More generally, in an explanatory use of argument, we try to make sense of something by deriving it (sometimes deductively) from premises that are themselves well established. With an inference to the best explanation, we reason in the opposite direction: Instead of deriving an observation from its explanation, we derive the explanation from the observation. That a hypoth- esis provides the best explanation of something whose truth is already known provides evidence for the truth of that hypothesis.
Once we grasp the notion of an inference to the best explanation, we can see this pattern of reasoning everywhere. If you see your friend kick the wall, you infer that he must be angry, because there is no other explanation of why he would kick the wall. Then if he turns away when you say, “Hello,” you might think that he is angry at you, if you cannot imagine any other reason why he would not respond. Similarly, when your car goes dead right after a checkup, you may conclude that it is out of fuel, if that is the best explanation of why your car stopped. Psychologists infer that people care what others think about them, even when they deny it, because that explains why people behave differently in front of others than when they are alone. Linguists argue that the original Indo-European language arose mil- lennia ago in an area that was not next to the sea but did have lakes and rivers, because that is the best explanation of why Indo-European languages have no common word for seas but do share a common root “nav-” that connotes boats or ships. Astronomers believe that our Universe began with a Big Bang, because that hypothesis best explains the background microwave radiation and spreading of galaxies. All of these arguments and many more are basi- cally inferences to the best explanation.
Solutions to murder mysteries almost always have the form of an infer- ence to the best explanation. The facts of the case are laid out and then the clever detective argues that, given these facts, only one person could possi- bly have committed the crime. In the story “Silver Blaze,” Sherlock Holmes concludes that the trainer must have been the dastardly fellow who stole Silver Blaze, the horse favored to win the Wessex Cup, which was to be run the following day. Holmes’s reasoning, as usual, was very complex, but the key part of his argument was that the dog kept in the stable did not bark loudly when someone came and took away the horse.
I had grasped the significance of the silence of the dog, for one true inference invariably suggests others. [I knew that] a dog was kept in the stables, and yet, though someone had been in and fetched out a horse, he had not barked enough to arouse the two lads in the loft. Obviously the midnight visitor was someone whom the dog knew well.2
Together with other facts, this was enough to identify the trainer, Straker, as the person who stole Silver Blaze. In this case, it is the fact that something didn’t occur that provides the basis for an inference to the best explanation.