Modeling Nature: The Scientific Data Revolution

技术发展使科学能够收集或产生大量复杂的数据集，这些数据集无法通过传统方法进行分析或理解。相反，计算机密集建模技术被用来通过生成数据模型来获得对数据的控制。数据通过进一步的建模被简化、精炼和操作。科学观察和实验的最终产物越来越多地以模型而不是理论的形式出现。模型对实验设计至关重要，是探测器和实验条件设计和控制的基础。在一个领域建立的模型被用作研究其他现象的探针。仿真模型提供了另一种实验数据来源。当代科学哲学对这种建模实践的关注相对较少。Suppe博士在这项研究中的第一个目标是提供对当代科学建模实践的详细理解。其次，他正在探索当代建模实践对理解科学知识的影响。第三，他从集合论结构之间的映射和嵌入关系的角度对科学建模进行了形式化的结构分析，将这种分析扩展到包含科学模型、科学理论和元数学模型的一般分析，并详细介绍了科学模型和理论之间的关系。第四，他将为科学和技术研究领域提供当代建模技术的适当描述和插图——这是科学家通常不会发表的。

Technological developments enable science to gather or produce massively complex data sets which cannot be analyzed or understood via traditional methods. Instead, computer-intensive modeling techniques are used to gain control of the data by producing models of data. Data are reduced, refined, and manipulated by further modeling. The end-products of scientific observation and experiment increasingly are in the form of models not theories. Models are essential to experimental design, underlying the design and control of detectors and experimental conditions. Models developed in one area are used as probes to investigate other phenomena. simulation models provide an alternative source of experimental data. Philosophy of science has paid relatively little attention to such a contemporary modeling practices. Dr. Suppe's first objective in this research is to provide a detailed understanding of contemporary scientific modeling practices. Second, he is exploring the implications of contemporary modeling practices for understanding scientific knowledge. Third he is providing a formal structural analysis of scientific modeling in terms of mapping and embedding relations between set-theoretic structures, extend that analysis to a general analysis that encompasses scientific models, scientific theories, and metamathematical models, and detail relations between scientific models and theories. Fourth, he is going to provide the field of Science and Technology Studies with suitable descriptions and illustrations of contemporary modeling techniques--which scientists typically do not publish.