I need to know a bit more about what you mean to give a proper answer. You can think of standard propositional logic as Boolean logic: the propositions are the fundamental variables taking the values true or false, and the sentential connectives are the logic gates.
Collingwood’s “revolution” involved making the fundamental variables bigger than propositions: the true and false values are assigned to complexes consisting of presuppositions, questions, and answers rather than simply to propositions.
Breaking propositions up into components involves a move from propositional logic to predicate logic. Thus e.g. the sentence “snow is white” is a single unit in propositional logic, which simply takes the value true or false. In predicate logic, we can explain the truth of the sentence in terms of functions and variables that are the sub-sentential components. “Snow” is the value of a variable x in the function |white(x)|, which yields truth if the value of x names something white and yields falsity if the value of x names something non-white. “Something is white” involves a higher-order function into which the function |white(x)| is placed, yielding truth if there is some value of x for which |white(x)| yields truth and yielding falsity if there is no value of x for which |white(x)| yields truth.
I’m sure there is a parallel in computer science for predicate logic, but I don’t know enough about computer science to say for sure. Is this what you were thinking of?