The method Ksenofontov used was to ask shamans and former shamans to describe their own election, in their own words, and to write down what they said. The result, published in 1930 as Legends and Stories about Shamans among the Yakuts, Buryats and Tungus, is a body of first-person narrative that almost nothing else in the Siberian literature can match.
The accounts follow a shape that has since become famous. The candidate falls ill and cannot recover; the illness is understood as a call that cannot be refused; in vision the body is taken apart by spirits, the flesh stripped, the bones counted and sorted, and an extra bone or a missing one determines what kind of specialist the person will be; the body is reassembled and the candidate recovers, now able to work. He also collected the counterpart material on the shaman's tree, the animal double, and the geography of the upper and lower worlds.
The year before he had published a study of what he called the cult of madness in Uralo-Altaic shamanism, arguing against reading shamanic illness as pathology and for reading it as an institution with its own logic. His material was translated into German in the 1950s and reached western readers that way, and a great deal of what general books say about shamanic initiation traces back through that translation to notebooks taken in Yakutia in the 1920s.