LINGUIST List 14.2751

Sun Oct 12 2003

Review: Syntax/Ling Theory: Hulk & Pollock (2001)

Editor for this issue: Naomi Ogasawara <>

What follows is a review or discussion note contributed to our Book Discussion Forum. We expect 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 review." Then contact Simin Karimi at


  1. Larry LaFond, Subject Inversion in Romance and the Theory of Universal Grammar

Message 1: Subject Inversion in Romance and the Theory of Universal Grammar

Date: Sun, 12 Oct 2003 16:37:34 +0000
From: Larry LaFond <>
Subject: Subject Inversion in Romance and the Theory of Universal Grammar

Hulk, Aafke C., and Jean-Yves Pollock, ed. (2001) Subject Inversion in
Romance and the Theory of Universal Grammar, Oxford University Press,
Oxford Studies in Comparative Syntax.

Announced at

Larry LaFond, Southern Illinois University Edwardsville


Hulk and Pollock's _Subject Inversion in Romance and the theory of
Universal Grammar_ brings together essays that address a number of
longstanding descriptive and theoretical concerns surrounding the
appearance of subjects on the rightmost periphery of sentences in
languages whose canonical word order is subject-verb-object (SVO).

H&P suggest that, while some consensus has been reached about subject
inversion in Romance (e.g., that it differs significantly from
Germanic subject inversion), at least three critical questions remain:
How do subject inversion constructions satisfy the Extended Projection
Principle (EPP)? In what structural position do we find the
postverbal subject? And how do the licensing requirements of
postverbal subjects in French differ from those in 'free' inversion
languages like Italian?

These questions are not novel, but H&P claim that refinements in the
theory of UG during the last couple decades have provided new
theoretical vantage points that now permit us to come closer to a
'truly explanatory' theory of subject Inversion. The chief
developments they note are the VP-internal subject hypothesis (Larson
1988; Sportiche 1988), the movement of verbs and subject DPs to
various layers of a split IP for the purpose of checking features
(Pollock 1989, 1997; Chomsky 1995), and a more fully articulated
nature of the left-periphery, where the Comp domain may house force,
focus, topic, and finiteness in a fixed hierarchical order (Rizzi

These developments take center stage in this volume, with each author
incorporating them into claims made about subject inversion. How
these authors account for such notions, however, reveals clear
differences in their assumptions. For example, some authors assume
foci are rightmost constituents for the purpose of receiving nuclear
stress, while others assume focus is a functional feature, heading its
own functional projection and attracting constituents to satisfy
feature checking, but a projection of the left periphery. The
disagreement between these approaches is left unresolved, although at
least one of the authors (Belleti, ch. 3) suggests a harmonizing

Following H&P's introduction, the volume commences with Barbosa's 'On
Inversion in Wh-questions in Romance' (ch. 2) in which Barbosa
suggests that a major difference between of Germanic and Romance
inversion is a parametric specification of the [wh] feature in these
languages: [+wh] is checked against the highest inflectional head in
Romance, but not in Germanic. Barbosa claims that in Romance Spec/IP
is an A'-position and the EPP and N-features of INFL are checked by
rich morphological agreement. This permits inversion constructions in
Romance, where subjects remain in situ in their VP-internal position,
and surface before the verb only when they are base-generated, as
dislocated topics or when subject to A'-movement to Spec/IP.

Belletti's 'Inversion as Focalization' (ch. 3), proposes that subjects
reside in a very low VP-external position in Italian, where subject
inversion is argued to be the result of the interplay of 'old' an
'new' information in the clause, with postverbal subjects carrying new
information. Feature checking comes into play in this analysis,
because focus is the head of a functional projection and must be
checked with a DP in its Spec. According to Belletti, however, the
subject does not move to Spec/IP to check an EPP feature, rather the
subject remains low in the clause and is licensed by focus. Belletti
further observes that VSO word order, permissible in Romance languages
such as Spanish and Romanian, is not accepted in Italian. This fact
follows if a functional projection headed by focus immediately
dominates the VP, blocking the movement of the object to its case-
position due to relativized minimality. That this order arises in
Spanish and Romanian, Belletti argues, is due both to the fact that in
the VSO order of these languages the postverbal subject is not
necessarily interpreted as new information, and to the presence of an
additional higher case-position in these languages, to which the
subject moves and is licensed by Case.

Costa's 'Marked vs. Unmarked Inversion and Optimality Theory' (ch. 4)
employs OT's constraint-based framework (Prince & Smolensky 1993) to
account for word- order possibilities in Italian. Costa and others
have previously argued that subject inversion may be the result of
relative rankings of the syntactic and discoursal constraints within a
grammar (Grimshaw & Samek-Lodovici 1995, Costa 1996, Speas 1997).
Here Costa adds a discussion of markedness and basic word order. Costa
demonstrates how the unmarked word orders of Portuguese, Greek, and
two varieties of Spanish are the result of differing rankings of the
same set of universal constraints. He then considers the asymmetry
between Portuguese and Italian with respect to subject inversion.
Costa's proposed ranking of constraints for Portuguese makes the
prediction that it is more important to satisfy discourse constraints
in this language than to license nominative case. It also predicts
that, while there may be scrambling via adjunction to VP, this
language permits no movement to Spec-AgrOP, and when subjects are not
the focus of the sentence they may raise to Spec-IP. The difference
between the rankings of Portuguese and Italian is suggested to be
slight: The Italian ranking moves OBJ-CASE higher than both ALIGNFOCUS
and STAY. This difference makes movement of the object to Spec-AgrOP
more important than satisfying both the discourse requirement to align
focus and syntactic constraints requiring A-movement to Spec-IP.
Thus, Costa suggests that an OT analysis provides a quite elegant and
economical account of language variation.

Kayne & Pollock ('New Thoughts on Stylistic Inversion,' ch. 5) and
Taraldsen ('Subject Extraction, the Distribution of Expletives and
Stylistic Inversion,' ch. 6) both suggest solutions related to
stylistic inversion, but are quite different in their conclusions.
Kayne & Pollock assert that subjects in French SI reside in a high,
dislocated position, above the IP, and cite as evidence both the types
of subject DPs which SI sentences permit and several facts concerning
extraction. Specifically, they argue that the wh-argument first
extracts to a position above the IP, to check the relevant [+wh]
feature, and then extraction of the subject occurs, moving it higher
in the left periphery than the wh-phrase. This two-part movement is
next followed by the preposing of the remnant IP to a yet higher
specifier position in the left periphery, with yet another movement of
the wh- phrase to a clause-initial position. The result is the
'inverted' surface word order of French SI. Kayne & Pollock see
certain advantages to positing this multiple leftward movement,
particularly in that it provides an account of observable facts that
were previously left unexplained.

Like Kayne & Pollock, Taraldsen (ch. 6) also posits that SI involves
Remnant Movement, but Taraldsen's focus is on the classical analysis
of the 'que' to 'qui' alternation in French. Taraldsen's new analysis,
one that critiques Rizzi's (1990) account, suggests that French 'qui'
is in fact 'que' plus 'i', with 'i' being a pure expletive, a pronoun
in the highest Spec-Infl position, having no number features. This
suggests commonality with languages such as Danish and Norwegian,
where such alternations also occur. For Taraldsen, strong features
are ones which must pied-pipe their host categories when they move;
weak features can move on their own. Taraldsen argues that French Infl
must check number and EPP, the number feature is considered weak
(permitting it to move to C sans pied-piping), but the number feature
of D is strong (demanding that the subject DP be checked at
Spec/CP). It is, therefore, the EPP feature that requires the presence
of expletive 'i' in Spec/IP. The result is that whenever subject
extraction occurs, it will involve 'que' plus 'i' to satisfy the
requirement for feature- checking.

Completing the volume is Zubizaretta's 'The constraint on preverbal
subjects in Romance interrogatives: A minimality effect' (ch. 7).
Zubizaretta's argument involves an analysis of clitics and strong Agr.
Zubizaretta suggests that in Romance verbal arguments merge into the
derivation with an abstract operator above TP, and there bind an
argument variable within VP. This 'Cl- operator' externalizes verbal
arguments with respect to tense associated with the verb. She holds
that preverbal subjects in Romance will appear in the specifier of
this Cl-operator, except when they are a focus. This analysis also
proposes a Q operator, which checks the interrogative force of
declaratives. If the Q-operator is co-indexed with a [+wh] DP, then it
binds variables in the clause. Zubizaretta concludes that subjects do
not appear between a wh-phrase and the inflected verb because to do so
would violate minimality, since the wh-variable is closer to the
Cl-operator than the Q- operator. This solution proposes that
differences among Romance languages are the result of parametric
options regarding the licensing of Q.


As the editors promise, _Subject Inversion in Romance and the Theory
of Universal Grammar_ successfully immerses the reader into the
ongoing discussion of some persistent problems in the theory of
grammar. Do these articles actually bring us, as H&P suggest, closer
to an explanatory theory of subject inversion? Yes, but still not
very close, since so many unanswered questions remain. For example,
beyond the observation that Romance subject inversion, unlike Germanic
V2, involves no V-to-C movement, agreement has not yet been reached on
whether the position of subjects is VP-internal or in some low
position immediately dominating the VP or in a higher position
still. Nor has consensus been reached on how subjects arrive (or
remain) in their positions. The solutions here lean either towards
complex machinations that account for the empirical facts at the
expense of elegance and economy, or simple mechanisms whose elegance
and economy cannot be denied, but whose ability to handle a fuller
range of data may be suspect. This may simply reflect less of a
critique of this particular volume and more a reflection of the state
of linguistic theory at the start of the 21st Century. The solutions
offered in this volume clearly spring from the different theoretical
vantage points and the differing conceptions of Universal Grammar that
inform them.

One important point of consensus among the authors in this volume is
that purely syntactic approaches, which fail to permit a consideration
of discoursal factors or features, are unlikely to adequately account
for the empirical facts of subject inversion. Interfaces between
syntax and discourse, and descriptions of the architecture in which
syntax and discourse interact, are at the heart of each of these
chapters. For some (e.g., Costa), instantiations of subject inversion
is the result of direct interaction and competition between syntactic
constraints and discoursal constraints. For others (e.g., Barbosa and
Belletti), different mechanisms syntactically encode discourse
features, with discoursal notions such as Topic and Focus features
heading distinct functional projections. While the authors agree that
notions such as Topic and Focus play important roles in understanding
subject inversion, consistent and rigorous definitions of those
notions are still needed. While most agree that certain positions
within the clause, e.g., the left or right periphery, have privileged
status, their attempts to provide detailed articulations of these
positions differ significantly. Other recent work suggests that even
the wide array of discoursal features discussed among these authors
may be inadequate for a fuller explanation of the nature of subject
inversion, the left (and right) periphery, and the encoding of Topic
and Focus. For example, while some of the authors here assume Topic
or Focus to be individual (or perhaps iterative) projections
attracting arguments (cf. Rizzi 1997), Poletto (2000) has suggested
that these projections are not unitary and may themselves be further
articulated into a number of projections that are the locus for
semantic differentiation. Such semantic differentiation provides a
potential challenge for work such as Belletti's, where classic
definitions of 'old' and 'new' information are used. We should
suppose, however, that if Poletto (2000) is on the right track, the
need for studies such as those provided in this volume should not be
dismissed, rather intensified.

This volume demonstrates, perhaps unwittingly, that the emerging
picture of subject inversion in Romance is still a messy one. It must
perhaps be so whenever one allows analysis of syntactic data to be
enriched (some might say contaminated) by discoursal and contextual
considerations. Nevertheless, while these authors have neither solved
all the questions surrounding inversion, nor even achieved consensus
among themselves, they have far more successfully provided carefully
nuanced discussions of the descriptive facts that an explanatory
theory will need to address. This alone is no small accomplishment,
and should sufficiently establish the value of H&P's volume.

As with many of the books in the Oxford Studies in Comparative Syntax
series, this volume as a whole would be quite challenging for
undergraduates, particularly since the authors employ differing
theoretical mechanisms to make their arguments. I could envision this
book's use in a high-level graduate seminar with students who have
already received advanced syntactic training. The book will, of
course, be useful for scholars working on subject inversion or related


Chomsky, N. (1995). The Minimalist Program. Cambridge, Mass: MIT Press.

Costa, Jo´┐Żo. (1997). Word order and constraint interaction. ms. ROA-
181. Rutgers Optimality Archive,

Grimshaw, Jane and Vieri Samek-Lodovici. (1995). Optimal Subjects.
University of Massachusetts Occasional Papers in Linguistics (UMOP),

Larson, R. (1988). On the double-object construction. Linguistic
Inquiry 19: 335-391.

Poletto, C. (2000). The higher functional field: Evidence from North
Italian Dialects. Oxford: Oxford University Press.

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

Pollock, J.-Y. (1997). Langage et cognition. Paris: Presses
Universitaires de France.

Prince, A. & P. Smolensky (1993). Optimality theory: Constraint
interaction in generative grammar. Ms., Rutgers University, New
Brunswick, N.J. & University of Colorado at Boulder.

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

Rizzi, L. (1997). The fine structure of the left-periphery. Haegeman,
L. (ed.) Elements of Grammar. Dordrecht: Kluwer.

Speas, M. (1997). Optimality Theory and null arguments. D. Archangeli
and D.T. Langendoen (eds.) Optimality Theory: An Overview, 171-99.
Malden, MA: Blackwell Publishers.

Sportiche, D. (1988). A theory of floating quantifiers and its
corollaries for constituent structure. Linguistic Inquiry 19: 425-449.


Larry LaFond is an Assistant Professor of English at Southern Illinois
University Edwardsville. He completed his Ph.D in linguistics from
the University of South Carolina in 2001. His research focuses on
second language acquisition of Romance and Germanic languages and
interfaces between syntax and discourse.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue