Lecture 13: From the mouths of babes
A site map of early language development
Stage
Age of Onset
Characteristics
| Babbling | 6-10 months | This is regarded as a tuning up of the phonological apparatus, termed "frobbing" by Pinker. |
 |
| One word stage | 10-20 months | Words refer to objects, actions, motions, routines, and modifiers. |
| The vocabulary explosion occurs around now. |
| Two word stage | 18-24 months | Consists of pairs of content words. Grammatical morphemes are missing. |
|
| The grammar explosion | 24-36 months | Pinker calls this "All hell breaks loose" because language changes so quickly during this period. |
| | |
Babbling
This is vocal play involving strings of syllables. Infants do produce
sounds that are not present in the surrounding language, but gradually
their babbling becomes more and more attuned to that language.
Babbling also occurs in the manual mode with deaf children of deaf parents
who acquire ASL as their native language. Petitto and Marentette studied
5 infants, 2 deaf and 3 hearing from 7 to 10 months of age. The deaf children
produced vastly more hand signs that correspond to handshapes used in Sign.
Further, their babbling developed through the same sequence the vocal babbling
of hearing infants and on a similar time course. Hearing children produce
"vocal jargon babbling"
(meaningless jargon babbling that sound like sentences) at 12-14 months; similarly
the deaf infants produced manual jargon babbling. They produced possible but
nonexistant forms in the ASL lexicon, mimicing its patterns of stress and and rhythm.
There was continuity between the syllabic forms used in the deaf infants'
manual babble and their first signs.
An interesting aside: If deaf children are presented with rudimentary, but
limited examples of Sign, they will learn to sign better than their parents.
The One Word Stage
No grammatical words like is, of, to or the are included.
The Two Word Stage
The two word stage consists of utterances of pairs of content words, such as:
more book
daddy do
want high
want car
want do
dog bark
there doggie
These utterances are not simply imitations of adult speech, rather they provide a good illustration of the systematic nature of young children's talk. Their main characteristic is the lack of grammatical morphemes, in this way they they are like the language of Broca's aphasics. Universal Grammar theory analyses the phrase structure of sentences to explain what is missing from two word utterances.
Take the simple sentence "A dog barked". The phrase structure of this sentence is analyzed as follows:
Children's first grammars consist only of lexical phrases built around content words. Around the age of two they acquire functional phrases built around grammatical morphemes within which they can fit their lexical phrases. This stage of language development is so rapid that it is known as "the grammar explosion". In fact development is more ordered than the explosion imagery (or Pinker's alternative "All hell
breaks loose") suggests. Roger Brown found that the first five grammatical morphemes acquired, and their usual sequence, are as follows:
- the -ing ending of the verb: Playing
- the prepositions in and on: On the table
- the plural -s ending on nouns: Three cats
- the past irregular form of verbs: He went
- the possessive 's ending on nouns: Jonathan's dinner
Between the ages of roughly two and three years children acquire the vast majority
of the grammar of the language they are learning.
The Grammar Explosion
Children vary greatly in the rate at which they undergo this grammar explosion. Consequently age does not give a good indication of the level of a child's linguistic development. A better measure is provided by the "mean length of utterance". This is measured by recording at least 100 consecutive utterances and counting the number of morphemes, content and grammatical, in each utterance.
The following chart shows some typical data from such an analysis:
When Ross was 18 months old he produced this sentence:
Yes, I did kick the ball.
Many of the achievements of children during this stage demonstrate a remarkable
degree of systematic rule following. Pinker has used the "wug test" to illustrate
the rule-bound nature of forming plurals. Other examples are provided by
negation and question formation. In negating sentences children pass through
three stages in a completely regular sequence.
- No at the front of the sentence: No I want juice
- Movement of the no to precede the verb: I no want juice
- Correct placement of negation: I don't want juice
For questions the sequence is:
- Why you eating?
- Why you are eating?
- Why are you eating?
Children go about the business of language acquisition in a very orderly way.
It is an active and constructive process.
How adults talk to children
They often use "babytalk" which consists of reduplicated syllables,
diminutive endings,
sometimes these words are based on old forms, e.g., "choo-choo train".
Parents often use nouns rather than first or second person pronouns;
e.g., "mummy" instead of "I". They base their speech on the here and now, it
is heavily contextualized. This somewhat mitigates the gavagai problem.
Last lecture I discussed the marked differences between speech and writing. Cook puts in well in "Inside Language":
"Because of the 'first draft' nature of speech, adults make 'mistakes': they get sentences muddled up; they stop in mid-stream; and so on."
When speech is directly addressed to children, however, virtually
all of the sentences are fully formed and correct.
When speech addressed to adults is analyzed into types of utterance
and compared with speech directly adressed to children,
the following marked differences emerge: speech-to-children has
twelve times as many questions, nine times as many commands,
and one third as many statements as speech-to-adults.
Although adult speech addressed to children is typically more grammatically
well formed, Pinker makes the important point that it is usually
not grammatically simpler.
The imitation-reinforcement or behaviorist theory of language acquisition
In "Language Basics" R.L. Trask has this to say about the imitation-reinforcement
theory:
"One of the most profound and indisputable achievements of linguistics in recent years
has been the demonstration that this imitation and reinforcement model is totally,
hopelessly, grotesquely wrong."
Correction of grammar is not usually very successful. The following rather sad example was recorded and reported by David McNeill:
Child: Nobody don't like me.
Mother: No, say 'nobody likes me'.
Child: Nobody don't like me.
(Eight repetitions of this exchange)
Mother: No, now listen carefully; say 'nobody likes me'.
Child: Oh! Nobody don't likes me.
Not surprizingly, given my academic interest in langauge, I have an example of myself enacting what the childless Pinker refers to as the "fussy parent" role:
In the movie "Toy Story", one of the main characters "Buzz Lightyear"
utters these memorable sentences. (Imagine the voice of Tim Allen)
Star Command? Do you read me? Why don't they answer?
Ross likes to put these words into his toy action figure's mouth thus:
Ross: Star Command? Are you read me? Why don't they answer?
Mum (emphatically): No. It's do you read me.
Ross (more emphatically): No. Are you read me?