A paper by Norris presents scientific observation as a highly inferential activity in which human senses have become secondary to instrumentation. Norris takes examples from elementary and secondary curricula to support his position. This paper examines Norris's contentions from two perspectives, that of cognitive psychology and that of the philosophy of science. It is argued here that Norris has discussed only one aspect of observation, theory-confirming observation. Another, pretheoretic or theory-building observation, is more closely related to the curriculum activities he criticized. From a cognitive psychology perspective the theory-confirming observation is of use to the expert, but for novices the absence of domain-specific knowledge and of sophisticated problem-solving processes may render theory-building observation more meaningful for instruction.