Refine
Year of publication
Document Type
- Doctoral Thesis (41) (remove)
Language
- English (41) (remove)
Has Fulltext
- yes (41)
Is part of the Bibliography
- no (41)
Keywords
- Semantics (2)
- Semantik (2)
- Spracherwerb (2)
- Adjective (1)
- Adjektiv (1)
- Alternative Questions (1)
- Alternativfragen (1)
- Architecture (1)
- Astell (1)
- Beckett (1)
Institute
- Neuere Philologien (41) (remove)
In narratology, a widely recognized method involves exploring the connection between implied authors and implied readers. It entails correlating abstract narrative components within a text to understand the conveyed message and the multitude of interpretations it can offer. The present study adopts an implied reader-oriented approach to analyze three selected novels from the twentieth and twenty-first centuries—one Nigerian, one Caribbean, and one Kurdish. The aim is to explore the potential readings within these texts, considering the hermeneutic process of critical reading. The selected texts include Chinua Achebe’s Things Fall Apart, (1958), Same Selvon’s The Lonely Londoners, (1956), and Karwan Kakesur’s The Channels of the Armed Monkeys, (2011). This approach closely examines the communication between the author and reader of the text, with a special focus on the varying levels of communication between the components of the narration, including fictional and implied fictional communication.
The implied fictional communication occurs between a narrative agent known as ‘the implied author’ and its fictional counterpart ‘the implied reader’ rather than between the real, flesh and blood authors and readers. I argue that this level of communication is coded, and the act of decoding it is part of the reading process performed by the reader. Certain texts can propose different and sometimes opposing readings which are initially and purposefully designed by the implied author and addressed to different implied readers. These readings are not necessarily the results of different real readers but rather incorporated ones predetermined by the implied author only to be acknowledged and uncovered by the readers. In other words, the latent meaning is and always was an integral part of the text and is not something created by the imaginative reader or critic. The core interest of my thesis lies in identifying prompts and suggestions within the narrative of the selected texts and ultimately understanding the readerships prestructured in them. Identifying the different readers within those texts will provide new reinterpretations that can add undetected values to the reading process and sometimes suggests opposing readings to how those texts have so far been read. Additionally, it is the objective of this thesis to propose new ways that readers can interact with reading literature that would result in a more aesthetic and entertaining reading experience besides providing ways to be more informed and aware of the cues certain narrative texts contain.
There have been numerous critical studies on both narratology and postcolonial or minority literatures; however, there has been little scholarly work that attempts to utilize narratology as a theoretical foundation for understanding postcolonial and minority fiction.
This study examines fictional texts from Nigerian, Caribbean, and Kurdish literature, employing the narratological concept known as ‘Multiple Implied Readers’. By incorporating concepts from Brian Richardson’s ‘Singular Text, Multiple Implied Readers’, and Peter J. Rabinowitz’s ‘authorial audiences’, I explore the various readerships that the texts could encompass. This exploitation may lead to the discovery of new readings, interpretations, and meanings that would otherwise remain undetected. These structures introduce provocative indeterminacies that challenge the reader’s synthesis of information into coherent configurations of meaning. Consequently, this approach not only enhances the reading experience but also opens doors to new interpretations of the text. In some cases, these interpretations could even dismantle prior understandings and propose entirely new readings.
The concepts of the implied author and implied reader have been studied before in relation to various disciplines of narratology. However, by applying them in conjunction with the relatively less researched subject of multiple implied readers, I aim to shed light on important aspects of these readings. This exploration could prove beneficial for literature students as well as critical readers of literary texts, revealing the potential of these texts to accommodate more than one implied reader within their narratives.
This dissertation is about case competition in headless relatives. Case competition is a situation in which two cases are assigned but only one of them surfaces. One of the constructions in which case competition takes place is in headless relatives, i.e. relative clauses that lack a head. This dissertation has two goals: (i) to give an overview of the data, and (ii) to provide an account for the observed data.
The grammaticality of a headless relative is determined by two aspects. The first aspect concerns which case wins the case competition. In all languages with case competition that I am aware of, this is determined by the case scale in NOM < ACC < DAT. A case more to the right on the scale wins over a case more to the left on the scale. This scale is not specific to case competition in headless relatives, but it can also be observed in syncretism patterns and morphological case containment. I show that that the case scale can be derived from assuming the cumulative case decomposition (cf. Caha 2009). A case wins over another case when it contains all features that the other case contains.
The second aspect of case competition in headless relatives concerns whether the winner of the case competition is allowed to surface when it wins the case competition. The winning case can be either the internal case required by the predicate in the relative clause, or the external case required by the predicate in the main clause. It differs from language to language whether they allow the internal and the external case to surface.
All language types I discuss allow for a headless relative when the internal and the external case match. The unrestricted type of language allows both the internal case and the external case to surface when either of them wins the case competition. Examples of this language type are Old High German, Gothic and Ancient Greek. The internal-only type of language allows only the internal case to surface when it wins the case competition, and it does not allow the external case to do so. An example of this language type is Modern German. The external-only type of language allows only the external case to surface when it wins the case competition, and it does not allow the internal case to do so. To my knowledge, there is no language that behaves like this. The matching type of language allows neither the internal nor the external case to surface when either of them wins the case competition. An example of this language type is Polish.
To account for the data, I set up a proposal that generates the attested patterns and excludes the non-attested ones. I let the variation between languages follow from properties of languages that can be independently observed. By investigating the morphology of the languages, I suggest differences between the lexical entries in the different languages. These different lexical entries ultimately lead languages to be of different types. In my proposal, I assume that headless relatives are derived from light-headed relatives. Light-headed relatives contain a light head and a relative pronoun. In a headless relative either the light head or the relative pronoun is deleted. The necessary requirement for deletion is that the deleted element (either the light head or relative pronoun) is structurally or formally contained in the other element.
I motivate the analysis for the internal-only type of language for Modern German, for the matching type of language for Polish and for the unrestricted type of language for Old High German. I first identify the morphemes that the light heads and relative pronouns in the languages consist of, and then I show to which features each of the morphemes correspond. The crucial difference between the internal-only type of language Modern German and the matching type of language Polish is how the phi and case features are spelled out. In Modern German they are spelled out by a phi and case feature portmanteau, and, in Polish, the same features are spelled out by a phi feature morpheme and a case feature morpheme. Old High German differs from the other two languages in that it has light heads and relative pronouns that are syncretic. I show how these differences in the morphology of the languages ultimately leads to different grammaticality patterns in headless relatives.
Comparing my account to others shows that all proposals account for the case facts using some kind of case hierarchy. The proposals differ in how they model the variation, both in the technical details of the proposal, but more importantly, also in empirical scope and predictions they make.
This thesis investigates the structure of research articles in the field of Computational Linguistics with the goal of establishing that a set of distinctive linguistic features is associated with each section type. The empirical results of the study are derived from the quantitative and qualitative evaluation of research articles from the ACL Anthology Corpus. More than 20,000 articles were analyzed for the purpose of retrieving the target section types and extracting the predefined set of linguistic features from them. Approximately 1,100 articles were found to contain all of the following five section types: abstract, introduction, related work, discussion, and conclusion. These were chosen for the purpose of comparing the frequency of occurrence of the linguistic features across the section types. Making use of frameworks for Natural Language Processing, the Stanford CoreNLP Module, and the Python library SpaCy, as well as scripts created by the author, the frequency scores of the features were retrieved and analyzed with state-of-the-art statistical techniques.
The results show that each section type possesses an individual profile of linguistic features which are associated with it more or less strongly. These section-feature associations are shown to be derivable from the hypothesized purpose of each section type.
Overall, the findings reported in this thesis provide insights into the writing strategies that authors employ so that the overall goal of the research paper is achieved.
The results of the thesis can find implementation in new state-of-the-art applications that assist academic writing and its evaluation in a way that provides the user with a more sophisticated, empirically based feedback on the relationship between linguistic mechanisms and text type. In addition, the potential of the identification of text-type specific linguistic characteristics (a text-feature mapping) can contribute to the development of more robust language-based models for disinformation detection.
This dissertation deals with the lexical, morphological, syntactic, and semantic properties of (VP )idioms and their behavior in combination with restrictive relative clauses, raising, constituent fronting, wh-movement, VP-ellipsis, pronominalization, the progressive form, verb placement, passivization, conjunction modification, and the N-after-N construction. It provides empirical evidence towards a combinatorial analysis of both semantically non-decomposable idioms (SNDIs) and semantically decomposable idioms (SDIs) and contributes to the (formal) formulation of such an account.
The Introduction (Chapter 1) first motivates why idioms are an exciting and challenging phenomenon and then gives a definition of the term idiom, a classification of idioms, and an overview of the wide spectrum of idiom analyses found in the linguistic literature.
Chapter 2, “Idioms as evidence for the proper analysis of relative clauses”, shows that the Modification Analysis beats the other two major analyses of restrictive relative clauses (RRCs), namely Raising and Matching, as (i) the latter two lead to a loss of numerous empirical generalizations in syntax and morphology, and (ii) contrary to the assumption in the literature, idioms in RRCs can, in fact, be licensed without literal syntactic movement of the RRC-head, which makes modification fully compatible with idiom reconstruction effects.
Chapter 3, “How frozen are frozen idioms?”, presents new empirical observations on the lexical, morphological, and syntactic flexibility of kick the bucket and displays that this idiom is not completely frozen with respect to its NP complement, the progressive form, and, in some contexts, even passivization. The chapter concludes that analyses of kick the bucket as a single lexical entry should be replaced by analyses of this and other SNDIs with a syntactically regular shape as consisting of individual word-level lexical entries that combine according to the standard rules of syntax.
This idea is taken up in Chapter 4, “The syntactic flexibility of semantically non-decomposable idioms”, which – based on the differences between English and German with regard to verb placement, constituent fronting, and passivization as well as a short outlook on Estonian and French – spells out a combinatorial analysis of SNDIs and augments it with a semantic analysis formulated in Lexical Resource Semantics, according to which some idiom parts make identical semantic contributions to the overall meaning of the idiom. The analysis further suggests that the syntactic flexibility of idioms is due to the semantic and pragmatic constraints on the involved constructions, rather than the syntactic encoding of the idioms.
Chapter 5, “Modification of literal meanings in semantically non-decomposable idioms”, reviews Ernst’s (1981) classical three types of idiom modification (internal, external, and conjunction) to then closely investigate the most challenging type, namely conjunction modification, in SNDIs. Based on naturally occurring examples of four SNDIs (two English, two German), it sketches an analysis in terms of two or more conjoined independent propositions, each of which can be the result of figurative reinterpretation. One of the propositions contains the idiomatic meaning, in (one of) the other(s), the meaning of the modifier applies to the literal meaning of the idiom’s noun.
Chapter 6, “Semantically decomposable idioms in the N-after-N construction”, offers a formal syntactic and semantic account of SDIs like pull strings in the N-after-N construction, as in Kim pulled string after string to get Alex into a good college. While the idiom contributes the type of entity at stake (‘string’ in the case of pull strings), N-after-N contributes that there are several instantiations of that type of entity and that these are subject to temporal or spatial succession. The chapter first summarizes the empirical properties of N-after-N, then provides an account of N-after-N in Head-driven Phrase Structure Grammar (HPSG), presents an updated version of the account of SDIs suggested in Chapter 2 within HPSG, and combines it with the HPSG account of N-after-N.
This dissertation investigates a special class of anaphoric form, yè, in Ewe known as the logophoric pronoun. This research makes a number of novel observations.
In the first chapter, I introduce the reader to the phenomenon under investigation as well as provide information on Ewe and its dialects and, methodology. In Chapter 2, I present the pronominal system of Ewe which is categorised into strong and weak forms following Cardinaletti & Starke (1994) and Agbedor (1996). The distribution of pronouns is outlined which sets the tone for an overview of logophoric marking. In this respect, I present variations in logophoric marking strategies cross linguistically and show that Ewe differs significantly from other pronouns in this category. In an effort to explain the deviant case of yè, I entertain the idea that yè is a pure logophoric pronoun in the sense of Clements (1975) and thus, its additional de re and strict interpretation does not imply non-logophoricity.
Chapter 3 demonstrates that yè is sensitive to contexts which portray the intention of an individual. Following Sells (1987), the antecedent of yè must have an intention to communicate. I broadly categorize logophoric contexts into reportative (direct-indirect speech) or non-reportative (speaker’s mental attitude, reporter’s observation or background knowledge of a situation). Based on this categorization, indirect speech report (Clements 1975), dis- course units such as a paragraph or an episode (Clements 1975), and sentential adjuncts such as purpose, causal and consequence clauses (Culy 1994a) are reviewed. The logophoric pro- noun occurs in the complement of attitude verbs (Clements 1975), also termed logocentric (à la (Stirling 1994)) or logophoric predicates (à la (Culy 1994a)) as well as with non-attitudinal verbs (e.g. va ‘come’ or wO ‘do’ as in sentential adjuncts). I argue contra Clements (1975) and Culy (1994a) that yè can occur with perception predicates. I further provide three new instances of non-reportative contexts which are compatible with yè namely, as-if clauses, benefactive na clauses and alesi ‘how’ clauses. I show, corroborating previous studies that contexts which are necessary for the licensing of yè include all of the aforementioned except causal clauses. Among these contexts, the complementizer be or regarding cases where there is no be, an element in C (due to the Doubly-Filled-Comp Filter (DFCF) c.f. Chomsky & Lasnik (1977)), is sufficient to license yè. Following Bimpeh & Sode (2021), yè is licensed by feature checking (in the spirit of von Stechow (2004)): be bears the interpretatble [log] feature which checks the uninterpretable [log] feature of yè. I include a redefinition of logophoricity as pertaining to Ewe.
Given the disparity found in the literature concerning the interpretation of yè: Ewedome (pronounce EVedome) has only de se readings (Bimpeh 2019); while ‘pure’ Ewe, Mina (variety of Ewe spoken in Togo) Pearson (2015), Danyi (O’Neill 2015) and Anlo (pronounced ANlO) (Satık 2019) has de re readings; chapter 4 aims at lending empirical support to the ungoing discussion by verifying the interpretation of yè. Two acceptability judgment tasks were conducted namely, truth value judgment task and binary forced choice task. The results corroborates Pearson (2012, 2015) and others’ discovery that yè has a de re interpretation in the Ewedome (contra Bimpeh (2019); Bimpeh et al. (2022)), Anlo and Tonu (pronounced TONu) dialects of Ewe.
In chapter 5, I discuss the relation between logophoricity (yè, yè a) and Control (PRO). I show that yè may be restricted to a set of verbs which obligatorily require the morpheme a ‘potential marker’ (Essegbey 2008), in subject position. This set of verbs are those that are known as control verbs c.f. (Landau 1999) in English. As a result of this restriction, research such as Satık (2019) claims that yè a is the overt instantiation of PRO in English. According to the Ewe facts, it appears as though on one hand, yè and PRO share similar properties in logophoric contexts and on the other hand, yè in combination with the potential marker, a also share properties with PRO in subject control environments. Against this background, I discuss the relation between yè, yè a and PRO and show that neither yè in isolation nor yè in combination with a, contrary to Satık (2019), is the overt instantiation of PRO. I clarify that the potential morpheme a is not cliticised or combined with the logophoric yè. The two forms are seperate morphemes. The potential marker a only shows up in control environments because a sub-class of verbs require it for grammaticality purposes. As such, the property of de se-ness does not come from yè by itself, yè a or a but rather from the sub-class of verbs which require the potential marker a...
James Joyce's Ulysses is treated as one of the most influential, paradigmatic texts of high modernism. Novels like Thomas Pynchon’s 1973 Gravity’s Rainbow and David Foster Wallace’s 1996 Infinite Jest, which equally raise claims to being the paradigms of their respective time, are perpetually compared to and measured against Joyce’s epic novel. However, novels like Ulysses, Gravity’s Rainbow and Infinite Jest are usually either grouped together due to their length, complexity and importance, to examine direct allusions in the texts or analyse a rather general “style” or to conversely stress the novels’ singularity and autonomy. I argue that not only can Joyce’s Ulysses, Pynchon’s Gravity’s Rainbow and Wallace’s Infinite Jest be meaningfully put in relation to one another but that their singularity and paradigmatic status in 20th century literature should be understood through the relationality of a Ulyssean Tradition. Novels like Gravity’s Rainbow and Infinite Jest can be fruitfully read in a Ulyssean Tradition. Their singular, paradigmatic aesthetic projects emerge from a reciprocal dialogue with Ulysses in their self-inscription into a Ulyssean Tradition. The intertextual connection of this Ulyssean Tradition is integrally constitutive of the autonomy through which these novels claim the status of singular representations of their respective human condition and thus epic paradigms of a new way of writing the world. By positioning themselves in the literary field alongside Ulysses as the received paradigm of modernism, Wallace in Infinite Jest and Pynchon in Gravity’s Rainbow legitimize their own, independent project and their own claims to paradigmaticness. The Ulyssean Tradition thereby becomes not only a way of writing,a nd this study not merely a study of literary influence, but also a way of reading that can generate new, independent readings through the relationality of a Ulyssean Tradition
Since Vietnamese is an isolating language, word order plays an important role in identifying the function of a particular word. Yet in some contexts word order may be flexible especially in the case of special information-structural settings. Discontinuous noun phrases constitute a specific case of non-canonical word order in Vietnamese.
I have conducted two read-speech experiments in order to find out whether there are prosodic or intonational effects in a comparison between continuous and discontinuous noun phrases in Vietnamese. In the first experiment, speakers from the Northern dialect were recorded and in the second experiment speakers from the Southern dialect. The results showed prosodic differences in the two word order conditions in both dialects. The duration of the classifier is significantly longer (p<0.001, ANOVA calculation) in the case of discontinuous noun phrases and the rising tone (sắc) is clearly articulated as rising. In the case of continuous noun phrases, the duration of the classifier is significantly shorter (p<0.001, ANOVA calculation) and a classifier with rising tone may lose its rising property. These prosodic effects are related to prosodic boundaries. In the case of discontinuous noun phrases, the classifier constitutes the prosodic boundary, whereas with continuous noun phrases, the (right) prosodic boundary occurs further to the right.
I assume that in Vietnamese there is generally a correspondence between syntactic and prosodic structure as in Selkirk (2011) and Féry (2017).
This means that for example the DP hai trái cam ‘two oranges’ (two CLF orange) is matched by a prosodic phrase, thus (hai trái cam)Φ. However, when the noun cam ‘orange’ is separated from the numeral-classifier complex, the noun and the classifier form a prosodic phrase on their own: (hai trái)Φ. It can thus be concluded that intonation effects in Vietnamese are not only present when expressing sentence modality and when changing the role of function words (Đỗ et al. 1998 and Hạ & Grice 2010), but they also play a role in word order change, as in discontinuous nominal phrases.
When it comes to syntactic aspects of discontinuous noun phrases, I discuss whether split constructions in Vietnamese involve movement as proposed by Trịnh (2011) or base-generation as put forward by Fanselow & Féry (2006). I argue for base-generation analysis since the second part of a discontinuous NP (remnant) may also occur outside of discontinuous noun phrases without its head noun and some discontinuous noun phrases do not have a continuous counterpart. My study confirms the connection between syntax and prosody.
The two parts of the discontinuous noun phrase form their own phrases syntactically as well as prosodically.
Attributive participle constructions in German behave like adjectives in terms of inflection and position, but keep their verbal arguments. They can be extended by adjuncts or arguments and these extended attributive present participles mainly occur in written language (Weber, 1994). As the same content can also be expressed in a relative clause (RC), I compare both constructions in order to find out under which conditions a participle construction could lead to processing difficulties and how this relates to RC processing.
Based on previous assumptions for production (e.g. Weber, 1971; Fabricius-Hansen, 2016), three potential factors on the comprehension of prenominal modifiers and RCs are investigated: modifier length, the internal structure and multiple levels of embedding. The hypotheses for an effect on modifier length are mainly based on two processing accounts that make opposite predictions under specific circumstances: memory-based accounts such as the dependency locality theory (DLT) (e.g. Gibson, 2000) and expectation-based accounts such as surprisal (e.g. Levy, 2008). An increase in modifier length results in more intervening material between determiner and noun for the participle construction, contrary to RCs where these elements are adjacent. This separation of the DP could increase memory load. Therefore, longer participles would slow down processing of the noun, while there should be no difference for RCs. Two acceptability judgment experiments showed a tendency for longer participle phrases to receive lower ratings. The modifier length was further investigated in online processing. Contrary to the predicted locality effect, self-paced reading data reveals an anti-locality effect for participle phrases, with lower RTs on the noun when additional material was present inside the modifier. This experiment was followed up by an eye-tracking experiment which replicated the anti-locality effect, but at the participle instead at the noun.
The second factor that was investigated is the argument structure of the participle (or RC verb). My hypothesis is that more “prototypical” adjectives in terms of syntactic structure and semantics are more acceptable and easier to process. Attributive participles are considered hybrids between verbs and adjectives (e.g. Fuhrhop & Teuber, 2000; L¨ubbe & Rapp, 2011) due to their modifier internal verbal function, but adjectival position and agreement with the noun. This double role could lead to difficulties, in particular with a more complex verbal structure. Therefore, the prediction was that the presence of an accusative object inside a participle phrase would lead to lower acceptability ratings and higher reading times in online processing. In the first two acceptability experiments, this prediction was borne out. In addition, an SPR experiment was conducted which manipulated the presence of either an accusative object or adjunct for participles (of verbs that could be used intransitively and transitively) and the corresponding RCs. The experiment showed an effect of the presence of an accusative object on the participle, with higher reading times if an object was present, compared to an adjunct. No such difference was found for the RC verb, which indicates that only participles are processed more slowly when there is an accusative object. An alternative explanation for this finding is the inherent imperfective aspect of the present participle: a direct object could change the event structure in such a way that the aspect no longer matches.
The third factor I investigate is an effect of double embedding on the acceptability of participle phrases and RCs. While double embedded participles are rated lower than double embedded RCs, there is a smaller decrease from single to double embedding for participles than for RCs, contrary to the predictions calculated by the metric of the DLT.
Overall, the results provide evidence for experience-based processing, but they cannot be explained by either memory- or experience-based accounts alone. The effect concerning the presence of an accusative object suggest that properties of the participle distinguish the construction from RCs and affect its processing. The thesis suggests that the latter effect needs to be investigated further in future research. Furthermore, the findings have implications for the role of attributive present participles in German and for hypotheses about similar constructions in other languages.
This work is about resumptive and non-resumptive relative clauses (RCs) in the three big Ibero-Romance languages: Spanish, Portuguese, and Catalan. In (1), the examined structures are exemplified for Spanish: (1a.) No conozco el hombre que viste _ ayer. “I don’t know the man that you saw yesterday.” (1b.) Es este el hombre que le enviaron el libro. “This is the man to whom they sent the book.” (1c.) Es este el hombre a quien le enviaron el libro. “This is the man to whom they sent the book.”
(1a.) displays a non-resumptive, or canonical, RC, which is characterized by the canonical use of a relativizing operator and a gap in the subordinate’s object position, a piece of evidence which has induced most of the generative literature to assume wh-movement of the relative operator in the sense of Chomsky (1977). The last two decades, however, have seen a big debate regarding the exact derivational analysis, starting with Kayne’s (1994) antisymmetry theory and the following focus on reconstruction and anti-reconstruction effects in RCs. This search for the correct starting site of the RC’s head noun has dismissed the original Head External Analysis (HEA) (Chomsky 1965, 1977) and led to the development of a Head Raising Analysis (RA) (Kayne 1994, Bianchi 1999, a.o.) and a Matching Analysis (MA) (Munn 1994, Sauerland 1998, a.o.). The discussion in this work argues that the data on reconstruction and anti-reconstruction effects are not sufficiently clear and reliable in order to adopt one of the head-internal analyses, i.e. a HEA or a MA. Instead, the work follows a variant of the HEA proposed for Portuguese by Rinke & Aßmann (2017), which adheres to standard assumptions about Romance syntax, and avoids the empirical problems that the other proposals have to face. Arguing that the HEA holds for all Ibero-Romance languages, this work also takes a stance in the debate around the categorical status of the relativizing element que and argues that it is always a D-element, and never of category C, i.e. there is no such thing as a relativizing complementizer (cf. also Kayne 2010, Kato & Nunes 2008, Poletto & Sanfelici 2018).
The work argues that wh-movement in a HEA fashion is the correct analysis also for resumptive relative clauses as in (1b., c.), which crucially lack a gap in argument position but show a resumptive pronominal element instead. Furthermore, it takes advantage of the fact that the choice of such genetically closely related languages like Spanish, Portuguese, and Catalan enables research to address the phenomenon under consideration from a microcomparative perspective, which is “the closest we can come … to a controlled experiment in comparative syntax” (Kayne 2005: 281-282). The descriptive literature suggests that, at least for Spanish and Catalan, there are two types of a resumptive RC structure available: a simple resumption as in (1b.), including mere que, and a complex resumption structure which displays a more complex relativizer like a quien in combination with a resumptive pronoun (1c.). However, a corpus study carried out for this work reveals that speakers of the three languages behave alike insofar as the only resumptive RC used in spontaneous speech is a simple-resumption structure, while complex resumption never occurs. Additionally, a multivariate analysis shows that in all three languages, grammatical case is the most important factor when it comes to the possibility of a resumptive structure in RCs: with a dative argument, simple resumption is obligatory, while for accusative and nominative arguments, resumption is optional. The discussion concludes that simple and complex resumption constitute different phenomena also on a structural level: the latter one is argued to be a subcase of clitic doubling, and therefore, receives an analysis along the lines of Pineda (2016), who argues against a dative alternation in Romance languages and locates the (non-)realisation of the dative clitic in a transitive clitic-doubling structure outside of syntax, it being a case of silent variation along the lines of Sigurðsson (2004) and Kayne (2005). From this perspective, it follows naturally that in Portuguese, complex resumption structures are ungrammatical. Simple resumption, on the other hand, which is a possible structure in all three languages, is argued to represent the phonological counterpart of “scattered deletion”, i.e. the preferred interpretation for an A’-chain according to Chomsky (2003): in the operator position SpecCP, every feature except for the operator feature is deleted, resulting in the phonological outcome que, while in the variable position, everything but the operator is interpreted, resulting in a pronominal element according to the argument’s phi-features.
This work takes a stance in the latest topics on generative analyses for relative clauses. Using not only theoretical considerations but conclusions drawn from empirical data on three languages, it offers a new perspective on pending questions and proposes to take a fresh look on supposedly outdated analyses.
This thesis examines the referential properties of prenominal possessive modifiers in Serbian. The focus of the investigation is on the configurations that have been claimed to violate Binding Principles B and C: lexical or pronominal possessives modifying a noun in subject position binding a pronoun or an R-expression in object position. Such constructions have been claimed to be ungrammatical in Serbian due to the alleged adjectival status of the possessive and its respective syntactic position as NP-adjoined (Despić 2013).
The present thesis takes up the ongoing debate about the categorial status of Serbian possessives as adjectives or determiners. Based on several arguments, such as word order, binding of anaphora, coordination, and the fact that they are typically represented by either nouns or pronouns, it is concluded that possessives rather behave like full noun phrases than adjectives. Therefore, I analyse possessives as DPs from a categorial point of view.
In a second step, the syntactic position of the possessives within the Serbian noun phrase has been investigated. Based on theoretical arguments (cf. Bašić 2004) and empirical evidence, I propose a structural position that would accommodate the binding facts and the referential possibilities in these configurations. In line with Kayne (1994), Bernstein and Tortora (2005) and Alexiadou et al. (2007), I assume that possessives occupy SpecAgrP in Serbian, where they move from their base position (SpecPossP). Thirdly, I question the (im)possibility of coreference with possessives in comparison to ‘typical’ binding constructions without possessives by providing empirical evidence from three experimental studies, showing that coreference between possessive modifiers and objects is indeed available in Serbian.
The results from Experiment 1 (a picture selection task) have shown that coreference between a lexical possessive and a (clitic or strong) pronoun is allowed in Serbian. Further, there is a tendency that the coreferential reading is preferred with clitics, while the disjoint reference is preferred with strong pronouns. The fact that coreference is possible, does not necessarily mean that it is always available as the only interpretation, but can be influenced by other (pragmatic) factors. The same is observed in Experiments 2 and 3 as coreference was chosen between pronominal possessives modifying a noun in subject position and R-expressions but rejected between pronouns and R-expressions in a forced-choice task, suggesting a structural difference – no c-command – in the former case. The results from the self-paced reading task corroborate this finding.
Importantly, the experimental results provide evidence that possessive configurations are not violating Binding Principles B and C. This implies that Serbian possessive constructions do not c-command out of the noun phrase, as predicted by the proposed syntactic analysis.
The findings from all experiments contribute to the bigger picture concerning the nature and behaviour of Serbian possessives and cast doubt on the cross-linguistic DG/AG parameter. Instead, the theoretical arguments and the empirical results from the experiments rather speak for a parallel structure of possessive noun phrases in Serbian and English and ultimately in favour of the Universal DP Hypothesis.