Painter Deryc T, Daniels Bryan C, Laubichler Manfred D
School of Complex Adaptive Systems, Arizona State University, Tempe, AZ, USA.
Max Planck Institute for the History of Science, Berlin, Germany.
Theory Biosci. 2021 Nov;140(4):391-399. doi: 10.1007/s12064-021-00359-1. Epub 2021 Nov 12.
The origins of innovation in science are typically understood using historical narratives that tend to be focused on small sets of influential authors, an approach that is rigorous but limited in scope. Here, we develop a framework for rigorously identifying innovation across an entire scientific field through automated analysis of a corpus of over 6000 documents that includes every paper published in the field of evolutionary medicine. This comprehensive approach allows us to explore statistical properties of innovation, asking where innovative ideas tend to originate within a field's pre-existing conceptual framework. First, we develop a measure of innovation based on novelty and persistence, quantifying the collective acceptance of novel language and ideas. Second, we study the field's conceptual landscape through a bibliographic coupling network. We find that innovations are disproportionately more likely in the periphery of the bibliographic coupling network, suggesting that the relative freedom allowed by remaining unconnected with well-established lines of research could be beneficial to creating novel and lasting change. In this way, the emergence of collective computation in scientific disciplines may have robustness-adaptability trade-offs that are similar to those found in other biosocial complex systems.
科学创新的起源通常是通过历史叙事来理解的，这些叙事往往聚焦于一小部分有影响力的作者，这种方法严谨但范围有限。在这里，我们开发了一个框架，通过对6000多篇文献的语料库进行自动分析，严格识别整个科学领域的创新，这些文献包括进化医学领域发表的每一篇论文。这种全面的方法使我们能够探索创新的统计特性，研究创新思想在一个领域预先存在的概念框架内倾向于何处产生。首先，我们基于新颖性和持久性开发了一种创新度量方法，量化新语言和新思想的集体接受程度。其次，我们通过文献耦合网络研究该领域的概念格局。我们发现，在文献耦合网络的边缘出现创新的可能性要大得多，这表明与成熟的研究路线保持不相关所带来的相对自由可能有利于创造新颖且持久的变化。通过这种方式，科学学科中集体计算的出现可能存在与其他生物社会复杂系统中类似的稳健性 - 适应性权衡。
