1.1.2.2. From Government and Binding Theory to the
Minimalist Program
The Government and Binding Theory is the second step in the
evolution of generative grammar and constitutes the foundation of MP. Although
the book devoted to this framework appears in 1981, some ideas developed within
the GBT originate from several papers committed before 1981. This theory is
mainly characterized by the modularization of the grammar. The grammar is split
into modules that have their own specific roles to play in government and
binding theory. The modules comprise: X-bar theory, theta theory, case theory,
control theory, bounding theory and binding theory.
The GBT focuses on the application of cross-linguistic
principles and parameters of Universal Grammar in the description of natural
languages. The principles are invariants of human language and are innate
whereas parameters are possible cross-linguistic variations. Many of the
principles restrict how movement is constrained. For instance, Subjacency
limits movement to crossing no more than two phrases of a particular kind, the
Structure Preserving Hypothesis states that transformations, i.e. movement, can
only move elements to positions that could be generated by means of Phrase
Structure rules. According to Borer (1984), parameters consist of choices of
feature specifications as the child acquires a lexicon (Chomsky 2004; 2007).
The computational system of every language is seen as the same. Thus, all
parameters are lexical and they account for the variety of languages. If the
child has evidence for gender in the language it hears, gender will be
included; if not, it won't be.
17
The model of grammar developed within GBT has four levels of
representation. Unlike the model of the Extended Standard Theory (1965) wherein
semantic interpretation was at the level of deep structure and phonetic
interpretation at the level of surface structure, these ones constitute each
one a level of representation. Thus, in addition to precedent levels found
within the Extended Standard Theory (deep structure and surface structure), we
have two others levels which are Logical Form (LF) and Phonological Form (PF).
Transformation rules are generalized by a single rule Move á (move
anything anywhere at anytime) which maps structures from the deep structure to
surface structure. Lexical items extracted from a lexicon are combined at
D-Structure. The sentence undergoes phrasal movement (move á) in order
to provide input to the S-Structure. Within the S-Structure the sentence is
factored into PF for sounds and shapes verifications and LF for semantic
interpretation. Within the S-S, several more rules must be satisfied, the theta
criterion and the case filter.
The Principles and Parameters theory approach was a key
development toward showing how natural language variation could be traced to a
more fundamental linguistic capacity via the setting of parameters (Chomsky and
Lasnik, 1993). Its refinements gave birth to the Minimalist Program. Chomsky
(1993) sets the basis of MP and develops it in subsequent works (Chomsky 1995,
1999, 2000, 2001, 2006, 2008, 2013...etc). The minimalist program is, as
Chomsky (2000:92) claims, a program, not a theory that seeks to discover to
what extent minimal conditions of adequacy suffice to determine the nature of
the right theory. It was conceived in order to overcome the tension between the
descriptive adequacy and the explanatory adequacy by making a balance between
them since the preceding framework was more descriptive than explanatory. The
central guiding principle in the MP is that language is an optimal solution to
map sound to meaning given constraints set by the sensorimotor system and the
Conceptual intentional system (the Strong Minimalist Thesis, Chomsky 2000). The
backbone of this program is the economy principle which, following Chomsky
(2000:99), seeks to eliminate anything unnecessary, that is superfluous
elements in representations and superfluous steps in derivations. This can be
captured graphically by contrasting the GBT model of representation against the
MP style as shown below:
18
Figure 4 : The GBT model of representation against the MP
style
|
GBT model of representation
|
MP Model of representation
|
|
Lexicon Phrase
Phonological Form
|
D-Structure
S-Structure
Structure Rules (X-bar)
Move á
Logical Form
|
Syntax
Spell
Phonological Form
|
Lexicon
Numeration
( MERGE, AGREE, MOVE)
-out
Logical Form
|
Source: Adapted from Tabe and Tamanji (2015)
As it is outlined on this diagram, the model of grammar
developed within the MP framework has two levels of representations, namely the
Logical Form (LF) and the Phonological Form (PF). These ones are interfaces,
the LF deals with the conceptual-intentional system meanwhile the PF has to do
with the sensory motor system. As one can notice, the two other levels of the
GBT (Deep-structure and Surface-structure) have been deleted since Chomsky
realized that they were operating in a similar fashion and were not forced by
interface conditions. Structures are built up through pure Merge and Move by
combining elements drawn from the lexicon. Interfaces are attained by three
major operations that hold in syntactic component namely Merge, Agree and Move.
Furthermore, Chomsky (2001, 2008) claims that derivations should proceed by
phases to avoid computational burden. He identifies CP and vP as phases and
once a phase is completed, it undergoes transfer at the interfaces. As far as
Move is concerned, Chomsky hammers that it should be implemented only as
«last resort» since it is costly. In addition, Long distance movement
should be avoided if need be or wherever necessary. Structure derivation within
MP follows a particular canal.
| 
