Language Acquisition Sınav Öncesi

Speech is like a sound river where there is no space between the sounds. How does a baby figure out where a word begins and ends?

There must be a starting point. Actually, there are two parts: segmentation (ses nehrini kelime kelime parçalara ayırdığımız phase) association (bu parçalarla doğru anlamları puzzle gibi birleştirmek. Örn: a cat is a fluffy animal in the real world.) LMNO örneği. While we are learning the alphabet, we hear the letter sounds so fast that we think that LMNO is actually a sound of alphabet, but in real, there are 4 distinct sounds of letters: L, M, N, O. In the beginning, the language which is in the baby’s world, is actually a slurry (bulamaç in Turkish) like this.

Segmentation (8 months only with sounds):

It is like a reading a long passage where there is no space between words like this: weareeating, we are going to get her (We are going together) We are going together and to get her are same as sounds, but we, adults, can understand these sentences by using in different contexts. How does a baby figure out this? At first, a baby does not know where to stop at first because it is not born with a lexicon. When he figures out this problem, he passes to new phase: Association (10-12 months)

Association (10-12 months) the mapping problem):

An association (mapping) problem occurs while a baby is trying to associate the things with real world referents. For instance, when baby hears “a rabbit” he thinks like this: Does rabbit mean a fluffy animal, white animal, long ears, jumping? Which one is correct? A child figure out this puzzle by trial-error (contexts). For instance, Association Attempt: a baby can call a red tomato as an apple because he/she codes all the red round objects as apples, but in time by way of using bootstrapping, he figures out that it is tomato not an apple. They are different objects. Babies can even figure out the meaning of abstract nouns such as happiness by using bootstrappings.

The Problems with Word-to-World Mapping

It explains why simply looking at the world isn't enough to learn words. Here are specific examples for each difficulty mentioned.

A. Scene Ambiguity (Multiple Objects) → Modi. Without extra cues (like pointing or eye gaze), the child cannot know which object is the target.

B. Abstract Nouns (Unperceivable Meanings) → "Idea" or "Tomorrow.". You cannot point to an abstract concept. "Word-to-world mapping" fails because there is nothing in the physical world to map to.

• A parent can hold up a cup and say "Cup." But how does a parent hold up a "tomorrow"?

C. Verbs: The "Time-Locked" Problem → Nouns are usually spoken while the object is present (Time-locked). Verbs are often spoken when the action is not happening. Did you spill, We are going to eat.

• Result: If the child tries to map the word "eat" to what they currently see, they might think "eat" means "Empty Table."

D. The Problem of Induction (Perspective Ambiguity)

This is the most complex philosophical problem. Even if the child knows which event is being talked about, the event itself can be viewed from two opposite angles. Agent or patient? A dog runs after a cat. The Ambiguity: This single event is simultaneously a "Chasing" event (from the dog's perspective) and a "Fleeing" event (from the cat's perspective). Why doesn't "chasing" mean "running away in front of"?

Bootstrapping

It is a code cracking method for a baby. Baby uses the parts that he knows already to figure out the other parts of language. For instance, he uses grammar of a sentence (syntax) to figure out meaning, or vice versa. There are three types of bootstrapping: Phonological (prosodic) Bootstrapping (using sounds and prosodic elements to figure out structure), Syntantic (using grammar to figure out the meaning), and Semantic Bootstrapping (using meaning to figure out the grammar). Babies can even figure out the meaning of abstract nouns such as happiness by using bootstrappings. UG must works with bootstrapping perfectly.

Phonological Bootstrapping (Prosodic Bootstrapping)

A baby is born without lexicon and grammar rules, so he needs something to build it. At first, when a baby is born, he uses music (sound and prosodic elements) to figure out the language. He categorizes languages according to their prosodic properties into three: stress-timed, syllable-timed, mora-timed. Prosodic Bootstrapping is using sounds and prosodic elements to figure out the structure (grammar). For instance, in English, there usually is a strong stress in the first syllable like DOC-tor, BA-by, so actually a child hears this sentence as: DOC-tor is checking the BA-by. Stress creates cues for him. If there is a stress like BA-by, then there is a signal that a new word is coming. There are also pauses and pitches like F0, which helps a child to figure out the structure. For instance, pauses are good cues for a baby while figuring out structure. He hears this sentence as: The boy [pauses] who ran fast [pauses] fell down. [F0] He was a good boy. In this scenario, thanks to pauses, he figures out the positions of subject (the boy), relative clause (extra information), and verb position (fell down). Also, by the help of F0 resetting pitch, he figures out that there is a new sentence as when there is an F0 pitch, there must be a new sentence in English.

• A baby acts like a statician

• Proof of phonological bootstrappings

• UG + Phonological Bootstrapping:

A baby acts like a statician

As babies are born without lexicon, they must rely on the statistics of sounds in a language to figure out word boundaries. There is a proof of this. There is an experiment on 8 months babies. A baby is listened to meaningless words bidakupatuda which are repeated in a loop. There are actually three meaningles words: bida, kupa, tuda. Statistically, after bi syllable there is 100% “da” as there are the parts of same word, but after “ku”, there is %33 possibility of “da” as they are parts of different words. When a baby is 10-12 months, things change as babies relies on metric segmentation of their native languages. For instance, in English, words are mostly trochaic (strong or weak) sylabble like BA-by, MOT-her. A baby figures out that if there is a sound like BA, then there is a new word like baby. A baby also relies on Phonotactic Constraints when he is 9 months. In languages there are phonotactic constraints (traffic rules of a language) which limits sound usage (sound combinations) in that language. For instance. There is %0 possibility a word like KN-ight because there is no /kn/ sound at the begining of an English word, so a baby can distinguish English from Dutch when he is 10-12 months by using this constraints as there is /kn/ sound at the beginning of a word in Dutch. There is also nearly no /vt/ or /ng/ (nasal n) sound at the beginning of a word in English, so when a baby hears love to, actually, he knows that they are two different words, or when he hears “I am going to..”, he knows that going and to are different words as in English, you cannot start a word with ng.

Proof of phonological bootstrappings

Natural pauses, syllable lengthening, F0 pitch resetting at the end of a sentence.

Studies show that babies are very sensitive to these cues. There is a study which is conducted with 6-7 months babies. When they are listened to sentences where there are natural pauses, they listen longer than unnatural pauses. Even, 3 days new-born baby can distinguish sound strings between words and sentences. So, a baby is born with understanding the sound boundaries ability. A 6 months baby can distinguish languages from different rhytmic class, but when he is 9 months, he even can distinguish English from Dutch because he starts hearing phonotactic constraints. Phonotactic constraints are like traffic rules which limits sound usage in baby’s native language. For instance, there is /kn/ sound at the beginning of a word in Dutch, but not in English (knight is exception because knight is not pronounced as knight in English. It is pronounced as night"

UG + Phonological Bootstrapping

While finding word boundaries (statistical learning) is powerful, it is insufficient to acquire complex grammatical rules. The child must combine Prosody + Statistics + Phonotactics constraints + UG. The word “ever”, for instance, a statistics fail because there is huge gap between ever and no as in the sentence “No person will ever come.”. How does a child figure out this complex sentence by using only statistics? Impossible.

Semantic Bootstrapping

According to Steven Pinker’s Theory, a child observes the real world and figures out the meaning. He uses this meaning knowledge to figure out the grammar of a language. For instance, when a child sees a dog is chasing a cat, he knows the agent (doer of the action of chasing: a dog) and patient (affected of the action: a cat), and when he hears a cat is chasing a dog, he creates bootstrapping in his mind. Then, agent (doer of an action) is usually at the beginning of a sentence, while a patient (affected one) is at the end of a sentence. A child becomes expert in time because he uses this bootstrapping in his language and also he follows some social cues such as eye-gazing and biases. For instance, think of a messy room where there are 10 toys. A child is wanted to find “modi”. Researcher looks at the box and says “modi”, he follows researcher’s eyes and finds “modi”. He figures out that “modi” is the toy in the box that researcher is looking at. He follows the researcher’s attention; not his own. And so, he acquires the grammar of his language perfectly thanks to this cues. He does not repeat, he acquires.

• Refutation of Steven Pinker observation theory: Blind childrens are also another powerful proof. Actually, the most powerful proof. While building a lexicon, a child must not observe the world as there is a proof from blind children (refutation of Steven Pinker). Blind children are born visually-impaired, so they cannot observe the world as others, but they still distinguish “see” from “look”

Syntactic Bootsrapping

Syntactic Bootstrapping is one of the most powerful tools for a baby while building a lexicon. A baby uses syntactic cueing while he is building new lexicon. A baby uses grammar of a sentence to figure out meaning of words, and so, he builds his own lexicon. There is a proof of this: kradding experiment. A child is watched two videos by using meaningless “kradding” word. In the first video, a duck is pushing a rabbit, but child hears “A duck is kradding a rabbit.”. When he hears this, he thinks kradding is a transitive verb because an agent (a duck) is doing an action (kradding) to something (a rabbit). This is a proof that he acquires English SVO sentence structure and the meanings by using syntactic bootstrapping. In the other video, a duck and a rabbit are wawing their hands, but a child hears “A duck and a rabbit are kradding.”. When he hears this, he figures out that kradding is intransitive verb in this context as there is no direct object. He codes kradding’s meaning as “waving hands” according to this context. This is proof of how a grammar is powerful meaning-building (lexicon-building) tool. Blind childrens are also another powerful proof. Actually, the most powerful proof. While building a lexicon, a child must not observe the world as there is a proof from blind children (refutation of Steven Pinker). Blind children are born visually-impaired, so they cannot observe the world as others, but they still distinguish “see” from “look”. How? Because they do not need to see the world, syntax is enough for acquiring. When they are wanted to tap the chair but not look at it, they tap passively, but when they are wanted to looking at the chair, they explores by using their hands actively. They must have learnt this difference by using syntax.

The problem of finding word forms arises

Because a speech is continuous (there is no break between words), a baby is born without lexicon, and there is segmentation ambiguity (lack of isolation), words are rarely taught one by one; they are in complex streams. For instance, a baby can hear this sentence as: We are going to get her (prepositional phrase), while and adult hears this as: We are going together (an adverb). A baby must figure out together is adverb in this context, not prepositional phrase like “to get her”.

How children solve problems

The whole object bias: A new word refers the entire object; not a part of it. (Rabbit: White fluffy jumping animal; not only long ears)

Social cues: Eye-gazing. Babies know that words refer to what the speaker is looking at, not what babies are looking at. They follow the eyes of speakers.

The mutual exclusivity Bias: Every object has only one name, and a baby eliminates other meanings by using this bias. Cup and handle example.

The Taxonomic Bias: Words label categories; not themes. Showing cow and get the word milk (thematic), and pig (taxonomic) example. They are both animals. Dax example.

Syntactic cueing: Grammatical markers alone tell the child if a word is an object or an verb (action). Show me a sib (dough), show me sibbing (action). Morphological marker -ing points out that this is a verb (action) not an object (noun).

Soruların cevabı:

Why is acquisition of verbs more difficult than acquisition of nouns? Explain syntactic cueing of verb meaning.

- Nouns are often concrete and observable in the real world. In contrast, acquiring verbs often requires more than just observation; a baby needs understanding the structure of the language (syntax) to figure out the meaning of the action (verb), especially for abstract concepts or when visual observation is insufficient as proven by the Blind Children example. Blind children, who cannot observe the world to figure out meanings, rely on syntax to distinguish verbs like "see" versus "look". When they hear see, they only tap the object passively, but when they hear the look, they tap the object actively (observing by hands). Children use syntactic cueing to distinguish between object (nouns) and action words (verbs). Syntactic cueing means grammatical markers alone can indicate to a child whether a word is an object or a verb. For example, there is an experiment about this. Children are asked to "show me a sib" versus "show me sibbing". The presence of the morphological marker "-ing" indicated to the child that the word was a verb (an action), whereas the lack of it indicated an object (dough).

Define and explain semantic bootstrapping of syntax. Is information from a single frame sufficient for the children to learn the meaning of the verb?

- According to Steven Pinker’s Theory, a child observes the real world and gets the meaning knowledge. He uses this meaning knowledge to figure out the grammar of a language. For instance, when a child sees a dog is chasing a cat, he knows the agent (doer of the action of chasing: a dog) and patient (affected of the action: a cat), and when he hears a cat is chasing a dog, he creates bootstrapping in his mind. Then, agent (doer of an action) is usually at the beginning of a sentence, while a patient (affected one) is at the end of a sentence. A child becomes expert in time because he uses this bootstrapping in his language and also he follows some social cues such as eye-gazing and biases. For instance, think of a messy room where there are 10 toys. A child is wanted to find “modi”. Researcher looks at the box and says “modi”, he follows his eyes and finds “modi”. He figures out that “modi” is the toy in the box that researcher is looking at. He follows the researcher’s attention; not his own. And so, he acquires the grammar of his language perfectly thanks to this cues. He does not repeat, he acquires. → No it is not sufficient. The mapping problem (association problem), Ambiguity of verbs (running tavşana mı ait yoksa eylemin kendisi mi), need for multiple contexts (verbs change in different situations like kradding), acquisition in time (babies become experts in acquistion in time by using bootstrapping) yüzünden değildir

Define and explain syntactic bootstrapping.

- Syntactic Bootstrapping is one of the most powerful tools for a baby while building a lexicon. A baby uses grammar of a sentence to figure out meaning of words, and so, he builds his own lexicon. There is a proof of this: kradding experiment. A child is watched two videos by using meaningless “kradding” word. In the first video, a duck is pushing a rabbit, but child hears “A duck is kradding a rabbit.”. When he hears this, he thinks kradding is a transitive verb because an agent (a duck) is doing an action (kradding) to something (a rabbit). This is a proof that he acquires English SVO sentence structure by using syntactic bootstrapping. In the other video, a duck and a rabbit are waing their hands, but a child hears “A duck and a rabbit are kradding.”. When he hears this, he figures out that kradding is intransitive verb in this context as there is no direct object. He codes kradding’s meaning as “waving hands” according to this context. This is proof of how a grammar is powerful meaning-building (lexicon-building) tool. Blind childrens: The most powerful proof. While building a lexicon, a child must not observe the world as there is a proof from blind children. Blind children are born visually-impaired, so they cannot observe the world as others, but they still distinguish “see” from “look”

Compare & Contrast rationalist and empiricist views by mentioning the relevant theories and their criticism.

- Bu soru language is acquired or learnt sorusu.

Syntactic structures are a projection of lexical properties. Explain.

- We can explain this concept through using Syntactic Bootstrapping, where the grammatical structure reveals the properties of the words within it. the grammar and structure of a sentence are determined by the specific requirements and definitions of the words (lexicon) used within it, particularly the verbs. Syntactic structures are projections of lexical properties because childrens are "reverse engineer" of this process: they use the sentence structure they hear to figure out the meaning and properties of unknown words. For example, in Kradding Experiment shows us that children understand that a sentence's structure reflects the verb's "lexical properties" (such as whether it is transitive or intransitive). When children hear "A duck is kradding a rabbit," they identify a Subject-Verb-Object structure. Because this structure "projects" a transitive relationship, the child infers that "kradding" must be a transitive verb (an action done to someone, like pushing). However, when children hear "A duck and a rabbit are kradding," they identify a structure with no direct object. This structure projects an intransitive property, leading the child to infer that "kradding" is an intransitive verb (an action done by the subjects, like waving hands). Briefly, the syntactic frame (the sentence structure) serves as a blueprint that reveals the hidden "lexical properties" (meaning and usage rules) of the verb.