PRINCIPLES AND PARAMETERS
A nativist theory of first language acquisition which reflects the current thinking of Noam Chomsky.
Chomsky takes the view that infants are born with an innate Universal Grammar (UG). This consists of a set of universal principles which characterise all (or nearly all) languages and a set of parameters, features which differentiate languages, usually on binary lines. A simple example of a parameter is the distinction between pro drop languages which permit the omission of a subject pronoun and languages where the subject pronoun is obligatory (cf.: Italian capisco and English I understand). Parameter setting is often represented on a + or - basis (e.g. + or - Pro-drop).
A child is thus innately endowed with knowledge about language in general which gives it a head start in cracking the code of speech. However, it is also endowed with a set of choices which have to be made in relation to the language to which it is exposed. Examples of innate principles are:
Structure dependency: the structure of all languages consists of hierarchically organised phrases.
the Projection Principle: syntactic structure is determined by entries in the lexicon (the choice of the verb GIVE entails the use of a particular syntactic pattern).
the Subjacency Principle: any constituent of a sentence that is moved (for example, to form a question or negative) can only cross one major boundary (a bounding node). Where such a major boundary falls is, however, determined by a language-specific parameter.
Binding Principles: unlike most pronouns, an anaphor (the term is used here for reflexives such as HIMSELF and reciprocals such as EACH OTHER) can only refer to an antecedent within the same sentence.
Examples of parameters set in response to the ambient language are:
the Null Subject (or Pro-drop) Parameter: whether the language does/ does not oblige the speaker to express a pronoun subject.
the Head Parameter: whether the head (major constituent) falls at the beginning or at the end of a standard phrase.
Bounding Parameters: restrictions on the way in which constituents can be moved (for example, in forming negatives and interroga tives).
the Adjacency Parameter: in effect, whether a transitive verb has to be followed immediately by its direct object or not. Compare English I like music a lot with French J’aime beaucoup la musique.
the Branching Parameter: whether the hierarchical structure of a sentence as shown in a tree diagram branches towards the left or towards the right. There are marked differences between right branching languages like English which standardly place the direct object after the verb and left-branching ones like Japanese where the object occurs before the verb.
A further parameter in which there has been much interest concerns preposition stranding. English accepts both pied-piping (About what were you talking?) and stranding (What were you talking about?). However, there are some quite complex restrictions on stranding, and native speakers disagree about the acceptability of sentences such as What meeting did she phone after? There are wide variations between languages as to whether they permit stranding or not.
Many attempts have been made to test the Principles and Parameters hypothesis empirically. A problem here lies in the fact that UG relates to an individual’s competence and not to their performance. An established procedure is therefore to ask subjects to make grammaticality judgements, i.e. to indicate whether for them, as users of the language, a sentence is grammatically acceptable or not. This method has been used to trace similarities between native speakers of a given language, to compare responses across languages whose parameters are said to be different and to compare the judgements made by monolinguals with those made by bilinguals or learners of a second language.
Attention has focused on the process of parameter setting. One issue is whether both parameters are neutral at the outset or whether one is the default or unmarked one. Some observational evidence suggests that infants acquiring their first language start off with an unmarked setting which has to be reset if it is not appropriate. For example, English-acquiring infants often omit pronoun subjects in their early productions, suggesting ‘pro-drop’ as a default setting. However, the ‘pro-drop’ example illustrates the dangers of using empirical data from language acquisition to support conclusions about UG. The data relates to the infants’ productions (i.e. to their performance); it is by no means clear to what extent this reflects the underlying competence which they have derived from inherited UG. The absence of a subject pronoun may reflect a cognitive inability to process more than two words rather than any preferential parameter setting.
Parameter setting also has important implications for research into second language acquisition. When a learner acquires a language whose settings are different from those of their L1, they have to reset established parameters. There is interest in whether some parameters are more difficult to reset than others. Is it easier to move from a marked to an unmarked setting (thus reverting to the situation afforded in infancy by UG)? Or is it easier to move from an unmarked to a marked?
See also: Chomskyan theory, Markedness, Universal Grammar
Further reading: Cook and Newson (1996); Crain and Lillo-Martin (1999); O’Grady (1997: Chap. 13); Radford (1990)
