PSYC 101 - midterm 2 (Schachner)

what makes humans unique?

ability to use symbols to represent thoughts, feelings, knowledge and communicate them to others

challenges to language development

perceiving individual sounds, understanding meanings of sounds, appreciate word order and its effect on meaning, interpret how it was said

Darwin’s take on language development

man has an instinctive tendency to speak

children @ ___ y/o have mastered basic structure of native language

5

components of language

phonology, syntax, semantics, pragmatics

phonology

units of sound that distinguish meaning (ie. lake vs rake)

syntax

rules for combining words (ie. “bit the man dog”)

semantics

understanding meaning of words (ie. did you “paint” if spilled some paint on the street?)

pragmatics

understanding conversational/social conventions and knowledge of how language is used (ie. sarcasm/rhetorical questions)

2 possible developments of phenome perception

1.) babies learn to tell different phenomes apart thru experience, get better with age
2.) babies learn NOT to tell apart the sounds that aren’t important in their language, get worse with age (perceptual narrowing)

perceptual narrowing

our experience shapes our perceptual abilities, become less sensitive to stimuli not common experienced

besides language where else does perceptual narrowing exist?

face, rhythm, and speech perception

logic behind Dr. Janet Werker’s conditioned head-turn method to test perceptual narrowing

expt: babies = interested in toys, train them to turn their heads when they hear a sound change by playing a sound repeatedly and then when the sound changes, immediately present the toy after

babies will eventually turn heads when sound changes, anticipating toy

results of Dr. Janet Werker’s conditioned head-turn method to test perceptual narrowing

age changes infants’ phenome perception —> worsens older they get

@ 6 mo. - can tell apart phonemes in 2 different languages (Hindi and Nthlakapmx) that don’t share phonemes w English

@ 10-12 mo. - can’t

at what age do babies start to coo?

6-8 weeks

sound perception and production at 6-8 wks old

cooing, increased awareness that vocalizations elicit responses from others, “dialogues” of reciprocal sounds

at what age do babies start to babble?

3-10 months, is different per language and conforms to sounds, rhythm, and intonation of language

where else is babbling important?

in animals that learn their calls, overall is important to learning process

prosody

characteristic rhythm, tempo, cadence, melody, and intonational patterns of language, exists prenatal sensitivity to it

how do infants use prosody?

before understanding words, use way phrase is said to figure it’s meaning (ie. “yes good boy” and “no don’t touch” = both encouraging if same in the same tone)

infant-directed speech (IDS)

distinctive speech mode adults adopt when talking to young children (ie. parentese, motherese), aka “musical” speech

consists of warm, affectionate tone, high pitch, extreme intonation, slower speech, and exaggerated facial expressions

application of IDS

IDS = common throughout world, but not speaking to infants

infants prefer IDS > speech directed to adults even when in wrong language/not directed at them

impact of musical speech/song

infant-directed song = very frequent in baby’s life, performance quality (expressiveness) depends on infant actually being present

songs can regulate infants’ mood (often adults too), powerful in calming infants, infants prefer singing > all other speech

infants listened to singing 2x longer than speech before fussing and crying

influence of worldwide lullabies on babies

across cultures, lullabies have similar acoustic features —> infants relax to unfamiliar foreign lullabies too

characteristics of early language experience

is more musical/emotive, includes both speech and song

implicit (unconscious) grammar

learned naturally, w/o explicit teaching

syntactic development @ ~12-18 mo.

1 word speech (ie. "Dada”) and some combined worked that child treats as 1 (ie. “iwant”)

syntactic development @ ~2 y/o

2-word phrases, makes telegraphic speech

telegraphic speech

lacks nonessential elements, short simplified sentences (ie. “give juice”)

Hirsh-Pasek Golinkoff, 1991 expt on 12-16 mo. olds

when shown 2 pics with descriptions differing only by word order, infants reliably looked at the pic that matches the spoken phrase

2 arguments about kids’ syntax knowledge

argument for = kids use stable word order (ie. eat cookie, not cookie eat)

argument against = kids could just be copying exact phrases, memorizing

Jean Berko-Gleason - test of kids’ application of syntactic rules

expt: the Wug test, asked 4 y/o to generate correct endings to novel words

results: kids applied grammatical rules to new, nonsense words

syntactic rule

a general principle about how words/word parts get combined

overregularization errors

children overapply rules, treating irregular words as regular (ie. he goed to the store)

Noam Chomsky and language acquisition claims

had a strongly nativist theory of language acquisition, claimed:

“poverty of the stimulus argument” - children don’t hear enough input to learn language

humans have a built-in, innate mechanism to learn syntax

a lot of knowledge is built-in —> “universal grammar”

language acquisition device only works in childhood, not adulthood

critical period for language

age matters in children learning language (spoken or signed)

sometime between 5-puberty, language acquisition starts becoming much more difficult and less successful (especially syntax/grammar)

2 evidences of critical period for language

1. Feral children - Genie = exposed to language at 13 and had trouble with syntax, when asked to “give the one that’s on blue” would pick the wrong one

2. 2nd language learners - immigrants from Korea and China to the US at different ages, if arrived before 7 y/o performance on English grammar test = just as god as native English speaker’s then declined if arrived at an older age, even when compared to people who’d been speaking English for the same number of years

Nicaraguan sign - context

1977-1983 - deaf children first grouped together put into oralist school —> didn’t learn sign language being taught but developed own

Nicaraguan sign - study findings

age matters - younger learners’ signs = more fluent and grammatical, will structure words and create grammar but adult learners don’t

input matters - those who learned less-structured language input were less fluent, later cohorts of kids made language more structured

Nicaraguan sign - significance/implications

children acquire and invent language in wide range of environments to communicate: structure it when input is chaotic & gradually create a new language when input is absent

deaf children of hearing parents need early intervention so have access to language input or will show big delays, parents should learn sign language for deaf child

innateness/nativist theories on language development

chomsky, pinker: have instinct to learn language = built into cognition, all human children learn language, strong similarities in how all language is structured, poverty of the stimulus argument

learning/constructivist theories on language development

m. tomasello: language can be learned without innate language-specific learning methods, language is constructed when there’s general-purpose learning mechs and the desire for effective communication

Nim Chimpsky

chimpanzee that was taught English, compared to humans:

learned slowly through extensive training, few words, highly repetitious, very limited learning of syntax

what’s the argument with large language models and syntax

LLM show evidence that syntax is learnable and syntax can be developed when trained to read lots, but LLMs need enough training and parameters

LLMs

a model based on neural networks trained using “unsupervised learning” where model learns to predict next word in a sentence based on previous words

Chomsky and LLMs

chomsky and others thought it wasn’t possible to learn syntax from observing language, LLMs disproved this

why do LLMs need so much more training?

Michael C. Frank: either of 4 reasons

1. people have something innate

2. need to be embedded in social environment for active/social learning

3. importance of multimodal learning, context

4. we’re basically approaching this wrong, evaluation differences

significance of LLMs

LLMs can inform our understanding of langauge development by considering similarities and differences from human learning

morphemes

smallest unit of meaning in a language (ie. dogs = dog + s = 2 morphemes)

challenge to semantic development

word segmentation - where words start and end, just attending to pauses don’t help

how do infants pick out individual words in pause-free speech even in the 1st days of life?

1. stress patterns of language (ie. English - stressed syllables typically begin words)

2. statistical learning to track which speech sounds tend to go together

statistical learning

learning and recognizing patterns

translational probabilities

how likely each syllable is to follow the previous one (ie. pretty baby, good baby, etc —> bee always after bay but what’s before bay changes so can guess bay-bee = 1 word)

saffin, aslin, & newport, 1996 expt on babies and statistical learning

expt: 8 mo. old hears a stream of speech syllables with no gaps between “words” of an artificial language so they can only find words by finding patterns

use violation of expectation but with sounds to test babies

test: play individual “words” and “nonwords” from fake language 1 by 1, speaker location is marked with a blinking light, measure how long infant looks at speaker triggering sound to keep playing

saffin, aslin, & newport, 1996 expt on babies and statistical learning - results

infants looked longer at fake, nonwords than fake words, infants were also sensitive to visual statistical regularities (ie. same test but on visual shapes instead)

learning word meanings - 6 month olds

know some common nouns, will look toward object you label

~7-18 mo. - learning word meanings

1st words spoken, vary though (usually people’s names, objects, events from everyday life)

young children overextend word meanings

will use a given word in a broader than appropriate context (ie. say ball, mean: ball, balloon, marble, egg, etc)

toddlers and learning word meanings

learn about 50 words then language explosion, ~18-24 mo. - learn 10-20 words/day

carey and bartlett, 1978 expt on: 2-3 y/o - fast mapping

children learn words after even just 1 exposure even if it’s only referred to indirectly

test: can you find the “koba?”, asking immediately, after 1 week/month and kids still remember

assumptions/biases children make to learn words

whole-object, shape, mutual exclusivity, syntactic context

whole-object assumption/bias

children assume word refers to whole object and not part, action, or property (expt was on 12-28 mo)

shape assumption/bias

children think a new word refers to objects of that same shape (expt was on 2 y/o)

mutual exclusivity assumption/bias

assume each thing will only have 1 name, infants infer what a new word means by ruling out objects already know (expt was on 1.5-2 y/o)

syntactic context assumption/bias

children use linguistic context in which novel words appear to infer meaning

syntactic bootstrapping

strategy where children use grammatical structure of whole sentences to figure out meaning/syntactic form of a word (ie. a sib = the container, some sib = the substance, he’s sibbing = the action)

social pragmatics

paying attention to social cues/context, social understanding = important in language acquisition

reference adult’s attention, emotions, intentionality (ie. assume labels refer to intended not accidental actions)

tomasello, strosbery, and akhtar, 1996 expt - pragmatics/word learning

expt: speaker says “Oh look a tulma” and kid knows to look at other’s cup and not own

toddler (1.5 y/o) uses speaker’s gaze and emotion to determine what novel word refers to

social cognition and language learning

kids with atypical social cognition (ie. w/ autism) have delayed language development

simultaneous bilingualism

multiple languages from birth

sequential bilingualism

acquisition of another language after acquisition of 1st language is well under way

phonology in simultaneous bilingualism

perceptual narrowing - maintains discrimination of sounds from both of their languages —> infants can tell apart 2 languages they’re learning based on sounds but unclear if can fully differentiate 2 phonologies of 2 languages

2 languages are learned somewhat independently, creating 2 lexicons at the same time and development of each language operate somewhat independently

simultaneous bilingualism - determinants of learning speed in each language

input amount, household composition, societal factors

bilingualism and language delay

language development is slower but not a delay, have similar vocabulary sizes to a monolingual just distributed across 2 languages

sequential bilingualism - proficiency depends on

characteristics of child, learning situation, age of acquisition

critical period effects on sequential bilingualism

those who acquired swedish earlier in life sounded more native like when judged by native-speakers

language attrition

switch from native language to 2nd language is so complete child can lose fluency/proficiency to first language

bilingualism’s clear advantages

communicating with different speakers, read books in original languages, access to multiple cultures

bilingualism and executive function

mixed results about whether bilingualism improves executive function, some yes some no

executive function components

inhibition (ignore distractions and stay focused), cognitive flexibility (easily switch perspectives, attention, etc), working memory (hold info in mind and manipulate/reason it)

tested with the marshmallow task

exec function - inhibition

ignore distractions and stay focused, resist making 1 response in favor of another

exec function - cognitive flexibility

flexibly switch perspectives, focus of attention, or response mappings

exec function - working memory

hold info in mind and manipulate/reason about it

alexis smith-flores - studies how children

develop a general understanding of emotions ←→ use others’ emotions to learn about specific people and things

alexis smith-flores - methods to study children

what infants look at, what children think/say, what children/infants do

common method to test what babies know

looking time method aka. preferential looking, habituation

purpose of surprise in infants when their expectations about how the world works gets violated

having right expectations helps you prepare for and intervene on the world, violation of expectations suggest a need for learning/exploration to be more accurate

stahl & feigenson, 2015 expt - infants’ learning/exploration

test on 11 mo. olds

expt: babies expect solidity, gravity, and continuity —> 2 treatments of either following expectation or violate by having obj: go through wall, floating, switch locations

then: babies = shown an obj (target) from the tx that violated/expected makes a sound + a new obj never seen before

test: play a sound and see which obj the infant looks to

results: infants prefer the new object under the expected tx and fail to associate sound with the new obj, infants prefer target obj and learn sound associated with it under the violated tx

test: give (target) obj for infant to play with and a novel obj for 1 min after the 2 tx

results: infants prefer new obj when under expected tx, but prefer target obj when under violated tx

significance of surprise in infants

expts show infants do surprise-induced learning and prompts and guides exploration

stahl & feigenson, 2015 expt - guide exploration

expt: show babies a train floating or going through a wall

results: will show more dropping or more banging behavior on the train toy depending on tx given

perez & feigenson, 2020 expt - explanations change infants’ behavior

test on 12 mo. olds

expt: see a truck (target) supposedly go through a solid wall, then shown either: truck did go through a wall OR wall had a hole

results: surprising outcome w explanation led to more preferential exploration of the novel distractor obj over target obj

explanation abolishes surprise-induced exploration

studies by perez, feigenson, and stahl - implications

show: violations provide opportunities to learn/explore world, infants learn from and explore surprising objs, infants integrate explanatory information into their information-seeking

kidd et al, 2012 expt - surprise and attention

expt: show infants a picture pop up behind different colored blocks in either: same color every time, pattern of alternating colors, or no pattern and random colors

results: higher looking-away probability during the really complex sequences

the goldilocks effect

infants prefer looking at and learning from complex but not overly complex things

they pay attention to learning progress and look away when they’re failing to learn from unpredictable events

does surprise always draw attention?

no, goldilocks effect

how do infants use surprise to learn?

surprise = response to a failure to predict how the world works, is a good indicator of good and optimal opportunities to learn

once surprise has been accounted for, exploration goes back to baseline

infants = efficient learners

mental theory development

built by children to understand how world works, starts with early-developing foundation (core knowledge) that’s built on by experience/learning

mental theories

intuitive ideas about why or how other ppl behave the way they do, predict and explain events

kids will test these theories and revise them

theory of mind (tom)

the ability to attribute beliefs, intentions, desires, knowledge, and other mental states to other people (ie. understanding other people’s preferences can be different from own)

repacholi & gopnik, 1997 expt - TOM

expt: expter deliberately chooses opposite food to the one baby likes then ask baby to give her one

results:

@14 mo. - seem to think everyone has the same preference they do —> will give what baby likes to expter

@18 mo. - understand that ppl have different preferences —> will give what expter seems to like