From: "M. A. Baker" Newsgroups: rec.folk-dancing Subject: Computer generated choreography Date: Fri, 29 Jan 1999 22:45:32 -0500 Message-ID: <36B2805C.52C0@tiac.net> mcl2@vms.cis.pitt.edu wrote: > Is this the first time a computer generated contra has been successfully > called at a real dance? My undergraduate thesis (MIT, 1976) was a computer program that generated random MWSD sequences. Each sequence started from a squared set, contained a series of moves, and resolved to an allemande left, or directly to a right and left grand. The program attempted to take into account proper use of hands, body flow, and avoid the use of overflow. In addition, the probabilities were adjusted to reduce getting into long series of, e.g. head men 1/4 right, side women u turn back, side men 1/4 left, heads partner trade, heads partner trade. While I was not a caller at the time, I did subject many of me fellow dancers through various walkthrus of the (never ending) output of this program. A s I remember, a few sequences were called at actual dances. All of the output was legal, most flowed, but some of the combinations were unusual enough that dancers found the material too hard. The issues of flow, timing, and what dancers will find diffucult are very different between contra and MW squares. A lot more could be said here. There are several computer programs which assist MWSD callers in writing new choreography. Most of these are checker pushing programs which know what calls are legal from each position, but have no knowledge of proper flow or use of hands. Some will show you the dancers as they move through the calls. Several of these programs can be asked to pick the next call at random from a selection of legal calls. One of them will use the text to speech feature of the Mac and "talk" random sequences. Clark Baker, Belmont, MA (using my wife's e-mail) cmb@nexen.com