TABLE OF CONTENTS The simplest validation experiment for this theory works like this, say for a system made from two components in series:
step 1. Each component is tested in isolation on each of its subdomains. This yields for each a measured vector of properties, one value for each subdomain.
step 2. The system property values are calculated for each subdomain of the first component using equation (6.2) and the measurements of step 1. This requires executing the first component on test data to measure the Tij matrix of equation (6.1).
step 3. Measurements are made on the actual system formed from the components, for each subdomain of the first component,
step 4. The vectors obtained in steps 2 and 3 are compared. The theory is supported to the extent that they agree. [11]
The purpose of a fundamental theory is primarily understanding. It seeks to describe, with just enough detail to capture a complicated phenomenon, what is really taking place. What does it mean to validate such a theory? Or put another way, as Karl Popper would insist to satisfy his definition of science [Popper (1959)], how could a fundamental theory be falsified, shown to be invalid?
The answer lies in the necessary simplifications and assumptions made in any theory, simplifications that are essential to its explanatory power. For our theory, to the extent that a component program does not behave as a step function on the subdomains chosen,...
