Brainscape's Knowledge GenomeTM


Top Flashcards (Certified)
All Flashcards

language and thought > midterm 3 > Flashcards

midterm 3 Flashcards

(58 cards)

Start Studying
1
Q

who are the experts at language learning

A
  • children
  • learnd + innate
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Timeline of language devlopment (1-4yrs)

A
  • 1 month
    -8 months
  • 1 yr
  • 2 yr
  • 3 yrs
  • 4 yrs
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Nature vs Nurture

A

Skinner (nurture)
- children are a blank slate

Chomsky (nature)
- language is innate, it is the mirror of the mind
- language labels what we are already know, so infants have conceptual categories before words

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

dimensions of developmental constraints (frank keil)

A
  • innate
  • acquired
  • domain specific (vision, language)
  • domain general (memory, attention)

innate + domain specific - chomsky
acquired + domain general (skinner)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

categorical perception vs statiscal learning

A

CP:
- babies hear phonemic contrasts
SL:
- babies track statistics of perceptual input
- babies track probabilites between sounds to detect patterns like word boundaries

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

categorical + voicing

A

place of articulation:
- pa, ta, da, ga

voicing:
- pa, ba

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

high amplitude sucking procedure

A
  • studying infant perception
  • infant given pacifier that plays sound when they suck
  • increase = noticed change
  • decrease = habituation
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Eimas (1971)

A
  • infants cam distinguish sounds from different categories w/ 40 ms voice onset
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

phonemic contrast + language exposure: Miyawaki et al.c 1975

A
  • language shapes perception
  • english speakers can distingusih /ra/ and /la/
  • japanese speakers can’t distinguish /ra/ and /la/ because japanese lacks this contrast
  • adults are primed to native language that they will not be able to recognize phoneme contrats in other languages
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Werker & Tees (1984) - conditioned head turn techniques

A
  • hindi and english speaking infants
  • perception of non-native phonemic contrasts by english-learning infants
  • 6-8 months can detect contrasts, after 12 months losses ability
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

selectionist model

A
  • we are born with ability to distinguish all phonemes
  • language input selects + preserves only what’s relevant for the language we grow up speaking
  • a process of LOSS not gain
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

chinchilla study: Kuhl + Miller (1975)

A
  • trained chinchillas to hear differences between /ba/ and /pa/
  • chinchillas are able to distinguish like human infants and adults
  • phoneme contrats may not be language-specific but auditory processing
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Magnet theory Kuhl (2000)

A

infants perceptual space starts universally but is warped by language experience

“Language magnets” are drawn to perception and shape it

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Adult perception and magnet effect kul & version (1996)

A

/ra/ and /la/ are categorized similarly and will be perceived more similarly than they actually are

native English speaking are less sensitive to distinguishing those phonemic differences

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Statistical Learning

A
  • word segmentation problem
  • infants cant tell when a word ends/next one begins
  • they rely on TRANSITIONAL PROBABILTIES
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Transitional probabilities (Saffran, newport, aslin, 1996) AI LANGUAGE STUDY

A

“pretty baby”
- tabidogowamilani
- pairs went together, other pairs did not

infants under 6 months could discriminate high prob pairs from part pairs after 2.5 minutes of exposure

stress patterns helped

ACQUIRED DOMAIN GENERAL

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

Transitional probability

A

likelihood that 1 syllable follows another

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

harmony bias study

A
  • adults learn 1/5 AI languages
  • either harmonic vs non-harmonic
  • adults prefer the harmonic patterns
  • children prefer prefer for harmonic patters because they like rule-bases systems, AKA consistency
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

Word knowledge

A

concept (meaning)
lexeme (pronunciation)
lemma (grammatical properties)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

John Locke (1960)

A

we show children an object then say the name, leading the child make a connection and link the two

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

Benjamin Whorf (1956)

A

the word is unstructured “kaleidoscopic flux”

Language determines thought and shapes infants’ perceptions

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

Piaget theory of development

He believed that infants do not understand objects existing indepently from their own actions

A

scheme: pattern of interacting with the environment, knowledge + behavior

  1. assimaltion
  2. adaption
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

Piaget stages of development

A
  • sensorimotor (0-2yrs)
  • preoperational (2-7yrs)
  • concrete operaional (7-12yrs)
  • formal operational (12+yrs)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

Piaget view on object knowledge

Study These Flashcards
A

0-4 months
4-9 months
9-12 months

25
A-notB error
object is hidden behind A multiple times then behind B
26
Baillargeon's object permanence study (1985,1987)
- rotating screen with box behind - either screen stops or moves through box
27
Liz Spelke: infants are born with innate principles
1. continuity 2. coherence 3. contact 4. gravity 5. intertia
28
Spelke, Kestenbaun + simmons (1993) moving rod study
4 month olds are shown a rod moving behind box - boring (continuous) - impossible (discontinuous)
29
Quinn Eimas and Rosenkrantz (1993) animal novelty study
- 3-4 month olds look at novel itemd - Different animals are shown and tested on their preference for new or familiar categories - Infants are mapping words to existing concepts
30
Word Learning Timeline Bergelson $ Swingley 2012
Infants 6–9 months can link words like “banana” or “milk” to objects At 18–20 months, mapping improves dramatically Conclusion: Word understanding begins earlier than expected
31
Semantic Uncertainty
- "vash" has too mnay referentials, semantic uncertainty to identify
32
Children as Intention Readers (Baldwin, 1993)
8-month-olds use adult gaze and body posture to disambiguate reference Joint attention is critical: 16-month-olds succeed only when adult is attending to the same object 18-month-olds succeed even when object is hidden (e.g., in bucket)
33
Mutual exclusivity
children assume that objects have one lable - dog and mipen Child might narrow down meaning based on what they already know Learning is guided by a mix of biases and logical inference
34
whole object bias (markman 1991/hanson + markman 2009)
words refer to whole objects
35
shape bias (landau, smith + jones, 1988)
Children generalize by shape, not color or texture
36
Gem vs junk Gillette et al. (1999), Medina et al. (2011), Trueswell et al. (2016)
Most is "junk": vague, ambiguous Some is "gems": precise and clear Gems (~15%): Clear referents, well-timed cues Junk (~60%): Many referents, few cues Intermediate (~25%): Mixed clarity
37
Human Simulation Paradigm
naive adult muted video of mother saying common words hear beep and if adults cant understand, than infants probably cant tell either
38
Global vs local learner
Global Learner: Stores all possible meanings from every situation. Later intersects them to find the word's true meaning. Local Learner: Makes a guess right away based on current info. Updates if necessary on future encounters.
39
Global Learning Model (Siskind, 1996)
In every situation: Store all possible meanings. Later: Find the overlap across situations.
40
local learning model
When hearing "cat," pick the most likely guess based on the current situation. If it fits in the future, keep it. If it doesn’t fit, revise your guess.
41
Yu & smith study (2007): global learning
Study Setup: Adults learned words for novel objects in ambiguous settings: 2 objects + 2 words, or 3 objects + 3 words. Test: Click on the correct object for a given word. Chance performance = 25%.
42
human stimulation paradigm: instead of beep, it is a nonsense word
Findings: Adults tended to remember only their previous guess, not the full set of possible referents. Accuracy on guessing depended mainly on whether their first guess had been correct. ✅ People commit to a meaning early and don't track everything.
43
children are local learners: Woodard, Gleitman, & Trueswell (2016)
Study with 2–3 year olds: Children heard a novel label like "mipen" for an unfamiliar object. Even if they made an early wrong guess, they stuck with it next time. Result: Children use a Propose-But-Verify (PbV) strategy: Propose one meaning immediately. If future evidence confirms it, stick. If not, guess again.
44
Dilution Effect & how to avoid
If you stored all possible meanings across all experiences: The meaning pool would get too diluted. Learner would never confidently identify what the word means. By proposing a single hypothesis and updating only when needed, learners stay focused.
45
Blind Individuals and Verb Learning
If vision is crucial, how do blind individuals learn verbs like "see" and "look"? Answer: They learn verbs just as well as sighted people! Study: Bedny et al. (2019) Blind and sighted adults rated verb similarities (e.g., "sparkle" vs. "glow"). Their ratings highly correlated—no major differences.
46
Why Do Children Learn Nouns First?
Conceptual Change Hypothesis: Babies’ minds can't handle complex ideas (like "cause," "see," "put") yet. Learning-from-Observation Hypothesis: Babies have complex ideas. It's just hard to see verbs happening.
47
Snedeker et al. (2007) – Adopted Children Study
Compared: Infants learning English from birth. International adoptees (6-year-olds) learning English after adoption. Finding: Both groups: First learn nouns, then verbs. Noun-first pattern is not about age — it's about learning from observation.
48
The "Information Change" Game
As vocabularies grow: Children start using additional sources of information. Learning shifts from just looking at the world to using: Social context Nouns mentioned Syntax (sentence structure)
49
Bootstrapping
Bootstrapping = Using partial knowledge to gain more knowledge. Syntactic Bootstrapping: Learning grammar helps learn verbs. Learning verbs helps strengthen understanding of grammar.
50
51
Yuan & Fisher (2009) Study - syntatic bootstrapping
24-month-olds listened to: "Jane blicked the baby." "The baby blicked." Babies inferred different actions based on sentence structure alone.
52
Negation
Type of Negation: prohibition, nonexistence & denial "No" and "not" are abstract. Even adults struggle to figure out negation just by looking at scenes.
53
Real-Time Parsing Example - adults
Example sentence: “Put the frog on the napkin into the box.” Adults initially think “on the napkin” tells where to put the frog. When they hear “into the box,” adults quickly realize their mistake and fix their plan.
54
Kindergarten-Path Effect
Young children make early parsing errors and cannot easily revise when new information comes. children use verb info and case making for real time parsing
55
Kindergarten-Path Effect Across Languages (Choi & Trueswell, 2011)
Tested Korean 4–5-year-olds on parsing tasks similar to the frog-on-the-napkin task. Set up ambiguous sentences in Korean. Tracked how children interpreted sentences in real time.
56
Why Do Children Have Trouble Recovering from Garden Paths?
cognitive control
57
evidence for cognitive control/parsing-control connection in children (woodard, pozzan and trueswell - 2015)
Parsing Task: Children listened to ambiguous sentences (like the "frog on the napkin" kind of sentences). Researchers tracked whether children could revise their interpretation once disambiguating information came in. Non-Linguistic Cognitive Control Task (e.g., Flanker Task): Tested children's inhibitory control (ability to suppress automatic responses). In a Flanker Task: Children saw rows of arrows and had to focus on the middle arrow and ignore distracting side arrows. Example: ➡️ ➡️ ➡️ ➡️ ➡️ (all match → easy)
58

language and thought (2 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2025 Bold Learning Solutions. Terms and Conditions