Le travail présenté ici est sous-tendu par deux observations majeures. Premièrement, la plupart des modèles proposés pour le français s’accordent sur l’existence de deux niveaux de structure prosodique: le syntagme accentuel et le syntagme intonatif. Deuxièmement, bien que l’existence d’un niveau additionnel de structure situé entre ces deux niveaux ait été proposé pour le français, les propriétés phonétiques et phonologiques de ce constituant n’ont pas clairement été définies. Dans cette thèse nous avons fourni des preuves de l’existence du syntagme intermédiaire (ip) à la fois en production et en perception de la parole. Grâce à cinq expérimentations menées dans le cadre de la phonologie de laboratoire, nous avons caractérisé les propriétés phonético-phonologiques de ce constituant et attesté de son rôle dans le traitement perceptif du langage. Les résultats obtenus en production montrent que l’ip est le domaine de l’abaissement des accents mélodiques en français. Sa frontière droite est marquée par un allongement pré-frontière ainsi qu’un accent de syntagme responsable du retour à la ligne de référence du registre. Les analyses menées en perception ont montré que les frontières droites du syntagme accentuel et du syntagme intermédiaire sont utilisées très tôt dans le processus de traitement syntaxique. Les indices phonétiques et phonologiques présents à ces frontières permettent aux auditeurs du français de construire des attentes sur la structure syntaxique des énoncés perçus. Une analyse séparée des différents types d’indices acoustiques a également montré qu’en l’absence de marquage tonal, les indices de durée semblent suffisants dans le but de marquer la frontière de syntagme accentuel. Un marquage conjoint de la frontière droite d’ip par les indices mélodiques et l’allongement pré-frontière semble au contraire nécessaire pour que les auditeurs du français perçoivent et utilisent cette frontière dans le traitement du langage. / The work described here is grounded by two major observations. Firstly, most of the French intonation models agree on the existence of two levels of prosodic phrasing: the accentual phrase and the intonation phrase. Secondly, although the existence of an additional level of structure ranked between these two levels has been proposed for French, the phonetic and phonological properties of this intermediate phrase (ip) have not been clearly defined. In this thesis we provide evidence for the existence of an intermediate level of phrasing in French through both speech production and perception studies. Results of five experiments conducted within the framework of laboratory phonology revealed specific ip phonetic and phonological properties and tested its role in the perceptual processing of language. The production studies showed that the ip is the domain of downstep in French and that its right boundary is marked by a phrase accent responsible for a return to the register reference line. Analyses conducted in perception showed that the accentual phrase and intermediate phrase right boundaries are used early in the syntactic processing. Phonetic and phonological indices at these boundaries allow French listeners to build expectations about the syntactic structure of spoken utterances. A separate analysis of different types of acoustic cues showed that without tonal marking, pre-boundary lengthening seems to be sufficient to mark the accentual phrase boundary. Joint marking through melodic and lengthening cues appears to be necessary to perceive and make use of the ip boundary in language processing.

Pianists often have trouble in determining where a phrase ends, or in other words, cadence identification. This is especially true of certain cadences that can be considered either as half cadences or authentic cadences. This analytically ambiguous cadential point can result in different performance decisions, so pianists should make informed decisions about what kind of cadence it is. This study aims to investigate such cadential ambiguity shown at points of phrase boundaries by focusing on Chopin's Piano Sonata No. 2 in B-flat minor, Op. 35, first movement. I offer both possibilities (a half cadence or an authentic cadence) at the phrase ending, suggesting a performance-related strategy based on each possibility. My objective is not to support only one cadential status, but to bring up the cadential problem from the analytical perspective and to demonstrate how cadence identification affects performance results. The dissertation is divided into two parts: analysis and performance, so it relies on a combined method of analytical terminologies and performance-related musical elements. In the analysis, the terminology of William Caplin is employed. The performance part refers to several method books written by prestigious piano pedagogues. After an introduction in Chapter 1, Chapter 2 reviews some literature on cadences. Chapter 3 specifically analyzes the first movement of Chopin's second sonata by means of Caplin's terminologies. Chapter 4 provides a performance-related method and Chapter 5 deals with a practical performance strategy.

L'analogie, c'est-à-dire une correspondance entre deux entités, est considérée une capacité de raisonnement importante. L'analogie proportionnelle, écrite $a:b::c:d$ et qui se lit ``$a$ est à $b$ ce que $c$ est à $d$'', en est un cas particulier où la correspondance tient de par la relation entre les éléments de deux paires d'objets. Le mémoire évalue certaines méthodes issues de l'usage de représentations distributionnelles vectorielles dans la résolution d'analogies proportionnelles verbales et les mène à leur prolongement naturel, la phrase. Nous ciblons la compétence de modèles de langue et des représentations qui peuvent en être extraites à la résolution d'analogies proportionnelles formées sur la base de relations syntaxiques, sémantiques, ou de connaissance encyclopédique. Peu d'ensembles de données existent pour les analogies de phrase et sinon comprennent pour la plupart des analogies au niveau de la forme, composées de phrases construites à partir de gabarits, ou bien variant peu dans les relations sémantiques qui tiennent entre les phrases. Nous construisons donc un ensemble de données contenant des phrases en paires relationnelles qui nous permet de construire des analogies en appariant deux paires. Nous essayons différentes variations de méthodes qui comportent un objectif de recouvrement par un modèle vectoriel. D'autres méthodes de résolution d'analogies proportionnelles sont explorées par voie de génération de texte. Nous expérimentons par le peaufinement du modèle de langue Flan-T5, pré-entraîné sur des paires instruction-réponse, sur nos analogies par une tâche séquence à séquence, ainsi que par l'incitation avec peu d'exemples en utilisant des versions de ce modèle en variant la capacité jusque dans la gamme des milliards de paramètres. En somme, la performance observée est faible pour toutes les tâches. Nous concluons, de l'utilisation de plongements de phrase, quelques mises en garde similaires à celles que l'on trouve avec la résolution d'analogies verbales par plongements lexicaux. Nos expérimentations génératives démontrent l'importance de données à la fois de bonne qualité et de bonne quantité, ainsi que le potentiel de l'apprentissage en contexte. Nous ajoutons à cela un aperçu qualitatif de la disparité entre l'habileté de modèles probabilistes entraînés pour prédire, à partir d'une instruction, la séquence correcte, et celle d'un modèle peaufiné par la méthode d'apprentissage par renforcement avec commentaires humains, à savoir ChatGPT. / Analogy, the correspondence between two things, has been hailed as an important reasoning capability. Proportional analogy, denoted $a:b::c:d$, read ``$a$ is to $b$ as $c$ is to $d$'' is a special case of this where a correspondence is made in the relation that holds between the elements of two pairs. This thesis evaluates methods originating in the recent use of distributional vector representations for solving four-part word analogies, bringing them to their natural extension, sentences. Few datasets of proportional sentence analogies exist, typically comprising purely formal analogies or sentences constructed by templates, and where semantic relations are typically limited in the variety we would hope to capture. Thus, for the purposes of our experiments, we curate a dataset of pairs of sentences for which a given relation holds and from which analogies can be constructed by matching pairs within a relation together. We target the analogy-solving ability of language models and representations derived therefrom, specifically as regards proportional sentence analogies formed on the basis of syntax, semantics, or encyclopedic knowledge. Different variations on previous methods are explored, all based on retrieval of the solution in a vector space model. Other methods of solving proportional sentence analogies by generation are attempted. We experiment with finetuning the instruction-trained Flan-T5 language model on sentence analogies as a sequence-to-sequence task, as well as prompting model checkpoints up into the billion-parameter range with few-shot examples. Overall performance at the task is poor in both settings. We find that similar caveats which apply to analogical reasoning with word vectors apply to sentence embeddings as well. Our generative experiments show the importance of data of suitable quality and quantity, as well the potential of in-context learning. Some qualitative insights are shown as to the disparity in task ability of instruction-trained probabilistic language models and one finetuned by reinforcement learning with human feedback, namely ChatGPT.

Thesis (DLitt (African Languages))--University of Stellenbosch, 2007. / Definiteness is generally viewed as a morpho-syntactic category. It is grammatically marked by articles in languages such as English, but not all languages have a formal grammatical encoding for this category. The nominal preprefix (NPP) in languages such as Dzamba is not an equivalent to the English article system; however, it displays a close association with definiteness. Northern Sotho is non-articled, and it does not possess the NPP in its grammatical system. As a grammatical category, definiteness is the grammaticalisation of a pragmatic category of identifiability, which is present in all languages. Identifiability, as a means of referent tracing, plays a vital role in communication. The main aim of this study is to investigate how the phenomenon of definiteness manifests itself in Northern Sotho. The introductory chapter describes the purpose and aim of the study, its theoretical approach and methodology, as well as its organisation. The second chapter presents an overview of the previous major works on definiteness. It begins with the literature on the category in general, and moves on to the literature on definiteness in African Languages. These previous studies agree on the central issues of this category. The speaker utters a definite noun phrase (NP) if he presupposes that the addressee will be in a position to locate and to identify the referent of the NP uniquely or inclusively. Chapter 3 examines noun phrases that are regarded as definite in Northern Sotho; and the factors that contribute to such a reading. Pragmatic factors, i.e. existential presupposition, transparent contexts and anaphoric reference make major contributions to the interpretation of a noun phrase as definite. Nominal determiners and quantifiers whose semantic content suggests locatability, uniqueness and/or inclusiveness give a noun phrase definite reference. Such determiners and quantifiers include the demonstrative, possessive with locative gona/ntshe, the universal quantifier, etc. Proper names and pronouns have unique reference. Chapter 4 investigates indefinite noun phrases in Northern Sotho. Bare noun phrases in this language such as mang (who), lefeela (nothing) and aretse (unknown thing/place) are incompatible with definiteness. Their semantic content suggests that their referent cannot or should not be uniquely identified. Nominal modifiers such as -ngwe (another/different/ a certain), -fe (who/which) and šele (another/different/strange) are also incompatible with unique identifiabilty and they, therefore, accord a noun phrase indefinite reference. Nouns with generic interpretation and nouns in idioms do not uniquely refer to particular individuals. Chapter 5 looks into the ambiguity of bare noun phrases in Northern Sotho. It examines such a phrase in the subject position, the object position and the complement position of prepositional phrases. Nominal modifiers such as the adjective, the relative and the possessive are incorporated into noun phrases to see how they affect the reading. The question of subject inversion (SI) is also investigated. Lastly opaque contexts are discussed, and the ambiguity created by opacity-creating operators is examined. The final chapter of the study presents the main findings.

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.

On January 14, 2004, President George W. Bush announced the Vision for Space Exploration (VSE). The goals of the vision include developing a new generation launch capability while completing assembly of the International Space Station (ISS) and retiring the Space Shuttle by 2010. In support of this goal, the Space Shuttle Program (SSP) initiated evaluation of hardware, infrastructure, and workforce skill mix needed to continue Space Shuttle flights until the projected 2010 retirement. The SSP also studied how NASA will deploy personnel from, and use the facilities of, the SSP to ensure that the Space Shuttle operates safely through its final flight, and to ensure personnel and facilities from the SSP are effectively transitioned to NASA’s exploration programs. NASA funding, like other federal agencies, is affected by various factors including domestic and international political environments, current and emerging technologies available to meet agency goals, and sustainability and potential economic return of federal expenditures. In this paper I will present a retrospective analysis of federal budget allocations to NASA as a percentage of the Federal Budget from years 1958 to 2010 (adjusted to 1979 dollars). The classic method for calculating net present value (NPV) is not well suited for projecting potential value of future R&D technologies. A quantitative analysis of R&D technologies transferred to private industry will be presented, as well as a description of a method of evaluating their significance will discussed relative to current budgetary considerations will likely for past, current and upcoming funding cycles. The opportunity at hand for NASA’s transition from SSP to Constellation in support of the VSE initiative is to advocate their culture as R&D innovators and emphasize the future benefit of increased funding. / text

This study examines phrase and larger formal structures in the first movement of Leo Brouwer’s Elegio de la Danza (1964), a work that draws on tonal and post-tonal traditions. By adapting key features of the tonal motive, as described by Douglass Green, and the tonal period, as proposed by Green and William Caplin, the model seeks to provide a tool for the discussion of phrases and larger forms in Brouwer’s work. An analysis of primary parameters, such as melody, harmony, and rhythm, provides the means to discuss how the composer articulates beginnings and endings of statements and responses, which are then grouped into antecedent and consequent phrases. These periods articulate large-scale sections, which outline a ternary formal design. Secondary parameters (dynamics, tempo markings, instrumental markings) further contribute to the identification of formal structures in Brouwer’s work.

No description available.

The object of this diploma thesis is an octalingual dictionary from 16th century from the collections of the Pedagogical Museum J.A. Komenského in Prague. The title page of this dictionary is missing and therefore the origin, the author and the range of users, who the dictionary was intended for, are not known. The aim of this diploma thesis was to confirm or disprove the hypothesis that the dictionary should have served tradesmen and travellers. The methods used were translation, language analysis of the phrase-book and the thematic analysis of the contents of the dictionary. The theoretical part of the diploma thesis deals with lexicography. It introduces the definition of a dictionary and various types of dictionaries with the most attention paid to translational dictionaries. Thereafter, the development of Czech lexicography from its beginning until nowadays is presented. The theoretical section also contains some facts about phrase-books. Furthermore, the time of 16th century is described with the aim to present the period when the dictionary was produced. Regarding the hypothesis mentioned above, the linguistic analysis has brought these results. The dictionary should have served not only tradesmen but a wide range of users because of the general vocabulary and because it is partly mentioned...

This thesis evaluates the translation quality of phrase-based machine translation system. It explains the translation error annotation scheme to manually annotate errors related to English to Urdu translation system. The primary goal of the thesis is to experiment with different heuristic in order to improve the translation quality based on through manual analysis of 200 test sentences. Different hueristics such as (1) pre-processing of input English, such as word reordering, (2) preprocessing the training corpus in order to improve word alignment, (3) using additional factors (in Moses factored translation) to better model target-side morphological coherence are applied and their impact on the translation quality is evaluated.