This dissertation is a grammar of Rena (ISO bez), a Bantu language spoken in southwestern Tanzania by approximately 600,000 people. Bena is largely undocumented, and though aspects of Bena grammar have been described, there is no usable, detailed treatment of the Bena language. Therefore the goal of this dissertation is provide the first detailed description of Bena that discusses phonology, morphology, and syntax. The analysis described in this grammar is based on data collected in the Njombe district of Tanzania during 2008 and 2009. Data throughout the grammar is taken from both elicitation and a corpus of 23 narratives. Though Bena is spoken by over half a million people, it is threatened by Swahili (the national language of Tanzania). Swahili's prominence in Tanzania has increased drastically since independence in 1961, and many (if not most) of the approximately 120 languages spoken in Tanzania are threatened by Swahili. Bena is no exception to this. The results of a sociolinguistic survey conducted in 2009 indicate that Swahili is having a significant impact on the Bena language. Therefore the writing of this dissertation comes at a crucial time. It provides a record of Bena at a time before too many features of the language are lost due to language contact. The first chapter provides an introduction to the Bena language and people. It also discusses results from the 2009 sociolinguistic survey which had the goal of clarifying questions on both the dialectal situation and the sociolinguistic vitality of Bena. The second chapter is devoted to phonetics and phonology. Of particular interest in this chapter are Bena's "predictable" tone system and the morphophonological process of imbrication (a type of coalescence in which multiple morphemes are interwoven together). The third chapter gives an overview of Bena word classes and provides a road map of the next several chapters of the grammar. Fourth is a description of Bena nominal morphology and other elements in the noun phrase. Like other Bantu languages, Bena uses a complex noun class system; Bena's 19 noun classes and the ways in which they are used are discussed in detail in this chapter. Following this is a description of Bena verbal morphology. Of particular interest in Bena is its tense aspect system--Bena distinguishes four separate past tenses and three distinct futures; these interact with five aspects. The second major focus of Chapter 5 is the use of a series of suffixes in verbal derivation. The sixth chapter of the grammar describes adverbs and other invariable words in Bena. Chapter 7 describes major aspects of Bena syntax. Because Bantu languages have rich morphological systems, most grammars of Bantu languages either give a fairly cursory treatment of syntax or they ignore it completely. This dissertation aims to fill that gap by providing a description of a Bantu language that is more balanced and acknowledges the significant roles played by both morphology and syntax. The final chapter highlights several features of Bena from a typological perspective and discusses areas in which further research on Bena has the potential to contribute significantly to Bantu linguistics.

This thesis tests the Interclausal Relations Hierarchy (VV: 209) vis-à-vis first hand data on eight complex verbal constructions of Sardinian, namely: 1. Campidanese ai / Logudorese-Nuorese àere a + infinitive, lit. 'to have to' + infinitive (hereafter also inf.), expressing future time reference; 2. Campidanese dèppi(ri) / Logudorese-Nuorese dèppere/dèvere + inf., lit. 'must' + inf., expressing both deontic modality and future time reference; 3. Campidanese fai / Logudorese-Nuorese fàghere a + inf., lit. 'to do to' + inf., in the sense of 'to be possible/allowed to do something'; 4. Campidanese fai / Logudorese-Nuorese fàghere + inf., lit. 'to do to' + inf., meaning 'make someone do'; 5. Campidanese lassai / Logudorese-Nuorese lassare/ (dassare) + inf. meaning 'let someone do'; 6. Campidanese torrai a, (po) / Logudorese-Nuorese torrare a + inf., lit. 'to return to', meaning both 'go back to + inf.' and 'do something again‟; 7. Campidanese andai / Logudorese-Nuorese andare a + inf., meaning 'to go to'; 8. Campidanese (am)megai / Logudorese-Nuorese (am)megai de/a + inf., lit. originally meaning perhaps 'pretend' or 'threaten' or 'have an aim', but nowadays meaning 'to be doing', 'to have the intention to'. The hierarchy ranks complex verbal constructions from the most cohesive to the least cohesive, both syntactically and from a semantic point of view. There is a a meaningful prediction of the hierarchy, i.e., the tightest syntactic linkage realizing a particular semantic relation should be tighter than the tightest syntactic linkage realizing looser semantic relations (VVLP: 483). Almost all the constructions respect this prediction, with the exception of megai de + infinitive. In this case it is possible that its syntax crystallised whilst its semantics developed further. In addition, our data display diatopic, i.e., geolinguistic variation. We use Virdis' (1988: 805) phonetic map to analyse our morphosyntactic data. We obtain a number of maps which show that morphosyntactic phenomena are in general more widespread than phonetic isoglosses, that is, they are common to the three main varieties of Sardinian: Campidanese, Logudorese and Nuorese.

A Division III examination in the School of Cognitive Science, Hampshire College, May 2010. Chairperson, Steven Weisler. / Typescript. Includes bibliographical references.

Le thème général de cette recherche est l'enseignement grammatical dans l'apprentissage de la lecture et de l'écriture. Ce travail s'inscrit dans le cadre du projet AMICAL (Architecture Multi-agents Interactive Compagnon pour l'Apprentissage de la Lecture), auquel participent linguistes et informaticiens. L'objectif de notre étude est d'analyser dans quelle mesure des travaux théoriques actuels en linguistique pourraient permettre de déterminer les connaissances sur la phrase à utiliser dans le cadre de l'enseignement grammatical à partir du début de l'apprentissage de la lecture. Pour atteindre cet objectif, nous observons l'évolution du concept de phrase en grammaire et en linguistique, puis nous examinons l'enseignement grammatical à l'école élémentaire. Ces observations nous conduisent à déterminer deux problèmes : la phrase est une unité difficile à définir et la grammaire scolaire telle qu'elle existe est sans cesse remise en cause, à la fois par les linguistes et par les didacticiens. Après avoir mis en évidence les problèmes scientifiques que l'on rencontre dans le discours grammatical scolaire, nous présentons une théorie linguistique - la Role and Reference Grammar - qui nous semble à même de répondre à nos besoins d'enseignement en nous fournissant des savoirs de référence (structure logique des prédicats, structure syntaxique stratifiée de la phrase, rôle des opérateurs, etc.) qui peuvent être utilisés dans l'objectif d'un enseignement grammatical visant la maîtrise de la communication verbale par la compréhension des liens entre sens et forme. C'est l'interaction entre ces deux éléments qui nous semble constituer le coeur de l'apprentissage grammatical. Enfin, à partir des éléments que nous fournissent les recherches en psychologie cognitive et la théorie à laquelle nous avons choisi de nous référer, nous proposons une organisation de la grammaire de phrase et explicitons les points théoriques qui nous semblent à même de constituer des objets à enseigner.

yes / This paper aims to demystify the hype and attention on chatbots and its association with conversational artificial intelligence. Both are slowly emerging as a real presence in our lives from the impressive technological developments in machine learning, deep learning and natural language understanding solutions. However, what is under the hood, and how far and to what extent can chatbots/conversational artificial intelligence solutions work – is our question. Natural language is the most easily understood knowledge representation for people, but certainly not the best for computers because of its inherent ambiguous, complex and dynamic nature. We will critique the knowledge representation of heavy statistical chatbot solutions against linguistics alternatives. In order to react intelligently to the user, natural language solutions must critically consider other factors such as context, memory, intelligent understanding, previous experience, and personalized knowledge of the user. We will delve into the spectrum of conversational interfaces and focus on a strong artificial intelligence concept. This is explored via a text based conversational software agents with a deep strategic role to hold a conversation and enable the mechanisms need to plan, and to decide what to do next, and manage the dialogue to achieve a goal. To demonstrate this, a deep linguistically aware and knowledge aware text based conversational agent (LING-CSA) presents a proof-of-concept of a non-statistical conversational AI solution.

Yes / This chapter encapsulates the multi-disciplinary nature that facilitates NLP in AI and reports on a linguistically orientated conversational software agent (CSA) (Panesar 2017) framework sensitive to natural language processing (NLP), language in the agent environment. We present a novel computational approach of using the functional linguistic theory of Role and Reference Grammar (RRG) as the linguistic engine. Viewing language as action, utterances change the state of the world, and hence speakers and hearer’s mental state change as a result of these utterances. The plan-based method of discourse management (DM) using the BDI model architecture is deployed, to support a greater complexity of conversation. This CSA investigates the integration, intersection and interface of the language, knowledge, speech act constructions (SAC) as a grammatical object, and the sub-model of BDI and DM for NLP. We present an investigation into the intersection and interface between our linguistic and knowledge (belief base) models for both dialogue management and planning. The architecture has three-phase models: (1) a linguistic model based on RRG; (2) Agent Cognitive Model (ACM) with (a) knowledge representation model employing conceptual graphs (CGs) serialised to Resource Description Framework (RDF); (b) a planning model underpinned by BDI concepts and intentionality and rational interaction; and (3) a dialogue model employing common ground. Use of RRG as a linguistic engine for the CSA was successful. We identify the complexity of the semantic gap of internal representations with details of a conceptual bridging solution.

This thesis presents a description and analysis of non-canonical case-marking of core arguments in Lithuanian. It consists of an introduction and six articles, providing historical and/or contrastive perspective to this issue. More specifically, using data from Lithuanian dialects, Old Lithuanian and other languages such as Icelandic, Latin and Finnic for comparison, the thesis examines the development and current state of non-canonical case-marking of core arguments in Lithuanian The present work draws on empirical findings and theoretical considerations to investigate non-canonical case-marking, language variation and historical linguistics. Special attention is paid to the variation in the case-marking of body parts in pain verb constructions, where an accusative-marked body part is used in Standard Lithuanian, and alongside, a nominative-marked body part in Lithuanian dialects. A common objective of the first three articles is to clarify and to seek a better understanding for the reasons for this case variation. The research provides evidence that nominative is the original case-marking of body parts in pain specific construction, i.e. with verbs, with the original meaning of pain, like skaudėti and sopėti ‘hurt, feel pain’. On the contrary, in derived pain constructions, i.e. with verbs like gelti with the original meaning of ‘sting, bite’ and diegti with the original meaning ‘plant’, accusative is the original case-marking of body parts. This accusative is explained by means of an oblique anticausative and it is argued furthermore that it is extended into the pain specific construction. The three last articles focus on the comparative and contrastive perspective. Their main results include the following: Lithuanian and Icelandic differ considerably in the frequency of using accusative vs. dative marking on the highest ranked argument. Accusative is more frequently used in Lithuanian while dative is dominant in Icelandic. The semantic fields of the dative subject construction have remained very stable, suggesting that the dative subject construction is inherited. It has, however, become productive in the history of Germanic, Baltic and Slavic. The similarities in Finnic and Baltic partiality-based object and subject-marking systems are due to Baltic influence. / <p>At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 1: In press. Paper 2: In press. Paper 3: In press.</p>