"Not that inferences are not important -- we rely in everyday life and in science as much on inferences as on reports -- in some areas of thought, for example, geology, paleontology, and nuclear physics, reports are the foundations, but inferences (and inferences upon inferences) are the main body of the science. An inference [...] is a statement about the unknown made on the basis of the known. We may infer from the material and cut of a woman's clothes her wealth or social position; we may infer from the character of the ruins the origin of the fire that destroyed the building; we may infer from a man's calloused hands the nature of his occupation; we may infer from a senator's vote on an armaments bill his attitude toward Russia; we may infer from the structure of land the path of a prehistoric glacier; we may infer from a halo on an unexposed photographic plate that it has been in the vicinity of radioactive materials; we may infer from the sound of an engine the condition of its connecting rods. Inferences may be carelessly or carefully made. They may be made on the basis of a broad background of previous experience with the subject matter, or no experience at all. For example, the inferences a good mechanic can make about the internal condition of a motor by listening to it are often startlingly accurate, while the inferences made by an amateur (if he tries to make any) may be entirely wrong. But the common characteristic of inferences is that they are statements about matters which are not directly known, statements made on the basis of what has been observed. (p.40-41)"