Twentieth-century psychology's infatuation with numbers would not have blossomed into an all-consuming love affair had it not been for the usefulness of statistical constructions based on group data. Statistical constructions represent work done on some raw material that already has a numerical form. However, that form involves more than the counting of heads. The individuals who are counted must also be endowed with count-able attributes. Where the attributes are physical, like height or weight, their transformation into numerical form depends on social practices developed long ago, but where one has to deal with psychological attributes the corresponding measuring practices represent a relatively recent development. This development was initially independent of the special use of statistical information that has occupied our attention so far. It is therefore necessary to step backward in time to consider the nature of the practices that made it possible to impose a numerical form on psychological at-tributes.

What is the difference between one and two?

Because it is rare for human experience and activity to take on a quasi-numerical form spontaneously, psychologists who insist on working with such forms must impose them on their subject matter. One way of doing this is to apply the numerical net post hoc, after the psychological event has taken place and has produced some record. The number of word associations with a particular thematic content can be counted, for instance, either for a particular individual or across individuals. This kind of quantification was attempted by Galton and practiced on a large scale by G. Stanley Hall.