The productivity of an element may correlate with the types of class membership it has. In English, for example, where a great many morphemes occur in both 'N' and 'V' classes ('a book', 'to book', 'a take', 'to take'), or are members of one and occur in sequences equalling the other ('lionize NNυ = V', 'preachment VVn = N'), the 'Nυ' and 'Vn' classes (including the zero members of each as in 'to book', 'a take') are productive. That is, given a member of 'N' which has no yet been recorded in 'V' position, or which has just come to be used in the language, there is a good chance of its occurring in 'V' position either with an accompanying 'Nυ' or without it (i.e. with the zero member of 'Nυ'). This may happen, though more rarely, even with the members of 'N' which have a paired but different member in 'V': e.g. 'to shoot', 'a shot', but also 'a young shoot'.
- Harris (1951), a pag.375, n.19

A question of some interest is that of productivity of elements: i.e. given an extremely large sample, with elements or classes 'A', 'B', 'C', etc., occurring with various members 'X₁', 'X₂', etc. of class 'X', which elements out of 'A', 'B', 'C' have a high probability of occurring with any new member 'Xո' of 'X', and which elements out of 'A', 'B', 'C' etc., do not? Those elements which have a high probability of occurring with any new 'Xո' are called productive in respect to 'X'.
- Harris (1951), a pag.374-375