This page was last updated on: February 23, 2005

Approaches to Language Acquisition
About Theories of Language Acquisition
Acquiring Discourse Skills
Child Directed Speech
Further Bibliography

About Theories of Language Acquisition


1 Week "WAAAAAAAA!" 
Translation - "I'm hungry!"
Christian learns that people pay attention to you when you make noises

6 Months "Babadadagugubaba" 
Translation - none.
Christian has learned that making noises is fun.

1 Year "Haargy!" 
Translation - "There's a  cat!" (The first cat he met was named Harvey, although this was a different cat)
Christian has learned that you don't always have to point to show people things.

2 Years "Captain Haddock hit Tintin.  Wine baba. No Talk!"
Translation - Captain Haddock's hitting Tintin  with a wine bottle. Stop reading this story! (I'm scared)
Christian has learned that language lets you be quite specific about what you feel and what you want.

5 Years "Mum I don't think I want to go through with this!"
Christian has learned the core structure if his native language and.. that even if you talk like the grown-ups, they still make you go to school.

How did Christian do that? At the moment there is no one completely adequate theory of  language acquisition, although some do better than others. A completely adequate theory has to account for these facts...


          at 6 weeks they coo
          at 6 months they babble
          at 1 year they produce their first word
          at 2 years they construct 2 word sentences
          at 5 years they have acquired the core grammar of their language, an incredibly complex system of
          abstract phonological, syntactic and semantic rules. They have a working vocabulary of 6000 words,
          which on average they have learned at the rate of 1 new word every 90 minutes.

2. Children are not explicitly taught how to do this. Their syntax is very rarely explicitly corrected, and attempts at such correction are almost invariably unsuccessful.

3. Children all over the world show a similar course of development  phonologically, syntactically, and semantically.

Examples of children replacing fricatives with plosives - 'stopping'
Adult                 Child                    Language
sundar               tundar                Hindi
hus                      hut                       Norwegian
zebra                debra                    English
fazik                    pazik                    Hungarian
sopa                    topa                     Spanish

Examples using the Meaning Relation: agent+action

ENGLISH           GERMAN                  FINNISH                SAMOAN
Bambi go            Puppe kommt           Seppo putoo        Pa'u pepe
                      [doll come]               [Seppo fall]             [fall doll]
Note that even at the two word stage, children tend to follow the word order in their language.

LANGUAGE                       WORD                    FIRST USE          LATER USES
English                                   "dudu"                         ball                    marbles, wheels, cement mixers
Russian                                "buti"                          ball                     toys, radishes, stone spheres

4. Children do imitate what they hear, but not in a straightforward way. Their 'imitations' obey their own internally constructed rule systems.

Dad: Say "jump" 
Child: Dup. Only Daddy can say  "dup"

Child: She holded the baby rabbits and we patted them.
Mum: Did you say she HELD them tightly?
Child: No. She holded them loosely.

Adult: Say "Fraser will be unhappy"
Child: Fraser unhappy.


Children with otherwise normal intelligence and specific problems in acquiring language, usually syntax.

Children with quite severe impairment to general IQ but who nevertheless acquire language, including complex syntax, quite successfully.

Theories tend to orient to either.... CHILD INTERNAL FACTORS  or  CHILD EXTERNAL FACTORS

How do theorists compare on the importance of child internal or child external factors? Below are some very schematic notes on a vast body of research and spanning many years. At the end of this section, you'll find some suggestions for reading further about their work.


The linguistic input was key - a model for imitation to be either negatively or positively reinforced. No essential difference between the way a rat learns to negotiate a maze and a child learns to speak.THE

The foundations for language are laid in the interaction between children and their caregivers. The special language directed to children provides a 'clearer signal' than that addressed to adults. (See also Child Directed Speech) and is potentially a key component in the acquisition process and a teaching tool.


For Chomsky the input a child receives provides not a teaching tool but the raw material for the child to work on. A trigger for the emergence of innate Principles and Parameters which are internal to the child.


Language capacity as a separte module of the mind, not the product of general cognition. Hman children are born with specific knowledge about universal properties of human language and the limits for the variation between languages.

The child's mind has specific stragies for processing language. However, the language capacity is not necessarily seen as a separate module.           

No specific language component in the child's mind, nor specific knowledge about language. Rather the child constructs knowledge about language from scratch using a brain that is uniquely adapted and predisposed to extract patterns and build ever more sophiticated and flexible representations of the world.


Although Chomsky's and Piaget's approaches to language acquisition are similar in that concentrate on child internal factors.  They are quite different in terms of the nature of the internal factors. Chomsky takes a nativist approach, while Piaget takes an empiricist one.

Language acquisition is explained, in part, by proposing that the child does not start from scratch but is born 'pre-wired' with certain types of knowledge which are required for successful language acquisition....

Chomsky proposes that the capacity to construct a SYNTAX of L1 is innate via a set of Principles & Parameters. (Also Karmiloff-Smith and late Cromer)
Bowerman, Clark, and Karmiloff-Smith propose that there may be innate constraints on many types of non-verbal conceptual thinking and categorization which are important for acquiring language.

Language acquisition can be explained without recourse to innate knowledge and mental structures.
Skinner proposes that  humans, like all animals learn from scratch through the environment  reinforcing some responses and suppressing others. Since the human brain is more complex, humans are capable of more complex learning, e.g. language.

Much of the early work on Child Directed Speech by Catherine Snow, and on mother child interaction by Jerome Bruner reflects a considerable sympathy with the empiricist approach
Piaget proposes that children construct their knowledge of language from scratch through actively engaging with the environment. The only innate element is a brain which is 'compelled' to constantly make sense of the world in all its aspects, including language.

Think of language acquisition as a door that needs to be unlocked by the child....

For nativists, the lock already comes with a set of keys. For the empiricists, the child needs to make the keys. For empiricists like Piaget, the child makes the keys virtually alone.....For empiricists like Skinner, Bruner, Snow, the arrival of the keys is a group effort strongly influenced by the environment and caregivers.


Acquiring Discourse Skills

Of course, there is far more to becoming a competent speaker of your language and a fully fledged member of your speech community than simply acquiring the syntax, phonology and vocabulary of your native language.  Children also need to acquire discourse skills.  Here is an exercise to get you thinking about what it means to be a competent speaker, in the sense that Dell Hymes uses the term

Imagine you are in a fancy restaurant with your 'significant other' and their parents.  You are meeting the parents for the first time.  Write down a list of the communicative 'disasters' that could occur.  Remember, we are talking about verbal communication, so exclude from this things like using bad table manners, or knocking over a full glass of red wine. Focus on the 'talk'.


Learning how to establish a discourse topic (See Keenan, E. & Schieffelin, B., 1976, 'Topic as a discourse notion: A study of topic in the conversations of adults and children, in C. Li (ed) Subject and Topic, NY: Academic Press).

It has been proposed that young children's discourse lacks 'recipient design' and thus requires much adult repair and effort to establish continuity in the conversation. During topic establishment, children often pause and wait for adult confirmation before continuing.  Initially, the adult seeks clarification from the child, but with age, children increasingly initiate their own repairs when they perceive that their utterance has not been 'successful'

Young children's initial difficulties stem from:

1. A limited attention span and being easily distracted by the environment around them. Thus, they may not realize the conversation is directed to them.

2. Immature pronunciation and a tendency to speak too softly making them difficult to understand.  By 34 months, though, children show in their self-repairs: increased precision of articulation, increased volume, use of contrastive stress

3. Trouble balancing the given and the new in topic establishment which can lead to either [a] 'talking down' where the speaker assumes that too much information is 'new' to the hearer. [b] 'obscurity' where the speaker assumes too much given information. Young children usually exhibit [b]. They rely heavily on contextual cues, gestures and naming.  Thus, before the age of 3, they tend to have trouble with non-situated discourse topics because gestures don't help. Until they can use pronouns, tense marking and subordination, it can be difficult for them to establish a referent for the topic of the discourse.

4. Immature syntax which can make it difficult to establish relations between referents and 'ego-centricity' (in the Paigetian sense) which makes the child assume that the hearer can simply 'piece together' the relationship between the various topics. Adults expect contingency and relevancy in discourse and specific markers from the speaker if this is not the case, .e.g. "to change the subject", or "on another point".

Learning how to keep the topic going  (See McTear, M. 1985, Children's Conversation, Oxford: Blackwell)

This involves learning how to make contributions to the conversation that both collaborate with and incorporate the overall topic and the previous utterance, thus increasing the coherence of the conversation. McTear observed the following developments at these ages

3;8  4;0        conversation consists largely of closed I/R pairs (Intiation/Response)

4;0  4;7        the acquisition of auxiliary verbs expressing doubt or possibility (might, may, could etc.) and
                   negation allows for greater topic continuity. The R increasingly becomes R/I

4;7  4;10      greater ability to encode justifications and causal relationships allows for even more continuing

Learning how to take turns
Observational studies of parents' conversations with their children have highlighted several common features in the way the interaction proceeds. Young children are usually perceived to be incompetent turn-takers with older speakers having expectations that their contributions will be irrelevant or delayed. The younger the child, the more likely their attempts to initiate a new topic will be ignored by older speakers and the more likely they are to be interrupted or overlapped (two speakers talking simultaneously). Some of this difficulty relates to the problems outlined above in learning how to establish a topic and some comes from children's relative cognitive immaturity. At first they find it difficult to monitor the conversation for transition-relevant points, i.e. points where it would be appropriate for them to have a turn.  However, as Susan Ervin-Tripp has pointed out, young children are considerably better at turn-taking in dyads (two party conversations).  Their difficulties become more marked when they have to monitor and participate in mutli-party conversations.  Both Bruner and Catherine Snow have pointed out the importance of the early interactions between babies and their caregivers in establishing turn-taking behaviour.  Important too for practicing turn-taking is the language play that young children indulge in with each other.

Learning how to use language politely (as determined by their particular culture), including learning how to produce and interpret socially appropriate directives. For more on this topic see Pragmatic Development


Child Directed Speech (CDS)

Child Directed Speech (CDS), sometimes called 'Baby Talk' or 'Motherese', is a special style used in speech to young children and has been extensively studied over the past 30 years.

It has several content characteristics:
(Peccei, J., 1999, 'Language and Age' in L. Thomas & S. Wareing (eds) Language, Society and Power) .
- calling the child by name, often using a 'pet' name or term of endearment
- shorter, grammatically simpler sentences
- more repetition
- more use of questions or question tags (That's nice, isn't it?)
- use of 'baby-talk' words
- expanding on and/or finishing a child's utterance

CDL also has a characteristic 'sound':
- higher pitch
- slower speed
- more pauses, particularly between phrases.
- clearer, more 'distinct' pronunciation
- exaggerated intonation (some words in the sentence heavily emphasised, and a very prominent rising tone
  used for questions)

Further characteristics of adult input to children, from a study by Newport, Gleitman & Gleitman, 1977
- only 1/1500 adult-to-child utterances is ungrammatical
- adult-to-child utterances are less than half the length of adult-to-adult utterances:
  MLU = 4.24 vs MLU = 11.94 (MLU is  Mean Length of Utterance)
- adult-to-child utterance contain fewer embeddings and conjoinings
- adult-to-child utterances are more redundant, given the situational context, i.e., the context makes the
  meaning of the utterance highly predictable.


These utterances contain a range of syntactic types: imperatives, questions, declaratives; the child is not presented with neat packages of structures
They contain many elliptical utterances, e.g., subjectless questions and imperatives; the child is not always presented with fully explicit structures.

The following chart outlining the main characteristics of CDS and their possible roles in the language acquisition process has been adapted from:
Clark, H and Clark, E., 1977, Psychology and Language, NY: Harcourt Brace Jovanovich