Language acquisition

The transcript shows a caregiver using clear features of CHILD-DIRECTED SPEECH (CDS) — the simplified, supportive register adults use with young children — which an interactionist would read as evidence for Bruner's LANGUAGE ACQUISITION SUPPORT SYSTEM (LASS).

The first feature is REPETITION. The caregiver repeats the key word 'doggy' five times and recycles the same simple structure ('there's the doggy', 'the doggy's running'). This repetition is a defining feature of CDS: it gives the child multiple, low-pressure exposures to a word and structure in a stable context, making the form easier to extract. An interactionist reads this as the caregiver actively scaffolding the input rather than leaving the child to learn from random speech.

The second feature is the frequent use of QUESTIONS focused on the HERE-AND-NOW. 'Where's the doggy? Is the doggy running?' are questions about a present, jointly-attended object (the picture). This invites the child to participate and ties language directly to shared, visible meaning — JOINT ATTENTION. The exchange is built around a ritualised picture-book ROUTINE, exactly the kind of 'format' Bruner argued provides the LASS.

These features support an interactionist account because they show that the child's input is NOT impoverished or random but finely tuned, repetitive and embedded in shared activity. For Bruner, it is precisely this social support system (the LASS) that enables language to develop — complementing, rather than replacing, any innate capacity (Chomsky's LAD).

The forms 'goed', 'seed', 'gived', 'mouses' and 'mans' are OVERGENERALISATIONS, or VIRTUOUS ERRORS: the child applies a regular grammatical rule to an irregular word. 'goed', 'seed' and 'gived' add the regular past-tense -ed to the irregular verbs go, see and give (went, saw, gave); 'mouses' and 'mans' add the regular plural -s to the irregular nouns mice and men.

These forms support Chomsky's NATIVIST theory because they cannot be explained by imitation. The child has never heard an adult say 'goed' or 'mouses' — no one models these forms — so the child cannot be copying them. The only possible source is that the child has internalised an abstract grammatical RULE ('to make a past tense, add -ed'; 'to make a plural, add -s') and is applying it PRODUCTIVELY, even to exceptions.

For a nativist, this is exactly what the LANGUAGE ACQUISITION DEVICE (LAD) predicts: the child is not a passive imitator but actively generates grammar from an innate capacity, extracting and over-applying rules rather than memorising each form. The 'errors' are therefore positive evidence of rule-learning, and support the view that grammatical knowledge comes partly from within the child rather than wholly from the input.

This transcript captures an interaction between a caregiver and a child of two years eight months (2;8), and it is best read through BOTH major theories of acquisition: the child's own forms supply evidence for Chomsky's nativist LANGUAGE ACQUISITION DEVICE (LAD), while the caregiver's responses are a clear example of Bruner's LANGUAGE ACQUISITION SUPPORT SYSTEM (LASS). The data ultimately supports a COMPLEMENTARY reading — an innate capacity operating through social support — rather than a simple either/or.

Taking the NATIVIST evidence first, the child's most telling form is the virtuous error 'I eated it all'. 'eated' applies the regular past-tense rule (-ed) to the irregular verb 'eat' (ate). The child cannot have heard an adult say 'eated', so this cannot be imitation; it can only come from an internalised RULE applied productively. For a nativist this is precisely the work of the LAD: the child generates grammar from an inner capacity, extracting and over-applying rules rather than memorising each verb. The child's other utterances also show emerging, rule-governed structure — the question 'where it go?' uses a wh-word and a clear questioning function, evidence of the kind of creative, generative grammar Chomsky attributes to an innate endowment. Such features fit the wider nativist case (the universality and regular sequence of stages, the speed of acquisition) that grammar is partly built-in.

The CARETAKER'S side of the transcript, however, is a textbook display of the LASS and child-directed speech. The caregiver uses simple QUESTIONS tied to the HERE-AND-NOW ('More what?', 'Do you want another banana?'), keeping language anchored to the shared, present situation (JOINT ATTENTION around the meal). More importantly, she scaffolds the child's grammar directly. 'You ate it all!' is a RECAST: she re-forms the child's 'eated' into the correct irregular 'ate' while preserving the meaning and praising the content ('Clever girl'), modelling the target form without overt correction. 'Where did it go?' is an EXPANSION of the child's telegraphic 'where it go?', supplying the missing auxiliary 'did'. For Bruner, these finely-tuned models — offered just ahead of the child's current grammar — are the support system that lets acquisition proceed.

This is where the two theories must be EVALUATED against each other. The strength of the nativist (LAD) account is its explanation of the virtuous error: 'eated' genuinely cannot be imitation, and the strong sequence/speed of acquisition across cultures is hard to explain by environment alone. Its weakness is that the LAD is UNOBSERVABLE and the transcript shows that the input is NOT as impoverished as the strong poverty-of-the-stimulus claim implies — this caregiver actively recasts, expands and tunes her speech. The strength of the interactionist (LASS) account is exactly this visible scaffolding (the recast of 'eated', the expansion of the child's question). Its weakness is that the LASS cannot easily explain WHY the child produced 'eated' in the first place — the caregiver never modelled it — or why children acquire grammar even where caregiver speech is sparse.

Weighing the data, the transcript supports BOTH and points to their reconciliation. The child clearly brings something internal to the task (the rule that produces 'eated', the generative question structure) — consistent with the LAD; yet that capacity is being met and shaped by a richly supportive interaction — the LASS. Bruner himself framed the LASS as the support that lets the LAD operate, and this exchange embodies that: an innate rule-forming child and a scaffolding adult, working together. The fullest account, then, is not LAD versus LASS but LAD THROUGH LASS — innate capacity realised through social support — and this short mealtime transcript is a near-perfect illustration of why the two theories are best read as complementary.

The statement asserts a strong version of Chomsky's nativism: that because the language a child hears is finite and flawed, the rich grammar children acquire MUST come from inside — an innate LANGUAGE ACQUISITION DEVICE (LAD). Using this transcript of a child of two years five months (2;5) and her caregiver, I will argue that the poverty-of-the-stimulus argument has real force but that the statement OVERSTATES it: the same data shows the input is far richer and more supportive than the argument assumes, so the evidence points to a COMPLEMENTARY account (innate capacity plus social support, the LASS) rather than to innateness 'proving' itself alone.

First, the case FOR the statement. The poverty-of-the-stimulus argument holds that input is fragmentary and error-prone yet children acquire an infinite, generative grammar, so knowledge must be partly innate. The child here shows exactly the creative, rule-governed productivity the argument points to: 'tower fall down', 'big one bottom' are her OWN constructions, not repetitions of the caregiver, and she combines words in largely correct order to express new meanings. She is generating utterances, not echoing. A nativist would add that this child, like children everywhere, is moving through the universal sequence of stages (here, telegraphic speech) at speed, regardless of the particular input — evidence Chomsky attributes to Universal Grammar accessed via the LAD. So far the statement looks well-founded.

However, the same transcript undermines the claim that the stimulus is genuinely 'impoverished'. The caregiver's speech is highly ORGANISED and supportive — a clear LANGUAGE ACQUISITION SUPPORT SYSTEM (LASS) in Bruner's terms. 'Oh no, the tower fell down!' is an EXPANSION/recast of the child's 'tower fall down', supplying the determiner and the correct irregular past tense 'fell'. 'You do it' recasts the child's non-standard 'me do it' into the standard pronoun. The caregiver also uses CHILD-DIRECTED SPEECH features — questions ('Shall we build it up again?'), the here-and-now focus, and a clear instructional model ('Put the big one at the bottom') that the child immediately re-uses ('big one bottom'). Far from being impoverished, this input is finely tuned, grammatically corrective and pitched just ahead of the child. The strong poverty-of-the-stimulus premise is therefore questionable on this very data.

This is the heart of the EVALUATION. The statement's logic only works if the stimulus really is too poor to support acquisition; but interactionist research (e.g. Snow on child-directed speech) and this transcript both show caregivers systematically simplifying, repeating, expanding and recasting. That weakens 'proves… must be innate' to something more modest. At the same time, the interactionist position has its own limit visible here: the caregiver never modelled the child's own combinations, and could not, by tuning alone, explain the child's capacity to generate novel, ordered utterances. Each theory explains part of the data and not the whole.

In conclusion, I agree only PARTLY with the statement. The poverty-of-the-stimulus argument captures something real — the child's generativity and the universal, rapid sequence of acquisition are hard to explain by imitation of input alone, which supports an innate capacity (the LAD). But it does not 'prove' that language 'must be innate', because the transcript shows the stimulus is not impoverished at all: the caregiver provides a rich, scaffolding LASS (expansion, recasting, child-directed speech). The most defensible reading is complementary — an innate rule-forming capacity realised through finely-tuned social support — which is precisely how Bruner positioned the LASS in relation to Chomsky's LAD. The statement is therefore an overstatement of a half-truth.

The claim presents the LAD and the LASS as RIVALS between which the data must adjudicate. I will argue the opposite: this transcript of a child of three years (3;0) shows that the two theories explain DIFFERENT aspects of the same acquisition process, and that the data, far from forcing a choice, positively demands BOTH. The claim mis-describes the relationship between Chomsky's nativism and Bruner's interactionism, which are best understood as complementary — the LASS being, in Bruner's own framing, the support that allows the LAD to operate.

Consider first what only the NATIVIST (LAD) account explains well. The child produces a cluster of virtuous errors: 'builded' (regular -ed on the irregular 'build'), 'it not falled' and 'I builded it good'. 'builded' and 'falled' cannot be imitation — the parent never says them; indeed the parent models the correct 'built' and 'didn't fall' immediately, yet the child reproduces 'builded' even AFTER hearing 'built'. This is decisive: the child is not copying the input but applying an internalised RULE ('add -ed for past tense') productively, overriding the correct model she has just heard. No purely interactionist account, which leans on the child taking up the caregiver's models, can easily explain a child IGNORING a freshly-modelled correct form in favour of her own rule-generated one. The persistence of 'builded' against the parent's 'built' is strong evidence for an inner, rule-forming capacity — the LAD.

Now consider what only the INTERACTIONIST (LASS) account explains well. The parent's speech is a model LANGUAGE ACQUISITION SUPPORT SYSTEM. 'You built a house!' RECASTS the child's 'I builded a house', modelling the irregular past tense while affirming the meaning. 'Why didn't it fall down?' EXPANDS the child's telegraphic question 'why it not fall?', supplying the auxiliary and do-support and demonstrating standard interrogative structure. The parent also uses CHILD-DIRECTED SPEECH (the praise 'Clever!', the tag question 'didn't you?', the here-and-now focus on the shared building). This finely-tuned, scaffolding input is exactly what Bruner's LASS describes, and what Snow's research on child-directed speech documents. A strong nativist account that dismisses input as merely 'triggering' an autonomous device cannot do justice to how precisely this parent pitches each model just ahead of the child.

The crucial EVALUATIVE point is that these are not competing explanations of the SAME phenomenon but explanations of DIFFERENT ones. The LAD explains the child's GENERATIVITY — her capacity to produce 'builded' and novel, ordered questions from rules no one taught her. The LASS explains the ENVIRONMENT that feeds and shapes that capacity — the recasts and expansions that supply the data the child's rule-system works on. The claim's error is to assume one phenomenon needing one winning theory. In reality, the transcript shows a rule-generating child (LAD) embedded in a scaffolding interaction (LASS) at the same moment: 'builded' (the device at work) and 'You built a house' (the support system at work) sit in adjacent turns.

It is worth conceding the claim's grain of truth: the STRONG forms of each theory do conflict. A hard nativism that treats input as nearly irrelevant is contradicted by the obvious richness of this parent's speech; a hard interactionism that treats the child as shaped wholly by input is contradicted by 'builded' surviving the parent's correct model. So the data does force us to reject the EXTREME version of each. But that is not the same as choosing between the theories themselves; rather, it pushes us toward their moderate, COMBINED form.

In conclusion, the claim is mistaken. The data does not force a choice between the LAD and the LASS; it requires both. The child's virtuous errors — especially 'builded' produced in defiance of the parent's modelled 'built' — evidence an innate, rule-forming capacity (the LAD), while the parent's recasting, expansion and child-directed speech evidence the social support system (the LASS) that nurtures it. Bruner devised the LASS precisely as a complement to Chomsky's LAD, and this transcript shows why: innate capacity and social support are partners in acquisition, not rivals. The only thing the data 'forces' us to reject is the extreme version of either theory in isolation.

This transcript of a confident four-year-old (4;2) is rich enough to test several theories of acquisition at once. Analysed across them, the data supports a NATIVIST core (Chomsky's LAD, with the critical period hypothesis as a biological backdrop), while the parent's responses show that an INTERACTIONIST mechanism (Bruner's LASS) is also operating. The fullest account is again complementary — innate capacity, biologically timed, realised through social support — and the critical period hypothesis must be deployed with appropriate caution.

The most striking features are the OVERGENERALISATIONS that dominate the child's speech: 'goed', 'swimmed', 'builded', 'buyed' (regular -ed on the irregular go, swim, build, buy → went, swam, built, bought). These virtuous errors are the classic NATIVIST evidence. The child has not heard 'buyed' or 'swimmed' modelled — in this very transcript the parent supplies 'swam' and 'bought' — yet she produces and even REPEATS 'buyed' after hearing 'bought'. This shows an internalised past-tense RULE applied productively and robustly, overriding correct input. For Chomsky this is the LANGUAGE ACQUISITION DEVICE at work: the child generates grammar from an innate capacity, not by imitation. The breadth of the over-application (four irregular verbs in a few sentences) underlines how systematic, rather than item-by-item, her grammar is — consistent with Universal Grammar.

The CRITICAL PERIOD HYPOTHESIS (Lenneberg, 1967) sits behind this nativist reading as a biological framing: the proposal that language is acquired within a maturationally-defined window in early childhood. At 4;2 this child is squarely within that window and acquiring grammar rapidly and effortlessly — consistent with the hypothesis that an innate capacity is active during this period. It is essential, however, to deploy this hypothesis with CAUTION. The transcript itself offers no direct evidence of a critical period — that evidence comes from elsewhere (creolisation; cases of late or deprived acquisition). The Genie case is often cited here, but it is a single, ethically fraught case confounded by extreme abuse and malnutrition, so it can only ever be SUGGESTIVE, never proof; over-claiming it is a known error. Cleaner support for an innate, time-sensitive capacity comes from creolisation (Bickerton), where children generate a full grammar from a structureless pidgin — adding structure absent from their input, much as this child generates 'buyed' that is absent from hers. So the critical period hypothesis is best presented as a plausible biological backdrop to the LAD, qualified by the recognition that the strong 'sharp window' version is debated (many prefer a gradually-tapering 'sensitive period').

The parent's turns, meanwhile, show the LANGUAGE ACQUISITION SUPPORT SYSTEM. 'You swam in the water?' and 'daddy bought you an ice cream?' RECAST the child's 'swimmed' and 'buyed' into the correct irregulars while sustaining the conversation; the questions keep a here-and-now, shared focus typical of CHILD-DIRECTED SPEECH. This is Bruner's LASS — finely-tuned, corrective-by-modelling input. Its presence is significant for EVALUATION: it shows the child IS receiving correct models, which makes her persistence with 'buyed' all the more telling as evidence of an internal rule (the recast does not immediately overwrite the rule). The interactionist account thus explains the supportive environment, while the nativist account explains why the child's own system resists it.

Discussing how FAR the theories account for the data: the nativist account (LAD), with the critical period as biological backdrop, accounts powerfully for the pervasive virtuous errors, the child's generativity, and the rapid, rule-governed acquisition at this age. The interactionist account (LASS) accounts for the supportive, scaffolding input the parent provides, and rightly cautions against treating the stimulus as impoverished. Neither alone is sufficient: the LASS cannot explain why the child manufactures 'buyed' against the modelled 'bought', and the LAD in its strong form underplays the obvious richness of the parent's recasts. The critical period hypothesis adds a maturational dimension but cannot be confirmed from a single transcript and must be hedged.

In conclusion, the theories TOGETHER account for the data well, individually only partially. The transcript's overgeneralisations are strong evidence of an innate, rule-forming capacity (the LAD), plausibly framed by a biologically-timed critical/sensitive period (Lenneberg), and supported at the population level by creolisation rather than by the over-cited Genie case. The parent's recasting and child-directed speech show Bruner's LASS operating simultaneously. The data therefore supports a complementary, appropriately cautious account: a child's innate, maturationally-timed capacity to form grammatical rules, realised and shaped through finely-tuned social interaction — with the critical period hypothesis accepted as a reasonable but not transcript-proven backdrop.