Indeed, this continuum of complexity leading into the extreme single agent case applies for MAS in general (see Figure 2). With communicating agents, systems can get arbitrarily complex and arbitrarily centralized until a single agent has all the control. Of course communication bandwidth may be prohibitively low to reach the extreme in a given domain.
However, even if it is possible to use the most complex system, it is not always wise to do so. In general, system builders should use the simplest possible multiagent systems according to Figure 2. This system designer's version of Occam's razor makes the system both easier to design and easier to use. Nonetheless, if needed, communication offers immense flexibility in MAS.
The fully general multiagent scenario appears in Figure 12. In this scenario, we allow the agents to be heterogeneous to any degree from homogeneity to full heterogeneity. The key addition is the ability for agents to transmit information directly to each other. From a practical point of view, the communication might be broadcast or posted on a ``blackboard'' for all to interpret, or it might be targeted point-to-point from an agent to another specific agent.
Figure 12: The general communicating MAS scenario. Agents can be
heterogeneous to any degree. Information can be transmitted directly
among agents as indicated by the arrows between agents. Communication
can either be broadcast or transmitted point-to-point.