Welcome to NUSAP net Glossary|
[ A |
[ T |
| [ 1 ]|
||The term 'normal science' was coined by T.S.Kuhn (1962), and Funtowicz and Ravetz (1990) referred to it when explaining their post-normal science'. In their words: "By 'normality' we mean two things. One is the picture of research science as 'normally' consisting of puzzle
solving within an unquestioned and unquestionable 'paradigm', in the theory of T.S. Kuhn (Kuhn 1962). Another is the assumption that the
policy environment is still 'normal', in that such routine puzzle solving by experts provides an adequate knowledge base for policy decisions. Of course researchers and experts must do routine work on
small-scale problems; the question is how the framework is set, by whom, and with whose awareness of the process. In 'normality', either science or policy, the process is managed largely implicitly, and is
accepted unwittingly by all who wish to join in."|
||Numerical error arises from approximations in numerical solution, rounding of numbers and
numerical precision (number of digits) of the represented numbers.
Complex models include a large number of linkages and feedbacks which enhances the chance
that unnoticed numerical artifacts influence the model behaviour to a significant extent.
The systematic search for artifacts in model behaviour which are caused by numerical error, requires
a mathematical 'tour de force' for which no standard recipe can be given. It will depend on the model
at hand how one should set up the analysis. |
To secure against error due to rounding of numbers, one can test the sensitivity of the results to the
number of digits accounted for in floating-point operations in model calculations. A pitfall here is pseudo precision.
||Acronym for Numeral Unit Spread Assessment Pedigree|
Notational system developed by Silvio Funtowicz and Jerry Ravetz to better manage and communicate
uncertainty in science for policy.
|[ 1 ]|