Department of Linguistics
University of Konstanz
tkoev [at] scarletmail [dot] rutgers [dot] edu
I obtained my PhD in linguistics from Rutgers University in 2013. After that, I was a Postdoctoral Researcher at the University of Stuttgart and the University of Düsseldorf. Currently, I am an Emmy Noether Fellow at the University of Konstanz. In the fall semester of 2018, I was a Visiting Scholar at MIT. I will be visiting Stanford in the Autumn quarter of 2022.
Here is my full Curriculum Vitae.
My research program is grounded in formal semantics and pragmatics and is informed by my interests in experimental linguistics, dynamic semantics, Bayesian confirmation theory, formal syntax, and Slavic and Germanic linguistics. Broadly speaking, I am interested in how semantic content is attributed to rational agents and to what degree. The attribution problem is about how different meaning dimensions (entailment, presupposition, implicature) and perspectives interact with each other. The degree problem is about what evidence is available in the context. These two problems have guided my research on parentheticality, evidentiality, modality, gradability, question bias, indefiniteness, and adverbial modification. The methodologies I employ in order to gain insights into these phenomena include traditional and experimental techniques of data collection as well as formal tools from logic and probability theory. My far-reaching goal is to establish a predictive framework of the various factors (operator scope, projectivity, evidence source, information status, modal force, etc.) that anchor semantic content, thus refining our understanding of how different types of content are integrated into a coherent whole.
2022. Parenthetical Meaning. Oxford Studies in Semantics and Pragmatics. OUP. [book] [chapter 1]
This book investigates the semantics and pragmatics of a representative sample of parenthetical constructions. These constructions are argued to fall into two major classes: pure vs. impure. Pure parentheticals comment on some part of the descriptive content of the root sentence but are otherwise relatively independent of it. Impure parentheticals modify components of the illocutionary force and affect the felicity or the truth of the root sentence. The book studies parentheticals from three theoretical viewpoints: illocutionary effects, scopal properties, and discourse status. It establishes and explicates the notion of parenthetical meaning in a formally precise and predictive dynamic semantic model. As a result, parentheticality is brought to bear on fundamental linguistic phenomena such as entailment and presupposition, binding and anaphora, evidentiality and modality, illocutionary force, and polarity.
(Copyright restrictions prevent me from posting online the entire book but the full pre-publication draft is available upon request.)
2021. "Parentheticality, Assertion Strength, and Polarity." Linguistics and Philosophy 44: 113-140. doi 10.1007/s10988-019-09285-4. [preprint] [paper]
Sentences with slifting parentheticals (such as The dean greeted the secretary, Jill said; Ross 1973) grammaticalize an intriguing interaction between truth-conditional meaning and speech act function. In such sentences, the assertion strength of the slifted clause (the non-parenthetical part of the sentence) is modulated by the parenthetical, which provides evidential support (Urmson 1952; Asher 2000; Rooryck 2001; Jayez and Rossari 2004; Davis et al. 2007; Simons 2007; Murray 2014; Maier and Bary 2015; AnderBois 2016; Hunter 2016). Starting with the idea that assertability comes in degrees (Lewis 1976; Davis et al. 2007), this paper develops a probabilistic update model that captures the role of parentheticality as a language tool for qualifying commitments. A crucial role here is played by the rule of Jeffrey conditionalization (Jeffrey 1990), which factors in the uncertainty of the parenthetical information itself. The model also derives certain effects of parenthetical modification not found in regular embedding constructions, including the fact that slifting parentheticals are limited to creating upward-entailing environments.
2018. "Notions of At-issueness." Language and Linguistics Compass 12(12): 1-16. doi 10.1111/lnc3.12306. [preprint] [paper]
Upon hearing the sentence Messi, who once scored a goal with his hand, won the Ballon d’Or, the addressee is likely to interpret the main clause as conveying the “main point” and view the appositive relative clause as contributing secondary information. The intuition that some part of the utterance conveys the main point has recently been discussed in formal semantics and pragmatics under the label of “at-issueness”. However, this label has been used in a variety of ways and there is often little clarity as to what is meant by it. This survey tries to clear things up by identifying and spelling out three specific notions of at-issueness, i.e. Q(uestion)-at-issueness, P(roposal)-at-issueness, and C(oherence)-at-issueness. After looking into what they say about similar kinds of data, I conclude that while these notions appear to capture facets of the same broad intuition, they are truly distinct. The paper also discusses potential connections of at-issueness to projection and commitment strength.
2017. "Quotational Indefinites." Natural Language and Linguistic Theory 35(2): 367-396. doi 10.1007/s11049-016-9344-x. [preprint] [paper]
This paper discusses one understudied variety of indefinites, which I call QUOTATIONAL INDEFINITES. Quotational indefinites are attested in languages like Bulgarian and German (see Cieschinger & Ebert 2011 on the latter), and are akin to Japanese wh-doublets (Sudo 2008; ms) and English placeholder words like whatshisface or so-and-so (cf. Clark & Gerrig 1990). The main claim of the paper is that quotational indefinites have a mixed semantics: they range over linguistic expressions yet make reference to both expressions and their denotations. These indefinites also require that the expressions they quantify over be of a certain type (a referential expression, a particular type of adverbial, etc.) and be uttered in a previous conversation. The formal analysis is framed in a two-dimensional semantics (Potts 2005; 2007) which cleanly separates the indefinite force and the reportative implications of sentences with quotational indefinites. This work uncovers important interactions between indefiniteness, quotation, and reportativity, and broadens our understanding of the typology of indefinites.
2017. "Evidentiality, Learning Events, and Spatiotemporal Distance: The View from Bulgarian." Journal of Semantics 34(1): 1-41. doi 10.1093/jos/ffv014. [preprint] [paper]
This paper defends the view that evidentiality need not be a semantic primitive but can rather be pragmatically derived from the spatiotemporal distance between the event described by the sentence and the "learning event", i.e. the event of the speaker acquiring the relevant evidence for her claim. While the empirical focus is on the Bulgarian evidential -l, this work adds to similar proposals about evidential markers in typologically unrelated languages (see Nikolaeva 1999, Fleck 2007, Speas 2010, Kalsang et al. 2013, Lee 2013). The view of evidentiality as a spatiotemporal distance undermines the claim that evidential sentences in Bulgarian have a modal force (see Izvorski 1997, Smirnova 2013) and correctly predicts that speakers are typically committed to the core proposition described by evidential sentences. The paper also discusses the not-at-issue discourse status and projection behavior of the evidential implication, suggesting that evidential meanings belong to the broader class of 'conventional implicatures', in the sense of Potts (2005). The formal proposal, couched in an update semantics, successfully captures not only the meaning of the evidential marker but also the discourse properties of evidential sentences in Bulgarian.
2015. "Experimental Evidence for the Truth Conditional Contribution and Shifting Information Status of Appositives" (with Kristen Syrett). Journal of Semantics 32(3): 525-577. doi: 10.1093/jos/ffu007. [preprint] [paper]
Appositive constructions (My friend Sophie, (who is) a classical violinist, performed a piece by Mozart) have stood at the center of debates concerning the range of possible meanings, and more specifically the status of not-at-issue entailments. However, it remains an open question what precisely their semantic and pragmatic contribution is to the sentence in which they appear. Here, we address this question head-on experimentally. We first investigate the information status of appositives and find that while nominal appositives (e.g. a classical violinist) and sentence-medial appositive relative clauses (e.g. who is a classical violinist) are largely not at issue, sentence-final appositive relative clauses can become at issue, as witnessed in their becoming the target of a direct rejection and being associated with subsequent questions. We then investigate the truth conditional contribution of appositives to sentences in which they appear, and find that whenever an appositive is false, participants judge the entire sentence False. Reaction times complement truth value ratings to demonstrate that this decision is largely automatic. We discuss possible reasons for the difference among appositive types and sentential positions, and propose that the pattern of results we observe and the strong similarity with conjunction can best be accounted for in a unidimensional semantics which treats appositives as dynamic conjuncts but which also relates linguistic form to the timing of making assertions in discourse.
2015. "An “Antiproviso Problem” for Appositive Relative Clauses." Snippets 29(4): 11-12. [paper]
The central observation of this note is that appositive relative clauses can trigger intermediate, i.e. weaker inferences. In the sentence below the appositive does not contribute to the at-issue content and projects in a weaker, conditionalized inference.
(i) If Jack buys a car, which will probably be a Volvo, his wife will be upset.
At-issue content: If Jack buys a car, his wife will be upset.
Projective inference: If Jack buys a car, it will probably be a Volvo.
I call this the “antiproviso problem” for appositive relative clauses, in reference to the “proviso problem” for presupposition (Geurts 1999).
To appear. "Bias in Question Tags." (with Cory Bill). In A. Benz et al. (eds.): Biased Questions: Experimental Results and Theoretical Modelling. Language Science Press. [draft]
We study various kinds of reverse-polarity tag questions in English, arguing that the speaker biases that such questions convey differ across three dimensions: optionality, strength, and polarity. We propose that the bias profile in each case primarily depends on the shape of the tag, while pointing at the central role of polarity focus and answer salience in generating these biases.
2019. "Adverbs of Change." In D. Altshuler and J. Rett (eds.): The Semantics of Plurals, Focus, Degrees, and Times. Essays in Honor of Roger Schwarzschild. Springer, 283–303. [paper]
Adverbs of change like quickly or slowly are known to give rise to a number of interpretations. For example, Selena ran quickly says that the rate of running is high while Selena quickly noticed the plane implies that the distance between the event of noticing the plane and some previous event is short. Existing accounts (e.g. Cresswell 1978; Rawlins 2013) take rate readings as primary but struggle to derive additional interpretations. By contrast, I argue that adverbs of change measure the temporal distance between two salient events (or event parts) that are compositionally or contextually available. The main claim of the paper is that adverbs of change have a single if underspecified semantics and that the different interpretations arise through interaction with aspectual and discourse structure.
(A shorter version appeared in Semantics And Linguistic Theory 27: 22-42, LSA under the title "Adverbs of Change, Aspect, and Underspecification." [paper])
To appear. "Impersonal Rules: The Case of General Prohibitives" (with Felix Frühauf, Hadil Karawani, Natasha Korotkova, Doris Penka, and Daniel Skibra). Sinn und Bedeutung 27.
2022. "Really: Ambiguity and Question Bias" (with Cory Bill). Sinn und Bedeutung 26. [paper]
We discuss two empirical puzzles about English really: (i) really is ambiguous between an intensifier use akin to very (cf. Zelda is really tall) and a conversational use that expresses definite certainty (cf. Zelda REALLY is tall); (ii) polar questions with conversational really convey a negative speaker bias towards the question prejacent (cf. Is Kai REALLY from Hawaii? ~> The speaker doubts that Kai is from Hawaii). We propose a single lexical entry according to which really combines with a gradable property P and states that the degree to which P applies meets all relevant standards. The ambiguity hinges on whether P ranges over degrees of individual properties (as in the case of intensifier really) or degrees of commitment (as in the case of conversational really). In addition, we propose to derive the question bias associated with conversational really from its obligatory contrastive focus marking, a feature that it shares with other polar elements that give rise to a similar effect. bias.
2021. "Verum Accent is VERUM, but not Always Focus" (with Cory Bill). Linguistic Society of America 6(1): 188-202. [paper]
This paper studies the occurrence of verum accent in declaratives and polar interrogatives. Verum accent exhibits two kinds of interpretational effect: (i) it requires an epistemic conflict across sentence types and (ii) it may also convey a negative speaker bias in polar interrogatives. We argue that the former effect is due to a presuppositional VERUM operator and that the latter effect arises from the possibility of said operator carrying polarity focus. Our proposal implies that verum accenting and polarity focus are two distinct phenomena that interact in interesting ways.
2021. "Believe is Strong but Subjective: Experimental Evidence from Hedging" (with Cory Bill and Maryam Mohammadi). Sinn und Bedeutung 25: 497-514, Open Journal Systems. [paper]
This paper contributes to the debate of whether the attitude verb believe has a weak or a strong semantics. According to Hawthorne et al. (2016), believe is weak and akin to the probability operator likely, typically receiving an “agent finds it more likely than not” interpretation. Alternatively, Koev (2019) proposes that believe conveys high certainty but qualifies this certainty as subjective or lacking evidence, in contrast with modals that convey high objective certainty, like sure (see also Lyons1977; Kratzer 1981; Nuyts 2001; Papafragou 2006; Portner 2009). We focus on the use of believe as a hedge (e.g. I believe the Giants will win, but I’m not sure they will) as allegedly the most convincing argument for the weak view, and argue that in fact it favors the strong-but-subjective view. We show experimentally that the availability of the hedging use of believe is affected by certain grammatical and discourse factors. Experiment 1 revealed that participants rate hedging sentences with combinations of third person/past tense/embedded features as less natural than canonical first person/present tense/main clause forms. In turn, Experiment 2 revealed that hedging sentences with at-issue prejacents are judged as more natural than sentences in which the belief component is at-issue. The observed variability posits a challenge to the weak view, which establishes a purely logical contrast in modal strength between likelihood vs. certainty. However, it is in line with the strong-but-subjective view, which establishes a contrast in modal content between certainty without evidence vs. certainty with evidence and predicts a more restricted distribution of the hedging reading.
2019. "Strong Beliefs, Weak Commitments." Sinn und Bedeutung 23(2): 1-18,
The standard Hintikkan semantics views believe as a universal quantifier over possible worlds (Hintikka 1969). This semantics (i) fails to capture the fact that believe is gradable (cf. partially believe or fully believe) and (ii) makes no predictions about the degree of certainty of the belief agent toward the prejacent. To remedy these problems, I propose a scalar semantics along the lines of Kennedy & McNally's (2005) analysis of gradable adjectives, arguing that believe is a maximum-degree predicate. While belief attributions are sometimes interpreted as hedges (e.g. I believe it's raining can be taken as a statement of uncertainty), I point out that such uses are restricted to contexts in which the belief component is not relevant to the question under discussion. Following up on a suggestion made in Chemla (2008), I propose that the weak sense of believe arises as an antipresupposition, a scalar inference derived through competition with a presuppositionally stronger know-competitor. Contra Hawthorne et al. (2016), I argue that the intuition of weakness is due not to reduced modal force but rather to the subjectivity of modal content, amounting to a situation in which the agent has full subjective confidence in the prejacent but fails to publicly commit to it.
2017. "Parentheticality, Assertion Strength, and Discourse." Sinn und Bedeutung 21: 679-694, semanticsarchive. [paper]
Sentences with so-called SLIFTING PARENTHETICALS (e.g. The dean, Jill said, flirted with the secretary; Ross 1973) grammaticalize an intriguing interaction between speech act function and conventional meaning, one that is not found in regular embedding constructions (e.g. Jill said that the dean flirted with the secretary). In such sentences, the main clause is independently asserted and at the same time interpreted in the scope of the parenthetical, which typically serves an evidential function. The discourse effect of this pragmasemantic set-up is that slifting parentheticals modulate the strength with which the main part of the sentence is asserted (Urmson 1952; Asher 2000; Rooryck 2001; Davis et al. 2007; Simons 2007; Maier & Bary 2015). Building on Davis et al. (2007), this paper proposes a probabilistic discourse model that captures the role of parentheticality as a language tool for qualifying speaker’s commitments. The model also derives two empirical properties that set apart slifting parentheticals from regular embedding constructions, i.e. (i) the fact that slifting parentheticals invariably express upward entailing operators and (ii) the fact that they usually do not occur in subordinate clauses.
2016. "On Quotational Indefinites." Sinn und Bedeutung 20: 412-423, semanticsarchive. [paper]
This paper discusses QUOTATIONAL INDEFINITES, an understudied variety of indefinites that is attested in languages like Bulgarian and German (see Cieschinger & Ebert 2011 on the latter), and are akin to Japanese wh-doublets (see Sudo 2008) and English placeholders like whatshisface or so-and-so (cf. Clark & Gerrig 1990). My major claim is that quotational indefinites existentially quantify over linguistic expressions and make reference to both expressions and their denotations. In addition, such indefinites require that the expressions they quantify over are of a certain type (a referential expression, a particular kind of adverbial, etc.) and originate in a previous conversation. This work uncovers important interactions between indefiniteness, quotation, and reportativity, and broadens our understanding of the typology of indefinites.
2016. "The Shiftability of German Appositive Relatives across Intensional Contexts: Two Experimental Studies." West Coast Conference on Formal Linguistics 33: 237-245, Cascadilla. [paper]
The research question that this paper is trying to answer is the following: What factors enable a perspective shift of appositives? I report two experimental studies from German that directly address this question. The experiments show (i) that verbs of saying are more likely to lead to shifted interpretations than other attitude predicates, and also (ii) that the German second subjunctive (Konjunktiv II), which has reportative uses, facilitates perspective shift. The major theoretical implication of this work is that secondary speech contexts are the prototypical appositive shifters.
2014. "Two Puzzles about Appositives: Projection and Perspective Shift." Sinn und Bedeutung 18: 217-234. [paper]
This paper solves two puzzles about the interpretation of appositive constructions in English: (i) it explains why appositives robustly PROJECT even though they are interpreted in situ with respect to order-dependent phenomena such as discourse anaphora; (ii) it analyzes certain non-projective readings of appositives as instances of PERSPECTIVE SHIFT, a phenomenon that bears striking similarities with the phenomenon of shifted indexical pronouns. To solve the first puzzle, I assume that appositives are interpreted in surface position but adopt a less standard mechanism of operator scope according to which operators can bind lexical predicates or other operators. Since appositives form ForcePs and Force operators cannot be bound, it follows that appositives project even when placed in the syntactic scope of a higher operator. With respect to the second puzzle, I argue that shifted appositives are evaluated with respect to a secondary speech context introduced by a verb of saying or inferred from the larger discourse. Since the two mechanisms of projection and perspective shift are independent, they can be given a uniform analysis.
2012. "On the Information Status of Appositive Relative Clauses." Post-proceedings of the Amsterdam Colloquium 18: 401-410, Springer. [paper][preprint]
Existing semantic theories of appositive relative clauses (ARCs) assume that ARCs contribute asserted but not at-issue content (Boer & Lycan 1976, Bach 1999, Chierchia & McConnell-Ginet 2000, Potts 2005, AnderBois et al. 2010, Murray 2010). In this paper I demonstrate that the information status of ARCs depends on their linear position in the clause: clause-medial ARCs are not at-issue whereas clause-final ARCs can behave like regular at-issue content. I propose a uniform one-dimensional semantics under which ARCs are conjuncts that can acquire at-issue status if the issue raised by the main clause has been terminated. The idea is formally implemented in Dynamic Predicate Logic (Groenendijk & Stokhof 1991) enriched with propositional variables (AnderBois et al. 2010).
2011. "Evidentiality and Temporal Distance Learning." Semantics And Linguistic Theory 21: 115-134, elanguage. [paper]
The grammatical category of evidentiality is traditionally defined as marking evidence type or related concepts (Anderson 1986, Willett 1988, Aikhenvald 2004). I argue against this received view as I show that evidential morphemes in Bulgarian mark the temporal distance between the time at which the speaker learned the described proposition and the topic time. I also demonstrate that Bulgarian evidentials represent projective/backgrounded content that is informative but does not affect the described proposition, which is plainly entailed. The latter fact especially has important typological and theoretical consequences. The proposal is formalized in a logic that extends Dynamic Predicate Logic by adding propositional variables (cf. AnderBois et al. 2010).
2011. "Definiteness as Agreement: Evidence from Bulgarian." West Coast Conference of Formal Linguistics 28: 133-141, Cascadilla. [paper]
I address the nature and distribution of the definiteness marker (DEF) in the Bulgarian DP. I argue that DEF is an inflectional suffix whose flexible distribution in the DP is due to what I call 'definiteness agreement'. Definiteness agreement is that part of the overall phi agreement in the DP which takes place between the D head and the highest/closest phi-agreeing element.
2013. Apposition and the Structure of Discourse. Unpublished dissertation, Rutgers University. [dissertation]
Work in Progress
"Question Bias from Polarity Focus." (with Cory Bill) [draft]
We argue that speaker bias in a subset of non-canonical polar questions in English is triggered by focus marking on some polar operator, such as (low or high) negation, VERUM, or conversational really. The proposed mechanism of bias generation is based on the assumption that, if a question partition cell is entailed and thus made salient by a contrasting focus antecedent, the speaker must be biased for that cell. While this mechanism generates a default speaker bias that is weak, obligatory and of the opposite polarity to that of the focus domain, in certain cases the semantics of the polar operator may independently lead to a raised degree of commitment to the question prejacent and thus may strengthen the bias. We also demonstrate how our focus-based account of speaker bias can be extended to capture the bias profiles of various reverse-polarity rising-tag questions.
"Believe is Strong but Subjective." [draft]
The verb believe is standardly analyzed as a universal quantifier over possibilities, i.e., as stating that the prejacent is true across all the attitude holder’s doxastic alternatives (Hintikka 1969). Although very popular, this semantics fails to capture the intuition that believe implies some sort of weakness on the part of the attitude holder regarding the prejacent proposition (cf. I believe that Kim is on vacation vs. I know that Kim is on vacation). In order to address this issue, I propose that believe carries a strong modal force (contra Hawthorne et al. 2016; Rothschild 2020) but conveys subjective modal content, where the latter part of the proposal builds on the distinction between ‘subjective’ and ‘objective’ epistemic modality (Lyons 1977; Kratzer 1981). I offer three main arguments in support of this view: (i) believe interacts with other epistemic expressions in ways predicted by its subjective nature, (ii) believe is closed under conjunction and thus must carry a strong force, and (iii) believe is a gradable predicate that takes the scale maximum as its default standard. Among other things, I demonstrate how the proposed account captures the initially surprising contrast in acceptability between I believe the Patriots will win, but I’m not sure and #I believe the Patriots will win, but I doubt it.
"Appositive Projection and Its Exceptions." [draft]
This paper has two major goals. The first is to offer a comprehensive account of the projection properties of appositive constructions. Appositives posit a challenge to traditional assumptions about form and meaning because they are interpreted in situ with respect to order-dependent phenomena like discourse anaphora but nevertheless escape the syntactic scope of entailment-canceling operators like negation or modals. Accounting for this pattern requires an innovative way of looking at propositional operators and the way they interact with appositives. The second goal of the paper is to address various claimed exceptions to the otherwise robust projection behavior of appositives. I argue that in some cases the construction under consideration is most likely not an appositive at all. In other cases, the observed non-speaker-oriented readings can be derived by pragmatic reasoning or are due to a perspective shift. Although genuine instances of semantically embedded appositives do seem to exist, I point out that such data have a limited empirical scope. I conclude that appositive projection is a pervasive phenomenon and is part and parcel of the semantics of appositives.