A prominent solution to the 'symmetry problem' allows implicatures to be computed from simple but... more A prominent solution to the 'symmetry problem' allows implicatures to be computed from simple but not from complex alternatives ('COMPLEXITY'; Katzir 2007). Recently Schwarz and Wagner (2024) have proposed a different mechanism for symmetry breaking ('BLOCKING'), arguing that it can, but COMPLEXITY cannot, account for cases of so-called 'simplex threats' in which the simple alternative is available but the expected implicature is unattested. This note provides a defense of COMPLEXITY. We show that it explains simplex threats once coupled with constraints on questions ('Partition by Exhaustification'; Fox 2019, 2020) and on assertability of sentences with contextually equivalent alternatives ('Fatal Competition'; Magri 2009, Bar-Lev and Fox 2023). We furthermore point out (following Schmitt and Haslinger 2025) that BLOCKING makes a wrong prediction for some cases.
This note explores Identity, the thesis that well-formedness coincides with meaningfulness in nat... more This note explores Identity, the thesis that well-formedness coincides with meaningfulness in natural language. Threats to Identity are posed by claims from both linguists and philosophers which include the following: (A) grammar can deceive us into thinking of a meaningless expression that it is meaningful; (B) only sentences are meaningful; (C) trivialities are ill-formed. I critically examine each of these claims and suggest responses to them which support Identity. I also discuss how Wittgenstein's Picture Theory of Language provides a way to reconcile Identity with empirical data that motivate (C).
To appear in the Proceedings of the 46th International Wittgenstein Symposium
Proposition 4.025 of the Tractatus describes translation as replacing each word in the original l... more Proposition 4.025 of the Tractatus describes translation as replacing each word in the original language text with its target language counterpart. This requires a one-one correspondence between vocabularies of the two languages, which never transpires in real life. I argue that Wittgenstein's remark should be read as articulating the theory of language which he promoted in the Tractatus: it tells us what translation looks like when that theory applies under ideal conditions. I then propose an account for the difference between translation in real life and translation as described by Wittgenstein. The account draws on the fact that language has a communicative function in addition to its representative function.
Vietnamese has a productive reduplication strategy where the reduplicant appears to the right of ... more Vietnamese has a productive reduplication strategy where the reduplicant appears to the right of the base and is segmentally identical to the base except that its last rhyme is -iec. In this note I attempt to account for some observations about -iec including the fact that it gives rise to ignorance inferences and is incompatible with classifiers. I propose that the semantics of iec parallels the pragmatics of disjunctions, and that the notion of similarity underlying the interpretation of -iec is contextual while that underlying the interpretation of classifiers is grammatical.
This note presents a series of contrasts pertaining to Vietnamese polar questions: (i) the subjec... more This note presents a series of contrasts pertaining to Vietnamese polar questions: (i) the subject can be definite but not quantificational; (ii) the subject can be plain but not only-focused; (iii) the modal adverb 'certainly' can follow but not precede verum focus. I argue that a monoclausal analysis, advocated in several previous works, will have difficulties accounting for these contrasts, and propose a bi-clausal analysis which explains them in a natural way. The explanation relies on the assumption of a general condition on questions, Partition by Exhaustification (PbE), in conjunction with some other independently motivated semantic and pragmatic constraints.
To appear in Translating and Interpreting the Tractatus, edited by Michael Beaney & David G. Stern, 2025
This note discusses translation in and of the Tractatus. Wittgenstein took translation to be word... more This note discusses translation in and of the Tractatus. Wittgenstein took translation to be wordfor-word. This view conflicts with reality. I propose to resolve the conflict by separating logical forms, which represent thoughts, from their externalization, which makes these thoughts perceptible to the senses. I also relate some problems I encountered when translating the Tractatus from German to Vietnamese.
The Connectives in Logic and Language. Lecture Notes in Computer Science, 2025
We propose a conservative analysis for conditional questions, i.e. those of the form if φ, Q wher... more We propose a conservative analysis for conditional questions, i.e. those of the form if φ, Q where φ expresses a proposition and Q a question. Our analysis retains the standard interpretation of if as a propositional operator and yields the right intuitions regarding the answers to these questions. Furthermore, we explain why and and or cannot embed questions the same way if does. We show how our account overcomes difficulties faced by previous theories, and discuss some open problems for future research.
This note presents a novel observation: exceptives do not tolerate cardinal determiners. It discu... more This note presents a novel observation: exceptives do not tolerate cardinal determiners. It discusses three analyses of exceptives that do not account for this observation: (i) von Fintel (1993) which takes exceptives to be modifers of predicates; Moltmann (1995) which takes exceptives to be modifiers of quantifiers; and (iii) Vostrikova (2021) which takes exceptives to be modifiers of clauses. It then proposes a slight modification of Vostrikova’s analysis which retains the virtues of the original and in addition accounts for the fact that exceptives do not tolerate cardinal determiners.
Conditional Questions (CQs), structures such as If it's raining, will Joanna leave? pose a puzzle... more Conditional Questions (CQs), structures such as If it's raining, will Joanna leave? pose a puzzle: they look like conditionals, but are interpreted like questions. Existing accounts in the literature all take the surface form of CQs at face value and treat them as matrix conditionals scoping over a question, and thus must employ sophisticated semantic machinery to produce a question meaning out of them (Isaacs and Rawlins 2008; Ciardelli et al. 2019). We develop and argue for an alternative theory on which CQs are, despite surface appearance, underlyingly matrix questions scoping over a conditional. The input to semantic interpretation is obtained as a result of syntactic reconstruction of the if-clause under the question-an independently needed mechanism. The theory does not necessitate any sophisticated semantics to handle CQs, and we show that it makes correct predictions with respect to diagnostics for reconstruction that surfacesyntax accounts don't make. We also argue that the extra semantic machinery employed by previous accounts is harmful in being at risk of over-generating unattested structures, a problem avoided by our account which only makes use of conservative semantics for CQs.
I consider two analyses of yes/no questions in Vietnamese. The 'monoclausal analysis' takes such ... more I consider two analyses of yes/no questions in Vietnamese. The 'monoclausal analysis' takes such questions to be of the form O(p), where O(p) = {p,¬p}, while the 'biclausal analysis' takes them to be of the form O(p)(q), where O(p)(q) = {p,q}. I argue in favor of the biclausal analysis on the basis of three observations: (i) subjects of yes/no questions cannot associate with only; (ii) subjects of yes/no questions cannot be quantifiers; and (iii) modal adverbs in yes/no questions can follow but not precede the polarity head. The argument relies crucially on the general requirement that answers to a question, once exhaustified, partition the context set.
Vietnamese has two types of NPIs, simple and complex, and two types of polar questions, yes/no qu... more Vietnamese has two types of NPIs, simple and complex, and two types of polar questions, yes/no questions and agreement questions. Simple NPIs can occur in both types of polar questions while complex NPIs can occur in yes/no but not in agreement questions. I propose an account for this fact using familiar ingredients of semantic and syntactic analyses. I then discuss some ways in which Vietnamese and English differ with respect to how distinctions in meaning align with distinctions in form.
The cross-linguistic research on "negative polarity items" (NPIs) not only reveals what contraint... more The cross-linguistic research on "negative polarity items" (NPIs) not only reveals what contraints are imposed by semantics on the output of syntax but also provides insights into how individual languages differ with respect to the way they satisfy these constraints. This note makes a small contribution to this enterprise: it discusses some differences in distribution between NPIs in English and their counterparts in Vietnamese. The discussion is preluded by a brief introduction to background concepts and assumptions. The Vietnamese data are presented as a challenge which motivates further thought and investigation. A sketch of an approach is provided at the end.
Forms of address must be prononimal in English but can be either pronominal or nominal in Vietnam... more Forms of address must be prononimal in English but can be either pronominal or nominal in Vietnamese. I propose to analyze this fact as a parametric difference: the two languages choose different ways to implement one and the same general principle of grammar. The analysis crucially relies on the hypothesis that some aspects of meaning which have traditionally been considered pragmatic are represented in the syntax.
I argue that there is tension in Wittgenstein’s position on trivialities (i.e. tautologies and co... more I argue that there is tension in Wittgenstein’s position on trivialities (i.e. tautologies and contradictions) in the Tractatus, as it contains the following claims: (A) sentences are pictures; (B) trivialties are not pictures; (C) trivialities are sentences. A and B follow from the “picture theory” of language which Wittgenstein proposes, while C contradicts it. I discuss a way to resolve this tension in light of Logicality, a hypothesis recently developed in linguistic research. Logicality states that trivialities are excluded by the grammar, and that under the right analysis sentences which look trivial are in fact contingent. The tools necessary to support Logicality, I submit, were not available to Wittgenstein at the time, which is what gives rise to his commitment to C. I end the paper by commenting on some points of contact between analytic philosophy and the generative enterprise in linguistics which are brought into relief by the discussion.
This paper discusses a set of observations, many of which are novel, concerning differences betwe... more This paper discusses a set of observations, many of which are novel, concerning differences between the adjectival modals "certain" and "possible" and their adverbial counterparts "certainly" and "possibly". It argues that the observations can be derived from a standard interpretation of "certain"/"certainly" as universal and "possible"/"possibly" as existential quantifiers over possible worlds, in conjunction with the hypothesis that the adjectives quantifiy over knowledge and the adverbs quantify over belief. The claims on which the argument relies include the following: (i) knowledge implies belief, (ii) agents have epistemic access to their belief, (iii) relevance is closed under speakers' belief, and (iv) commitment is pragmatically inconsistent with explicit denial of belief.
There are two schools of thoughts on exceptives. The "Fintelians" take exceptives to be modifiers... more There are two schools of thoughts on exceptives. The "Fintelians" take exceptives to be modifiers of the NP argument of the determiner, while the "Anti-Fintelians" take them to be something else. I present the observation that exceptives do not tolerate cardinal determiners. I then discuss the problem it poses for two Anti-Fintelian analyses and propose a Fintelian account. The main idea of the account is that exceptives introduce subdomain alternatives.
I present novel observations about iterated questions, i.e. questions about questions, and propos... more I present novel observations about iterated questions, i.e. questions about questions, and propose an analysis. The conclusions I argue for are the following: (i) speech acts are represented in the grammar; (ii) speech act recursion is possible but is limited to at most two levels; (ii) declarative questions are questions about an assertion act. I also show that assuming speech acts in the grammar can help systematize some puzzling differences between matrix and embedded sentences with respect to their pronunciation.
Linguistische Arbeitsberichte 96: Gisbert Fanselow’s Contributions to Syntactic Theory, 2024
The theory of chain linearization I propose in a number of works predicts a typology in which Ger... more The theory of chain linearization I propose in a number of works predicts a typology in which German would exemplify a type if some cases of incomplete category fronting in this language do not involve VP remnant movement. This is precisely what Gisbert Fanselow argues for in one of his papers. In this note, I present this argument in its dialectical background and respond to some issues which arise from it.
Polar questions in Vietnamese consist of an affirmative sentence followed by a negation particle.... more Polar questions in Vietnamese consist of an affirmative sentence followed by a negation particle. Modern Vietnamese has three negation particles, but only two can occur in this function. This note proposes an account for this gap. The account is premised on the analysis of questions as sets of alternatives, and draws on facts of diachronic change gleaned from historical texts.
Sự tồn tại của văn học hư cấu là một thách thức đối với quan niệm kinh điển về ngôn ngữ như một c... more Sự tồn tại của văn học hư cấu là một thách thức đối với quan niệm kinh điển về ngôn ngữ như một công cụ trao đổi thông tin. Về mặt khái niệm, cung cấp thông tin có nghĩa là mô tả thế giới thực. Một tác phẩm hư cấu được sáng tác và tiếp nhận trong ý thức hỗ tương giữa người nói, tức tác giả, và người nghe, tức độc giả, rằng nó không mô tả thế giới thực. Mặc dù vậy, ngôn ngữ trong tác phẩm hư cấu không vận hành như thể chức năng thông báo của nó đã được vô hiệu hoá. Bài này đưa ra một cách nhìn về tác phẩm hư cấu trong đó mâu thuẫn nói trên được giải toả. Theo cách nhìn này, tác phẩm hư cấu, về bản chất, vẫn mang tính tả thực. Nó không vẽ ra một thế giới không thực, mà nói cho chúng ta biết thế giới thực sẽ ra sao dưới một sự thay đổi tối thiểu. Bài viết cũng đưa ra một lập luận ngôn ngữ học để cho thấy cách ta tiếp nhận tác phẩm hư cấu có điểm chung với cách ta hiểu câu điều kiện, tức câu có hình thức "nếu p thì q".
Uploads
Papers by Tue Trinh