Oh so then perhaps you can link to these 1940-1960 statistics that somehow neatly and consistently segmented out computing roles into easy to define categories despite the fact that it was a new field and the lines between subfields were and always have been changing? Got a link handy?

Oh weird, apparently no such link