The grammar of modification is highly complex and raises numerous questions about the relation between meaning and form. This dissertation provides a study of how modified noun phrases are interpreted and examines the consequences of these results for the syntax of the nominal domain. The discussion centers on two types of modification: superlatives and stacked modification. The data comes primarily from English, but other languages are also discussed. There is initial evidence that the main claims hold across a wide range of languages. The common view on superlatives is that they have two types of interpretations which are the result of a scope ambiguity and that the contrast between them needs to be captured by means of syntactic devices. Contra this standard approach I propose a saliency theory of superlatives which claims that there is no categorical difference between these two interpretations and where the variation in the meaning of superlatives is purely pragmatic in nature. Under this view the meaning of superlatives is a function of the properties of the surrounding discourse and the context-sensitivity of superlatives is subsumed to the more general phenomenon of context-dependency in the interpretation of natural language quantifiers. The saliency theory differs from other analyses that have adopted a discourse approach in that the so-called comparative reading does not depend on the presence or interpretation of focus. Previous approaches to multiple adjectives analyzed their order in terms of the semantics of individual adjectives. I present a new set of data which shows that this is insufficient and propose an explanation that takes into account the meaning of the whole nominal phrase. This result has consequences for how the architecture of grammar should be conceived. In particular, it shows that principles of syntactic well-formedness can sometimes be sensitive to compositional semantic interpretation, as well as pragmatic information. This is in contradiction to many contemporary approaches to grammar where the semantic component has no influence on the syntactic one. / text

Este trabalho pretende analisar os diferentes empregos de adjetivação no romance modernista Memórias sentimentais de João Miramar, de Oswald de Andrade, bem como a expressividade alcançada por eles dentro de seu universo textual. Procura, também, investigar a contribuição dos usos dos adjetivos para o traço de estilo mais marcante da obra: a brevidade. / This research intends to analyse different adjective uses in the modernist novel Memórias sentimentais de João Miramar, by Oswald de Andrade, as well as the expressiveness reached by them in their textual universe. It will also try to investigate the contribuition of the adjective uses to the novel more characteristic style aspect: conciseness.

The present study is a corpus-based study which examines social roles constructed in the British nineteenth-century courtroom. To discover the prevalent social roles in British nineteenth-century society the present study focuses on premodifying adjectives characterizing men and women. The method of classification is through semantic domains. The study shows that the social roles of men and women are more similar than the findings of previous research have demonstrated.

Vietovardis Lietuva rašytiniuose šaltiniuose minimas nuo 1009-ųjų metų. Iš Lietuvių kalbos žodyno (LKŽe) buvo išrinkti 58 dariniai turintys kamieną lietuv- ir suklasifikuoti kategorinėmis klasėmis (kalbos dalimis). Juos sudaro 31 daiktavardis, 11 būdvardžių ir 16 veiksmažodžių. Iš daiktavardžio Lietuva išvesti 5 dariniai. Pastebėta, kad vienintelį darinį lietuvis išvestą iš daiktavardžio Lietuva supa didžiausias semantinis ir darybinis giminiškų žodžių laukas. Jį sudaro 22 dariniai (15 vedinių ir 7 dūriniai): 18 daiktavardžių, 1 būdvardis ir 3 veiksmažodžiai. Tolesnės darybos rezultatas yra dar 31 darinys: 10 daiktavardžių, 9 būdvardžiai ir 12 veiksmažodžių. Iš vietovardžio Lietuva sudaryti dariniai kartais laikytini darybiniais sinonimais arba darybiniais variantais: sudarytos 4 DS ir 3 DV eilės. Sudarytos 4 leksinių sinonimų eilės. LKŽe neužfiksuoti deminutyvai, išskyrus daiktavardį lietuvaitis, -ė. / The name of Lithuania in hand-written sources is mentioned since 1009. From Lithuanian dictionary (LKŽe) it has been selected 58 word-formations having a root “lietuv-“ and divided in classes on categories (in parts of speeches). It includes 31 noun, 11 adjectives and 16 verbs. From a noun Lietuva are formed 5 word-formations. It is established that unique word-formation “lietuvis“ formed from noun Lietuva surrounds a huge field of related words. It is made by 22 word-formations (15 – derivatives and 7 compounds): 18 nouns, 1 adjective and 3 verbs. The result of further formation is 31 word-formation: 10 nouns, 9 adjectives and 12 verbs. Word-formations formed from a word “Lietuva“ sometimes are considered as the formed synonyms or formed cases: are created 4 DS and 3 DV turns. It was created 4 turns of lexical synonyms. There are no more diminutives in the Lithuanian dictionary (LKŽe) except the noun lietuvaitis, -ė.

Este trabalho pretende analisar os diferentes empregos de adjetivação no romance modernista Memórias sentimentais de João Miramar, de Oswald de Andrade, bem como a expressividade alcançada por eles dentro de seu universo textual. Procura, também, investigar a contribuição dos usos dos adjetivos para o traço de estilo mais marcante da obra: a brevidade. / This research intends to analyse different adjective uses in the modernist novel Memórias sentimentais de João Miramar, by Oswald de Andrade, as well as the expressiveness reached by them in their textual universe. It will also try to investigate the contribuition of the adjective uses to the novel more characteristic style aspect: conciseness.

ESL grammar books have lists of present and past participial adjectives based on author intuition rather than actual word frequency. In these textbooks, the –ing and –ed participial adjectives derived from transitive verbs of state and emotion are presented in pairs such as interesting/interested, boring/bored, or surprising/surprised. This present study used the Corpus of Contemporary American English http://corpus.byu.edu/coca/ (COCA) to investigate the overall frequency of participial adjectives in use as well as their frequency within certain varieties of contexts. The results have shown that among most frequently used participial adjectives there are not only the participial adjectives derived from transitive verbs of psychological state, such as interesting/interested, but also the participial adjectives derived from transitive verbs of action with their intransitive equivalents, such as increasing/increased. The data also revealed that many participial adjectives lack corresponding counterparts and thus cannot be presented in –ing/-ed or -en pairs (e. g., existing, ongoing, concerned, supposed). Finally, a majority of the differences between participial adjectives, including the differences between present (-ing) and past (-ed or -en) participial adjectives, are reflected in their collocations. This study suggests that a new approach of teaching participial adjectives along with their collocations in relation to their frequencies in particular contexts can help second language learners develop awareness of how and when these participial adjectives should be used to convey an individual’s intended meaning in a native-like manner

In the present study, the linguistic performance of 11 Swedish high-school students of Spanish as second language (L2)/foreign language (FL) of 3 different levels, between 16-18 years old, has been measured through analysis of experimental elicited speech samples from a picture description task, designed to elicit placement of adjectives and dimensional adjectives in Spanish. The task provided 3 images with phrases in the source language and the participants were asked to translate it into the target language (TL). Subsequently, 9 images without text were presented and the students were asked to describe the images in the TL. The intention was to measure the grammatical accuracy concerning the placement of adjectives and the possible transfer –from Swedish to Spanish– of semantical markers regarding dimensional adjectives. The results from the performance analysis of the learners responses indicate that in beginner learners of Spanish, the semantic markers of dimensional adjectives have been transferred from Swedish to Spanish. One example that reflects this is the predominant use of the form largo ‘long’ instead of alto ’high/tall’ regarding the height of a person. The beginners also exhibit transfer in the placement of adjectives. These types of errors found in the speech of beginner students were not salient to the same extent in more advanced students.

This dissertation develops a semantic model of gradable adjectives such as ‘tall’, ‘good’, ‘big’, ‘heavy’, etc., within a formal semantic theory of locatives we call Locative Structure Semantics (LSS). Our central hypothesis is that gradable adjectives are, semantically, a species of locative expression. The view of gradable adjectives as locatives is inspired by the vector-based semantic models of Vector Space Semantics (VSS), as well as the notion of perspective or point of view, as found in Leonard Talmy’s research on spatial expressions (Talmy [153]) and the tradition of Situation Semantics (cf. Barwise and Perry [9, p. 39]). Following Barwise and Seligman [11], we construe the contextual variability that characterises gradable adjectives in terms of shifts in cognitive perspective. We argue that perspectives are a formal part of a semantic representational structure that is shared by expressions from several different domains, which we refer to as a locative structure (L-structure). The notion of an L-structure is influenced by Reichenbach’s notion of tense, and can be thought of as a generalisation of the Reichenbachian notion of tense to the realm of concepts. Reichenbach [134] proposed that each temporal expression is associated with three time points: a speech point, S, an event point, E, and reference point, R, where E refers to the time point corresponding to the event described by the tensed clause, S is (usually) taken to be the speaker’s time of utterance, and R is a temporal reference point relevant to the utterance. In LSS we extend this tripartite scheme to locative expressions in general, to which we assign a ternary structure comprising a Perspective, a Figure, and a Ground, represented symbolically as P, F, and G, and which are generalisations of the Reichenbachian S, E, and R, respectively. We show that a formal semantics based on L-structures enables us to capture important crosscategorial similarities between gradable adjectives, tenses, and spatial prepositions.

The use of gradable adjectives and adverbs represents an important means of expressing comparison in English. The grammatical forms of comparatives and superlatives are used to express explicit orderings between objects with respect to the degree to which they possess some gradable property. While comparatives are commonly used to compare two entities (e.g., “The blue whale is larger than an African elephant”), superlatives such as “The blue whale is the largest mammal” are used to express a comparison between a target entity (here, the blue whale) and its comparison set (the set of mammals), with the target ranked higher or lower on a scale of comparison than members of the comparison set. Superlatives thus highlight the uniqueness of the target with respect to its comparison set. Although superlatives are frequently found in natural language, with the exception of recent work by (Bos and Nissim, 2006) and (Jindal and Liu, 2006b), they have not yet been investigated within a computational framework. And within the framework of theoretical linguistics, studies of superlatives have mainly focused on semantic properties that may only rarely occur in natural language (Szabolsci (1986), Heim (1999)). My PhD research aims to pave the way for a comprehensive computational treatment of superlatives. The initial question I am addressing is that of automatically extracting useful information about the target entity, its comparison set and their relationship from superlative constructions. One of the central claims of the thesis is that no unified computational treatment of superlatives is possible because of their great semantic complexity and the variety of syntactic structures in which they occur. I propose a classification of superlative surface forms, and initially focus on so-called “ISA superlatives”, which make explicit the IS-A relation that holds between target and comparison set. They are suitable for a computational approach because both their target and comparison set are usually explicitly realised in the text. I also aim to show that the findings of this thesis are of potential benefit for NLP applications such as Question Answering, Natural Language Generation, Ontology Learning, and Sentiment Analysis/Opinion Mining. In particular, I investigate the use of the “Superlative Relation Extractor“ implemented in this project in the area of Sentiment Analysis/Opinion Mining, and claim that a superlative analysis of the sort presented in this thesis, when applied to product evaluations and recommendations, can provide just the kind of information that Opinion Mining aims to identify.

The aim of this thesis is to examine preferences for personality traits within the spectrum of extroversion and introversion. Based on previous research, the hypothesis of this study was that extroversion was valued higher than introversion. The differences between men and women’s preferences were also examined. The material being used was based on research on the fundamental lexical hypothesis and its semantic similarity to the Big Five. Seventy two students filled in a questionnaire where they were asked to rate 18 personality adjectives as positive or negative. The results indicate a preference for extroversion which supports the hypothesis. However, no significant difference between men and women could assuredly be established. / Syftet med detta examensarbete är att undersöka preferenser för personlighetsdrag inom spektret extroversion och introversion. Med utgångspunkt från tidigare studier ställdes hypotesen att extroversion värderas högre än introversion. Dessutom undersöktes om preferenserna skiljer mellan kvinnor och män. Materialet som användes var baserad på studier om den fundamentala lexikala hypotesen och dess semantiska likhet till the Big Five. Sjuttiotvå studenter fick besvara en enkät där de ombads bedöma 18 personlighetsadjektiv som positiva eller negativa. Resultatet visar en generell preferens för extroverta personlighetsadjektiv vilket stödjer studiens hypotes. Någon säker skillnad mellan mäns och kvinnors preferenser gick inte att iaktta.