Several studies have shown that women tend to use more personal pronouns and therefore show more involvement with the reader. This paper examines the differences between male and female editors’ letters in magazines. The study applied the method of corpus linguistics in order to examine forty editor’s letters twenty from the male-targeted magazine Gentlemen’s Quarterly and twenty from the female-targeted magazine Harper’s Bazaar. First person singular and second person singular pronouns were examined to determine whether the female editor showed more involvement with the reader than the male editor. The result shows that the male editor from the Gentlemen’s Quarterly shows more involvement with the reader than the female editor from Harper’s Bazaar, which clashes the findings of previous studies.

L'acquisition de connaissances relatives aux constructions verbales est une question importante pour le traitement automatique des langues, mais aussi pour la lexicographie qui vise à documenter les nouveaux usages linguistiques. Cette tâche pose de nombreux enjeux, techniques et théoriques. Dans le cadre de cette thèse, nous nous intéressons plus particulièrement à deux aspects fondamentaux de la description du verbe : la notion d'entrée lexicale et la distinction entre arguments et circonstants. A la suite de précédentes études en traitement automatique des langues et en linguistique nous faisons l'hypothèse qu’il n’y a pas de distinction marquée entre homonymes et quasi-synonymes ; de même, nous posons qu’il existe un continuum entre arguments et circonstants. Nous proposons une chaîne de traitement complète pour l'acquisition de schémas prédicatifs verbaux en japonais à partir d'un corpus non étiqueté de textes journalistiques. Cette chaîne de traitement intègre la notion d'argumentalité au processus de création des entrées lexicales et met en œuvre une modélisation de ces deux continuums. La ressource produite a fait l'objet d'une évaluation comparative qualitative, qui a permis de mettre en évidence la difficulté des ressources linguistiques à décrire de nouvelles données, plaidant par là même pour une lexicologie s'inscrivant dans le cadre épistémologique de la linguistique de corpus. / Lexical knowledge acquisition of verbal constructions is an important issue for natural language processing as well as lexicography, which aims at referencing emerging linguistic usages. Such a task implies numerous challenges, technical as well as theoretical. In this thesis, we had a closer look at two fundamental aspects of the description of the verb: the notion of lexical item and the distinction between arguments and adjuncts. Following up on studies in natural language processing and linguistics, we embrace the hypothesis that there is no clear distinction between homonyms and quasi-synonyms, and the hypothesis of a continuum between arguments and adjuncts. We provide a complete approach to lexical knowledge acquisition of verbal constructions from an untagged news corpus. The acquisition process makes use of the notion of argumenthood, and builds models of the two continuums. Our lexicon has been evaluated on a qualitative and comparative basis. Siding with lexicography anchored in the theoretical framework of corpus linguistics, we show the difficulty of using lexical resources to describe as yet unseen data.

This dissertation addresses an unanswered question in Arabic psycholinguistics. Arabic words are characterized by their nonconcatenative structure, in which a consonantal root that encodes the main semantic content is interleaved with a derivational pattern (called "binyan", pl. "binyanim"), which is typically vocalic but may also contain consonantal elements and contributes grammatical information. The canonical example of the Semitic root and binyan system is the combination of root /ktb/ which denotes the broad semantic sense of "writing" with verbal binyan /CaCaC/ (with Cindicating a root consonant) to form [katab] "he wrote" and with nominal place binyan/maCCaC/ to form [maktab] "office". Although significant work has been done on the psycholinguistic reality of Arabic morphemes by exploring various phonological, morphological and semantic features across numerous experimental modalities in both the visual and auditory domains (Boudelaa and Marslen-Wilson, 2004, 2005, 2011), no study has investigated the roles of base/morpheme frequency and surface/word frequency and their implications for underlying morphological structure in the lexicon of Arabic as has been done for English, Dutch, and Finnish (Baayen et al., 1997; Alegre and Gordon, 1999; New et al., 2004; Taft, 1979, 2004). Competing models of word recognition propose various integrations of morphology. Whole-word models suggest that there are no separate representations for morphemes, and that the co-activation of related words can be attributed to similarity in form and meaning (Norris and McQueen, 2008; Tyler et al., 1988). Decomposition models posit that words are recognized by accessing the words' constituent morphemes (Meunierand Segui, 1999; Taft et al., 1986; Wurm, 2000). Hybrid models incorporate multiple pathways to recognition. Words are either recognized holistically or by their constituent morphemes depending on multiple factors (Balling and Baayen, 2008; Taftand Nguyen-Hoan, 2010; Lopez-Villasenor, 2012). Of most relevance to the current study is the role of the productivity of the words' derivational affixes: words with unproductive affixes are processed holistically whereas words with productive affixes are processed as a function of their morphemes. This dissertation presents results from four auditory lexical decision experiments performed with native Jordanian speakers in Amman, Jordan, and provides evidence that binyan productivity determines whether the frequency of the base morpheme affects the speed of recognition. By manipulating root and word frequency for three binyanim, one more productive and two less productive, I provide evidence that verbs in the productive binyan are fully decomposable during lexical access and verbs in less productive binyanim are recognized holistically. For a more productive binyan, I examine Binyan I of the form /CaCaC/, and two less productive binyanim are Binyan VIII of the form /iCtaCaC/and Binyan X of the form /staCCaC/. These results together support a hybrid model of lexical access in which some words are recognized via decomposition into the morphemes they are composed of, and others are recognized by their whole word form. These results are consistent with those of Balling and Baayen (2008); Taft and Nguyen-Hoan (2010); Bertramet al. (2000), among others, as derivational affix productivity is the deciding factor determining whether a word will be recognized holistically or decomposed during lexical access.

In this multi-methodological (Gee, 2011; Hult & D. Johnson, 2015) study, I examine Richard Ruiz's (2014) original concept of the epiphenomenal nature of language in language policy and planning (LPP) across social scales (Hult, 2013) in Hong Kong. While research in Hong Kong has focused on interactions between schools, teachers, students, parents, business, and the government, the work on non-profits and non-governmental organizations (NGOs) remains a neglected social scale. Addressing this gap, I examine the role of a local NGO, Hong Kong Unison (Unison), as a social actor involved in the negotiation of language-in-education policies for the city's ethnic minority students. Through the collection of one decade of publically accessible documents, I created a corpus of Unison's work. Corpus linguistics approaches and a wider-angle perspective to critical discourse analysis (cf. Tian, 2006, 2008) were combined in order to highlight salient patterns and discourses within the data (cf. Baker, 2016). Corpus and discursive analyses indicate that Unison is primarily involved in transforming language policies through their active role in increasing public awareness about the social, political, and educational difficulties ethnic minority students encounter in Hong Kong. Furthermore, the NGO’s ideologies reveal the epiphenomenal nature of LPP. Epiphenomenality reflects how decisions made about language are influenced and shaped by non-linguistic phenomena. Unison's negotiation of LPP demonstrates how their decisions about language are connected to issues of equality, justice, economic opportunity, educational attainment, and social advancement. These ideologies manifest themselves in dialogue across social scales, demonstrating Unison's impact in negotiating LPP in Hong Kong. This study, while providing more insight into LPP research by examining the role of a local NGO, continues to raise questions on how to best understand how multiple scales intersect in the policymaking process, and how the epiphenomenal nature of language shapes decision-making.

(in English): French grammars define the word class adverbs as a very heterogeneous set distinguished particularly by its invariability and intransitivity. The aim of this thesis is to study the valency structure of certain French adverbs forming such a subset which to certain extent denies aforesaid definition. Therefore, the definition of adverbial valency and the characteristic of "inheritance" and "fading" of adverbial valency potential after a transformation from verbs and adjectives will be given. The outcomes of the theoretical part will then be verified by analysing the data presented in the electronic dictionary le Petit Robert completed by qualitative and quantitative analyses based on two written French corpuses: Frantext and L'Est Républicain. To describe the "inheritance" and "fading" of valency potential, the same method will be used to depict the set of the adverbs derived from the adjectives, which take prepositional phrases headed by the preposition de as their complement. This work will be based not only on the theory of general linguistics and grammar, but also on the knowledge of mathematical and corpus linguistics. Klíčová slova (anglicky): valence, adverbs, French, corpus linguistics.

This article presents the findings of a corpus linguistic analysis of the hashtags #mansplaining, #manspreading, and #manterruption, three lexical blends which have recently found widespread use across a variety of online media platforms. Focusing on the social media and microblogging site Twitter, we analyze a corpus of over 20,000 tweets containing these hashtags to examine how discourses of gender politics and gender relations are represented on the site. More specifically, our analysis suggests that users include these hashtags in tweets to index their individual evaluations of, and assumptions about, "proper" gendered behavior. Consequently, their metadiscursive references to the respective phenomena reflect their beliefs of what constitutes appropriate (verbal) behavior and the extent to which gender is appropriated as a variable dictating this behavior. As such, this article adds to our knowledge of the ways in which gendered social practices become sites of contestation and how contemporary gender politics play out in social media sites.

No ensino de língua nacional, concordância é um dos tópicos em cujo aprendizado observa dificuldade por parte dos discentes, principalmente pelo grande número de regras facultativas das gramáticas, que muitas vezes não levam em conta o uso formal real da língua. Este trabalho visa a descrever esse uso, a partir da observação de um corpus do caderno opinião de jornais de grande circulação, confrontando os resultados com as prescrições da norma gramatical escolar, a fim de separar, em tais prescrições, a parte aproveitável da não coincidente com a realidade do corpus, se for o caso. Pretende-se, dessa forma, contribuir para a boa qualidade do ensino da língua portuguesa nos níveis fundamental e médio, especificamente no que se refere à concordância / In first language teaching, topics related to verbal agreement present a high degree of difficulty for students to master, mainly because of optional rules that only emphasize a strictly grammatical teaching and disregards reality in the stantard use of language.Through the analysis of a language corpus extracted from O Globo and Folha de São Paulo, the two most notorious newspapers in Brazil, this work aims at verifying trends in the standard use of language as to verbal agreement, comparing the corpus data with rules prescribed by traditional grammarians. The final goal is to demonstrate that it is possible to minimize the overwhelming number of optional rules whose main application is to establish paradigms that hinders the performance of students proficient in their language.In short, the research was conducted both qualitatively and quantitatively, and is based on prescriptive and descriptive approaches to language teaching and also on the author's experience in educational institutions. The conclusion is that, with the help of this comparative study, the construction of a quality teaching can be achieved through more interaction

A presente pesquisa tem por objetivo investigar como a palavra paz é entendida, em termos de conceito, pelo Conselho de Segurança da Organização das Nações Unidas. Para tanto, são analisados trinta e sete relatórios oficiais produzidos pelo Conselho de Segurança, no período de agosto de 1994 a junho de 2009, acerca das missões de paz realizadas em trinta e uma regiões/países que apresentavam ameaça à paz e à segurança internacionais durante aquele período. De acordo com a Conselheira Gilda Santos Neves, chefe da Divisão das Nações Unidas do Ministério das Relações Exteriores, em seu texto O Brasil e a Criação da Comissão para a Consolidação da Paz (2008), a paz é algo que se consolida e não se constrói. Tal posição norteia a presente pesquisa, uma vez que o objetivo aqui é mapear as expressões linguísticas realizadas através da palavra paz. As bases teóricas desta pesquisa encontram-se fundamentadas na teoria da metáfora cognitiva, de Lakoff e Johnson (1980), bem como no estudo de Deignan (2005) em seu livro intitulado Metaphor and Corpus Linguistics, que visa a fornecer os benefícios que a abordagem cognitiva de metáforas pode obter através da análise de corpora digitalizados. Após compilar os relatórios do Conselho de Segurança e prepará-los para serem lidos pelo programa computacional WordSmith Tools 3.0, foram extraídas todas as ocorrências da palavra paz dos referidos relatórios. Das 686 ocorrências geradas, foram deixadas para análise somente aquelas com sentido metafórico e, no total, nove esquemas conceptuais foram construídos. A pesquisa feita sugere que, para o Conselho de Segurança, a paz é algo profundamente desejado tanto pela população das zonas de conflito quanto pela comunidade internacional. No entanto, a paz não é facilmente construída ou estabelecida. Alcançar a paz implica seguir um processo com diferentes etapas, ou seja, com início, meio e fim, bem como superar obstáculos e retrocessos que surgem no meio do caminho. Para tanto, diversos investimentos têm de ser feitos por todos aqueles envolvidos e realmente interessados na paz mundial. Por fim, vê-se que a visão da Conselheira Gilda Santos Neves, de acordo com as metáforas aqui analisadas, está correta, já que, conforme apontam os resultados do presente estudo, o conceito de paz, para o Conselho de Segurança, não é o de algo a ser construído do zero / The present research aims at investigating how the word peace is understood, in terms of meaning, by the Security Council of the United Nations. In order to do so, thirty seven official reports written by the Security Council, from August 1994 to June 2009, about the peace missions in thirty one areas/countries which presented threat to international peace and to international security during that period are analyzed. According to the Counselour Gilda Santos Neves, Head of UN Division, Ministry of External Relations, in her book O Brasil e a Criação da Comissão para a Consolidação da Paz (2008), peace is something which is consolidated, not built. This statement guides the present research, as the aim here is to map the linguistic expressions realized through the word peace. This research draws on the theory of the cognitive metaphor, by Lakoff and Johnson (1980), as well as the study proposed by Deignan (2005) in her book entitled Metaphor and Corpus Linguistics, which aims at providing the benefits the cognitive metaphor approach can obtain through the analysis of digitalized corpora. After compiling the reports of the Security Council and preparing them to be read by the software used, all the occurrences of the word peace were extracted. From 686 occurrences found, only the ones with metaphorical meanings were taken into consideration and nine conceptual metaphors were created altogether. The research suggests that, for the Security Council, peace is deeply desired not only by the population of the conflict zones but also by the international community. In spite of this fact, peace is not easily built or established. Achieving peace implies following a process with different phases, as well as overcoming obstacles and drawbacks which appear in the middle of the course. Therefore, several investments must be made by everyone involved and interested in global peace. Finally, it can be concluded that, according to the metaphors analysed here, the point of view of the Counselor Gilda Santos Neves is right, as the results of this study show that the concept of peace, for the Security Council, is not of something which must be built from its very beginning

Using the combined approach of Variationist Sociolinguistics and Cognitive Linguistics, this thesis undertakes the classification and analysis of certain prepositions in spoken Quebec French. The study examines 21 interviews that make up part of the Corpus de français parlé au Québec (CFPQ). The aim of this thesis is to examine the use of the variables expressing the concept of ‘possession’, and those equivalent to English before/in front of and after/behind. These three variables are represented as (POSS), (ANTE) and (POST). An initial quantification of the variants is carried out, which establishes the contexts of production, and helps determine the areas of linguistic analysis to be explored. For the (POSS) variable, the data is examined in terms of linguistic factors such as the reference of the possessor, the avoidance of hiatus, and inalienable/alienable possession. Interpersonal variation is also considered, including age and gender in addition to level of education. From the Cognitive Linguistic perspective, we investigate ‘reference point theory’ and how it can shed light on the alternation between the variants. The (ANTE) and (POST) variables are studied in terms of the type of reference (i.e. locative or temporal), the locating noun category, and the age, sex, and level of education of the speakers. The Cognitive Linguistic theory of ‘subjectification’ is also considered for these two variables. For the (POSS) variable, the reference of the possessor and the level of education are seen to be important factors for the use of possessive à. In addition, the ‘reference point theory’ contributes to our understanding of the use of this variant. With the (ANTE) and (POST) variables certain variants are seen to be employed both with and without an overt complement. The variant devant is predominantly found in contexts involving narrative discourse, and the variants en avant and en avant de are preferred for locative reference. Once again, the Cognitive Sociolinguistic approach highlights the possibility that the difference in variant choice is linked to the speakers’ cognitive construal of the situation.

Second language (L2) learners notoriously have trouble using articles in their target languages (e.g., a, an, the in English). However, researchers disagree about the patterns and causes of these errors. Past studies have found that L2 English learners: *Predominantly omit articles (White 2003, Robertson 2000), *Overuse the (Huebner 1983, Master 1987, Parrish 1987, Tarone & Parrish 1988, Thomas 1989, Ionin 2003), or *Overuse a (Leung 2001). Previously proposed explanations of the causes of article errors include: *Learners have incorrect or incomplete semantic representations (Tarone & Parrish 1988, Hawkins & Chan 1997, Goto Butler 2002, Ionin 2003), or *Learners have complete, correct semantic representations for articles, but difficulty choosing the lexical form during production due to stress on mental processing or phonological limitations (Lardiere 1998, Bruhn de Garavito & White 2000, White 2003, Goad, White, & Steele 2003). Prior studies have focused on articles, which identify discourse relationships, but have not considered other morphemes that do so as well, such as pronouns and demonstratives. Furthermore, they have focused on L2 errors in isolation and not in the context of a full discourse or contrasted with first language (L1) input. This study examined the use of articles and other discourse morphemes in 20 L1 and 20 L2 English essays. L2 essays were produced by L1 Chinese and Korean writers at two proficiency levels. The essays' noun phrases (NPs) were marked for part-of-speech, co-reference, syntactic position, and other discourse-relevant features. L2 errors were identified and categorized. Frequency data showed that L2 proficiency level more often indicated significant differences in discourse construction than L1. No significant difference between L2 and L1 writers was when considering all articles together. Breaking this down, students used a/an significantly less than L1 writers, but the use of the was not significantly different. In contrast, the error analysis showed most L2 mistakes being made in the use of the, with almost none in the use of a/an. Together the frequency and error data give a richer understanding of discourse and article use in L2 production.