In current presence or availability systems, the method of presenting a user's state often supposes an instantaneous notion of that state - for example, a visualization is rendered or an inference is made about the potential actions that might be consistent with a user's state. Drawing on observational research on the use of existing communication technology, we argue (as have others in the past) that determination of availability is often a joint process, and often one that takes the form of a negotiation (whether implicit or explicit). We briefly describe our current research on applying machine learning to infer degrees of conversational engagement from observed conversational behavior. Such inferences can be applied to facilitate the implicit negotiation of conversational engagement - in effect, helping users to weave together the act of contact with the act of determining availability.