Starting with the seminal work of Klima & Bellugi (1966) and Bellugi (1967), young English-speaking children have been observed to pass through a stage at which their negative utterances differ from those of adults. Children initially use not or no, whereas adults use negative auxiliary verbs (don't, can't, etc.). To explain the observed mismatches between child and adult language, the present study adopts Zeijlstra's (2004, 2007, 2008a, b) Negative Concord Parameter, which divides languages according to whether they interpret negation directly in the semantics with an adverb, or license it in the syntactic component, in which case the negative marker is a head and the language is a negative concord language. Our proposal is that children first hypothesize that negation is expressed with an adverb, in keeping with the more economical parameter value. Because English is exceptional in having both an adverb and a head form of negation, children must also add a negative head (i.e. n't) to their grammar. This takes considerable time as the positive input that triggers syntactic negation and negative concord is absent in the input for standard English, and children must find alternative evidence. The Negative Concord Parameter accounts for an intricate longitudinal pattern of development in child English, as non-adult structures are eliminated and a new range of structures are licensed by the grammar.