LMB Exam 2

Vocal folds/cords

a pair of loosely attached flaps that vibrate as air passes through them

Minimal Pairs

a pair of words that have different meanings, but all of the same sounds with the exception of one phoneme -tan vs. man

Glide

a sound produced when you obstruct the airflow only mildly, allowing most of it to pass through the mouth -W, and Y

Bilabial

a sound that is produced by obstructing airflow at the lips -"p"

Affricative

a sound that is produced when you combine an oral stop and a fricative together

Liquid sounds

a sound that is produced when you let air escape over both sides of your tongue -l, r *differs depending on tongue placement (front or back of mouth)

Fricative

a sound that is produced when your tongue narrows the airflow in a way that produces a turbulent sound "s" "f" and "z"

Alveolar

a sound whose place of articulation is the alveolar ridge, just behind the teeth -"t"

Velar

a sound whose place of articulation is the velum (the soft tissue at the back of the roof of your mouth -"k"

Over-extension

mapping new words into categories that are too general -Referring to all animals as "doggie"

Under-extension

mapping new words into categories that are too specific -If a child is told a carnation is a flower, then she might not realizes that daisies, roses, etc. can also be referred to as flowers (might not be able to call them anything)

Jean Berko Gleason Study: Wug

-Showed kids a bird and said "this is a wug" -Then showed them two and said to the kids "there are two..." -The kids would respond "wugs" -Showed that they could tag on the right plural morpheme when faced with a new word that they couldn't possibly have heard before

Inflectional affixes

serve as grammatical reflexes or markers, the presence of which is dictated by the grammatical rules of a language -ed and -ing change the tense of a verb Create words from existing ones

Lack of Invariance Problem

speech signal is ambiguous there's no 1-1 mapping between acoustics and phonemes implication-- you can't rely on sounds to understand anything, you must use a combination of your ears and context: top-down information

Telegraphic Speech

speech that preserves the correct order of words in sentences, but drops many of the small function words such as "the, did, or to" **do this by age 2

Phonotactics

statistics of sound sequences in a language ex: the sequence "bk" is rare inside words in english... if you hear "bk" there's probably a word boundary

Whole Object Bias

the (theoretical) assumption by babies that a new word heard in context of a salient object refers to the whole thing and not its parts, color, surface, substance, or the action the object is involved in

Is CP Innate/Unique to Humans?

**Peter Eimas experiment claimed that it was, but they only tested humans, so this could not be proven

Gavagai Problem

- When a tribesman sees a rabbit go by, he screams "Gavagai!" -People might assume this means "rabbit" but could also mean other things (white, furry, running, etc.) **can compare to toddlers-Do they know what parents are talking about when they point things out?

"zav" studies

-Asked children to apply to word "zav" to different things --Told that a Dalmatian is a "zav" --Then asked if other dogs are also "zavs" **Will usually say that the other dogs are not "zavs" -If they provided 3 different examples of a "zav" (i.e. Dalmatian, terrier, and collie), they would be more willing to shift their interpretation of the word to a basic-level category (dog-- when other types of dogs shown, they would say they were "zavs") -On the other hand, if they gave 3 examples of the word "zav" with all Dalmatians, they would be convinced it applied to a subordinate-level category (Dalmatian) **Results show that children don't always match words to the right category the first time they hear them-- they will gather more evidence until they can figure it out

Stress Patterns in Infants

-At 7.5 months- can only distinguish words that have emphasis on first syllable --No trouble slicing words with trochaic pattern from -----Don't recognize iambic words

Jenny Saffran et.al artificial language experiment: results

-Babies showed distinction between words and past-word sequences Listened to past-words longer -Babies kept track of transitional probabilities:

Pros/cons of using different category levels

-Basic level- more commonly used, still provides you enough information in most cases -Subordinate- helps you compare two different kinds (of dogs) -Superordinate-level- hard to tell what you're talking about, need to be more specific in a lot of cases **kids need to learn how to map words into the right categories, tend to make mistakes

How does bilingualism affect individuals?

-Because they carry an extra vocabulary, they experience more competition than monolinguals Increased competition makes language processing slower and harder -Evidence that bilinguals show less efficient word recognition compared with monolinguals -Also experience interference across languages in speaking as well as understanding language, and this is reflected in poorer performance in picture-naming tasks, even when the bilinguals are speaking in their first and dominant tongue -Upside of bilingualism-Build up more cognitive control abilities

Experience and CP

-CP starts very early, but also affected by experience (different cultures) -curves for different cultures are different -- different languages influence speech perception i.e. [p] in spanish/french and [b] in english

Jenny Saffran et.al artificial language experiment: design

-Familiarized 8-month-old infants with unbroken 2 minute strings of flatly intoned, computer-generated speech -Speech followed repeated consonant vowel structure (English rule- vowel comes after consonant, can be grouped in a large number of ways) -Stimuli was represented in a way that represents an artificial language -Each word consisted of 3 consonant-vowel syllables -Uninterrupted, 2 minute sound stream -Only consisted of 4 words- repeated

Katherine Estes Study- Babies attaching words to meanings

-Study with 17-month-old babies -First heard a 2.5 minute stream of artifical language -Immediately after, babies were exposed to a novel word-learning task using the switch paradigm -Half of the new words in the switch task corresponded to word units in the artificial language the babies had just heard -The other half were sequences of sounds that had occurred just as often in the artificial language but that straddled word boundaries (part-words) -Babies were able to learn the associations between pictures and words, but only for the sound units that represented word units in the artificial language, showing that they were applying the results of their segmentation strategies to the problem of mapping sounds to meaning -Results suggest that small children are able to draw on their stores of segmented units to help them in the difficult task of matching words and meanings

Connectionist Computer Model

-Treats verb stems and past-tense forms as bundles of sound that are associatively linked together in memory -Models job is to learn to predict which sounds make up the past-tense form, given a particular verb stem --If you give the machine the word "grind" it should figure out that the past tense is "ground" even though it has never heard that word before **Uses previous statistical data to figure out what the past tense of this verb should be -Beginning- has random answers because it has had minimal exposure -After its been exposed to new words, starts to -**Will recognize that certain consonants appear in present and past tense of categories of irregular verbs **Also have similar sounds/vowels Reflects gradual learning of irregular past-tense forms through a form of statistical analogy

Language Cues and experience

-variation in language experience affects word learning -huge variation in amount of input to kids higher SES--> more input of language-->more known words

Experiment for familiar words hypothesis

-Used head-turn preference paradigm -Showed that 6-month-old would look longer at familiar words than random words in test phase when these words were played next to their own name or "mommy" in familiarization phase -Shows that babies at this age can segment words in some instances

Voicing

-Whether/when vocal folds are vibrating as you utter a consonant -Vocal folds can constrict loosely or tightly -Air comes up through lungs passes through these flaps- if they are loose or tight this creates low or high pitch -Vowels are always produced while vocal folds are vibrating, but consonants can vary --Z,V,D made by vibrating vocal folds -S,F,T are not

Head-Turn Preference Paradigm- Experimental Design

-baby sits on lap of parent who listens to music with headphones (prevents parent from hearing stimuli and giving cues) -video camera records baby's response/unseen observer watches -one loudspeaker is at the front of the room (light is flashed at the front to draw child's attention to a neutral point) and two are on either side of the room **artificial language is created- some syllables are stressed and some are not hypothesis: infants can use word stress to group syllables into possible words/find word boundaries **should react differently to a real word (one with stress) Familiarization: 1. Have infant become familiar with sound 2. Train baby to expect that sounds can come from the speaker on left or right wall 3. Tightly lock together the head-turn behavior to the infant's auditory attention Test phase: -light will flash on one of the sides- baby's head will turn toward the light, and a test word will be played on repeat until the infant looks away -the experimenter will record how long the baby looks

How are consonants made in the vocal tract?

-by pushing air out of lungs, through larynx and the vocal folds (cords) -airflow passing through the vocal tract has to be blocked at some point above the larynx -Location where blockage occurs influences the sound that's made

Learning Theory

-child gets enormous info from environment that constrains word learning ex: joint attention Gavagai problem: word learning is impossibly hard in isolation, but not with lots of other cues

Joint Attention Experiment

-child needs to learn word while also following gaze -can't learn new word if speaker is pointing to another object

Mutual Exclusivity Experiment

-children presented with two objects- an apple and a screwdriver -asked children to point to the "dax" -since they don't know what a screwdriver is, and they know what apples are, they will point to the screwdriver

helping low income families speak more to their children: using LENA word recorders to track parents' and childrens' speech

-children won recording device- detects parents and kids amount of TV watched -2 programs combine recordings with intervention from parents -home visits (expensive) -charts provide feedback to parents (like a fitbit)

Ambiguity in perceiving consonants

-consonants are short and quiet, can't hear them well without vowels -consonants can be identified during the transition between the consonant and the vowel

What do people base guesses on when trying to figure out what is said (lack of invariance problem)

-experience with the language(if its your native language you can guess better) -knowledge of the speaker's accent -visual information (McGurk Effect) -knowledge of context (demo 1)

Is Mutual Exclusivity an effect of experience?

-for children learning multiple languages, objects can have more than one name "ball" vs. "pelota" experiment results -18 months: --monolingual children use ME --bilingual children show marginal use of ME --trilingual children don't use ME

Examples of ways to help parents from low income families speak more to their children

-give books to parents (not that helpful) -pediatrician visits, maternity nurses- can inform parents -telling parents its "nutrition for language"-- changes perspective -harlem childrens zone: good but hugely expensive *provide resources to kids (same that middle class children get)

In Class Activity- brief exposure experiment

-heard a string of made up words -then asked if we heard certain strings of words in the recording (write down yes or no) **to demonstrate how babies can detect regularities in the syllables/phonotactics of what they're hearing

What is CP good for?

-helps you deal with lack of invariance -you perceive continuous variation as discrete categories (i.e. [b] [d] and [g] -improves discrimination between categories -hides variation within a category

Examples of differences in phoneme vs. allophone in other languages (english, korean/japanese, spanish, thai/hindi, mandarin)

-in english [l] and [r] are two different phonemes (if you change the phoneme you change the word) -in korean and Japanese [l] and [r] are two allophones of a single phoneme-- doesn't change the meaning of a word in english [p^h] and [p] are allophones of a single phoneme -in spanish, theres no [p^h], only [p] -in thai/hindi [p^h] and [p] are different phonemes -mandarin 2 phonemic categories (like English) BUT make a meaningful distinction between voiceless aspirated and unaspirated sounds rather than voiced and voiceless ones (unlike English) -Differences between th and /t/ is obvious to them -Differences between /t/ and /d/ is not

CP in Development

-infants born able to distinguish all categories that any language uses **adults are poor at distinguishing categories that their language doesn't use -Babies start with fine detailed perception, then learn to refine, categorize the important ones-- their own language's phoneme contrasts **CP is learning not to pay attention to detail (loss starts at 10-12 months) -babies can form intuitions about where words start and end based on the rules of their native language **mandarin doesn't break up word boundaries in convenient ways-- There must be enough flexibility in their perceptual systems to adapt to the categories as defined by their particular language --Mandarin babies will eventually ignore the distinction between voiced and voiceless sounds

Connectionist Claim

-kids are relating present and past tense in a network -when there are only a few words-> easy to keep them separate in network -as they learn more, begin to get confused and patterns interfere with each other -takes longer to learn the words with rate patterns: these are irregulars

Rule-based claim

-kids start out memorizing patterns for plural/past tense -kids learn rules for plural and past tense -kids apply the rules everywhere -kids eventually re-learn exceptions

Janet Werker Experiment- English speaking babies hearing Thompson Salish

-language has a contrast between velar and uvular ejective [k] vs. [q] -used conditioned head turn: baby trained that change in sound means that toys on the side will light up -baby hears many [k]'s and then a [q] -baby will turn head when they hear the q -only infants can do this- old people can't distinguish the difference

Why not memorize 500+ coarticulation patterns?

-many patterns sound very similar -the patterns vary with where the phoneme in the word is (p vs. p^h) -patterns vary with rate of speech (fast speaker--vowels get compressed more) -patterns vary with the speaker (accent, rate of speech, etc.)

McGurk Effect

-mismatch between visual and auditory information -shows how (visual) context can influence speech perception video: lips making different movement than the sound being made: changes perception sound=ba vision= ga brain thinks the man is saying "da" **this is because the lips make the sound "b" and the tongue goes back to make the sound "g" so the brain averages the sounds-- "d" is made by both the tongue and the teeth

Babies and Language

-need to figure out different units of words- what sounds indicate beginning/end of words in sentences that are spoken to them by their parents **Spoken speech doesn't have clear breaks between words like written language does

Artificial language experiments

-use an artificial language: made up language with a small amount of words -words are repeated syllable sequences like "dah-bu-zee" or "tee-bo" -the repetitions create the statistics: babies learn that "da" goes with "bu-zee" and "tee" goes with "bo" -let infant hear speech long enough to detect patterns -test whether infant differentiates between "words"- repeated syllable sequences and "non-words"-- non-repeated sequences of syllables

Speech Perception (brain vs. reality)

-perception always different from reality -what you're consciously aware of is only a fraction of what's actually in the world and what's actually going through your brain -system is designed to hide things from your conscious awareness -as a result- can't use intuition about how these processes work: have to use experiments

helping low income families speak more to their children: texting parents of 3-5 year olds

-pilot study in san fran: texts in english, spanish, chinese were sent 3 texts a week -1 fact -1 tup -1 goal Results: -parents reported they talked to their children more compared to control could have some bias-- not accurate -based on self report of parents -didn't actually look at the results of the children's language improvement

Hypothesis 1: Non-Speech Sounds- Violin Strings Experiment

-plucked vs. bowed violin strings (short vs. long tone onset times--TOT) -vary in TOT on continuum -tested people in an experiment--showed CP shows that CP can't be unique to speech

Marc Hauser study: monkeys and speech segmentation

-replicated Saffran's study -Showed that monkeys turned their head more often when they heard a part-word than when they heard a word

Strategies for word segmentation

-start at the ends of utterances, where there are pauses, and work inwards -look for stressed syllables (louder)-- there's often (but not always) a boundary before a stressed syllable -use phonotactics

Which view is right?

-still a source of debate -if you'e learning rules and exceptions, it's not really clear how exceptions could interfere with rules -if you're learning the statistics throughout, then common exceptions should interfere with regulars

Hypothesis 2: Categorical Perception in Animals- Chinchilla Experiment

-test if an animal responds categorically to VOT Continuum -found an animal that has human-like hearing: chinchillas Experiment: Patricia Kuhl and James Miller (1971) Chinchilla Experiment -Chinchilla in cage with loudspeaker on top -get food reward on left side for [ba] -avoid shock on right after [pa] -see which way chinchilla will run for different VOT's- 0,10,20,30,40,50, and 60 ms VOT -found that chinchillas also perceived consonants categorically along VOT boundary similar to humans

Word Segmentation/Guessing

-there are no spaces between spoken words- separated by consonants like p, t, k, rather than silence between words -its a guessing game: need to divide up acoustic signal at most likely points for word boundaries (certain songs can have confusing lyrics)

Innate Knowledge Theory

-whole object bias -mutual exclusivity -child uses gestures, lots of input to make educated guess -joint attention constraints the possibilities

Segmentation and experience

-words you know -stress patterns in your language -phonotactics of your language **very hard for a non-native speaker

Two possibilities for why its hard to distinguish between aspirated and unaspirated sounds

1. Differences at some points in the VOT continuum are inherently easier to hear than distinctions at other points --Between 10 and 30 ms is easier than between 30 and 60 ms 2. Our perceptual system has become tuned to sound distinctions differently, depending on whether those distinction are allophonic or phonemic in nature --What we hear doesn't only have to do with objective acoustic differences between sounds, but also by the role that the sounds play within the language system

Two views on over-regularization

1. rule-based-- think that there's two different parts of the brain: words/rules and lexicon think that children memorize rules for regular words and irregular words **ad -ed for past tense, add -s for plural lexicon: exceptions that don't follow the rules (mice, went, etc.) 2. connectionist -not two separate parts -all in a complex network -relate knowledge of singular/present tense to plurals/past tense *nodes connect the two neurons and they're all connected (not two different areas) -think that rules aren't stored

When do kids master syntax of language?

2.5: form full sentences, but they're still simple Kindergarten- master syntax of language **Kids never violate basic word pattern rules i.e. subject in front of a verb

Bilingualism and Cognitive abilities experiment with A-not-B test

Agnes Kovacs and Jacques Mehler (2009) -Tested 7 month old babies being raised in monolingual and bilingual households -Used A-not-B test --See object hidden in location, watched as its moved to a different location -Need to find object in new location- suppress the knowledge of the old location -Can only reliable accomplish this at 18 months Used simplified version- looked at one side of a screen and then had to learn to look at the other side --Rewarded with a treat Babies from bilingual households- better at redirecting attention to other side -Monolingual babies more likely to fail at the task

Peter Eimas habituation experiment

Allows researchers to figure out if babies can distinguish new sounds from ones they already know -babies suck on pacifier connected to computer to test how much they sucked (pressure sensitive meter) **infants as young as 1 month tested -experimenters play a sound repeatedly until baby gets bored of the sound (sucking less) -experimenters then play new sound and see if the baby starts sucking faster (if they do, this indicates tha they heard the distinction phonemic change= different stimulus from different phonemic categories (e.g. 20 ms to 40 ms VOT) acoustic change=different stimulus, from same adult phoneme category (e.g. 20 ms to 0 ms VOT) Control= no change in stimulus

Why don't babies' representations of meaningful words contain all the phonetic detail of the strings of sounds that they segment from speech? Second Idea

Babies are prone to error when trying to retrieve words from memory -Confuse words that are similar to each other --Still have knowledge of words that sound alike, but have trouble remembering which is which **children vulnerable to sound-based confusion --Even adults struggle with this Katherine White study- Adults given made up names for geometric shapes, got confused children vulnerable to sound-based confusion

Artificial Language

Constructed to have certain specific properties for the purpose of testing an experimental hypothesis: strings of sounds correspond to "words" which may or may not have meaning, and whose combination may or may not be constrained by syntactic rules

David Pisoni TOT Experiment/ABX Discrimination Task

Created stimuli using two distinct tones and varying the number of milliseconds that elapsed between the onsets of the two tones-- tone onset time (TOT) ABX Discrimination Task: a test procedure in which subjects hear two different stimuli followed by a third which is identical to one of the first two. The subjects must then decide whether the third stimulus is the same as the first or the second -Play sounds that are 20ms apart, then play sounds that are 30 ms apart, then play one that's 20ms apart again- see if they can match it to the correct one Results: found that people are good at distinguishing between TOT at 20 ms-- similar to VOT -can distinguish between 20 and 30 but can't between 40 and 50 or 0 and 10 **at 20ms people can distinguish between two different sounds

Aspiration with English speakers

English speakers unconsciously know aspiration rule -very hard to avoid following the rule -what gives people foreign accents

Hypothesis 3: Categorical Perception for Visual Stimuli, Experiment

Experiment: used morphing of images of faces (has intermediate steps) -face perception -continuum of facial features -morphed between pairs of famous faces: (JFK/Clinton, Pete Townsend/Sly Stallone) -pictures morphed in 10% increments -participants had to guess identity based on image presented Results: -show categorical perception

Word-break experiments with infants

Familiarization phase: heard sentences that contained a target word - "bike" Test phase: saw how long infants were interested in the word "bike" compared to other words that they didn't hear during familiarization phase Results -7.5 month old babies looked longer at "bike" than other words --Shows that they learned there was a distinction -6 month old babies didn't seem to have this ability

Whole Object Bias with Gavagai Debate

Gavagai=Rabbit -not ears, hopping, etc. -children name the whole object first, and name the parts of it after

Data testing critical/sensitive period hypothesis: Ferel Children

Genie -abused, tied to a chair, punished for vocalizing, no normal interaction, no cognitive stimulation -was discovered and removed from parents at age 13 when found: -could learn words but not grammar -left hemisphere was not stimulated during childhood so it got smaller/reprogrammed for other things -after 3 years in English environment, phoneme perception was good (could distinguish between lamb and lamp) and rhyme perception was good (could point to picture that rhymed with ball) -also good at meaning: hot/cold, big/little BUT never progressed complex sentences, and omitted grammatical morphemes

Whole object bias experiment with part/whole

George Hollich et. Al-- babies have a sense of what kinds of things words attach to Experiment phase -Experimenter will hold an object in front of a baby that has 2 parts (one boring and one interesting) that detach and reattach -Called the object a made up name while separating the two pieces repeatedly Test phase -Put the whole object and then one of the pieces next to each other. Asked baby to look at the (made up word) Results -Babies looked at the whole object rather than just the piece of it **to make sure the babies weren't looking longer at the whole object because it was more interesting, they tested for this **Found that babies looked at the whole object and just the part of the object for the same amount of time

Why phonemic differences can make learning a new language difficult

Hard for new English speakers to tell the difference between two consonants if there's no phonemic difference for them in their native language --shows that a lot of mental structures go into it

Hypothesis on familiar words with infants

Infants can break long streams of sound into familiar units

Example of babies and phonotactics

baby hears- doyouwantyourbottle/teddy/duck *hear the word boundary of b, t, d

Morphology

It analyzes the structure of words and parts of words, such as stems, root words, prefixes, and suffixes. "sebbing"-- know its a verb because of the "ing"

Bilingualism and Cognitive abilities experiment: "Simon Task"

Measured the performance of adults 30-90 on cognitive interference task known as the "Simon task" -Red or green square pops up on a screen, people ave to press a button on the right to indicate a green square or a button on the left to indicate a red square -Sometimes the square is on the opposite side of the button : Have to ignore squares location and still press the correct button -- respond more slowly when this happens, shows interference Results: -Bilinguals outperform monolinguals at all ages -Difference more pronounced in their 60s

Subordinate-level categories

More specific categories comprising words that encompass a narrow range of referents -Dalmatian

Data testing critical/sensitive period hypothesis: critical period and American Sign Language (ASL)

Newport Study research is trying to distinguish time and experience Subjects: -adult signers of ASL -all deaf from birth -age of first exposure to ASL: birth-20 -all signers using ASL for 30 years at a time for testing Research Question: does age of acquisition matter? Results: -got less answers correct when acquisition was later in life -shows possibility of critical period Problems with experiment: -children who acquire ASL at early age have way more learning experience/exposure to the language-- practicing it more often than older adults

Why is compositionality important?

Non-compositionality would seriously hinder the communicative usefulness of sticking words together into longer units -Too complicated to try to figure out what a person is trying to express --i.e. "red dax"-- don't know what a dax is, why is it red, etc. -Can't predictably compute meaning

Why don't babies' representations of meaningful words contain all the phonetic detail of the strings of sounds that they segment from speech? First Idea

One idea- difference between info that babies can pay attention to while processing language in the here and now, and the info they commit to long-term memory in the form of a stable lexical representation (info committed to long-term memory about the sound and meaning properties of words, and certain constraints on their syntactic combination) -Babies only remember sounds that they need to distinguish words from one another --Would need a lot of sound detail to distinguish thousands of words in their vocabulary that they will have as adults --Doesn't need to know the distinction between /b/ and /d/ at first, but will later learn that this is a helpful distinction

other rule-based view

Other researchers think children could be memorizing even the regular forms and then extending the past-tense forms to new verbs by analogy

Bilingualism and Alzheimer's

Patients have loss of cognitive control Bilingualism offers some protection against this-- delays progression of symptoms

Whole Object Bias Experiment with sticks

Philip Kellman and Elizabeth Spelke Study (1983) -3 month old babies shown two sticks poking out of the top and the bottom of a screen -If they moved together, the babies thought that the sticks were attached- would be surprised when the screen was removed and they saw two different objects (would stare at it longer) -Weren't surprised if the objects moved separately -Also know: objects can't disappear, can't pass through each other, can't move unless they come into contact with other objects

Juan Toro and Josep Trobalon study: rats and speech segmentation

Rats don't have any communication skills- rules out if speech segmentation is related to language Found that they were able to use statistical regularities to differentiate between words and non-words Shows that speech segmentation is a cognitive skill not related to language

Michael Ullman Study-- rule based evidence (brain structure

Regular and irregular complex words reflect same distinction as the one between declarative and procedural memory -Ventral system- memorizing words -Dorsal system- "how to" knowledge -Temporal lobe- accessing memorized info -Frontal regions- grammatical tasks Alzheimer's/Parkinson's patients -have extensive degeneration of temporal region, frontal region and basal ganglia are fine -Patients can still produce grammatical speech but can't retrieve info -opposite for Parkinson's ERP Study -brain wave activity different for different tasks

VOT graph- results

Results show categorical perception because several VOT values are all treated the same, then there is a very sharp shift in responses from few P responses to many P responses.

Difference in Pitch

Some languages: makes a phonemic difference- can change the meaning of the word --High or low, rises then falls or falls then rises, etc.

Data testing critical/sensitive period hypothesis: Johnson and NewportCritical periods in second language learning

Subjects: speakers of East Asian languages, non-native speakers of english -all lived in US 5 years at time of test -all similar social backgrounds at time of test -varied in age they had english exposure Results -score declined with people who arrived to the US later in life --started to decline around 8-10

Differences in brain structure depending on the language a person speaks

Tonal differences on words- more left hemisphere activity with mandarin speakers, more right hemisphere activity with English speakers

Vocal Tract/Acoustics

Vibrations produce sound waves that are shaped into different speech sounds by the rest of the vocal tract -different positions of your vocal tract change the acoustic signals that are perceived by your ear -acoustic signal--> understanding

When is P aspirated?

When p is the first sound of the syllable it is aspirated **no aspiration for "spot" or "spit"

What would graph look like without CP

Without categorical perception, people would gradually perceive more and more "p"-ness in the stimuli, and the responses would be on a diagonal straight line starting at about zero "P" responses on the left (at 0ms VOT) and going up with increasing VOT

Phonemic inventory

a list of different phonemes in a language

Forced choice identification task

a way to test for categorical perception -Have people listen to many examples of speech sounds and indicate which one of the two categories each sound represents ("pa" vs. "ba") -Speech sounds created in a way that varies in the VOT in small increments -Extreme ends (-20 and 60 ms)- people are in a agreement over what the sound it -In the middle- unsure of what the sound is -At a certain VOT, people's perception of the sound will change --at that point, it is clear to the person what they are hearing --after that point, more unsure

Peter Eimas habituation experiment: Results- Categorical Perception or Continuous?

acoustic, phonemic and control showed that they noticed a change CP: If they have categorical perception, they should be oblivious to differences between certain sounds but acutely sensitive to differences between other sounds that fall on different sides of a critical boundary Continuous: If they perceive continuously, they should always hear sound B as different, and increase sucking any time sound B is introduced **evidence supports CP: -rate of sucking goes up when two sounds straddle VOT boundary of ~25 ms -otherwise, they're oblivious to VOT

Infixes

affixes "shoehorned" into the middle of a word Only in some languages, NOT IN ENGLISH

Suffixes

affixes attached at the end of a word -able, -ed, -ing

Prefixes

affixes attached at the front of a word Un-, pre-

Derivational affixes

affixes that transform a word of one category into a word of a different category or significantly change the meaning of the word Own + er= owner, pre + view = preview

Manner of Articulation

airflow can be stopped partially or completely

Mutual Exclusivity

an object can only have one label **child pays attention to patterns (of naming)-- keeps track of these statistics

Unaspirated stop

an unvoiced oral stop without aspiration, produced with a relatively short VOT

Duality of Patterning

concept that language works at two general levels, with units of sound combining into meaningful units (usually words) and these meaningful units combine into a larger pattern of meaningful syntactic units **children need to learn sounds, words, and eventually sentences

Brief Exposure and Infants

babies can detect regularities in the syllables/phonotactics of what they're hearing -can use phonotactic info to detect patterns at 7 months -foundation to word learning -not just unique to speech-- can also use for face recognition and music

Plural Morphemes

can only attach the plural morpheme to the end of something that is a word in its own right ** When certain words are paired with other words, they can't be plural anymore i.e. "lifeboat salesgirl" can not turn into "lifeboats salesgirl"

Is Mutual Exclusivity uniquely human?

can test in 2 ways: -mutual exclusivity in non-humans -ask if humans learn that objects have 1 name answer: no-- experiment with dog: -dog had names for all of its toys -put a toy it didnt know and asked for it (random name) -dog found the toy by deduction

why is Genie studies bad science

cant answer questions like-- -is this evidence of a specifically linguistic critical period or general brain development? -is this evidence of biologically timed change in brain or a change in brain plasticity through experience? **lack of experience is itself a form of experience that changes brain development

Revised Hypothesis 2/Testing with Crickets

categorical perception if part of mammal auditory system tested this hypothesis with different types of animals/non-mammals: Crickets -invertebrates -rely on instinctual processing of sounds -chirp frequencies are meaningful and important for crickets --10kHz chirps indicate a possible mate --40 Hz chirps indicate a threat -designed experiment with loudspeaker- crickets either hope toward or away from speaker depending on sound results: -show that they do have categorical perception (similar results to human experiments) new hypothesis- any species with hearing has CP

Aspiration and Accents

certain letters are aspirated in some languages, but not in others. examples: spanish vs. english 1. spanish word "peso" should not be aspirated, but an english speaker trying to attempt this word may pronounce it wrong because they would expect it to be aspirated 2. "beso" and "peso" are both not aspirated in spanish **an accent helps produce these aspirations

U shaped learning

children learn in stages of regularization (chart in U shape for % correct words said on y axis and number of words learned on x axis) at the beginning= children use correct forms in the middle= make errors later on= eventually correct again (go/went vs. goed)

Learning words vs. learning grammar (chomsky vs. skinner)

chomsky: innate universal grammar skinner: all stimulus-response

Head-Turn Preference Paradigm (what is it)

compares how long babies keep their heads turned toward different stimuli **6-7 months: swivel head to show that they are paying attention to something -looking time: turn their head toward something they find interesting Math= TP- frequency (XY)/frequency (X) **TP of Saffran Study -TP of syllable pairs within words: 1.0 -TP of syllable pairs across word boundaries: 0.33

Unvoiced/Voiceless

described a sound that does not involve simultaneous vibration of the vocal folds; in a voiceless stop followed by a vowel, vibration happens only after a lag (more than 20 milliseconds)

What causes critical/sensitive periods?

different possibilities: -change in motivations or emotional response? (i.e. duck imprinting) -growth in memory or cognitive abilities? (i.e. less is more hypothesis-- smaller memory is beneficial, lets you focus on more important things) -change in neural circuits through experience? --the learning itself changes brain circuits/plasticity --learning one system reduces ability to learn another --if so, not just passage of time, but amount of experience

Noun-noun compounds and Semantics

dont work in the same way as sentences -Houseboat is a boat that's also a house -But a housewife is not a wife that's also a house Doesn't fully follow a semantic pattern, but has some common semantic relations= Some have a part-whole relationship i.e. Computer screen, care engine Second noun is a thing for the benefit of the first i.e. Baby carriage, dog bed, cat toy *these types of words need to be memorized-- no rule for them (follows the view on the debate the language is memorized)

Long Vowels vs. Short Vowels

english: an be used to create emphasis "no way" vs. "nooo waaay" other languages: can change the whole meaning of the word

Over-regularization is resistant to correction

even if an adult corrects a child, they may understand what the adult is saying when they use the correct form, but they resist to use the correct form

Compounding

gluing together two independent words into one unit so that the new unit acts as a single word

Case

grammatical markers that signal the grammatical role (subject, direct object, indirect object, etc.) of a noun within a given sentence Types of endings= nominative, accusative, dative **different languages use different endings/markers for words-- changes meaning

Top-Down Information

how our brains make use of information that has already been brought into the brain by one or more of the sensory systems.

Combining visual and language cues (picture with bowl/confetti example)

if a baby looks at a picture of a person mixing confetti in a bowl, certain cues can be given to the child to try to figure out what the word is (noun or verb) -"look! a seb!" would be the bowl -"look!" "seb!" would be confetti -"look!" "sebbing!" would be the action of mixing

nasal stop

if you lower the velum (soft tissue at the back of the roof of your mouth) in a way that lets the air pass through your nose "n" "m" and "ng"

Productivity

in linguistics, a process that can be applied very broadly to a large set of lexical items, rather than being restricted to a small set of words -McDonalds language example

Analogy/Rule-based view

in regard to forming complex words, a process of comparison in which similarities between members of pairs or sets of word forms are taken as a basis for the creation of another word form -Hen is to rooster as aunt is to... uncle -Slip is to slipped as wrap is to... wrapped **Difference in sounds= difference in meaning **Apply same sound difference to the next word can also be applied to irregular verbs ring is to rang as sing is to sang Problem-- will extend these rules to some irregular verbs that don't use same endings bring--> brang, shook--> shuck

Theory: Infants and Word Segmentation

infants have an innate ability to store language statistics -even with limited language experience, babies start finding word boundaries

Two Theories with word learning

innate knowledge vs. learning

Phonotactic Constraints

language specific constraints that determine how the sounds of a given language may be combined to form words or syllables -Different rules for different languages- how consonants and vowels can be strung together *at 9 months- babies have some knowledge of the templates for proper words in their language --experiments with head turn preference have been done: Look for different amounts of times at English words and Dutch words- show that they know the difference

transitional probabilities

likelihood that one particular syllable will be followed by another specific syllable -Easy to predict the second syllable of a word if you know the first syllable --Only likely to predict the second syllable if they're within the same word --If they are two different words, it is hard to predict

Affixes

linguistic units that can't stand on their own but have predictable meanings when attached to a stem morpheme

30-million word gap

more exposure=better learning -repetition helps kids solidify sounds and meanings Kids from poorest families generally heard much less talk than children from more economically privileged homes -lower income kids- 30 million words fewer than kids of higher class

Is statistical learning a specialized human skill for language?

no- animals can also segment speech

Is CP Unique to Speech?

no: non speech sounds are also perceived categorically Hypotheses if CP is not special: 1. humans have CP for non-speech sound 2. animals have CP 3. we have CP for visual stimuli

Critical Period Hypothesis

normal language acquisition must begin in a certain time window or it will never be "native" -not just "earlier is better." claim is that there's a biological phenomenon, like puberty, that shuts off language acquisition after a certain time point -term "sensitive period" is more neutral about possible reason why earlier is better

Nouns/Adjectives and Semantics

nouns and adjectives follow a semantic pattern: any time you join an adjective with a noun, the adjective serves the purpose of identifying a property that the noun possesses-- fully compositional

Stop consonant:2 types

oral stop nasal stop

Forced choice identification task: Class Activity

played "pa/ba" sounds at different VOTs. Had to guess whether the speaker was saying "ba" or "pa" chart- the percent of "pa" increased as the VOT increased shows that more VOT= "pa" sound (because "ba" has 0 VOT)

Hearing speech out of context demo (activity in class)

played a bunch of words without seeing the person say the words results -the class mainly understood the nouns -shows that sound isn't sufficient for understanding acoustic information without context (only 40% would get it right when in a quiet room) -more evidence for lack of invariance problem -shows importance of top-down information

Jenny Saffran et.al artificial language experiment: test phase

researchers noted how much attention the babies paid to actual "words" they heard in the familiarization phase and compared it with the attention the babies paid to three-syllable sequences that also occurred in the speech but that straddled "word" boundaries

Trochaic Stress Pattern

syllable emphases pattern in which the first syllable is stressed

Iambic stress pattern

syllable emphasis pattern in which the first syllable is unstressed

Coarticulation

the articulation of two or more speech sounds together, so that one influences the other. every transition depends on which consonant and which vowel is said **the mouth is already prepared to say the rest of a word after the first sound it made example: kee vs. coo **no sound to indicate the k phoneme (difference between k and c)

Compositionality

the concept that there are fixed rules for combining units of language in terms of their form that result in fixed meaning relationships between the words that are joined together -The way the parts of a sentence are put together makes it possible to predict the meaning **Can put different subjects or different verbs into the structure of a sentence i.e. sally punched danny is different than danny punched sally

Lack of Invariance

the fact that the speech signal is not definitive about what phonemes you're hearing

Basic Level Categories

the favored midlevel category of words that strike a balance between similarity among members of the category and distinctiveness from members of other categories **You use the word dog more than Dalmatian (specific) and mammal/animal (broad)

Voice Onset Time (VOT)

the length of time between the point when a stop consonant is released and the point when voicing begins "ba"--> 0 ms "pa"--> 60 ms

Semantics

the meaning of a sentence; the system of rules for interpreting the meaning of a sentence based on its structure -You can combine the same assortment of words and create new meaning

Superordinate-level categories

the more general categories of words that encompass a wide range of referents -Animal, mammal

Aspiration

the puff of air released in speech: the letter "p" can sound different depending on if is aspirated or not "p" is aspirated sometimes

Morpheme

the smallest bundles of sound that can be related to some systematic meaning **Children learn meanings of morphemes, not words

Syntax

the structure of a sentence- a set of rules or constraints for how the units of language can be put together

Allophones

two or more similar sounds that are variants of the same phoneme-- when sound differences don't fundamentally change the identity of a speech unit example: -changes in loudness -aspiration

Phoneme

units of speech that if changed, changes the meaning of the word "pat" vs. "bat" --"p" and "b" are different phonemes

Aspirated Stops

unvoiced oral stops with a longer VOT

Switch Task

used for studying how children map sounds to meaning Tested whether children pay attention to fine details of sound in learning new object-word mappings Habituation Phase -two objects are each paired up with a novel word that is not an actual English word but obeys its Phonotactic rules (would make sense being a word- not random letters) -To learn associations, babies watch pictures of objects one at a time, accompanied by their associated labels spoken though a loudspeaker -Trials repeated until babies show signs of habituation (become familiar with the stimuli- looking time declines) Test Phase -Labels that go with objects (during the habituation phase) are swapped -Compare this to the first trial Results: -If baby links the correct label to the picture of each object, they should register surprise when the wrong label is paired with the familiar picture --Should look longer at "switched" trials -See this in babies 14 months of age or older --Shortly after babies say first words -Don't notice the switch if words are very similar to each other --"bih" and "dih"-- think they're the same -Can recognize different between /b/ and /d/ in categorical perception, but can't recognize difference when using these sounds for meanings of words

30-million word project

used word recorders that provided feedback results: improvement in amount of words spoken problem with data- -small sample -at first showed improvement but eventually the effect of intervention died out after 4 months (like how people get bored of the fitbit)

Categorical Perception

what you know about the sound structures of your language imposes sharp boundaries, so you categorize sounds that fall within a single phoneme category (even if they're different in various ways) as the same, whereas sounds hat straddle phoneme category boundaries clearly sound different **divides stimulus into discrete categories (no longer continuous points that are equally perceptible

Oral Stop

when the air is fully blocked in the mouth and not allowed to leak out through the nose "p" "t" and "k"

Voiced Oral Stop

when vibration of vocal folds happens simultaneously with the release of articulators (~20 milliseconds) --B in the word Ban

What makes something a word?

when you start building structures out of linguistic units, it can be slotted into those particular spaces that are reserved for words -Can be a combination words like "lifeboat salesgirl"