Twitter, the microblogging service in which users share ideas in140-character increments (or 280-character increments after late September 2017), hosts approximately 500 million new tweets each day. While the field of social network analysis is quite advanced, only limited research has been conducted into discipline-specific networks (Espinos, 2017 and Espinos, 2015). Across the museum field, colleagues have used Twitter as a primary discovery source for professional networks and literature since the birth of the service, but more recently, we’ve begun witnessing conversational phenomenon emerging as an important factor for knowledge sharing. These conversations, known as "Twitter canoes" have been only lightly researched to date, but are prominent among the museum community. This paper will use two such canoes that occurred in 2017 between museums professionals to explore mechanisms for mapping tweets, methods for extrapolating meaning, and the possible future uses of Twitter canoes in museum work. The paper will also suggest potentially meaningful network metrics that might help identify and automate the detection of these conversations among twitter users. The authors anticipate that these methods could also be usefully extended to other social networks.