By recognizing science as a set of local practices if becomes possible
to 'decentre' it and develop a framework within which all knowledge s
ystems can be equitably compared. It is argued that all knowledge trad
itions are spatial in that they link people, sites and skills. In orde
r to ensure the continued existence of the diversity of knowledge trad
itions rather than have them absorbed into the great imperialist archi
ve we need to enable disparate knowledge traditions to work together t
hrough the creation of a third space in which the social organization
of trust can be negotiated. (C) 1997 Elsevier Science Ltd.