This paper applies principles derived from ethnocriticism to help explain differential outcomes with different methods used to analyze the work of Directory Assistance telephone operators in a large US telecommunications company. The work of Directory Assistance operators provides a subtle case of computer-supported cooperative work. Collaborative work between operator and customer is supported and shaped by digitized-voice and database technologies. Our work also involved the introduction of additional voice-recognition technologies to this human-to-human collaboration. In a previous paper, we used methods from participatory design to show that knowledge work is a major component of the operators‘ conversations with customers. By contrast, other research using formal cognitive task analyses had described operators‘ work as routine and as involving no active problem solving. How had evidence that we had found so compelling been invisible to other analysts? I analyze the concept of ’’invisible work‘‘ as an attribute not of the work, but rather of the perspectives from which that work appeared to be invisible. Ethnocritical heuristics help us to contrast the analytical methods and their outcomes.