W5 Nativist approach in early multi-word processings ✅ Flashcards
What is the assumptions of the nativist approach?
- Assume that grammar processes the relationships between abstract variables. (like computer)
- Assume that grammatical categories and rules are prioritised since birth (Universal grammar)
- Predict that the acquisition of a particular aspect of grammar should have an all-or-nothing quality. (child can utilise the grammar)
What is the nature of UG: Principles and Parameters
- All the possible rules for languages are innate.
- Grammar is universal (UG) – the rules of grammar apply in all languages.
- Different rules for grammar have specific constraints (parameter) across language.
-> Example of parameters: word-order and subject-usage (some language don’t need subject)
-> Children need to figure out the parameter in their target languages
What is the theoretical advantage & issues of UG?
Advantages:
- Avoids problem of explaining how children acquire complex grammatical rules
- Allows a unified theory of acquisition across languages whilst explaining how languages differ.
Problems:
- Parameters not clearly specified (how many, which part is coded by parameter)
- Unclear how children avoid setting parameters incorrectly (why make mistakes if mechanism is innate?)
- Bi/Multilinguism - how do children set >2 versions of same parameter?
What are the evidence for principle and parameter of UG? (3)
- Children’s early utterances (usually) observe adult word order -> set parameter.
- Children are creative in word production from early on -> applying rules of grammar.
- Preferential studies: children 2yrs and younger can understand word order (some evidence)
What are three kind of evidence against the nativist approach?
- Children display limited knowledge of S-V-O word order in production and act-out studies
- Naturalistic data studies provide evidence of partial, lexically-specific knowledge in a grammatical category -> failure to generalise
- Many studies show very close relation between what children hear, the frequency, and what/when they learn
-> supporting evidence for constructivist approach
What is meant by Radford’s maturational model?
WHY? To account for observed limitations in word production and generalisation failure in grammatical category
HOW IT WORKS:
- Lexical Stage (~20 months):
* utterances mainly consist of content words (e.g. nouns, verbs, adjs)
* some corresponding adult utterance omitted - Functional Stage (~24 months):
* Child’s innate grammar ‘matures’
* Activation of more complex grammatical components in utterances (e.g. auxiliary verbs, determiners and inflections)
What are the definitions of auxiliary verbs, determiners and inflections?
- Auxiliary: modality, certainty, futurity (e.g. can, will)
- Determiners: distinguish definite and indefinite referents (a vs. the)
- Inflections: mark verb tense and agreement (e.g. watch/watched; I watch/he watches)
Theoretical advantages & problems of the maturational model + evidence?
Advantages:
- Explains why early utterances are not fully grammatical
- Allow for development -> fits with evidence
- Generalisable to developing children with disabilities (deaf, blind)
Problems:
- Difficult to identify precise point in development
- Children show some usages of complex grammatical functions in earliest stages, although inconsistent and varies across languages
- At around 24 months, children’s use of many functional words related to lexical frames
-> first stage still influences the 2nd stage
What is meant by the Linking Problem and its solution (Semantic Bootstrapping)?
- Problem - how can children link up innate knowledge of grammatical categories to words they are hearing -> caregiver don’t label new words as noun or verbs
- Solution: semantic bootstrapping
* Assume that grammatical categories and rules are innate
* Children use semantics to map words from the input to these innate categories by using innate Linking Rules.
* Children can also link semantic roles (agent/patient) to syntactic role (subject/object)
* Examples:
person, thing (semantic) = noun (syntax)
action, change of state (semantic) = verb (syntax)
What is meant by distributional analysis as a solution for linking words to syntactic categories? (extra help for semantic bootstrapping)
Problem(s):
* Not all syntactic categories fit to one semantic label (e.g. not all verbs are actions - believe, want)
* Not all subject are agents
Solution: distributional analysis
* Determine word order from prototypical sentences
* Then apply knowledge of word order to work out grammatical category of more abstract terms.
-> basically applying sentence structure of obvious grammatical categories to figure out novel terms
Advantages and problems for semantic bootstrapping?
Advantages: Explains…
- how children break into innate system
- why early utterances follow adult word order
- how children learn non-semantically obvious verbs and nouns, etc.
Problems:
- many children’s early lexically-specific utterances are not semantically prototypical -> unlikely to be based on innate knowledge
- doesn’t work for passive sentences
What is the overall conclusion for the nativist approach?
WHAT: another approach to explain how children learn to put words together into sentences
HOW:
- assumes children operates with innate knowledge specific to grammar
- differences between grammars in languages is encoded in principles and parameters
- continuity (fully innate) vs. maturational (some parts of the grammatical system ‘switch on’ later in development
KEY PROBLEMS
- the linking problem (connect language and innate grammar)
- define specificity between innate maturation vs. development from gradual learning
