997). The former accounts for the relationship of words to the world, the latter for the relationship of words among themselves. The distinction may simply be a classification of patterns of behaviour involved in ordinary use of the lexicon. Recent research in neuropsychology and neuroscience, however, suggests that the distinction might be neurally implemented, i.e., that different cognitive (...) architectures with partly distinct neural realizations might be responsible for cognitive performances involving inferential and referential aspects of semantics. This hypothesis is strongly consistent with patient data supporting the notion of a functional double dissociation between inferential and referential abilities, and with a set of direct cortical mapping studies and neuroimaging experiments suggesting that inferential and referential abilities are underpinned by at least partly different regions of the human brain (e.g., Marconi et al. 2013; review in Calzavarini 2017). The initial hypotheses formulated in the setting of the philosophy of language, along with the neuropsychological experimental evidence (about how referential and inferential competences may be neurally instantiated) can be the input to computational modelling activities involving the inferential and the referential aspects of lexical competence. The aim of the talk is to offer a critical discussion of the kind of formalisms that can be used to model the two aspects of lexical competence, and of the main difficulties related to the use of these computational techniques. -/- Inferential (=symbolic) vs referential (=connectionist) formalisms. (shrink)

In this paper I will be concerned with the question of the extent to which semantics can be thought of as a purely formal exercise, which we can engage in in a way that is neutral with respect to how our formal system is to be interpreted. I will be arguing, to the contrary, that the features of the formal systems which we use to do semantics are closely linked, in several different ways, to the interpretation that we give to (...) those formal systems. The occasion for this question, and the main example that I will use to illustrate my answer to it, is the close relationship between the formal systems employed in recent statements of apparently competing accounts of epistemic modals with the dynamic, expressivist, and relativist theoretical paradigms. The structure of the paper will be straightforward. In part 1, I will briefly introduce four theories of epistemic modals – one dynamic theory, two expressivist theories, and one relativist theory. Then in part 2 I’ll show that one expressivist theory is formally equivalent to the dynamic theory, that the other is formally equivalent to the relativist theory, and that the two expressivist theories are themselves essentially notational variants. I’ll use these facts to pose our central question: if these theories have so much formally in common, then doesn’t that suggest that we can separate the task of constructing a formal semantics from the task of deciding between competing interpretations of it? Finally, in part 3 I’ll answer that question in the negative. There are at least three reasons why formal semantics cannot be separated from questions of interpretation that are illustrated by the theories I introduce in part 1. (shrink)

This paper presents reasons against semantic relativism. Semantic relativism is motivated by intuitions that are presumed to raise problems for traditional or contextualist semantics in contested domains of discourse. Intuition-based arguments are those based on competent speakers’ putative intuitions about seeming faultless disagreement, eavesdropper, and retraction cases. I will organize the discussion in three parts. First, I shall provide a brief introduction to the intuition-based arguments offered in favor of semantic relativism. Second, I shall indicate that there (...) are ways for contextualism to explain the (appearance of) intuitions that support semantic relativism. Third, I shall review some experimental results and independent arguments that put into question the appeal of semantic relativism. (shrink)

Semantic blindness is the inability to recognize semantic features of terms one can competently use. A theory that implies semantic blindness incurs a burden to explain how one can competently use a term without realizing how the term works. An argument advanced in favor of epistemic relativism is that its main competitors, contextualism and subject-sensitive invariantism, imply that speakers suffer from semantic blindness regarding ‘knows’ while relativism has no such implication. However, there is evidence that relativism (...) also implies semantic blindness regarding ‘knows,’ apparently crippling the case for relativism. In this paper I argue that the semantic blindness that affects relativism is not a problem at all. First, the blindness is not as widespread as it appears. It does not prevent ordinary speakers from expressing important epistemic truths. Further, I provide an error theory for relativism that has three features that render it unproblematic: there is evidence independent of relativism that people make this error, relativism predicts this error; it is not an ad hoc rescue, the error only occurs in rare and obscure situations. People are fallible and finite, and assuming relativism is true, they make mistakes exactly where we should expect them to. (shrink)

According to the philosophical tradition, translation is successful when one has substituted words and sentences from one language with those from another by cross-linguistic synonymy. Moreover, according to the orthodox view, the meaning of expressions and sentences of languages are determined by their basic or systematic role in a language. This makes translating normative and evaluative discourse puzzling for two reasons. First, as languages are syntactically and semantically different because of their peculiar cultural and historical influences, and as values and (...) norms differ across cultures, it is unlikely that languages will have synonymous evaluative and normative expressions. If translation is only successful by cross-linguistic synonymy, it would seem that we will not be able to translate the value theoretic claims of persons from radically different cultures. But it is with such persons that dialogue on evaluative matters is imperative, to resolve ethical and axiological differences that could be the root of conflict. Second, as the orthodox account of meaning renders it linguistically relative, it is unlikely that expressions across languages will be cross-linguistically synonymous. Thus, on the Orthodox account of translation, translation is indeterminate (as W.V.O. Quine has argued) or impossible (as Jacques Derrida has argued). In this dissertation I argue for a novel theory of meaning and translation based on innovations in the translation studies literature and my prior work in cross cultural research, which I call Text-Type Semantics or TTS. TTS explains how translation is successful while affirming radical cultural and linguistic diversity. It treats disciplinary concerns as the neutral criteria to calibrate translation. On the basis of TTS I argue that we need what I call the “Quasi-Indexical” account of thick and thin concepts (or QI) to translate normative and evaluative discourse. I argue that QI and TTS succeed where competing accounts in the moral semantics literature (such as Non-Analytic Naturalism and Expressivism) fail. The argument also shows that the relativization of truth (in philosophy and beyond) to languages and cultures is mistaken. (shrink)

A generative grammar for a language L generates one or more syntactic structures for each sentence of L and interprets those structures both phonologically and semantically. A widely accepted assumption in generative linguistics dating from the mid-60s, the Generative Grammar Hypothesis , is that the ability of a speaker to understand sentences of her language requires her to have tacit knowledge of a generative grammar of it, and the task of linguistic semantics in those early days was taken to be (...) that of specifying the form that the semantic component of a generative grammar must take. Then in the 70s linguistic semantics took a curious turn. Without rejecting GGH, linguists turned away from the task of characterizing the semantic component of a generative grammar to pursue instead the Montague-inspired project of providing for natural languages the same kind of model-theoretic semantics that logicians devise for the artificial languages of formal systems of logic, and “formal semantics” continues to dominate semantics in linguistics. This essay argues that the sort of compositional meaning theory that would verify GGH would not only be quite different from the theories formal semanticists construct, but would be a more fundamental theory that supersedes those theories in that it would explain why they are true when they are true, but their truth wouldn’t explain its truth. Formal semantics has undoubtedly made important contributions to our understanding of such phenomena as anaphora and quantification, but semantics in linguistics is supposed to be the study of meaning. This means that the formal semanticist can’t be unconcerned that the kind of semantic theory for a natural language that interests her has no place in a theory of linguistic competence; for if GGH is correct, then the more fundamental semantic theory is the compositional meaning theory that is the semantic component of the internally represented generative grammar, and if that is so, then linguistic semantics has so far ignored what really ought to be its primary concern. (shrink)

The logicality of language is the hypothesis that the language system has access to a ‘natural’ logic that can identify and filter out as unacceptable expressions that have trivial meanings—that is, that are true/false in all possible worlds or situations in which they are defined. This hypothesis helps explain otherwise puzzling patterns concerning the distribution of various functional terms and phrases. Despite its promise, logicality vastly over-generates unacceptability assignments. Most solutions to this problem rest on specific stipulations about the properties (...) of logical form—roughly, the level of linguistic representation which feeds into the interpretation procedures—and have substantial implications for traditional philosophical disputes about the nature of language. Specifically, contextualism and semantic minimalism, construed as competing hypotheses about the nature and degree of context-sensitivity at the level of logical form, suggest different approaches to the over-generation problem. In this paper, I explore the implications of pairing logicality with various forms of contextualism and semantic minimalism. I argue that to adequately solve the over-generation problem, logicality should be implemented in a constrained contextualist framework. (shrink)

Type-logical semantics studies linguistic meaning with the help of the theory of types. The latter originated with Russell as an answer to the paradoxes, but has the additional virtue that it is very close to ordinary language. In fact, type theory is so much more similar to language than predicate logic is, that adopting it as a vehicle of representation can overcome the mismatches between grammatical form and predicate logical form that were observed by Frege and Russell. The grammatical forms (...) of ordinary language sentences consequently may be taken to be much less misleading than logicians in the first half of the 20th century often thought them to be. This was realized by Richard Montague, who used the theory of types to translate fragments of ordinary language into a logical language. Semantics is commonly divided into lexical semantics, which studies the meaning of words, and compositional semantics, which studies the way in which complex phrases obtain a meaning from their constituents. The strength of type-logical semantics lies with the latter, but type-logical theories can be combined with many competing hypotheses about lexical meaning, provided these hypotheses are expressed using the language of type theory. (shrink)

Arguments from disagreement often take centre stage in debates between competing semantic theories. This paper explores the theoretical basis for arguments from disagreement and, in so doing, proposes methodological principles which allow us to distinguish between legitimate arguments from disagreement and dialectically ineffective arguments from disagreement. In the light of these principles, I evaluate Cappelen and Hawthorne's [2009] argument from disagreement against relativism, and show that it fails to undermine relativism since it is dialectically ineffective. Nevertheless, I argue that (...) an alternative challenge to relativism based on disagreement is available. More generally, I argue that semantic theory is not answerable to data stemming from ‘loaded’ philosophical principles regarding the nature of disagreement. Rather, semantic theorists will exhaust their dialectical responsibilities regarding disagreement if they can demonstrate consistency with a minimal account of the concept. (shrink)

The thesis deals with the concept of truth and the paradoxes of truth. Philosophical theories usually consider the concept of truth from a wider perspective. They are concerned with questions such as - Is there any connection between the truth and the world? And, if there is - What is the nature of the connection? Contrary to these theories, this analysis is of a logical nature. It deals with the internal semantic structure of language, the mutual semantic connection (...) of sentences, above all the connection of sentences that speak about the truth of other sentences and sentences whose truth they speak about. Truth paradoxes show that there is a problem in our basic understanding of the language meaning and they are a test for any proposed solution. It is important to make a distinction between the normative and analytical aspect of the solution. The former tries to ensure that paradoxes will not emerge. The latter tries to explain paradoxes. Of course, the practical aspect of the solution is also important. It tries to ensure a good framework for logical foundations of knowledge, for related problems in Artificial Intelligence and for the analysis of the natural language. Tarski’s analysis emphasized the T-scheme as the basic intuitive principle for the concept of truth, but it also showed its inconsistency with the classical logic. Tarski’s solution is to preserve the classical logic and to restrict the scheme: we can talk about the truth of sentences of a language only inside another essentially richer metalanguage. This solution is in harmony with the idea of reflexivity of thinking and it has become very fertile for mathematics and science in general. But it has normative nature | truth paradoxes are avoided in a way that in such frame we cannot even express paradoxical sentences. It is also too restrictive because, for the same reason we cannot express a situation in which there is a circular reference of some sentences to other sentences, no matter how common and harmless such a situation may be. Kripke showed that there is no natural restriction to the T-scheme and we have to accept it. But then we must also accept the riskiness of sentences | the possibility that under some circumstances a sentence does not have the classical truth value but it is undetermined. This leads to languages with three-valued semantics. Kripke did not give any definite model, but he gave a theoretical frame for investigations of various models | each fixed point in each three-valued semantics can be a model for the concept of truth. The solutions also have normative nature | we can express the paradoxical sentences, but we escape a contradiction by declaring them undetermined. Such a solution could become an analytical solution only if we provide the analysis that would show in a substantial way that it is the solution that models the concept of truth. Kripke took some steps in the direction of finding an analytical solution. He preferred the strong Kleene three-valued semantics for which he wrote it was "appropriate" but did not explain why it was appropriate. One reason for such a choice is probably that Kripke finds paradoxical sentences meaningful. This eliminates the weak Kleene three valued semantics which corresponds to the idea that paradoxical sentences are meaningless, and thus indeterminate. Another reason could be that the strong Kleene three valued semantics has the so-called investigative interpretation. According to this interpretation, this semantics corresponds to the classical determination of truth, whereby all sentences that do not have an already determined value are temporarily considered indeterminate. When we determine the truth value of these sentences, then we can also determine the truth value of the sentences that are composed of them. Kripke supplemented this investigative interpretation with an intuition about learning the concept of truth. That intuition deals with how we can teach someone who is a competent user of an initial language (without the predicate of truth T) to use sentences that contain the predicate T. That person knows which sentences of the initial language are true and which are not. We give her the rule to assign the T attribute to the former and deny that attribute to the latter. In that way, some new sentences that contain the predicate of truth, and which were indeterminate until then, become determinate. So the person gets a new set of true and false sentences with which he continues the procedure. This intuition leads directly to the smallest fixed point of strong Kleene semantics as an analytically acceptable model for the logical notion of truth. However, since this process is usually saturated only on some transfinite ordinal, this intuition, by climbing on ordinals, increasingly becomes a metaphor. This thesis is an attempt to give an analytical solution to truth paradoxes. It gives an analysis of why and how some sentences lack the classical truth value. The starting point is basic intuition according to which paradoxical sentences are meaningful (because we understand what they are talking about well, moreover we use it for determining their truth values), but they witness the failure of the classical procedure of determining their truth value in some "extreme" circumstances. Paradoxes emerge because the classical procedure of the truth value determination does not always give a classically supposed (and expected) answer. The analysis shows that such an assumption is an unjustified generalization from common situations to all situations. We can accept the classical procedure of the truth value determination and consequently the internal semantic structure of the language, but we must reject the universality of the exterior assumption of a successful ending of the procedure. The consciousness of this transforms paradoxes to normal situations inherent to the classical procedure. Some sentences, although meaningful, when we evaluate them according to the classical truth conditions, the classical conditions do not assign them a unique value. We can assign to them the third value, \undetermined", as a sign of definitive failure of the classical procedure. An analysis of the propagation of the failure in the structure of sentences gives exactly the strong Kleene three-valued semantics, not as an investigative procedure, as it occurs in Kripke, but as the classical truth determination procedure accompanied by the propagation of its own failure. An analysis of the circularities in the determination of the classical truth value gives the criterion of when the classical procedure succeeds and when it fails, when the sentences will have the classical truth value and when they will not. It turns out that the truth values of sentences thus obtained give exactly the largest intrinsic fixed point of the strong Kleene three-valued semantics. In that way, the argumentation is given for that choice among all fixed points of all monotone three-valued semantics for the model of the logical concept of truth. An immediate mathematical description of the fixed point is given, too. It has also been shown how this language can be semantically completed to the classical language which in many respects appears a natural completion of the process of thinking about the truth values of the sentences of a given language. Thus the final model is a language that has one interpretation and two systems of sentence truth evaluation, primary and final evaluation. The language through the symbol T speaks of its primary truth valuation, which is precisely the largest intrinsic fixed point of the strong Kleene three valued semantics. Its final truth valuation is the semantic completion of the first in such a way that all sentences that are not true in the primary valuation are false in the final valuation. (shrink)

In this paper, I propose a virtue-theoretic approach to semantics, according to which the study of linguistic competence in particular, and the study of meaning and language in general, should focus on a speaker's interpretative virtues, such as charity and interpretability, rather than the speaker's knowledge of rules. The first part of the paper proffers an argument for shifting to virtue semantics, and the second part outlines the nature of such virtue semantics.

The paper by Fritz Hamm, Hans Kamp and Michiel van Lambalgen (in what follows abbreviated as ‘HKL’) is a very rich one. Not only does it contain a wealth of empirical and formal insights concerning the analysis of tense and aspect, planning and causality, and other phenomena, it also contains some penetrating remarks concerning the scope and method of semantic theory. It is the latter aspect of the paper that I want to make a few comments on in what (...) follows. (shrink)

What factors are involved in the resolution of scientific disputes? What factors make the resolution of such disputes rational? The traditional view confers an important role on observation statements that are shared by proponents of competing theories. Rival theories make incompatible (sometimes contradictory) observational predictions about a particular situation, and the prediction made by one theory is borne out while the prediction made by the other is not. Paul Feyerabend, Thomas Kuhn, and Paul Churchland have called into question this account (...) of theory-resolution. According to these philosophers, substantially different and competing scientific theories are semantically incommensurable: those theories do not share a common observation language. Two charges have been leveled against the semantic incommensurability theories. The first is that it ignores that some semantic features of observational terms (e.g., their reference) can be expressed by proponents of competing theories. The second is that the semantic incommensurability thesis is self-defeating. In this paper I will argue that both of these charges are true but not for the reasons usually given. (shrink)

The “problem of intentionality” from the vantage point of a representational understanding of mind is explaining what thoughts and beliefs are and how they guide behaviour. From an anti-representationalist perspective, on the other hand, on which cognition itself is taken to be a kind of action, intentionality is a capacity to engage in behaviour that is meaningfully directed toward or about some situation. That these are not in fact competing insights is obscured by the representational/anti-representational framing of the debate. This (...) paper begins the work of shifting the conversation in two ways: (1) by arguing that it is the commitment to internal representations, not the acknowledgement of a role for representation per se, that is problematic; and, (2) by describing an alternative, externalist, representational approach that draws on extended, embodied, enactive insights. (shrink)

This essay makes the case for, in the phrase of Angelika Kratzer, packing the fruits of the study of rational decision-making into our semantics for deontic modals—specifically, for parametrizing the truth-condition of a deontic modal to things like decision problems and decision theories. Then it knocks it down. While the fundamental relation of the semantic theory must relate deontic modals to things like decision problems and theories, this semantic relation cannot be intelligibly understood as representing the conditions under (...) which a deontic modal is true. Rather it represents the conditions under which it is accepted by a semantically competent agent. This in turn motivates a reorientation of the whole of semantic theorizing, away from the truth-conditional paradigm, toward a form of Expressivism. (shrink)

John McDowell has often emphasized the fact that the use of langauge is a rational enterprise. In this paper, I explore the sense in which this is so, arguing that our use of language depends upon our consciously knowing what our words mean. I call this a 'cognitive conception of semantic competence'. The paper also contains a close analysis of the phenomenon of implicature and some suggestions about how it should and should not be understood.

In this paper I argue that the idiosyncrasy of linguistic competence fosters semantic conceptions in which meanings are taken for granted, such as the one that Quine calls ‘uncritical semantics’ or ‘the myth of the museum’. This is due to the degree of automaticity in the use of language which is needed for fluent conversation. Indeed, fluent conversation requires that we speakers instinctively associate each word or sentence with its meaning (or linguistic use), and instinctively resort to the (...) conceptual repertoire of our language, without calling into question that the meaning of a particular word, or the conceptual repertoire of our language, could have been different than they are. This habit of taking meanings for granted, inherent to our linguistic ability, sometimes interferes with our semantic research, hampering it. In order to illustrate this problem, I pinpoint four places in Quine’s work where, despite his acknowledged analytical rigour, and despite his congenital aversion to the habit of taking meanings for granted, he himself appears to slip into this habit, inadvertently. (shrink)

According to Emma Borg, minimalism is (roughly) the view that natural language sentences have truth conditions, and that these truth conditions are fully determined by syntactic structure and lexical content. A principal motivation for her brand of minimalism is that it coheres well with the popular view that semantic competence is underpinned by the cognition of a minimal semantic theory. In this paper, I argue that the liar paradox presents a serious problem for this principal motivation. Two (...) lines of response to the problem are discussed, and difficulties facing those responses are raised. I close by issuing a challenge: to construe the principal motivation for BM in such a way so as to avoid the problem of paradox. (shrink)

In this article, I defend an account of linguistic comprehension on which meaning is not cognized, or on which we do not tacitly know our language's semantics. On this view, sentence comprehension is explained instead by our capacity to translate sentences into the language of thought. I explain how this view can explain our capacity to correctly interpret novel utterances, and then I defend it against several standing objections.

The linguistic meaning of a word in a language is what fully competent speakers of the language have a grasp of merely in virtue of their semantic competence. The meanings of words sometimes change over time. ‘Meat’ used to mean ‘solid food’, but now means ‘animal flesh eaten as food’. This type of meaning change comes with change of topic, what we are talking about. Many people interested in conceptual engineering have claimed that there is also meaning change (...) where topic is retained. For example, they claim that the meanings of ‘fish’ and ‘pasta’ have undergone such change, and that the meaning of ‘marriage’ would change this way after gay marriages become legal and widely accepted. In this paper, I relate two sets of relatively independent literatures: mainstream philosophy of language and conceptual engineering to argue that on a plausible and widely accepted Minimalist view of meaning that is part and parcel of anti‐descriptivism, none of the above sorts of cases involve meaning change with topic retention. I do this by showing how to distinguish minimalism about meaning from the related theses of externalism and anti‐individualism about intension and how to separate meaning from intension in a way that allows meaning and topic to remain the same despite changes in intension. The larger lesson is that much like we should not disregard the boundary between the narrowly meaning‐related (‘semantics’) and the more broadly communication‐related (‘pragmatics’), and we should not disregard the boundary between the former and the more broadly thought‐related, conceptual or cognitive (‘cognition’). (shrink)

I discuss some problems faced by the meaning‐inconsistency view on the liar and sorites paradoxes which I have elsewhere defended. Most of the discussion is devoted to the question of what a defender of the meaning‐inconsistency view should say about semantic competence.

Competing theories on the semantics of group pejorative terms (also known as‘slurs’)comprise both advocates and opponents to the Identity Thesis (IT), according to whichthese terms and their neutral counterparts do not differ in semantic value. In the oppo-nents’camp, Christopher Hom has offered an argument based on substitution of slurs andneutral counterparts that both supports his semanticist approach and cast doubts on all IT-based approaches to slurs. We aim to point to a dilemma triggered by this argument based on evidence (...) showing that substitution of some words (including but not restricted to slurs)for non-problematically co-referential pairs may fail to preserve truth-values in some linguistics contexts. (shrink)

It is commonly held that there are internal links between understanding and assent such that being semantically competent with an expression requires accepting certain sentences as true. The paper discusses a recent challenge to this conception of semantic competence, posed by Timothy Williamson (2007). According to Williamson there are no understanding-assent links of the suggested sort, no internal connection between semantic competence and belief. I suggest that Williamson is quite right to question the claim that being (...) semantically competent with an expression e requires accepting a certain sentence S as true. However, Williamson does not merely wish to reject this version of the understanding-assent view, but the very idea that the connection with belief provides constitutive constraints on linguistic understanding and concept possession. This further move, I argue, is very problematic. Giving a plausible account of semantic competence requires accepting that there are constitutive links between understanding and assent, although these links should be construed holistically rather than atomistically. (shrink)

This paper motivates two bases for ascribing propositional semantic knowledge (or something knowledgelike): first, because it’s necessary to rationalize linguistic action; and, second, because it’s part of an empirical theory that would explain various aspects of linguistic behavior. The semantic knowledge ascribed on these two bases seems to differ in content, epistemic status, and cognitive role. This raises the question: how are they related, if at all? The bulk of the paper addresses this question. It distinguishes a variety (...) of answers and their varying philosophical and empirical commitments. (shrink)

Veltman’s test semantics and developments thereof reject the canon about semantic contents and attitude ascriptions in favor of dynamic alternatives. According to these theories the semantic content of a sentence is not a proposition, but a context change potential (CCP). Similarly, beliefs are not taken to be relations between agents and propositions, but agents and CCPs. These deviations from the canon come at the cost of an elegant explanation about the correctness of belief. Standardly, it is taken that (...) the content of a belief is correct just in case the content of that belief is true. The proponent of the test semantics cannot appeal to this explanation since they hold that certain contents, namely epistemically modalized contents, do not express propositions, and are neither true nor false. After motivating the need for an account of the correctness of belief within the dynamic framework, I argue that the few extant accounts in the literature are unsatisfactory. A novel account is proposed that avoids the problems of competing views. (shrink)

As an empirical inquiry into the nature of meaning, semantics must rely on data. Unfortunately, the primary data to which philosophers and linguists have traditionally appealed—judgments on the truth and falsity of sentences—have long been known to vary widely between competent speakers in a number of interesting cases. The present article constitutes an experiment in how to obtain some more consistent data for the enterprise of semantics. Specifically, it argues from some widely accepted Gricean premises to the conclusion that judgments (...) on lying are semantically relevant. It then endeavors to show how, assuming the relevance of such judgments, we can use them to generate a useful, widely acceptable test for semantic content. (shrink)

What considerations place genuine constraints on an adequate semantics for normative and evaluative expressions? Linguists recognize facts about ordinary uses of such expressions and competent speakers’ judgments about which uses are appropriate. The contemporary literature reflects the widespread assumption that linguists don’t rely upon an additional source of data—competent speakers’ judgments about possible disagreement with hypothetical speech communities. We have several good reasons to think that such judgments are not probative for semantic theorizing. Therefore, we should accord these judgments (...) no probative value for the development of a semantics for our moral terms. Such judgments can no longer be presumed to put pressure on theories according to which our moral expressions share a semantics with ordinary, descriptive terms. Many rivals to pure, Descriptivist theories count among their advantages the ability to accommodate these judgments. If these judgments have no probative value, such theories lose an important source of support. (shrink)

Floridi and Taddeo propose a condition of “zero semantic commitment” for solutions to the grounding problem, and a solution to it. I argue briefly that their condition cannot be fulfilled, not even by their own solution. After a look at Luc Steels' very different competing suggestion, I suggest that we need to re-think what the problem is and what role the ‘goals’ in a system play in formulating the problem. On the basis of a proper understanding of computing, I (...) come to the conclusion that the only sensible ground-ing problem is how we can explain and re-produce the behavioral ability and function of meaning in artificial computational agents. (shrink)

In philosophy of language, a distinction has been proposed by Diego Marconi between two aspects of lexical competence, i.e. referential and inferential competence. The former accounts for the relation-ship of words to the world, the latter for the relationship of words among themselves. The aim of the pa-per is to offer a critical discussion of the kind of formalisms and computational techniques that can be used in Artificial Intelligence to model the two aspects of lexical competence, and (...) of the main difficulties related to the use of these computational techniques. The first conclusion of our discussion is that the dis-tinction between inferential and referential semantics is instantiated in the literature of Artificial Intelli-gence by the distinction between symbolic and connectionist approaches. The second conclusion of our discussion is that the modelling of lexical competence needs the advent of hybrid models integrating symbolic and connectionist frameworks. Our hypothesis is that Conceptual Spaces, a framework devel-oped by Gärdenfors more than fifteen years ago, can offer a lingua franca that allows to unify and gener-alize many aspects of the representational approaches mentioned above and to integrate “inferential” (=symbolic) and “referential” (=connectionist) computational approaches on common ground. (shrink)

According to non-cognitivism, moral sentences and judgements do not aim to represent how things morally are. This paper presents an empirical argument against this view. We begin by showing that non-cognitivism entails the prediction that after some reflection competent ordinary speakers’ semantic intuitions favor that moral sentences and judgements do not aim to represent how things morally are. At first sight, this prediction may seem to have been confirmed by previous research on folk metaethics. However, a number of methodological (...) worries lead us to doubt this interpretation. We, therefore, conducted a psychological study that alleviates these worries as far as possible. It turned out that competent ordinary speakers’ reflective semantic intuitions dominantly fail to favor that moral sentences do not aim to represent how things morally are. This challenge to non-cognitivism is defended and supplemented by considering deflationary theories of moral truth and middle ground theories in the cognitivism/non-cognitivism debate. (shrink)

According to the accepted translation-failure interpretation, the problem of incommensurability involves the nature of the meaning-referential relation between scientific languages. The incommensurability thesis is that some competing scientific languages are mutually untranslatable due to the radical variance of meaning or/and reference of the terms they employ. I argue that this interpretation faces many difficulties and cannot give us a tenable, coherent, and integrated notion of incommensurability. It has to be rejected. ;On the basis of two case studies, I find that (...) the confrontations between many classical incommensurable languages are not confrontations between two untranslatable languages with different distribution of truth values, but rather the confrontations between incompatible fundamental presuppositions at the ontological level. We can always identify a truth-value gap between two incommensurable languages. Such a truth-value gap indicates a communication breakdown between the two language communities on the one hand, and is caused by the incompatible fundamental presuppositions underlying them on the other. ;I thereby identify the truth-value functional relationship between sentences, instead of the meaning-referential relationship between terms, as the dominant semantic relation between two incommensurable languages. According to my presuppositional interpretation of incommensurability, the real secret of incommensurability lies in the ontological setup of two competing presuppositional languages. When two presuppositional languages with incompatible factual commitments encounter with each other, the confrontation leads to a truth-value gap, and consequently a communication breakdown between them. Formally put, two scientific languages are incommensurable when core sentences of one language, which have truth values when considered within its own context, lack truth values when considered within the context of the other due to an ontological gap between them. ;The presuppositional interpretation makes many significant contributions to the discussion of the issue of incommensurability and the related metaphysical and epistemological issues: It confirms the existence of the phenomenon of incommensurability and makes it metaphysically and epistemologically significant. It establishes the tenability and integrity of the notion of incommensurability. It avoids many alleged unattractive epistemological and metaphysical consequences of the translation-failure interpretation. (shrink)

The meaning of morpheme (a minimal unit of linguistic significance) cannot diverge from what it is taken to mean. But the meaning of a complex expression can diverge without limit from what it is taken to mean, given that the meaning of such an expression is a logical consequence of the meanings of its parts, coupled with the fact that people are not infallible ratiocinators. Nonetheless, given Chomsky’s distinction between competence (ability) and performance (ability to deploy ability), what a (...) complex expression means does, after a fashion, coincide with what it is taken to mean: to the extent that speaker-performance approximates to speaker-competence---i.e. to the extent that people are able to operationalize their linguistic competence---what speakers take complex expressions to mean coincides with they in fact mean; and herein lies an answer to the question “what is linguistic meaning?” that holds with respect to both simple and complex expressions. (shrink)

In this paper, a novel experimental task is developed for testing the highly influential, but experimentally underexplored, possible worlds account of conditionals (Stalnaker, 1968; Lewis, 1973). In Experiment 1, this new task is used to test both indicative and subjunctive conditionals. For indicative conditionals, five competing truth tables are compared, including the previously untested, multi-dimensional possible worlds semantics of Bradley (2012). In Experiment 2, these results are replicated and it is shown that they cannot be accounted for by an alternative (...) hypothesis proposed by our reviewers. In Experiment 3, individual variation in truth assignments of indicative conditionals is investigated via Bayesian mixture models that classify participants as following one of several competing models. As a novelty of this study, it is found that a possible worlds semantics of Lewis and Stalnaker is capable of accounting for participants’ aggregate truth value assignments in this task. Applied to indicative conditionals, we show across three experiments, that the theory both captures participants’ truth values at the aggregate level (Experiments 1 and 2) and that it makes up the largest subgroup in the analysis of individual variation in our experimental paradigm (Experiment 3). (shrink)

Syntax has to do with rules that constrain how words can combine to make acceptable sentences. Semantics (Frege and Russell) concerns the meaning of words and sentences, and pragmatics (Austin and Grice) has to do with the context bound use of meaning. We can hence distinguish between three competing principles of translation: S—translation preserves the syntax of an original text (ST) in the translation (TT); M—translation preserves the meaning of an ST in a TT; and P—translation preserves the pragmatics of (...) an ST in a TT. A prominent form of P is functionalism defended by linguists and translation theorists (J.R. Firth, Eugene Nida, Susan Bassnett and many others) and historically was defended by philosophers (Russell, Ogden and Richard) but abandoned by philosophers and criticized by Wittgenstein. If we adopt M, then a TT will always say exactly what the ST says, and hence all subsequent TTs, even alternative ones produced via M, will be consistent with each other. But if we adopt P, in contrast, we have no reason to believe that the TTs will say what the ST does, and moreover they can contradict each other. If such contradictory translations are produced on the basis of the totality of empirical evidence, it results in what Quine called the indeterminacy of translation. Yet, P is not easy to reject. In many cases, translation in accordance with M where the meaning to be preserved is linguistic results in TTs that are failures. In contrast to a language focused approach to semantics, I close by following a lead in the translation theory literature of identifying text-types (genres) as a tool for identifying translatable content in an ST. To individuate text-types I identify them with disciplines, as elucidated by the 2nd century Patañjali’s Yoga Sūtra. This allows for the definition of textual meaning as the discipline relative pragmatics of an ST and further for translation to proceed by way of M, while taking the intuitions that motivate P seriously. Translations that preserve textual meaning will not only have the same meaning as each other but will be pragmatically felicitous. (shrink)

There is a standard quantificational view of generic sentences according to which they have a tripartite logical form involving a phonologically null generic operator called 'Gen'. Recently, a number of theorists have questioned the standard view and revived a competing proposal according to which generics involve the predication of properties to kinds. This paper offers a novel argument against the kind-predication approach on the basis of the invalidity of Generic Excluded Middle, a principle according to which any sentence of the (...) form ⌜Either Fs are G or Fs are not G⌝ is true. I argue that the kind-predication approach erroneously predicts that GEM is valid, and that it can only avoid this conclusion by either collapsing into a form of the quantificational analysis or otherwise garnering unpalatable metaphysical commitments. I also show that, while the quantificational approach does not validate GEM as a matter of logical form, the principle may be validated on certain semantic analyses of the generic operator, and so, such theories should be rejected. (shrink)

The necessity of origin suggests that a person’s identity is determined by the particular pair of gametes from which the person originated. An implication is that speculative scenarios concerning how we might otherwise have been had our gametic origins been different are dismissed as being metaphysically impossible. Given, however, that many of these speculations are intelligible and commonplace in the discourses of competent speakers, it is overhasty to dismiss them as mistakes. This paper offers a way of understanding these speculations (...) that does not commit them to incoherence but aims to make the best sense of what they are expressing. Using the philosophical framework of two‐dimensional semantics, it proposes that the speculative scenarios are best analysed as epistemic possibilities, rather than as metaphysical possibilities. It then explores some implications of this analysis for the ethical challenges associated with the non‐identity problem. (shrink)

Against the received translation-failure interpretation, this book presents a presuppositional interpretation of incommensurability, that is, the thesis of incommensurability as cross-language communication breakdown due to the incompatible metaphysical presuppositions underlying two competing presuppositional languages, such as scientific languages. This semantically sound, epistemologically well-established, and metaphysically profound interpretation not only affirms the tenability of the notion of incommensurability and confirms the reality of the phenomenon of incommensurability, but also makes some significant contributions to the discussion of many related issues, such as (...) the notion of conceptual schemes, the notion of truth-value status and truth-value conditions, and the issue of cross-language understanding and communication. (shrink)

The Galilean explanatory style is characterized by the search for the underlying structure of phenomena, the positing of "deep" explanatory principles, and a view of the relation between theory and data, on which the search for "crucial data" is of primary importance. In this paper, I trace the dynamics of adopting the Galilean style, focusing on the science of episodic memory. I argue that memory systems, such as episodic and semantic memory, were posited as underlying competences producing the observable (...) phenomena of memory. Considered in idealized isolation from other systems, episodic memory was taken to underlay the ability of individuals to remember events from their personal past. Yet, in reality, memory systems regularly interact, standing in many-to-many relations to actual memory tasks and experiences. Upon this backdrop, I explore a puzzle about the increasing prominence of the notion of autonoetic consciousness in Tulving's theory of episodic memory. I argue that, contrary to widespread belief, the prominence is not best explained by the purported essential link between autonoetic consciousness and episodic memory. Rather, it is explained by the fact that autonoetic consciousness, hypothesized to uniquely accompany episodic retrieval, was considered a source of crucial data, predictable only from theories positing a functionally distinct episodic memory system. However, with the emergence of a new generation of theories, positing wider memory systems for remembering and imagination, the question of the relation between episodic memory and autonoetic consciousness has been reopened. This creates a pressing need for de-idealization, triggering a new search for crucial data. (shrink)

Paul Pietroski has developed a powerful minimalist and internalist alternative to standard compositional semantics, where meanings are identified with instructions to fetch or assemble human concepts in specific ways. In particular, there appears to be no need for Fregean Function Application, as natural language composition only involves processes of combining monadic or dyadic concepts, and Pietroski’s theory can then, allegedly, avoid both singular reference and truth conditions. He also has a negative agenda, purporting to show, roughly, that the vocabulary of (...) standard truth conditional semantics is far too powerful to plausibly describe the linguistic competence of mere human minds. In this paper, I explain some of the basics of Pietroski’s compositional semantics and argue that his major objection to standard compositionality is inconclusive, because a similar argument can be mounted against his own minimalist theory. I argue that we need a clear distinction between the language of the theorist---theoretical notation---and the language whose nature we are trying to explain. The theoretical notation should in fact be as expressively powerful as possible. It does not follow that the notation cannot be used to explain mere human linguistic competence, even if human minds are limited in various ways. (shrink)

Most philosophers are familiar with the metaphysical puzzle of the statue and the clay. A sculptor begins with some clay, eventually sculpting a statue from it. Are the clay and the statue one and the same thing? Apparently not, since they have different properties. For example, the clay could survive being squashed, but the statue could not. The statue is recently formed, though the clay is not, etc. Godehart Link 1983’s highly influential analysis of the count/mass distinction recommends that English (...) draws a distinction between uncountable “stuff” and countable “things”. There are two mereological relations, related in specific ways. Our primary question here is whether an empirically adequate account of the mass/count distinction really does require distinguishing “things” from “stuff”, and thus postulating two corresponding mereological relations, or if instead positing only one sort of entity and corresponding mereological relation is sufficient, as other semantic theories would have it. This question is meant to be one of what we call natural language mereology. We are asking about the mereological commitments of English, or perhaps competent speakers of English, and not about ultimate reality as such. There is no pretense that we will definitively solve the metaphysical puzzle of the statue and clay. (shrink)

Extended abstract presented at the AISC 2018 Conference, 15th International Conference of the Italian Association of Cognitive Science, Pavia.

According to the inferential view of language comprehension, we hear a speaker’s utterance and infer what was said, drawing on our competence in the syntax and semantics of the language together with background information. On the alternative perceptual view, fluent speakers have a non-inferential capacity to perceive the content of speech. On this view, when we hear a speaker’s utterance, the experience confers some degree of justification on our beliefs about what was said in the absence of defeaters. So, (...) in the absence of defeaters, we can come to know what was said merely on the basis of hearing the utterance. Several arguments have been offered against a pure perceptual view of language comprehension, among others, arguments pointing to its alleged difficulties accounting for homophones and the context-sensitivity of ordinary language. After responding to the arguments against the perceptual view of language comprehension, I provide a new argument in favor of the perceptual view by looking closer at the dependence of the justificatory qualities of experience on the notion of a defeater as well as the perceptual nature of language learning and language processing. (shrink)