Press releases serve as a major influence on public opinion of a politician, since they are a primary means of communicating with the public and directing discussion. Thus, the public’s ability to digest them is an important factor for politicians to consider. This study employs several well-studied measures of linguistic complexity and proposes a new one to examine whether politicians change their language to become more or less difficult to parse in different situations. This study uses 27,500 press releases from the US Senate between 2004–2008 and examines election cycles and natural disasters, namely hurricanes, as situations where politicians’ language may change. We calculate the syntactic complexity measures clauses per sentence, T-unit length, and complex-T ratio, as well as the Automated Readability Index and Flesch Reading Ease of each press release. We also propose a proof-of-concept measure called logical complexity to find if classical Boolean logic can be applied as a practical linguistic complexity measure. We find that language becomes more complex in coastal senators’ press releases concerning hurricanes, but see no significant change for those in election cycles. Our measure shows similar results to the well-established ones, showing that logical complexity is a useful lens for measuring linguistic complexity.

This study is an examination of how a word’s syntactic category, word length, and usage frequency might impact a speaker’s use of communicative pause. Previously collected between and within utterance language samples from 21 people with aphasia (Harmon, 2018) were evaluated in this study. Participants consisted of 11 individuals diagnosed with mild or very mild aphasia and 10 individuals with moderate aphasia;15 who exhibited fluent subtypes and 6 non-fluent subtypes of aphasia. Data from the Corpus of Contemporary American English (COCA) was used to code the word frequency and syntactic category of each word in the language samples. Generally, speakers with both non-fluent and fluent aphasia produced more monosyllabic words of very high frequency with a greater percentage of function words than content words. Analyses revealed no significant correlations between the pause duration for either the word length or word frequency for either group of speakers. In relation to syntactic category, no significant differences in pause duration were found between content and function words in the between utterance condition. However, non-fluent speakers preceded content words with significantly shorter pause durations within utterances when compared with the function words. Due to differences in sample sizes between the speaker and syntactic groups, non-parametric statistics were used for some comparisons. In addition, this study does not fully account for the influence of fillers and incomplete words. Despite these limitations, this study will contribute to the research regarding communicative speech pause in speakers with aphasia and provide insight into more useful diagnostic and treatment strategies.

Languages are culturally transmitted through a repeated cycle of learning and communicative interaction. These two aspects of cultural transmission impose (at least) three interacting pressures that can shape the evolution of linguistic structure: a pressure for learnability, a pressure for expressivity, and a pressure for coordination amongst users in a linguistic community. This thesis considers how these sometimes competing pressures impact linguistic complexity across cultural time. Using artificial language and iterated learning experimental paradigms, I investigate the conditions under which complexity in morphological and syntactic systems emerges, spreads, and reduces. These experiments illustrate the interaction of transmission, learning and use in hitherto understudied domains - morphosyntax and word order. In a first study (Chapter 2), I report the first iterated learning experiments to investigate the evolution of complexity in compositional structure at the word and sentence level. I demonstrate that a complex meaning space paired with pressures for learnability and communication can result in compositional hierarchical constituent structure, including fixed combinatorial rules of word formation and word order. This structure grants a productive and productively interpretable language and only requires learners to acquire a finite lexicon and a finite set of combinatorial rules (i.e., a grammar). In Chapter 3, I address the unique effect of communicative interaction on linguistic complexity, by removing language learning completely. Speakers use their native language to express novel meanings either in isolation or during communicative interaction. I demonstrate that even in this case, communicative interaction leads to more efficient and overall simpler linguistic systems. These first two studies provide support for the claim that morphological and syntactic complexity are shaped by an overarching drive towards simplicity (or learnability) in language learning and communication. Chapter 4 reports a series of experiments assessing the possibility that the simplicity bias found in the first two studies operates at a different strength depending on the linguistic level. Studies in natural language learning and in pidgin/creole genesis suggest that while morphological variation seems to be highly susceptible to regularisation, variation in other syntactic features, like word order, appears more likely to be reproduced. I test this experimentally by comparing regularisation of unconditioned variation across morphology and word order in the context of artificial language learning. I show that language users in fact regularise unconditioned variation in a similar way across linguistic levels, suggesting that the simplicity bias may be driven by a single, non-level-specific mechanism. Taken together, the experimental evidence presented in this thesis supports the hypothesis that the cultural and cognitive pressures acting on language users during learning and communicative interaction - for learnability, expressivity and coordination - are at least partially responsible for the evolution of linguistic complexity. Specifically, they are responsible for the emergence of linguistic complexity which maximises learnability and communicative efficiency, and for the reduction of complexity which does not. More generally, the approach taken in this thesis promotes a view of complexity in linguistic systems as an evolving variable determined by the biases of language learners and users as languages are culturally transmitted.

It has been assumed that mathematics testing indicates the development of mathematics concepts, but the linguistic demands of assessment have not been evaluated, especially for children with mild intellectual disabilities. 244 children (grades 2 – 5) were recruited from a larger reading intervention study. Using a multilevel longitudinal SEM model, baseline and post-intervention time points were examined for the contribution of item linguistic complexity, child language skills, and their potential interaction in predicting item level mathematics assessment performance. Item linguistic complexity was an important, stable, and negative predictor of mathematics achievement with children’s language skills significantly and positively predicting mathematics achievement. The interaction between item linguistic complexity and language skills was significant though not stable across time. Following intervention, children with higher language skills performed better on linguistically complex mathematics items. Mathematics achievement may be related to an interaction between children’s language skills and the linguistic demands of the tests themselves.

Essa pesquisa, desenvolvida de acordo com os pressupostos da pesquisa-ação, teve como objetivo, analisar os benefícios de dois Programas de Produções de Narrativas Escritas, em escolares da terceira série do Ensino Fundamental. Para a amostra, foram selecionados 60 escolares de duas classes da terceira série de uma mesma escola estadual de São Paulo. Foram programados encontros semanais, num total de quatorze, sendo utilizada uma história de livro infantil por semana. Para a primeira turma (Grupo A), a pesquisadora contava a história selecionada e para a segunda turma (Grupo B) a mesma história era lida, compartilhando-se esta leitura com a classe por meio de recurso audiovisual, enfatizando a marcação de diálogos por meio da entonação e prosódia. Em seguida, cada escolar produzia uma narrativa escrita com base no tema presente no texto apresentado. Antes e após a realização do Programa de Práticas de Narrativas Escritas, os escolares desenvolveram produções escritas livres a partir de um tema proposto. As produções escritas foram analisadas qualitativamente e quantitativamente, o que caracterizou uma pesquisa mista. Foram adotados os critérios das Competências Comunicativas (Genérica, Enciclopédica e Lingüística), adaptados do lingüista francês Maingueneau (2002). Estes dados foram classificados e comparados nos momentos inicial e final do programa e receberam tratamento estatístico. Observou-se que, em ambos os grupos, os escolares apresentaram evolução na produção de narrativas. Na comparação entre os programas, os dois grupos apresentaram progressos, sendo que o Grupo B apresentou maior evolução estatisticamente significante no desempenho dos sujeitos, nas três competências comunicativas analisadas, com destaque para a Competência Lingüística. Todos os estudantes se beneficiaram dos programas, inclusive aqueles que apresentavam dificuldades para elaborar narrativas escritas. Conclui-se que o programa de leitura compartilhada dos livros infantis constitui-se em uma estratégia mais eficiente para auxiliar o estudante a desenvolver melhores produções escritas, pois além ser um evento de exposição ao letramento, conta com recursos prosódicos que exploram a complexidade lingüística, necessária para motivar o estudante a envolver-se com a leitura e a produção textual. / This research was developed according to the postulations of the action research and aimed to analyze the benefits of two Programs of Written Narrative Production in children from the third grade of an elementary school. The sample was composed by 60 school children from two third grades of a same State school of São Paulo. Fourteen weekly meetings were scheduled and one story from a child book was used per week. In the first group (Group A) the researcher read the selected story and in the second group (Group B) the same story was read combined with a written audiovisual resource, emphasizing dialogues markers through intonation and prosody. Then, each school children produced a written narrative based on the theme of the presented text. Before and after the Written Narrative Production Program, the school children developed free written productions from a proposed theme. The written productions were analyzed qualitative and quantitatively, characterizing a mixed research. The Communicative Competence criteria were adopted (Generic, Encyclopedia and Linguistic) adapted from the French linguist Maingueneau (2002). Data were classified and compared in the beginning and in the end of the program, and were statistically analyzed. It was observed that both groups of school children presented improvement in the narrative production. Comparing the two programs, both groups improved; Group B presented greater statistically significant improvement in the three communicative competences analyzed, especially in the Linguistic Competence. All school children benefited from both programs, including those who presented difficulties to develop written narratives. It was concluded that the combined reading program of child books was a more efficient strategy to assist the student to develop better written productions, since, besides being an event of exposition to literacy, it counts on prosodic resources that explore the linguistic complexity necessary to motivate the student to get involved with reading and textual production.

Essa pesquisa, desenvolvida de acordo com os pressupostos da pesquisa-ação, teve como objetivo, analisar os benefícios de dois Programas de Produções de Narrativas Escritas, em escolares da terceira série do Ensino Fundamental. Para a amostra, foram selecionados 60 escolares de duas classes da terceira série de uma mesma escola estadual de São Paulo. Foram programados encontros semanais, num total de quatorze, sendo utilizada uma história de livro infantil por semana. Para a primeira turma (Grupo A), a pesquisadora contava a história selecionada e para a segunda turma (Grupo B) a mesma história era lida, compartilhando-se esta leitura com a classe por meio de recurso audiovisual, enfatizando a marcação de diálogos por meio da entonação e prosódia. Em seguida, cada escolar produzia uma narrativa escrita com base no tema presente no texto apresentado. Antes e após a realização do Programa de Práticas de Narrativas Escritas, os escolares desenvolveram produções escritas livres a partir de um tema proposto. As produções escritas foram analisadas qualitativamente e quantitativamente, o que caracterizou uma pesquisa mista. Foram adotados os critérios das Competências Comunicativas (Genérica, Enciclopédica e Lingüística), adaptados do lingüista francês Maingueneau (2002). Estes dados foram classificados e comparados nos momentos inicial e final do programa e receberam tratamento estatístico. Observou-se que, em ambos os grupos, os escolares apresentaram evolução na produção de narrativas. Na comparação entre os programas, os dois grupos apresentaram progressos, sendo que o Grupo B apresentou maior evolução estatisticamente significante no desempenho dos sujeitos, nas três competências comunicativas analisadas, com destaque para a Competência Lingüística. Todos os estudantes se beneficiaram dos programas, inclusive aqueles que apresentavam dificuldades para elaborar narrativas escritas. Conclui-se que o programa de leitura compartilhada dos livros infantis constitui-se em uma estratégia mais eficiente para auxiliar o estudante a desenvolver melhores produções escritas, pois além ser um evento de exposição ao letramento, conta com recursos prosódicos que exploram a complexidade lingüística, necessária para motivar o estudante a envolver-se com a leitura e a produção textual. / This research was developed according to the postulations of the action research and aimed to analyze the benefits of two Programs of Written Narrative Production in children from the third grade of an elementary school. The sample was composed by 60 school children from two third grades of a same State school of São Paulo. Fourteen weekly meetings were scheduled and one story from a child book was used per week. In the first group (Group A) the researcher read the selected story and in the second group (Group B) the same story was read combined with a written audiovisual resource, emphasizing dialogues markers through intonation and prosody. Then, each school children produced a written narrative based on the theme of the presented text. Before and after the Written Narrative Production Program, the school children developed free written productions from a proposed theme. The written productions were analyzed qualitative and quantitatively, characterizing a mixed research. The Communicative Competence criteria were adopted (Generic, Encyclopedia and Linguistic) adapted from the French linguist Maingueneau (2002). Data were classified and compared in the beginning and in the end of the program, and were statistically analyzed. It was observed that both groups of school children presented improvement in the narrative production. Comparing the two programs, both groups improved; Group B presented greater statistically significant improvement in the three communicative competences analyzed, especially in the Linguistic Competence. All school children benefited from both programs, including those who presented difficulties to develop written narratives. It was concluded that the combined reading program of child books was a more efficient strategy to assist the student to develop better written productions, since, besides being an event of exposition to literacy, it counts on prosodic resources that explore the linguistic complexity necessary to motivate the student to get involved with reading and textual production.

The right ear advantage (REA) phenomenon has been utilized in clinical and research settings to study auditory processing disorders and linguistic lateralization. Previous research has established that the REA is not reliable in its measures within or between individuals. This is likely due to the influence of other variables, such as neuromaturation and attention. One variable that has not been studied in depth in this context is linguistic complexity. It was hypothesized that stimulus conditions with levels of linguistic complexity would elicit corresponding levels of temporal lobe activity. Understanding and controlling the variables that affect the REA will increase the reliability of the measure. Twenty right handed, neurotypical individuals aged 18-29 participated in a functional magnetic resonance imaging (fMRI) study that identified the regions and the extent of activation involved in listening to dichotic syllables, words, and sentences. Three durations of speech babble corresponding to the mean duration of the syllables, words, and sentences were used as control stimuli. Participants listened to dichotic stimuli and reported the stimulus they heard best during an fMRI scan. Reaction time (RT), ear preference, and fMRI data were recorded simultaneously and analyzed post hoc. Behavioral results showed that words had the shortest RTs and the greatest REA; syllables and sentences were similar to each other for both measures. Significant main effects were found in brain regions known to be involved in cognitive control of attention and linguistic processing. Words were associated with significant activation differences for ear preferences and minimal frontal lobe involvement for right ear preference. Syllables caused the least activity in the frontal lobe regions and less voxel activity in the temporal lobes than syllable-length babble. Sentences had the greatest voxel activity in the frontal and temporal lobe regions. It was concluded that words would best reflect the REA in clinical and experimental designs. Words had minimal involvement of frontal lobe regions indicating minimal cognitive control of attention and the largest discrepancies in activation patterns between right and left ear preferences that showed less cognitive power to process right ear stimuli in a dichotic listening situation.

La communication humaine vise principalement à transmettre de l'information par le biais de l'utilisation de langues. Plusieurs chercheurs ont soutenu l'hypothèse selon laquelle les limites de la capacité du canal de transmission amènent les locuteurs de chaque langue à encoder l'information de manière à obtenir une répartition uniforme de l'information entre les unités linguistiques utilisées. Dans nos recherches, la stratégie d'encodage de l'information en communication parlée est connue comme résultant de l'interaction complexe de facteurs neuro-cognitifs, linguistiques, et sociolinguistiques et nos travaux s'inscrivent donc dans le cadre des systèmes adaptatifs complexes. Plus précisément, cette thèse vise à mettre en évidence les tendances générales, translinguistiques, guidant l'encodage de l'information en tenant compte de la structure des langues à trois niveaux d'analyse (macrosystémique, mésosystémique, et microsystémique). Notre étude s'appuie ainsi sur des corpus oraux et textuels multilingues dans une double perspective quantitative et typologique. Dans cette recherche, la langue est définie comme un système adaptatif complexe, régulé par le phénomène d'auto-organisation, qui motive une première question de recherche : "Comment les langues présentant des débits de parole et des densités d'information variés transmettent-elles les informations en moyenne ?". L'hypothèse défendue propose que la densité moyenne d'information par unité linguistique varie au cours de la communication, mais est compensée par le débit moyen de la parole. Plusieurs notions issues de la théorie de l'information ont inspiré notre manière de quantifier le contenu de l'information et le résultat de la première étude montre que le débit moyen d'information (i.e. la quantité moyenne d'information transmise par seconde) est relativement stable dans une fourchette limitée de variation parmi les 18 langues étudiées. Alors que la première étude propose une analyse de l'auto-organisation au niveau macro- systémique, la deuxième étude porte sur des sous-systèmes linguistiques tels que la phonologie et la morphologie : elle relève donc d'une analyse au niveau mésosystémique. Elle porte sur les interactions entre les modules morphologique et phonologique en utilisant les mesures de la complexité linguistique de ces modules. L'objectif est de tester l'hypothèse d'uniformité de la complexité globale au niveau mésosystémique. Les résultats révèlent une corrélation négative entre la complexité morphologique et la complexité phonologique dans les 14 langues et vont dans le sens de l'hypothèse de l'uniformité de la complexité globale d'un point de vue typologique holistique. La troisième étude analyse l'organisation interne des sous-systèmes phonologiques au moyen de la notion de charge fonctionnelle (FL) au niveau microsystémique. Les contributions relatives des sous-systèmes phonologiques (segments, accents, et tons) sont évaluées quantitativement en estimant leur rôle dans les stratégies lexicales. Elles sont aussi comparées entre 2 langues tonales et 7 langues non-tonales. En outre, la distribution interne de la charge fonctionnelle à travers les sous-systèmes vocaliques et consonantiques est analysée de façon translinguistique dans les 9 langues. Les résultats soulignent l'importance du système tonal dans les distinctions lexicales et indiquent que seuls quelques contrastes dotés d'une charge fonctionnelle élevée sont observés dans les distributions inégales de charge fonctionnelle des sous-systèmes dans les 9 langues. Cette thèse présente donc des études empiriques et quantitatives réalisées à trois niveaux d'analyse, qui permettent de décrire des tendances générales parmi les langues et apportent des éclaircissements sur le phénomène d'auto-organisation. / The main goal of using language is to transmit information. One of the fundamental questions in linguistics concerns the way how information is conveyed by means of language in human communication. So far many researchers have supported the uniform information density (UID) hypothesis asserting that due to channel capacity, speakers tend to encode information strategically in order to achieve uniform rate of information conveyed per linguistic unit. In this study, it is assumed that the encoding strategy of information during speech communication results from complex interaction among neurocognitive, linguistic, and sociolinguistic factors in the framework of complex adaptive system. In particular, this thesis aims to find general cross-language tendencies of information encoding and language structure at three different levels of analysis (i.e. macrosystemic, mesosystemic, and microsystemic levels), by using multilingual parallel oral and text corpora from a quantitative and typological perspective. In this study, language is defined as a complex adaptive system which is regulated by the phenomenon of self-organization, where the first research question comes from : "How do languages exhibiting various speech rates and information density transmit information on average ?". It is assumed that the average information density per linguistic unit varies during communication but would be compensated by the average speech rate. Several notions of the Information theory are used as measures for quantifying information content and the result of the first study shows that the average information rate (i.e. the average amount of information conveyed per second) is relatively stable within a limited range of variation among the 18 languages studied. While the first study corresponds to an analysis of self-organization at the macrosystemic level, the second study deals with linguistic subsystems such as phonology and morphology and thus, covers an analysis at the mesosystemic level. It investigates interactions between phonological and morphological modules by means of the measures of linguistic complexity of these modules. The goal is to examine whether the equal complexity hypothesis holds true at the mesosystemic level. The result exhibits a negative correlation between morphological and phonological complexity in the 14 languages and supports the equal complexity hypothesis from a holistic typological perspective. The third study investigates the internal organization of phonological subsystems by means of functional load (FL) at the microsystemic level. The relative contributions of phonological subsystems (segments, stress, and tones) are quantitatively computed by estimating their role of lexical strategies and are compared in 2 tonal and 7 non-tonal languages. Furthermore, the internal FL distribution of vocalic and consonantal subsystems is analyzed cross-linguistically in the 9 languages. The result highlights the importance of tone system in lexical distinctions and indicates that only a few salient high-FL contrasts are observed in the uneven FL distributions of subsystems in the 9 languages. This thesis therefore attempts to provide empirical and quantitative studies at the three different levels of analysis, which exhibit general tendencies among languages and provide insight into the phenomenon of self-organization.

Cette thèse en linguistique théorique s’inscrit dans un cadre comparatiste. La première partie de notre thèse est un essai de théorisation de la mesure de la complexité linguistique. Nous y proposons une redéfinition de la théorie de la marque, socle de notre recherche transversale, qui prend en compte (1) les différentes définitions qu’elle a reçues au cours de son développement, (2) les apports que peuvent représenter d’autres modèles théoriques et (3) les critiques qui lui ont été faites jusqu’à notre présente étude. La seconde partie de notre thèse est un essai d’application de la théorie de la marque qui a pour objectif de tester son degré de validité dans plusieurs composantes du langage et dans des analyses intra- et intersystémiques de quelques langues indo-européennes et de l’océan Indien. Il s’agit aussi de cerner ses limites et de présenter les principes fonctionnels avec lesquels elle est en concurrence. / This thesis in theoretical linguistics is set within a comparative framework. The first part is an attempt to theorize the measurement of linguistic complexity. We endeavour to redefine markedness theory – on which our cross-disciplinary research is based – while taking into account (i) the different definitions given to the theory since its creation, (ii) the benefits other theoretical models can bring to this theory, and (iii) the criticisms the theory has received. The second part of our thesis is an attempt to apply markedness theory in order to test its degree of validity in several components of language and in intraand inter-systemic analyses of some Indo-European and Indian Ocean languages. We also aim at identifying the limits of this theory and presenting other competing functional principles.

This project uses corpus linguistics and geostatistics to test the sociolinguistic typological theory put forward by Peter Trudgill on the history of Norwegian. The theory includes several effects of societal factors on language change. Most discussed is the proposal that ‘intensive’ language contact causes simplification of language grammar. In the Norwegian case, the claim is that simplificatory changes which affected all of the Continental North Germanic languages (Danish, Swedish, Norwegian) but not the Insular North Germanic Languages were the result of contact with Middle Low German through the Hanseatic League. This suggests that those simplificatory changes arose in the centres of contact with the Hanseatic League: cities with Hansa trading posts and kontors. The size of the dataset required would have made it impossible for previous scholars to test this prediction, but digital approaches render the problem tractable. I have designed a 3.5m word corpus containing nearly all extant Middle Norwegian, and developed statistical methods for examining the spread of language phenomena in time and space. The project is made up of a series of case studies of changes. Three examine simplifying phonological changes: the rise of svarabhakti (epenthetic) vowels, the change of /hv/ > /kv/ and the loss of the voiceless dental fricative. A further three look at simplifying morphological changes: the loss of 1.sg. verbal agreement, the loss of lexical genitives and the loss of 1.pl. verbal agreement. In each case study a large dataset from many documents is collected and used to map the progression of the change in space and time. The social background of document signatories is also used to map the progression of the change through different social groups. A variety of different patterns emerge for the different changes examined. Some changes spread by contagious diffusion, but many spread by hierarchical diffusion, jumping first between cities before spreading to the country at large. One common theme which runs through much of the findings is that dialect contact within the North Germanic language area seems to have played a major role: many of the different simplificatory changes may first have spread into Norwegian from Swedish or Danish. Although these findings do not exactly match the simple predictions originally proposed from the sociolinguistic typological theory, they are potentially consistent with a more nuanced account in which the major centres of contact and so simplifying change were in Sweden and Denmark rather than Norway.