A system for synchronizing multiple modalities is described. A user may use multiple modalities, such as voice and browser, to interact with data on a network, such as the World Wide Web. All of the modalities may be synchronized so that all are updated when the user enters information in just one. A method of communicating between devices includes receiving a request for first-modality data that includes first content, and sending a message in response to receiving the request, the message including information allowing the request of second-modality data that includes second content overlapping the first content. Another method includes requesting first data for a first modality, the first data including first content, and automatically requesting second data for a second modality, wherein the second data includes second content that overlaps the first content.