Application interoperability and data exchange are desirable goals, but conventional system design practices make these goals difficult to achieve, since they create heterogeneous, incompatible conceptual structures. This conceptual incompatibility increases system development, maintenance and integration workloads unacceptably. Conceptual data independence (CDI) is proposed as a way of overcoming these problems. Under CDI, data is stored and exchanged in a form which is invariant with respect to conceptual structures; data corresponding to multiple schemas can co-exist within the same application without loss of integrity. The use of CDI to create domain-independent applications could reduce development and maintenance workloads and has potential implications for data exchange. Datasets can be merged without effort if stored in a conceptually-independent manner, provided that each implements common concepts. A suitable set of shared basic-level archetypal categories is outlined which can be implemented in domain-independent applications, avoiding the need for agreement about, and implementation of, complex ontologies.