Crombie, Hacking and the Emergence of Two New Styles: Data-Intensive and Network-Relational Reasoning

The classical taxonomy of Western scientific styles introduced by Crombie and reinterpreted by Hacking has framed how historians and philosophers understand the plurality of scientific reasoning. Yet contemporary scientific practice has expanded beyond the six established styles, revealing new ways of generating truth, constructing objects and defining evidence. We argue that two emerging modes of inquiry qualify as additional styles within the Crombie–Hacking framework. The first is the data-intensive style, whose knowledge production relies on large datasets, algorithmic extraction of patterns and high-dimensional stability rather than explicit theories or controlled experiments. It creates new scientific objects such as embeddings, latent structures and predictive signatures. Still, it introduces its own epistemic norms grounded in robustness, cross-validation and scale. The second emerging mode is the network-relational style, which treats relations rather than intrinsic properties as the basis of explanation. It creates objects such as motifs, connectomes, influence pathways and multilayer networks. Still, it establishes truth through structural coherence and relational stability across independent observations. Both modes satisfy Hacking’s criteria for a scientific style: they generate new questions, create new ontological domains, authenticate themselves through internal standards and support statements that were previously impossible. By comparing these two styles with the classical six, we argue that the Crombie–Hacking taxonomy requires expansion to capture the epistemic transformations of contemporary science. The result is a more inclusive and pluralistic account of scientific rationality that reflects the current practices in biology, neuroscience, physics, epidemiology and data science.