lecture 6 - intro to psycholinguistics Flashcards
-psycho neurolinguistics
-psychology of language include
-Cognitive Neuroscience of Language
the order od differnt disciplines
linguistics
psychology
pathology/brain lesion studies
cognitive neuroscience
computational approaches, AI (language models)
study of language began from -pathology - brain lesion studies
history
Greece V B.C.
Alcmaeon of Croton - the
brain, not the heart, was the
organ that ruled the body.
Greece 400 B.C.
Hippocrates of Kos - the brain
is the seat of intelligence
Wernicke’s aphasia
a language disorder, makes it harder for you to understand words and communicate
-Speech is grammatically and phonetically ‘normal’
- But is semantically meaningless, mixing words, repetitive
● Nonsense ‘invented’ words that sound like real words
● The comprehension is impaired
Broca’s aphasia
● Speech production is not grammatically and phonetically
‘normal’, much simpler
● The preserved speech is meaningful, mixing words,
repetitive
● Word-finding problems
● No nonsense ‘invented’ words
● The comprehension is not impaired
Evidence for speech comprehension and production was
supported by…
supported by partially separate brain networks
Evidence for speech comprehension and production being supported by partially separate brain networks led to…
Led to perception and production being studied separately
perception and production rely on the same memory system
explain
word finding problems in both types of aphasia.
● Since grammar is linked to Broca’s aphasia, led some to believe
that grammatical processing is ‘separate’ from other language
functions
what is the skinner vs chomsky table
● Skinner’s Verbal Behaviour: Language is a learned behaviour
(like any other)→ when children learn to speak, they are reinforced
by positive responses and inhibited by negative resources.
● Chomsky’s 1959 review: behaviourism cannot explain natural language learning: (1) input is too poor/noisy; (2) poor grammar is rarely corrected yet learnt perfectly; (3) complex language unique to humans. Proposed Universal Grammar, suggested that
language is innate/genetic? - language module in the brain
criticism about skinner,chmosky debate
idea that language is unique to humans sidelines the evolutionof language and complex animal communication, overemphasis on grammer
Evolution of language – usage-based accounts:
gestures + vocalisations → abstraction
(we know humans evolved to learn and speak more complex language)
animal communication ‘dog buttons’
Morden technology allows much closer
study of animal communication
● Christina Hunger – speech pathologist invented ‘dog buttons’ after working with non-verbal children
dog buttons
● A lot can be expressed
without complex grammar:
simple words/commands,
phrases that describe
situations, abstract words (not
random)
● Large video database study by
Bastos et al., 2024
● Human-like fast mapping:
words learnt from single
exposure
the era of language models
Moving away from the origins/innateness to mechanistic
models of how language is acquired/processed/produced.
● Increasingly replying on experimental approaches, including
neuroimaging
● Increasingly computational in nature
cohort model
● As individual sounds
come in → brain
activates multiple
candidates
● Once sufficient input
is present - Word is
recognised
(Uniqueness point)
● Clear structure,
influential
what is the connectionist model
● Informed by cognitive and
computer science
● Neuro-like structure
● As each new chunk of input is
presented, activity spreads
along the network
connections, changing the
activation values in the
processing layers
● Inspiration for current
neural nets like Chat GPT
ai computational language models
● Closely linked to connectionism
● Learn aspects of language without
instruction - unsupervised learning
● An a mathematically plausible model
of language processing?
● Cannot learn grammar?
● Modern psycholinguistics
