The growth of digitization in the cultural heritage domain offers great possibilities to broaden the boundaries of historical research. With the ultimate aim of creating social networks of person names from news articles, we introduce a person name disambiguation method that exploits the relation between the ambiguity of a person name and the number of entities referred to by it. Modeled as a clustering problem with a strong focus on social relations, our system dynamically adapts its clustering strategy to the most suitable configuration for each name depending on how common this name is. Our method's performance is on par with the state-of-the-art reported for the CRIPCO dataset, while using less specific resources.