Over several years the authors of this article have had intensive discussions to find common ground in the topic they both specialize in: dance. Egil Bakka is a Norwegian ethnochoreologist, specializing in Nordic traditional dance/folk dance; Gediminas Karoblis is a Lithuanian philosopher, specializing in phenomenology and ballroom dancing. Our starting point was a philosophical question about the notion of dance knowledge and a shared worry about the empirical basis for many academic works on dance that other colleagues have also pointed to (Hoerburger 1959; Lange 1983; Adshead-Lansdale 1994; Grau 1998; Farnell 1999; Fügedi 2003). Bakka then brought up the widespread reservation against the use of film/video for the documentation and analysis of dance.1 We continued with a wish to clarify to ourselves the epistemological basis for research in dance, and somewhere along the way we started writing this article. We experienced that a dialogue in which methodological issues in dance research were confronted with philosophical scrutiny brought about a number of interesting perspectives. We hope that our exercise may be of interest to a broader audience. The aim is to explore how our different disciplinary points of departure-philosophy and ethnochoreology-can be brought to interact in creating a deeper understanding of our topic, rather than comparing the disciplines or discussing their differences.
Our writing started out with Bakka innocently sketching a description of fieldwork in Numedal, Norway, which was meant to be an illustrative introduction. It was left as a rough draft. When we returned to revise it, we realized that it illustrated a number of epistemological problems that we had dealt with in the meantime; therefore, we leave it here in its "innocent" form and will return to it in the discussions towards the end of the article.
Ethnochoreological fieldwork in Numedal, November 1985
Egil Bakka's description, based on fieldnotes from 1985, was as follows:
The musicians on the stage are playing the third dance this evening, a reinlender,2 and the dance floor of the little countryside community house in eastern Norway is filled with dancers, mostly in their sixties and seventies. They are back in the arena where they picked up the art of dancing as teenagers. It was a necessity for them to master social dancing to be part of the social life in the small community, and no teaching of this repertoire was offered. What they picked up has served them well through life, and it still comes in handy at weddings and at dance parties for their age group. The dancers typically know six or seven such dances, called by local names, including vals, springar, hamburg, rei(n)lender, masurka, fokstrot, and tango.3
Numedal is, according to Bakka, one of those places where the fairly square, conventional structure of the reinlender seems to have been broken up and changed into a free flow of elements. The conventional structure has two bars of promenade (four beats-i.e., two step periods of two beats each) and two bars of turning (four beats); these two elements are repeated again and again in strict accordance with the musical periods.4 The free flow of elements can consist of step patterns covering three beats, which immediately breaks up the regularity. The promenade then becomes longer than two bars and often then becomes quite freely improvised, as is also the case for the length of the turning (Bakka et al. 1987:5). In our dance event, most of the dancers do this freely improvised version, but one couple in their nineties do the dance with the conventional square structure (16 mm film, "Nore og Uvdal," 1985, Rff archives). Two couples choose to change to totally different techniques of dancing for a while. One technique has a separate name, steger.5 A "young" couple in their early fifties starts doing some swing or rock-'n'-roll motives as a variation (filming logs 1985, Rff archives) (figure 1). …