Language Acquisiton

The study of language acquisition examines how knowledge of language emerges in early infancy and how it grows. The framework adopted is the generative theory of Universal Grammar (UG). This theory posits that human beings are innately endowed with a system of richly structured linguistic knowledge which guides infants in analyzing incoming linguistic stimuli. This approach allows for clear and falsifiable predictions about children's linguistic competence.

1.1 Reflections on the Course of Language Acquisition

Human language acquisition is an astonishing process; children accomplish putting words in the correct order and constructing complex sentences in about three years.

Acquiring a first (native) language is an effortless achievement occurring under the following crucial conditions:

• Without explicit teaching: Systematic instruction is not required.

• On the basis of positive evidence (i.e., the utterances they hear).

• Under varying circumstances and in a limited amount of time (by about age 5, most constructions are mastered).

• In identical ways across different languages (e.g., babbling starting at 6–8 months and first words at 10–12 months).

The Role of Negative Evidence:

• Parents’ corrections (negative evidence) are rare, generally noisy, and appear to be ineffective at improving children's grammatical behavior.

• For instance, a child resisted correction eight times regarding the use of double negation (Nobody don't like me).

• Children have the best chance of success by relying solely on positive evidence, which is abundantly available.

1.2 The Logical Problem of Language Acquisition

This problem, characterized by the argument from the poverty of the stimulus, highlights that children attain very rich linguistic knowledge (encompassing an infinite number of sentences) despite hearing only a finite number of sentences. Since children rely only on positive evidence and are not told which sentences are ill-formed (e.g., they never produce sentences violating abstract properties, even though these properties cannot be deduced from input), the linguistic knowledge must come from an innate source.

1.3 The Notion of Grammar

To know a language means possessing grammar, a finite system represented in the mind/brain. It is a mental generative procedure that uses finite means to generate an indefinite number of sentences.

Grammar provides the tools to:

• Establish acceptability: Determine whether a sentence is licit (e.g., condemning Dog a old a bone ate).

• Assign structural representations: These representations are constrained in specific ways and allow humans to cope with linguistic ambiguity. For example, the sentence John kept the car in the garage has two interpretations depending on whether the car in the garage forms a single constituent.

1.4 Constraints

Constraints are linguistic principles that forbid specific word sequences, operations, or sound/meaning pairings. They are universal (holding across languages) and inviolable.

• Constraints on Form: Block ungrammatical constructions (e.g., principles prohibiting forms like How does John regret that John behaved?).

• Constraints on Meaning: Limit the meaning assigned to acceptable sentences (e.g., Principle C of the binding theory, which bans an anaphoric interpretation of a pronoun in certain structural configurations, as seen in He danced, while John was singing).

1.5 Where Does Knowledge of Language Come From?

Alternative hypotheses like imitation, reinforcement, and association fail to explain language acquisition.

• Failure of Imitation and Reinforcement: Children produce novel utterances (e.g., negative questions like What does he doesn't eat?) and overregularized forms (which they haven't heard, like goed or singed) that go beyond the input, demonstrating they are attempting to discover abstract rules. Since corrections are ineffective and input is based on meaning, not grammar, reinforcement is deemed insufficient.

• Failure of Association (Connectionism): Although association mechanisms might help segment speech, they struggle to explain abstract and intricate aspects of linguistic knowledge, such as the constraints discussed in section 1.4 or how to handle linguistic ambiguity.

• The Innateness Hypothesis: This leads to the conclusion that linguistic knowledge is innate. The Universal Grammar (UG) is the genetic equipment necessary for acquisition, defining the range of possible human languages.

1.5.5 The Principles-and-Parameters Model: UG is composed of two types of constraints:

1. Principles: Encode the invariant/universal properties of languages (e.g., the constraint governing pronoun interpretation).

2. Parameters: Encode the properties that vary from language to language, acting like switches to be set by environmental input. An example is the pro-drop or null subject parameter (governing whether the sentential subject must be phonologically expressed, set positively for Italian, negatively for English).

Acquisition is a selective process; the child selects the appropriate values for parameters based on the input, rather than being explicitly taught (an instructive theory).

1.6 The Critical Period

Innate behaviors are often associated with critical periods during which the ability to acquire competence reaches its peak before declining.

• Evidence comes from individuals deprived of early language exposure (like Genie, who showed limited syntax even after years of rehabilitation).

• Congenitally deaf individuals exposed to American Sign Language (ASL) after birth did not attain native mastery of morphology as well as those exposed from birth.

• Non-native speakers exposed to a second language after age 7 typically fail to achieve native mastery of morphology and syntax.

Chapter 2: First Steps into LanguageIntroduction

Infants are born processing speech stimuli and display surprising sensitivity to acoustic cues for units like syllables, phonemes, and words. They start as universal learners, and experience narrows this perceptual sensitivity in a selective process.

2.1 The Quest for the Native Language

2.1.1 Language Discrimination:

• Newborns (as young as 4-day-olds) can distinguish their native language from a foreign language (e.g., French from Russian).

• The High-amplitude sucking procedure (HAS) is used, measuring increased sucking rates when infants detect a change in stimuli.

2.1.2 Discrimination between Two Foreign Languages:

• Newborns can also distinguish between two foreign languages they have not been exposed to (e.g., 4-day-old French learners distinguishing English from Italian or Japanese).

• However, this ability declines rapidly; 2-month-old English learners struggle to discriminate certain foreign pairs (like French from Russian/Japanese).

2.1.3 The Source of Discrimination:

• Infants rely on prosodic information (suprasegmental structure, like stress and intonation) for discrimination.

• If stimuli are played backward (destroying prosodic coherence but preserving phonemic and word-level prosody), infants fail to discriminate languages.

2.1.4 The Rhythmic Hypothesis:

• Infants categorize languages based on their rhythmic properties (stress-timed, syllable-timed, mora-timed).

• The fundamental unit of representation in early development is the syllable (or vocalic nucleus), which carries acoustic information about length, intensity, and height. Infants can count syllables (and thus vowels) in a word.

• Experiments show infants detect a change when presented with sentences belonging to a different rhythmic class (e.g., Italian/Spanish to Dutch/English).

2.2 Learning the Phonemic Contrasts of the Native Language

• Universal Capacity: Infants start able to distinguish a wide variety of sound contrasts, including non-native ones (e.g., Hindi and Salish contrasts for English learners).

• Functional Reorganization: This extraordinary capacity rapidly declines. By 10–12 months, infants are no longer able to distinguish non-native sound contrasts.

• This developmental change is called the functional reorganization of the sound space and is considered a gain for the learner, as it restricts the search space to the sounds relevant to the native language.

2.3 Infants’ Speech Production

2.3.1 Maturation of the Vocal Apparatus:

• The infant’s vocal tract initially resembles that of apes (high larynx). Around 4 months, the larynx descends, and the vocal tract shape approximates the adult configuration, allowing for the production of speech sounds.

2.3.2 Babbling:

• Begins around 6–8 months. It involves specific sounds, syllable organization, and lacks associated meaning.

• Forms include canonical babbling (sequence of the same CV syllable, e.g., bababa) and variegated babbling(sequence of different CV syllables).

2.3.3 Amodal Language Capacity:

• Deaf children exposed to sign language produce manual babbling that is similar in timing and structure to vocal babbling in hearing infants.

• This suggests that the human capacity for language is amodal—sensitive to the particular units, structures, and regularities of natural languages, regardless of the modality of expression (speech or sign).

2.3.4 Babbling and Word Production:

• Babbling is considered a fundamental step in word development.

Summary of Linguistic Development (Chapters 1 and 2)