LINGUIST List 15.649

Fri Feb 20 2004

Review: Syntax: Boeckx & Grohmann, ed. (2003)

Editor for this issue: Naomi Ogasawara <naomilinguistlist.org>


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 Sheila Dooley Collberg at collberglinguistlist.org.

Directory

  1. Olga Zavitnevich-Beaulac, Multiple Wh-Fronting

Message 1: Multiple Wh-Fronting

Date: Fri, 20 Feb 2004 11:52:20 -0500 (EST)
From: Olga Zavitnevich-Beaulac <c3002er.uqam.ca>
Subject: Multiple Wh-Fronting

Boeckx, Cedric and Kleanthes K. Grohmann, ed. (2003) Multiple Wh-
Fronting, John Benjamins, Linguistik Aktuell/Linguistics Today 64.

Announced at http://linguistlist.org/issues/14/14-2331.html


Olga Zavitnevich-Beaulac, Universit� du Qu�bec � Montr�al (UQAM)

INTRODUCTION 

The original work by Wachowitcz (1974) and, in particular, the study
by Rudin (1988) have lead to the appearance of a new topic in research
on wh-questions: multiple wh-fronting. In the recent years, and
especially with an advent of Minimalism, the puzzling phenomenon of
multiple wh-movement has generated a number of hypotheses that offer
different solutions to the existing linguistic fact (see among others
Cheng (1991), Grewendorf (1999), Simpson (1999, 2000)). The book,
edited by Cedric Boeckx and Kleanthes K. Grohmann, is a collection of
12 papers (including the Introduction) that further develops this
topic. The main question that linguists attempt to answer concerns the
nature of multiple wh-movement. Minimalism does not allow superfluous
steps and operations. If movement of a single wh-phrase suffices to
check strong/uninterpretable feature, why do languages move all wh-
expressions clause-initially? The contributors to this volume explore
this issue on a large number of cross-linguistic data offering their
solutions to the problem. A number of related issues are examined as
well, among them Superiority Constraints, Wh-Focus interaction,
multiple specifiers, extended CP field, remnant IP/TP movement.

DESCRIPTION 

The introduction by Cedric Boeckx and Kleanthes K. Grohmann (pp.1-15)
offers a brief overview of research on multiple wh-fronting. The
authors illustrate why the case of multiple wh-fronting gets so much
attention in minimalist studies, namely the four reasons for a
''successful marriage'' between multiple wh-fronting and minimalism.
The introduction also includes summaries of the articles contained in
the book. Since a number of papers deal with the problem of
Superiority, it is interesting to have the authors' observation that
although Superiority is viewed as one instance of generalised
Relativized Minimality effects, it is found only in the case of
wh-questions as opposed to other types of A'-movement (e.g. Topic,
Focus).

In ''Symmetries and asymmetries in multiple checking'' (pp.17-26)
Cedric Boeckx extends Chomsky's (2000) insight of Case and
phi-features being the two sides of the same coin, to Focus/Wh
relation claiming that the two features, although not identical, have
the same function. He further contends that focus feature being
interpretable on the Probe (but not on the Goal) can be checked
multiply. The opposite holds for the wh-feature: it is interpretable
on the Goal, but not on the Probe. The uninterpretable wh-feature of
the Probe can only be valued once by the closest element, from which
follows the Superiority effects in wh- feature checking. Thus the
difference between Focus and Wh-features is attributed not to their
intrinsic properties, but to their placement. Crucial for this
proposal is the distinction between Match and Agree, the latter
incorporating the former plus a Valuation procedure. Move can take
place under Match independently of Agree.

In ''On wh-islands and obligatory wh-movement contexts in South
Slavic'' (pp.27-50) Zeljko Boskovic presents more evidence that
multiple wh- fronting (MWF) languages, namely Serbo-Croatian (SC) and
Bulgarian, do not present a uniform phenomenon. This time the author
re-examines the case of wh-island extraction. First, Boskovic
demonstrates that Bulgarian is not totally immune from wh-island
effects as claimed by Rudin (1988): wh-arguments, but not wh-adjuncts,
can be extracted out of wh-islands. Secondly, he shows that some non
MWF languages (e.g. Swedish) exhibit the same pattern in wh-island
extraction as Bulgarian. Moreover, Boskovic illustrates that in
certain instances SC allows multiple filled Spec SP. However, the
presence of multiple Specs does not make it possible to escape
wh-island effect in SC. Based on this discussion Boskovic suggests
excluding wh-islands extraction criterion as a qualifying ground for
MWF dichotomy. Yet Boskovic fails to answer the question of why
Bulgarian and SC exhibit different behaviour in wh- argument
extraction out of wh-island. He leaves the answer to be determined
and, instead, reiterates his previous claim that all the differences
between the two languages can be traced to a single distinction in the
''lexical properties of the interrogative C-head, more precisely, the
PF affix status of the Bulgarian interrogative C'' (p.27).

Molly Diesing in ''On the nature of multiple fronting in Yiddish''
(pp. 51-76) examines two wh-question strategies in Yiddish (single
vs. multiple fronting). Diesing illustrates that although both
strategies convey the same (pair-list reading in all cases) reading
the two differ in their syntactic properties: multiple, but not
single, wh-fronting exhibits Superiority effects. Diesing suggests
that multiple wh- fronting is a genuine instance of wh-movement to
[+multiple Spec, CP] that can only originate from an A-position. The
closest to [Spec, CP] wh-phrase must move to Spec CP first hence
Superiority constraint. Single wh-fronting in matrix clauses is a
result of wh-scrambling to IP, which is immune to Superiority. The
wh-feature is checked a la Boskovicvia LF insertion of the C-head.
Embedded questions have [- multiple Spec, CP] and wh-movement in
embedded questions is a subject to Superiority. Diesing concludes
that Yiddish presents a hybrid type of MWF languages as it does not
properly fit in any specific class yet exhibits certain
characteristics peculiar of several of them.

In ''On the morphosyntax of wh-movement'' (pp.77-98) Marcel den Dikken
proposes that wh-movement in English is an instance of Focus movement
(at least in root clauses). The author presents the following
arguments. First, observing that in matrix clauses Topic can precede
wh-phrases (''?To Mary, what should we give?'') den Dikken concludes
that a wh-phrase appears lower than [Spec,CP], presumably in the Focus
position. Modifying Kayne's (1998) original proposal that non-wh
focus phrases can undergo focus movement followed by the remnant TP
movement (VP-preposing) den Dikken suggests that in multiple
wh-questions ''in- situ'' wh-phrase moves to a Focus position first
and then TP moves over it bringing the other wh-phrase to a
structurally higher position. In single wh-questions a wh-phrase also
moves to [Spec, FocusP], however remnant TP movement does not happen
because TP would be a harmful intervener between [+wh] C and the
wh-phrase in [Spec, FocusP]. Considering obligatory wh-relative
fronting den Dikkenin adopts the view that relatives are predicates of
NPs they are construed with. In order to establish a predicate
relation operator movement of a wh- phrase to embedded Spec, CP is
necessary. The trigger of movement is the morphological wh-feature of
C. Den Dikken's argument is built in part on the assumption that
multiple wh-questions can receive only a single pair echoic reading
when the fronted wh-constituent is a wh-the- hell expression, as in
(1):

(1)	 ''?Who the hell is in love with who?'' 

The author seems to overlook the following situation: a speaker
providing new information says that ''John is in love with Mary, Jim
is in love with Sue and Bill is in love with Monika'', a hearer
totally lost in this information posits (1). It is a totally
legitimate question which requires pair-list reading in spite of the
presence of ''wh-the-hell''.

Kleanthes K. Grohmann (''German is a multiple wh-fronting language!''
(pp. 99-130)) considers constructions with multiple interrogation in
German and suggests that this language should be included in the
typology of MWF languages. Grohmann observes that German patterns
with Bulgarian and Rumanian in that in all contexts of multiple
wh-questions obligatory pair-list reading is achieved. This,
according to the Hagstrom-Boskovic's approach, implies that all
wh-phrases undergo genuine wh-movement to CP as opposed to mere
wh-fronting or wh- scrambling. Indeed the author presents syntactic
evidence for A'- movement of both wh-phrases (intervention effects,
parasitic gap, weak crossover). However, unlike Bulgarian, German
never exhibits Superiority effects. Superiority condition can be
escaped if wh- phrases are D-linked. Grohmann illustrates that this
is the case in German where multiple wh-phrases must be D-linked.
This means that multiple wh-questions in German can never be a
''true'' request for ''new information''. Thus Grohmann concludes
that in multiple wh-questions German moves both wh-phrases to two
distinct focus positions within CP.

Youngmi Jeong (''Deriving Anti-Superiority effects: Multiple wh-
questions in Japanese and Korean'' pp. 131-140) offers a minimalist
explanation of Anti-Superiority and additional wh-effects. The author
rejects the view (Watanabe 1992, Saito 1994) that these effects come
from the violation of Empty Category Principle, but claims that they
are determined by three factors: no adjunction can proceed to the
right (Kayne 1994); in multiple wh-questions more than one, but not
all wh- phrases should front (Pesetsky 2000); CP-field includes
several functional projections (Rizzi 1997). Building on these
assumptions Jeong illustrates why the structure in (2) results in
ungrammatical derivation and why grammaticality is restored when an
additional wh- phrase is added, as in (3):

(2) *Naze nani-o anata-wa katta no? 
 why what-Acc you-Top bought Q
 ''Why did you buy what?''
(vs. Nani-o naze anata-wa katta no?)

(3) Dare-ga naze nani-o katta no?
 who-Nom why what-Acc bought?
 ''What did who buy why?''

Contrary to the general assumption that Japanese and Korean are wh-in-
situ languages, Jeong claims that wh-phrases in these languages do
undergo wh-movement to the functional domain. An exception is 'why',
which is base-generated in the functional projection.

In ''Conjoined questions in Hungarian'' (pp.141-160) Anik� Lipt�k
considers wh-questions that contain multiple wh-phrases, but refer to
a single incident, such as ''Who saw Mary and when?''. Lipt�k
illustrates that Hungarian has two distinct patterns and the two
''cannot be lumped together under a uniform syntactic analysis''
(p.157). In the first type all wh-phrases occur preverbally, while in
the second type the first wh-phrase appears preverbally, but the
others are in a postverbal position introduced by a co-ordinator �s
'and'. Lipt�k rejects previous elliptical clausal co-ordination
analysis (B�nr�ti 1992) that assumes that the two derivations are
parallel and that conjoined wh-phrases always result from the
co-ordination of two clauses with ellipsis in either the first or the
second conjunct. Instead she suggests that preverbal conjoined
questions involve clause internal co-ordination of wh-phrases, hence
the co-ordination relation is established within one and the same
clause. In contrast, postverbal conjoined questions are instances of
multiclausal co-ordination accompanied by ellipsis. Arguing for an X'
theoretic approach to co-ordination Lipt�k claims that asymmetric
relation holds between the two conjuncts (the two wh- phrases within
the same clause or the two conjoined clauses). The first conjunct
c-commands the second but not vice versa. However, this conclusion is
reached based on a single case of wh-subject and which- phrase
interaction, as in (4) and (5):

(4) a) Ki besz�lt �s melyik (pro) bar�tj�r�l? 
 who-Nom talked-3sg and which friend-poss.3sg-about
 ''Who talked and about which friend of his?''
 b) *Melyik (pro) bar�tja besz�lt �s kir�l?

(5) a) Ki �s melyik (pro) bar�tj�r�l besz�lt?
 who-Nom and which friend-poss3sg-about talked-3sg
 ''Who talked and about which friend of his?''
 b)*Melyik (pro) bar�tja �s kir�l besz�lt?

Notice that asymmetry disappears if both wh-phrases are adjuncts, as
in (6) and (7) (I am thankful to Edit Jakab for Hungarian data):

(6) Mi-vel t�rte be az ablak-ot �s mikor?
 what-with broke-3sg PREF the window-acc and when
 ''With what did he break the window and when?'' 
 Mikor t�rte be az ablakot �s mivel?
 ''When did he break the window and with what?''
(7) Mivel �s mikor t�rte be az ablak-ot?
 what-with and when broke-3sg PREF the window-acc
 ''With what and when did he break the window?''
 Mikor �s mivel t�rte be az ablakot?

In other words, as noted by the author, there are still many questions
that need to be addressed.

Persian wh-questions seem to posit a problem for the Minimalist
theory, as this wh-in-situ language allows optional wh-scrambling,
scrambled wh-phrases being subject to Superiority Constraint.
Moreover, multiple wh-questions require pair-list reading. These
problems are resolved by Ahmad R. Lotfi in ''Persian wh-riddles''
(pp.161-186). Adopting Hagstrom's (1998) proposal Lotfi assumes that
Persian has a phonetically null Q-marker which originates with the
lowest wh-phrase and undergoes subsequent movement to the functional
domain resulting in pair-list reading in all cases. Scrambled
wh-phrases move to a lower than CP projection being driven by Focus
feature. Superiority effects are caused by different reasons. Thus
subject/object violation does not result from Superiority constraint
per se but from Case assignment requirement: object cannot move/be
scrambled from the case assigning verb unless case requirement is
satisfied in an alternative way (by ra- morpheme that cliticised to an
object as -ro/o). As for ''true'' Superiority cases Lotfi proposes
that a wh-adjunct cannot cross a wh- object argument. Moreover the
timing of scrambling plays a crucial role: wh-arguments scramble
before Q-marker movement, while wh-adjuncts scramble after Q-movement.
Hence the two types of movement (wh- argument vs. wh-adjunct) differ
not in their properties, but in the time when move occurs in relation
to the Q-marker.

There are a number of questions that arise and are left unaccounted,
thus if ''adjuncts need no case-marking to license their scrambling''
(p.182) and scrambling is common in Persian, why can wh-adjuncts not
be scrambled over wh-subject:

(8) *Koja ki raft?
 where who went
 ''Who went where?''

Another question concerns the landing site of scrambled wh-objects.
It seems that they can appear in both A and A' positions, i.e. before
and after subjects, the unmarked order being subject +place/time
+object +verb:

(9) Chi-o ki xarid?
 what-case who bought
 ''Who bought what?''

(10) Armin chi-o koja xarid?
 Armin what where bought
 ''Where did Armin buy what?''

Lara Reglero's paper ''Non-wh-fronting in Basque'' (pp.187-227)
presents an account of wh-question formation in Basque. Basque
employs single and multiple wh-fronting strategies, both subject to
Superiority Constrain. Wh-movement in this language is argued to be
an instance of focus-movement (Ortiz de Urbina 1995). Assuming close
connection between Topicalization and D-Linking Reglero suggests that
in case of single wh-fronting the ''in-situ'' wh-phrase is D-linked or
Topicalized, hence it undergoes Topic movement to a position following
the verb, which in turn moved to the Focus Projection. Fronted
wh-phrase, not being D-linked moves to Focus position preceding the
verb. For multiple wh-fronting questions she claims that closest to
the verb wh- phrase is focused, while the wh-phrase to the left of it
is topicalized. Thus all wh-phrases move to the left periphery, both
Topic and Focus are hosted in the same projection, which Reglero calls
DeltaP.

Some aspects of the analysis raise certain questions, thus it is not
clear why the Topic feature of wh-phrases must be checked obligatory,
but checking of the Topic feature of non-wh-elements is optional (it
can be scrambled). Is such factor as ''discourse related'' a
sufficient ground for one head being able to license two distinct
functional features such as Topic and Focus? Moreover, can the same
head have different specification: Attract-all-discourse and
Attract-1Topic? And a more general question: how far can we stretch
the notion of ''discourse related''?

To test whether a wh-phrase is D-linked Reglero uses Pesetsky's (1987)
'the hell' strategy and den Dikken's application of this test:
aggressively non-D-linked 'the hell' is incompatible with D-linked
wh-phrase. However, Reglero disregards one property of 'the hell'
expression pointed out by den Dikken (same volume): 'the hell' needs
to be licensed and this licensing occurs in the Focus projection. In
other words it is not the case that leftmost wh-phrase must be
D-linked, it is the case that 'the-hell' expression must be focused
licensed.

The paper ''Malagasy as an optional multiple wh-fronting language''
(pp. 229-254) by Joachim Sabel discusses different wh-question
strategies employed by this Western Austronesian language. Malagasy
exhibits full and partial wh-movement in single wh-questions, and it
allows wh-in- situ as well. In multiple questions in Malagasy one
wh-phrase can be fronted and the other(s) remain in situ,
alternatively all wh- expression can stay in-situ or be fronted.
Sabel argues that full wh- movement in Malagasy is triggered by a
strong Focus feature, when this feature is selected for computation
being assigned to respective functional heads and wh-elements in the
numeration. If a numeration contains a wh-expletive which is directly
merged in C, then wh-movement will be partial as the relevant feature
in matrix C is checked by expletive. Accordingly, if no [+strong]
Focus feature is selected, no wh-movement will occur, as no feature
checking is required. For the case of multiple wh-fronting Sabel
suggests that similar to Bulgarian, wh-phrases in Malagasy prior to
A'-movement form a wh-cluster and raise to the left periphery as a
single unit. Sabel also discusses different constraints existing in
single and multiple wh-questions in Malagasy. Thus surface
optionality in wh-question formation in Malagasy is the result of
different numerations. However, the optionality question is not
resolved completely, as one language allows two instantiations (strong
and weak) of the same feature.

In ''Multiple wh-fronting in Serbo-Croatian matrix questions and the
matrix sluicing construction'' (pp.255-284) Sandra Stjepanovic
examines multiple wh-fronting in SC. Stjepanovic argues that SC has
two focus positions: one located above TP and the other between TP and
VP; and that wh-movement in SC is driven by the Focus feature. She
further illustrates that Superiority effects found in some short
distance null C matrix questions (that is, context where Superiority
should not be observed) result from sluicing with multiple wh-remnants
or multiple sluicing. Stjepanovic first presents evidence that
elliptical constructions with wh-questions in SC are results of
multiple sluicing, but not gapping, and then demonstrates that
sluicing of multiple wh- remnants is wh-movement to Spec, CP followed
by IP deletion. The fact that wh-phrases move to Spec, CP explains
their behaviour regarding Superiority effects.

DISCUSSION 

The book is interesting in that it provides a large amount of
empirical data on multiple wh-fronting including widely discussed
languages, such as Bulgarian, Serbo-Croatian and Hungarian, as well as
less studied languages such as Basque, Malagasy, Persian, Yiddish.
Those who do research in this area can find a number of interesting
cross-linguistic facts. All the papers deal explicitly with the
phenomenon of multiple wh-fronting addressing the issue from different
perspectives. In fact many papers are interrelated, as the authors
come to similar conclusions yet based on the analysis of different
linguistic phenomena and on different lines of argumentation. In that
I agree with the authors that papers in the volume ''are best
appreciated when put together'' (p.8). Thus both den Dikken and
Grohmann entertain the view that traditionally classified single
wh-movement languages, such as English and German, in fact move all
wh-phrases to the left periphery. Similarly in Lotfi's and Diesing's
analyses obligatory pair-list reading in multiple wh-questions in
Persian and Yiddish results not from genuine wh-movement, but wh-focus
fronting with Q-marker being attached to the lowest wh-phrase. Again
the two authors use different arguments to build their hypothesis.

All the papers in the book point to the existing interaction between
wh-questions and focus, thus confirming Horvath's (1986) original
insight that wh-movement can be driven by the focus consideration.
Another area of research extensively discussed in the book is the
extended structures of the left periphery (Rizzi 1997). It is obvious
that it needs further articulation (see Rizzi 2001) as it appears that
more elements move to the functional domain than it was originally
thought. The book offers up-to-date analyses of the topic,
incorporating the latest hypotheses developed in the Minimalist
theory. It contributes to our understanding of the mechanism of
multiple wh- questions. Moreover it refines the existing typology of
multiple wh- questions. The contributors also raise a number of
questions, hence identifying issues for further research in the area.

Having said that, I have to admit that proposals presented in the
papers differ in their novelty, originality and quality. Although all
the papers clearly articulate the problems they want to address and
the goal they aim to achieve, in my view, not every contributor is
successful in convincing the reader that the advocated approach is on
the right track.

There is one general comment regarding the use of the linguistic data.
Historian E. H. Carr once wrote: ''The facts (in our case language
data) are really not at all like fish on the fishmonger's slab. They
are like fish swimming about in a vast and sometimes inaccessible
ocean; and what the historian (or linguist) catches will depend,
partly on chance, but mainly on what part of the ocean he chooses to
fish in and what tackle he chooses to use -- these two factors being,
of course, determined by the kind of fish he wants to catch'' (p.18).
In linguistics the situation is even more complicated than in history
as language allows great creativity. ''Creative, exploratory
expressions are made constantly by speakers ... to facilitate
communication as in changes to avoid ambiguity or to foster easier
identification of discourse roles'' (Harris and Campbell 1995:54).
The vast majority of such expressions are never repeated, yet they
still can be registered by researchers. Since there are no ''strict
criteria'' for language acceptability, very often grammaticality
judgements are at the mercy of a researcher.

In a few papers of the book the reader can come across comments like
''data are not crystal-clear'' (Diesing p.69), ''there seems to be
disagreement among speakers...'' (Grohmann, p.127), ''neither sentence
is impeccable'' (den Dicken p.96). Of course, the fact is that in real
life we do not deal with ''ideal speakers'' and ''ideal language''.
However, couple of time there was an impression that the data was
pushed to accommodate a hypothesis. Thus den Dikken uses a modified
example from Kayne (1998) ''*I spoke to only John.'', which is starred
as fully ungrammatical. Yet in the original paper Kayne notes that
evaluating the sentence ''?John spoke to only Bill.'' ''speakers vary
substantially in their judgements, ranging from fully acceptable to
fully unacceptable'' (p.148). The worst in this respect is Reglero's
paper. In the body of the text, on page 194, the author gives example
(23) from Basque ''Nork zer erosi du?'' (''Who bought what?''). It is
followed by endnote 9 ''note that some speakers disallow
(23)''(p.221). The sentence in (23) is not even marked ''?'' as
marginally acceptable. Reglero further presents an ungrammatical
sentence in (24): ''*Zer erosi du nork?'' (What did who buy?). Again
there is an endnote: ''According to Etxepare and Ortiz de Urbina (24)
is grammatical'' (p.221). If you do not know Basque than you are in
trouble! In endnote 29 the author refers the reader to the data in
her previous work, however admitting that the data ''are also
inconclusive''. She comes to a conclusion that wh-phrases do not move
to Spec, SP in Basque based on the fact that one (!) out of three
informants allowed a single-pair answer. Of course the question
arises about the validity of a hypothesis which is built on such
doubtful data. This in no sense means that one should not explore
beyond the ''pure grammatical'' and put forward daring proposals,
maybe just provide more data whose acceptability status is higher than
''marginally acceptable''.

There are a couple of minor errors:
In Boeckx' paper there is a mix up with Japanese proper names in
example 6 (p.19): ''Hanako introduces professor Tanaka to Mary''.
Analysing it on the next page Boeckx notes: ''if an object honorific
marker surfaces on the verb, we obtain the odd interpretation that
Taro respects Mary'' (indeed odd!).Jeong in his paper refers the
reader to a syntactic tree represented in (16), which never appears in
the article.

REFERENCES:

Boskovic, Zeljko (1997) Superiority Effects with Multiple Wh-Fronting
in Serbo-Croatian. Lingua 102, pp.1-20.

Boskovic, Zeljko (1999) On Multiple Feature Checking: Multiple Wh-
Fronting and Multiple Head Movement. In Samuel David Epstein &
Norbert Hornstein (eds.) Working Minimalism. Cambridge, Mass: MIT
Press, pp.159-187.

Boskovic, Zeljko (2002) On Multiple Wh-Fronting. Linguistic Inquiry,
Vol.33, number 3, pp.351-383.

Carr, E. H. (1986) What is History -- The George Macaulay Trevelyan
Lectures delivered in the University of Cambridge, January - March
1961, London: Macmillan Press.

Cheng, Lisa (1991) On the Typology of Wh-Questions. Ph.D. Diss. MIT,
Cambridge, Mass.

Grewendorf, G�nther (1999) The Additional Wh-Effect and Multiple Wh-
Fronting. In David Adger, Susan Pintzuk, Bernardette Plunkett & George
Tsoulas (eds.) Specifiers: Minimalist Approach. New York & Oxford:
OUP, pp.187-201.

Harris, Alice & Campbell, Lyle (1995) Historical Syntax in Cross-
Linguistic Perspective. CUP.

Horvath, Julia (1986) Focus in the Theory of Grammar and the Syntax of
Hungarian. Dordrecht: Foris.

Kayne, Richard (1998) Over vs. Covert movement. Syntax 1, pp. 128-191.

Kiss, Katalin �.(1995) Focus is a Non-Uniform Phenomenon. In Inga
Kohlhof, Susanne Winkler & H.Bernhard Drubig (eds.) Proceedings of
the Goettingen Focus Workshop 69, pp.175-196.

Rizzi, Luigi. (1997) The Fine structure of the Left Periphery. In
Liliane Haegeman (ed.) Elements of Grammar: Handbook of Generative
Syntax. Dordrecht: Kluwer Academic Publishers, pp.281-337.

Rizzi, Luigi (2001) On the Position 'Int(errogative)' in the Left
Periphery of the Clause. In Guglielmo Cinque & Giampaolo Salvi (eds.)
Current Studies in Italian Syntax: Essays Offered to Lorenzo Renzi.
Elsevier Amsterdam, pp.287-296.

Rudin, Catherine (1988) On Multiple Wh-Questions and Multiple Wh-
Fronting. Natural Language and Linguistic Theory 6, pp.445-501.

Simpson, Andrew (1999) Wh-Movement, Licensing and the Locality of
Feature-Checking. In David Adger, Susan Pintzuk, Bernardette Plunkett
& George Tsoulas (eds.) Specifiers: Minimalist Approach. New York &
Oxford: OUP, pp. 231-247.

Simpson, Andrew (2000) Wh-Movement and the Theory of Feature Checking.
John Benjamins, Amsterdam.

ABOUT THE REVIEWER 

Olga Zavitnevich-Beaulac is a Postdoctoral Research Fellow in the
Asymmetry Project in the D�partement de Linguistique at UQAM. She
received her Ph.D. from the University of Cambridge, UK. Her research
interests include theoretical and comparative syntax with a focus on
syntax and morphology of wh and yes/no-questions, as well as problems
of optionality in natural languages.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue