ROLES OF INPUT 2: WHEN OUTPUT EXCEEDS INPUT BCS259: Lecture 16 3/23/2017
3-minute no laptop time
Today, 1. Learn about what pidgins and creoles are and how they come about 2. Discuss how language learners take (some times) inconsistent input and generate more regularized and systematic output 3. Understand that children are not faithfully memorizing and mimicking the input; their learning is guided by more general biases
Language learning vs. language making A common language learning situation: parent/care-taker to child Adult s language --> correct end-state model children create forms by themselves with reference to adult forms (e.g., English past tense forms) Here the input plays a dual-role: 1. source of information 2. end-state model
Language learning vs. language making example video
Language contact Historically, there are a number of situations in which some languages develop for trade or through use as a lingua franca transmission largely adult to adult There is input but no end-state goal (model) Pidgin: Language created by adults from different language backgrounds who need to communicate with each other.
Pidgins Hawaiian Pidgin language: English spoken by immigrant workers from Japan, Korea, and the Philippines who worked for English speakers
Pidgins Superstrate a high prestige language in a contact situation Substrate a low prestige language in a contact situation Ifu laik meiki, mo beta make taim, mani no kaen hapai. if like make, more better die time, money no can carry If you like/want to build (a temple), you should do it before you die. you can cannot take it with you
Pidgin characteristics 1. Reduced (simplified) phonology Tok Pisin Examples: olgeta < Eng. all together (means: all/every ) oltaim < Eng. all the time (means: always ) 2. Reduced (simplified) syntax Tok Pisin Example: [M]i stap gutpela. Mi no gat trabol I PAST-CONT good. I not have trouble. (because) I was always good, I did not make trouble.
Pidgin characteristics 3. Semantically broad words antap (<Eng. on top ) - lukluk antap = look up - putim antap long tebol = put on the table - Mi antap long yu = I am superior to you 4. Strong L1 influences kote, motete, awl friend give, ne? Buy, take-back, all friend give, final-particle [They] buy [presents], take [them] back, and give [them] to all their friends, don t they? Hall$(1943)!
Pidgin Summary small [ lexicon ] used in limited circumstances (e.g. trade) no [ native ] speakers (everybody has L1) simplified [ phonology ] and [ syntax ] lots of variation (many competing forms) Random variability (Different speakers and different groups of speakers use words and other expressions differently)
The second generation and after: Creoles
Creoles Creole : A creole is a language that once was a pidgin but which subsequently became a native language for some speakers Syntactically more complex than pidgin Larger lexicon Occasional constructions from pidgin are used consistently Function words and morphemes are developed Complex constructions are developed
From pidgins to creoles Tok Pisin: Papua New Guinea (Most vocabulary derived from English) Creolized pidgin developed tense markers Used sentence adverb baim bai to express time baim bai ( by and by = future) > bai > be (obligatory marker next to verb) Bed ia bai bruk na em bai kam gen The bed will break and he will come again
Creole characteristics Longer sentences, simplified word forms Tok Pisin Creolization Example: Fes tru, disla man na meri First true, this man and woman blem, ol sa lukautim of him, they HAB take care of pikinini blol gutpla. child of them good. In the first one, this man and his wife, they are taking good care of their child Bickerton*(1981)!
Creole characteristics Longer sentences, simplified word forms Tok Pisin Creolization Example: Fes tru, disla [<dispela] man na meri First true, this man and woman blem [<bilong en], ol sa [<save] lukautim of him, they HAB take care of pikinini blol [<bilong ol] gutpla [<gutpela]. child of them good. In the first one, this man and his wife, they are taking good care of their child Bickerton*(1981)!
Creole Summary Used for many, every-day, circumstances Larger lexicon Grammatically more consistent across speakers L2 Variation is eliminated (regularlization of variability) More concise word forms Has native speakers!
Why are creoles more regularlized? Creoles tend to have more features that other existing languages have (e.g., tense, case) Language learners = regularlizers Language Bioprogram Hypothesis (Bickerton, 1984) The capacity for language creation in creolization is the same capacity that underlies language acquisition Humans have an innate core knowledge about the structural properties human language have (domain-specific knowledge) in accord with the generativist approach
Interim summary Can learners create language when there is no end-state model is presented? Language learners are not just faithfully reproducing the input They sometimes take non-systematic input (e.g., pidgin), restructure it, and output something more complex and regularized (e.g. creole) Universal Grammar Language Bioprogram hypothesis??? (Thursday, next week)
Home Signs Home sign: A basic communication system created within a family that involves at least one linguistically, but not socially isolated, non-hearing individual. Gestural input from caretakers of home signers do not form the basis of child home sign system (Goldin Meadow & Mylander, 1983): Homesigners seem to invent a system on their own.
Home Signs Most deaf children (90-95%) born to hearing parents Parents usually don t sign, so child does not receive natural linguistic input Parents / relatives may develop a novel signed system with the child Parents usually produce individual gestures Child will string together rudimentary signs in a fluid and consistent manner
Home Sign systems Home signers distinguish nouns and verbs Home signers do not always use the same word order of their care-takers. (e.g., Child home signers in Taiwan and in the US use an order like Jar twist you instead You twist (the) jar ) Children also develop basic grammars (Tense, Aspect, Designation of Subject) Parents don t tend to pick up on these inventions
This suggests kids are doing something different from adults...is this bioprogram? Senghas (2003) notes similarity of Home Sign to Pidgins simple grammar small lexicon isolated signs Home Signs used to communicate between parties that don t share a language Can Home Sign become creolized?
Nicaraguan Sign Language (NSL) In 1978, the Nicaraguan government opened the nation s first public school for non-hearing children. Focused extensively on lip-reading (and then Spanish-based sign language), but the students signed to each other on the playground A sign language emerged! 1st cohort: students entering in the 1970s 2nd cohort: students one generation later, entering school in the 80-90s 3rd cohort: students two generations later, entering school in the 2000s
Nicaraguan Sign Language (NSL)
Nicaraguan Sign Language (NSL) First cohort signers: Each learner had their own home sign system -- No common language Developed pidgin sign languages to communicate with each other Second cohort signers: Exposed to pidgin created by first cohort Developed a creole sign langauge
Counting 1-10 1st cohort
Counting 1-10 2nd cohort
Counting 1-10 3rd cohort
Nicaraguan Sign Language (NSL) 1st cohort: slower similar to a non-signers counting on the fingers 2nd and 3rd cohorts faster more compact signs more systematic
Regularization Rotated reference frame Senghas(&(Coppola,(2001(
Regularization Second cohort signers all rotated: highly regular and systematic langauge use Senghas(&(Coppola,(2001(
Complexity Senghas (1995) 25 speakers of Nicaraguan sign language Year of Entry (before 1983 vs. 1983 or later) Age at Entry (before 1983 vs. 1983 or later) young: 0-6 years medium: 6-10 years old: 10+ years Story-retelling task
Complexity: How many arguments? Verbs with 2+ Arguments young medium old before 1983 1983 or later year of entry I ran. (1-argument) I met her. (2-arguments) I gave her the book. (3-arguments) Younger children produce more arguments (Senghas 1995)
Complexity: Verb inflections Inflections per verbs young medium old before 1983 1983 or later year of entry walk walking (verb forms) walked Younger children produce more verb forms (Senghas 1995)
Complexity: Verb agreement Agreement per Verbs young medium old before 1983 1983 or later year of entry I/You smilie. vs. He smiles. I am smiling. vs. He is smiling. Younger children produce more verb forms (Senghas 1995)
Are children special kinds of learners? What is going on here? Younger children seem to be the driving force of language creation. (They make new patterns and retain more complex patterns.) Critical period? or more domain-general statistical learning ability? Children make inconsistent usages consistent They regularize the input they get Even in the absence of the end-state model!
learning from reduced/inconsistent input What we ve seen: Children of pidgin speakers develop creole Children learning from pidgin based on homesign systems acquire more complex sign language Children of non-native speakers develop native (or very near native) proficiency In every case, output exceeds input Remaining question: Where does this systematicity come from? (more on next Thursday)
Summary 1. Learn about what pidgins and creoles are and how they come about Pidgin: Language created by adults through communicating with each other simpler vocabulary and structure a tremendous amount of variability Creole: Language created by children who learn pidgin as their input more complex yet systematic grammar regularized
Summary 2. Discuss how language learners take (some times) inconsistent input and generate more regularized and systematic output Children (especially those younger than 10) drive language creation and change: They generate more grammatically complex and highly regularized patterns out of inconsistent input Consistent pattens emerge in different creole (spoken + sign) languages such as tense, aspect, person and number
Summary 3. Understand that children are not faithfully memorizing and mimicking the input; their learning is guided by more general biases Adult s input does not necessarily work as an endstate model. Children induce systematic patterns even when the input contains noise and inconsistencies When do they deviate from the input? What s guiding the language learning and making? Universal grammar or something else?
By Tuesday & Thursday 1. Do your homework! Due Tuesday (Ask us questions. Think about possible research questions) 2. Tuesday: Brainstorming about project ideas Last year s presentations can be found here: https://languagedevelopment.wordpress.com/ 3. Fedzechkina et al., (2012) on Thu + In-class review session for Exam 2. 4. Study guide for Exam 2 will be uploaded by tomorrow night.