Electronic discussion groups provide a convenient forum for individuals to share their experiences of chronic illness. The language use of individual participants, and the way their language shifts over time, may provide implicit indications of important shifts in sense-of-self. This paper relates experience with application of the Hyperspace Analogue to Language (HAL) model for automatic construction of a dimensional model from a corpus of text. HAL is applied to 17 months of discussion on a closed list of 20 women coping with chronic illness. The discussion group was moderated for a focus the phenomenon of ?Transition? - how people can learn to incorporate the consequences of illness into their lives. The current phase of research focuses on identification of clusters of words that can represent key aspects of Transition. The HAL models for two participants have been analyzed by experts in Transition to form candidate clusters. These clusters are then used as a basis for contrasting the language usage of an individual participant over time as compared to the entire corpus. We have not yet found a reliable basis for identifying transitions in an individual based on their entries into a discussion forum, although the clusters may have some inherent value for introspection on individual experiences and Transition in general. We report challenges for interpretation of the HAL model related to the correlation of dimensions and the impact of group dynamics.