Richard Schmidt's Noticing Hypothesis (1990, refined 1995, 2001) makes a single sharp claim: input does not become intake without conscious noticing. The hypothesis directly contradicts Krashen's position that acquisition is entirely subconscious, and it has become one of the most empirically productive ideas in SLA.
The hypothesis grew out of Schmidt's own diary study of his Portuguese acquisition in Brazil. Tracking what he produced against what he remembered noticing in input, he found a striking correlation: forms he had consciously registered in conversation were the ones that began appearing in his speech. Forms he had been exposed to but not noticed remained absent from production. Schmidt & Frota's (1986) "Notice the Gap" paper formalized this observation into a testable claim.
| Strong form | Weak form |
|---|---|
| Noticing is necessary for acquisition | Noticing is facilitative for acquisition |
| Without conscious registration, no input becomes intake | Acquisition can proceed without noticing, but more slowly and less reliably |
Most current research supports the weak form. Both versions agree that selective cognitive processing of input is acquisition-relevant; the disagreement is whether subjective awareness is strictly required.
Schmidt (2001) carefully distinguished two levels of awareness, and the hypothesis applies to the lower level:
Acquisition requires noticing, not understanding. Learners begin acquiring forms long before they can articulate the governing rule.
The hypothesis is narrower than it is usually reported to be. What Schmidt (1990, 2001) claimed must happen is the conscious registration of a form in the input. That is a different act from "noticing the gap", where a learner compares their own interlanguage output against the target and spots the mismatch. Both ideas trace back to the same Schmidt & Frota (1986) diary study, but they are separate constructs, and Schmidt held them apart. Folding the two together (a common slip in teacher-training accounts) inflates the hypothesis into a claim about output and self-correction that it never made. Gap-noticing belongs with the Output Hypothesis; the Noticing Hypothesis proper is about what gets in through the input in the first place.
The Noticing Hypothesis dismantles Krashen's non-interface position. If conscious noticing is required for input to become intake, then conscious processing is constitutive of acquisition, not merely a separate "learning" system that runs alongside it. This opens the door to a weak interface position: explicit instruction and form-focused activities work because they direct attention and increase the probability of noticing.
Schmidt softened the hypothesis over the years, and critics argue the later versions quietly concede the point. By 2001 he had narrowed noticing to the registration of "elements of the surface structure of utterances", and aligned his position with usage-based and emergentist accounts of learning rather than rule-based ones. But those accounts (N. Ellis 2005 puts it most plainly) hold that the bulk of acquisition is implicit, statistical, and unconscious, with conscious attention needed only to flag a handful of "fragile" forms the unconscious system tends to skip. Once noticing shrinks to that, the original strong claim that there is no intake without conscious noticing has little work left to do. Geoff Jordan, drawing this line of criticism together, concludes the necessity claim is simply false: conscious attention to form can help with some features but is not a precondition for learning a second language.
Schmidt's 2001 retreat granted part of this. He accepted that some learning happens without awareness while holding that more learning, of more kinds of feature, happens with it. The live question is no longer whether noticing ever matters, but whether it is necessary, as the original banner claimed, or just one helpful route among several.
The hypothesis provides the rationale for a wide family of techniques: