Germans, and the ignorant Slavs, and the honest Chinese, and the untrustworthy Japanese, and so on and so on. All these are generalizations drawn from samples, but the samples are selected by a method that statistically is wholly unsound. Thus the employer will judge labor by the most troublesome employee or the most docile that he knows, and many a radical group has imagined that it was a fair sample of the working class. How many women's views on the "servant question" are little more than the reflection of their own treatment of their servants? The tendency of the casual mind is to pick out or stumble upon a sample which supports or defies its prejudices, and then to make it the representative of a whole class.
A great deal of confusion arises when people decline to classify themselves as we have classified them. Prophecy would be so much easier if only they would stay where we put them. But, as a matter of fact, a phrase like the working class will cover only some of the truth for a part of the time. When you take all the people, below a certain level of income, and call them the working class, you cannot help assuming that the people so classified will behave in accordance with your stereotype. Just who those people are you are not quite certain. Factory hands and mine workers fit in more or less, but farm hands, small farmers, peddlers, little shop keepers, clerks, servants, soldiers, policemen, firemen slip out of the net. The tendency, when you are appealing to the "working class," is to fix your attention on two or three million more or less confirmed
