The interface debate asks one deceptively simple question: can explicit knowledge, the conscious rules a learner can state, ever turn into implicit knowledge, the unconscious system that drives fluent, spontaneous use? How a theory answers fixes what it believes teaching can achieve, which is why this argument sits underneath almost every dispute about grammar instruction. Three positions divide the field.
Krashen's Monitor Model draws a hard wall between two separate systems. Acquisition is subconscious, fed by comprehensible input much as a child acquires a first language; learning is the conscious study of rules. On the non-interface view the two never meet: learned knowledge cannot become acquired knowledge. Its only job is to run the Monitor, an editor that checks and repairs output after the acquired system has produced it, and only when the speaker has time, is focused on form, and knows the rule. It polishes performance but never builds competence, so explicit grammar teaching is largely beside the point.
DeKeyser, drawing on the cognitive psychology of skill learning, argues the reverse. Through the right kind of practice, explicit knowledge can become automatic. Skill-acquisition theory traces a path from declarative knowledge (knowing that a rule holds), through proceduralisation into procedural knowledge (knowing how to apply it in performance), to automatisation, where use becomes fast, accurate, and almost effortless. The practice has to fit: effects are largely skill-specific, and proceduralisation works best when practice resembles real use. This is the engine room of PPP. Present the rule, practise it, produce it freely is a sequence that only makes sense if the strong interface is real.
Ellis rejects both extremes. Explicit knowledge does not convert directly into implicit knowledge, but it is far from idle: it helps indirectly by feeding the processes that do build the implicit system. Knowing a feature makes a learner likelier to notice it in the input and to register the gap between what they say and the target form, which raises intake and drives restructuring. Ellis adds a developmental brake: instruction can help a feature become implicit only when the learner is ready for it. Explicit teaching, on this account, primes attention and processing rather than pouring rules straight into fluent use.
Each position licenses a different classroom. Non-interface makes form-focused instruction nearly futile and hands everything to input. Strong interface licenses explicit teaching plus production practice, and with it PPP and the whole present-then-drill tradition. Weak interface keeps instruction in play but pushes it toward awareness-raising and meaning-embedded focus on form rather than decontextualised rule-and-drill. Evaluating any method that teaches grammar comes down, in the end, to which interface assumption it quietly rests on.