The primary concern of this piece of work is to reconsider scientific realism debate in the philosophy of science. Accordingly, the overall aim is to come up with the clues of a viable scientific realist attitude in the face of anti-realist interpretations of scientific theories. To accomplish this aim, I make use of two modified versions of scientific realism, that is, &lsquo / epistemic structural realism&rsquo / and &lsquo / entity realism&rsquo / . Epistemic structural realism is a realist position of which proponents claim that the only knowable part of the reality is the structure of it which is expressed by the mathematical equations of our best scientific theories. On the other hand, according to entity realism, the only assured knowledge obtained from scientific theories is the existence of theoretical entities posited by these theories. I argue that a combination of the properly construed versions of these two positions might fulfill the afore-mentioned aim of this thesis.

History is descriptive. Epistemology is conceived as normative. It appears, then, that a historical approach to epistemology, like historical epistemology, might not be epistemically normative. In our context here, epistemology is not a systematic theory of knowledge, truth, or justification. In this thesis I approach epistemic justification through the vantage point of practice of science. Practice is about reasoning. Reasoning, conceived as the human propensity to order perceptions, beliefs, memories, etc., in ways that permit us to have understanding, is not only about thinking. Reasoning has to do with our actions, too: In the ordering of reasoning we take into account the desires of ourselves and others. Reasoning has to do with tinkering with stuff, physical or abstract. Practice is primarily about skills. Practices are not mere groping. They have a form. Performing according to a practice is an activity with a lot of plasticity. The skilled performer retains the form of the practice in many different situations. Finally, practices are not static in time. Practices develop. People try new things, some of which may work out, others not. The technology involved in how to go about doing things in a particular practice changes, and the concepts concerning understanding what one is doing also may change. This is the point where history enters the picture. In this thesis I explore the interactions between history, reasoning, and skills from the viewpoint of a particular type of epistemic justification: a priori justification. An a priori justified proposition is a proposition which is evident independent of experience. Such propositions are self-evident. We will make sense of a priori justification in a context of regarding science as practice, so that we will be able to demonstrate that the latter accommodates the normative character of science.

The topic of this diploma thesis is the position of scientific realism presented in the framework of naturalized philosophy of science. The aim is to clarify this position and to show that if one denies realism, scientific practice does not make sense. For this purpose main focus is first devoted to the key parts (metaphysical realism, semantic realism, and epistemological realism) which constitute the scientific realism. Next, a detailed analysis of the arguments against and in favor of realism is offered, and concrete examples taken from the sciences are used to illustrate key points. Space is also devoted to the analysis of a physical theory of heat of the 18th and 19th centuries in connection with an antirealist argument directed at the history of science. Also, one of the few fully elaborated antirealist positions, constructive empiricism, is presented and critically evaluated. In a similar fashion, this thesis pays attention to a specific form of realism, called entity realism. Though the resulting image is a thoroughly realist position, this position strives to accurately capture the numerous nuances of the scientific practice, offering a fresh perspective on some of the traditional views.

Dans ce travail, nous analysons le concept de manipulation à travers une théorie élargie qui inclut le réalisme des entités et la théorie causale de la manipulation. Dans une seconde étape, nous appliquons cette théorie à plusieurs expériences scientifiques. Cette thèse est articulée autour de quatre parties : Dans la Partie I nous exposons le concept de manipulation. Nous examinons le réalisme structurel en mettant l’accent sur ses difficultés concernant particulièrement les incertitudes ontologiques de la physique. Cet examen du réalisme structurel nous permettra de développer un réalisme des entités qui est susceptible de constituer une alternative au réalisme structurel.La partie II est consacrée aux aspects clés du réalisme des entités, c’est-à-dire ses fondements, ses arguments ainsi qu’une définition aussi précise que possible du concept de manipulation.Dans la partie III, nous élaborons une théorie générale de la manipulation/intervention qui est basée sur le réalisme des entités et la théorie de Woodward sur l’intervention, l’invariance et la profondeur de l’explication des évènements causaux. Nous allons développer trois éléments : 1-un système des conditions de l’intervention, 2- une typologie de l’intervention, 3-l’intervention probabiliste. Ce dernier type de l’intervention est applicable à l’expérience EPR. Dans la partie IV- Nous appliquons cette théorie à cinq expériences scientifiques : l’expérience EPR, l’expérience de l’électrodynamique quantique en cavité, l’expérience sur l’effet Zeeman, l’expérience du Tevatron (collisions de protons) et la dernière est non des moindres, est une expérience d’astrophysique (ondes gravitationnelles). Dans chacune de ces expériences, notre théorie a trouvé une application réussie. Toutes les conditions de l’intervention sont applicables sur les cinq expériences et une typologie bien définie de l’intervention a été élaborée et qui met en évidence les résultats de l’intervention dans les différentes séquences de ces expériences.Par conséquent, notre théorie de la manipulation reflète une utilité authentique et réelle dans différent domaines de la physique et de l’astrophysique. Un pareil résultat est important tenant compte du rôle de la manipulation expérimentale pour la science moderne. / In this work, we analyze the concept of manipulation through a comprehensive theory which includes the entity realism and the causal theory of manipulation. At the second stage, we implement this theory to several scientific experiments. It is divided into four parts:In Part I, we will have a long journey until we will reach the true concept of manipulation. We analyze the structural realism and we explore its difficulties regarding ontological problems of physics. Our survey of structural realism enables to open a new way to developing entity realism instead of structural realism.Part II will be dedicated to the core aspect of entity realism such as its foundations, its arguments and our aim to reach a global definition of manipulation concept.In Part III I will achieve my aim which is elaborating a global theory of manipulation/intervention based on entity realism and the Woodward theory of intervention, invariance and depth explanation of causal events. We will develop three items : 1- a system of intervention conditions 2-a typology of intervention, 3-probabilistic intervention. Such type of intervention is applicable to EPR experiment. In Part IV, we implement our theory to five scientific experiments: EPR experiment, electrodynamics cavity experiment, Zeeman effect experiment, Tevatron experiment (collision of protons) and last but not least an astrophysical experiment (gravitational waves).In each case, our theory is successful. All intervention conditions are applicable to our five experiments and a well defined typology of intervention is developed based on intervention results through different stages of those experiments. Therefore, the theory of manipulation shows a real and genuine application in different domains of physics and astrophysics. Such result is important regarding the role of manipulation in modern science.