Saman is an Evenki word. It reached Russian through accounts of Siberia in the seventeenth century, entered German and French in the eighteenth, and by the twentieth had been applied to healers, mediums, diviners and ecstatic specialists on every inhabited continent.
The generalisation was not accidental. Mircea Eliade's Shamanism: Archaic Techniques of Ecstasy, published in French in 1951, presented shamanism as a single archaic technique underlying religion everywhere, and the book's influence is the main reason the word means what it now means in English. Ronald Hutton and Andrei Znamenski have both traced how the picture was assembled, from a small number of early reports, read as though they described a universal type, by scholars who had not been to Siberia.
The cost of the generalisation is specific. It puts one people's word onto other peoples' offices, and it flattens offices that are structurally different. An Evenki shaman, an Inuit angakkuq, a K'iche' Maya ajq'ij, a Korean mudang, a Dine hataali and a Sakha oyuun are not variants of one thing. They differ in how the office is entered, what it is for, whether trance is involved at all, and what the practitioner is understood to be doing. Several of them do not travel to any other world, which is the feature the general definition is built on.
This atlas uses shaman for the Tungusic and Siberian context, and each tradition's own word elsewhere, wherever one is documented. Where a word is not documented, the entry says what is known and stops.
The practice sold in Europe and North America as shamanism is a further step removed. Its main modern source is Michael Harner's core shamanism, set out in 1980 and explicitly designed as a synthesis with the cultural specifics removed. Whatever it is, it is not the office this entry describes.