My approach to text analysis is to refer the text to an analytic apparatus which, in my case, produces a description of the referrent activity. This activity will, of course, operate at a number of different levels of analysis as well as in numerous texts at any given level. The fractal nature of my language of description entails that the same schema can be applied independently of the level of analysis (see, also, Dowling, 1999). The relationship between levels of analysis can be considered in terms of the strength of the institutionalisation of the lower level in terms of the higher.

Where the institutionalising level is a pedagogic activity, then the strength of institutionalisation will be strong and texts at the lower level of analysis will be incorporated into a formal or strongly classified pedagogic regime. High discursive saturation practices may be described as discourse, and low discursive saturation practices as competence-oriented skills.

Where the institutionalising level is an exchange activity, then the strength of institutionalisation will be weak and texts at the lower level of analysis will be incorporated into an informal or weakly classified pedagogic regime. High discursive saturation practices may be described as ideolect, and low discursive saturation practices as performance-oriented skills.