The Great Train Robbery has been used time and time again as a reference to blatant criminal acts that have escaped punishment, and should by now have lost its attraction. Unfortunately this does not seem to be the case, at least as regards the context of this paper. Reference to the Great Train Robbery still constitutes a temptation to those of us who like calling a spade a spade, while not really enjoying sticking our necks out alltogether too far. Be that as it may, this paper attempts to clarify the manner in which theories and knowledge in educational research have functioned to either universalise specific knowledge or have pulled off one of the most audacious coups in modern history. The purpose of this paper is to present arguments for both of these alternatives and to discuss the relative merits of both, as a means of understanding the philosophical underpinnings of educational research. This discussion is framed within the context of a study of a specific knowledge base, which has been produced over the years 1939-1995. That which is known as the Malmö Longitudinal Study began in 1938, when a schoolteacher, Siver Hallgren, began research for his graduate degree in Psychology at the University of Lund in Sweden. Hallgren's work and a selection of the seminal texts in the Malmö Longitudinal Study knowledge base have been analysed using discourse analysis and have been mapped, as Social Cartography mappings. Further analysis has been conducted, in order to clarify i) how the concept of knowledge has been conceptualised in the texts themselves and to ii) ascertain the meaning of these conceptualisations, seen from a historical and philosophical perspective. The results of the analysis are discussed in relation to the works of Ernst Cassirer and Michael Lynch, philosophers whose works are deemed to be both similar and dissimilar. Both the similarities and dissimilarities in their "ways of seeing" knowledge and science are considered to be of interest, especially as regards the all-important role they allot concepts and concept formation. While Cassirer proposes that precision is the aim of concepts, Lynch proposes the use of fluid concepts. While, on the one hand, Cassirer views science as a cumulative process, the purpose of which is the Truth, Lynch asserts that, in a plural world, it is necessary for space to be made available for both Truth and several truths. Both, however, agree as to the importance of concepts for knowledge and science. Focusing on the conceptualisations of knowledge in this body of educational research, the paper discusses whether or not its "way of seeing" can be considered in much the same manner as many have viewed the Great Train Robbery: "It isn't right, but you've got to give them credit for getting away with it!" Or, if one can view this "way of seeing" knowledge and science as the universalisation of specific knowledge, making it serious business, with far-reaching ethical consequences.