Habil.-schrift Univ. Zürich, 1999. / Die einzelnen Texte sind zwischen 1992 und 1998 in unterschiedlichen Publikationen erschienen.

This thesis is concerned with the syntactic and semantic properties of a dispositional middle construction. Building on new empirical observations, I argue for a novel derivational account of the middle construction, namely that it has a complex eventive structure. This proposal contrasts with existing analyses that can be divided into syntactically uniform approaches and semantically uniform approaches. Crucially, existing analyses have failed to address the fact that only transitive activity and accomplishment verbs undergo middle formation. I use this observation to argue that an eventive cause, or a CAUSE head, is a defining structural feature of a middle construction. Furthermore, I argue that the middle construction does not project the understood agent in syntax and the understood theme is base-generated in the internal argument position. The desired surface structure arises via movement of the understood theme, which creates a linearizable structure. The complex eventive structure argued for in this thesis not only captures the syntactic properties of a middle construction, but it also serves as the source of a dispositional interpretation, which is one of the characteristic properties of a middle construction. / Thesis / Master of Science (MSc)

<p>Event-related potentials (ERPs) have been used extensively in the scientific research of cognitive processing such as language comprehension. Specific responses, such as the negativity called N400 (Kutas & Hillyard, 1980), have in the literature typically been associated with semantic violations in sentences. Another electrophysiological response, the positive P600 waveform, has mostly been associated with syntactic and morphological violations. However, recently, the P600 has been reported also in connection with semantic violations (Kuperberg et al, 2003; van Herten, 2004; Osterhout, 2004). The present research further explores the neurophysiological correlates of processing sentences with semantic and morpho-syntactic violations. It tests the functional interpretations of the P600 component, which has been proposed to reflect syntactic error detection, context updating, or syntactic reanalysis or repair. I contrasted semantic and syntactic possessive violations. The semantic violation conditions (i.e. <em>The mother borrowed the <strong>car’s daughter</strong> for work yesterday</em>), morphosyntactic violations (i.e. <em>The mother borrowed the <strong>daughter car</strong> for work yesterday) </em>and double violation sentences (i.e. <em>The mother borrowed the <strong>car daughter</strong> for work yesterday) </em>were derived from the control condition, (i.e. <em>The mother borrowed the <strong>daughter’s car</strong> for work yesterday). </em>I explored whether the P600 component may index more general processes than ones related to syntactic error detection. An N400 was seen to our semantic manipulation, i.e. <em>The mother borrowed the <strong>car’s daughter</strong> for work yesterday.</em> However, none of the conditions produced a P600.</p> / Master of Science (MSc)

Studies of African American English (AAE) structure have historically placed significantemphasis on its system of tense and aspect, and have done so for good reason. In the interest of developing a comprehensive descriptive analysis of the variety’s syntactic and semantic features, research on the syntactic constructions and functional grammatical items that distinguish it from other English varieties continues to bring about new insights into the different elements that make up a system of tense and aspect, as well as how these elements interact with other parts of the grammar—not only in AAE but crosslinguistically. One of these elements is the verbal marker <i>BIN</i>, which situates part of an event in the remote past, as shown in (1).<div><br></div><div>(1) Jane <i>BIN</i> saw that movie.</div><div>‘Jane saw that movie a long time ago.’<br></div><div><br></div><div>This paper further investigates both the function of and restrictions on the aspectual marker <i>BIN </i>in African American English (AAE) using acceptability judgment data collected in an online survey of AAE speakers. With this study, I aim to contribute to thetheoretical description of the verbal system of AAE (L. J. Green, 1993) and its system of tense and aspect. The judgment task will identify patterns of acceptability surrounding the following two factors: event type and whether the verb receives progressive or past tense marking. Using a generative-constructivist semantic framework (Ramchand, 2008), I hypothesize that the semantic information represented by the aspectual marker BIN will either allow or disallow certain combinations of event structure and progressiveness, and these restrictions may be demonstrated to be systematic according to the erb classes proposed byRamchand (2008). Additionally, based on the survey data and the approach to the decomposition of event structure regarding Outer and Inner aspect proposed by (Travis, 2010), I will propose that restrictions on <i>BIN </i>and ambiguity between structures containing<i> BIN </i>can be accounted for syntactically based on the configurations of both grammatical and lexical aspect.<br></div><div><br></div>

This dissertation investigates intra-clausal relations, namely, the relations that obtain between the elements in a clause. This investigation is based on a detailed study of copular sentences in Russian and Italian. In particular, three types of intra-clausal relations are investigated here: phrase-structural relations, thematic relations, and case relations. / With respect to phrase-structural relations, it is argued that not all syntactic structures are asymmetrical. Rather, it is proposed that under certain conditions---when the two input phrases have the same features---Merge will result in a symmetrical structure. This requirement for matching features leads to a more parsimonious analysis of equative sentences where the interpretation derives directly from the syntactic structure, without postulating a special "identity copula". / As for thematic relations, it is claimed that there is no one-to-one correspondence between thematic positions and structural positions (contra the strong version of UTAH, Baker 1988). Instead, a more flexible theory of thematic relations is proposed. It is also proposed that theta-assignment is not a necessary condition for DP interpretation. Rather, a DP can be interpreted if it establishes a certain relationship with another theta-marked DP. This analysis extends to Left Dislocation, Pronoun Doubling and sound like -construction. / Finally, case relations are said to be tied to thematic relations. A version of the Visibility Condition is thus argued for. It is maintained that non-argument DPs---namely, those that are merged as neither complements nor specifiers of a lexical head---need not be case-marked in syntax at all and appear with the morphological default (i.e., nominative) marking. The alternative "agreement in case" analysis of NOM-NOM sentences is argued against; various conceptual and empirical problems for this analysis are identified and discussed. / The analysis developed in this dissertation accounts for a number of properties of copular sentences, including their interpretation, case-marking patterns, and such syntactic properties as extraction, inversion, binding possibilities and unaccusativity diagnostics.

No description available.

In Japanese linguistics, multiple nominative sentences, passives and causatives have been widely discussed both in the fields of traditional grammar and generative grammar. Some arguments, mainly in generative grammar, are overviewed; and some counterarguments are presented. Furthermore, these constructions, which have been treated separately from each other in the grammar, are discussed together. Some interactions of these constructions, which have various types, are shown, with focus on one particular type in each construction. This particular type involves a relationship of two NPs, namely, an NPx -no NP relationship (NP's NP). In addition, with the utilization of bound morphemes such as sase (the causative morpheme), the lexical approach, which is a recent trend in generative grammar, is taken and defended.

This thesis discusses rule incompatibility phenomena (i.e., ungrammaticality resulting from two or more applications of movement rules in particular ways) and their related theoretical problems. The discussion concludes that the transformational component is exempted from many tasks which have been assigned to it, and the component should be reduced to a great extent. Rule incompatibility phenomena cannot be accounted for well by devices in the transformational component, such as the SD and SC of transformational rules, rule ordering, or constraints on rule application. The task of explaining the phenomena must be transferred to S-structure or the LF-component. A proposal is made that the phenomena are to be handled by constraints on binding theory, which operates at S-structure. Under the assumption of the free indexing theory, instances of rule incompatibility involve ambiguity in binding, or phrases not eligible to be binders. Thus, they are ruled out by the Unique Ā-Binding Constraint (UĀBC), which states that Ā-bound expressions must be uniquely bound to binders, and the Constraint on Incomplete Ā-Binders, which states that incomplete phrases cannot Ā-bind expressions. These constraints release the transformational component from the task of explaining the grammaticality judgments. Some cases of rule incompatibility can be accounted for either by the wh-island constraint or by the UĀBC. But, Chomsky's bounding theory must be revised, and given a revised bounding theory, the wh-island constraint does not hold. Nonexistence of the wh-island constraint indicates that rule incompatibility phenomena must be accounted for by the other alternative (i.e., the UĀBC), and that the transformational component is exempt from the wh-island constraint. Of the two possible indexing theories, the free indexing theory can handle rule incompatibility phenomena, but the other theory (the simultaneous indexing theory in our terminology) cannot. The choice of the free indexing theory releases transformational rules from the task of assigning indices to traces. It is furthermore suggested that the transformational component may be exempt from even the Subjacency Condition and the rule Move-α, and therefore, may be abandoned completely.

No description available.

No description available.