The epistemological affordances of technologies such as the Internet and computers are – yet again – offering promising and threatening opportunities to reshape humanistic research. The large digitization efforts within humanities has created new kinds of ‘big data’ textual source materials only a ’mouse click away’ (e.g. Google books, JSTOR or the Bodleian Digital Library). This socio-technical development presents new epistemological challenges for research within various humanities disciplines. To aid this effort, some researchers are turning to new kinds of (digital) data-mining methods to tackle this complexity. The subject of this study, topic modeling (TM) is such a digital humanities method. The presentation systematically surveys academic applications of topic modelling – an algorithm that parameterizes word concurrences – within historical research. The aim is to answer questions such as; what are the stated benefits of TM, whether there is qualitative differences between TM and traditional methods, and what new epistemological challenges TM creates for historical research? Our starting point is 2004 with the first peer-reviewed historical article and end point in 2013 with the publication of a special journal issue on applications of TM. Our preliminary results show that TM indeed affords new possibilities of innovative qualitative approaches in historical research. However, for all practical purposes TM is, as of yet, not a ‘black-boxed technology’ as many of its key variables still lack general agreed upon standards. This incorporation of TM within historical studies appears to be analogues to earlier developments in disciplines such as; human geography or psychology. These earlier introductions of quantitative tools and methodologies into previously qualitatively dominated disciplines ultimately changed the character of these disciplines. If this will occur within historical studies or humanities remains to be seen.