This paper has reviewed and discussed the state of progress of IS research using representation theory. We have argued that models of representation provide a mature theoretical basis for scholars researching conceptual modelling artifacts and activities. However, the state of progress in this particular area of research has mostly remained at the level of assessing various language capabilities and the argument in this paper is that the underlying representation theory provides a fruitful basis to transcend this level of investigation to study further phenomena of interest associated with conceptual modelling. In fact, it is time researchers moved outside the confines of traditional studies. Some prior research, such as the work by Gemino and Wand (2003, 2005), serve as stimulating examples for studies that use the richness of the theory to derive research hypotheses over and above the level of language capabilities. In our own research we aim to further transcend this research horizon and to progress the state of maturity and dissemination of representation theory and representational analysis. With this aim in mind, we have outlined an example of research that links propositions and findings from representation theory with other theories in our discipline, namely the issue of acceptance of IS artifacts.

In future work we aim to continue along two lines of investigation in particular. Firstly, we will continue our work on the acceptance of modelling languages. At present we are developing and testing a measurement scales inventory to conduct empirical studies on the adoption and acceptance of modelling languages in order to test the hypothesised relationships in our research model shown above. Secondly, in a related stream of research, we will continue our work on the effect of representational deficiencies on the quality of the models produced. As part of this work, we have evaluated and empirically confirmed representational deficiencies of a selected process modelling language and communicated our findings to the developers of that language in order to influence a revision of the language specification. After distribution of the revised process modelling language specification, we will, as a last step in the research, assess and compare the quality of business process models produced using the revised modelling language with those produced using the pre-revision version of the language.

As a concluding remark, we would like to add that we have found the research method of representational analysis very useful in understanding and exploring the challenges related to conceptual modelling and we expect this type of research to continue to give stimulating input to both academic and practical work in the area of conceptual modelling in the future.