Beginning with birth, typically developing children use strategies to communicate, and the functions of their language change with maturation and interaction. Since communication cannot exist if both parties do not participate, it is important to not only study the ability of the child, but also the behavior of the adult. Numerous studies have examined the behavior of the parent or other domestic adult, but few have included the study of teacher behaviors.

This study investigated teachers' response to four types of nonverbal communication attempts made by infants. The gestures included: deictic, affect signaling, object-related, and conventional. The type of response was also documented as facilitating dialogue or discouraging dialogue. Thirty infant teachers and/or teacher assistants from 11 centers in Northeast Tennessee were videotaped for 30 minutes. Videos were coded to determine which of the defined functions teachers were most likely to respond to and if the response facilitated or discouraged dialogue. The results showed that overall, teachers responded to 25% of nonverbal attempts; 75% of the infants' nonverbal attempts teacher offered a non-facilitative response or missed the gesture. Pearson correlations determined that there were relationships between children's attempt to communicate and teachers' response in all four types of nonverbal communication, including deictic, r (30) = .659, p .001; affect signaling, r (30) = .917, p = .001; object-related, r (30) =.848, p = .001; and conventional, r (30) = .794, p = .001. There were several relationships between the number of nonverbal attempts by children and teachers offering a facilitative responses including affect signaling, r (30) = .776, p = .001; object-related, r (30) = .635, p .001; and conventional, r (30) = .514, p = .004, but not with deictic attempts. There were relationships between the number of nonverbal attempts by children and teachers offering discouraging responses in all 4 types of nonverbal language, including deictic, r (30) = .706, p .001; affect signaling, r (30) = .630, p = .001; object-related, r (30) = .582, p =.001; and conventional, r (30) = .439, p = .015.