1. Universal Grammar and the Innateness Hypothesis

• What is Language?
  • Knowledge
  • Use
• Why is it the way it is?
  • An arbitrary system; consistent with the structure of the mind
  • Designed for efficence communication and processing
• Why do only humans have it?
  • Specified in the human genome
  • By-product of human intelligence and social complexity
• How do children get language?
  • Born with an innate predisposition for acquisition
  • General learning strategies
• What is the best kind of evidence for testing theoretical predictions?
  • Native speaker acceptability judgements (backed by other data)
  • Data from naturalistic usage (corpus)

To achieve descriptive adequacy

• Description: account for what is grammatical in a given language
• Generalisation: predict what is grammatical in a given language
• Account for variation across languages

To achieve explanatory adequacy

• To account for how language is acquired in spite of the Poverty of the Stimulus
  • The input a child receives contains random errors, only positive evidence, and is finite. 
  • The output is complete language; there must therefore be innate support. 
• To do so with elegant theory (minimal independent assumptions/ theoretical machinery)

GRAMMAR IS a system of unconscious principles that determine what is possible in a language.

GRAMMAR ISN'T a finite inventory of memorised sentences.

This assumes that the brain overgenerates language; however, if a principle is violated, the sentence is always ungrammatical. 

I-LANGUAGE: internal, unconscious knowlegde of what is possible and impossible in language. 

• The Linguist aims to make this knowledge explicit on the basis of:
  • Attested data (actual production)
  • Acceptability judgements from native speakers

E-LANGUAGE: external, actual manifestations of language.

This shows that performance and competence are not equal.

They believe that:

• No innate predisposition for language.
• Language is subsumed under other cognitive abilities. 
• Competence is equal to performance. 
• Knowledge of language rises from input and exposure. 
• Languages vary more and are more subject to socio-cultural factors. 

UG is:

• A system of principles common to all languages. 
• Principles vary according to parameters: binary options that determine possible variation between languages. 
• Available only to humans (because it is assumed to be innate). 

Example parameter: syntactic subjects can be realised as non-overt pronouns in some languages. 

Example principle: determiner recursion is impossible in any language. 

A linguist must therefore work out what data from positive evidence is required to trigger acquisition with the help of UG.

According to Principles and Parameters Theory, the following are innate:

• Grammatical categories.
• X-Bar Trees.
• Principles and Parameters. 

Phrase Structure Rules are:

• Abstract economical rules that generate all the possible sentences in a language. These rules combine words > phrases > sentences.
• Allow recursion (embedding of phrases). 
• Account for the modularity of the system (small trees plug into larger trees). 

Sample DP PSR:   

• DP > D'
• D' > D NP
• NP > N'
• N' > N

This provides a generalised mechanism for generating phrase structures across different categories in different languages. There is only one phrase structure rule for all categories and this is a part of UG:

• XP > (WP) X'
• X' > X' (ZP)
• X' > X (YP)

• XP = Maximal Projection
• X' = Intermediate Projection
• X = Head
• YP = Complement of X
• ZP = Adjunct to X
• WP = Specifier of XP

X-Bar Trees use binary branching (two nodes maximum). This allows a more constrained grammar and a more economical, elegant model, which helps explanatory adequacy.

• The position hosting auxiliaries is called INFL or I (Inflection).
• Inflection is a functional head. 
• Inflection is the head of the sentence.
• If I is the head of the sentence, then a sentence is just another type of phrase; an IP (Inflection Phrase). 

• Subject DPs are located in the Spec of IP and agree with I. 
• Any clause containing a subject and verb is an IP. 
• I may be realised as AUX, an Affix, the infinitve to, or the abstract features on V.
• Auxiliaries are subject to head movement. 
• IPs can be inside other IPs.

• Complementisers are another category of functional head. 
• Certain verbs can take complementiser phrases (CPs) as complements, the heads of which in turn take IPs are complements. 
• Spec(CP) is also the position that hosts wh-pronouns in wh-movement.

We know what phrase can be accompanied by a phrase within an XP because of the requirements of the head. These requirements are encoded in the lexicon.

The lexicon is a list of words (or lexemes) in a language containing information about:

• The meaning of the word.
• Their lexical category. 
• Their argument structure and c-selectional properties. 

Argument structure is what arguments are required by the head and how many. 

• External arguments are subjects. 
• Internal arguments are complements. 

C(complement)-Selection (subcategorisation) lists which complements are required by the head. 

• Heads c-select constituents (phrases), not words. 
• Argument structure includes subjects, c-selection doesn't.
• Complements are linked with a particular syntactic position; arguments aren't. 
• Complements are specified for phrasal category; arguments aren't.
• A constituent can be both.
• Complements are part of phrase structure (c-selection), arguments are part of argument structure.