Periodic Reporting for period 1 - MODEL TRANSFER (Model Transfer and its Challenges in Science: The Case of Economics)
Okres sprawozdawczy: 2023-04-01 do 2025-09-30
Scientific research is characterized by strong disciplinary specialization that often manifests in highly abstract models tailored to particular target systems. At the same time, a salient feature of contemporary science is that there is increasing interaction across different fields. One way this interaction manifests is in the transfer of models. Beyond their relevance to scientific practice, such model transfers play a crucial role because they can act as catalysts for the advancement of science. As such, they raise core philosophical questions such as why models are transferable across domains, how they can address fundamentally distinct problems, and how they contribute to scientific progress.
Despite an increasing interest among historians and philosophers of science in the phenomenon, model transfer has not yet received its due attention. This project will therefore undertake a comprehensive philosophical investigation of model transfer, its challenges, and its implications for scientific progress in order to place this practice at the center of the agenda in philosophy and history of scientific modeling. We do so by combining approaches from philosophy and history of science with computational methods that are themselves new to philosophy. For our case studies, we focus primarily on economics.
We study model transfer on three levels:
Methodological and conceptual tools (SP1): Current studies of model transfer rely primarily on the case study methodology, which faces limitations like selection bias and underdetermination. We overcome these limitations by substantiating systematic analysis with a combination of detailed historical research and quantitative-empirical mappings and visualizations of model transfer processes. We will also critically reflect on this integrative methodology to better understand the potentials and limitations of our approach for philosophy of science generally.
Philosophical investigation (SP2): Current literature focuses on why models are transferable, emphasizing template-based approaches. While those approaches serve as an ideal starting point for thinking about model transfer, they require more elaboration to serve as the basis for a comprehensive philosophical investigation of model transfer, its challenges, and its relationship to scientific progress. Furthermore, there is no consensus on what exactly is transferred or what features enable transfer. Additionally, the modification of models during transfer and their impact on theoretical frameworks require clarification. Finally, there are no comprehensive analyses of the obstacles to transfer, such as structural differences between source and target domains and resistance to transfer due to methodological or epistemological commitments. Our team addresses all of these issues.
Relation to scientific progress (SP3): While model transfer is acknowledged as significant, its precise role in enabling progress remains unclear. The project systematically analyzes not only the epistemic benefits of model transfer in specific cases but the more general consequences for a domain, particularly in the context of model-based disciplines like economics.
Despite an increasing interest among historians and philosophers of science in the phenomenon, model transfer has not yet received its due attention. This project will therefore undertake a comprehensive philosophical investigation of model transfer, its challenges, and its implications for scientific progress in order to place this practice at the center of the agenda in philosophy and history of scientific modeling. We do so by combining approaches from philosophy and history of science with computational methods that are themselves new to philosophy. For our case studies, we focus primarily on economics.
We study model transfer on three levels:
Methodological and conceptual tools (SP1): Current studies of model transfer rely primarily on the case study methodology, which faces limitations like selection bias and underdetermination. We overcome these limitations by substantiating systematic analysis with a combination of detailed historical research and quantitative-empirical mappings and visualizations of model transfer processes. We will also critically reflect on this integrative methodology to better understand the potentials and limitations of our approach for philosophy of science generally.
Philosophical investigation (SP2): Current literature focuses on why models are transferable, emphasizing template-based approaches. While those approaches serve as an ideal starting point for thinking about model transfer, they require more elaboration to serve as the basis for a comprehensive philosophical investigation of model transfer, its challenges, and its relationship to scientific progress. Furthermore, there is no consensus on what exactly is transferred or what features enable transfer. Additionally, the modification of models during transfer and their impact on theoretical frameworks require clarification. Finally, there are no comprehensive analyses of the obstacles to transfer, such as structural differences between source and target domains and resistance to transfer due to methodological or epistemological commitments. Our team addresses all of these issues.
Relation to scientific progress (SP3): While model transfer is acknowledged as significant, its precise role in enabling progress remains unclear. The project systematically analyzes not only the epistemic benefits of model transfer in specific cases but the more general consequences for a domain, particularly in the context of model-based disciplines like economics.
Methodological and conceptual tools (SP1): The research team has investigated model transfer cases, including agent-based models, rational choice models, and foundation models, using traditional historical methods, such as oral history and case study methodology. We have also begun to apply state-of-the-art computational methods, such as network analysis and LLMs, and will do so more extensively in the coming years. We have already made methodological contributions that reflect on the usefulness and limitations of (empirical) network analysis in philosophy of science. We expect that by showcasing the development and application of these digital humanities methods in philosophy and history of science, the methodological toolbox of philosophy of science will be significantly extended. Insights from this work have informed research in SP2 and SP3, leading to an account of model transfer as model construction, a taxonomy of transfer processes, a typology of challenges to model transfer, and an analysis of the unit of transfer to provide an account of progress in model-based fields.
Philosophical investigation (SP2): The team has explored model transfer through two key philosophical dimensions: the unit of transfer and the transfer process, particularly in the context of economics. This research critically assesses and extends the template-based account of model transfer by focusing in particular on the model construction process involved in model transfer, examining the importance of the concept of a formal template, and evaluating the usefulness of formal templates for analyzing model transfers in economics. Additionally, the team is conducting a systematic analysis of the epistemic challenges that arise from model transfer and their potential harms to scientific communities. This research results in a taxonomy of transfer processes and a typology of challenges that hinder successful model transfer.
Relation to scientific progress (SP3): Our research group has initiated work on exploring the relationship between model transfer and progress, with particular attention to questions of unification. We are developing a novel account of scientific progress that emphasizes the creation and accumulation of formal templates for model construction. We anticipate that this framework will be especially valuable for understanding progress in mathematically oriented areas of the social sciences—such as economics—where traditional philosophical accounts often fall short.
Philosophical investigation (SP2): The team has explored model transfer through two key philosophical dimensions: the unit of transfer and the transfer process, particularly in the context of economics. This research critically assesses and extends the template-based account of model transfer by focusing in particular on the model construction process involved in model transfer, examining the importance of the concept of a formal template, and evaluating the usefulness of formal templates for analyzing model transfers in economics. Additionally, the team is conducting a systematic analysis of the epistemic challenges that arise from model transfer and their potential harms to scientific communities. This research results in a taxonomy of transfer processes and a typology of challenges that hinder successful model transfer.
Relation to scientific progress (SP3): Our research group has initiated work on exploring the relationship between model transfer and progress, with particular attention to questions of unification. We are developing a novel account of scientific progress that emphasizes the creation and accumulation of formal templates for model construction. We anticipate that this framework will be especially valuable for understanding progress in mathematically oriented areas of the social sciences—such as economics—where traditional philosophical accounts often fall short.
Our philosophical accounts place the phenomenon of model transfer at the heart of the philosophical agenda. Our framework provides a fresh perspective on longstanding epistemic questions about models and progress in model-based fields. Our historical research of model transfers and their challenges in economics opens a new perspective on the recent history of economics and its organization as a scientific field. The cases we examine yield new insights into model transfer in science more broadly, providing a much-needed bridge between philosophy of science, Integrated History and Philosophy of Ccience, and the social sciences.
By combining philosophical-conceptual results with detailed historical and empirical insights, our research sets a new standard for multi-method approaches in HPS, general philosophy of science, and philosophy of science in practice. The methodological tools we develop are broadly applicable to future studies of model transfer, fostering ongoing, systematic dialogue between scientific practice and the formation of philosophical concepts. In doing so, we advance the program of empirical and digital philosophy of science.
Finally, we expect the results of this project to have significant implications for society as a whole, offering a fresh perspective on the organization of science, the role of models in advancing knowledge, and the nature of interdisciplinary research. These topics are highly relevant to a wide range of stakeholders, including policymakers, research institutions, and the public. A better understanding of how scientific models shape knowledge production and cross-disciplinary collaboration will inform decision-making, enhance public trust in science, and foster more effective research policies.
By combining philosophical-conceptual results with detailed historical and empirical insights, our research sets a new standard for multi-method approaches in HPS, general philosophy of science, and philosophy of science in practice. The methodological tools we develop are broadly applicable to future studies of model transfer, fostering ongoing, systematic dialogue between scientific practice and the formation of philosophical concepts. In doing so, we advance the program of empirical and digital philosophy of science.
Finally, we expect the results of this project to have significant implications for society as a whole, offering a fresh perspective on the organization of science, the role of models in advancing knowledge, and the nature of interdisciplinary research. These topics are highly relevant to a wide range of stakeholders, including policymakers, research institutions, and the public. A better understanding of how scientific models shape knowledge production and cross-disciplinary collaboration will inform decision-making, enhance public trust in science, and foster more effective research policies.