66 PETER ROBERTS

 One of the consistent observations made by colonial writers is that the
 slaves danced for hours, a fact that encouraged commentators to classify
 these events as 'plays' with various scenes. Another observation was that
 the singing of the slaves was mostly ex tempore. Taken together, these
 observations suggest that it is better not to regard each 'scene' as an
 independent dance, but instead as a variable part of a whole, and survivals
 should therefore not be seen as separate (parts of) dances. Features or motifs
 of courtship dances, commemorative (battle) dances, and spirit possession
 dances survived with different adaptations and in different combinations in
 various locales according to the exigencies of each situation. Distinct names
 given by commentators to 'units' served to separate and reify them, but this
 separation may be more a feature of academic analysis than a representation
 of the conceptions of the performers themselves.
 Today, there is little or no folk knowledge of the dance that Labat described
and called the calendar; it is the image of the stickfight calenda that is almost
exclusively known today in the islands where it is part of their cultural history.
This seems unfortunate, for it means that there is general ignorance about
the links between the calenda and the more famous dances known worldwide,
such as the rhumba and the fandango. The fact that there is little historical
knowledge of the calenda also means that major features of the Caribbean
heritage in dance remain obscure. The same can be said for the way in which
each community of slaves in the Caribbean selected from their ancestral
culture and put together their own 'plays.'
 What seems remarkable in this tracing of the life of the calenda is that a
word used in so many different countries by so many different people
preserved an appearance of uniformity in spite of a web of connections and
variants as well as substantial changes over time. I submit that this was
because Labat's initial written version of the word was adopted in a canonical
way and repeated unquestioningly. This was a case of the written word being
used by writers whose knowledge was limited to create a non-representative
reality. Furthermore, the idea of cultural uniformity was not accidental; it
was part of a philosophy about presumably 'uncivilised' people. Note the
words of Moreau de Saint M6ry in 'Danse:'

 The dance among civilised people is subject, like almost all other aspects of
 their culture, to the caprices of fashion, while simple or savage people preserve
 a dance in an almost invariable form. A greater sum of ideas offering more
 combinations, variety in all areas can hardly but be the attribute of a more
 perfected people; and perhaps the dances of diverse peoples may serve, on a
 graduated scale, to identify their degree of civilisation. (1789: 172-3)

 Contrary to what Moreau de Saint M6ry claims about uniformity, the
survival of different motifs in different sequences and structures is what
characterises the dances in the various places described by writers. It is
this unpredictability of creolisation that has always confounded cultural and
linguistic analyses.