This fundamental duality manifests in the different vocabularies used to express what the worker does and what gets done. These two vocabu laries are based on a difference in perception. W hat the w orker does is the vocabulary of behavior, of worker action. W hat gets done is the vocabulary of results and reflects technology. In no way can we understand human behavior at work by using the vocabulary of results. This was the major problem with the 1940 edition of The D ictionary o f occu pational titles; its job definitions were almost entirely in the language of results. Thus when counseling workers according to their behavioral experience and pre sumed potential for placement, there were no satisfactory behavioral indi cators to validate a match. An intensive study of the kinds of inferences being made from the descriptive language of results led to the recognition of the duality.