hibiting the greatest variation according to the country's degree of economic development.
In a provocative article, Hatt suggests that it may be unrealistic to think of the occupational structure as a single continuum and to ask people to rank a wide range of occupations. By analyzing the ranking patterns in a nationwide study of occupational prestige he found that some occupations fell into one scale and others into another. He concluded that there are broad occupational families (situs) which contain similar occupations. Thus it is feasible to compare occupations within the military situs, the political or industrial situs, but not occupations which belong in different situs. More research needs to be done in order to test the usefulness of this scheme.
In the remainder of the chapter short articles are presented which provide scales of occupational prestige. All of these scales have been used frequently to measure both occupational prestige and occupational mobility. Probably the scale used most often is that devised by Alba Edwards and used by the United States Census. Its categories are found in Miller's selection. The other scales are presented here because they are frequently used by social scientists, their utility depending upon the research task at hand. Smith's scale has an advantage over the others because it has more occupations that are generally well known; Warner's classification is perhaps most useful in measuring general "social class position" in small communities. The North-Hatt scale has the advantage of being methodologically the most precise and based upon a national sample. Hatt's situs scale has not been sufficiently refined for general research purposes. The reader should become well acquainted with these scales because they are referred to in subsequent chapters dealing with occupational mobility and related phenomena.
Paul K. Hatt
The purpose of this paper is to present a theory and to suggest a method of occupational classification usable in the study of social stratification.2 This necessarily entails the description of some currently unmet problems and an evaluation of existing classifications. To apply a classification to a totally different problem from the one for which it