Unlike other animals who typically produce their signals as single units not combined with other signals (cf. Seyfarth & Cheney, 1997), humans produce words in combination with other words. The deaf children in our study are no exception-the children frequently combine their gestures with one another, and use those combinations to convey different meanings. For example, a deaf child combines a point at a toy grape with an “eat” gesture to comment on the fact that grapes can be eaten, and at another time combines the “eat” gesture with a point at the experimenter to invite her to lunch with the family. Moreover, and equally important, the deaf children’s gesture combinations function like the sentences of early child language in a number of respects. To be specific, the gesture combinations:

• convey the same types of meanings that children learning conventional languages convey in their early sentences;

• are characterized by predicate frames comparable to those underlying the sentences of early child language; and

• are characterized by surface regularities that mark who does what to whom and that are comparable to marking devices found in early child language.