Using the technique of implicational scaling DeCamp [DeCamp, D. 1971. “Towards a generative analysis of a post-creole continuum”, in Hymes, D. 1971. Pidginization and Creolization of Languages. Cambridge: Cambridge University Press.] showed that there were constraints on the occurrence of basilectal, acrolectal and mesolectal forms […] From a more technical perspective we can say that the working assumption behind implicational scaling as used by DeCamp, Bickerton and other creolists is that the scale consists of a series of ranked isolects (ie systems of individual speakers), or lects. Each (iso)lect is the output of a grammar (which is part of a polylectal grammar), is invariant, and differs from the one immediately next to it with respect to a single feature or rule in the panlectal grammar.
- Romaine (1988), Pag. 163

[…] the technique of implicational scaling is appropriate only for variables which can be unidimensionally or linearly ordered. If that were not so, then implicational relations would break down.
- Romaine (1988), Pag. 179