PSYC 336 MIDTERM 2

what are the two levels at which duality of patterning operates at?

1) meaningless units of sound combine to create meaningul units (words)

2) meaningful units (words) combine to create larger meaningful syntactic units

habituation

decreased response to a stimulus after repeated presentation of that stimulus

stager & werker: switch task results

• infants exposed to pictures of different objects while speaker plays a novel word (“lif)

• they were more interested when the label was DISSIMILAR to the stimuli

  • “lif” vs. “neem”

• didn’t notice the switch if words were SIMILAR

  • “bih” vs. “dih”

lexical representations 

information that is commited to long-term memory about the sound and meaning properites of word and certain constraints on their syntactic combination

what are the two possibilities explaining the switch task results?

1) babies do not store detail information about sounds in lexical representations

2) babies do store detail information, but they are bad at the retrieval process

preferential looking paradigm study

infants were shown two photos and asked to look at one of the objects

• they will look at the corresponding item 

• used to assess pre-verbal infants 

what occurs when a baby has become habituated to an auditory stimulus

after repeated presentations of the same auditory stimulus, the baby spends less time orienting toward the stimulus when it is played again

• than it did during past presentations of the same stimulus 

whole object bias

assumption by babies that a new word heard in the context of a salient object refers to the WHOLE thing

Holllich et al. (2007): results

• examined how the act of naming an object impacts how babies visually examine the object and its parts 

when the object was named “modi”, infants looked longer at the WHOLE object than part of it

• when asked to find the “modi” 

mutual exclusivity bias

one-to-one correspondence between object categories and linguistic labels 

byers-heinlein & werker (2009): mutual exclusivity bias results

• study examining monolinguals, bilinguals, and trilinguals

17-18 month old bilinguals are less likely to show mutual exclusivity bias

• trilinguals show this bias even less

• exposure to more than one language reduces bias

categories: word levels

• superordinate level

• basic level

• subordinate level

superordinate word level

general word 

• “animal”

basic word level

the favored mid-level category of words that strike a balance between similarity among members of the cateory and distinctiveness from members of the other categories

• “cat”

subordinate word level 

more specific

• “ragdoll”

under-extension

mapping new words onto categories that are too specific

• a ginger cat is NOT considered a “cat” because it is ginger

over-extension

mapping new words onto categories that are too general

• a dog is considered a “cat” because it is fluffy and has fur 

true or false

young children are less reluctant than adults to extend a new word to a broad category

false

• adults have strong assumptions and are more likely to do so 

latourette, chan, & waxman (2023): results

• plushies either had different names or the same name (dax)

distinct names → looked longer at novel stimuli

consistent names → indifferent looking times

• if they are forming a category, they will look for similarities when they have the same name 

associative learning

learning process by which associations between two stimuli are made as a result of experiences in which the two are paired

• think of pavlov’s dog

true or false

children understand that language is rooted in highly social behaviour

true

baldwin et al. (1996): study results

• infants play with a novel object alone while the speaker phone plays “dawnoo! there’s a dawnoo”

infants fail to link the word ‘dawnoo’ with the novel object 

tomasello & barton (1994): study results

• experimenter looks through different cones with objects inside to find the toma

when the experimenter expressed an excited “Ah”, infants were able to recognize what the “toma” is 

koenig & woodward (2010): test results

• study examining how children prefer to learn from reliable speakers

infants in the inaccurate condition were less likely to recognize the right object

• reveals that children prefer to learn from reliable speakers

how many words in the following sentence actually map onto an object category?

Mommy is stirring porridge in the pot.

1: pot

gilette et al. (1999): study results

• adults were asked to guess meaning of words based on video clips of parents interacting with toddler 

45% accuracy for nouns 

15% accuracy for verbs 

• adults struggle to guess verbs, especially if they lack linguistic contexts that typically accompany with the words 

true or false

verbs do not come specified for argument structure 

false

argument structures

syntactic frames that provide info about how many objects/participants are involved in each event and what kind of objects/participants are involved

intransitive verb

only one participant

• ex. sleep

transitive word

two participants, the subject, and the object of the action

• ex. kick → the baby kicked Jane

ditransitive word

three participants, the subject, the obejct, and the indirect object

• ex. send → he sends her a letter 

syntactic bootstrapping

using syntactic properties of words to identify aspects of meaning that words are likely to convey

niagles (1990): study results

• infants watched 2 salient actions: arm waving and pushing while the experimenter said a statement

“the duck is gorping the bunny” → looked longer at pushing scene

“the duck and the bunny are gorping” → looked longer at the arm-waving scene 

true or false:

by 13-14 months, they expect words that sound like nouns but not words that sound like adjectives refer to categories 

true 

true or false

by 2 years, they still aren’t able to differentiate between common nouns and proper names

false

true or false

by 3-4 years, they are able to infer that adjectives are used to communicate properties of objects 

true 

30 million word gap

by the time children enter kindergarten, children from high SES hear about 30 million more words than children from low SES

true or false

quantity of input is not strongly correlated with size of children’s vocabulary

false

schneidman et al. (2013): study results

• examined total numbers of words compared between single-speaker and multi-speaker families

• overall number of words children heard at 2.5 years did not predict vocabulary size at 3.5 years old

• number of words directed at 2.5 year old children predicted their vocabulary size at 3.5 years 

infant directed speech

baby talk

• high pitch, slower, exaggerated vowels and consonants

• babies prefer to listen to this 

true or false:

toddlers learn novel words better than IDS over Adult Directed Speech

true and false!

• there’s conflicting evidence as toddler’s ability to learn novel words might be specific to a western context

• asian cultures typically do not use baby talk yet they still have a decent vocabulary size

characteristics of high quality input

• contingency

• referential transparency

contingency

the extent to which a caregiver’s response is connected in timing and meaning to the child’s words or communcative actions

• context dependent

referential transparency

extent to which words are connected in obvious way to obejcts in joint attention

cartmill et al. (2013): study results

• adult subjects guessed target words when word was replaced by a beep in audio clip 

referential transparency score of parents predicted their child’s vocabulary 3 years later 

weisleder & fernald (2013): study results

• infants were given a decide to wear that recorded everyday speech they overheard

• child-directed input had larger vocabularies

• they were also faster to look at a picture that matched the spoken word than those who had less child-directed speech

root/free morpheme

individual words that can exist as is

• ex. cat

bound morpheme

cannot exist by itself

• inflectional & derivational

inflectional morpheme

bound morpheme mostly related to grammar

• ex. cat-S

• ex. fly-ING

derivational morpheme

bound morpheme that can make changes to the meaning of the word

• ex. teach-ER

how to create new words

1) compounding

2) derivational affixes

3) inflectional affixes

true or false:

children often produce the correct forms of exceptions first, then overgeneralize by using rule-like forms 

true 

two systems argument for rule-like vs. exceptions 

regular forms → assemble parts automatically

irregualr forms → memorized 

one system argument for rule-like vs. exceptions

solve by analogy

kids also overgeneralize the irregular forms 

semantic priming

hearing or reading a word partially activates other words that are related in meaning to that word 

lexical decision task

participants read strings of letters on a screen and decide as quickly as possible whether the string of letters forms a real word or not 

results of lexical decision task

it took longer to recognize a real word when it follows a word related in meaning

• ex. nurse → doctor

visual world paradigm results

• Yee & Sedivy (2006)

• participants are asked to click on the hammer while looking at pictures of different objects

participants were briefly drawn to “nail” 

facilitation

processes that make it easier for word recognition

• how activation in one word will make activation of another word faster

inhibition

processes that make word recognition more difficult

• primed with “stiff” → slower to recognize “still” 

neighborhood density effects

people take longer to recognize words that come from dense neighborhoods than words that come from sparse neighborhoods 

• ex. sling > stench

spreading activation network

activation “flows” throughout the network along these links

• hearing “lion” will also excite the connected words “cat” and “tiger”

mediated semantic priming

“stripes” activated due to it being related to “lion” via some intervening word

decay function

the rate at which information fades in memory

• information that has become activated gradually returns to a baseline of actvation 

competition effects

occurs most when presence of related wordsslow down or impedes the recognition of target word

excitatory connections

connections along which activation is passed from one unit to another

• the more active a unit becomes, the more it increases the activation of a unit it is linked to 

inhibitory connections

connections that lower the activation of connected units so that the more active a unit becomes

• the more it suppresses the activation of a unit it is linked to 

localist interpretation

discrete letter/sound units

• assumes that a word is everything that’s represented by a node 

• assume it contains meaning, semantic features, speech sounds, and syntactic representation 

distributed view

words could be captured as bundles of features with no intervening word nodes 

homographs

words that spelled the same but have different meanings

homonyms

words that are spelled and sound the same but have different meaning

polysemous

words that can refer to more than one related word sense

exclusive access hypothesis

argues that only one or the other meaning is activated

exhaustive access hypothesis

everything related will be activated

bottom up approach 

sound level → lexical level → semantic features

• flow of activation is one direction only

• separate decision mechanism to select the most contextually appropriate word

bi-directional approach

context → semantic features → lexical level

sound level → lexical level → semantic features 

crossmodal priming task results

• participants watch a black screen as a sentence plays, when the priming word occurs, a word flashes on the screen

immediately after the primed word:

• recognition time is the same for homonyms

3 syllables after primed word

• recognition time is the same for unrelated homonym and word

in a cross-modal priming experiment, participants hear the sentence: “the store ran out of change, so lisa went to the bank yesterday to get more change”

if the visual target word appears 5 syllables after the primed word “bank”, what result would you expect

participants would recongize the word “river” as quickly as “cake”

cohort competitiors

words with overlapping onsets

cohort model

multiple cohort competitors become active immediately after the word onset and are gradually narrowed down to a single candidate as sounds unfold 

• uniqueness point

incremental language processing

as we hear a word unfold, we start forming guesses about what it might be before it’s finished

spivey & marian (1999) results

• tested cohort competition effects for cross-language cohort competitors 

hearing russian instructions

• more likely to look at english cohort competitors

hearing english instruction

• no clear effect of cross-language competition

TRACE model

streams of sounds are continuously fed into the word recognition system WITHOUT any need for identificaiton of the left edges of words 

• rhyme competitors will be at a disadvantage

logographic writing system

symbols represent meaning

syllabary writing system

mapping symbols onto syllables

alphabets writing system

letters represent sounds

phonemic awareness

conscious recognition of phonemes as distinct units 

boustrophedon writing system

read left to right and then right to left

saccades

when/where our eye gaze jumps

fixations

when/where our eye gaze stay still

dual-route model

two pathways that links graphemes with speech

• predicts that irregular or exceptional words take longer

assembled phonology route (dual-route model)

graphemes are “sounded out” against their corresponding sounds, beginning at the left edge of the word

• works for “regular” words and words we have never seen before

direct route (dual-route model)

directly match the visual input with its lexical entry and get the correct pronunciation that way

• works for irregular words 

connectionist model

single route

• 3 units (orthography, semantics, phonolgy)

• any unit can affect any other unit

consider the prounication of the spelling pattern “-ave” in english (e.g. gave, pave, wave, behave). according to a connectionist model of word reading, the word “have” would 

have weak connections between orthographic and phonological units

compositionality

meaning of a whole sentence is composed out of the meanings of its parts and the structure of the sentence

• predictability of meaning from the way the parts are put together 

constitutent

syntactic category consisting of a word or a group of words that clump together and functionas a single unit within a sentence 

knowledge of structure is …

• generative

• hierarchical

• allow for recursion

one-word(holophrastic) stage (12-18 months)

use of a single word to express a complex idea

• “food” → give me food)

• “up” → pick me up

two-word stage (18-24 months)

use two word combinations to from simple phrases

• “more milk”

• “no cookie”