early stages of development, the superstrate will be the EL, not the ML, it is predicted that most of these lexemes will be content morphemes. However, superstrate morphemes may turn up as system morphemes in the following way: content morphemes may be borrowed and then used with new meanings, as system morphemes.
Further discussion of pidgin and creole development is beyond the scope of this study; but these possibilities suggest that researchers should consider the role of CS.
When there is extensive borrowing (beyond the incorporation of cultural B forms into a language), some explanation beyond citing 'bilingualism' or 'contact' is needed. In communities exhibiting CS, specifically where there is a motivation for CS as the unmarked choice, an explanation suggests itself. Such switching includes ML + EL constituents in most cases, and these constituents consist of morphosyntactic frames from the ML, filled with content morphemes from either the ML or the EL.
Scenario I. Content morphemes are borrowed into the L1, which is maintained as the main language of the community. As long as the maintained language is the ML during CS, CS offers no easy means for EL morphosyntax to affect the ML, but it does provide a mechanism for EL content morphemes to enter the ML.
Scenario II. This outcome takes incorporation of EL content morphemes into the ML a step further. If many EL content morphemes are used in CS over a period of time, relexification from the EL occurs. Again, largely content morphemes are involved.
Scenario III. This outcome may arise through CS in a community with a good deal of intergroup communication. CS may become a type of lingua franca in such communities, with the ML for this CS as the first language of the most socio-politically dominant of the groups (in terms either of numbers or of socio-political prestige). System morphemes and phrase-structure rules from the ML may be gradually borrowed into the speakers' Lls (the ELs).
Scenario IV. This outcome rests on a more radical CS mechanism. Socio-political conditions are such that speakers 'turn over' the ML in their CS patterns: what was the EL becomes the ML. The result is that they will now use the morphosyntax from the 'new' ML in their ML+EL constituents in CS. This 'turnover' sets the stage for borrowing the new