This paper critically analyzes problems and issues related to learning with open, technology-based environments. Theoretical assumptions that underlie learning in this type of environment are clarified in the first section. The second section addresses imitations in adaptive capabilities of technology, including: the learner relies upon feedback that is contingent upon learner actions; when learners maintain naive beliefs and fragmented understandings, the system must be able to present consequences and data that can be perceived as consistent or inconsistent; learners may not perceive the implications of such responsive data; and interactions may be distorted and misunderstood. The third section covers the importance of shared meaning as a necessary means toward increasing both partners' (i.e., the learner and the system) understanding each other's point of view at important points during the interaction. Sources of misunderstanding are then discussed, including: perceptual limitations of visual cues; biased and confounded meanings; failure to share meanings within system boundaries; and incongruent meta-level approaches. (Contains 44 references.) (DLS)