Previous research has shown that the possible area of confusion between make and do is restricted to monotransitive use. Consideration of a random sample shows that if examples with complements of low lexical content are excluded, occurrences of make largely outnumber those of do. Consideration of the make examples leads to discussion of a number of issues including the dominance of mental over material process, the incidence of nominalized verbs as complements, and the reasons (semantic, syntactic or pragmatic) that might lead a speaker to choose make+complement rather than the cognate simple lexical verb (to make a promise vs. to promise).