<p>This thesis consists of four papers in domain theory and a summary. The first two papers deal with the problem of defining effectivity for continuous cpos. The third and fourth paper present the new notion of an admissible domain representation, where a domain representation D of a space X is λ-admissible if, in principle, all other λ-based domain representations E of X can be reduced to X via a continuous function from E to D. </p><p>In Paper I we define a cartesian closed category of effective bifinite domains. We also investigate the method of inducing effectivity onto continuous cpos via projection pairs, resulting in a cartesian closed category of projections of effective bifinite domains. </p><p>In Paper II we introduce the notion of an almost algebraic basis for a continuous cpo, showing that there is a natural cartesian closed category of effective consistently complete continuous cpos with almost algebraic bases. We also generalise the notion of a complete set, used in Paper I to define the bifinite domains, and investigate what closure results that can be obtained. </p><p>In Paper III we consider admissible domain representations of topological spaces. We present a characterisation theorem of exactly when a topological space has a λ-admissible and κ-based domain representation. We also show that there is a natural cartesian closed category of countably based and countably admissible domain representations. </p><p>In Paper IV we consider admissible domain representations of convergence spaces, where a convergence space is a set X together with a convergence relation between nets on X and elements of X. We study in particular the new notion of weak κ-convergence spaces, which roughly means that the convergence relation satisfies a generalisation of the Kuratowski limit space axioms to cardinality κ. We show that the category of weak κ-convergence spaces is cartesian closed. We also show that the category of weak κ-convergence spaces that have a dense, λ-admissible, κ-continuous and α-based consistently complete domain representation is cartesian closed when α ≤ λ ≥ κ. As natural corollaries we obtain corresponding results for the associated category of weak convergence spaces.</p>

This thesis consists of four papers in domain theory and a summary. The first two papers deal with the problem of defining effectivity for continuous cpos. The third and fourth paper present the new notion of an admissible domain representation, where a domain representation D of a space X is λ-admissible if, in principle, all other λ-based domain representations E of X can be reduced to X via a continuous function from E to D. In Paper I we define a cartesian closed category of effective bifinite domains. We also investigate the method of inducing effectivity onto continuous cpos via projection pairs, resulting in a cartesian closed category of projections of effective bifinite domains. In Paper II we introduce the notion of an almost algebraic basis for a continuous cpo, showing that there is a natural cartesian closed category of effective consistently complete continuous cpos with almost algebraic bases. We also generalise the notion of a complete set, used in Paper I to define the bifinite domains, and investigate what closure results that can be obtained. In Paper III we consider admissible domain representations of topological spaces. We present a characterisation theorem of exactly when a topological space has a λ-admissible and κ-based domain representation. We also show that there is a natural cartesian closed category of countably based and countably admissible domain representations. In Paper IV we consider admissible domain representations of convergence spaces, where a convergence space is a set X together with a convergence relation between nets on X and elements of X. We study in particular the new notion of weak κ-convergence spaces, which roughly means that the convergence relation satisfies a generalisation of the Kuratowski limit space axioms to cardinality κ. We show that the category of weak κ-convergence spaces is cartesian closed. We also show that the category of weak κ-convergence spaces that have a dense, λ-admissible, κ-continuous and α-based consistently complete domain representation is cartesian closed when α ≤ λ ≥ κ. As natural corollaries we obtain corresponding results for the associated category of weak convergence spaces.

Neste trabalho desenvolve-se um estudo sobre os Espaços Coerentes Gerados por Conjuntos Básicos, dotados de uma estrutura adicional. Por estrutura adicional entende-se uma estrutura algébrica, de ordem pontual, de medidas, topológica e lógica. Estes espaços, denotados por , constituem uma subcategoria dos Espaços Coerentes, cujos objetos, ordenados pela inclusão, são conjuntos coerentes constituídos por subconjuntos do conjunto básico, os quais estão relacionados pela relação de coerência induzida, que estrutura a teia deste espaço. Os morfismos desta categoria são as funções de objetos geradas por funções básicas. As propriedades algébricas e relacionais destas funções básicas, externas ao processo de construção, ao se propagarem, passam a influenciar na verificação das propriedades internas das funções de objetos. Contudo, este trabalho não é um estudo categórico. A metodologia adotada utiliza a linguagem simples e intuitiva da Teoria dos Conjuntos, que possibilita a visualização e a análise dos relacionamentos existentes, não apenas entre os morfismos que envolvem os objetos totais ou parciais desta categoria, mas também das estruturas ou pré-estruturas externas que os formam, representados pelas funções de tokens e funções básicas. Mostra-se que as funções de objetos são totais e bem definidas, alem de serem monótonas e continuas neste espaço. Entretanto a análise da estabilidade, e consequentemente da linearidade esta associada a injetividade das funções básicas. Uma das características mais importantes da construção proposta e o desenvolvimento de um sistema de representação linear para funções localmente lineares, com a definição do espaço coerente A* gerado pelo produto de subteias. Neste espaço, as funções de objetos são lineares e coincidem com os morfismo da categoria dos espaços coerentes. Além disso, mostra-se que A* e isomorfo ao espaço coerente gerado pelo produto direto dos sub-espaços, ПĄ. Desta forma, toda transformação definida para um tipo de dado estruturado a partir de um conjunto básico enumerável tem uma representação linear, constituída pelos morfismos da categoria dos espaços coerentes. A existência da representação linear para as funções elementares garante a existência da representação linear para outras funções derivadas destas. Apresenta-se ainda uma especificação desta construção, introduzindo-se o Espaço Coerente de Intervalos Racionais, IIQ. Na busca de uma aplicação compatível com uma abordagem computacional, em especial para Análise Real, mostra-se que, em IIQ, cada função real elementar esta identificada com uma função de objetos linear, definida a partir da correspondente função elementar racional. Dentre as funções que foram analisadas destacam-se: a exponencial, a logarítmica, a potência, a potência estendida, a raiz n-ésima, as funções trigonométricas como seno, cosseno e tangente e suas correspondentes funções inversas, como também a função polinomial. Verificou-se que todas estas funções de objetos são totais, bem definidas, ou pertencem ou possuem uma representação linear na categoria COSP-LIN dos espaços coerentes, alem de serem fechadas para os objetos totais e quasi-totais deste espaço, sendo possível estabelecer o correspondente par-projeção para cada uma delas. / In this work the Coherence Spaces Generated by Basic Sets with additional structure are studied. By additional structure one means an algebraic, topological and logical structure with a punctual order and a measure system. These spaces, indicated by A, are a subcategory of the category of Coherence Spaces, whose objects, ordered by inclusion, are coherent sets formed by the induced web coherence relation. The morphisms of this category are the functions of objects generated by basic functions. The algebraic and relational properties of these basic functions - external to the construction process - are propagated and cause important influences in the verification of the internal properties of the functions of objects However, this research is not a categorical study. The methodology uses the simple and intuitive language of the Set Theory, which allows the visualization and the analysis of the existing relationships, not only among, the morphisms of the total and partial objects of this category, but also among their structures or pre-structures, represented by the functions of tokens and basic functions. It is shown that the functions of objects are total and well defined. They are also monotone and continuous. However the stability and the linearity of the functions of objects depend on the fact if the basic functions are injective or not. One of the most important features of this construction is the development of a linear representation system for the local linear functions, by the definition of a coherence space A*, which is generated by the subweb product. In this space the functions of objects are linear and therefore they are the morphisms of the category of Coherence Spaces. Moreover, it is proved that A* is isomorphic to the coherence space generated by the directed product of the subspaces, denoted by ПĄ . Then, for each transformation defined for a structured data type considering a denumerable basic set there exists its related linear representation. The existence of a linear representation for elementary functions guarantees the existence of a linear representation for others derived functions. As an application of this construction, the Coherence Space of Rational Intervals, denoted by IIQ, is introduced. In order to show an application which is compatible to a computational approach, specially for the real analysis, each elementary real function is identified with a linear function of objects, defined considering the related elementary rational function. Some of the analyzed functions are the exponential, the logarithmic, the power , the extended power, the root, the trigonometric (sine, cosine and tangent and their relates inverses), and the polynomial functions. It is proved that all of these functions of objects are total and well defined. Moreover, either they belong to the category COPS-LIN of the coherence spaces or they have a linear representation in the same category. It is also possible to define a related projection pair for each one of them.

Neste trabalho desenvolve-se um estudo sobre os Espaços Coerentes Gerados por Conjuntos Básicos, dotados de uma estrutura adicional. Por estrutura adicional entende-se uma estrutura algébrica, de ordem pontual, de medidas, topológica e lógica. Estes espaços, denotados por , constituem uma subcategoria dos Espaços Coerentes, cujos objetos, ordenados pela inclusão, são conjuntos coerentes constituídos por subconjuntos do conjunto básico, os quais estão relacionados pela relação de coerência induzida, que estrutura a teia deste espaço. Os morfismos desta categoria são as funções de objetos geradas por funções básicas. As propriedades algébricas e relacionais destas funções básicas, externas ao processo de construção, ao se propagarem, passam a influenciar na verificação das propriedades internas das funções de objetos. Contudo, este trabalho não é um estudo categórico. A metodologia adotada utiliza a linguagem simples e intuitiva da Teoria dos Conjuntos, que possibilita a visualização e a análise dos relacionamentos existentes, não apenas entre os morfismos que envolvem os objetos totais ou parciais desta categoria, mas também das estruturas ou pré-estruturas externas que os formam, representados pelas funções de tokens e funções básicas. Mostra-se que as funções de objetos são totais e bem definidas, alem de serem monótonas e continuas neste espaço. Entretanto a análise da estabilidade, e consequentemente da linearidade esta associada a injetividade das funções básicas. Uma das características mais importantes da construção proposta e o desenvolvimento de um sistema de representação linear para funções localmente lineares, com a definição do espaço coerente A* gerado pelo produto de subteias. Neste espaço, as funções de objetos são lineares e coincidem com os morfismo da categoria dos espaços coerentes. Além disso, mostra-se que A* e isomorfo ao espaço coerente gerado pelo produto direto dos sub-espaços, ПĄ. Desta forma, toda transformação definida para um tipo de dado estruturado a partir de um conjunto básico enumerável tem uma representação linear, constituída pelos morfismos da categoria dos espaços coerentes. A existência da representação linear para as funções elementares garante a existência da representação linear para outras funções derivadas destas. Apresenta-se ainda uma especificação desta construção, introduzindo-se o Espaço Coerente de Intervalos Racionais, IIQ. Na busca de uma aplicação compatível com uma abordagem computacional, em especial para Análise Real, mostra-se que, em IIQ, cada função real elementar esta identificada com uma função de objetos linear, definida a partir da correspondente função elementar racional. Dentre as funções que foram analisadas destacam-se: a exponencial, a logarítmica, a potência, a potência estendida, a raiz n-ésima, as funções trigonométricas como seno, cosseno e tangente e suas correspondentes funções inversas, como também a função polinomial. Verificou-se que todas estas funções de objetos são totais, bem definidas, ou pertencem ou possuem uma representação linear na categoria COSP-LIN dos espaços coerentes, alem de serem fechadas para os objetos totais e quasi-totais deste espaço, sendo possível estabelecer o correspondente par-projeção para cada uma delas. / In this work the Coherence Spaces Generated by Basic Sets with additional structure are studied. By additional structure one means an algebraic, topological and logical structure with a punctual order and a measure system. These spaces, indicated by A, are a subcategory of the category of Coherence Spaces, whose objects, ordered by inclusion, are coherent sets formed by the induced web coherence relation. The morphisms of this category are the functions of objects generated by basic functions. The algebraic and relational properties of these basic functions - external to the construction process - are propagated and cause important influences in the verification of the internal properties of the functions of objects However, this research is not a categorical study. The methodology uses the simple and intuitive language of the Set Theory, which allows the visualization and the analysis of the existing relationships, not only among, the morphisms of the total and partial objects of this category, but also among their structures or pre-structures, represented by the functions of tokens and basic functions. It is shown that the functions of objects are total and well defined. They are also monotone and continuous. However the stability and the linearity of the functions of objects depend on the fact if the basic functions are injective or not. One of the most important features of this construction is the development of a linear representation system for the local linear functions, by the definition of a coherence space A*, which is generated by the subweb product. In this space the functions of objects are linear and therefore they are the morphisms of the category of Coherence Spaces. Moreover, it is proved that A* is isomorphic to the coherence space generated by the directed product of the subspaces, denoted by ПĄ . Then, for each transformation defined for a structured data type considering a denumerable basic set there exists its related linear representation. The existence of a linear representation for elementary functions guarantees the existence of a linear representation for others derived functions. As an application of this construction, the Coherence Space of Rational Intervals, denoted by IIQ, is introduced. In order to show an application which is compatible to a computational approach, specially for the real analysis, each elementary real function is identified with a linear function of objects, defined considering the related elementary rational function. Some of the analyzed functions are the exponential, the logarithmic, the power , the extended power, the root, the trigonometric (sine, cosine and tangent and their relates inverses), and the polynomial functions. It is proved that all of these functions of objects are total and well defined. Moreover, either they belong to the category COPS-LIN of the coherence spaces or they have a linear representation in the same category. It is also possible to define a related projection pair for each one of them.

Neste trabalho desenvolve-se um estudo sobre os Espaços Coerentes Gerados por Conjuntos Básicos, dotados de uma estrutura adicional. Por estrutura adicional entende-se uma estrutura algébrica, de ordem pontual, de medidas, topológica e lógica. Estes espaços, denotados por , constituem uma subcategoria dos Espaços Coerentes, cujos objetos, ordenados pela inclusão, são conjuntos coerentes constituídos por subconjuntos do conjunto básico, os quais estão relacionados pela relação de coerência induzida, que estrutura a teia deste espaço. Os morfismos desta categoria são as funções de objetos geradas por funções básicas. As propriedades algébricas e relacionais destas funções básicas, externas ao processo de construção, ao se propagarem, passam a influenciar na verificação das propriedades internas das funções de objetos. Contudo, este trabalho não é um estudo categórico. A metodologia adotada utiliza a linguagem simples e intuitiva da Teoria dos Conjuntos, que possibilita a visualização e a análise dos relacionamentos existentes, não apenas entre os morfismos que envolvem os objetos totais ou parciais desta categoria, mas também das estruturas ou pré-estruturas externas que os formam, representados pelas funções de tokens e funções básicas. Mostra-se que as funções de objetos são totais e bem definidas, alem de serem monótonas e continuas neste espaço. Entretanto a análise da estabilidade, e consequentemente da linearidade esta associada a injetividade das funções básicas. Uma das características mais importantes da construção proposta e o desenvolvimento de um sistema de representação linear para funções localmente lineares, com a definição do espaço coerente A* gerado pelo produto de subteias. Neste espaço, as funções de objetos são lineares e coincidem com os morfismo da categoria dos espaços coerentes. Além disso, mostra-se que A* e isomorfo ao espaço coerente gerado pelo produto direto dos sub-espaços, ПĄ. Desta forma, toda transformação definida para um tipo de dado estruturado a partir de um conjunto básico enumerável tem uma representação linear, constituída pelos morfismos da categoria dos espaços coerentes. A existência da representação linear para as funções elementares garante a existência da representação linear para outras funções derivadas destas. Apresenta-se ainda uma especificação desta construção, introduzindo-se o Espaço Coerente de Intervalos Racionais, IIQ. Na busca de uma aplicação compatível com uma abordagem computacional, em especial para Análise Real, mostra-se que, em IIQ, cada função real elementar esta identificada com uma função de objetos linear, definida a partir da correspondente função elementar racional. Dentre as funções que foram analisadas destacam-se: a exponencial, a logarítmica, a potência, a potência estendida, a raiz n-ésima, as funções trigonométricas como seno, cosseno e tangente e suas correspondentes funções inversas, como também a função polinomial. Verificou-se que todas estas funções de objetos são totais, bem definidas, ou pertencem ou possuem uma representação linear na categoria COSP-LIN dos espaços coerentes, alem de serem fechadas para os objetos totais e quasi-totais deste espaço, sendo possível estabelecer o correspondente par-projeção para cada uma delas. / In this work the Coherence Spaces Generated by Basic Sets with additional structure are studied. By additional structure one means an algebraic, topological and logical structure with a punctual order and a measure system. These spaces, indicated by A, are a subcategory of the category of Coherence Spaces, whose objects, ordered by inclusion, are coherent sets formed by the induced web coherence relation. The morphisms of this category are the functions of objects generated by basic functions. The algebraic and relational properties of these basic functions - external to the construction process - are propagated and cause important influences in the verification of the internal properties of the functions of objects However, this research is not a categorical study. The methodology uses the simple and intuitive language of the Set Theory, which allows the visualization and the analysis of the existing relationships, not only among, the morphisms of the total and partial objects of this category, but also among their structures or pre-structures, represented by the functions of tokens and basic functions. It is shown that the functions of objects are total and well defined. They are also monotone and continuous. However the stability and the linearity of the functions of objects depend on the fact if the basic functions are injective or not. One of the most important features of this construction is the development of a linear representation system for the local linear functions, by the definition of a coherence space A*, which is generated by the subweb product. In this space the functions of objects are linear and therefore they are the morphisms of the category of Coherence Spaces. Moreover, it is proved that A* is isomorphic to the coherence space generated by the directed product of the subspaces, denoted by ПĄ . Then, for each transformation defined for a structured data type considering a denumerable basic set there exists its related linear representation. The existence of a linear representation for elementary functions guarantees the existence of a linear representation for others derived functions. As an application of this construction, the Coherence Space of Rational Intervals, denoted by IIQ, is introduced. In order to show an application which is compatible to a computational approach, specially for the real analysis, each elementary real function is identified with a linear function of objects, defined considering the related elementary rational function. Some of the analyzed functions are the exponential, the logarithmic, the power , the extended power, the root, the trigonometric (sine, cosine and tangent and their relates inverses), and the polynomial functions. It is proved that all of these functions of objects are total and well defined. Moreover, either they belong to the category COPS-LIN of the coherence spaces or they have a linear representation in the same category. It is also possible to define a related projection pair for each one of them.

Des études expérimentales examinant le rôle socialisateur de l’autorité parentale dans des contextes de transgression persistante de règles ont démontré les avantages d’utiliser la conséquence logique plutôt que d’autres stratégies d’autorité (la punition, le raisonnement et l’absence d’autorité). À l’aide d’une méthodologie par vignettes expérimentales et d’un échantillon de 214 adolescents (M = 15,28 ans), la présente étude a bonifié ces recherches en comparant ces mêmes stratégies dans un contexte de transgression à facettes multiples. Spécifiquement, le rôle modérateur des perceptions des adolescents de l’enjeu sous-jacent à la transgression (personnel c. non-personnel) sur leurs croyances quant au potentiel socialisateur des stratégies d’autorité a été évalué. Lorsque les adolescents ont catégorisé la transgression comme un enjeu non-personnel, les résultats des études antérieures ont été reproduits. En effet, la conséquence logique a été jugée comme étant aussi efficace que la punition à prévenir une transgression future (et plus efficace que le raisonnement et l’absence d’autorité) et comme la stratégie la plus acceptable. De plus, contrairement à la punition, la conséquence logique n’a pas été perçue comme frustrant plus l’autonomie que le raisonnement. En comparaison, les adolescents ayant catégorisé la transgression comme un enjeu personnel ont perçu la conséquence logique moins favorablement, laissant place au raisonnement comme stratégie d’autorité préférable. Les implications de ces résultats pour la socialisation des adolescents et les pratiques parentales optimales sont discutées. / Experimental studies focusing on the socialization role of parental authority exertion in persistent rule-breaking contexts involving non-personal issues have recently shown the advantages of using logical consequences over alternative strategies (mild punishments, reasoning and no-authority). Using an experimental vignette approach and a sample of 214 adolescents (Mage = 15.28 years), the present study extended these findings by comparing the same parental interventions in a hypothetical rule-breaking setting involving a multifaceted issue. Specifically, and based on research anchored in social domain theory, we evaluated how adolescents’ perceptions of the issue underlying the multifaceted transgression (personal vs. non-personal) moderated their beliefs regarding authority exertion strategies. When adolescents perceived the transgression as a non-personal issue, past results were replicated and enhanced. Adolescents perceived the logical consequence as at least as effective as the mild punishment to prevent future transgressions (i.e., more so than reasoning and no-authority) and as the most acceptable strategy. Furthermore, contrary to the mild punishment, they did not perceive the logical consequence as more autonomy-thwarting than reasoning. In contrast, adolescents who categorized the transgression as a personal matter rated the logical consequence less favorably, leaving reasoning as a preferred form of intervention. Implications for optimal parenting are discussed.

Os números naturais parciais são informações parciais sobre números naturais e w. As propriedades matemáticas do domínio de números naturais parciais e de funções que envolvem este domínio são estudadas via o use da teoria dos domínios de Scott. A manipulação formal de naturais parciais é estudada mediante a utilização de um calculo-ג tipado com constantes. Relações com a teoria da recursão são estudadas. É mostrado como funções continuas entre naturais parciais podem representar processos interativos, possivelmente perpétuos. / Partial natural numbers are informations about natural numbers and w. Mathematical properties of the domain of partial natural numbers and functions involving this domain are investigated with the aid of Scott domain theory. A typed ג-calculus is introduced for investigating formal manipulation of partial natural numbers. Relations with recursion theory are investigated. It is shown how continuous functions on natural numbers can represent (possibly perpetual) interactive processes.

Os números naturais parciais são informações parciais sobre números naturais e w. As propriedades matemáticas do domínio de números naturais parciais e de funções que envolvem este domínio são estudadas via o use da teoria dos domínios de Scott. A manipulação formal de naturais parciais é estudada mediante a utilização de um calculo-ג tipado com constantes. Relações com a teoria da recursão são estudadas. É mostrado como funções continuas entre naturais parciais podem representar processos interativos, possivelmente perpétuos. / Partial natural numbers are informations about natural numbers and w. Mathematical properties of the domain of partial natural numbers and functions involving this domain are investigated with the aid of Scott domain theory. A typed ג-calculus is introduced for investigating formal manipulation of partial natural numbers. Relations with recursion theory are investigated. It is shown how continuous functions on natural numbers can represent (possibly perpetual) interactive processes.

Os números naturais parciais são informações parciais sobre números naturais e w. As propriedades matemáticas do domínio de números naturais parciais e de funções que envolvem este domínio são estudadas via o use da teoria dos domínios de Scott. A manipulação formal de naturais parciais é estudada mediante a utilização de um calculo-ג tipado com constantes. Relações com a teoria da recursão são estudadas. É mostrado como funções continuas entre naturais parciais podem representar processos interativos, possivelmente perpétuos. / Partial natural numbers are informations about natural numbers and w. Mathematical properties of the domain of partial natural numbers and functions involving this domain are investigated with the aid of Scott domain theory. A typed ג-calculus is introduced for investigating formal manipulation of partial natural numbers. Relations with recursion theory are investigated. It is shown how continuous functions on natural numbers can represent (possibly perpetual) interactive processes.

L'objectif de cette thèse est de modéliser les changements en chaîne historiques dans les théories phonologiques modernes. Martinet (1952, 1955, 1970) distingue deux principales catégories de changements en chaîne: les chaînes de propulsion et les chaînes de traction. Les chaînes de propulsion seront traitées dans la première partie de cette thèse (chapitres 2 et 3). Afin de modéliser les chaînes de propulsion, un modèle existant est utilisé: La théorie de préservation du contraste de Łubowicz (2003, 2012). Concernant la modélisation des chaînes de traction, qui sera principalement traitée dans la seconde partie de le thèse (chapitres 4 et 5), une nouvelle théorie sera proposée, basée sur les travaux phonétiques de De Boer (2001). Cette proposition est basée sur la notion des domaines piliers. Le chapitre 1 s'ouvre sur l'observation que ni les théories dérivationnelles orientées vers l'input – type SPE – ni les théories orientées vers l'output, comme la théorie de l'optimalité de Prince & Smolensky (TO) sont en mesure de formaliser les changements en chaîne. Dans les théories dérivationnelles, il est possible d'obtenir une bonne description des changements en chaîne. Cependant, afin de faire cela, nous devons briser les liens unissant les différentes étapes des changements en chaîne. Par conséquent, nous perdons l'interdépendance systémique entre les stades, ce qui est au cœur des changements en chaîne. Concernant la théorie de l'optimalité, sa structure stricte sur deux niveaux se révèle problématique dans le cas des changements en chaîne. Dans un changement en chaîne schématique comme /a/ → [e] > /e/ → [i], il est impossible pour la TO, dans la même analyse, de faire en sorte que [e] soit un candidat optimal et /e/ un candidat bloqué. Dans le chapitre 2, nous analysons et comparons plusieurs amendements à la théorie de l'optimalité proposés pour rendre compte des changements en chaîne. Beaucoup d'entre eux ne sont pas satisfaisants dans le cas des changements en chaîne car ils ne parviennent pas à modéliser la force systémique des changements en chaîne. La seule théorie qui semble accomplir cet objectif est la théorie de préservation du contraste (CPT) de Łubowicz. Cette théorie est basée sur les notions de contraste, préservation du contraste et elle évalue des scénarios au lieu de candidats individuels. La préservation du contraste est accomplie à travers la transformation du contraste. La notion de transformation du contraste signifie que les niveaux de contraste du système pré-changement sont préservés dans le système post-changement mais leur nature change. Une observation sur la CPT est que Łubowicz l'avait originellement conçue pour les changements en chaîne synchroniques. Dans le chapitre 2, nous démontrons que la CPT est en mesure de modéliser la force systémique qui donne aux changements en chaîne diachroniques leur cohérence. Ainsi, dans le chapitre 3, nous décidons de tester la validité de la CPT sur de véritables changements en chaîne historiques. Nous avançons deux études de cas sur deux changements en chaîne. Le premier, bien connu, est le changement en chaîne vocalique du Grand Changement Vocalique Anglais (English Great Vowel Shift). L'autre changement en chaîne est la Seconde Mutation Consonantique allemande (High German Consonant Shift). Nous démontrons que la CPT est en mesure de rendre compte de ces deux changements en chaîne. Au terme du chapitre 3, la CPT est testée sur une autre catégorie de changements en chaîne : les changements en chaîne observés dans l'acquisition de la langue maternelle. Le changement en chaîne étudié est s → θ → f. Ce changement en chaîne est différent des deux autres car il n'y a pas de nouveau niveau de contraste créé. Le dernier stade du changement en chaîne se conclut par une fusion. Encore une fois, la CPT peut rendre compte de ce changement en chaîne de manière satisfaisante. / The aim of this thesis is to formalise historical chain shifts within modern phonological theories. Martinet (1952, 1955, 1970) distinguishes two main categories of chain shifts: push chains and pull chains. Push chains will be dealt with in the first main part of this thesis (chapters 2 and 3). For modelling of push chains, an existing model is used: Łubowicz's (2003, 2012) Contrast Preservation Theory. For modelling pull chains, which will be the focus of the second part of this thesis (chapters 4 and 5), a new theory will be proposed, based on the phonetic work by De Boer (2001). This proposition is based on the notion of the pillar domains (domaines piliers). Chapter 1 starts with the observation that neither the derivational, SPE-like, theories or output driven theories like Prince's & Smolensky's Optimality Theory (OT) are able to formalise chain shifts properly. Within the derivational theories, it is possible to get a correct description of a chain shift. However, in order to do that, we have to break the links between the different stages of the chain shift. Therefore, one loses the systemic interdependence of the different stages, which is the essence of the chain shift. As regards to Optimality Theory, it is the two-level structure of the theory which proves problematic for chain shifts. In a theoretical chain shift such as /a/ → [e] > /e/ → [i], it is impossible for OT to make, in the same analysis, [e] an optimal candidate and /e/ a non-optimal one. In chapter 2, we analyse and compare several output-driven propositions that have been made to account for chain shifts. Many of them are not satisfactory because they do not manage to model the systemic motivation of chain shifts. The only theory that seems able to complete this objective is Łubowicz's Contrast Preservation Theory (CPT). This theory is based on the notions of contrast, contrast preservation and it evaluates scenarios rather than individual candidates. Contrast preservation is achieved through contrast transformation. The term contrast transformation indicates that the contrast levels of the input system are preserved but that their nature is different in the output system. One observation about CPT is that Łubowicz designed it for synchronic chain shifts. In chapter 2, we prove that CPT is able to model the perceptible coherence in diachronic chain shifts. In chapter 3, thus we decide to test the validity of CPT on actual historical chain shifts. It is tested on two cases. One is the well-known vocalic chain shift the Great Vowel Shift in English. The other one is the Second German Sound Shift (or High German Consonant Shift). We demonstrate that CPT is able to account for both of these chain shifts. At the end of chapter 3, CPT is tested on another type of chain shifts, i.e. chain shifts in first-language acquisition. The studied chain shift is s → θ → f. This chain shift is different from the other two because there is no new contrast level created. The last stage of this chain shift ends with a merger . Once again, CPT can account for this chain shift in a coherent way.