Another approach to the logical discontinuity between answer domains is through the study of paradoxes. For Solomon Marcus: "Paradoxes occur when two different levels of knowledge, of language, of communication, of reality, of human behaviour, etc. are seen as one level, are mixed, are superposed, are combined, or are confused." (149) He gives 18 pairs of levels which demonstrate a variety of paradoxes of which some are well-known to specialists.

To clarify the semiotic difficulties involved, Marcus groups them into four types:

(a) Semiparadox: A against B, but not necessarily B against A (e.g. "Mary gave birth to a child and got married") (b) Paradox: A against B, and B against A. If "against" is a logical negation, for example, this results in a logical paradox (e.g. when something is simultaneously good and bad) (c) Semiantinomy: A against B, and A for B, where "for" is a binary relation which is inverse with respect to the relation "against" (e.g. the well-known claim of Epimenides the Cretan that "All Cretans are always liars") (d) Antinomy: where (A, B) and (B, A) are both semiantiomies, such that the first term of a dichotomy both opposes and needs the second term, with the terms attracting and rejecting each other.

Marcus and Tataram have applied these distinctions in the analysis of 60 interacting global trends noted by the Goals, Processes and Indicators of Development Project (150). They argue:

"When dealing with the contemporary world, a basic step is to learn how to progress from a descriptive to an evaluative analysis, from what is directly perceived to what is scientifically understood, although such an understanding may sometimes surprise the intuitive perception....Many such trends are organized in opposite pairs, but their contradictory nature is much more richer and perfidious than what these binary oppositions reflect."

Of special interest with respect to the later arguments of this paper is the manner in which they show that antinomic relations emerge when semiparadoxical and paradoxical pairs of trends are associated as cyclic sequences of trends. Although they do not explore this feature explicitly in their analysis of an influence-trends matrix, it highlights an essential feature of the dynamic associated with paradox. It is especially valuable, given the implications for comprehension, that they have related this cyclic process to development trends.

The difficulty with any such approach is that the very logic of the method employed disguises the full force of the paradox and of the hiatus it engenders in any univocal communication. It effectively prevents the insertion of the engendering elements into the same framework, unless they are denatured and converted into symbolic entities, as in the case of the Marcus initiative.