LINGUIST List 12.974
Sat Apr 7 2001
Review: Martin et al (eds) Step by Step
Editor for this issue: Simin Karimi <siminlinguistlist.org>
What follows is another discussion note contributed to our Book Discussion
Forum. We expect these discussions to be informal and interactive; and
the author of the book discussed is cordially invited to join in.
If you are interested in leading a book discussion, look for books
announced on LINGUIST as "available for discussion." (This means that
the publisher has sent us a review copy.) Then contact Simin Karimi at
simin
linguistlist.org or Terry Langendoen at terry
linguistlist.org.
Directory
,
Message 1:
Date:
From: <>
Subject:
Roger Martin, David Michaels, and Juan Urigereka (eds.),
(2000), Step by Step: Essays on Minimalist Syntax in Honor
of Howard Lasnik, MIT Press, Cambridge, Massachusetts.
357pp, $42.95.
Reviewed by Ahmad R. Lotfi, Azad University
This collection of essays on minimalist syntax covers a
number of current issues in linguistic theory including
movement, wh-questions, quantifier scope, temporal relations,
relativization, and also representational and derivational
approaches to language. The book is dedicated to Howard
Lasnik for his twenty-fifth anniversary at the University of
Connecticut in 1997. Naturally, the reader witnesses his
presence in all essays in this collection. Most probably,
however, that will be true for ANY other collection of
minimalist papers, too! Thank you, Howard, for everything!
CONTRIBUTORS:
Andrew Barss, Zeliko Boskovic, Noam Chomsky, Hamida
Demirdache, Hiroto Hoshi, Kyle Johnson, Roger Martin, Keiko
Murasugi, Javier Ormazabal, Mamoru Saito, Daiko Takahashi,
Juan Uriagereka, Myriam Uribe-Etxebarria, Ewa Willim.
CONTENTS:
Introduction: Some Possible Foundations of the Minimalist
Program (Roger Martin and Juan Uriagereka)
1. Minimalism and Asymmetric Wh-Interpretation
(Andrew Barss)
2. Sometimes in [Spec, CP], Sometimes in Situ
(Zeliko Boskovic)
3. Minimalist Inquiries: The Framework (Noam Chomsky)
4. The Primitives of Temporal Relations (Hamida Demirdache
and Myriam Uribe-Etxebarria)
5. How Far Will Quantifiers Go? (Kyle Johnson)
6. Japanese Complex Noun Phrases and the Antisymmetry Theory
(Keiko S. Murasugi)
7. A Conspiracy Theory of Case and Agreement
(Javier Ormazabal)
8. The Japanese Light Verb Construction and the Minimalist
Program (Mamoru Saito and Hiroto Hoshi)
9. Move F and Raising of Lexical and Empty DPs
(Daiko Takahashi)
10. On the Grammar of Polish Nominals (Ewa Willim)
SYNOPSES AND EVALUATIONS:
Introduction
Some Possible Foundations of the Minimalist Program
(Roger Martin and Juan Uriagereka)
Synopsis
Martin and Uriagereka elaborate on two minimalist theses:
(a) a "weak minimalist thesis", or *methodological
minimalism*, which is not controversial in sciences as it
is a search for "simple and nonredundant theories of the
world (or Occam's razor)" (p.1), and (b) a "strong minimalist
thesis", or *ontological minimalism*, which is concerned
with the question of the "optimal design" of language faculty
(FL) rather than the issue of "best theory" for it. There is
nothing new in the Minimalist Program as far as (a) is
concerned. The program is innovative in that it heavily
relies on the strong minimalist thesis, which has nothing to
do with the weaker thesis after all: "[o]ur explanation may
be good, mediocre, or terrible, and it would still be true
that FL works as well as it does, perhaps optimally as just
suggested" (p.2).
Lasnik and Saito's (1984) account of movement is an example
of a theory good enough both in methodological elegance and
empirical coverage. Despite that, Chomsky rejects it in
favor of a methodologically less elegant theory of movement
(which happens to be less adequate in empirical terms, too,
as it does not systematically explain the adjunct/argument
asymmetries any more) because of his reliance on the strong
thesis. For Lasnik and Saito, "Affect Alpha" (instead of
"Move Alpha") affords movement, deletion, and insertion. They
also allow back-and-forth movements. Chomsky, on the other
hand, rejects such a thing, and, as a result, his theory
"needs a new set of axioms, usually referred to as 'last
resort' "(p.10).
Critical evaluation
Chomsky's (now classic) assertion that the generative
linguist should develop a theory of language for an idealized
native-speaker in a homogeneous speech community was
essentially a *methodologically* minimalist move that could
bring about an elegant theory of language divorced from non
linguistic factors necessarily influencing one's linguistic
performance. Optimal design of FL is apparently conceivable
only for such an idealized speaker of language because in
the real-time world, economy is always constrained by such
factors as the listener's demand for the speaker to remain
maximally distinct in what she produces: while the speaker
is interested to get rid of the so called "PF extra-baggage"
as soon as possible, the listener, on the other hand,
prefers the speaker to keep the PF burden on her shoulders.
The trade-off between economy and distinctness seems to have
no place in the Minimalist account of FL optimality of design.
Hence, what Martin and Uriagereka consider as ontological
minimalism is conditional upon methodological minimalism,
which makes the program run into trouble as it implies
that methodology has motivated a priori the design of the
very phenomenon it aims to study! Ontological minimalism, in
this sense, is the consequence of methodological minimalism,
and, as a result, it makes more sense to say it has every
thing to do with the weak minimalist thesis.
Chapter One
Minimalism and Asymmetric Wh-Interpretation
(Andrew Barss)
Synopsis
Barss examines two positions on wh-questions recurrent in
the literature: (1) wh-scope assignment via movement
so that all wh-expressions move to A-bar positions before LF
to be interpreted as restricted interrogative quantifiers
(Chomsky 1981, Higginbotham and May 1981, Huang 1982, and
Lasnik and Saito 1992 among others), and (2) wh-scope
assignment and interpretation without movement as such
expressions are interpreted in situ (Pesetsky 1987, Cinque
1990, Reinhart 1995, and Chomsky 1995 among others). He
hypothesizes that both mechanisms are needed in order to
afford a minimalist account of LF. Barss considers the data
on multipair reading and superiority, crossed binding
asymmetry, asymmetry in wh-QNP interpretations, and
comparative superlative to support his hybrid theory of
wh-interpretation.
Critical evaluation
Barss's account of wh-interpretation seems not to meet the
condition of falsifiability on empirical theories. Barss's
hybrid theory is always valid (and as a result, insignificant)
as any empirical support for either of the positions
mentioned earlier will also support his account of wh
interpretation while any empirical challenge to any of
these two cannot force Barss into a corner. His analyses
are not always informative either. He notices that
(contrary to Chomsky's (1995) Shortest Movement Condition)
both (22) and (23) below are grammatical:
(22) Which man do you think helped which woman yesterday?
(23) Which woman do you think which man helped yesterday?
According to Barss, the paradox "disappears if we permit the
occurrence of [+wh] to be optional on those wh-expressions
that need raise to [Spec, C] for interpretation. [...] Here,
since only one of the interrogative phrases bears the feature
[+wh], the violation of the Shortest Movement Condition is
only apparent" (p.39). However, he does not explain why in
each case the expression that bears [+wh], and not the other,
must do so. This is as informative as saying both (22) and
(23) are grammatical because we've got some grammatical
option here.
What would happen to Barss's analysis if English allowed
(22) but not (23), or vice versa? All one had got to do then
was drop "optional" and say one or the other wh-expression
MUST bear [+wh]. Apparently, we do not even need the Shortest
Movement Condition any more. It will suffice to say which
wh-expression bears [+wh] so that it can move to [Spec, C].
Chapter Two
Sometimes in [Spec, CP], Sometimes in Situ
(Zeliko Boskovic)
Synopsis
Boskovic examines French as a "mixed language", in contrast
with English, Bulgarian, and Japanese, in the sense that it
displays both wh-fronting and wh-in-situ patterns in its
multiple questions. Although wh-fronting is always possible,
the other pattern is rather constrained as it is not allowed
in embedded questions. French and Serbo-Croatian (SC) are
parallel in that "[w]here French does not have syntactic
wh-movement, SC does not exhibit Superiority effects" (p.
57). They differ in that SC wh-phrases must necessarily
undergo fronting when they do not move overtly to [Spec, CP]
while they remain in situ in French.
Boskovic's LF C-insertion analysis of such mixed languages
assumes that strong features may be inserted at LF provided
that they are checked and deleted immediately. The com
plementiser itself is inserted at the root of the tree as a
legitimate instance of Merge, which cannot take place in
embedded positions as it must necessarily expand structure.
The strong [+wh] feature, however, is inserted later at LF.
It prevents the wh-in-situ pattern as it would either
require the covert expansion of the embedded question via
Merge or the overt movement of the wh-phrase. The overt
movement proves to be obligatory in embedded questions as
merge cannot take place under such circumstances. The LF
C-insertion analysis is superior to the optionally strong/
weak [+wh] feature analysis since the former assumes that
[+wh] feature is always strong in French and SC.
Critical evaluation
Boskovic's LF C-insertion analysis is an interesting solution
to both theoretical and empirical problems. The solution is
still minimalist in that it allows lexical insertion to take
place at LF only under certain conditions: "[I]f an NP such
as *John* is inserted in LF, the derivation crashes because
LF cannot interpret phonological features of *John*. If, on
the other hand, *John* is inserted in PF, PF will not know
how to interpret the semantic features of *John*" (p.55).
Boskovic then concludes that "[t]he only way to derive a
legitimate PF representation and a legitimate LF represen
tation is for *John* to be inserted before the level of
S-Structure" (p.55). It follows that only PF insertion of
semantically superfluous lexical items and LF insertion of
phonologically null elements are allowed in the standard
minimalist analyses.
Despite that, it is still possible to conceive of a
radically minimalist model (Lotfi, ms.) in which there is
only one single interface level where phonetic, formal, and
semantic features are all interpreted by the relevant
external systems so that compatibility between a feature,
say the formal feature Q, and some external system like C-I
(and also its incompatibility with A-P), will ensure a
legitimate interpretation. In such a "unitarianist" model of
language, a derivation crashes iff at least one of its
features (no matter whether semantic, formal, or phonetic)
is left uninterpreted when the process of interpretation at
Semantico-Phonetic Form (SPF) is over: hence no need to worry
about the question of LF-PF mapping. The model can explain
(among other things) *John*'s problem of insertion (and also
the empirical questions raised in this chapter) as well as
Boskovic's without necessitating the split of the interface
into PF and LF.
Chapter Three
Minimalist Inquiries: The Framework
(Noam Chomsky)
Synopsis
Chomsky's Minimalist Inquiries: the Framework--although
still exploratory like The Minimalist Program (1995)--is
intended to be a major rethinking of MP issues and 'a clearer
account and further development of them' (p.89). A set of
fresh terms (e.g. EPP-feature and Agree) are introduced in
order to explain the complexities of the functioning of
language faculty. The chapter develops Chomsky's MP thesis
of movement according to which elements move in the fulfillment
of certain morphological requirements. Chomsky assumes
movement--or "dislocation", the term Chomsky prefers here--to
be an (apparent) "imperfection of language" or a "design flaw"
which makes the strong minimalist thesis untenable. Chomsky
assumes two examples (of such imperfections) to be
uninterpretable features of lexical items and the dislocation
property.
The role of the minimalist program for the syntactic theory
remains to be the discovery of the mechanisms that force
dislocation. Chomsky considers the uninterpretable inflectional
features as the devices employed to yield dislocation.
In other words, he thinks these two imperfections might
reduce to one, the dislocation property while dislocation
might itself be required by design specifications. Chomsky
recognizes that "[g]iven some apparent property P of
language", it is possible that "(a) P is real, and an
imperfection", "(b) P is not real, contrary to what had been
supposed", and "(c) P is real, but not an imperfection ..."
(p. 112). Although the strongest conclusion (c) is less
likely than others to be correct, it is the most interesting
one, and as a result, a novel question worthy of asking.
Critical evaluation
Apart from terminology, Chomsky's thesis of movement developed
here does not seem to be significantly different from the
original ideas discussed earlier in his MP. It is rather an
attempt to recast MP terminology in new ones without a serious
attempt to define them clearly. Chomsky seems to recognize
this strategy himself in note 108 (while discussing the
featural composition of AGR): "In MP, it (that Agr consisting
only of uninterpretable phi-features should be disallowed,
Lotfi) could be avoided only by recourse to the (dubious)
distinction between deletion and erasure" (p.151). In absence
of clearly stated definitions and empirically motivated
distinctions, there will always remain a chance for one's use
of new terms and distinctions to prove later to be recourse to
dubious ones.
Chomsky seems to dispense with the concept strength altogether
saying: "The concept strength, introduced to force violation of
Procrastinate, appears to have no place. It remains to
determine whether the effects can be fully captured in
minimalist terms or remain as true imperfections" (p. 132).
Despite that, he immediately introduces a new term--EPP-
feature-- which is functionally similar (at least as far
as movement is concerned) to strength as formulated in MP,
and a new operation--Agree-- in order to explain the
mechanisms underlying movement:
The operation ... Move, combining Merge and Agree(,) ...
establishes agreement between alpha and F and merges P(F)
(generalized "pied piping") to alphaP, where P(F) is a
phrase determined by F ... and alphaP is a projection headed
by alpha. P(F) becomes SPEC-alpha. ... All CFCs (core
functional categories) may have phi-features (obligatory
for T, v). These are uninterpretable, constituting the core
of the systems of (structural) Case-assignment and
"dislocation" (Move). ... Each CFC also allows an extra
SPEC beyond its s-selection: for C, a raised wh-phrase; for
T, the surface subject; for v, the phrase raised by Object
Shift (OS). For T, the property is the Extended Projection
Principle (EPP). By analogy, we can call the corresponding
properties of C and v EPP-features, determining positions
not forced by the Projection Principle. EPP-features are
uninterpretable ... though the configuration they establish
has effects for interpretation. (pp. 101-2)
He then formulates the configuration (22) below for CFCs "with
XP the extra SPEC determined by the EPP-features of the
attracting head H" (p.109):
(22) alpha = [XP [ (EA) H YP ]]
Typical examples of (22) are raising to subject (yielding
(23A)), Object Shift (OS, yielding (B), with XP= DO and t its
trace), and overt A'-movement (yielding (C), with H = C and XP
a wh-phrase ... :
(23) a. XP - [T YP]
b. XP - [SU [ v [V t ]]]
c. XP - [C YP]
The EPP-features of T might be universal. For the phase heads
v/C, it varies parametrically among languages and if available
is optional. ... [T]he EPP-feature can be satisfied by Merge
of an expletive EXPL in (23a), but not in (23b) and (23c)"
(p. 109).
The LI thesis of movement (like his MP thesis) merely states
that things move simply because some mysterious EPP-features
up there make them move (as strong features did in his MP
account of the thesis). And by EPP-features he means those
features we understand must be there because of the raising of
an element to the new position. Since choice of Move over
Agree follows from presence of EPP-features, and since such
features are uninterpretable ones presumably doomed to
deletion in the course of derivation, we are once more left
with the question of why they should be there after all.
Chomsky's allusion to "certain semantic properties" involving
dislocated structures seems to have something to do with such
functionalist theories as parsing or theme-rheme structure in
explaining the why of movement. Chomsky has set himself on
the exploration of the mechanisms involved in movement. Then
one may wonder how the nature could anticipate (if it did)
our future need to such (then useless) uninterpretable
features as that part of the computational mechanism we will
happen to employ later when we want to move things for
meaning's sake. One possibility is that such features evolved
later to take care of our already existing needs to
communicate meaning. The other possibility, which is more in
line with the ideas expressed in Gould (1991) and Uriagereka
(1998), is to consider uninterpretability an exaptation--a
property of the language faculty that was NOT adapted for
its present function, i.e. affording movement so that certain
semantic effects are achieved, but later co-opted for that
purpose. Uninterpretability as an adaptation must not be
particularly attractive to Chomsky as it implies that
uninterpretable features, which are illegible to the C-I
system, are still semantically motivated in origin.
Uninterpretability as an exaptation, on the other hand, makes
the proposal less falsifiable than ever.
Chapter Four
The Primitives of Temporal Relations
(Hamida Demirdache and Myrian Uribe-Etxebarria)
Synopsis
This article is an attempt to reduce "the grammar of Tense
and Aspect to the same set of universal substantive and
structural primitives" (p. 157). For Demirdache and Uribe-
Etchebarria, tense relates UT-T (the time of utterance) and
EV-T (the event time) with T as the head of TP with two
time-denoting phrases as the arguments. T can have the
meaning of WITHIN (present: UT-T WITHIN EV-T), AFTER
(past: UT-T AFTER EV-T), or BEFORE (future: UT-T BEFORE
EV-T). Aspect, on the other hand, is concerned with the
internal temporal constituency of the event. It follows that
the assertion time (AST-T) as "the time interval in the event
time of the VP that Aspect focuses" (p.161) is a key term for
the description of aspects. Aspect is then different from
tense in that the former relates EV-T to AST-T while the
latter relates AST-T to UT-T:
TP
/\
UT-T T'
/\
T-0 AspP
/\
AST-T Asp'
/\
Asp-0 VP
/\
EV-T VP
Rejecting the Reichenbachian analyses of tense and aspect,
the authors consider these spatiotemporal relations to be
those of central/noncentral coincidence comparable with
prepositions expressing similar spatiotemporal relations,
hence:
Central coincidence
(a) [I ////// F] (where I and F stand for initial and
final bounds of the event) e.g. Progressive
Noncentral coincidence
(b) //////[I F] e.g. Prospective
(c) [I F]////// e.g. Perfect
Critical evaluation
The article possesses genuine theoretical merits. Moreover,
the model is NOT incompatible with minimalist accounts of
language. Despite that, it is not a minimalist essay in
the strict sense of the word. The major thing the model has
in common with minimalism is some economy flavor accompanying
their Constraint on Aspect Recursion--no vacuous
viewpoint shifting is allowed. Apart from this, the model
seems to be more at home with Bybee, Perkins, and Pagliuca's
(1994) accounts of tense and aspect. It seems to be a good
example of cases where both formalist and functionalist
accounts of language converge.
Chapter Five
How Far Will Quantifiers Go?
(Kyle Johnson)
Synopsis
Focusing on the current accounts of Quantifier Raising (QR),
Johnson tries to reveal the empirical inadequacy of the
idea that QR can be reduced to A-movement (Hornstein 1995).
More specifically, Hornstein suggests that both subject
lowering and object raising are at work in order for the
object quantifier to gain wider scope than a local subject
in (1b) below:
(1) a. [IP Someone [VP loves everyone]].
b. [IP everyone-1 [IP someone [VP loves t-1]]]
c. [IP someone-2 [IP everyone-1 [IP t-2 [VP loves t-1]]]]
He argues that on Hornstein's account an object must be
narrower in scope than any other sentential element outside
VP as the object moves into the first available specifier
position above VP. According to Johnson, however, this is
not true as (13) is still ambiguous due to *many of the
questions* having wider/narrower scope that *not*:
(13) Gary didn't answer many of the questions on the exam.
Instead, Johnson suggests that QR is like scrambling in
German and Dutch: (a) QR moves objects out of embedded
infinitival clauses, (b) a complementizer in the infinitival
clause blocks the operation, (c) there is no QR when the
embedded clause is not infinitival, and (d) scrambling can
also relocate adjuncts as QR does.
Critical evaluation
Johnson's is a genuinely minimalist analysis as it tries to
collapse an operation (QR) into another independently
motivated one (scrambling). Despite that, it is not clear
why QR must prove to be similar to this specific operation
and nothing else.
For Chomsky, movement--an imperfection of language--takes
place in the fulfillment of certain morphological requirements
- uninterpretable features; another imperfection of
language. While the existence of some morphological
constraints on scrambling is NOT inconceivable after all,
it does not necessarily mean that scrambling is morphologically
motivated in the Chomskyan sense of the word.
Johnson's analysis raises certain expectations with regard
to some common (morphological) source for both QR and
scrambling, which are not satisfied in this article. Even
at the level of constraints, it is not clear at all how QR
and scrambling can share featural characteristics. Then
Johnson's analysis fails to go beyond a mere analogy drawn
between QR and scrambling. It seems to be far from
"collapsing" one operation to another.
Chapter Six
Japanese Complex Noun Phrases and the Antisymmetry
Theory
(Keiko S. Murasugi)
Synopsis
Japanese "relative clauses" are known to have very peculiar
properties. Murasugi argues that relative clauses are not
allowed in the language. Instead, the language employs pure
sentential modifiers to modify the nominal heads extensively,
and, as a result, can express relative clause meaning via
pure complex NPs. But why is it so?
Kayne (1994) proposes the structure in (1) below for N-final
relative clauses derived by (a) the movement of the relative
head to [Spec, CP], and (b) the movement of IP to [Spec, DP].
(1) [DP [IP ... t-i ...]-j [D' D [CP NP-i [C' C t-j]]]]
Movement to [Spec, CP] is in general A-bar movement. As a
result, it proves to be a violation of the Proper Binding
Condition, which is not allowed in Japanese "relative clauses"
due to the finiteness of their main verbs. Relative clauses
are simply impossible in Japanese because of this nonmovement
property of Japanese "relatives".
Critical evaluation
Murasugi's chapter is quite well-organized, empirically rich
and theoretically intriguing. Her use of the evidence from
child Japanese to support her theory-internal analysis is
especially interesting. The essay is not minimalist in the
strict sense of the term, however. It is written within the
principles-and-parameters general framework with heavy
reliance on Kayne's Antisymmetry Theory. No use is made of
other concepts and mechanisms recurrent in the standard
minimalist literature such as those introduced in Chomsky
(1993-2000). The article is unique in that it is the only
essay in this collection with no single reference to
Chomsky's work.
Chapter Seven
A Conspiracy Theory of Case and Agreement
(Javier Ormazabal)
Synopsis
Ormazabal focuses on the relation between case checking and
movement. Perlmutter (1971) had already noticed the well-
formed combination of clitics in Romance as 1DAT+3ACC and
how it contrasted with the ill-formed 3DAT+1ACC. Ormazabal
also notices the well-formed 1ABS+3DAT in Basque, contra
Bonet's (1994) descriptive generalization according to
which if DAT, then ACC/ABS = 3rd person.
Two competing proposals--Albizu's (1997) morphological
analysis (Person-Case Constraint) and the Minimal Link
Condition (MLC)--are examined to find out how well each of
them accounts for such cooccurrence restrictions. Ormazabal
finds Albizu's account conceptually redundant, and favors a
syntactic account in terms of the MLC. Based on some empirical
evidence from Romance, which reveals a contrast between
strong and weak pronouns, he concludes that it is the feature
[+animate], and not Case, that can explain such phenomena.
"[T]he movement of arguments from the VP shell is not driven
by Case-checking considerations, but by the need to check an
animacy feature: only [+animate] arguments are forced to
move from their base-generated position" (p.248).
Critical evaluation
Ormazabal's interesting empirical evidence for the role
[+animate] plays in movement raises questions on the
sustainability of minimalist thesis of movement and also
checking theory. In standard minimalist accounts, it is an
uninterpretable feature that is typically checked and
finally deleted to prevent crash at interface levels.
Actually Chomsky (this volume) collapses these two so-called
imperfections of language, namely uninterpretability and
dislocation, into one so that the former brings about the
latter. Now Ormazabal tells us that it is not an uninterpretable
feature that initiates movement but the interpretable
[+animate]: "unlike syntactic Case, which is a theory-
internal construct, features like [+animate] point to a
conception of checking operations where the feature checked
is interpretable in either the target or the attracted
element, and it survives to one of the interface levels"
(p.254).
The proposal boils down to the claim that movement takes
place after all in the fulfillment of some checking requirement,
which is not a challenge to Chomsky's thesis of
movement. The point of departure from the standard theories
of movement then proves to be the status of the feature (to
be checked) in terms of interpretability. This raises two
problems for Ormazabal's account, one conceptual, and the
other technical. For Chomsky, interpretable features need
not be checked (1995:289) because it is always the uninterpretable
feature that is an illegitimate object at interface
levels, hence to be checked and deleted. But for Ormazabal
the urge for checking [+animate] is so compelling that
checking this interpretable feature via movement becomes a
forced move. The conceptual question is what this checking
is after all that forces such a move. If checking is not
the deletion of an uninterpretable in order to prevent the
crash of the derivation at interface levels, what is it? The
technical problem is that in his examples for animacy
effects, it must be the [+animate] feature of the moved
element (the strong pronoun/1st and 2nd person pronouns) that
is always checked through movement rather than that of the
target: in the grammatical sentences with a cliticized
pronoun (which does not exhibit the [+animate] feature)
the [+animate] feature of the target (if there is such a
feature there after all) is necessarily left unchecked due
to the absence of any other element with the [+animate]
feature to do the checking. This is more compatible with
Chomsky's (now outdated) formulation of Greed as here alpha
moves to satisfy its own checking requirements. Then
Ormazabal's use of the term "attracted element" is rather
odd.
Chapter Eight
The Japanese Light Verb Construction and the Minimalist
Program
(Mamoru Saito and Hiroto Hoshi)
Synopsis
In a Japanese light verb construction one or more arguments
of the theta-assigning noun appear as clausal arguments of
the non-theta-assigning light verb *su* while the others
remain within the projection of the noun:
(1) Mary-ga John-ni [NP toti-no zyooto]-o sita.
Mary-NOM John-to land-GEN giving-ACC did
'Mary gave a piece of land to John.'
According to Grimshaw and Mester's (1988) argument transfer
analysis, the noun *zyooto* 'giving' transfers its agent
(Mary) and goal (John) theta-roles to the verb *sita* (= su
+ ta 'past') while the theme argument *toti* 'land' remains
within the NP headed by the theta-role-assigning noun. They
note the following restrictions on such constructions:
(2) a. At least one internal theta-role of the noun must be
assigned to an argument outside the NP.
b. If a theta-role T is assigned outside the NP, then all
theta-roles that are higher than T in the thematic
hierarchy must also be assigned outside the NP.
Saito and Hoshi believe that this analysis "is a description
of a fact rather than an explanation" (p.273). Moreover, "it
is not clear why the thematic hierarchy must be observed,
after argument transfer takes place, among the arguments of
two independent theta-role assigners" (p.274) namely the
theta-role assigning noun and the light verb *su*. Instead
they argue for an LF incorporation analysis in which the
noun *zyooto* is assumed to assign its theme theta-role
within the NP and then raise to the position of *su* in LF
and discharge its agent and goal theta-roles there. This
new analysis supports Chomsky's (1995) minimalist program
according to which D-Structure, S-Structure and the
Projection Principle are eliminated. It also favors the
Last Resort condition on movement.
Critical evaluation
While the criticisms Saito and Hoshi level at Grimshaw and
Mester's argument transfer analysis are not unreasonable,
it is somehow doubtful that their own LF incorporation
analysis is not a more abstract and less accessible
description of the same facts that Grimshaw and Mester's
analysis tries to capture. That theta-roles are formal
features of the theta-assigners which enter checking
relations is apparently a descriptive label itself rather
than a genuine explanation of why things work that way.
Actually what Grimshaw and Mester assume to have taken place
at the abstract and invisible interface level of D-Structure
is now shifted in position to another interface level equally
abstract and invisible. Although the analysis helps to
dispense with D-Structure, it does so at the price of
interpreting theta roles after Spell-Out in positions different
from the base. The conceptual problem is that this makes
PF-LF mapping increasingly difficult to explain. In Grimshaw
and Mester's analysis, on the other hand, the transfer takes
place at D-Structure, hence a part of the common heritage
of PF and LF.
Chapter Nine
Move F and Raising of Lexical and Empty DPs
(Daiko Takahashi)
Synopsis
The major question Takahashi addresses in this essay is why
a category alpha (and not some feature F to be checked) must
move into the checking domain. As Takahashi put it, Chomsky
(1995) assumes two conditions with regard to the movement
of F:
(1) F carries along just enough material for convergence.
(2) Words whose features are isolated or scattered may not
be subject to PF rules, making the derivation crash.
Condition (2), or Takahashi's PF Integrity Condition, makes
two predictions. First, "since the PF Integrity Condition
is simply inapplicable in LF, LF movement should be
restricted to feature movement" (p.298). Second, "if alpha, a
category containing a feature F that is about to enter into
a checking relation, is phonologically null ..., then
movement of F alone out of alpha should be possible ... .
This leads to the expectation that other things being equal,
empty categories should behave differently from their
lexical counterparts when they contain features that need to
undergo checking" (p.298). In this chapter Takahashi uses the
data from a raising construction in Japanese to confirm the
second prediction.
(1) F carries along just enough material for convergence.
(2) Words whose features are isolated or scattered may not
be subject to PF rules, making the derivation crash.
Condition (2), or Takahashi's PF Integrity Condition, makes
two predictions. First, "since the PF Integrity Condition
is simply inapplicable in LF, LF movement should be
restricted to feature movement" (p. 298). Second, "if alpha, a
category containing a feature F that is about to enter into
a checking relation, is phonologically null ..., then
movement of F alone out of alpha should be possible ... .
This leads to the expectation that other things being equal,
empty categories should behave differently from their
lexical counterparts when they contain features that need to
undergo checking" (p. 298). In this chapter Takahashi uses the
data from a raising construction in Japanese to confirm the
second prediction.
There are two types of empty pronominal subjects in
Japanese: one type is called argumental pro as it receives a
theta-role, and is exemplified in (3) below (his 8); the
other is quasi-argumental pro as it does not bear any theta-
role (e.g. in (4) below (his 16):
(3) pro Hanako-o sikatta.
Hanako-ACC scolded
'I/WE/You/He/She/They scolded Hanako.'
(4) pro sigureta.
rained
'It rained.'
Takahashi observes that quasi-argumental pros cannot raise
while argumental ones behave like lexical DPs. Consequently,
argumental pros raise to control PRO while quasi-argumental
ones cannot:
(5)
pro-1 [PRO-1 biiru-o nomi nagara] Hanako-o sikari
beer-ACC drinking while Hanako-ACC scolding
hazimeta.
began
'I/We/You/He/She/They began scolding Hanako while drinking
beer.' (with the meaning that at some point during
drinking beer, I/we/you/he/she/they began scolding
Hanako)
(6)* pro-1 [PRO-1 sigure nagara] t-1 fubuki hazimeta.
raining while snowing began
'It began snowing while raining.' (with the meaning that
at some point during the rainfall, snow began to fall)
Based on this asymmetry in raising between these non-
argumental pros and their lexical counterparts, Takahashi
argues that the asymmetry follows from the Move F hypothesis.
Apparently, the raised F lacks some qualification
necessary to control. The other possibility is that features
cannot control at all. Whatever the case, it confirms the
second prediction as null subjects (quasi-argumental pros)
differ from their lexical counterparts in raising. "Although
argumental pros appear not to conform to the prediction
of (the) ... analysis, ... they do not undermine it, since
the seemingly unexpected cases may not involve raising at
all" (p.314).
Critical evaluation
The central question here seems to be why the feature F
contained in the element alpha cannot raise without the PF
material pied-piped to it *prior to Spell-Out* as raising F
without the PF material at LF is already taken to be borne
out in the literature Takahashi mentions in his chapter.
Takahashi's focus on null subjects is an attempt to control
this extraneous factor of PF features. He correctly
hypothesizes that under such circumstances "the movement of F
alone out of alpha should be ... obligatory ... even in
overt syntax" (p. 298) as empty categories have no PF
features at all to be pied-piped to their formal features. As
a result, one expects the feature F of a null subject to
raise when needed without making the sentence ungrammatical.
However, it is not quite clear how Takahashi takes the
impossibility of raising quasi-argumental pros to support the
hypothesis that a phonetically null element may have its
formal features raised. On the contrary, this could mean
that a quasi-argumental pro (nor one of its formal features)
can raise in overt syntax even if no PF features are pied-
piped to it. (6) above (with the interpretation specified) is
ungrammatical NOT because features cannot control but that
they do not get scattered/isolated even then. In other words,
argumental pros, quasi-argumental pros and lexical DPs
disallow some formal feature of themselves to raise on its own.
They differ in that DPs and argumental pros raise while
quasi-argumental ones don't.
Chapter Ten
On the Grammar of Polish Nominals
(Ewa Willim)
Synopsis
Willim investigates the relevance of the DP analysis of
nominal structure to the grammar of nominals in Polish,
an articleless language with rich nominal inflection for
gender, number, and morphological case. The empirical data,
including the absence of lexical articles, and no clitic
attaching to the genitive subject, Willim concludes that
"there is neither functional morphology to support the
functional head D in Polish, nor evidence for a
phonologically unrealized but syntactically active D head
in the language" (p.330).
In languages with the D paradigm, extraction from NP takes
place via [Spec, DP] so that features can be checked in the
configuration headed by D. It follows that such extractions
are sensitive to the Specificity Condition. Willim again find
no direct evidence on the availability of D as "a head
providing a position targeted by a phrase undergoing overt
movement in Polish" (p.332).
Longobardi (1994) assumes that the feature [+/- R] marks
the element in D position. For proper nouns and pronouns
the value of the feature is positive as they refer to
objects. In Italian, [+R] is strong while "[i]n English [+R]
is weak, banning overt N-to-D raising ..." (p.337). Empirical
data from Bemba and Basque suggest that (in)definiteness of
reference is not the function of D. It follows that the
existence of the DP layer at LF is not logically necessary
for Polish nominals. "Rather, (in)definiteness of reference
may be directly related to the properties of syntactic
configurations in which nominals find themselves at LF in
Polish" (p.340).
Critical evaluation
To my relief, Willim does NOT conclude that although DP is
missing in Polish overt syntax, there is some DP projection
at its LF so that the highest level of generalizability is
guaranteed for our minimalist theories and mechanisms!
Although Willim does not dismiss LF as the level at which
(in)definiteness and also the reference of nominals are
determined, she rejects the necessity of resorting to
D-insertion at LF and also the existence of some DP layer
there for doing such things. Instead, the properties of LF
syntactic configurations in which nominals (and not Ds) are
inserted will suffice.
References
Albizu, P. 1997. Generalized Person-Case Constraint: A case
for a syntax-driven inflectional morphology. In A.
Mendikoetxea and M. Uribe-Etxebarria, eds. Theoretical Issues
on the Morphology-Syntax Interface, 1-34. ASJU, San Sebastian.
Bonet, E. 1994. The Person-Case Constraint: A morphological
approach. In H. Harley and C. Phillips, eds. The Morphology-
Syntax Connection, 32-52.
Bybee, J., R. Perkins, and W. Pagliuca. 1994. The evolution
of grammar: Tense, aspect, and modality in the languages of
the world. Chicago: University of Chicago Press.
Chomsky, N. 1981. Lectures on Government and Binding.
Dordrecht: Foris.
Chomsky, N. 1995. The Minimalist Program. Cambridge, Mass.:
MIT Press.
Cinque, G. 1990. Types of A-bar Dependencies. Cambridge,
Mass.: MIT Press.
Gould, S. J. 1991. Exaptation: A crucial tool for
evolutionary psychology. Journal of Social Issues 47, 43-65.
Grimshaw, J. and A. Mester. 1988. Light verbs and theta-
marking. Linguistic Inquiry 19, 205-232.
Higginbotham, J. and R. May. 1981. Questions, quantifiers,
and crossing. The Linguistic Review 1. 41-80.
Huang, C-T. J. 1982. Logical relations in Chinese and the
theory of grammar. Doctoral dissertation, MIT, Cambridge,
Mass.
Kayne, R. 1994. The Antisymmetry of Syntax. Cambridge, Mass.:
MIT Press.
Lasnik, H., and M. Saito. 1984. On the nature of proper
government. Linguistic Inquiry 15, 235-289.
Lasnik, H., and M. Saito. 1992. Move Alpha. Cambridge,
Mass,: MIT Press.
Longobardi, G. 1994. Reference and proper names. Linguistic
Inquiry 25, 609-665.
Lotfi, A. R. (ms.) Semantico-Phonetic Form: A unitariainist
grammar. Available on Arizona Minimalist Syntax Archives at
<http://www.minimalism.linguistics.arizona.edu/AMSA/PDF/
AMSA-186-1200.txt>.
Perlmutter, D. 1971. Deep and Surface Structure Constraints
in Syntax. New York: Holt, Rinehart and Winston.
Pesetsky, D. 1987. Wh-in-situ: Movement and unselective
binding. In E. Reuland and A. ter Meulon, eds., The
Representation of (In)Definiteness. Cambridge, Mass.: MIT Press.
Reinhart, T. 1995. Interface strategies. (OST Working Papers
of Theoretical Linguistics 95-002.) Utrecht Institute of
Linguistics OTS, Utrecht University.
Uriagereka, J. 1998. Rhyme and Reason: An Introduction to
Minimalist Syntax. Cambridge, Mass.: MIT Press.
Reviewer:
Dr. Ahmad R. Lotfi, Chair of the English Department, Azad
University at Esfahan, Iran, where he teaches linguistics
to graduate students of ELT. His research interests lie in
minimalist syntax, second language acquisition studies in
generative grammar, and Persian linguistics.