Article Preview
Top1. Introduction
We could speak about three different paradims in the historical evolution of science (see, for example, [Nieto de Alba, 1998] or [Morin, 1995]). Each paradigm is characterized by a way to think and to focus the analysis of the reality. These three paradigms, nevertheless, are not substitutive, but complementary. This evolution has affected to Economy like all other sciences. In this work we analize the fundamental principles that sublie in each one of the mentioned paradigms, as well as the reasons for its sprouting and the implications for the Economy, as much at theoretical level as at practical level, making special reference to the area of the Analysis of Economic Time Series
1.1. Characterization of the Evolution of the Conception of the Reality: of the Determinist Paradigm to the Statistical Paradigm
Thus, at first, within the scientific development derived from the Newtonian mechanics, scientists worked under a Deterministic Paradigm. The Principle of Strong Causality governed this paradigm. This Principle maintained that the same consequences follow inexorably from the same causes (this is the Laplace’s demon). However, Laplace also remarked the importance of the probability. The probability arises as something necessary when scientists take conscience of the impossibility of complete knowledge of interacting causes as the number of implied variables increases. For this reason, the paradimg evolves to the Statistical Paradigm. The Principle of Weak Causality substitutes the Principle of Strong Causality. This new Principle states approximately the same consequences follows from approximately the same causes. In the new paradigm, universal determinist laws are not enunciated. Statistical laws substitute deterministic laws, stating that, in average, the behavior of the analyzed variables could be explained by means of a universal law. The prediction capacity is mainteined, but now in probabilistic terms, using statistical inference. The Deterministic and Statistical Paradigms coexists, each of them applied to different fields. The first one for the systems (simple) with few degrees of freedom and the second one for the systems (complex) with many degrees of freedom. At this time, complexity was conceived as merely quantitative, due to the sum of a high number of degrees of freedom.