This article addresses the relation of the 'unknown' to the 'known' at
two periods of European thought. At the time of Columbus the unknown
was linked to the known through the notion of 'discovery' within a giv
en canon of knowledge: note is taken of certain intellectual implicati
ons. The article then moves to the present-day with its greater possib
ility of freeing the unknown from necessary linkage to the known: here
attention is given to the notion of 'invention' contra discovery and
of referentiality contra canonicity. Thence the positions of Columbus
and Evans-Pritchard are compared and aspects of the praxis of anthropo
logy put under brief critical review, respecting the paradox - implici
t throughout the article - of knowing the unknown.