II. Hume’s critique of causality

This leads us to the theory of the relativity of causality and chance: causal relativity.

Specifically, it is postulated that we can never know - completely, finally and unambiguously - whether or not the laws of nature are causal, statistical or some unknown mixture of the two.

a) Hume’s analysis of human intelligence is radically false

The Formality of Reality: Xavier Zubiri’s Critique of Hume’s Analysis of Causality
Photo provided by

Table 1. Mapping of Hume’s Analysis of Causality to Zubiri’s Critique

Therefore, there is actually no empirical evidence for the existence of some causal law L which acts the same way (in all particulars) at all times.

It may be objected that L may be inferred when each X is approximately similar, since each Y will likewise be approximately similar.

c) The locus of causality is formality, not content of impressions

What was Hume’s target? What notion or theory did he wish to demolish? Why was it important for him to do so? Causality as a notion had developed from the pre-Socratics through the Greeks, culminating in Aristotle’s analysis of it into four types. Further work was done by the Islamic philosophers, and later by the Medieval philosophers, though few substantive changes were made by them to the basic idea. The classical notion of causality and its principle characteristics, used by virtually all philosophers until Hume’s time, is summarized in Figure 1. Even Hume assumed that this notion of causality was monolithic, and sought not to refine or modify it, but to discard it altogether and replace it with a totally different notion.

d) Functionality is the correct replacement for or interpretation of causality
Photo provided by

Having approached Hume’s account of causality by this route, ..

Zubiri, with his accustomed thoroughness, has analyzed Hume’s philosophy in order to extract what important insights lie therein. Zubiri agrees with Hume and the other British empiricists that prior to development of any epistemology, it is first necessary to go one level deeper in order to fully analyze human intelligence. Such analysis is not a new theory, but is (or should be) only a pure and rigorous description of that intelligence. He also agrees that this analysis must be the basis of any theory of causality, and in particular, Zubiri agrees with Hume that "causes", in some metaphysical sense, are not given in experience. Finally, he agrees that there are significant problems with the "classical" notion of causality, which is in need of a complete overhaul. Nonetheless, he believes that Hume’s analysis of causality is fatally flawed in respect to all three aspects of it given above. The thrust of Zubiri’s critique centers upon five major areas:


The implication of this is quite clear: insofar as Hume’s theory of the intelligence requires causality to be deterministic, it is erroneous, as is his theory of causality.

Hume’s analysis of causality ..

Zubiri observes that that on which we base our knowledge, whether at the level of primordial apprehension or at the higher levels, is not constant conjunction. What was the idea behind causality in classical philosophy? It was to express a particular type of relationship between two things (or events, or processes). This relationship has the characteristics discussed above in section III. This may or may not have been adequate in the time of Aristotle, and up through the medieval period. But we now know that things can be related in many more ways than can be adequately described by the deterministic paradigm of classical causality. To describe this situation, Zubiri borrows a term (and idea) from mathematics: that of . Functionality, sometimes describable only in mathematical language, is a much more general way of describing relationships among things. These relationships may be among more than two things, and may involve statistical ideas. Functional relations may or may not involve causality in the traditional sense, or Hume’s version, constant conjunction; functionality is a much broader concept, capable of supporting inferences such as counterfactual conditionals which are beyond the range of constant conjunction. Furthermore, functional relationships may be—and indeed often are—statistically based, for which constant conjunction as an explanation is hopelessly inadequate. Functional relations exist for all three levels of intelligence, beginning with primordial apprehension.