| AUTHOR: Pereltsvaig, Asya
TITLE: Copular Sentences in Russian
SUBTITLE: A Theory of Intra-Clausal Relations
SERIES TITLE: Studies in Natural Language and Linguistic Theory 70
YEAR: 2007 (paperback edition in 2008)
Andrew Carnie, University of Arizona
The syntax of copular constructions and other instances of non-verbal predicates
has been a puzzle at least since Aristotle. In particular, understanding the
apparent differences among copular sentences that express statements of identity
(equative sentences) and those expressing memberships in classes (predicational
sentences) is a particular challenge given that the two constructions are often
similar but rarely wholly identical. Asya Pereltsvaig's book _Copular Sentences
in Russian_ is the most recent contribution to this discussion, and an important
one as well.
Pereltsvaig's book is both a careful investigation of the facts of Russian
copular constructions and an important advance theoretically. She proposes a
view of equative sentences (such as ''John is the doctor''), where the two DPs
stand in symmetric opposition to one another, i.e. the structure of such
sentences starts with a symmetric [DP DP] constituent. By contrast, she argues
that predicational sentences are formed by an NP (rather than DP) complement to
a predicational light verb which proxies the introduction of a subject. These
proposals have important consequences for modern Minimalist thinking. As
discussed below, if such proposals are correct they force us to revise our
understanding of the Antisymmetry proposal and make use of highly refined
versions of theta and case theories.
Chapter 1 starts with a quote from Conan Doyle's _The Adventures of Sherlock
Holmes_ ''It is dangerous to theorize without data, my friend Watson.'' This is
excellent advice for all linguists, especially with the current climate in
Minimalism, where data seems to sometimes take a backseat to discovering
elegance in theory. Pereltsvaig's book is an exception in this regard: it is
filled with both interesting data and important theorizing. In addition to
outlining the basic theoretical assumptions of the work, this first chapter lays
out the empirical cornerstone around which the rest of the book is written.
Russian has, among other similar constructions, two major patterns that appear
with the verb _byl_ 'be'. The first takes two nominative arguments and has
essentially an equative/identity interpretation. In the other construction, one
of the arguments appears in the instrumental case and has a predicative
interpretation. (Examples from Pereltsvaig, p 7)
1) a) Lenin byl Vladimir Uljanov Equative
Lenin.nom was Vladimir Ulyanov.nom
'Lenin was Vladimir Ulyanov'
b) Gnomy byli suščestvami rabotjaščimi Predicative
Gnomes.nom were creatures.inst laborious.inst
'(The) gnomes were laborious creatures.'
Pereltsvaig claims that the pattern in (1a) -- the NOM-NOM pattern -- involves a
symmetrically constructed small clause that is the complement of a functional
projection, which in turn has no thematic structure and is realized by _byl_.
One of the DPs moves to the specifier of TP for EPP reasons, disrupting this
symmetric pattern (similar to the proposal made by Moro 2000).
2) NOM-NOM pattern
[TP DP-i T [FP byl [DP t-i DP]
In the pattern in (1b) -- the NOM-INST pattern -- the predicate is an NP (not a
DP), and is the complement of a light verb _byl_ rather than to a homophonous
functional category. The subject argument is inserted in the specifier of this
light v. Simplifying Pereltsvaig's diagram slightly for ease of discussion, this
results in a structure like that in (3):
3) [TP DP-i T [vP t-i byl NP]
Her analysis relies on the distinction between DPs and NPs, so Chapter 2 deals
with the thorny controversy of whether Russian has DPs or not, despite its lack
of determiner morphology. Pereltsvaig departs from the view in Baker (2003) that
the distinction between nouns and other categories is the association of a
referential index with only nominal categories. Instead, she adopts the view,
common in the Distributed Morphology literature, that referentiality is a
feature of determiners. Using evidence from coordination, demonstratives,
pronominalization and numerals, Pereltsvaig argues that the post copular
nominals in the NOM-NOM pattern are DPs but those in the NOM-INST pattern are
NPs. This chapter also includes a short discussion of the ''reversibility'' of
arguments in both NOM-NOM and NOM-INST patterns, and argues that the movements
involved are of different types.
The third chapter argues that symmetric structures are not only necessary,
contra the Antisymmetry hypothesis, but are a predicted consequence of the
labeling part of Merge. Given the options of asymmetric projection, union and
intersection of categorial features, the first is found when the two items being
merged exhibit a mismatch in features (such as the merging of a verb with a
nominal complement). The other two are the consequence of merging two
identically categorized elements (such as two DPs in equative constructions).
Pereltsvaig gives related evidence for symmetric Merge in Edo serial verbs, and
from the range of features that express agreement in various kinds of Russian
Chapter 4 is the meat and potatoes of the book and is perhaps the best part of
the work. Its topic is the way in which thematic relations are assigned to
arguments both in normal verbal clauses and in the various kinds of copular
constructions. Pereltsvaig assumes that thematic requirements hold at LF and
uses the three methods of thematic transmission proposed by Higginbotham (1985):
(i) theta-marking (essentially equivalent to the traditional notion of thematic
assignment from a predicate to one of its arguments, augmented by the
requirement that this can only happen under a sisterhood relation), (ii)
theta-identification (the discharge of thematic structure by sharing a feature
with a sister or c-commanding head) and finally (iii) theta-binding (where a
c-commanding head absorbs the theta-role. The mechanics of this are irrelevant
here, but essentially this last mechanism has the effect of the determiner
saturating the argument structure of a predicative NP so that the nominal can
function as an argument.) Theta binding is the relevant mechanism for the
NOM-NOM pattern where we have two referential DPs. Theta identification and
marking are used in the NOM-INST pattern. Building upon Fiengo & May's (1994)
theory of indexing, Pereltsvaig argues that coreference requirements between DPs
in the NOM-NOM is due to a loosened definition of binding. Pereltsvaig claims
that there is no ''identity'' copula in these constructions at all, the identity
happens through the familiar effects of the binding theory. The treatment of the
NOM-INST pattern is more complex. Pereltsvaig uses the bifurcation of theta
roles that Samek-Lodovici (2003) proposes, whereby theta roles are composed of
two parts: a thematic index (the contentful part of a theta role) and argument
variable (the slot in the argument structure occupied by the argument). NP
predicates come into the structure with ''deficient'' theta structure, and as such
cannot license arguments without support. The support is provided by a light
verb, which has an argument variable without a thematic index. It acquires a
thematic index from the N head via the process of theta-identification. This
explains a number of mysterious properties about copular constructions including
the non-locality of the theta role assignor and the fact that DPs are disallowed
with this predicative v. (A mystery to me is why indefinite DPs are possible in
this construction in English, however). The analysis also explains why equative
constructions are always individual level, but predicatives allow both stage and
individual level interpretations (the lack of an equative predicate means that
there is no place for an event argument in the LF of the clause), and certain
pragmatic restrictions on the use of the two types of clauses.
The final content chapter in the book addresses case relations in copular
constructions in Russian. Pereltsvaig adopts the view that abstract Case is an
argument-licensing/identification requirement that holds at Logical Form (LF)
and is universal. Morphological case, by contrast is a surface requirement of
Phonetic Form (PF) and is subject to language-specific restrictions. Pereltsvaig
argues that the second DP in the NOM-NOM construction is simply the result of
default morphology applying, because nominals in Russian require case
morphology, without a structural case they receive the default nominative.
While I'm willing to buy this account of case, I was a little puzzled by the
fact that Pereltsvaig spends several pages trying to argue that nominative is
the universal (or at least typologically most common) default case. Since
surface morphological case can be language-specific I wasn't entirely clear on
why it was relevant that there are other languages that seem to use the
nominative as the default. This is particularly puzzling to the speaker of
English, for whom the accusative is the clear default. This doesn't take away
from Pereltsvaig''s argument, but it was an unusually flawed piece of
argumentation in an otherwise tightly argued text.
With respect to the NOM-INST pattern, recall that in this construction,
Pereltsvaig argues there is a real copular verb. She claims that the
instrumental case here is a quirky or inherent case associated particularly with
this verb (and other copula-like verbs). She further argues that the fact that
instrumental case is used here (rather than say dative or genitive) is not
accidental. It due to the fact that dative case is linked exclusively to
affected objects and goals, and genitive is tied to quantitative objects,
whereas instrumental is the default inherent case used in all other
circumstances. The chapter ends with an interesting extension of the analysis to
case marking with Russian depictive secondary predicates.
Chapter 6 provides a useful summary of the main claims of the book.
This book is a revised version of the author's (2001) McGill University
Dissertation. Despite the fact that in the ever-changing dynamic of Generative
Grammar this work is relatively old, (for example, it is pre-phase theoretic),
the theoretical and empirical content is still very relevant for both Slavicists
and theoretical syntacticians. The author has clearly gone to a great deal of
effort to update and polish the work for this book-level treatment. I note,
however, two research programs on copulas, which probably deserved some
additional discussion in the revisions for this book. One is the set of papers
by Heycock and Kroch (1998, 1999, 2002), and the other is Adger and Ramchand (2003).
While the non-syntactician will be challenged by theoretical details of this
work, most working minimalists will find the tight argumentation, logical
organization and clear and easy tone of the book refreshing. The ideas flow well
into one another and for the most part the theoretical proposals are given the
right amount of contextualization. Each chapter ends with a very useful
point-form summary of the main claims of that section and one is never left
wondering where the discussion is going. Chapter 4 in particular is a tour de
I was less impressed by chapter 3, which I thought was less carefully argued. In
this chapter, Pereltsvaig sets up a number of strawman positions about
Antisymmetry and then demolishes them. It wasn't clear to me however, that any
of the strawman positions she takes (such as the claim that SVO should be
typologically the most common order) are ever actually made by proponents of
Antisymmetry. But this chapter is the exception. Another minor complaint is
about the nature of the Extended Projection Principle (EPP), which Pereltsvaig
claims to be sensitive to referential elements only, flying in the face of, for
example, the fact that expletives can satisfy this constraint.
The theoretical treatment is a real advance. Pereltsvaig draws together
theoretical tools from morphology, semantics, and syntax to explain the complex
behavior of these kinds of sentences. While the idea that there is no equative
copula is not new (dating at least as far back as Jespersen 1924), Pereltsvaig's
take on the question is very novel. Rather than trying to completely assimilate
equatives into a predicative syntax (as done, for example, in Heggie 1988 among
others), she argues for a clear structural distinction between equatives and
predicatives, but suggests that this contrast finds its source precisely in the
absence of a verb in the former cases.
As one of the defenders of the notion that there is an equative predicate
(Carnie 1995, 1997, 2000), I feel compelled to point out that Pereltsvaig does
not discuss the extensive literature on the equative/predicative distinction in
the Semitic and Celtic languages (including the Carnie and Adger & Ramchand
articles mentioned above, Zaring (1996), Rouveret (1996), Hendrick (1994,1996),
Doherty (1996), Doron (1986) and Fassi Fehri (1993) among many others), In these
languages, equatives are marked by the *addition* of morphological structure
(typically in the form of a pronoun) and we find less overt morphological
structure in the predicatives. This would at least at first glance appear to be
exactly the opposite prediction of Pereltsvaig's analysis, so it's a shame the
data aren't discussed. Although interestingly, I suspect these data are in fact
entirely compatible with the analysis given in her book. The fact that the extra
material is a pronoun is compatible with her claim that the symmetric DP
structure is directly the complement of a functional head.
Another area that I wish had been addressed was the evidence that the two DPs in
equatives do, in fact, show asymmetries. For example, extraction from within the
first DP has been shown to be more difficult than that of the second (Moro
1997). This again appears to be prima facie evidence against a symmetric
approach. Although, again ironically, I think it is easily accounted for in
Pereltsvaig's system as the symmetries only exist at the earliest levels of
These complaints should not detract from this very interesting monograph. The
scholarship here is of a very high quality and the book will be of great value
not only to Russian linguists but also to researchers in the syntax, semantics
and morphology of copular constructions cross-linguistically.
Adger, David and Gillian Ramchand (2003) Predication and Equation. _Linguistic
Inquiry_ 34: 325-359.
Baker, Mark (2003) _Lexical Categories: Verbs, Nouns and Adjectives_. Cambridge:
Cambridge University Press
Carnie, Andrew (1995) Non-verbal Predication and head-movement. Ph.D.
Carnie, Andrew (1997) Two types of Non-Verbal Predication in Modern Irish.
_Canadian Journal of Linguistics_ 42: 57-73
Carnie, Andrew (2000) On the notions XP and Xo. _Syntax_ 3: 59-106
Doherty, Cathal (1996) Clausal structure and the Modern Irish copula. _Natural
Language and Linguistic Theory_ 14: 1-46.
Doron, Edit (1986). The pronominal 'copula' as agreement clitic. _Syntax and
Semantics_ 19 (_The syntax of pronominal clitics_, ed. Hagit Borer). New York:
Academic Press. 313-332.
Fassi Fehri, Abdelkadar (1993) _Issues in the Structure of Arabic Clauses and
Words_. Dordrecht: Kluwer.
Fiengo, Robert and Robert May (1994) _Indices and Identity_. MIT press.
Heggie, Lorie (1988). The Syntax of Copular Structures. Ph.D. Dissertation,
University of Southern California
Hendrick, Randall (1994) The Brythonic Copula and Head Raising. In _Verb
Movement_, ed. David Lightfoot and Norbert Hornstein. Cambridge, Cambridge
Hendrick, Randall (1996) Syntactic Effects of Suppletion in the Celtic Copulas.
In Borsley and Roberts (eds). _The Syntax of the Celtic Languages: A comparative
perspective_. Cambridge University Press. 75-96.
Higginbotham, James (1985) On Semantics. _Linguistic Inquiry_ 16: 547-593
Jespersen, Otto (1924) _The Philosophy of Grammar_. London: Allen and Unwin.
Moro, Andrea (1997) _The Raising of Predicates_. Cambridge University Press.
Moro, Andrea (2000) _Dynamic Antisymmetry_. MIT Press.
Rouveret, Alain (1996) Bod in the present tense. In Borsley and Roberts (eds).
_The Syntax of the Celtic Languages: A comparative perspective_. Cambridge
University Press. p 125-170
Zaring, Laurie (1996) Two ''be'' or not two ''Be'': Identity predication and the
Welsh Copula. _Linguistics and Philosophy_ 19:103-142
ABOUT THE REVIEWER
Andrew Carnie is an associate professor of linguistics at the University of
Arizona. He wrote a dissertation on copular constructions in Irish and has other
interests in phrase structure, copular constructions, argument hierarchies and
case systems as well as the syntax of Celtic and other VSO languages. He has a
well-known introductory syntax textbook (_Syntax: A Generative Introduction_)
from Wiley Blackwell, and has two books from Oxford University Press in 2008
(_Constituent Structure_ and the forthcoming _Irish Nouns_).