Introduction - Extreme Events: A Physical Reconstruction and Risk Assessment - page 13

discovering, developing and refining those aspects of reality that are amenable
to mathematical analyses (Ziman, 1978). It is powerful science and has been at
theforefront of human intellectual achievement for many decades, but it does
not traditionally take a historical view of nature. The geological approach to
examining nature on the other hand is one of 'taking the world as it is' which
can be thought of as a type of realism (Baker, 1994). The focus in geology is in
deriving hypotheses from nature rather than applying elegant theories. Baker
suggests that the geological approach involves synthetic reasoning according
to classical doctrines of commonsensism, fallibilism and realism. The geology
approach does not seek to provide answers to puzzles about the sequence of
past and possibly future events by imposing limiting assumptions upon the
real world. Too often engineers impose limitations upon their investigation of
thenature of natural hazards because they do not feel comfortable adopting a
geological and/or historical approach to the problem. By using prehistoric data
as a test of the initial assumptions made in natural hazard risk assessments,
we adopt a more scientific approach to the problem. In other words, we seek
to discover nature through research and not halt our attempts at discovery by
adopting limiting assumptions. Prehistoric data is a powerful ally in the quest
to reduce risk from natural hazards as it helps to test these initial assumptions.
Technically though, the prehistoric data does not always record natural hazards
because these events only become a hazard when they affect humans. If they
occur in isolation from humans they can be better referred to as extreme natural
events. Extreme events, as registered in prehistoric records therefore, are real as
they have occurred in a temporal and spatial sequence that is discoverable.
Concluding comments
The dilemma presented here exists because different professionals, usu-
ally engineers and natural scientists (often geo-scientists), see the nature of haz-
ards from different perspectives. The engineering approach is to develop a model
of this behaviour which is seen to remain constant over time. The natural sci-
entist's view is that the behaviour of hazards may change over time and no one
particular slice of time is necessarily reflective of the nature of that hazard.
Engineers typically regard the evidence from natural long-term records as too
imprecise to be meaningful. So they are often not interested in incorporating
such information into their models. However, the natural scientist's view is that
while this evidence may be less precise than the engineer's data, it is nonetheless
more than sufficiently accurate to test the assumptions upon which the engi-
neer's model rests. This scenario and its typical consequences can be explained
by a simple story.