Children’s early attempts at syntax have been described in the recent language development literature in terms of “pivot grammar.” The pivot-open class distinction is discussed in the light of the author’s more recent research that inquired into the semantic intentions that underlie early sentences. When utterances were examined along with context and behavior in the speech events in which they occurred, certain underlying conceptual relations could be identified. It is concluded that the “pivot grammar” account is only a superficial characterization of the form and distribution of linguistic elements in early two-word utterances. It is suggested that a more productive model of early language development to use for evaluation and treatment of language pathology would need to specify the semantic relations among objects and events that are coded by syntax.