Transcript Generative Grammar, Minimalism, and Language Change

Generative Grammar, Minimalism,
and Language Change
Elly van Gelderen
ASU CS Seminar
16 February 2015
Outline
Brief history of Generative Grammar
structure/hierarchy over linear order
non-imitative
Introduction to the Minimalist Program (MP)
attribute as little as possible to UG
How language change informs the MP:
unidirectional change tells one how the
language learner builds her/his grammar
Generative Grammar of the 1950s
Noam Chomsky and the generative model he
develops offer an alternative, starting in the late
1950s, to descriptive linguistics and behaviorism
and bring about a revolution in the fields of
linguistics, psychology, and cognitive science.
Chomsky continues to stress descriptive adequacy
and explanatory adequacy to emphasize how the
language faculty is represented in and acquired by
humans. BEA: why is the FoL the way it is?
Poverty of the Stimulus
The focus is on the mind of the language
learner/user (the competence) and ceases to be
on the structures present in the language
produced (the performance). The input to
language learning is seen as poor, a
phenomenon known as the ‘poverty of the
stimulus'. The basis for this phenomenon is that
speakers know so much more than what they
have evidence for from the input.
Chomsky (1975: 36).
The innate language faculty, when "stimulated
by appropriate and continuing experience, …
creates a grammar that creates sentences with
formal and semantic properties”.
Model of language acquisition/change
(based on Andersen 1973)
Generation n
UG
+
experience
=
I-language n
E-language n
+ innovations
Generation n+1
UG
+
experience n
=
I-language n+1
E-language n+1
Internal Grammar
Structure over linear order
Reflexives
(1) He voted for himself
(2) *They wanted him to vote for themselves.
Who is `himself’ in:
(3) The uncle of Obama voted for himself.
Subject-Aux Inversion
(4) Elephants that can swim are happy.
(5) Are elephants that can swim happy?
 (6) * Can elephants that swim are happy?
Locality of movement
(1)
Who did she think she was dealing with?
(COCA 2007)
(2) *Who did she ask when John met?
(3) I heard (the report) that she met him.
> (4) Who did I hear that she met?
> (5) *Who did I hear the report that she met?
Reanalysis is crucial
(1) Paul said, "Starting would be a good thing to do.
How would you like to begin?“ (COCA 2010 Fiction)
(cartoon is on Handout)
Phrase Structure and Transformations
(1)
a.
Sentence
NP + VP
b.
VP
Verb + NP
(Chomsky 1957 : 27)
(2) If S1 is a grammatical sentence of the form
NP1 Aux V
NP2,
then the corresponding string of the form
NP2 Aux + be + en
V
by + NP1
is also a grammatical sentence.
(Chomsky 1957: 43)
X’, Principles and Parameters improve
this system:
(1)
CP
C’
C
TP
T’
T
VP
V’
V
(2) Headedness, pro-drop, etc
...
Chomsky (2014: Interview 4 March)
Between early and late GG: “There are two crucial
things that remain the same. One is recognition
that the core property of language is an unbounded
system for generating hierarchic structures with
dual interpretations and the other is that it should
be constructed within a biological framework”.
Emphasis on MERGE (=UG) and features
(parametrized)
Minimalist Program: reduce UG
Three Factors
“(1) genetic endowment, which sets limits on the
attainable languages, thereby making language
acquisition possible; (2) external data, converted to
the experience that selects one or another
language within a narrow range; (3) principles not
specific to FL [the Faculty of Language]. Some of the
third factor principles have the flavor of the
constraints that enter into all facets of growth and
evolution.... Among these are principles of efficient
computation”. (Chomsky 2007: 3)
Economy
Locality = Minimize computational burden (Ross
1967; Chomsky 1973)
Use a head = Minimize Structure (Head
Preference Principle, van Gelderen 2004)
Late Merge = Minimize computational burden
(van Gelderen 2004, and others)
The latter two can be seen in terms of Feature
Economy
Borer-Chomsky-Conjecture
Parameters now consist of choices of feature
specifications as the child acquires a lexicon
(Chomsky 2007).
Baker, while disagreeing with this view of
parameters, calls this the Borer-ChomskyConjecture (2008: 156):
"All parameters of variation are attributable to
differences in the features of particular items
(e.g., the functional heads) in the lexicon."
Types of minimalist features
The semantic features of lexical items (which
have to be cognitively based)
The interpretable ones relevant at the
Conceptual-Intentional interface.
Uninterpretable features act as `glue’ so to
speak to help out merge. For instance, person
and number features (=phi-features) are
interpretable on nouns but not on verbs.
Formal features are interpretable and
uninterpretable
(Chomsky 1995: 277):
airplane
Interpr. [nominal]
[3 person]
[non-human]
Uninterpr [Case]
build
[verbal]
[assign
accusative]
[phi]
Merge and AGREE
(1)
TP
T’
T
[u-phi]
[i-pr] DP
VP
many buffaloes
[i-3] [i-P]
V’
V
live
PP
in this room
Current Generative Model
Lexicon
N(arrow) S(yntax)
Interfaces PHON
External
systems:Sensorimotor
SEM
Conceptual-Intentional
Semantic and formal overlap:
Chomsky (1995: 230; 381) suggests: "formal
features have semantic correlates and reflect
semantic properties (accusative Case and
transitivity, for example)."
I interpret this: If a language has nouns with
semantic phi-features, the learner will be able
to hypothesize uninterpretable features on
another F (and will be able to bundle them
there).
Radford (2000): in acquisition from + > -
If semantic features are innate, we need:
Feature Economy
(a) Utilize semantic features: use them as for
functional categories, i.e. as formal features
(van Gelderen 2008; 2011).
(b) If a specific feature appears more than
once, one of these is interpretable and the
others are uninterpretable (Muysken 2008).
Features and grammaticalization
Grammaticalization is a change from semantic to
formal features with (optional) loss of
phonetic features.
For instance, a verb with semantic features, such
as Old English will with [volition, expectation,
future], can be reanalyzed as ‘ll having only
the grammatical feature [future].
Cycles tell us which features matter
Subject and Object Agreement (Givón)
demonstrative > third ps pronoun > agreement > zero
noun > first and second person > agreement > zero
noun > noun marker > agreement > zero
Copula Cycle (Katz)
demonstrative > copula > zero
third person > copula > zero
verb > copula > aspect
Noun Cycle (Greenberg)
demonstrative > definite article > ‘Case’ > zero
noun > number/gender > zero
And about processing/economy
Negative Cycle (Gardiner/Jespersen
see van der Auwera)
a negative argument > negative adverb > negative
particle > zero
b verb > aspect > negative > C
(negative polarity cycle: Willis)
CP Cycle
Adjunct AP/PP > ... > C
Future and Aspect Auxiliary
A/P > M > T (> C)
V > ASP
Negative Cycle in Old English
450-1150 CE
a.
no/ne
early Old English
b.
ne
after 900, esp S
c.
(ne) not
d.
not >
(na wiht/not)
after 1350
-not/-n’t after 1400
Neg Cycle in terms of structure
NegP
Neg’
Neg
ne
VP
V
DP/AP
no thing
and in terms of features
DP in the VP
semantic
> Head Neg
> [u-neg]
>
>
Specifier of NegP
[i-neg]
>
negative affix
and then renewal is needed from another lexical
element
The Subject Cycle
A. demonstrative > third person pronoun > clitic
> agreement
B. noun/oblique pronoun > first/second pron >
clitic > agreement
"agreement and pronominalization ... Are
fundamentally one and the same
phenomenon“ (Givón 1978: 151).
Just a few examples
The Basque verbal prefixes n-, g-, z- are identical to the pronouns
ni ‘I’, gu ‘we’, and zu ‘you.’ (Gavel & Henri-Lacombe 1929-37),
As early as the 19th century, Proto Indo-European verbal endings
-mi, si, -ti are considered to arise from pronouns (e.g. Bopp
1816).
Hale (1973: 340): in Pama-Nyungan inflectional markers are
derived from independent pronouns: “the source of
pronominal clitics in Walbiri is in fact independent pronouns”.
Mithun (1991): Iroquoian agreement markers derive from ProtoIroquoian pronouns
Haugen (2004: 319): Nahuatl agreement markers derive from
pronouns.
The various cycles in terms of features
The cycle of agreement
noun > emphatic > pronoun > agreement > 0
[sem]
[i-phi] [i-phi]/[u-phi] [u-phi]
The cycles of negation
Adjunct/Argument Specifier
semantic >
[i-NEG]>
Head (of NegP)
[u-NEG]
>
Modal Cycle
Verb
>
[volition, expectation, future]
AUX
[future]
affix
--
Acquisition, Sign Language, ...
Unidirectional change in sign language
e.g. Aronoff et al; Fisher & Gough; Pfau & Steinbach:
V>ASP, N > AGR,
and L1 Acquisition
e.g. Brown (1973); Josefsson & Håkansson (2000)
Interlanguage: debate as to features
Lardiere (2007), Hawkins (2005), Tsimpli et al (2004)
Pre-human features: place, duration, negation ...
New developments
Eliminate the features that trigger movement
and `blame’ movement on the resolution of a
labeling conflict:
{XP, YP} : one of the 2 must move....
Wrap-up
Brief overview of current generative linguistics
Differences between early – current program:
Simpler but still focus on hierarchical structures
and movement through MERGE.
My interest: Gradual, unidirectional change
provides a window on the language faculty
Current Developments
Some references
Chomsky, Noam 1965. Aspects of a Theory of Syntax. The
Hague: Mouton.
Chomsky, Noam 1995. The Minimalist Program. Cambridge:
MIT Press.
Chomsky, Noam 2013. Problems of Projection. Lingua 130: 3349.
Chomsky, Noam 2014a. Problems of Projection: Extensions.
ms.
Chomsky, Noam 2014b. Transcript of interview by Naoki Fukui
and Mihoko Zushi (March 4, 2014)
Gelderen, Elly van 2011. The Linguistic Cycle. Oxford: Oxford
University Press.
Gelderen, Elly van 2013. Clause Structure. Cambridge:
Cambridge University Press.
Ott, Dennis & Radek Šimík 2015. Alternatives to formal
features. DGfS Workshop description.