Tamara Dobler

Speech liberalism holds that the linguistic output of generative AI is continuous with human speech and should therefore be treated as meaningful in broadly the same way. Strong versions of this view take the fluent behaviour of chatbots to justify attributing genuine linguistic and cognitive agency, while nevertheless advocating caution when questions of moral or legal responsibility arise. I argue that this gives too much evidential weight to successful conversational interaction, where speaker intentions largely recede into the background. Once AI-generated speech contributes to serious harm, however, those intentions become central. More broadly, the paper challenges a way of theorising about AI speech that builds classifications of linguistic agency primarily from successful conversational interaction. I develop two independent but complementary arguments. The first is ethical–legal. I argue that classifying chatbot output as speech is not normatively neutral: within existing free-speech doctrine, it risks generating responsibility and liability gaps in cases of AI-mediated harm. The second is epistemic. I argue that the behavioural evidence supporting strong speech liberalism is systematically distorted by anthropomorphic bias, reinforced by AI design practices. The paper concludes that theories of AI speech should not be built exclusively from successful conversational exchanges and that, until a broader evidential case is made, a more conservative approach to AI speech is warranted in high-stakes contexts.

Should conversational AI be treated as constitutionally protected speech? Recent litigation, most notably Garcia v. Character Technologies, Inc., presents this question in a new form. While the plaintiff characterises Character.AI as a commercially distributed product whose design gives rise to ordinary principles of safety and liability, the defendants argue that chatbot outputs constitute protected speech under the First Amendment. We argue that Garcia presents a problem of legal underdetermination between competing concepts of speech rather than a dispute resolvable by analogy alone. Drawing on conceptual ethics, teleological approaches to legal interpretation, and recent work on legal construction, we develop a normative methodology for choosing between competing legal concepts. Applied to conversational AI, this methodology requires asking which concept of speech best advances the constitutional purposes of the First Amendment. We argue that those purposes include not only expressive freedom, listeners’ interests, and innovation, but also responsibility and legal redress, and that they do not all support the same concept of speech. Extending constitutional speech coverage to conversational AI places pressure on doctrines governing harmful speech that continue to presuppose intentional human speakers, while offering comparatively little additional protection for the constitutional values traditionally served by the First Amendment. We therefore argue that courts have good reason to prefer a concept of speech that does not initially treat AI-generated outputs as constitutionally protected speech but instead allows disputes concerning conversational AI to be addressed through product law.

ABSTRACT This paper examines Wittgenstein’s rule-following argument in relation to Chomskyan generative grammar, focusing on selectional restrictions. While Wittgenstein views meaning determining rules as context-dependent and socially embedded, Chomsky treats linguistic rules as computational and internal to the mind. I argue that Wittgenstein’s critique does not apply to most syntactic rules but poses a challenge for selectional restrictions, which rely on world knowledge. If lexical features depend on external factors, selectional restrictions cannot be purely internalist. To preserve internalist semantics, I propose a modification of Pietroski’s (2018) model in which meaning rules function automatically, eliminating interpretative choice. This approach reconciles internalism with linguistic flexibility while maintaining its explanatory power within a generative framework.

On the most common interpretation of occasion-sensitivity what varies cross-contextually is the truth-conditional content of representations. Jerry Fodor argues that when extended to mental representation this view has some problematic consequences. In this paper I outline an approach to occasion-sensitivity which circumvents Fodor’s objections but still maintains that the aspect of thought that guides deliberation and action is occasion-sensitive. On the proposed view, what varies cross-contextually are not truth conditions but rather the conditions for accepting a representation as true relative to a practical goal that is pursued on an occasion. I show that although the proposal entails an error theory this theory is not problematic since it is meant to compensate for the over-generating nature of semantic competence, namely, the fact that not all of the representation’s truth-makers are conducive to a given contextually salient goal.

Two Conceptions of Wittgenstein's ContextualismHow should we understand Wittgenstein's proposals that "the meaning of a word is its use in the language" (Wittgenstein 1953, §43) and that a name only has a meaning in a language-game (ibid. §49)? Are they incompatible with occasion-invariant semantics? In this paper I present two leading interpretations of Wittgenstein's contextualism: James Conant's meaning-eliminativism (ME) and Charles Travis's meaning-underdetermination (MU). I argue that, even though these two interpretations are very similar, the latter gives a more nuanced account of Wittgenstein's contextualism which does not involve a commitment to the claim that words have no meaning outside immediate contexts of use.

Peter Hacker defends an interpretation of the later Wittgenstein's notion of grammar, according to which the inherently general grammatical rules are sufficient for sense-determination. My aim is to show that this interpretation fails to account for an important contextualist shift in Wittgenstein's views on sense-determination. I argue that Hacker attributes to the later Wittgenstein a rule-based, combinatorial account of sense, which Wittgenstein puts forward in the Tractatus. I propose that this is not how we should interpret the later Wittgenstein because he insists that particular circumstances of use play a necessary role in determining the boundary between sense and nonsense.

In this paper I propose a partition semantics for sentences containing objective predicates that takes into account the phenomenon of occasion-sensitivity associated with so-called Travis cases. The key idea is that the set of worlds in which a sentence is true has a more complex structure as a result of different ways in which it is made true. Different ways may have different capacities to support the attainment of a contextually salient domain goal. I suggest that goal-conduciveness decides whether some utterance of a sentence is accepted as true on a particular occasion at a given world. The utterance will not be accepted as true at a world which belongs to a truth-maker which is less conducive to a contextually salient goal than other truth-makers. Finally, the proposed occasion-sensitive semantics is applied to some cases of disagreement and cancellability.

It is a dominant view in the philosophical literature on the later Wittgenstein that Chomsky’s approach to the investigation of natural language stands in stark contrast to Wittgenstein’s, and that their respective conceptions of language and linguistic understanding are irreconcilable. The aim in this paper is to show that this view is largely incorrect and that the two approaches to language and its use are indeed compatible, notwithstanding their distinctive foci of interest. The author argues that there is a signifi cant correspondence in at least five different areas of their work, and that once we pay attention to these there will be lesstemptation to see Wittgenstein and Chomsky as enemies.

In this article, I propose an inquisitive approach to semantic underdetermination using the model of issue resolution to describe how occasion meanings are determined in the process of pragmatic enrichment. I appeal to “Travis cases” to motivate the account of semantic underdetermination based on alternative ways for some objectato be F. When interpreting a sentence, we look how to narrow down the space of metalinguistic alternatives and achieve the state where a metalinguistic issue is resolved. I suggest that, in doing this, we primarily rely on contextual domain goals, preferring those alternatives which are conducive to the goal.

The paper argues for a model of polysemy based on the blueprint offered by Paul Pietroski whereby the meaning of a lexical item is an instruction to fetch a concept from an address. We show that the bare idea of fetching admits of a deep construal, where a concept is fetched, and a shallow construal, where the instruction merely links a lexical item to an address without automatically retrieving anything from the address; retrieval only occurs when the item is embedded within a syntactic structure. We offer considerations in favour of the shallow construal, which is consistent with a root conception of lexical items.

Building on Wittgenstein’s ideas, I defend a brand of pluralism that associates words with conceptual families and appeals to this notion in the course of philosophical problem solving. I argue that certain problems that the received view of conceptual engineering (‘improvement by replacement’) faces can be more easily overcome if we adopt a pluralist perspective. I show that the proposed approach can circumvent the problem of topic discontinuity, whilst also avoiding the threat of trivialisation, since it can easily accommodate both coarse-grained and fine-grained topics. Finally, I argue that my variant of pluralism is compatible with de novo conceptual engineering insofar as it allows that the paradigm associated with a word may shift overtime thus creating an opportunity for new candidate members to join a conceptual family.

In his post-Tractatus work on natural language use, Wittgenstein defended the notion of what he dubbed the autonomy of grammar. According to this thought, grammar - or semantics, in a more recent idiom - is essentially autonomous from metaphysical considerations, and is not answerable to the nature of things. The argument has several related incarnations in Wittgenstein?s post-Tractatus writings, and has given rise to a number of important insights, both critical and constructive. In this paper I will argue for a potential connection between Wittgenstein?s autonomy argument and some more recent internalist arguments for the autonomy of semantics. My main motivation for establishing this connection comes from the fact that the later Wittgenstein?s comments on grammar and meaning stand in opposition to some of the core assumptions of semantic externalism. nema