An introduction to Second Language Acquisition; a survey of key theories, evidence, and debates in SLA research

Popular Beliefs About Language Learning

Before looking at the formal research, it is worth pausing on the beliefs that circulate informally among teachers, learners, and the wider public. Friends, family members, students, and parents often hold strong intuitions about how second languages are learned: that children learn languages effortlessly while adults struggle; that some people simply have “a gift” for languages and others do not; that exposure alone is enough, or conversely that explicit grammar teaching is essential; that mistakes are a sign of poor learning rather than a normal and necessary part of it. Second language acquisition (SLA) research exists, in part, to test these popular beliefs against systematic evidence – and, as the following sections show, many of them turn out to be oversimplified or simply wrong.

L1 versus L2 Acquisition

A foundational question in the field is how far second language (L2) acquisition resembles first language (L1) acquisition. Children acquiring their mother tongue do so universally and apparently effortlessly, reaching native competence within a few years regardless of intelligence or formal instruction. L2 learners, by contrast, vary enormously in their ultimate success, rarely achieve native-like competence in adulthood, and seem to depend on factors – motivation, aptitude, instruction, age of onset – that play little or no role in L1 acquisition. These contrasts raise the central question that has driven SLA research since the 1960s and 1970s: is L2 acquisition fundamentally the same kind of process as L1 acquisition, simply operating under less favourable conditions, or is it a qualitatively different process altogether, perhaps drawing on general problem-solving and memory systems rather than the specialised language faculty thought to underlie child language acquisition?

Early SLA Research: Language Transfer

The earliest systematic accounts of SLA grew out of the observation that learners’ errors often seem to echo the structure of their first language. A French learner of English who asks “Do you speak Italian?” with French-influenced word order, or who transfers French syntax into English sentences more broadly, appears to be drawing directly on L1 knowledge rather than building L2 knowledge from scratch. This observation became the basis for the Contrastive Analysis Hypothesis (CAH), most closely associated with Robert Lado’s Linguistics Across Cultures (1957). The CAH proposed that where the learner’s L1 and the target L2 are structurally similar, acquisition will be easy; where they differ, acquisition will be difficult. On this view, a systematic, point-by-point comparison of two languages should allow researchers to predict in advance which features of the L2 will cause difficulty, how serious that difficulty will be, what specific errors learners will produce, and how persistent those errors will prove to be.

The CAH had genuine appeal: it offered a principled, predictive framework that could be applied directly to syllabus and materials design. It is also not without empirical support – language transfer clearly does explain a good number of L2 errors, particularly in the early stages of learning, when learners have little L2 knowledge to fall back on and so default more heavily to L1 patterns. Over time, however, several serious problems with the hypothesis emerged. First, as learners’ proficiency develops, transfer-related errors tend to diminish, even though the structural distance between the L1 and L2 has not changed – something the CAH, in its strong form, cannot easily explain. Second, and more damagingly, the bulk of the errors L2 learners actually produce do not seem to result from L1 transfer at all. Strikingly similar errors turn up again and again among learners from very different language backgrounds, suggesting that something other than – or in addition to – cross-linguistic comparison is driving the acquisition process. These findings pushed researchers to look beyond contrastive analysis for an explanatory framework.

Early SLA Theories: Behaviourism

The CAH itself rested on a broader theoretical foundation: behaviourist learning theory, which dominated psychology and applied linguistics through the middle of the twentieth century. Behaviourist accounts of language development rested on three core tenets. First, language is learned through imitation, practice, and habit formation, much like any other behaviour. Second, reinforcement – praise for correct imitation, correction of errors – is essential for learning to occur; without feedback from the environment, habits cannot be properly established. Third, learning an L2 is held to be different in kind from acquiring the mother tongue, because the habits already established in the L1 actively interfere with the new habits required by the L2 – the theoretical basis for the transfer-based predictions of the CAH.

Each of these tenets came under sustained criticism as researchers began collecting and analysing real learner data in the 1970s. If language were simply a matter of imitation and habit, learners should produce only the forms they have heard, yet learner speech is full of creative, non-target forms that could not have been imitated from any input – utterances such as “How many years have you teached?”, “Yes, he cans”, or “He’s a lessoner”. These forms show learners over-applying regular grammatical patterns (the regular past tense ending, the third-person -s, productive derivational morphology) to irregular or non-existent cases. They are systematic errors, not random noise, and they could not have come from imitating native speakers, who do not produce them. This kind of error evidence was one of the strongest pieces of evidence against a purely imitation-and-habit account of language development.

A second strand of evidence against behaviourism came from observation of learners in the so-called silent period: many learners, particularly young children acquiring an L2 naturalistically, go through an extended phase in which they produce very little spoken language while continuing to process and apparently internalise the input around them. A purely behaviourist account, in which production and external reinforcement drive learning, struggles to explain how substantial linguistic progress can occur during a period with little or no productive practice taking place at all.

The third and perhaps most influential strand of counter-evidence came from morpheme studies, beginning with work by Heidi Dulay and Marina Burt in the mid-1970s and developed further by researchers such as John H. Schumann and Nathalie Bailey. These studies tracked learners’ accuracy in producing a specific set of English grammatical morphemes – the plural -s, the progressive -ing, the copula be, the auxiliary be, the articles the and a, irregular past tense forms, the third-person singular -s, and the possessive enclitic ‘s. The striking finding was that learners from markedly different L1 backgrounds tended to acquire these morphemes in a broadly similar order, regardless of the order in which the morphemes happened to be taught or the structure of the learners’ first languages. This “natural order” of acquisition was very difficult to square with a behaviourist, transfer-based account, since the CAH would predict that acquisition order should vary depending on the learner’s L1. The consistency of the order instead suggested that some internal, language-specific process – not simply habit formation shaped by input and L1 interference – was guiding the course of acquisition.

Creative Construction and Interlanguage

Taken together, the error data, the silent period, and the morpheme studies pointed towards a very different picture of L2 acquisition: not habit formation shaped by reinforcement, but a creative construction process. On this view, the learner actively analyses the language input they are exposed to and forms hypotheses about its underlying structure, much as a scientist forms and tests a hypothesis against evidence. These hypotheses are then tested against further input, and depending on the outcome, confirmed, rejected, or revised. The language that results from this ongoing process of hypothesis formation and revision – the learner’s own developing linguistic system, distinct from both the L1 and the target L2 – was termed the learner’s interlanguage, a concept introduced by Larry Selinker (1972) building on earlier related notions such as Pit Corder’s “transitional competence” and William Nemser’s “approximative system”. The interlanguage concept reframed learner errors not as failures or evidence of interference, but as natural, often systematic by-products of an active and creative learning process – evidence of a learner grammar in its own right, with its own internal logic, rather than a defective copy of the target language.

Innatism: A Response to Behaviourism

The accumulating evidence against behaviourist accounts of language development converged with a broader theoretical shift already underway in linguistics, prompted above all by Noam Chomsky’s critique of behaviourist accounts of L1 acquisition and his proposal of an innate Universal Grammar (UG) – a set of abstract principles common to all human languages, built into the human mind. Nativist, or innatist, approaches to SLA extended this reasoning to second language learning. The central argument is that the input learners are exposed to is too impoverished, too inconsistent, and too limited to account for the rich and abstract grammatical knowledge learners eventually display; this is the so-called “poverty of the stimulus” argument. Certain properties of language, on this view, simply could not be induced from input alone, and must instead be innate. Input is therefore reduced to the role of a trigger: it activates and shapes an internal language-processing capacity that the learner already possesses, rather than supplying, piece by piece, the knowledge the learner ends up with.

A number of UG-based theories of SLA were developed through the 1970s and 1980s exploring exactly how much of UG remains available to L2 learners, and whether L1 settings of UG parameters transfer into L2 acquisition. Of these various theoretical proposals, the one that has remained most influential – not least because it set out to explain SLA as a whole, rather than one narrow grammatical phenomenon – is Stephen Krashen’s theory of second language acquisition, developed across the late 1970s and through the 1980s. Krashen’s theory is influenced by the innatist, UG-oriented spirit of the period, while occupying something of a middle ground: rather than committing to a strong nativist position in which virtually everything is innate, Krashen’s framework represents a kind of compromise, sometimes referred to as the weak innate hypothesis, in which innate predispositions and exposure to input work together to drive acquisition.

Krashen’s Theory of Second Language Acquisition

Krashen’s model, developed initially with Tracy Terrell and set out across a series of publications in the late 1970s and 1980s – most comprehensively in Second Language Acquisition and Second Language Learning (1981) and The Input Hypothesis: Issues and Implications (1985) – is built around five interrelated hypotheses.

1. The Input Hypothesis. There is, on Krashen’s account, only one genuine route to language acquisition: exposure to comprehensible input. If input contains structures slightly beyond the learner’s current level of competence – input that Krashen formalises as i+1, one step ahead of the learner’s existing system – and that input is nonetheless comprehensible to the learner (typically through context, gesture, or simplification), the new structure will be acquired.
2. The Natural Order Hypothesis. Grammatical features of the target language are acquired in a predictable sequence. This sequence is essentially fixed, and – echoing the morpheme-study findings discussed above – is not altered by the order in which forms are presented in the classroom or by how much instructional emphasis they receive.
3. The Acquisition–Learning Hypothesis. Competence in an L2 can develop through two entirely distinct processes: subconscious acquisition, which resembles the way children acquire their L1, and conscious learning, which involves explicit knowledge of rules. Crucially, Krashen argues these two systems are entirely separate, and learning cannot convert into acquisition (a claim often referred to as the “non-interface” position). Only the acquired system, on this view, is available for the kind of spontaneous, fluent communication that characterises real-time language use.
4. The Monitor Hypothesis. The consciously learned system is not entirely without a role: it can function as a “monitor”, or mental editor, that checks and adjusts output from the acquired system. This monitoring role, however, is only available under three fairly restrictive conditions: the learner must know the relevant rule, must be focusing on linguistic form rather than on meaning, and must have sufficient time to apply the rule – conditions rarely all met simultaneously in spontaneous conversation.
5. The Affective Filter Hypothesis. Finally, Krashen proposes a metaphorical “affective filter”: a psychological barrier, raised by negative attitudes, anxiety, low motivation, or other unfavourable emotional states, that can prevent otherwise comprehensible input from being processed and acquired. A learner who is anxious or unmotivated may be exposed to plenty of suitable input without it ever “getting through” to the acquisition device.

Krashen’s framework proved enormously influential in language teaching, motivating comprehension-based and communicative approaches that prioritise rich, comprehensible exposure over explicit grammar instruction. It has also attracted substantial criticism – particularly the sharp acquisition/learning distinction, the vagueness of the i+1 construct, and the difficulty of testing several of the hypotheses empirically – but its core insight, that comprehensible input is necessary for acquisition, remains widely accepted even among researchers who reject other parts of the model.

The Value of Interaction

If comprehensible input drives acquisition, an obvious next question is how that input becomes comprehensible in the first place – and conversational interaction turns out to play a central role. Naturally occurring conversations between native speakers (NS) and non-native speakers (NNS) show a recognisable pattern: native speakers do not simply talk to learners as they would to other native speakers, and the resulting back-and-forth often does real interactional work even when it looks, on the surface, like small talk. A native speaker noting that a non-native speaker is from the Canary Islands and is met with a brief confirming response, or a native speaker checking on a non-native speaker’s plans for the day, are mundane exchanges, but they exemplify the kind of scaffolded, responsive talk through which input gets adjusted to a learner’s level in real time.

Michael Long’s influential modified input hypothesis, developed from the early 1980s onward, examined exactly these patterns of native-speaker speech modification in conversation with lower-level non-native speakers. Long identified several consistent forms of modification: linguistic simplification (shorter sentences, simpler vocabulary, less complex structures), a slower rate of delivery, comprehension checks (confirming that the listener has understood), and repetition or paraphrase of material that was not understood the first time. Critically, this modification is not purely one-directional. Non-native speakers also actively request modification of the input addressed to them – asking for repetition, clarification, or simplification when something is not understood – and in doing so help to negotiate input that is appropriately pitched to their own level. On the basis of these observations, Long argued that interactional modification promotes acquisition precisely by ensuring, jointly and dynamically, that the input learners receive is both comprehensible and appropriately calibrated to their developing competence – extending Krashen’s input hypothesis from a property of input in isolation to a property that emerges through interaction.

Input, Intake, and Noticing

A further refinement to input-based accounts of acquisition concerns what happens, cognitively, between exposure to input and the eventual acquisition of new language. Richard Schmidt’s noticing hypothesis holds that comprehensible input on its own is not sufficient for learning: learners must consciously attend to linguistic form within that input for acquisition to take place. Mere exposure, however extensive, will not by itself convert input into acquired knowledge; some degree of conscious noticing is required to turn input into what is sometimes called intake – the portion of input that the learner actually processes and integrates into their developing system.

Within this noticing framework, Richard Schmidt and Sylvia Frota (1986) distinguish a process they call noticing the gap: this occurs when a learner perceives a discrepancy between a target-language form they encounter in the input and the corresponding form in their own current interlanguage, prompting a potential revision of their internal grammar. Merrill Swain (1998), working within her broader output hypothesis, identifies a related but distinct phenomenon she calls noticing the hole: this occurs not while processing input, but while trying to produce language, when learners realise that they lack the linguistic resources needed to express what they want to say. Both processes converge on the same underlying mechanism – conscious attention to linguistic form, whether triggered by a mismatch noticed in input or by a gap noticed in one’s own output – and both can lead the learner to notice a discrepancy between their current interlanguage and the target language, a discrepancy that is taken, across much of this research tradition, to be a necessary condition for new language to be acquired.

Link to clickable infographic created by Claude 

This is great for revision and each card expands – A few things built into it )by Claude, not myself):

• Clickable accordion cards — each theory card expands to show a fuller explanation, key evidence, and sub-elements (Krashen’s five hypotheses each get their own tile, for instance)
• Colour-coded by theory family — coral for behaviourist/CAH, purple for interlanguage and noticing, teal for input/interaction, blue for Krashen
• Timeline spine with era labels (1950s–60s → 1970s → 1980s → 1990s) running down the left
• Keyboard accessible — all cards work with Enter/Space for screen reader users
• Warning banner on the CAH card flagging its empirical problems
• Responsive — adapts to narrower screens, including the Krashen hypothesis grid

 

References

Corder, S. P. (1967). The significance of learners’ errors. International Review of Applied Linguistics in Language Teaching, 5, 161–170.

Dulay, H., & Burt, M. (1974). Natural sequences in child second language acquisition. Language Learning, 24(1), 37–53.

Krashen, S. D. (1981). Second Language Acquisition and Second Language Learning. Oxford: Pergamon Press.

Krashen, S. D. (1985). The Input Hypothesis: Issues and Implications. London: Longman.

Lado, R. (1957). Linguistics Across Cultures: Applied Linguistics for Language Teachers. Ann Arbor: University of Michigan Press.

Long, M. H. (1983). Native speaker/non-native speaker conversation and the negotiation of comprehensible input. Applied Linguistics, 4(2), 126–141.

Schmidt, R. (1990). The role of consciousness in second language learning. Applied Linguistics, 11(2), 129–158.

Schmidt, R., & Frota, S. (1986). Developing basic conversational ability in a second language: A case study of an adult learner of Portuguese. In R. Day (Ed.), Talking to Learn: Conversation in Second Language Acquisition (pp. 237–326). Rowley, MA: Newbury House.

Selinker, L. (1972). Interlanguage. International Review of Applied Linguistics in Language Teaching, 10(1–4), 209–231.

Swain, M. (1998). Focus on form through conscious reflection. In C. Doughty & J. Williams (Eds.), Focus on Form in Classroom Second Language Acquisition (pp. 64–81). Cambridge: Cambridge University Press.