© 2019 Estonian Literary Museum, Estonian National Museum, University of Tartu ISSN 1736-6518 (print), ISSN 2228-0987 (online) 29 Journal of Ethnology and Folkloristics 13 (1): 29–49 DOI: 10.2478/jef-2019-0003 FOLK-TALE NETWORKS: A STATISTICAL APPROACH TO COMBINATIONS OF TALE TYPES* JULIEN D ’HUY PhD reseacher African Worlds Institute (IMAF) Centre Malher 9, rue Malher 75004 Paris, France e-mail: dhuy.julien@yahoo.fr ABSTRACT This paper is an atempt to study combinations of tale types using a networks approach and calculating the centrality index of each type (degree, betweenness, eigenvector centrality). The network of tale types seems to take the form of a ‘small world’ with a few types serving as bridges between highly connected sets of tale types. The centrality of each type also seems to depend more on its age than on how widespread it is. KEYWORDS : tales of magic • tale types • combinations • network • centrality INTRODUCTION A tale type is a particular narrative schema in which episodes and narrative motifs are organised in a sufciently stable way. It is above all a tool for classifying and study- ing oral transmission stories, i.e. in the Aarne–Thompson–Uther (ATU) classifcation. The tale type system is very useful for thinking about folk tales in general because it allows structure where diversity would otherwise only be apparent. Yet folk tales are rarely pure and rarely belong to a clearly determined type. The folklorist François- Marie Luzel (1887: 418) said: I have reproduced these [i.e. the stories] exactly as I have collected them, in order to give an idea of the way certain storytellers, believing to increase the interest of their tales, modify and mix them, to please the listeners. The more a tale is long and flled with marvels and tests, and the more it has of success, usually, to the public of the winter evenings. The longer the tale and the more it is flled with marvels and tests, the more successful it usually is at the audience of winter evenings. * The author thanks the reviewers for their constructive comments which greatly contributed to improving this paper, as well as the editorial team of the Journal of Ethnology and Folkloristics for their invaluable work during the editing process.