The Inference of Common Cause Nat- uralized Aviezer Tucker Inferences of common causes from their effects are ubiquitous in science and everyday life: Evolutionary biologists infer the existence of extinct species from similarities between their descendants, and their properties from those of fossils. Detectives and historians infer descriptions of past events from testimonies. Textual critics infer properties of lost texts from present exemplars. Geneticists and comparative linguists infer the migra- tion routs of our pre-historic ancestors from the present geographic distri- butions of genes and languages. Teachers infer plagiarism from similarities between exams. Many philosophers (Reichenbach (1956); Sober (1988); Cle- land (2002); Tucker (2004)) agree that much of our knowledge of the past is founded on inferences of common causes from their present effects. The philosophical literature about the inference of common causes origi- nated in Reichenbach’s attempt (Reichenbach (1956)) to deduce a principle of common cause from the second law of thermodynamics. I argue first that this attempt failed. Most philosophers did not pursue Reichenbach’s deductive project. Instead, they attempted to explicate principles of infer- ence of common cause in the logical positivist prescriptive and unempirical senses of the term. It is difficult to debate philosophical explications be- cause they can fall on their normative aspect when confronted with empirical counter-examples. Rather than refute, empirical evidence can demonstrate the irrelevance of these explications, as the history and sociology of science demonstrated the irrelevance of much of the logical positivist explication of science, spurring the introduction of new philosophies of science. I attempt to do so here. I show that explications of Reichenbach’s common cause principle confused the inference that some common cause existed without specifying its properties with the inference of a concrete common cause with a unique set of properties. They also did not distinguish the inference of common cause types from the inference of common cause tokens. I clear these confusions and argue that no exclusively a-priori probabilistic silver bullet exists for the description of the inference of common cause. Addi- tional assumptions and theories about the transmission of information in