Fourth, both the computer and the user may have initiative. For example, a pop-up window can be presented to confirm a risky selection.

Therefore, the design of feedback, affordances, and cues is central to dialogue-based interaction.

Third, in dialogue, both the computer and the human participate in establishing a shared context. The computer does not simply receive a message; it also communicates the effects of that message.

Context, as we will see, can be understood through the concept of a state: The feasible communication acts and their effects are conditioned by the state of the partner.

In other words, in dialogue, acts of communication are conditional: The meaning of a turn depends on the communication context.

Siri's reply and the selection of "compose" would make no sense without the context provided by the preceding exchange.

Second, both dialogues in Figure 18.1 evolve in a turn-based manner. Each turn redefines the communication context.

For instance, the seven-stage model of interaction proposed by Norman [600] applies to all modalities of interaction; Section 18.1 explains this model in detail.

In Figure 18.1, two nominally different types of interaction are shown: speech-based and graphical. Both are forms of dialogue. This means that the concepts of dialogue are applicable across modalities.

This broad definition has several immediate and important consequences for HCI. First, dialogue, as a form of interaction, is not limited to speech and language even though this is often our first interpretation of the term "dialogue."