This paper elucidates that for rigorous analysis of many engineering struct
ures, three seemingly unrelated concepts, namely probability, fuzziness and
anti-optimization, ought to be utilized. This is in order to take into acc
ount the uncertainty in the actual values of the basic variables or on the
models employed to evaluate the structure-expected behaviour. In this case,
it is possible and reasonable to use the three paradigms in a combined man
ner. Some examples are given to illustrate these ideas. (C) 2001 Elsevier S
cience Ltd. All rights reserved.