Empirical Inference of Numerical Information into Causal Strategy Models by Means of Artificial Intelligence - Distributed Artificial Intelligence, Agent Technology, and Collaborative Applications - page 293

However, if we abandon the restriction to correlation as a concept for the proof of causality and a
measure of association, it seems to be possible to infer further causal knowledge from empirical data
and therefore improve the quality of the model base significantly. Hence, this chapter proposes an ap-
proach to automatically prove managerial cause-and-effect relations and to approximate the unknown
causal function underlying these associations.
Causality Concepts
Since one main objective of this chapter is to provide an approach to proof hypothetically assumed
cause-and-effect relations between managerial variables, it is necessary to define necessary and sufficient
conditions for the concept of causality in this environment. As it has been outlined in the introduction,
the mere correlation of two variables seems to be insufficient for this purpose. Moreover, causality per
se cannot be observed or tested by objective means. According to Kant it is a synthetic judgment a priori
(Schnell et al., 1999, p. 56). Causality must therefore be regarded as an assumption about the connection
between cause and effect made by the human mind and based on a variety of experiences rather than
some kind of natural phenomenon which can be observed in an objective manner.
As a consequence of the lack of observability of causal relations as outlined above, there has been a
broad scientific dispute within the philosophy of science about the concept of causality. The beginning
of this discussion can be traced back to the Humean regularity theory (Sondhauss, 1998): This notion
of causality is founded mainly on the two concepts of contiguity and temporal succession (Hume 1748),
i.e. two events always have to occur within temporal and/or spacial limits and the effect must follow
the cause in time. Although this theory does not seem to be fully sufficient to explain the concept of
causality 1 , it introduces an important property which is part of almost every causality theory: Causal
relations are usually regarded as asymmetric associations of two events.
However, there are (at least) two objections regarding this basic theory of causation:
Firstly, representatives of probabilistic causality theories bring forward the argument that the deter-
ministic association between cause and effect as an integral part of the regulatory theory represents a too
rigorous characteristic. According to their notion, the occurrence of a cause only raises the probability
of an effect but does not necessarily imply it.
Secondly, the Humean regulatory theory is criticized because of temporal precedence being the
only characteristic to establish the necessary asymmetric property of a causal relation as for example
Brady (2002) explains:
The Humean theory does even less well with the asymmetrical feature of the causal relationship because
it provides no way to determine asymmetry except temporal precedence. [...] Causes not only typically
precede their effects, but they also can be used to explain effects or to manipulate effects while effects
cannot be used to explain causes or to manipulate them. (p. 18)
Approaches to overcome these shortcomings lead to the notion of causality as it is proposed by in-
terventionistic theories. The central ideas of these theories are driven by human action as the definition
of Gasking (1955) shows:
The notion of causation is essentially connected with our manipulative techniques for producing results.
(p. 483)