… universal lawfulness is consistent with individual exceptions …
This is from Causality and Modern Science: Third Revised Edition by Mario Bunge (1959; 1979):
… There are laws. And the principle of universal lawfulness, a stronger postulate, may be taken to read thus: Every single event is lawful, i.e. is determined in accordance with a set of objective laws — whether we know the laws or not. Or, again, Every single fact is the locus of a set of laws. Note that, in this wording, the principle of universal lawfulness does not assert that facts are determined by laws, but in accordance with laws, or simply lawfully. Thereby the idealistic doctrine is avoided, according to which natural and social laws are not the immanent form of facts, but prescribe them ab extrinseco. There is no Rule of Law, laws do not determine anything: they are the forms or patterns of determination — and this is one of the reasons why determinacy is not synonymous with lawfulness. Thus, for example, the constraints to which a dynamical system is subjected contribute to the determination of its motion; but Gauss and Hertz’s principle of least constraint does not determine the motion along the corresponding least-curvature path: it is just the form of the action of the constraints in motion.
The principle of lawfulness, however, does not require that every individual phenomenon should always occur in the same way whenever certain conditions are fulfilled; universal lawfulness is consistent with individual exceptions, with occurrences in a given low percentage of cases.
… the target of my arrows will not be the causal principle but only the claim that causation is the sole category of determination and that, as a consequence, the causal principle enjoys an unlimited validity. In fewer words: I will not argue against the notion of causation but against causalism.
… Causation (efficient and extrinsic) is only one among several categories of determination; there are other types of lawful production, other levels of interconnection, such as statistical, teleological, and dialectical determinacy.
In real processes, several categories of determination concur. Purity in types of determination (such as purity of causation) is as ideal as any other kind of purity.
The causation category, far from being external to other categories of determination, is connected with them. Thus multiple causation leads to statistical determinacy, the latter may in turn lead to quantitative self-determination, and reciprocal causation is interaction or interdependence.
The causal principle holds approximately in certain domains. The degree of approximation is satisfactory in connection with certain phenomena and very poor with regard to others.
It seems to me that Bunge repeatedly contradicts himself. Nevertheless I am interested in his approach. To be continued.