The progression of neuroscience relies on the discovery of structure in the brain. From the discovery of neurons to the structure of the potassium channel, and, in recent years, the repeated observation of remarkable geometric structure in the distributed activity of neural populations. What this population-level structure does is not written on it for anyone to read, generally speaking; many statistical and theoretical tools have had to be developed for interpretation. In these chapters, I benefit from and contribute to the growing set of tools for parsing geometries. First, my collaborators and I studied the representation of syntax in (at the time) state-of-the-art language models. Second, we sought to understand why certain geometries emerge in artificial networks. Third, we model the geometry of working memory representations to try and find why 'swap errors' occur. Finally, we offer a new framework and method for discovering discrete structure in continuous representations.

Orientador: Sonia Maria Lazzarini Cyrino / Tese (doutorado) - Universidade Estadual de Campinas, Instituto de Estudos da Linguagem / Made available in DSpace on 2018-08-21T04:51:12Z (GMT). No. of bitstreams: 1 TeixeiradeSousa_Lilian_D.pdf: 3506329 bytes, checksum: e5635d7c31b15e173507503856816bd4 (MD5) Previous issue date: 2012 / Resumo: A negação sentencial no Português Brasileiro (PB) pode ser realizada através de três tipos diferentes de estruturas considerando o número e a posição de partículas negativas: [Neg V], [Neg V Não] e [V Não]. Essa distribuição é bastante rara nas línguas naturais, já que, geralmente, as línguas apresentam apenas uma estrutura para expressar negação sentencial e, algumas vezes, uma estrutura adicional com função discursiva. Alguns pesquisadores associam essa alternância de formas a um processo de mudança linguística conhecido como Ciclo de Jespersen (Schwegler 1991, Furtado da Cunha 1996), enquanto outros explicam a alternância em termos de estrutura informacional (Schwenter 2005; Cavalcante 2007, 2012). Neste estudo, buscamos descrever a ocorrência das estruturas na língua, através de testes de aceitabilidade e preferência, com o objetivo de formular uma análise dentro do quadro teórico da sintaxe gerativa que considerasse tanto as características distributivas quanto as possibilidades de interpretação dessas estruturas. Na literatura linguística é bastante recorrente a afirmação de que apenas a estrutura [Neg V] é livre de restrições sintáticas, o que tem servido de argumento para caracterizá-la como a negação padrão do PB. As outras duas formas, porém, apresentam fortes restrições sintáticas, não sendo ambas possíveis em perguntas-Qu ou em encaixadas temporais. Em nossa análise, verificamos que, além das diferenças entre forma padrão e formas marcadas, as estruturas [Neg V Não] e [V Não] apresentam distinções quando comparadas, por isso, as consideramos, assim como Biberauer & Cyrino (2009), como fenômenos diversos. [V Não], como observamos, não é possível em nenhum tipo de oração encaixada ou com o preenchimento de sujeitos não definidos ou ainda para veicular informação nova, estando restrita a contextos responsivos. [Neg V Não], por outro lado, não é compatível apenas com infinitivas ou encaixadas temporais e não apresenta restrições quanto ao tipo de sujeito e nem quanto ao tipo de informação, se nova ou dada. Também entoacionalmente as estruturas apresentam distinções, enquanto o não de [V Não] não é prosodicamente proeminente, o não final de [Neg V Não] pode ou não ser prosodicamente proeminente. Uma vez que tanto [Neg V] quanto [Neg V Não] podem veicular informação nova e dada, tratamo-las como negações semânticas. Considerando, entretanto, as restrições de [Neg V Não] em sentenças não temporalmente marcadas ou coocorrendo com conjunções subordinativas temporais, passamos a relacioná-la à categoria T. Tendo em vista, ainda, a incompatibilidade dessa estrutura em contextos narrativos, em que não há proposições, mas o sequenciamento de eventos, definimos essa estrutura enquanto um tipo de negação semântica com escopo sobre proposições em oposição à [Neg V], que teria escopo sobre situações. [V Não], por ocorrer unicamente em contextos responsivos e não ser capaz de licenciar itens de polaridade, é interpretada não como uma negação semântica, mas como uma estrutura com a função pragmática de foco. Do ponto de vista teórico, considerando a proposta de Fases, relacionamos a estrutura [Neg V] à fase V, [Neg V Não] à fase T e [V Não] ao CP, periferia à esquerda da fase T / Abstract: Sentential negation in Brazilian Portuguese (BP) may be accomplished through three different kinds of structures, considering the number and position of the negative particles: [Neg V], [Neg V Não] and [V Não]. This distribution is quite rare in natural languages, which usually feature only one structure to express sentential negation, with an optional structure to convey a discourse function. While some researchers have identified this alternation between forms with a process of linguistic change known as Jespersen's Cycle (Schwegler 1991, Furtado da Cunha 1996), some others explain this variation in terms of information structure (Schwenter 2005; Cavalcante 2007, 2012). In this study, we seek to describe the occurrence of these structures in BP by means of acceptability and preference tests, with the objective of formulating an analysis within the theoretical framework of generative syntax that would consider both distributive characteristics and possible interpretations of these structures. In the linguistic literature, it is frequently claimed that the [Neg V] structure is the only one devoid of syntactic restrictions. Such an idea has been presented as an argument for its character as the standard negation form in BP. On the other hand, the other two forms do have strong syntactic restrictions, once they are not both possible in wh-questions or in embedded time clauses. In our analysis we have noticed that, in addition to the differences between the standard form and the other ones, there is a distinction between the [Neg V Não] and [V Não] structures. Therefore we consider them to be separate phenomena, likewise Biberauer & Cyrino (2009a, 2009b): [V Não], as we have observed, is impossible in all kinds of embedded clause, with undefined subjects or even when conveying new information, as it is restricted to response contexts; [Neg V Não], on the other hand, is only incompatible with infinitive or embedded time clauses, and is unrestricted as to the type of subject or type of information conveyed, that is, whether new or given. These structures also contain intonational distinctions: while the 'não' in [V Não] is not prosodically prominent, the final 'não' in [Neg V Não] may or may not be prosodically prominent. Given that both [Neg V] and [Neg V Não] can convey both new and given information, they are treated as semantic negations. By taking into account that [Neg V Não] is not acceptable either in infinitival sentences or in sentences introduced by temporal conjunctions, we put forth that the second 'não' is merged in the T domain. Besides, given the incompatibility of this structure with narrative contexts in which there are no propositions but rather a sequence of events, we have defined this structure as a kind of semantic negation taking scope over propositions as opposed to [Neg V], which takes scope over situations. And because [V Não] occurs solely in the context of response and does not allow for polarity items, we have defined it not as a semantic negation, but as a structure marking pragmatic focus. From a theoretical standpoint, considering Phase Theory, we have related the [Neg V] structure to the V phase, [Neg V Não] to the T phase and [V Não] to the CP, the left periphery of the T phase / Doutorado / Linguistica / Doutora em Linguística

Selon la théorie du syntagme nominal développée par Wilmet (2003), la notion de caractérisant est une notion fonctionnelle qui désigne tous les « accompagnateurs » du nom (ou déterminants) qui modifient l’extension du nom. Cette notion est indépendante des catégories morpho-syntaxique puisqu’on trouve parmi les caractérisants des adjectifs, des syntagmes prépositionnels, des noms, des adverbes, des propositions relatives voire des phrases entières.<p>Les linguistes du français se sont surtout intéressés à la position absolue de l’adjectif (antéposition ou postposition au nom), mais peu à leur ordre relatif. Il était intéressant d’étendre le point de vue à tous les caractérisants parce que le mélange de caractérisants de différentes natures, en particulier la séquence relative de l’adjectif et du complément du nom, pose d’intéressantes questions linguistiques. La notion fonctionnelle montre également sa valeur dans un cadre comparatiste, car différentes langues ne rendent pas le même concept avec la même catégorie morpho-syntaxique.<p>Notre théorie est que la séquence des caractérisants, tant en antéposition qu’en postposition, est régie par une hiérarchie de critères morpho-syntaxiques et sémantiques, en particulier par leur valeur classificatrice, descriptive ou spécificatrice. On a souvent classé les adjectifs en fonction de leur appartenance à une classe sémantique ontologique (couleur, forme, matière…). En réalité, de très nombreux adjectifs et caractérisants n’entrent pas dans ces catégories et ce type de classification n’est pas le premier critère à l’œuvre dans l’ordre des mots. <p>Le syntagme nominal apparaît dès lors comme structuré en différentes couches concentriques autour du nom ;il est délimité en antéposition par les quantifiants et en postposition par les caractérisants spécificateurs qui lui font en quelque sorte pendant. On observera dès lors d’intéressants phénomènes de sens et d’acceptabilité grammaticale dans le jeu des quantifiants et des caractérisants. Inversement, la position relative d’un caractérisant influe sur sa valeur. On pourrait résumer ces effets de sens par la formule :on dit d’abord ce que c’est, ensuite comment c’est, et enfin lequel c’est. De manière très générale donc, on observe que l’orientation des déterminants se fait selon un axe intrinsèque-extrinsèque ou objectif-subjectif.<p>L’étude d’expressions dans d’autres langues et dans des domaines spécialisés (cuisine, appellations officielles incluant des adjectifs géographiques, localisation de logiciels et chimie organique) permet de valider cette hypothèse tout en montrant que l’ordre des mots est un phénomène de génération, propre à chaque langue car la traduction modifie la nature morpho-syntaxique et peut modifier la valeur des caractérisants. <p>La partie technique de la thèse a exploité des techniques de traduction assistée par ordinateur, de traduction automatique et de traitement du langage, elle a fait appel aux langages de balisage standards de la famille XML pour la représentation des corpus et des règles ainsi que pour la réalisation des procédures. Les corpus spécialisés ont été constitués par alignement de corpus monolingues ou par traduction. Ils ont tous été mis au format XML ;les règles de traduction ont été formalisées dans le même format et elles ont été implémentées en XSLT. La formalisation des corpus en assure la portabilité et facilite les recherches de structures grammaticales sur un corpus catégorisé. Les corpus parallèles sont en outre d’une grande aide pour les traducteurs. Enfin, l’automatisation permet de valider les règles linguistiques proposées. / Doctorat en philosophie et lettres, Orientation langue et littérature / info:eu-repo/semantics/nonPublished

Doctorat en philosophie et lettres / info:eu-repo/semantics/nonPublished

Thesis (MA (African Languages))--University of Stellenbosch, 2005. / This study examines data from communication-theoretic refusal in isiZulu, relating to pragmatic theorists’ argument that every human interaction to a large extent carries with it an element of threatening one or both participants’ face. The speech act of refusal has been identified as one of the most effective speech acts to be employed for this phenomenon in this study. Every speech act is influenced by contextual, cultural and many other background factors associated to age, gender and rank, to mention a few, which contribute towards the composition of speech act. The issues relating to the theory of politeness prompted this study in the extent to which politeness plays a role in refusals in educational contexts of Zulu speaking students. The study has evaluated Brown and Levinson’s (1987) theory of politeness as a universal phenomenon against the findings of my data, in that the universality advanced by these theorists does not quite fit with this study. The variable percentages in this data, especially in refusals, indicate that politeness does not always exist in terms of positive and negative face in the participants’ mind but other considerations related to the goal of the speech act are the main source of the speech act. In this study, refusal in the Zulu context has demonstrated that politeness has been employed as a strategy for encoding distance between speaker and solicitor.

Thesis (PhD)--Stellenbosch University, 2001. / English Abstract: This research project examines the syntax of negative sentence constructions in Sesotho and explores implications of the empirical data on negation within the framework of the Minimalist Program. According to Minimalist Program, language consists of a lexicon and Computational System where the operations Merge and Move generate sets of structural descriptions, it is driven by the principle of economy, which entails that movement should take place only when necessary for the purpose of Case feature checking. It is also concerned with the expansion of syntactic structures in terms of X-bar theoretic properties, where functional category gives full categorical status. This project further considered the morphology of negation in Sesotho where Sesotho data is refined within the Lexeme-morpheme base morphology, where bound grammatical morphemes such as the negative morphemes in Sesotho are defined in terms of morphological spelling operations. Within these two theoretical frameworks, sentence constructions which realize negation by means of negative morphemes over a full range of tense, aspect and mood distinctions were examined, including sentence construction that realize negation in subordinate clauses as well as those that entail issues of topic and focus, and scope of negation. Constituent negation, with constituents such as subjects, objects and adjuncts is also investigated. Various ways of negating constituents in Sesotho were examined: they are cleft sentences, pseudo-cleft sentences, sentence construction with locative AgrS [ ho ] and sentence constructions with subject inversion as well as sentences where constituents are negated within the VP. All these methods were considered with regard to noncopulative and copulative verbs. The issue of focus which entails plain focus, restrictive and contrastive foci, the issue of topic, which entails stage topics, modified topics and multiple topics are investigated and interpreted within the Focus Structure theory, a theory which assigns focus structures to sentences of the language. Negative sentences are assigned two focus structure viz. main focus structure and the subordinate focus structure. Finally, this study also looks into the scope of negation in Sesotho, where negative criterion (Neg-criterion) stipulates various positions of negative morphemes or words which in turn determines the scope of negation. Scope of negation is also examined in terms of focus, with constituents such as NP's, NP Lac's and NP's in subordinate clauses. / AFRIKAANSE OPSOMMING: Hierdie navorsingsprojek ondersoek die sintaksis van konstruksies met negatiewe sinne in Sesotho en gaan die implikasies na van die empiriese data oor ontkenning binne die raamwerk van die Minimalistiese Program. Volgens die Minimalistiese Program, bestaan taal uit In leksikon en In rekenaarsisteem waar die bewerkinge Saamsmelt en Skuif versamelings strukturele beskrywings genereer, dit word aangedryf deur die beginsel van ekonomie wat inhou dat Skuif slegs plaasvind wanneer dit nodig is vir die doel van die nagaan van die Kasus kenmerk: dit het ook te doen met die uitbreiding van sintaktiese strukture in terme van die teoretiese kenmerke van die X-balk, waar die funksionele kategorie volle kategoriale status verkry. Dié projek het ook verder die morfologie van ontkenning in Sesotho oorweeg waar die Sesotho data verfyn is binne die Lekseem-Morfeem Basis Morfologie, waar grammatikale morfeme soos die negatiewe morfeme in Sesotho gedefinieer word in terme van morfologiese spellingsbewerkings. Binne hierdie twee teoretiese raamwerke, is daar ondersoek ingestel na sinskonstruksies wat negatief realiseer deur middel van negatiewe morfeme binne In volledige reeks van onderskeidinge van tyd, aspek en modus, insluitende sinskonstruksies waar die negatief realiseer in die bysin, asook daardie negatiewe wat gaan oor sake soos Topiek en Fokus, en die omvang van ontkenning. Konstituent ontkenning is ondersoek waaronder konstituente soos onderwerpe, voorwerpe en adjunkte. Verskeie wyses om konstituente in Sesotho te ontken is ondersoek, naamlik: klef en pseudo-klef sinne, sinskonstruksie met die lokatiewe kongruensie [ho] en sinskonstruksies met subjeksinversie asook sinne waar konstituente ontken word binne In werkwoordfrase. AI hierdie wyses is nagegaan met nie-kopulatiewe en kopulatiewe werkwoorde. Die probleem van fokus is ondersoek wat insluit eenvoudige fokus, beperkende en kontrastiewe fokus, asook die probleem van topiek wat verskeie soorte topiek insluit. Hierdie ondersoek is gedoen en uiteindelik geïnterpreteer binne die Fokus Sruktuur teorie, In teorie wat fokusstrukture toeken aan sinne van In taal. Negatiewe sinne word twee fokusstrukture toegeken, naamlik: die hoof fokus struktuur en die onderskikkende fokus struktuur. Laastens het die studie ook gekyk na die omvang ('scope') van ontkenning in Sesotho waar die negatiewe kriterium verskeie posisies van negatiewe morfeme of woorde stipuleer wat dan weer die omvang van ontkenning bepaal. Die omvang van ontkenning is ook ondersoek in terme van fokus waar kontrastiewe fokus die uitbreiding van omvang bepaaloor konstituente soos naamwoordfrases, lokatiewe naamwoordfrases en naamwoordfrases in onderskikkende klouse.

Thesis (PhD)--Stellenbosch University, 2002. / ENGLISH ABSTRACT: The study focused on three descriptive nominal modifiers which specify the attributes of nouns, i.e. the morphological adjective, the relative clause and the descriptive possessive. The morphological adjective appears in an adjectival phrase, which has to consist of a determiner and an adjective. The adjective must have agreement with the head noun in an NP. The adjective root may appear with nominal suffixes such as -ana and -gadi, it can be reduplicated, it may be transposed to other categories and it may even be compounded. The AP may also occur in predicative position as well as in comparative clauses. The relative clause may have the same semantic properties as the adjective. The relative clause in Setswana consists of a determiner in the position of the complementizer followed by an lP. Such an lP may have a copulative or non-copulative verb. Attention in this study has focused on the nominal relative, which appears as a complement of a copulative verb. These nominal relative stems have been divided into two sections, i.e. a section in which the nominal relative stems may not appear in a descriptive possessive construction and a second section where these stems may also appear as a complement of the possessive [a]. The semantic features of these nominal relative stems have been isolated and it is clear that they show a wide variety of semantic features. This type of relative clause represents the most prolific category, which specifies the attributes of nouns. The third category, which displays the semantic feature of an attribute of a noun, is the descriptive possessive construction. The syntactic and semantic structure of this type of phrase has been investigated. A wide variety of complements of the possessive [a] have been isolated in Setswana and some semantic features have received specific attention, i.e. group nouns and partitives. / AFRIKAANSE OPSOMMING: Die studie fokus op drie deskriptiewe nominale bepalers wat die attribute van naamwoorde spesifiseer nl. die morfologiese adjektief, die relatief en die deskriptiewe possessief. Die morfologiese adjektief kom voor in 'n adjektieffrase wat bestaan uit 'n bepaler en 'n adjektief. Die adjektief moet klasooreenstemming hê met die kernnaamwoord in 'n naamwoordfrase. Die adjektiefstam kan voorkom met nominale suffikse soos ana en gadi, dit kan geredupliseer word, oorgeplaas word na ander kategorieë en selfs samestellings vorm. Die adjektieffrase kan ook voorkom in 'n predikatiewe posisie sowel as in vergelykings. Die relatief kan dieselfde semantiese eienskappe hê as die adjektief. Die relatief in Setswana bestaan uit 'n bepaler in die posisie van die komplementeerder gevolg deur 'n infleksiefrase. So 'n infleksiefrase kan 'n kopulatiewe of nie-kopulatiewe werkwoord bevat. Die aandag in hierdie studie het gekonsentreer op die nominale relatief wat voorkom as 'n komplement van 'n kopulatiewe werkwoord. Hierdie nominale relatiewe stamme is verdeel in twee afdelings nl. 'n afdeling waarin die nominale relatiewe stamme nie kan voorkom in 'n deskriptiewe possessiewe konstruksie en 'n tweede afdeling waar hierdie stamme ook kan voorkom as 'n komplement van die possessiewe [a]. Die semantiese kenmerke van hierdie nominale relatiewe stamme is geïsoleer en dit is duidelik dat hulle 'n wye verskeidenheid van semantiese kenmerke het. Hierdie tipe relatief verteenwoordig In baie wye keuse t.o.v. die attribute van naamwoorde. Die derde kategorie wat die semantiese kenmerk van 'n attribuut van 'n naamwoord vertoon, is die deskriptiewe possessiewe konstruksie. Die sintaktiese en semantiese struktuur van hierdie tipe frase is nagegaan. 'n Groot verskeidenheid komplemente van die possessiewe [a] is geïsoleer in Setswana en sommige semantiese kenmerke het spesifieke aandag gekry nl. groepnaamwoorde en partitiiewe.

Switch-Reference (SR) is a phenomenon in which the coreferentiality of two (or more) subjects in a complex sentence is indicated by a morphological device. The purpose of this dissertation is to discuss recent work which deals with SR within the Government and Binding Theory, and propose an alternative analysis to it. The framework I will adopt for such an alternative analysis of SR is Categorial Grammar. A basic notion underlying Categorial Grammar is that an expression is divided into a functor and an argument, and each functor and argument are further divided into a functor and an argument until the division reaches to an undividable element. Given the assumptions that a functor and its argument must be compatible and that a functor has some subcategorization properties, I argue that "Agreement" phenomenon (subsuming agreement and disagreement) can be handled insightfully. Furthermore, I propose that such a treatment of "Agreement" can be extended to SR systems in general if we consider the "same subject" and "different subject" phenomena as cases of agreement and disagreement, respectively. I claim that a composite in which a SR morpheme appears forms a functor which takes another composite as its argument, and that the relation between the functor and its argument and the relation between some parts of the functor and its argument are characterized as "agreement" or "disagreement": The functor and the argument must be compatible as assumed above, and the nature of compatibility (whether "agreement" or "disagreement") is controlled by the subcategorization properties of the SR morpheme associated with the functor (i.e., if "same subject", the relation is agreement, and if "different subject", it is disagreement). By treating SR in this fashion, I intend to provide a unified analysis for apparently different SR systems in three diverse languages, namely, Tairora, Hopi, and Warlpiri.

Approved for public release, distribution is unlimited / The CyberCIEGE project seeks to create an alternative to traditional Information Assurance (IA) training and education approaches by developing an interactive, entertaining commercial-grade PC-based computer game/virtual laboratory. CyberCIEGE will provide a robust, flexible and extensible gaming environment where each instance of the game is based on a fully customizable scenario. These scenarios are written in the CyberCIEGE Scenario Definition Language. Unfortunately, the trade-off for flexibility, extensibility and fully customizable scenarios is syntax complexity in the scenario definition language. This thesis will solve this real world problem by showing that the complexity of scenario definition language syntax can be managed through a software tool. This thesis will develop such a tool and further demonstrate that progress can be made toward automating scenario generation. / Civilian, Federal Cyber Service Corps, Naval Postgraduate School

Faculty of Humanities School of Literature, Language and Media University of the Witwatersrand A Master’s Dissertation / This study investigates how Hausa, a West Chadic language (Afro Asiatic phyla) remodells loanwords from English (Indo – European) to suit its pre-existing phonology. Loanword adaptation is quite inevitable due to the fact that languages of the world differ, one from another in many ways: phonological, syntactical, morphological and so on (Inkelas & Zoll, 2003, p. 1). Based on this claim, receptor languages therefore employ ways to rephonologize new words borrowed into their vocabularies to fit, and to conform to native structure demands. Hausa disallows complex onsets, preferably operates open syllables and avoids consonant clustering in word-medial positions as at its best can tolerate no more than a single consonant at a syllable edge (Clements, 2000; Han, 2009). On the contrary, English permits complex onsets as well as closed syllables (Skandera & Burleigh, 2005). Such distinctions in both phonologies motivate for loanword adaptation. Hausa therefore employs repair strategies such as vowel epenthesis, consonant deletions and segmental substitutions and/or replacements (Newman, 2000; Abubakre, 2008; Alqhatani & Musa, 2014) to remodell loanwords. For analytical purposes, this research adopts theoretical tools of Feature Geometry (FG) (Clements & Hume, 1995) and Optimality Theory (OT) (Prince & Smolensky, 2004) to clearly illustrate how loanwords are modified to satisfy Hausa native demands (Kadenge, 2012). Vowel epenthesis in Hausa involves two main strategies: consonantal assimilation and default insertions. During consonantal assimilation, coronal and labial segments spread place features unto the epenthetic segment in the process determining the vowel type and/or quality, while in the case of default insertions, fresh segments are introduced context independently. Concerning segmental substitutions, most notably are English consonants /p/ and /v/ maximally replaced with similar ones, [f] and [b] that exist in Hausa on the basis that former and latter segments share same phonation features