In this book I argue that a reason for the limited success of various
studies under the general heading of cybernetics is failure to
appreciate the importance of con- nuity, in a simple metrical sense of
the term. It is with particular, but certainly not exclusive, reference
to the Arti cial Intelligence (AI) effort that the shortcomings of
established approaches are most easily seen. One reason for the relative
failure of attempts to analyse and model intelligence is the customary
assumption that the processing of continuous variables and the
manipulation of discrete concepts should be considered separately,
frequently with the assumption that continuous processing plays no part
in thought. There is much evidence to the contrary incl- ing the
observation that the remarkable ability of people and animals to learn
from experience nds similar expression in tasks of both discrete and
continuous nature and in tasks that require intimate mixing of the two.
Such tasks include everyday voluntary movement while preserving balance
and posture, with competitive games and athletics offering extreme
examples. Continuous measures enter into many tasks that are usually
presented as discrete. In tasks of pattern recognition, for example,
there is often a continuous measure of the similarity of an imposed
pattern to each of a set of paradigms, of which the most similar is
selected. The importance of continuity is also indicated by the fact
that adjectives and adverbs in everyday verbal communication have
comparative and superlative forms.