Process 3

In the case of the source of illumination of the moon cited earlier, for example, students must be made explicitly conscious of the fact that they see the extent of illumination growing steadily as the angular separation between the moon and sun increases, up to full illumination at a separation of 180°. This direct observation leads, in turn, to the inference that what we are seeing is reflected sunlight.

In working up to the concept of "oxygen" (without any prior mention of this term at all) with a group of elementary school teachers some years ago, I had them do an experiment in which they heated red, metallic copper in an open crucible and weighed the crucible periodically. What they saw happening, of course, was the copper turning black and the weight of crucible and contents steadily increasing. When I walked around the laboratory and asked what they had observed so far, many answered, "We observed oxygen combining with the copper." When I quizzically inquired whether that was what they had actually seen happening, their reaction was one of puzzlement. It took a sequence of Socratic questioning to lead them to state what they had actually seen and to discern the inference that something from the air must be joining the copper to make the increasing amount of black material in the crucible. It had to be brought out explicitly that this "something from the air" was the substance to which we would eventually give the name "oxygen." What they wanted to do was to use the technical jargon they had acquired previously without having formed an awareness of what justified it.

This episode illustrates the importance of exposing students to repeated opportunity to discriminate between observation and inference. One remedial encounter in one subject matter context is not nearly enough, but opportunities are available at almost every turn. Mendel's observations of nearly integral ratios of population members having different color and size characteristics must be separated from inference of the existence of discrete elements controlling inheritance. In the study of literature, analysis of the structure of a novel or poem must be distinguished from an interpretation of the work. In the study of history, primary historical data or information cited by the historian must be separated from the historian's interpretation of the data.

A powerful exercise once employed by some of my colleagues in history was to give the students a copy of the Code of Hammaurabi accompanied by the assignment: "Write a short paper addressing the following question: From this code of laws, what can you infer about how these people lived and what they held to be of value?" This exercise obviously combines exposure to both processes 1 and 3.