LINGUIST List 12.3131

Tue Dec 18 2001

Review: Cinque & Salvi, Studies in Italian Syntax

Editor for this issue: Terence Langendoen <terrylinguistlist.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 siminlinguistlist.org or Terry Langendoen at terrylinguistlist.org. Subscribe to Blackwell's LL+ at http://www.linguistlistplus.com/ and donate 20% of your subscription to LINGUIST! You get 30% off on Blackwells books, and free shipping and postage!

Directory

  1. Manideepa Patnaik, Review: Current Studies in Italian Syntax

Message 1: Review: Current Studies in Italian Syntax

Date: Tue, 18 Dec 2001 10: 12: 22 -0800 (PST)
From: Manideepa Patnaik < com">manideepapyahoo. com>
Subject: Review: Current Studies in Italian Syntax

Cinque, Guglielmo, and Giampaolo Salvi, ed. (2001) Current
Studies in Italian Syntax: Essays Offered to Lorenzo Renzi.
North-Holland, hardback ISBN 0-08-043874-1, xii+326pp,
$91.00, North-Holland Linguistic Series: Linguistic
Variations 59.

Manideepa Patnaik, Harvard University

The sixteen contributions which make up this volume
are representative of the research currently carried
out in Italy on Italian and, more generally, Romance
syntax (in generative tradition). The essays were
specially collected to pay homage to Professor Lorenzo
Renzi, a scholar who has since the 1960s promoted and
shaped the study of Italian syntax in Italy, both
through his own work and through a collective
enterprise that culminated in the publication of the
Grande Grammatica Italiana di Consultazione (3 vol. ,
Bologna, II Mulino, 198-1995). Most of contributors to
this volume were engaged in that enterprise as young
unemployed linguists, and are now among the most
prominent specialists in the field of Italian syntax.
The target audience of this book includes syntactitians,
Romance linguists, and especially students of the
Italian language.


CONTENTS
1. Syntactic intervention effects on Italian polarity items
 (P. Acquaviva)
2. Speculations on the possible source of expletive
negation in Italian comparative clauses (A. Belletti)
3. The position of topic and focus on the left periphery
(P. Beninca)
4. Aspect prefixes in verbal periphrases in Italian and
other Romance languages (F. Benucci)
5. "Propulsive" tenses in modern Italian fiction prose
(P. M. Bertinetto)
6. A second thought on emarginazione: distressing vs.
"right dislocation" (A. Cardinaletti)
7. "Restructuring" and the order of aspectual and root
modal heads (G. Cinque)
8. The birth of a functional category: from Latin ILLE
to the Romance article and personal pronoun (G. Giusti)
9. Romance causatives and dynamic antisymmetry
(M. T. Guasti, A. Moro)
10. A note on clitic doubling in French
(R. S. Kayne)
11. Either "subject oriented" or merely sentential
(L. Lonzi)
12. The syntax of object clitics: si in Italian dialects
(M. R. Manzini, L. M. Savoia)
13. Complementizer deletion and verb movement in standard
Italian (C. Poletto)
14. On the position "int(errogative)" in left periphery
of the clause (L. Rizzi)
15. The two sentence structures of Early Romance
(G. Salvi)
16. Evidence for a null locative in Italian
(C. M. Tartora)

ANALYSIS
The contributions by Bennica, Poletto and Rizzi
approach the fine structure of the left periphery of
the Italian clause, shedding new light on the
hierarchy of the functional projections that make up
the CP "field". Rizzi provides evidence for an
Interrogative Phrase projection (hosting se 'whether'
and perche 'why') distinct from, and lower than, his
(1997) ForceP, and higher than his (1997) FocusP and
the projection to which wh-phrases move (in embedded
questions). Poletto discusses evidence for a subject
position within the CP "field", in the broader context
of complementizer deletion in Italian. Beninca provides
arguments that TopicP is invariably higher than
FocusP, despite certain appearances, and that
exclamative wh-phrases are hosted in a projection
higher than TopicP.

The contributions by Cardinaletti, Tartora, and Guasti
and Moro, all deal with different word order problems
from an " antisymmetric" point of view. Cardinaletti
provides evidence for treating the marked orders VS, O
and VO, S (with the last constituent "distressed")
quite differently one from the other: the former from
the focusing of the subject of the subject in situ
(within VP); the latter from a right-dislocation of
the subject, in compliance with Antisymmetry. Tartora
argues for the necessity of distinguishing two types
of VS orders with unaccusatives verbs, depending on
the presence of an overt (or covert) goal. Guasti and
Moro analyze certain word order and other properties
to the Italian causative construction as consequences
of a particular interpretation of Antisymmetry
recently proposed by Moro.

The contributions by Benucci, Bertinetto, and Cinque
relate instead to the functional portion of the
clause, with special reference to tense, aspect and
modality. While Bertineto focuses on narrative uses of
tense, and the apparent replacement in contemporary
literary Italian of the simple past with other tenses,
Benucci and Cinque concern themselves with the
structural representation and order of these
functional notions in the clause. Cinque attempts to
provide an ordering of aspectual and modal heads that
is more finely grained that the one proposed in his
previous (1999) work, with evidence of a novel kind,
based on a new analysis of "Restructuring". Benucci
traces the diachronic development of a number of
aspectual heads in Romance expressed as prefixes on
verbs and documents that existence of " prefix
climbing" in restructuring contexts.

Also concerned with functional projections and
dianchrony are Giusti's and Salvi's contributions.
Giusti analyses the rise of articles and clitics from
Latin demonstratives via a process that reinterprets a
phrase in specifier position as an X0 in the
adjacent head position. Salvi discusses a little known
subordinate sentence type of early Romance, later
superseded by the Verb Second structure of the main
clauses) interpreting it as a residue of Latin
sentence structure.

The contributors by Belletti and Acquaviva deal with
different aspect of negation in Italian. Belletti
approaches the question of expletives negation in
(subjunctive) comparative clauses, deriving it, and
the degree phrase piu 'more', via movement from the
discontinuous negative elements non...piu 'not...
anymore' in the comparative clause. Acquaviva analyses
the licensing conditions of polarity items in Italian
(including overtly negative phrases), suggesting that
like in English they must be found to their licensing
operator at LF.

Lonzi's contribution considers a number of
subject-oriented adverbs in Italian analyzing their
interaction with passive, negation and other
structural properties.

The contributions by Kayne, and by Manzini and Savoia,
deal with the syntax of clitics. Kayne's proposal that
structurally case-marked pronominal in French must be
doubled by a clitic sheds new light on many aspects of
the syntax of tonic and clitic pronouns in French, as
well as in Italian and other Romance varieties.
Manzini and Savoia discuss the status of si in its
different interpretations and consider its ordering
within the clitic cluster in many different dialectal
varieties.

Paolo Acquaviva's chapter discusses in some
detail if and to what extent Italian polarity
items (including N-words) are subject the same
semantic/pragmatic intervention effect called
Immediate Scope Constraint (ISC) documented by
Linebarger 1987 for English polarity items; second, to
explore the significance of the resulting pattern for
the general issue of polarity sensitivity. With
respect to these two related questions, the data
examined suggest, respectively, that the scopal
intervention effect proposed by Linebarger does in
fact hold of Italian dependent items and N-words
alike, and that the ISC must be recognized as a
property of polarity licensing rooted in the LF
representation. The various other constraints on licensing
(chiefly, but not only, Subjacency effects) and the
typology of licensers are not directly relevant to the
discussion, which centers on the effects triggered by
quantifiers and other operators intervening between a
licenser and a polarity item.

Despite appearances, there are good reasons to accept
that the ISC holds on Italian polarity items just the
way it does in English. Polarity items must be
adjacent to their licensing operator at LF in a
syntactic sense: complementizers selected by
adversative predicates, which do not interfere with
the interpretive scopal relations of quantifiers,
count as interveners. This has two consequences:
First, Linebarger (1987) was correct in viewing the
ISC as a syntactic constraint on LF, given that the
ISC is not a special condition on polarity items but
just one instance of a pervasive syntactic respects,
although they are obviously different in other
important respects. This applies in the same way to
N-words and dependent items like alcunche or
chicchessia, which are not marked as negative:
Therefore the negative characterization is irrelevant
for the polarity item-like quality of N-words. In a
different sense, the negative characterization is
irreverent also for another distinction brought out by
the ISC, namely, that between polarity items that can
and cannot be existentially quantified. Polarity
items, then, are not just restricted variables, or
indefinites obligatorily in the scope of a licensing
operator, but can be quantified so long as the
semantic and pragmatic conditions on their licensing
are met. This conclusion is perhaps not surprising in
connection with English polarity items, but when
applied to Italian N-words it acquires a new
importance.

The fact that N-words have a different distribution
depending on whether they can or cannot be
existentially quantified implies that they undergo
at LF, hence in the syntax, the "movement"
transformations other indefinites undergo when
existentially quantified. This is different from a
generalized movement to Spec NegP for all and only
negative words (Romance and Germanic), as envisaged,
for example, by Beghelli & Stowell (1997). The
evidence indicates that whether or not all [+negative]
elements raise to NegP, some N-words and some
non-negative polarity items are existentially bound
within the scope of negation. Such elements raise at
LF, which means that existentially quantified N-words
construct with the operator binding then a syntactic
relation with properties of an A-bar chain. How this
relation I subsumed by a general theory of covert
LF-movement, and hoe it interacts with relations
linking all N-words to the negative operator, over and
above existential quantification, is questions has
been left for future research.

Belletti's chapter sketches out a proposal for an analysis
of expletive negation. A large class of comparative clauses
in Italian, identifiable with those introduced by the
wh-word being in the subjunctive mood and reduced
small clause comparatives introduced by the
complementizer allow for the presence of the negative
marker non. Although negation is present in this kind
of comparatives the clause or small clause where it
appears doesn't have a negative value. Indeed, this
kind of negation is often referred to as "expletive
negation". Belletti hypothesizes with adequate
evidence that the subjunctive is the licenser of the
negative marker, yielding its expletive value.
Following various works (notably Pollock 1997), it has
been assumed that clause structure contains a ModP
projection, where "mood" is indicated, normally
realized in the verbal inflection. It is precisely the
modal operator, that is present in the Spec/negP.
Which moves to Spec/NegP and in so doing licenses the
negative marker non through the establishment of a
Spec-Head agreement relation. The subjunctive modal
operator and the negative marker have in common an
irrealis feature. This feature would be at the very
source of the possibility that the required agreement
relation be correctly established. It has been
hypothesized that this feature is at the source of the
"potentially" flavor, typically associated with the
presence of the expletive negation.

The negative marker can also appear in comparative
small clauses introduced by the complementizer. These
comparative small clauses can be attributed the same
analysis as the quanto-comparatives. This is done by
making the assumption that a full, although
non-overtly realized, AgrP clause structure is
attributed to the same clause, also including a NegP
projection. The hypothesis is given further support by
the observed presence of the complementizer che as the
introducer of the small clause, since a complementizer
systematically selects a clause, i. e. an AgrP. Accordig
to the adopted analysis, a NegP headed by the negative
marker non is the source of the degree adverb filling
the spec/DegP of the compared phrase. As no lexical
verb is present within the small clause, the negative
marker (phonologically) attaches to the following
constituent. Since the degree adverb moves to Spec
position of the matrix DegP to check the degree
feature in the head of the DegP, it is immediately
clear why the degree adverb cannot be interpreted as a
negative adverb of the matrix clause. Since the degree
adverb undergoes one checking operation in the matrix
Spec/DegP it cannot further move to Spec/NegP to
establish the agreement relation with negative head,
necessary for its interpretation as a negative adverb
to become available. No such problem involves the
first occurrence of the degree adverb since the degree
feature in the head of the matrix DegP is checked by
the second occurrence of the adverb. There exists in
Italian a kind of a parasitic substandard comparative
utilizing a relative construction rather than
quanto-movement. The mod of the relative/ comparative
clause is systematically the indicative. In this case,
predictably no expletive negation is possible

Paola Beninca assumes a theory of grammar that
conceives syntactic structure as a theoretical
object representing the hierarchical relations
between linguistic elements; its output is linear
sequence of unambiguously ordered elements.
This paper deals with elements that appear on the left
periphery of the CP field. She first sums up the main
lines of Rizzi's theory and then discusses some points
on his analysis, showing that if further data of
Italian will be taken into account then a more
constrained theory and presumably a more faithful
picture of "fine structure of the CP" can be obtained.

The data includes all the elements that can be assumed
to be located in CP, i.e. interrogative pronouns and
phrases; relative pronouns and phrases; exclamative
phrases; thematised elements(possibly binding a
resumptive clitic); focalized elements (with
contrastive intonation). All of these elements occupy
the Spec of a functional projection and they do not
appear to be in complementary distribution except in
strict V2 languages, such as German and Dutch. They
can generally appear together, and therefore it is not
possible to assume a single projection in the left
periphery, or rather a single specifier; various
mechanisms have been proposed in order to account for
the fact that, although certain restrictions are
established, many Spec positions appear to be
available. Solutions such as recursion of CP,
adjunction to the Spec or multiple specifiers, would
not lead us to expect any ordering among different
elements, unless this was due to independent reasons.

If one has to establish a strict ordering between the
elements that can appear in the left periphery, one
has to hypothesize a fine structure with "labeled"
positions, in a way to be determined. The other
elements that are located in CP, whose nature is
clearly that of a head are: complementisers, i. e. , of
generic subordination, of yes/no interrogative, of
infinitival clause. etc; the inflected verb, which
reaches a C head in specific types of sentence
(primarily questions), or in all main clauses in V2
languages. As a conclusion she draws a map of the
syntactic elements in the left periphery that does not
differ very much from the one sketched by Rizzi
(1997). The major findings of this project is, she has
pointed out that TopP is not the only projection where
a Theme can be hosted; Hanging Topic has precise
characteristics and appears to be higher than TopP
than has been labeled as Dis (course) P. She has
provided evidence for a localization of exclamative
wh-: and it seems likely that this location may be the
Spec of ForceP, both for a syntactic and semantic
reason. They wanted most to argue for one thing in
general, namely, that the X-bar module is rigidly
repeated and the positions are very strictly allotted
to a constituent with certain characteristics of a
syntactic or pragmatic nature; it is not necessary,
for the time being to hypothesize the possibility of
adjoining Spec positions to an already present Spec;
contrastive intonation does not reveal a syntactic
position activated, on the contrary, it is able to
attribute a contrastive interpretation to a
constituent, even if it is not moved to Topic. The aim
of this paper was to support a syntactic treatment of
the CP functional structure and to narrow the
possibilities open to it.

The analysis developed in Benucci's paper is centered
on the syntax of aspect prefixes in old and
modern Romance verbal periphrases (with special
stress on Italian, French and Piedmontese). It is based
on and further develops the ideas incorporated in the
extremely rich sentential functional structure
suggested by Cinque (1997). If the results achieved by
Franco Benucci are correct then this work can
constitute an independent supporting evidence for
Cinque's proposals. This scheme allow one to account
in a satisfactory way for other phenomena that occur
in the same syntactic contexts, such as those related
to the so called passivising 'si' in early and modern
Romance languages (which depend on the
presence/absence and on the intrinsic features of
AgrsP in both the matrix and the embedded clause of
the periphrasis, cf. Rizzi 1976,Benucci 1990), the
French phenomenon, a special case of Quantifier
Floating occurring in periphrastic contexts with
Cinque's analysis of sentential functional structure
and may therefore be included in the scheme that is
developed in this paper.

Bertinetto notices the not infrequent"local"
shifts in "core" function from the Simple Past
to other tenses, such as the "epic" Present, the
Compound Past or even) although to a lesser extent)
the Pluperfect. As far as the latter two tenses are
concerned, this innovation obviously presupposes stage
of aoristicization of these tenses in the spoken
language (for the relevant varieties of Italian).
Second, the author notice the emergence of a
'strategic" use of the tenses, consisting in the more
or less systematic deployment of different tenses to
obtain a given "color" in specific sections of the
narrative. Finally, the authors found examples
betraying the attempt, on the part of certain authors,
to render the sense of disorientation afflicting
contemporary mean and women, through the dissolution
of the unitary temporal-aspectual perspective from
which the events are observed.

Viewed against the background of tradition narrative
techniques, the various solutions analyzed in this
paper might paper different instances of "derivation
from the norm". However, this poses serious problems
of interpretation. The question that immediately arises
is: what is the norm contemporary writer start from?
To some extent, it must be the old one, still alive to
a considerable degree, as witnessed by the far from
negligible number of modern writers that adopt it.
However, it is also possible that the example of
various "deviant" twentieth century works may now
offer a more stratified image of the norm. If Francis
bacon is right in saying that literature is a "kind of
contract of error between the deliverer and the
receiver", there is little doubt that he definition of
the literary norm is one of the essential aspects,
probably the most complex, of this "contract".
The calling into question of the undisputed dominance
of a single unmarked narrative tense (the aoristic
Past), the adoption of "strategic" procedure in the
build-up of the narrative texture, and above all the
abolition of the unique perspective point, all suggest
an explicitly multi-level structuring of the
narrative, which allows for permanent ambiguity in the
perception of the narrative focus. As Fludernik (1996)
notes, this is one of the most pervasive
characteristics of modern fictional prose, as shown
for instance by the frequent use-and multiple
manifestations-of free indirect speech.

Cardinaletti shows that in spite of the apparently
similar prosodic and pragmatic properties exhibited by
these two sentences, there are many differences
between VSO and VOS. This suggests that emarginazione
does not correspond to a uniform syntactic process. On
the distribution of quantifies constituents, of
binding phenomena, and of agreement patterns, the
author has shown that the so-called emarginazione
construction corresponds to two different structures
depending on the syntactic function of the
marginalized constituent. In VSO the object is
distresses in its base position; in VOS the subject is
right-dislocated" (base-generated in a position
structurally lower than the clause. This is exactly
what is expected under the antisymmetric approach of
Kayne 1994), where there is no post-object position
for the subject (i. e. , no rightward specVP). The
analysis is conformed by the semantic and prosodic
properties of the two constructions: in VSO the
subject is contrastively focused, and is assigned
stress by the Emphatic/Contrastive Stress Rule; in VOS
the object can be no contrastively focused, and is
assigned stress by the Nuclear Stress Rule. Nothing, of
course, prevents the subject from being distressed.
The distinction between distressing and Right
Dislocation has an important implication. It supports
the proposal that post-verbal focused constituents
occur in their base-position inside VP. An alternative
proposal, which involves a rightward focus position
above VP to which focused elements are moved, would be
forced to analyze all materials following the focused
constituent as right-dislocated, thus failing to
capture the asymmetries.

Cinque's chapter argues that if functional
affixes and particles are interpreted as the
overt realization of distinct functional heads,
there is reason to posit the existence of a
substantial number of distinct aspectual heads ordered
among each other. Cinque presents some facts, internal
to just one language, Italian, which appears to offer
some evidence for ordering these heads among each
other. No existing analysis of "restructuring" offers
a natural account of why the transparency effects
characteristic of this phenomenon occur across
languages with just the classes of modal, aspectual
and movement verbs. The general expectation of the
assumption, that if a verb may either be generated
(and licensed) as the head of the VP, or, when it
"lexicalises" a particular head, directly in that head
position, both the monoclausal nature of the
phenomenon and the membership of the verb in the
"restructuring" class can be naturally derived; and
that if the various functional heads of the clause are
rigidly ordered then "restructuring" verbs should
display a rigid relative order among each other when
transparency effects obtain or when they are licensed
not as lexical verbs, but as "functional" verbs
generated in specific functional head is generally
fulfilled. By exploiting the rigidity in the relative
order of the "restructuring" verbs he finds out some
evidence to determine the relative position of a
number of aspectual and root modal heads that had
remained undermined in Cinque 1999. In particular,
this paper integrates into the partial order proposed
there the functional heads corresponding to
ASPconative, ASPfrustrative/sucesss, ASPinceptive,
ASPpredispositional, and ASPdelayed
(or finally), and refines the positions of the root
modal heads within the overall hierarchy proposed in
Cinque 1999. The revised portion of the hierarchy is:

ASPhabitual>ASPdelayed (or finally)
>ASPpredispositional>ASPrepetive (I)>
ASPfrequentive (I)>MODvolition>ASPcelerative
(I)>ASPterminative>ASPcontinuative>ASPperfect>ASPretrospective
>ASPproximative>ASPdurative>ASPprogressive>ASPconative>ASPcompletive
(I)>Voice>ASPcelerative (II)>ASPinceptive (II)
ASPcompletive (II)>ASPreceptive (II)>ASPfrequentive (II)

The goal of Giusti's paper is to provide a
formal development of the Latin demonstrative
ILLE into two different categories, namely
the definite article and the personal pronoun's
third person singular form. This paper shows how
Renzi's (1997) proposal, which captures the
correlation among many observable facts found across
Romance Romance languages, can be straightforwardly
represented in the recent minimalist framework
developed by Chomsky (1992,1995). First section
reviews and slightly revises Renzi's (1997: I-II)
proposal of analyzing the three categories of
Demonstrative, Personal Pronoun and Article as a
bundle of semantic and syntactic features. The partial
difference in features to be found among the three
categories is the reason for the different structural
positions occupied by the three elements. The second
section follows Renzi (1997: 12-15) in taking the
development of ILLE as an example of a more general
process of "grammaticalization" in the sense of
Millett (1912) and recently Lehman (1982). This
process turns a lexical element into a functional one.
It has been shown that "grammaticalization" in the
generative terms reduces to the reanalysis of a
constituent in a functional Specifier as being filler
of the adjacent head.

As the name suggests, Guasti and Moro's article deals
with Romance causatives within the framework of
dynamic antisymmetry. Elaborating on the previous
accounts they propose that causative verbs take a
small clause as complement in analogy with copular
construction. They suggest that the surface order of
causatives is determined by the search of an
antisymmetric geometry. In causatives based on
unergative verbs, the verbal head moves out of the
small clause to break the symmetric configuration. In
causatives based on transitive verbs, the whole VP
must move and a prepositional complementizer must be
present to ensure the realization of an antisymmetric
configuration. No movement has to take place in
causatives based on unaccusative verbs, because the
configuration is not symmetric to begin with. The
first section of this article discusses the main facts
concerning Romance causatives . The second section
presents the dynamic antisymmetry framework and in the
third section they exemplify it through a discussion
of copular constructions. In the final section they
apply the dynamic antisymmetry framework to Romance
causatives and motivate the approach by proposing a
new analysis of Romance causatives by assuming that
the material following the causative verb is a small
clause and that DA holds. They derive the distribution
of the preposition a in Italian causatives from the
necessity of breaking a POS and show that there is no
exception to the fact that the predicative relation is
established uniformly between a DP and a VP in
causatives.

Richard Kayne writes the tenth chapter A Note on
Clitic Doubling in French. French normally has
verb-object order as in:

(1) Jean connait Marie 'John knows Marie'
(2) Jean parle de Marie 'John speaks of Mary'

In (2), the object can be replaced by a personal pronoun:

(3) Jean parle de moi 'J speaks of me'

In (1), the corresponding sentence is deviant.

(4) *Jean connait moi 'John knows me'

However, a clitic counterpart is well formed in
this case:

(5) Jean me connait 'John me knows'

Kayne (1975) took the deviance of (4) to directly reflect
the obligatory character of the movement operation
involved in the derivation of (5). That approach did
not attach syntactic importance to the difference in
form between the moi of (3) /(4) and the me of (5).
Kayne (2000) argued that moi and me differ in that moi
is bimorphemic and me not. If moi is bimorphemic and me
monomorphemic (m- with a phonologically epenthetic
-e), then it becomes difficult to think of (5) as
simply corresponding to (4) plus movement, which might
have been expected to yield

(6) *Jean moi connait

rather than (5). In this paper Kayne focuses rather on
the question how to exclude (4) if (4) is not the
exact non-movement counterpart of (5). If (4) does not
"underlie" (5), then it is no longer possible to
interpret the deviance of (4) as simply reflecting the
failure to apply the movement operation needed to
derive (5). In the present perspective, the problem
with (4) is not that moi has failed to move, nor that
there is a similar well-formed sentence with a clitic
instead of moi, but rather that moi has been doubled.

The proposal is: Pronominal arguments that are
structurally Case-marked in French must be doubled by
a clitic.

What now makes (4) deviant is that it is missing the
clitic imposed by the proposal made above. French
cannot have a direct object argument moi, but only of
the form moi m(e), which would yield

(7) Jean me connait moi 'John me knows me'

Similarly, the following contrasts:

(8) *Jean parle a moi
(9) Jean me parle a moi 'John speaks to me'

The deviance of (8) relative to (9) is interpretable
in terms of the proposal made above, i. e. , dative
(non-oblique) moi must be doubled. The following sections
take up issues such as gapping, subjects, modified
pronouns, quantifiers, quantifiers with covert non-clitic
pronouns, third person restrictions on covert
non-clitic pronouns, and extension to covert subjects
to test the requirement that French structurally case
marked pronouns must be doubled by a clitic. This
requirement suggests that the subject in the second
conjunct of a gapping sentence does not have
structural case. The doubling requirement in question
sometimes rules out non-doubling sentences whose
doubling and clitic counterparts are also not
acceptable; this poses a problem for Cardinaletti and
Starke's (1994) structure minimization proposal for
such sentences. French non-doubling sentences with a
clitic (and no corresponding non-clitic) can contain
pro (in addition to the clitic) only in third person
cases, with implications for the null subject
phenomenon in other languages, for past participle
agreement and for apparently bare quantifiers.

In Lidia Lonzi's analysis of the so-called
"subject-oriented sentence adverb", left adjunction to
maximal projection appears preferable to the
generation in Spec position. If this analysis
proves correct, the generation in Spec cannot be the only
legitimate treatment for adverbs. In her view,
the left adjunction for subject-oriented adverbs is
grounded in this analysis of an inherent parenthetical
status, which has three major consequences in syntax.
First of all, by a reasonable assumption concerning
left adjunction, left-adjoined adverbs must be outside
the scope of negation, independently from their
specific position with respect to NegP. Secondly, no
licensing criterion is required. Thirdly, there is no
risk of unmotivated adverb movement. The number of
possible positions in the string can parallel the
number of places made available by maximal, ultimately
lexical, projections, a traditional datum which does
not exclude the importance of studying the relative
order of various adverbs, including the parenthetical
ones, in one and the same "position", along the lines
of Cinque.

The article by Manzini and Savoia also deals with clitics.
They discuss the status of si in its different
interpretations and consider its ordering within the clitic
cluster in many different dialectal varieties. They expect
that every object clitic occupy its own CI projection,
characterized by a particular set of features and
ordered in a particular way with respect to all other
positions. The problem that faces them is to determine
which features exactly are realized by object clitics
and how they structure themselves in a syntactic
hierarchy. The conception of object clitic positions
that they develop in this paper makes use of thematic
characterizations, or of aspectual ones.

Cecilia Poletto examines some cases of complementizer
deletion and propose a V-to-C analysis for them. Though
it is not easy to reduce all cases of complementizer
deletion to V-to-C, it has been shown that at least in some
cases this is a viable hypothesis. A split-CP analysis
combined with V-to-C movement gives the tools to account
for many interesting facts, both in standard Italian and in
the NIDs. The semantic feature attracting the verb into CP
layer is not always the same: for embedded subjunctive it
is a [-realis] feature, for disjunctive sentences it
probably is an operator feature connected with the
null operator in the SpecC position.

Luigi Rizzi provides evidence for an Interrogative Phrase
projection (hosting se 'whether' and perche' 'why')
distinct from, and lower than, his 1997 ForceP, higher
than his 1997 FocusP and the projection to which
wh-phrases move (in embedded questions). In this
analysis, se, the complementizer introducing embedded
yes/no questions in Italian, fills a position in the C
system which is lower than Force, the position filled
by the declarative complementizer che, but higher than
Foc and the position filled by Wh-elements in embedded
questions. Perche' and other Wh-elements corresponding
to higher adverbials can fill the position of Spec of
Int (at least when construed locally); this explains
why such elements can occur with a following focus in
both main and embedded questions; it also helps us to
understand why these elements do not trigger
obligatory I-toC movement in the main questions.

Giampaolo Salvi discusses a little known
subordinate sentence type of early Romance (later
superseded by the Verb Second structure of main
clauses) interpreting it is as a residue of Latin
sentence structure.

Finally, Christina M. Tortora argues for the necessity
of distinguishing two types of VS orders with
unaccusative verbs, depending on the presence of an
overt (or covert) goal.

REFERENCES
Beghelli, Filippo and Timothy Stowell (1997).
Distributivity and negation. In Anna Szabolcsi (ed. )
Ways of Scope Taking. Dordrecht: Kluwer. 71-107

Benuncci, Franco. (1990). Destrutturazione. Classi
verbali e costruzioni perifrastiche nelle lingue
romanze antiche e moderne. Padova: Unipress.

Chomsky, Noam. (1992) A minimalist program for
linguistic theory. MIT Occasional Papers in
Linguistics, 1. (Published (1993) in Kenneth L. Hale
and Samuel Jay Keyser 9eds. ) The View from Building
20: Essays in Linguistics in Honor of Sylvain
Bromberger. Cambridge, (Mass. )

Cinque, Guglielme. (1999). Adverbs and Functional
Heads: A Cross-Linguistic Perspective. New York &
Oxford: Oxford University Press (circulated in manuscript
form in 1997)

Fludernik, Monika. (1996). Linguistics and literature:
Prospects and horizons in the study of prose. Journal
of Pragmatics, 26, 583-611

Kayne, Richard S. (1994). The Antisymmetry of
Syntax. Cambridge, Mass. : MIT Press.

Lehmann, Christian (1982). Thoughts on
Grammaticalization. A Programmatic Sketch. University
of Cologne.

Linebarger, Marcia (1987). Negative polarity and
grammatical representation. Linguistics and Philosophy
10,325-387.

Meillet, Antoine. (1912). Linguistique historique and
linguistique generale. Paris: Champion. (New edition:
1948. )

Pollock, Jean-Yves. (1989). Verb movement, Universal
Grammar and the structure of IP. Linguistic Inquiry,
20,365-424

Pollock, Jean-Yves. (1997). Notes on clause structure.
In Liliane Haegeman (ed. ) Elements of Grammar.
Dordrecht: Kluwer, 237-279

Renzi, Lorenzo. (1997). Fissione di lat. ILLE nelle
lingue romanze. In: Gunter Holtus, Johannes Kramer and
Wolfgang Sweckhard (eds. ). Italica et Romanica.
Festschrift fur Max Pfister zum 65.
Geburtstag. Tubingen: Niemeyer.

Rizzi, Luigi. (1976). Ristrutturazione. Rivista di
Grammatica Generativa, 1. 1, 1-54.

Rizzi, Luigi (1990) Relativized Minimality. Cambridge,
Mass: MIT Press

Rizzi, Luigi. (1997). The fine structure of the left
periphery. In Liliane haegeman (ed. ) Elements of
Grammar: Handbook of Generative Syntax. Dordrecht:
Kluwer, 281-337


ABOUT THE REVIEWER
Dr. Manideepa Patnaik is at present an associate in
the department of Sanskrit and Indian studies at
Harvard University. She is currently working on
Complementizer system. Besides this, she is trying to
finish some of her other projects such as: A
comparative lexica of Western Austro-Asiatic
Languages, A reference grammar of Oriya, and A
reference grammar of Juang.
com">Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue