Introduction to Acquisition

What would lead linguists to believe that language is largely innate?

• The "poverty of the stimulus"
  • Different children receive drastically different input.
  • Speech to children (like all speech) contains performance errors, but these are not marked as such for the child.
  • In trying to figure out what a word or a sentence means, many different interpretations are possible.
• The lack of negative evidence (the generative / nativist position)
  • Language acquisition means in part learning the right grammar. Each grammar is compatible with some forms and incompatible with other forms. The data that children use in coming up with candidate grammars are examples of grammatical forms provided by speakers of the language.
  • Each of the child's hypotheses about the correct grammar is a generatlization over the forms she has heard.
  • But children may over-generalize; that is, they may get the grammar wrong by hypothesizing one that is compatible with some ungrammatical forms.
  • If children over-generalize, they apparently need negative evidence to correct their overgeneralizations. That is, they need to be told that certain forms are ungrammatical as well as that certain forms are grammatical.
  • But children apparently either do not receive negative evidence or ignore it when they do.
• Given the poverty of the stimulus and the lack of negative evidence, the only way a child could learn language is to have innate help, constraints on the possible grammars and meanings that can be learned: the Language Acquisition Device or Universal Grammar. Particular constraints have been proposed for each domain within language.

What might be innate?

• Cognitive predispositions which are specific to language, as opposed to cognition in general
• Cognitive predispositions specific to particular domains within language (e.g., syntax, phonology)

Why do some question strong innateness and modularity?

• The input is much richer than it appears at first glance; caregivers may even unconsciously sequence the input.
• Children may receive indirect negative evidence; that is, people may fail to respond appropriately when they produce ungrammatical forms.
• There may be other types of learning mechanisms which do not require explicit negative evidence.
• Innateness says little about the stages in language development.