Causation as Explanation

Citation: Michael Scriven (1975) Causation as Explanation. Nous 9(1): 3-16 (RSS)
Internet Archive Scholar (search for fulltext): Causation as Explanation
Download: http://www.jstor.org/stable/2214338
Tagged:

Summary

Like Salmon (1997), Scriven believes that causation is indispensable in contemporary science, especially in fields such as psychology, education, social sciences, biosciences and medicine, where the analysis of causal relationships are crucial aspects of the study. He also believes that causation as a respectable scientific method of explanation has not received the full attention that it deserves. This is due to many scientific philosophers stating that any reference to causation is a sign of an immature science. These same scholars will also state that science should seek to describe rather than explain, and also that value-free science is the ideal. These contradicting statements exemplify the love-hate relationship that philosophers have with explanation, causation and evaluation.

Scriven promotes an approach that analyzes scientific language without using linguistic terms from other epistemological exercises, essentially an “ordinary language approach”. This reduces any linguistic baggage that might be associated with, misunderstood, or ignored by the adoption of that term into an unfamiliar field. This can be explicitly seen even in the use of the term "causation" itself. Causation has been used in scientific arguments by scientists unfamiliar with the philosophical background of the term (see re linguistic terms above), therefore it is a strike against the analysis to have unreliable authorities using terms they are not fully familiar with.

Causation approaches that can deal with concepts without making errors are preferable to other forms of explanation which cannot. It is better to have a more vague less precise explanation that is valid rather than an explanation that is more precise but error-ridden.

Scriven promotes a hybridized version of causation and the DN model of scientific explanation. While he believs that it is reasonable to propose that causes are not necessary conditions for their effects - causes themselves do not constitute the sufficient conditions to explain the effect. Therefore you must be able to at least assume or propose a general law/premise that provides the sufficient conditions in combination with the cause. This echoes the requirement of the DN model of a natural law or theory in which to draw expectations. The example that Scriven uses is hyperbolic in its absurdity: All people who are born will die, therefore birth is the "cause" of death. It is clear, however, that other factors "cause" death, and that will birth is not the cause of everyone's death, it is both a necessary and sufficient condition for it. Salmon's (1997) argument for relevance and counterfactual arguments is particularly salient for this point.

Scriven, then, defines cause as an explanatory factor (of a particular kind), and causation is the relation between explanatory factors (of this kind) and what they explain (1975:11). He also eloquently states, that the explanation is whatever would allow a qualified reader/listener to fully understand the fact/phenomenon.

References Cited Salmon, Wesley C. "Causality and explanation: A reply to two critiques." Philosophy of Science (1997): 461-477. Scriven, M. (1975). "Causation as Explanation." Nous 9(1): 3-16.