In input processing models of learning, input constitutes the data provided to the machine for processing. For language learners, it is the spoken or written language samples which are available to the learner as evidence about how the language works.
Three factors are important:
Input is often equated with positive evidence, that is, well-formed sentences which learners hear in the environment, as opposed to negative evidence, which may include corrective feedback or explicit instruction. While most theories of second language acquisition emphasise the essential role of input in acquisition, for some, like generative SLA and emergentist models (Ellis), input is a necessary and sufficient condition: exposure to the language is all that is required.
Such a radical position on the role of input in second language acquisition is defended by Krashen in his input hypothesis: for him, second language acquisition occurs through the processing of comprehensible input which is slightly beyond the learner's current competence (i + 1). In this view, instruction is unnecessary.
The only contribution that classroom instruction can make is to provide comprehensible input that might not otherwise be available outside the classroom.
The interaction hypothesis takes a more measured view, according a role in acquisition for interaction and learner output. Interactional analysis makes mention of modified input, or simplified language produced by a more proficient to less proficient speakers (mother to child: motherese, baby talk; native to nonnative speaker: foreigner talk):
Speech tends to be slower (and sometimes even louder); intonation is exaggerated; syntax tends to be simpler [?] ; lexical items tend to be simpler (often reflecting the more frequently used words in a language.
However, not all input can be comprehended and acquired at once. If input consists of the language sample available to the learner, then intake (Corder 1967) is that part of input which the learner attends to and thus learns from. Input processing theory is concerned with how learners convert input into intake. This model suggests that 'structured input,' which allows learners to attend to important features of the input without either explicit rule-learning or practice, is a necessary and sufficient condition for second language acquisition (VanPatten, 1994, 2002).
Read and reflect: Sequences of acquisition