Coordinators: Steven Foley and Pranav Anand

Spring 2016

May 13

Erik Zyman (UCSC): “Hyperraising to object and the mechanics of Agree”

In English and other familiar languages, A-movement can occur out of an infinitival clause, but not out of a finite clause:

(1a) Sue₁ seems [__₁ to be embezzling money].

(1b) *Sue₂ seems [__₂ is embezzling money].

On one prominent analysis, this is because an element must bear an unvalued feature to be an eligible goal for Agree (the Activity Condition, Chomsky 2000, 2001). On this view, in (1b)—unlike in (1a)&emdash;Suegets Case in the embedded clause, and, having no more unvalued features, becomes invisible to higher A-probes such as matrix T. Here, I present novel data illustrating the Janitzio P’urhepecha (JP) “accusative + complementizer” construction (e.g., ‘They want Xumo-

that build a house’) that strongly suggest the Activity Condition is not an inherent constraint on Agree (cf. Nevins 2004). I show that the accusative DP is in the matrix and truly A-moves out of the embedded CP (i.e., this is hyperraising, not prolepsis or object control). Crucially, a hyperraised accusative DP can be linked to a nominative floated quantifier in the embedded CP&emdash;showing that nominative Case is available in the embedded CP, but its subject A-raises out of it nonetheless, challenging the Activity Condition.

The findings have further theoretical consequences. First, I argue that existing analyses of hyperraising (e.g., Martins & Nunes 2010, Carstens & Diercks 2013, Halpert 2016, Petersen & Terzi to appear) cannot extend to the JP case, and propose that JP allows hyperraising to object but English does not because in JP, but not in English, the finite embedding C can optionally bear the feature [uD[

]] (cf. Cable 2012). Secondly, I argue that JP hyperraising to object can be accounted for straightforwardly on an altruistic (target-driven) analysis of movement (Chomsky 2000, 2001, 2004, McCloskey 2001), but not under Greed (Bošković 1995, 2002, 2007) or Labeling (Chomsky 2013, 2015).

April 22

Erik Maier (Berkeley): “As above, but below: Karuk directional suffixes as low applicatives”

Karuk, a Hokan language of Northern California, has a distinctive set of over 50 verbal directional suffixes used to express the Path and/or Ground of a motion event, a subset of which were analyzed by Macaulay (2004) as high applicatives (cf. Pylkkänen 2008). In this talk, I present a previously undescribed restriction wherein telic verbs cannot combine with the suffixes and argue that, while this restriction cannot easily be accounted for in a high applicative analysis, it follows neatly from an analysis of these suffixes as PathP complements to the verb in Ramchand (2008)'s system of VP decomposition, specifically as a result of the Path-Result complementarity built into that system. This requires the suffixes to reside low in the structure, and as such they constitute a new type of low applicative afforded by Ramchand (2008)'s system, one with semantics more akin to Pylkkänen (2008)'s high applicatives, severing the connection Pylkkänen (2008) makes between low or high structural position and types of applicative semantics. Some preliminary remarks on the semantics underlying the observed Path-Result complementarity follow. Throughout the talk, I will also discuss the difficulties of diagnosing telicity in the Karuk language and field situation, and present a novel test to do so, based on insights into the default temporal interpretation of bounded and unbounded verbs from Smith (2007) and Mucha (2013).

April 8

Karl DeVries (UCSC)

Sentences like (1) have played a modest role in the literature on partitives, appearing in lists of sentences counter-exemplifying the partitive constraint (roughly, that the inner DP must be definite).

(1) That book could belong to one of three people (Ladusaw 1982).

There are two strategies for reconciling bare cardinal partitives with the partitive constraint. Ladusaw (1982) argues that the inner cardinal is a specific indefinite and Barker (1998) suggests that bare cardinal partitives can also be used when inner cardinal exhausts the restrictor set; (2) has such an interpretation.

(2) Sybil is one of three people Otis admires.

Sentences like (3) suggest that matters are more complex. When a bare cardinal partitive appears in the scope of a universal quantifier it can give rise to a cumulative reading.

(3) Every student read one of three papers. (i.e. every student read one paper and three papers were read overall)

Sentence (3) does not require that the inner cardinal be specific nor does it require that there be only three (contextually salient) papers. I develop a compositional account of cumulative readings using an extension of First Order Logic with Choice (Brasoveanu & Farkas 2011) and discuss how cumulative readings fit into larger debates about the status of the partitive constraint.

Winter 2016

March 11

Filippa Lindahl (University of Gothenburg): “Swedish relative clauses: Very weak islands”

The mainland Scandinavian languages allow movement out of relative clauses, a phenomenon known as Relative Clause Extraction (RCE). In this talk, I present results from my ongoing dissertation project. Based on a collection of examples from conversation and radio, I give an overview of the environments in which RCE occurs, and which types of phrases are typically allowed to move out of RCs in Swedish. Most extraction in spontaneous usage consists of topicalization or relativization, but interrogative wh-movement and it-clefting out of RCs are also possible. Adjuncts are usually not extracted, but this is only a tendency; it is possible to extract adjuncts that are contrastive or deictic (denoting a specific point in time, for instance). On the other hand, it is impossible to form why-questions that question an RC-internal reason.

This suggests that Swedish RCs are a type of weak island (cf. Cresti 1995, Szabolcsi 2006, Ruys 2015). But Swedish RCs are even more transparent than well-known weak islands, in that they do not block functional readings of questions. Since Swedish RCs are opaque for certain types of phrases, namely why and certain other adjuncts, we cannot simply say that they are non-islands; but semantic approaches like Cresti 1995 and Ruys 2015 are too restrictive for Swedish, since these are specifically designed to explain why functional readings are blocked. Swedish relative clauses thus show that islands aren’t just strong or weak, but that they can be very weak.

January 8

Deniz Rudin (UCSC): “Decomposing Assertion: Epistemic Possibility and the Pragmatics of Update”

Disagreement over might-claims has been held to be problematic for the standard semantics of epistemic modals (Kratzer 1977, 1981), in which might-p is true iff there is a p-world epistemically accessible from the world of evaluation. If the speaker is making the claim relative to their own accessibility relation, then why can an addressee disagree with a might-claim if they believe its prejacent to be false? When a might-claim is rejected, what’s being rejected is not that the prejacent is epistemically accessible to the speaker, but that the prejacent is epistemically accessible to the rejector. Yalcin (2011) and Swanson (2011) hold that such data invalidate the Kratzerian paradigm; Stephenson (2007), von Fintel & Gillies (2011) and Yanovich (2014) pursue various modifications of the standard Kratzerian semantics to try to account for disagreement data while preserving Kratzer’s insights. I argue that disagreement data are in fact completely unproblematic for the standard Kratzerian semantics if thought about from the perspective of a neo-Stalnakerian formal pragmatics in the vein of Farkas & Bruce (2010), in which an assertion of pcomprises a proposal for all interlocutors to commit to p. In a pragmatics in which each interlocutor evaluates the content of an assertion relative to their own information state before deciding whether or not to commit to it, it follows naturally that addressees can disagree with might-claims on the basis of their own epistemic state, regardless of the epistemic state that licensed the assertion in the first place—given the standard Kraterzian semantics for epistemic modality, the pragmatics automatically derives the effects that the accounts above attempt to stipulate into the semantics. I go on to argue that updates with might-claims should not be modeled as intersecting the context set with a set of worlds, and show how non-intersective update conditions for various theories of the semantics of epistemic modality can be derived from the same Stalnakerian formula that derives intersective updates for non-modalized propositions.

Fall 2015

December 4

Andreas Walker (Konstanz): High and low readings in counterfactuals

The interpretation of counterfactual conditionals, such as (1), has been shown to depend on a notion of similarity (Lewis 1973): we interpret the conditional in possible worlds that are as similar as possible to the actual world while differing from it in making the antecedent of the conditional true.

(1) If John had owned Platero, he would have been happy.

In this talk, I explore how this similarity relation interacts with the presence of indefinites in the antecedent. It has been argued (van Rooij 2006, Walker & Romero to appear) that there are two readings (which we call high and low)in what are called “counterfactual donkey sentences”, which exhibit the indefinite-pronoun structure of donkey sentences on top of their counterfactuality, such as (2): one in which we interpret the conditional in the most similar world where John owns a donkey, regardless of that donkey's identity, and one in which we have to consider the most similar world for any possible donkey that could be owned by John.

(2) If John had owned a donkey, he would have beaten it.

I will sketch how to derive these two readings in both current major accounts of donkey sentences, dynamic semantics and D-type theory. I will then return to the question of whether this provides a satisfactory explanation of why and when these readings arise and explore a related phenomenon: counterfactuals with disjunctive antecedents show a similar variability in readings (Alonso-Ovalle 2009). This provides a possible avenue towards a pragmatic account of the high/low distinction that generalizes over both indefinites and disjunction.

November 12

Jason Merchant (UChicago): Problems of pride and lust: Nonlocal selection?

Since Aspects, selection or strict subcategorization—both in syntax and morphology--has been restricted to complements or sisters. This fact follows as a theorem from recent formulations of Merge (and from its equivalents in other frameworks). In a theory where roots can select for their complements, the locality of selection can also provide a satisfying account of cross-categorial uniformity of selection (e.g. in "rely on, reliance on, reliant on”). This talk explores a set of data that are problematic for the claim that selection is always local; the data come from categorially nonuniform selection in English, Dutch, German, and Greek (e.g., “to pride oneself on, my pride in, proud of”, as well as historically divergent selectional properties of nominal vs. verbal “lust”), diptotic prepositions in German, and nonlocal contextual allomorphy in the Greek verb. I formulate a mechanism for joint selection, and argue that joint selectors must from a single span: any contiguous set of terminals in an extended projection. I conclude with a discussion an apparent problem from pseudopassives and the systematic nonexistence of pseudomiddles.

October 23

CUSP Practice Talks

Hitomi Hirayama (UCSC): Japanese contrastive wa and ignorance inferences

In this talk, I will investigate the Japanese contrastive marker wa with respect to ignorance inferences it can give rise to, pointing out that this particle shows similarities to superlatives like at least. In particular, I propose that the contribution of wa is best captured by using inquisitive semantics (Groenendijk and Roelofsen, 2009). Under this analysis, wadenotes a set of alternatives and makes the hearer pay attention to them in the same way as might does so for possibilities (Ciardelli et al. 2009).

Karl DeVries, Karen Duek, Kelsey Kraus, Deniz Rudin, and Adrian Brasoveanu (members of UCSC's LaLoCo): The semantics of corrections

At first glance, correcting mistakes while speaking seems like purely a performance issue, of little interest to theoretical linguists. We think this phenomenon deserves a second glance.

(1) Anders, uh ..., [Andrew]F broke the window.

We will refer to the portion of the utterance preceding the disfluency as the anchor, and the portion following it as the correction. A first analysis might be that the anchor plays no role in the interpretation of the correction. We argue that this initial assumption deserves to be revisited based on data such as (2) and (3).

(2) Every boy brought, uh ..., [should]F have brought a water bottle.

(3) Every boy brought, uh ..., he [should]F have brought a water bottle.

Prima facie, these utterances might involve partial deletion of the anchor and incorporation of remaining material into the correction: in (2) the anchor's partial VP is deleted and its subject attached to the beginning of the correction; in (3) the subject of the correction must also be deleted so that the correction can receive its proper interpretation relative to every boy. We argue that an account along these lines fails to capture the relation between anchor and correction. First, the anchor influences the focus structure of the correction:

(4) a. ??Every boy brought a book. . . I'm sorry, I mean [should]F have brought a water bottle

b. Every boy brought a book. . . I'm sorry, I mean [should]F have brought [a water bottle]F

As indicated in (1)-(3), focus is obligatory on the locus of the correction. (4) demonstrates that if there are multiple loci of correction, there must be multiple foci. If the anchor were merely cannibalized for spare parts, its content could not influence the focus structure of the correction.

Second, sentences like (3) are subject to the same constraints as telescoping phenomena.(3) has a counterpart in (5), a simplified version of an example from Roberts (1987):

(5) Every boy walked up to the speaker. He shook her hand and returned to his seat.

(6) a. *No boy walked up to the speaker. He shook her hand...

b. *No boy brought, uh, he should have brought a water bottle.

These restrictions are incompatible with a delete-and-copy account of corrections. We provide a dynamic account of correction phenomena that involves full, incremental interpretation of both the anchor and the correction, while crucially distinguishing between the interpretation of an utterance and the contribution of its assertion to the discourse, as in Anderbois et al (2015). In our account, both anchor and correction are fully interpreted, allowing the content of the anchor to play a role in the interpretation of the correction. However, only the correction is asserted. On this view, telescoping becomes possible because the entire utterance is composed via iterated dynamic conjunction, making it possible to interpret singular pronouns in the correction relative to quantifiers in the anchor. The wellformedness of the entire utterance is determined based on a relationship of focus inclusion between the anchor and the correction, namely whether there exists a possible completion of the anchor that is a member of the focus semantic value of the correction.

October 2

Hidekazu Tanaka (Okayama University): "Why (*to)?"

In this talk, I show that the impossible sequence why to in English gets ameliorated when the infinitive marker to is elided. An account is developed that assimilates the observation to a much better known syntactic fact: island repair. To establish this, I will examie two constructions that remain poorly studied in the literature; coordinated wh-infinitives and antecedent contained sluicing.