When Weber uses the notion of probability in his definitional statements--for example, in defining a relationship as existing "in so far as there is a probability that" a certain norm of behavior will be adhered to--he responds to similar considerations. Probability is here taken to mean that in all likelihood men involved in a certain context will orient their behavior in terms of normative expectations. But this is always probable and never certain because it can also be assumed that for some actors the chains of causality peculiar to their unique social relationships will lead to departure from the expected probability.

It is convenient to distinguish two directions in Weber's view of causality--historical and sociological. "Historical causality determines the unique circumstances that have given rise to an event. Sociological causality assumes the establishment of a regular relationship between two phenomena, which need not take the form 'A makes B inevitable,' but may take the form 'A is more or less favorable to B.' " The quest for historical causality asks the question: What are the causes of the Bolshevik revolution? The search for sociological causality involves questioning the economic, the demographic, or the specifically social causes of all revolutions or of particular ideal types of revolutions.

The quest for historical causes, Weber pointed out, was facilitated by what has been called mental experiments. When we learn that two shots fired in Berlin in 1848 started the revolution of 1848, we must ask whether the revolution would have taken place had these shots not been fired. If we conclude that it would have started in any case, we can rule out these shots as causes of the subsequent revolutionary development. When we ask whether the Battle of Marathon was a major causal event for the subsequent history of Hellenic civilization, we must perform the mental experiment of envisaging Greece dominated by the Persians. Such an experiment will convince us that had the Athenians lost the battle, a Persian Greece would have been a basically different society. We can then conclude as to the probability that the outcome of the Battle of Marathon, by guaranteeing the independence of the city-states, was indeed a major causal factor in the subsequent development of Greek civilization.

The assessment of the historical significance of an historical fact will begin with the posing of the following question: In the event of the exclusion of that fact from the complex of the factors which are taken into account as co-determinants, or in the event of its modification in a certain direction, could the course of events, in accordance with general empirical rules, have taken a direction in any way different in any features which would be decisive for our interest?
To determine sociological causality, Weber argues, also requires operating within a probabilistic framework. This type of generalization attempts to establish, for example, that the emergence of capitalism required a certain type of personality largely shaped by the preachments of Calvinist divines. The proof of the proposition comes when, either through mental experiment or through comparative study in other cultures, it is established that modern capitalism could probably not develop without such personalities; therefore, Calvinism must be considered a cause, though emphatically not the cause, of the rise of capitalism.

This example calls attention to the fact that Weber's methodological reflections served as a tool in his substantive investigations. Yet he was not concerned with methodology for its own sake and, like many another scientist, he did not always follow his own methodological guidelines. Contrary to his nominalistic stress on the acting person as the unit of analysis, he advanced a theory of stratification based largely on structural explanations rather than on a subjective theory of class distinctions.

When explaining the decline of the Roman Empire, he focused on structural changes in Roman agriculture. More importantly still, Weber's life-long preoccupation with the increase of rationality in the modern world was to a considerable extent based on structural considerations, as witness his stress on the separation of the household from the business enterprise as a harbinger of economic rationalization. In all these instances, Weber also provides illustrations pointing to changing motivations of historical actors, yet on balance, structure seems more important than motivation.

Though a number of other examples could be cited where Weber did not apply his methodological injunctions, many more instances in his work reveal that he put his methods to brilliant use in his substantive analysis.

From Coser, 1977:224-226.