Martin (1992: 323):
For purposes of analysing scientific English it is useful to draw a distinction between expectancy relations and implicational ones. Whereas in most fields probability modalises the relation between activities (one is likely to follow another), in science an attempt is made to construct reality in such a way that one activity implies another (the relationship between the two is causally modulated).
 Expectancy relations are concerned with interpersonal assessments, whereas implicational relations are concerned with logical inference.
 This grammatical metaphor misconstrues probability as the Agent of modalisation instead of one of its types:
in most fields
modalises (‘makes modal’)
the relation between activities
In SFL theory, the relation between two activities, in the sense of two figures in a sequence — i.e. semantics, not "field" — is logical (expansion or projection), not interpersonal (modality).
 These are category errors. Implication is distinct from cause — A implies B ≠ A causes B — and cause (ideational meaning) is distinct from modulation (interpersonal meaning).