When studying interthinking, we are in the odd, but interesting, position of having to use the process we are trying to understand in order to understand it. That is, as in the scientific study of any complex phenomenon, researchers have to gather observations of it, find ways of describing it in words (using metaphors, narratives and other modelling tools) and then, by a process of rational debate, agree upon the best explanation of what they have found. What is more, researchers have to use their own knowledge as language users and interthinkers to interpret the data they gather. So what methods do we have at our disposal? Earlier chapters have shown how several well-established types of research-conversation analysis, ethnography, computer-based text analysis, experimental action research and so on-have already provided valuable insights into how we use language to think together. Different methods can make their own distinctive contribution to understanding, and in cross-disciplinary research there is no reason why researchers should not combine them.13 But we do need to use methods which do justice to conversation as an interactive, continuing process of making meaning. Thinking together with language is a problematic process, and the constructing of common knowledge involves regular monitoring and repair by those involved. Through recording and analysing language as it is used in the accomplishment of everyday activities, we can begin to understand how people give continuity to their shared understanding, and to explain how and why they succeed or fail. My favourite metaphorical image for this kind of analysis is watching a school of dolphins from a moving ship (not least because this has the advantage of associating a very laborious, desk-bound process with a much more relaxed and leisurely one). If we notice that a particular dolphin has a white mark or other distinctive feature, each time an animal with that feature appears we assume it is the same one, travelling between sightings under the surface of the sea. In this way, the occasional sightings of individual animals can tell us about the continuous, co-ordinated activities of the school as a whole. For discourse analysts, key words, language patterns and topics appear and reappear in continuous stretches of language like surfacing dolphins. Some conversational dolphins may regularly appear together, suggesting that their relationship is significant. Others, even some apparently prominent members of the school, seem to get lost along
the way. Analysts can use these observations to build models of the process of interthinking, and so begin to explain how people develop shared, coherent lines of thought and follow them through to achieve practical outcomes-or, just as importantly, how they fail to do so.