Tuesday, February 10, 2009

The Third Scientific Revolution

The first scientific revolution, forged and championed by such luminaries as Francis Bacon, Galileo Galilei and others. The first scientific revolution emphasised the gathering of and reliance upon empirical evidence. Observations would be quantified, measured, isolated, and repeated. The first scientific revolution bequeathed to the world the first scientific method: observe a phenomenon, identify independent and dependent variables, create an experiment, alter the independent variable(s) and observe any changes to the dependent variables, draw causal conclusions. Repeat and expand as necessary to form and confirm a theory.

The second scientific revolution, begun in the social and life sciences: statistical analysis. When a phenomenon becomes impossible to recreate in a controlled, laboratory environment, the investigator may instead choose to observe the phenomenon repeatedly in an outside environment. Data are collected, tabulated, and analysed as a statistical universe. Correlations are noted and causal conclusions are drawn. Studies are repeated and expanded to form and confirm theories.

The third scientific revolution, made possible by microcomputing: scientific modelling. A set of conditions regarding a phenomena are observed. Those data are used to formulate a set of initial conditions in an abstract, computational model. The model is allowed to operate, and consequent conditions of the model are derived. Those consequent conditions are checked against observed consequent conditions of the phenomenon being modeled. If the modeled consequential conditions closely match the observed consequential conditions, then the operations of the model serve as the basis for drawing causal conclusions and the formulation of theories.

1 comment:

Anonymous said...

The real problem, I'd think, would be with the creation of the model without knowing *why* the observed activity occurs in the first place. The underlying cause might be something beyond your data-gathering, causing you to wrongly attribute a presumed cause to an observed effect.

The benefit of the field work / experiment is that you can diddle with the actual world, instead of the artificial one. Sure, if your scientific model took into account *everything* (which is impossible, or close enough to it), you'd be right on the mark. I just don't have that much faith in the models.