Please enable JavaScript.
Coggle requires JavaScript to display documents.
Developmental Language Disorder/Specific Language Impairment (Hedenius et…
Developmental Language Disorder/Specific Language Impairment
What is DLD/SLI?
Diagnostic Criteria
What is lang?
Grammar, phonology, semantics
"nuts + bolts" to learn words + put them together
Beyond speech but prior to communication
Distinguish from
Speech
Speech sounds (physically speaking) - motor movements + generating
Not necessarily communication difficulties
Communication
Understand social rules, non/literal lang.
Can be distinguished
Diagnosed by
SLT
Ed. Psych
Clnical Psych
Paediatricians
Psychiatrists
GPs
Teachers
Delay vs disorder?
Developing later than peers but catch up + typical when do so (DELAY)
Significant, severe, persistent, negative consequences social and education (DISORDER)
DLD = 'unlikely to meet socially-define functional expectations either currently or in the future because of current/future lang abilities'
Criteria
Vocab issues
Issues with sentence structure
Speed of online use of above
NOT due to
Sensory impairment
Motor dysfunction
Another medical condition
Intellectual disability or global developmental delay
Assess both expressive and receptive lang
Terminology
DLD = Specific lang impairment, primary land disorder, lang disorder, dysphasia
Prototypical forms of impairment and adult outcomes
Prototypical
Morphology
Certain aspects of syntax
(v often) phonology - speech sounds
Not global delay - differences across aspects
Also diff across lang
Particular difficulties WITHIN lang > not all components of lang equally affected
Outcomes
Psychosocial outcomes
Anxiety disorders
Drug abuse & antisocial behaviour (SLCN 5yr)
Low self esteem & social problems (receptive lang dis)
Prevalence SLCN amongst male prisoners V high
Social participation
Durkin et al
Children(N=139) 7yrs > retest at 16yrs
Literacy
Acaemic achievement
Friendships
Emotional health
8% no difficulties, 41% difficulties in 2, 32% difficulties in 3
2 areas > 90% literacy/academic achievement
3 areas > 86% the above + difficulties with friendships
Clinical Markers
Non-word repetition (NWR)
DLD worse than children matched on overall lang level (younger)
Regular past tense - very sensitive and specific marker in school age english speaking children with DLD
Heritability
NWR linked to chromosome 16q (replicated in other cohorts)
Expressive lang score on Clinical Evaluation of Lang Fundamental + chromosome 19q (inconsistent replication - only NWR)
DLD/SLI
Heterogeneous
Not 'just dyslexia' - can have dyslexia without having DLD
Theories of Lang Deve.
1) Linguistic nativist (grammar-based) accounts
Domain specific modular theories
Grammar & lexical learning = separate
Grammar learning underpinned by Universal Grammar (underlying hierarchical rules universal to all lang)
Innate cog modules do specific things
Pinker (1999) & Clahsen (1989) - dual route (words and rules)
Rice (2000) - extended optional infinitive hypothesis
Van dey Lely (2005) - deficit in construction of certain grammatical principles
Predictions
If grammar is problematic
Regular morpho-syntatic forms predicted to be diff - compared to irregular (word learning rather than systematic rules)
Irregular forms should be not as effective (word-learning system)
Actual rule should be impaired for ALL lang (for DLD)
2) Usage based theories
Lang can be learned from
Shared intentionality
Understand that they are interlocutor/collaborator are jointly engaged 'together' in activity, understand shared goals/intentions
Earliest instantiation > engagement in true joint attentional frame
Statistical learning
Ability to implicitly learn the probabilities with which particular context predict the occurrence of certain items (transitional probabilities)
Distributional analysis
Patterns in things we are exposed to
Children using
Infants are very good at learning these kinds of distributional probabilities (Kirkham et al., 2002)
Distributional learning must be at least one of the mechanisms infants use to learn the phonology and vocabulary of their lang
Early acquisition through stats learning
Children could have simply implicitly learnt that 'is' will be followed by 'ing' and that 'can' is unlikely to be closed followed by 'ing'
This type of distributional/statistical learning could still be playing a very important role in helping children learn about word categories (e.g. nouns tend to be preceded by a/the, verbs often have ed/ing on the end)
Early learning of syntax based around lexical schemas
Goal: map SVO onto Agent-Action-patient
Process of form-to-meaning mapping (similar word learning, prototype formation)
Starting point = rote-learning of common phrases
Lexically-based phrases 'find X'; 'Where X'
Gradual schema abstraction and generalisation based on semantic similarity
Theoretical account + predictions for DLD
Evidence problematic for Linguistic Nativist accounts
Area of morpho-syntax on which SLI/DLD particularly impaired varies cross-linguistically
Factors
Low perceptual salience
3rd person singular -s vs. Italian 3rd person singular -a or -e (Leonard et al 1987)
Italian SLI children don't omit singular
DO freq omit 3rd person plural morphemes
Frequency
E.g. English reg n pl +s easier than 3rd ps sg
Oetting & Rice (1993)
Regularity and complexity
E.g. word order not particularly impaired in English DLD but it is key feature of German DLD
Lindner & Johnson (1992)
Evidence supporting statistical learning deficits
Syntax learning
The participants are not told the underlying rules
Learning phase: exposure to the strings usually occurs under a cover task, such as memorisation of those strings
Artificial grammar learning studies involve ps being exposed to strings of letters, tones, or other sequences generated by a rule-system (which specifies which orders can/not occur)
Test phase: ps are asked to judge whether novel strings are un/grammatical
The strongest evidence that statistical learning might play a role in child acquisition of grammar comes from artificial grammar learning studies
Statistical learning can be used to build categories (so this is a logical argument against the linguistic nativist (modular) assumption that grammatical categories are innate)
Statistical leanring is not domain-specific
Typical Deve
Implicit statistical learning related to degree of syntactic priming in typically-developing primary school aged children
Related to comprehension of passives and relative clauses
How children learn to generalise or constrain their use of argument structure (i.e. the mapping of sentential frames to event-based meanings)
DLD as stats impairment?
and artificial grammar learning
Even when the stimuli consist of non-verbal tone
Statistical learning abilities correlate vocabulary abilities in children with DLD
DLD impaired relative to typical control on serial reaction times tasks
Procedural Deficit Hypothesis
Ullman & Pierpont (2005)
Procedural: acquisition, consolidation and automatisation > particularly motor and perceptual
Declarative: acquisition and storage of facts and events (e.g. including word meanings)
Claim DLD = impairment of neurological system responsible for procedural learning
Learning of implicit sequences
Learning probabilistic rules
Categorisation
Working memory
Procedural learning used for rule-governed aspects of lang (syntax, morphology and phonology)
Explicitly predicts unimpaired vocabulary
NB: this also includes irregular past tense forms
Explicit vs. implicit learning
Implicit learning tasks - not conscious
Explicit learning - deliberate
Problems
Children with DLD perform badly with the irregular past tense (supposedly learnt through rule system)
Children with SLI/DLD are typically slower to learn vocab
Evidence which is problematic for all theories
Some evidence re selective attention (Spaulding et al. 2008)
DLD is NOT the same as ADHD
Sustained attention (including visual sustained attention (Ebert & Kohnert, 2006)
Impairments not easily accounted for
Hedenius et al. (2011)
Grammar predicts procedural learning and consolidation deficits in children with SLI
Procedural deficit hypothesis - SLI explained by abnormalities of brain structure > procedural memory
Predicts impairments of procedural memory - underlying grammatical deficits
Consolidation and longer term procedural sequence learning in children w/SLI
Alternating Serial Reaction Time (SLI vs typically developing) - initial learning session + ~3 days later test for consolidation/learning
Both = initial sequence learning, did not differ in longer term learning
Only TD = clear consolidation
When re-categorized on grammar deficits = clearer pattern
Grammar impaired/normal = initial sequence learning, only normal grammar = consolidation/longer-term learning
Grammar impaired = lose any sequence knowledge from initial session
Suggests that consolidation + longer term procedural learning are impaired in SLI