Modeling data complexity in public history and cultural heritage

The publication by Galleries, Libraries, Archives and Museums of metadata about their collections is fundamental for the creation of our shared digital cultural heritage. Yet, we notice, these digital collections are, on one hand, of little use to scholars (because of the inconsistent quality of the published records), and, on the other hand, they fail to attract the interest of the general public (because of their dry content). These problems are exacerbated by the current move towards public history, where citizens are no longer just passive actors, but play an active role in contributing, maintaining and curating historical records, leading some to question the trustworthiness of collections in which non scholars have the ability to contribute. The core issue behind all these problems is, we argue, a (doomed) search for objectivity, often caused by the fact that data models ignore the derivative and stratified nature of cultural objects, and allow only one point of view to be expressed. In turn this forces the publication of bowdlerized records and removes any venue for the expression of disagreement and different opinions. We propose an approach named contexts to solve these issues. The adoption of contexts makes it possible to support multiple points of view inside the same dataset, allowing not only multiple scholars to provide their own possibly contrasting points of view, but also making it possible to incorporate additions, corrections and more complex kinds of commentaries from citizens without compromising the trustworthiness of the whole dataset.