Due to the rapid accumulation of digital and digitized cultural data, we are witnessing the emergence of culture analytics in the humanities, a research field which combines Digital Humanities and humanities domain expertise (culture, history, and language) with computational and data-intensive methods. Culture analytics pursued so far has neglected to identify the fundamental processes of culture, instead it has followed mainstream data science by relying heavily on machine learning to model data. While machine learning is extremely important for predictive modeling and task automation, it primarily uses a black box approach, which is ’blind’ to the mechanisms and principles that underlie cultural data. This black box makes it hard to translate data models back to humanistic research questions and prevents the humanities from contributing to a new data-driven understanding of culture.

To overcome mainstream limitations and advance culture analytics, we propose to build the Culture Analytics Network (CAN) that links renowned applied mathematicians from China and US with Danish humanities researchers who share an interest in culture analytics. The purpose of CAN is to propose a calculus of culture that circumvents the black box by identifying fundamental processes of culture through applications of concepts and state-of-the-art models from complexity science to big cultural data. The outcome of CAN will be essential for developing valid measures of culture’s impact on sociopolitical dynamics and for showcasing the potential of humanities domain expertise in the age of big data.