one-to-one correspondence of research categories with research categories is broken

[ernestguevarra]

this is a similar issue to the country - region correspondence although this is more complicated in that there are non-unique sub-categories that map to different categories.