This dissertation shows that the generalizations that speakers project from the lexical exceptions of their language are biased to be natural and output-oriented, and it offers a model of the grammar that derives these biases by encoding lexical exceptions in terms of lexically-specific rankings of universal constraints in Optimality Theory (Prince & Smolensky 1993/2004). In this model, lexical trends, i.e. the trends created by the phonological patterning of lexical exceptions, are incorporated into a grammar that applies deterministically to known items, and the same grammar applies stochastically to novel items. The model is based on the Recursive Constraint Demotion algorithm (Tesar & Smolensky 1998, 2000; Tesar 1998; Prince 2002), augmented with a mechanism of constraint cloning (Pater 2006, 2008b). Chapter 2 presents a study of Turkish voicing alternations, showing that speakers replicate the effects that place of articulation and phonological size have on the distribution of voicing alternations in the lexicon, yet speakers ignore the effects of vowel height and backness. This behavior is tied to the absence of regular effects of vowel quality on obstruent voicing cross-linguistically, arguing for a model that derives regular phonology and irregular phonology from the same universal set of OT constraints. Chapter 3 presents a study of Hebrew allomorph selection, where there is a trend for preferring the plural suffix [-ot] with stems that have [o] in them, which is analyzed as a markedness pressure. The analysis of the trend in terms of markedness, i.e. constraints on output forms, predicts that speakers look to the plural stem vowel in their choice of the plural suffix, and ignore the singular stem. Since real Hebrew stems that have [o] in the plural also have [o] in the singular, Hebrew speakers were taught artificial languages that paired the suffix [-ot] with stems that have [o] only in the singular or only in the plural. As predicted, speakers preferred the pairing of [-ot] with stems that have [o] in the plural, i.e. speakers prefer the surface-based, output-oriented generalization. Chapter 4 develops the formal theory of cloning and its general application to lexical trends, and explores its fit with the typologically available data. One necessary aspect of the theory is the “inside out” analysis of paradigms (Hayes 1999), where the underlying representations of roots are always taken to be identical to their surface base form, and abstract underlying representations are limited to affixes. An algorithm for learning the proposed underlying representations is presented in a general form and is applied to a range of test cases.

In this dissertation, I examine the effects of lexical accent on the perception of intonational prominence in Japanese. I look at how an F0 accent peak is perceived relative to another flanking F0 peak in the same utterance with respect to perceived intonational prominence. Through four experiments, I show that the lexical prosodic structure plays a significant role in the perception of intonational prominence. I first show that two distinct perceptual processes are at play in the perception of relative perceived prominence in Japanese: accentual boost normalization and downstep normalization. Accentual boost normalization normalizes the accentual boost of an accented word. In this process, the extra F0 boost assigned by a lexical accent does not count as part of the F0 peak’s excursion that contributes to the perceived prominence of the F0 peak. I demonstrate that when an accented word and an unaccented word are perceived as having the same prominence, the accented word has a higher F0 peak value than the unaccented word does. Downstep normalization compensates for the production effect of downstep, a pitch range compression phenomenon after a lexical accent. Experiments show that for an F0 peak to be perceived as having equivalent prominence to a preceding F0 peak, the second peak is always lower in F0 when the first word is accented than when it is unaccented. This suggests the existence of a perceptual process that normalizes the effect of downstep. I then examine the nature of accentual boost normalization and downstep normalization and show that they refer to two distinct types of lexical accent property when they are applied. One is the phonetic F0 contour shape that is characteristic of accented words. The other is the phonological lexical accent information that is uniquely specified for accented words. The experimental results show that the perceptual effects of the normalization processes are seen when only the phonological lexical accent information of a word is present with its F0 contour shape being ambiguous as well as when the same word is acoustically manipulated into different F0 contour shapes.

This dissertation examines the syntax and semantics of clausal complements. It identifies semantic underpinnings for some syntactic properties of the arguments of propositional attitude verbs. The way clausal arguments compose with their embedding predicates is not uniform and semantic differences emerge from the syntactic context clausal arguments appear in. Three case studies are taken up: clausal arguments of nouns, dislocated clausal arguments (sentential subjects and topics), and infinitival complements with overt subjects (AcI constructions). Chapter Two assembles evidence to support Stowell’s (1981) claim that the clausal complements of nouns are modifiers. It is shown that the clausal complements of nouns behave like adjuncts in their ability to bleed condition C (Kuno 2004 and Jacobson 2003, and explored here further). The compositional strategy used to compose attitude nouns with their arguments, following Kratzer (2006), is shown to account for this behavior and to be commensurate with observations made by Grimshaw (1990). I then show how the modifier status of clausal complements of nouns is determined by the way in which nominals are formed from clause-taking verbs. Chapter Three examines another complementation strategy, found with fronted clauses. New data from binding is provided in support of Koster’s (1978) hypothesis that clauses do not move. Specifically, fronted clauses fail to show the effects of syntactic reconstruction. An analysis, making crucial use of de re attitude ascription, is offered to account for ‘apparent’ binding into fronted clauses. Chapter Four makes the case for enriching the meanings of clausal complements. By examining some new patterns with accusative with infinitive (AcI) constructions (such as I see him to be a fool), I argue for decomposing certain doxastic attitude verbs, putting the introduction of alternatives into the complement. Here too the role of de re attitude ascription is shown to play a crucial role. It is argued that AcI constructions involve de re attitude ascription, with added constraints (determined by the lexical content of the embedding verb) on the nature of the Acquaintance Relation (Kaplan 1968, Lewis 1979). Several predictions about the kinds of verbs that can participate in AcI are borne out.

This thesis is concerned with the description and analysis of two semantically different types of definite articles in German. While the existence of distinct article paradigms in various Germanic dialects and other languages has been acknowledged in the descriptive literature for quite some time, the theoretical implications of their existence have not been explored extensively. I argue that each of the articles corresponds to one of the two predominant theoretical approaches to analyzing definite descriptions: the 'weak' article encodes uniqueness. The 'strong' article is anaphoric in nature. In the course of spelling out detailed analyses for the two articles, various more general issues relevant to current semantic theory are addressed, in particular with respect to the analysis of donkey sentences and domain restriction. Chapter 2 describes the contrast between the weak and the strong article in light of the descriptive literature and characterizes their uses in terms of Hawkins's (1978) classification. Special attention is paid to two types of bridging uses, which shed further light on the contrast and play an important in the analysis developed in the following chapters. Chapter 3 introduces a situation semantics and argues for a specific version thereof. First, I propose that situation arguments in noun phrases are represented syntactically as situation pronouns at the level of the DP (rather than within the NP). Secondly, I argue that domain restriction (which is crucial for uniqueness analyses) can best be captured in a situation semantics, as this is both more economical and empirically more adequate than an analysis in terms of contextually supplied C-variables. Chapter 4 provides a uniqueness analysis of weak-article definites. The interpretation of a weak-article definite crucially depends on the interpretation of its situation pronoun, which can stand for the topic situation or a contextually supplied situation, or be quantificationally bound. I make a specific proposal for how topic situations (roughly, the situations that we are talking about) can be derived from questions and relate this to a more general perspective on discourse structure based on the notion of Question Under Discussion (QUD) (Roberts 1996, Büring 2003). I also show that it requires a presuppositional view of definites. A detailed, situation-semantic analysis of covarying interpretations of weak-article definites in donkey sentences is spelled out as well, which provides some new insights with regards to transparent interpretations of the restrictors of donkey sentences. Chapter 5 deals with so-called larger situation uses (Hawkins 1978), which call for a special, systematic way of determining the situation in which the definite is interpreted. I argue that a situation semantic version of an independently motivated type-shifter for relational nouns (shifting relations (⟨e, ⟨e, st⟩⟩) to properties (⟨e, ⟨st⟩⟩)) brings about the desired situational effect. This type-shifter also applies to cases of part-whole bridging and provides a deeper understanding thereof. Another independently motivated mechanism, namely that of Matching functions, gives rise to similar effects, but in contrast to the type-shifter, it depends heavily on contextual support and cannot account for the general availability of larger situation uses that is independent of the context. The anaphoric nature of the strong article is described and analyzed in detail in chapter 6. In addition to simple discourse anaphoric uses, I discuss covarying interpretations and relational anaphora (the type of bridging expressed by the strong article). Cases where uniqueness does not hold (e.g., in so-called bishop sentences) provide crucial evidence for the need to encode the anaphoric link between strong-article definites and their antecedents formally. The resulting dynamic analysis of strong-article definites encodes the anaphoric dependency via a separate anaphoric element that is incorporated into a uniqueness meaning. Finally, remaining challenges for the analysis are discussed, in particular the existence of strong-article definites without an antecedent and a puzzling contrast between the articles with respect to relative clauses. The final chapter discusses some loose ends that suggest directions for future work and sums up the main conclusions.

This dissertation is about the two related notions of subsumption and distinctness and the roles they play in the semantics of the particles only and too, respectively. In chapter 2 I formulate the distinctness requirement introduced by too in terms of the distinctness relation, which is taken to be a relation between sets of events. The chapter examines the nature of the distinctness requirement and, by doing so, the semantic nature of too. Two issues are discussed: (1) In which sense is too focussensitive? (2) Is the distinctness requirement a presupposition? I provide some evidence that too is conventionally associated with focus, in the sense of Beaver and Clark (2008), and that it is a presupposition trigger, in the sense of Heim (1983, 1992). I then formulate, building on Heim (1983, 1992) and others, a dynamic account of too as a presupposition trigger. Chapter 3 is dedicated to the notion of subsumption, which plays a crucial role in the semantics of only. It opens with an overview of Kratzer's (1989) account of subsumption within the framework of situation semantics. I then argue that Kratzer's account does not apply to modal sentences, in particular to counterfactuals. I propose that subsumption should be analyzed as Strawson-entailment with respect to a subset of W, i.e. as weak Strawson-entailment. The notion of Strawson-entailment was put to use by von Fintel (1997) in accounting for several phenomena related to counterfactuals. It will be shown to play a crucial role in a general account of subsumption as well. Throughout chapter 3 it is assumed that when only combines with a counterfactual, it takes as argument the proposition expressed by that counterfactual. In chapter 4 I try to be more faithful to the syntax, where only is a sister of the antecedent or the consequent, depending on its location in the sentence. I sketch an analysis where only, while not combining with the entire counterfactual, still has as its domain of quantification a set of counterfactual propositions.

This dissertation explores case and verbal agreement in Icelandic. Case and agreement generally pattern together, but there are exceptional instances in which case and agreement come apart. In Icelandic, verbs agree with Nominative DPs. However, in some constructions, agreement with a Nominative is optional. In the standard account of case and agreement (Chomsky 2000), both types of features are determined simultaneously via the same syntactic operation. The standard theory, therefore, predicts that case and agreement should pattern the same way, and that neither should be optional. Moreover, based on fieldwork conducted at the University of Iceland, I present data that has not heretofore been reported. I argue that the likelihood of agreement depends on the type of construction. My research builds on other work which addresses optionality in Icelandic agreement (e.g. Sigurðsson and Holmberg 2008). This dissertation makes a substantial contribution to the literature on Icelandic agreement in that the rate of agreement across various types of constructions has not been examined. I illustrate that this type of optionality is not only robust, but also systematic. This dissertation contributes to the larger literature on case and agreement in several important ways. First, I argue for a departure from the standard proposal that case and agreement are established via the same syntactic operation. I propose that it is possible for the probe which assigns case to be in a relationship with a DP, even though the probe which establishes agreement is not in a relationship with that DP. Second, I provide empirical support for Multiple Agree. I argue that the survey findings reported in this dissertation provide evidence that a probe can enter into a relationship with more than one goal. Third, I provide empirical evidence for the optionality of Multiple Agree. I argue that agreement is optional only in constructions in which there is an item intervening between T and the Nominative, and Multiple Agree is, thereby, required in order for an agreement relationship to be established.

This dissertation investigates the articulation of aspect in African-American English (AAE). Its primary goal is the development of a formal semantics of AAE simple V-ed sentences that explains their compositional interpretation and relationship to done V-ed sentences. Building largely on the valuable works of Green (1993; 1998), Déchaine (1993), Dayton (1996), the work herein supports the conclusions that AAE simple V-ed sentences such as The frog done jumped are ambiguous, having both past perfective and present perfect readings, and that AAE done V-ed sentences such as The frog done jumped are unambiguously present perfect. Further, it identifies a distinction in meaning between AAE simple V-ed perfects and done V-ed perfects. This distinction makes untenable analyses of the simple V-ed ambiguity in which a silent done is responsible for contributing perfect aspect to the present perfect versions of these sentences. Instead, this work traces the ambiguity to the presence of a covert present tense operator found in the present perfect (but not past perfective) versions of simple V-ed sentences, and the interaction of this operator with the -ed morpheme. In the proposed analysis, single AAE -ed morpheme unambiguously denotes a temporal relation of precedence, contrasting with the two distinct Standard American English (SAE) morphemes often notated as -ed and -en and often argued to denote past and perfect respectively. When it interacts with a covert present tense operator, AAE -ed contributes its precedence relation to the domain of aspect, resulting in the perfect aspect relation (situation time precedes topic time). When it is the highest tense/aspect marker in a sentence, it contributes its precedence relation to the domain of tense, resulting in the past tense relation (topic time precedes utterance time). On the proposed analysis, -ed thus makes the same semantic contribution to simple V-ed sentences and done V-ed sentences on all of their readings. One theoretically interesting result of this investigation is the finding that a semantically unambiguous operator may contribute sometimes to the interpretation of aspect and sometimes to the interpretation of tense.

This dissertation focuses on a phenomenon that I call vowel intrusion . There are cases where a vowel can be heard between two consonants, yet the phonology behaves as if no vowel is present. These “intrusive vowels” are non-syllabic, and native speakers are often unaware of their existence. I argue that intrusive vowels are a percept resulting from the organization of articulatory gestures. When two consonant gestures have little overlap with one another, there is an acoustic release between them; vowel gestures typically overlap neighboring consonants considerably, and it is possible for an overlapping vowel gesture to be heard in this period of release. Intrusive vowels are not segments. They behave unlike true epenthetic vowels. A topological survey reveals that vowel intrusion happens in consonant clusters that contain a sonorant or a guttural, and that it is always the vowel adjacent to the sonorant or guttural that is heard during the release. Intrusive vowels occur primarily in heterorganic clusters, especially next to geminates; they often disappear at fast speech rates, and in some languages, they occur only within or only between syllables. I argue that these characteristics are best explained in a theory that uses Articulatory Phonology representations (Browman & Goldstein 1986 et seq.). I develop a theory called Timing-Augmented Surface Phonology (TASP), cast within the framework of Optimality Theory. TASP contains constraints on the alignment of neighboring gestures (Gafos 2002) and on the permitted degree of overlap between different gestures. The theory requires a segmental representation as well as a gestural representation. Syllables organize segments rather than gestures, and that inter-segmental gestural alignment is universally non-contrastive. The same gestural framework describes both the short, schwa-like intrusive vowels often described as “excrescent”, and also a longer type found in Scots Gaelic and Hocank (Winnebago), in which the vowel is heard in two long parts on either side of the sonorant. In the latter cases the sonorant and vowel together behave like a bimoraic nucleus, and are adjoined in a structure similar to vocalic diphthongs. The theory also has implications for the analysis of Hocank accent.

This dissertation addresses two questions: (a) how do we explain the limited set of argument structure types? and (b) how do lexical structures relate to syntactic structures? My work is situated within the approach of Hale and Keyser (2002, henceforth H&K), whose purpose is to ascertain the role that structural factors play in the behavior of lexical items. As for the first question, H&K argue that argument structures are restricted by the combination of primitive lexical categories defined solely by structural properties. I point out that the restrictive power of this theory is undermined by allowing unrestricted recursive combination of such primitive units. By restricting this type of lexical recursion, this problem disappears. As for the second question, I argue that two adicity-changing processes, transitivization and detransitivization, are key to understanding the relation between lexical and syntactic structures. I investigate these processes in Catalan and Spanish. I argue that there are two types of verbal affixes: one plays no role in the argument structure of the verb (but refers instead to its aspectual properties), while the other is responsible for the aforementioned processes. I argue that the affixes of the second type correspond to a functional category that bridges over from lexical to syntactic structures. After identifying these two types of affixes, I re-analyze some data (from O'odham, Navajo, Miskitu, Ulwa, and Yaqui) that H&K find problematic. Finally, I address an important extension of H&K's theory, the “manner index”. Such index is either proximate (bound by the internal argument), or obviative (bound by other than the internal argument). In H&K's theory, the type of index associated with the root determines a particular structure, which in turn determines the behavior of the corresponding verb. By breaking the determination relation between index and structure, I derive two typologies. First, I obtain a four-way paradigm by combining the structural type (“put”/“get”) and the index type (obviative/proximate). Spanish prepositional verbs present the predicted patterns. Second, I account for two alternations: unergative-unaccusative in Italian, and unergative-transitive in English. Given the explanatory power of the obviative/proximate index, I conclude that it is a welcome addition.

This dissertation examines the semantic interpretation of various types of DPs in so-called concealed-question (CQ) constructions, as Bill’s phone number in the sentence John knows Bill’s phone number. The peculiar characteristic of DP-CQs is that they are interpreted as having the meaning of an embedded question. So, for instance, the definite Bill’s phone number from the example above can have the same meaning as the embedded question what Bill’s phone number is. Building on previous proposals from Heim (1979) and Romero (2005), I defend the hypothesis that CQs denote individual concepts (IC-approach). The main result of the dissertation is that (a) it provides genuinely new analyses for several types of CQs that seemed problematic for existing analyses, including quantified and indefinite CQs (John knows every book that Mary read this summer/a doctor who can treat your illness ), and (b) it shows that the IC-approach can deliver the right results if we allow quantifier raising and adopt the copy theory of movement (Chomsky 1995) and Fox’s trace conversion mechanism (Fox 1999, 2002). Chapter 1 introduces initial data on CQs and briefly discusses dissimilarities between concealed questions and their embedded question counterparts. In Chapter 2, I introduce Heim (1979) and Romero (2005)’s analysis of definite - CQs as denoting individual concepts (IC-approach). Following up on Nathan (2006), I show that the IC-approach can be extended to account for CQ-meanings of quantified DP-objects, under the assumption that the NP-CQ is shifted into a predicate of meaningfully sorted individual concepts (an assumption that was not required to account for CQ-meanings of definite descriptions). As discussed extensively in the course of the chapter, the assumption that common nouns must in some cases denote predicates of individual concepts has found independent motivation in the literature (Montague 1973, Nathan 2006, Romero 2007, among others). Therefore, the proposed extension of Heim and Romero’s analysis to the quantified cases is fairly uncontroversial. In Chapter 3, I discuss some problems for the IC-Approach. First, I show that the analysis of quantified CQs laid out in Chapter 2 cannot be extended to quantified CQs with non-relational NPs. Second, I discuss the problematic ambiguity between pair-list readings and set readings (Heim 1979, Roelofsen and Aloni 2008) and propose that such ambiguity should be traced back to the systematic ambiguity between transitive and intransitive meanings of relational nouns. In this way, I argue that the failure of accounting for set readings under the IC-approach is just another symptom of its inability to account for non-relational NP-CQs, and that the two problems should be unified. Finally, I discuss the challenge presented by indefinite CQs with non-relational nouns. In Chapter 4, I propose an amendment to the IC-approach that accounts for the problems presented in Chapter 3. The solution relies on the copy theory of movement (Chomsky 1995) and Fox’s trace conversion mechanism (Fox 1999, 2002). Overall, The main point of the chapter is to show that once we have an account for quantified CQ-readings with non-relational NPs, all the other challenges can also be taken into account. Finally, I propose that one further amendment is necessary to account for pair-list readings with relational nouns that are not functional.