Exam 2 Study Guide
Exam 2 Study Guide PSY/LING 34
Popular in Language Development
Popular in Linguistics and Speech Pathology
This 65 page Study Guide was uploaded by Texana Sonnefeld on Sunday March 8, 2015. The Study Guide belongs to PSY/LING 34 at University of Arizona taught by Louann Gerken in Spring2015. Since its upload, it has received 120 views. For similar materials see Language Development in Linguistics and Speech Pathology at University of Arizona.
Reviews for Exam 2 Study Guide
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 03/08/15
Exam 2 Study Guide Lecture 5 Lexical segmentation 4 multipleselect questions Finding words in continuous stream of speech Figure out what words mean Given a cue for mapping Cues to lexical segmentation don t need exposure to know how to use them and Cues you do need exposure to learn how to use from language Lecture 6 Lexical mapping 3 multipleselect questions and 1 short answer graph Shape bias graph on quiz as child vocab increases the shape bias also increases when they know about 50 words they increase suggests that shape bias is learned shape color 03975 texture 05 025 0 025 2650 51 1 00 nouns in vocabulary Lecture 7 Morphosyntax 6 questions 3 multipleselect questions 1 multiple choice and 2 matching questions count morphemes sentence types Syntactic constituents What s true How we nd themwhat they are Clause vs phrase clause can stand on its own subject verb there are prosodic markers at the end of phrases and clauses end of clause marker are stronger and the same across all languages while phrase ends is more speci c to particular language mnemonic clause Santa clause fat lVlorpheme counting 3 kinds of sentences active transitiveintrans and passive match sentence type to sentences one is more than once How children gure who did what to whom in sentences Big bird quotgorpedquot the ball to cookie monster and big bird quotgorpedquot cookie monster the ball graph don t have to explain 9 multiple choice question about the graph Lecture 8 Four debates 3 multiple choice 1 multipleselect and 2 matching stages of past tense learning assertions of theories They re about past tense errors Order the stages of past tense productions Theories about subject less sentences 2 questions they are opposite each other answer wise Matching question match theory to what it says 2 examples are from yellow bottle other example is from past tense errors Phonology perceptionproduction asymmetry like the quot squot phenomenon Don t know which section this is from Table of Contents Lecture Subject Pagels Mental Lexicon 23 Lexical Development 46 Segmentation Problem 78 91 5 Lecture 5 Lexical 1620 Segmentation Language speci c stress 2123 patterns 2324 Conflicting Cues 2526 27 Demo of Mapping problem 29 M31 problem and ill12 solutions Solution 1 mm 1319 perceptual boundaries interact with language Solution 2 Words refer to 2026 categories I I 6 Lexical Solution 3 Social 25733 Mapping constraints mutual excluswnty and eye gaze 3437 Solution 5 Statistics of the 3840 Situation 4146 4 Table of Contents Lecture 7quot Dverview of MorphoSjmtactic Development What is morphology What is Syntax What is gnomesyntax 24 Three components of mambosyntax 610 Component 2 Syntactic Categories l ll39lr Component 3 Thematic Roles includes Lecture 3 Four Debates 1331 The four debates 1 How far do children generalize il 31 2 Can Associative models account for over generalization errorsquot i 4 3 Do children make generalizations for which they have no evidenceI 54 4 Do children39s Mpg syntactic errors re ect possible human grammars 6 Morphological Case Errors Hamel2m 3945 SHIPPED 45 Lecture 5 Slides Some solutions to the segmentation problem NO LEARNING LEARNING REQUIRED more ep U Ut reqwre earning Int e anguage e ore Adj acency to Singleword grammatical utterances morphem es doggy the you know a word is following Monosyllabic words Languagespeci c at the ends of stress patterns every language differs utterances lengthen words at end of sentence e S I C Statistical information phonotactics which syllables occur with other syllables Which can occur consonant and vowel placement Words at the ends of utterances Note that in an utterance like Look at the dogquot the end of the word dog is easy to nd Middle class Western parents do put words that they want to highlight in utterance nal position And children are better able to identify words in this position However this may not be a universal Feature of child directed speech Can identify where is the doggy at 18 months in a picture Less better at finding the dog when it is phrased as where is the doggy in the picturequot in the middle Adj acency to grammatical morphemes 24months old Studies of children39s ability to associate a word with an appropriate picture depending on the preceding grammatical morpheme Find the dog For me 335139 Performance using the the to help them find what it is they are looking for Find dog For me Find was dog For me Find gub dog For me Is the effect due to segmentation or syntactic processing Languagespeci c stress patterns 9montholds learninq Enqlish prefer the Sw stress patterns over wS stress patterns 6month olds DO NOT do that But do they use stress in segmentation 75montholds familiarized using HPP with passages containing Sw kingdom hamlet vs wS words guitar device could pick out Sw words couldn t tell wS words apart from each other At test infants heard lists of words that did or did not occur in familiarization Infants listened longer to familiarized Sw words By 10 months infants listened longer to familiarized words regardless of stress pattern Conclusions from conflicting cues Statistical cues may play a role in segmentation very early before infants know much about the word structure of their language Once infants know a few words they begin to depend on languagespeci c stress phonotactic information Summary of the segmentation problem Words are not separated From each other in the speech stream Infants and adults have available a variety of cues For quotding words languagespecific cues bebies use cues to help identify words some words requnre learning In the language Some cues are languageneutral eg singleword utterances Others require experience with the particular language eg stress Learners change their weighting of cues over development as they discover which ones are most reliable What is the Mental Lexicon A stored representation of the words we know p t w 39dmade p Word an arbitrary sound pattern that is consistently associated with a particular meaning Words in the mental lexicon have at least three types of information stored Magquot noun or verb grammatical information syntactic category gender n Lecture 6 Slides Old Idea Whole Object Assumption Words refer to whole objects Children take the label hammer to refer to the whole hammer Parents seem to assume that children make this assumption They say things like that39s a dog and that39s his tailquot That is they label whole objects rst and then label parts in a more marked way Pointing out the thing itself then its parts second A clear case of learning in mapping Humans are able to perceive all of the different Spatial arrangements However children need to ignore some Spatial differences in service of learning the prepositions in their language Sound familiar Being shaped as you learn your language to pay attention to certain aspects but ignore spatial differences Words promote categorization at 14 mos mi 80 l 70 chanting oftime 60 random they picked animal 50 picked category 40 col r 30 Y 20 j U i i l i L 39No Word Adjective Mum J I on chose 1 animal below chance EXquot 1 Yaxis is category choices 0 l4montholds treat the novel noun as referring to the object category eg animal and the novel adjective as referring to the property choice at random 0 Infants in the no word39 condition don39t show signi cant evidence of forming a category choice random Mapping errors Underextension using one word for only one object Using a word For a smaller set of referents than adult usage duck used only For rubber bathtub duck in their bathtub doggie used only For the Family pe r their doggie Ovia rt 1982 study of underextensions Children 1019 months Shown live hamster or rabbit and taught name Shown pictures and toys in same category as original or similar pictures and toys Children painted or looked at item when named 1519 month olds far superior to younger children measured how often they would pick out things that belonged to that category they pointed to the objects like they were only the first thing they saw and not others they refer to categories over time labels are learned over time Mapping errors Overextension Using a word for a larger set of referen rs fhan adul r usage doggie used to refer 139o all animals rickfockquot used lo refer ro all round rhings Overexfensions can fell us abouf fhe informa rion children use 1390 generalize Three fype of information have been no red cluster of features Mutual exclusivity way to solve the gavagi problem a u m i O n New words refer to previously unlabeled referents Each referent has only one label If a child is presented with a familiar tool hammer and a previously unlabeled tool Funnel she will think that the new word refers to the new fool Give methefunnel uses mutual exclusivity to give the correct object because they know that that is not the hammer Mutual exclusivity Six assumption problem 2 A Children and adults reliably assume I m talking about the whisk since I could ask For the thing my uncle I childremhmkthatevew gave me IF thats what I wanted object has a label they are thinking about how meme arenkely Paul Bloom and others argue that we use our 55 gnggfgeCa39bd understanding of other people and their likely behavior theory of mind to learn words On this view children with autism should have dif culty with word learning The Theory of Mind view contrasts with the idea that we have innate constraints speci c to learning words Exploring Social Constraints ability to follow adults attention Baldwin study oF joint attention with 1617 and 1819 month old infants Two toys one in bucket one in Front of child Two conditions Followin labelling ExPerimenter and child are looking toy in front of child joint attention ExPerimenter labels I Oy in front of chidif they were both paying attention they would label it Discrepant labelling Experimenter looking in bucket and child looking at toy in Front of them violates joint 39 39 39 do they think what they are attention ExPerImenter labels toy in bucket looking atwhanhe personis labenng ON QUIZ 2 Where does the shape bias come From This graph shows that as the number of nouns referring to solid objects in the child s 1 vocabulary increases so does 075 fexfure categorization based on shape 0395 025 0quot Therefore It appears that the shape bias is learned 0 C 025 2650 51 100 Could other constraints be learned as well eg whole object taxonomic nouns in vocabulary The Statistics of the Situation Quine was right When you hear a word refer to a single example you entertain several possible referents Eg dog and Dalma an first example in the slide above But just three different examples are enough to help you narrow down your hypothesis spacesecond example This supports Hypothesis Selection from generalizations based on experience using statistics of environment to figure it out Syntactic bootstrapping EnglishSpeaking children39s vocabularies contain many more nouns than verbs Some researchers have argued that verbs are harder to learn than nouns Most of the constraints discussed so Far have been proposed to account For noun acquisition whole object constraint You might imagine that extracting verb meaning From the world is dif cult Syntactic bootstrapping What makes it easier to guess the verb meaning the second time is that the sentence structure helped you out We will talk more about the sentence structure when we get to syntax The idea that sentence structure helps you learn verb meaning is called Syntactic Bootstrapping learning one thing based on something else Bootstrapping means that you learn one thing in this case verb meaning based on something else in this case sxanX order of the words Summary of the mapping problem A word can have an in nite number of referents Several cues to mapping are available Natural perceptual boundaries interact with language Words refer to categories Social constraints mutual exclusivity and eye gaze Learning to learn the shape bias Statistics of the situation Syntactic Bootstrapping Lecture 7 Slides What is morphosyntax Both morphology and syntax play a role in sentence interpretation In some languages like Cree the burden oF Creemorphmogv inFormation Falls heavily on morphology while E 9 s y aquot in others like English it Falls heavily on syntax We will talk about them as one system From the point oF view oF language development Three components of morphosyn rax 1 S e n t e n c 3 role that the personobject plays in the sentence syntactic constituents that contain a minisentence subject and verb 3 S V V After he chased the old yellow cat my dog took a na9 2clauses syntactic constituents that can be replaced by a sinqle word The old yellow cat likes chocolate milkquot likes itquot He doesquot The old yellow cat likes chocolate milk Cues ro syntactic cons ri ruenfs How do we find clauses and phrase a Few les in sentences After he chased the old yellow ca r my dog nappedquot The old yellow cal drinks chocola re milk Lengthening and pitch changes occur on cat in bo rh sentences al rhough more so at the end of a clause than a phrase Clauses are more universal Phrases are more language specific Infants use prosody to nd constituents Results Both 6montholds and 9montholds listen longer to natural than unnatural passages when natural pauses occurred at clause boundaries 9montholds but not 6montholds listen longer to natural than unnatural passages when natural pauses occurred at phrase boundaries Correlation with language typology More natural to know this from an early age Babies will need to learn based on their target language 2 Part of figuring out the meaning of the sentence is finding out the parts of the sentence The most important part of a phrase is called the head A phrase is named after the lexical category of its head If a phrase has only one word it must be the head Some lexical categories include noun verb adverb adjective preposition Thedet catnoun NP drankverb the lnilk NP VP Children and infants use distributional cues An illustration of distributional cues in English Table 74 Illustration of Some Distributional Cues for Nouns and Verbs in English the dog the tree the cat the kiss the shoe the ball a dog a tree a cat a kiss a shoe can chase can catch can eat can throw can wash can tickle is chasing is catching is eating is throwing is washing Test of sensitivity to distributional cues ifa child says the ball does she also say a ball Distributional learning by infants Table 7 5 Russian Feminine and Masculine Nouns Each with Two Case Endings Feminine Nouns polkoj rubashkoj ruchkoj knigoj korovoj polku rubashku ruchku vannu knigu Masculine Nouns uchitel39ya stroitel39ya zhitel ya korn ya pisar a nlt pisar yem uchitel39yem stroitel39yem zhitel39yem medved yem korn ye Note Bolded words were withheld during familiarization and comprised the grammatical test items An apostrophe after a consonant indicates that the consonant is palatalized in Russian Ungrammatical words were vannya korovyem medevedoj pisaru 17month olds learn the pattern but 12montholds do not learned the pattern in the 2 minute experiment can figure out the categories 3 Thematic Roles Each NP in a sentence has a particular role noun phrase that it plays in the sentence meaning called its thematic role Thematic Role Brief Description Agent someone or something that performs an action Theme someone or something to which an action is done Experiencer someone who experiences a sensation or emotion Recipient someone or something who receives something Location a place where something ends up at the end of an ac on Sentence Type NP 1 Subj NP2 NP3 Oscar wept Activesubject Agent intran itive no noun g er verb Oscar liked Activesubject Experiencer Theme Lucinda transitive noun after verb Oscar kissed Active Agent Theme Lucinda transitive Lucinda was Passive Thteme A egt 39 agent and theme IFS noun 88 n noun kissed by Oscar arempped Lucinda was Reduced Theme agen l missmg kissed passnve Oscar gave Dative Agent Recipient Theme 39 no preposition first noun second noun third noun Luanda a bOOk between 1st and 2nd Oscar gave a ToDative Agent Theme Recipient bOOk tO Lucinda ggfgzsd39tggbetween same information just different order Oscar put a book Locative Agent Theme Location on the table gives location Cues to Thematic Roles Word order Particular verbs lst NP is usually the agent The 2nd NP is the theme For pour Mom poured The dog chased the cat salt into the shaker Sentence type passive dative etc The 3rd NP is theme For ll Mom lled the lst NP is the theme in shaker with salt passive sentences Case marking her vs she in English but we don t use it overridden by word order in English not used in English not in languages like Russian Using sentence tyge to assign thematic roles Recall that in active sentences the subject is usually the agent The dog chased the cat The cat spilled the milk But in passive sentences the subject is the theme The dog was chased by the cat The milk was spilled by the cat IF children knew this relation they could decide what kind of sentence they were hearing and assign thematic roles Using Earticular verbs to assign thematic r 4yrold error Can I ll some salt hasn t figured out the verbs assign with thematic roles Can I pour some salt into the I bearquot Can I ll the bear with Pogggg salt Similar data shown in comprehension up to about 45 years Which one shows lling Which one shows pouring FILLING Summary of Assigning Thematic Roles By age two year children can use word order at least in transitive sentences and case marking to assign thematic roles they gummthe However they continue to struggle with details by around age4 sentences that are different From the most Frequent types eg intransitives passives etc Children also struggle with encoding the speci c thematic roles associated with Speci c verbs Lecture 8 Slides 4 Debates 1How far do children generalize 2Can associative models work can you get by without grammar 3Do children make generalizations For which they have no evidence 4Do children s errors reflect human languages 1 How far do children generalize Associationism not very far until they have a lot of input examples No Grammar generalization is based on particular words 10 slarf generalization on specifics ofinput Hypothesis Testing and Triggering children are generalizers but the bases differ for the two theories gene39a39izmguesses This debate often focuses on verb learninq 1 How Far do children generalize Verb Island Hypothesis Associationism they aren t generalized into all verbs won t assume learn verbs one at a time hit also means run Young children rst learn thematic roles associated with particular verbs They later collapse across similar verbs and create more abstract categories like subject direct object etc Eg giver GIVE givee given I gave Zelda a bookquot caller CALL callee callen I called Rodney a foolquot thrower THROW throwee thrown I threw him an egg qivercallerthrower become agents and subjects verbs need to be quotputtogether you can make inferences about where a verb falls in a brand new sentence Evidence against verb islands in production doing it based on the restrictions on their language Conwe amp Demufh 2007 but good at generalizing in this structure 1 3yearolds 08 dative Bert pilked Ernie the ball g double object 39939Ve 06 g Olnnovaiwe double dative Bert gorped the ball to Erniequot 36 0 39M dM donate 0 39 white new form preposmonaq donated a statue to the museum right hiaCk for m they learned Id nated mu um a statue wrong Children frame on dou e 02 object form qeneralized to 0 preposition and vice versa giggled one verb in one Domomw pmsmnd Model Form 33905 inward Preposmonal Form39 when they learned pilked they equally used both versions but did it more in the other way the properties of English doubleobJects do not work wuth certain verbs eg Bert donated Ernie the ban if they learned in the prepositional way they usually used the model version added syntacticsemantic transparency because of the presence of a preposition How Far do children generalize Summary Although there is some con icting evidence it appears that children as young as 2 years amp generalize across verbs However their willingness to do so may depend on the amount and potentially type of evidence that they have been given and also the particulars of the task less willing to generalize in the first type of example but in the last two they are 2 Can Associative models account For over generalization errors can t account with that without grammar Past tense error model goed instead of went Hypothesis Testing and Triggering say no 2 Can associative models work models that don t assume they learn a grammar One of the primary examples of children39s ability to generalize beyond the input they get are past tense errors like goed Learning appears to come in three stages pretty pity REG RESSIONS when learning phonology of the language pretty There are two main accounts of how these errors arise Past tense errors Grammatical view Past tense simpli ed present tense ending in voiced phoneme except d plus d present tense ending in voiceless phoneme except t plus t present tense ending in t or d plus lad Children apply this rule to all Forms even those that are exceptions in adult English eg goed39 instead of went39 Associative models cannot easily account For this Past Tense Errors Associationist View The past tense rule is represented in the mind on the Hypothesis Selection and Triggering views not on the Associationist view Within the Associationist view implemented in computer model There are no phonological rules Children store connections between a present and past tense version of each word Given this how can children possibly make over generalization errors Past tense errors Summary The past tense rule is represented in the mind on the Hypothesis Selection and Triggering views not on the Associationist view Is the Associationist view a work in progress more work is needed to get rid of weird errors or Fundamentally flawed the weird errors are a sign of a deep problem that requires a grammar 3 Do children make generalizations For which they have no evidence Associationism and Hypothesis no learningfrom environment T g g n g s yesu innate they are born with knowledge to form generalizations 3 Do children make generalizafions rha r aren39t in the inpu r 18monfholds Infermodal Preference Looking Procedure Look A yellow bo r rle Infants think that one refers to the entire NP 350 300 250 200 Ufamiliar Yel39ow blue 150 I novel 100 050 Mean Looking Time seconds 000 1 Control Anaphoric look more at the yellow bottle Fig 1 Mean Intuit time Iin second to the Mo test images in each Cmdition From Lidz et al 2003 What infants know about syntax but couldn39t have learned Experimental evidence For syntactic structure at 18 monthsquot But how could they learn Don t have to have it as builtin knowledge you can learn it from your environment that Look a yellow bottle Can you find another one could be taken as another bottle or another yellow bottle Because a yellow bottle is also a bottle the learner doesn t get any direct evidence that one refers speci cally to yellow bottle However if you only encounter situations in which one refers to adjective noun the whole NP and never to just the noun you have good evidence for the hypothesis that one refers to the whole NP Summary of learning errors From no input Children show a remarkable ability to interpret language the way that adults do From an early age They might achieve this ability with innate knowledge of language Triggering They might instead use indirect evidence lack of another one referring to just the noun Hypothesis Selection 4 Do children39s morphosyn racfic errors reflec r possible human grammars Associa rionism and Hypothesis Selection say i r39s possible rha r they won39t Triggering says rhey musf This deba re has an impac r on how we can inferpre r errors eg Subjec rless sem ences wan r ice creamquot Pronoun case errors her likes i r Children s subjectless sentences Examples Want ice cream Push him Triggering 1 Languages either require overt subjects eg English or not eg Spanish 2 Children are born with their prodrop parameter set to allow subjectless sentences Subjectless sentences Inputbased account But sentences like It39s rainingquot have been in the child39s input From birth Why do they suddenly notice them Furthermore children omit many parts of words and sentences including subjects These omissions might re ect dif culties in production not an incorrect grammar Three studies Subjectless sentences Summary Englishspeaking children s omissions of subjects may be might due to an incorrect settinq of an innate parameter However several studies suggest that Children attempt to produce subjects consistent with their input more aboutthe limitations They may omit SUijC I S due I39O constraints on the th I d tth g arfmaa2 t22 e an no 8 lenqth and stress patterns of sentences they produce So the problem may in Fact relate more to limits on what they can say and not the grammar 4 Debates Summary Is there a grammar Children are amazing generalizers contrary to verbisland hypothesis GrammarIess39 approaches to generalization may be Fatally awed supporting Hypothesis Selection or Triggering against Associationism Is there innate linguistic knowledge Anaphoric one prodrop and pronoun case data are explainable in a number of different ways The approaches WITH a grammar seem to do better so Hypothesis selection and triggering There IS a grammar a set of generalizations and rules and not what they take away with innate knowledge HYPOTHESIS SELECTION SEEMS TO BE SUPPORTED THE MOST
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'