EDITORS: Petra Sleeman and Harry Perridon
TITLE: The Noun Phrase in Romance and Germanic
SUBTITLE: Structure, variation, and change
SERIES TITLE: Linguistik Aktuell/Linguistics Today 171
PUBLISHER: John Benjamins
YEAR: 2011
Jennifer Culbertson, Center for Language Sciences, University of Rochester, NY
SUMMARY
This book is a collection of papers presented at the conference “Variation and
Change in the structure of the noun phrase in Germanic and Romance: Autonomous
developments or result of language contact?” The chapters are separated into two
parts, the first focused on variation and the second on change. The topics
covered are fairly broad, including the syntax and semantics of various
components of the DP, cross-linguistic variation and variation within a given
language, contact-induced change, and grammaticalization.
The book begins with an introductory chapter by Perridon and Sleeman covering
similarities and differences across the Romance and Germanic languages with
respect to definiteness marking, the position of adjectives, the function of
weak adjective declensions (in Germanic), the evolution of genitives, and the
emergence of determining possessives (in Germanic). The general goal of the
chapter is to raise the question of how new nominal categories and structural
options emerge -- in some cases they are present in the mother language, in
others they arise independently in various daughter languages -- and how they
are integrated into the DP. For example, although the Romance languages
apparently inherited definiteness marking from Late Latin, the Germanic
languages instead each developed it after having become separate languages. The
authors present the (apparently near simultaneous) development of definiteness
marking system as an opportunity to examine possible motivations for this
emergence, sources of definite articles and affixes, and effects of the new
system on the coherence of the DP as a whole.
The first chapter, “Scaling the variation in Romance and Germanic
nominalizations” by Artemis Alexiadou, Gianina Iordăchioaia & Florian Schäfer,
proposes a number of structural distinctions among nominalization structures in
Germanic and Romance in order to derive systematic differences in their
behavior. The authors show that although all nominalizations mix both verbal and
nominal features, behaviors associated with particular nominalizations in
Romanian, Spanish, English, and German fall on two scales -- one ranging from
less to more “nominal”, the other less to more “verbal”. The micro-variation
found, they argue, is constrained by the presence of particular verbal and
nominal layers within the DP which are each compatible with only some behaviors
on each scale. For example, most verbal nominalizations (e.g. Spanish verbal
infinitives) feature tense and aspect projections which make licensing of
nominative case and occurrence with modal or auxiliary verbs possible. On the
other hand, the presence of nP in more nominal nominalizations (e.g. English
nominal gerunds) makes adjectival modification possible and leads to the
expression of gender features.
The second chapter, “What all happens when a universal quantifier combines with
an interrogative DP” by Robert Cirillo, investigates the structure and
derivation of interrogative DPs which co-occur with a universal quantifier.
Cirillo assumes a stranding analysis of floating universal quantifiers,
originally proposed by Sportiche (1988) and Giusti (1990), whereby ‘all’ in a
sentence like “The students have all read the book” is base-generated as the
head of QP and stranded when the DP it dominated moves to a higher position.
Presenting data first from Germanic, in which universal quantifier +
interrogative DP combinations seem to closely parallel universal quantifier +
non-interrogative DP, Cirillo suggests a unified analysis of the two whereby the
quantifier selects the interrogative DP. However, he then presents data from
German, American English, Swedish, British English, and Romance which suggest,
on the contrary, that the universal quantifier is actually the complement of the
interrogative DP. He argues in favor of this analysis over a unified one,
showing that stranding can be obligatory, optional, or impossible depending on
the language, and that in some cases universal quantifiers and interrogative DPs
cannot be combined at all.
The third chapter, “Micro-diversity in Dutch interrogative DPs” by Norbert
Corver & Marjo Van Koppen, describes variations in the use of ‘wat voor ’n N’
(shown in 1 below) within and across dialects of Dutch, and provides a syntactic
analysis to derive it. The localization of variation relates to the possibility
of subextracting the wh-word ‘wat’, yielding a discontinuous phrase (lit. “what
have you for a car bought?”); some dialects do not permit it, but if a dialect
does permit it, the non-split pattern is necessarily allowed as well.
(1) Wat voor ’n auto heb je gekocht?
What for a car have you bought
The authors assume Bennis et al.’s (1998) analysis of the standard Dutch
construction in (1) as involving a predicate relation between the pronoun ‘wat’
and the noun, represented as a DP-internal small clause. The surface position of
‘wat’ in the left periphery is derived via predicate displacement within the DP
-- ‘wat’ moves from predicate position to Spec,DP. Differences in the
availability of the split construction across Dutch dialects are due, the
authors argue, to whether the wh-word wat moves to the edge-position in the DP
or not, which in turn depends on properties of the D head which they show to
differ across dialects.
The fourth chapter, “Noun phrase structure and movement” by Johanna L. Wood &
Sten Vikner, provides an account of differences in position, agreement
morphology, and semantics of etymologically related words ‘so’ and ‘such’ in
English, Danish, and German. The authors show that across these three languages,
both ‘so’-type words and ‘such’-type words can occur in pre- and post-article
position, and can modify either the entire DP or the AdjP only. They argue that
in the post-article position, these words are derived as attributes (they are in
prototypical adjective position). However, they argue against this analysis for
pre-article ‘so’ and ‘such’ words, instead advocating that they involve
predicate raising from a right-NP-adjoined small clause. This is supported, for
example, by the fact that in pre-article position these words are inflected only
in languages which inflect predicate adjectives (not in German for example). The
authors conclude by discussing some changes to these words in German, for
example ‘so’ has widened its scope from expressing degree to kind as well --
unlike in English ‘so’ can occur without an adjective, modifying the DP (“so ein
Hotel”).
The fifth chapter, “A unified structure for Scandinavian DPs” by Susanne
Lohrmann, is concerning with determiner doubling, definiteness marking, and
adjectival inflection in Scandinavian. In Swedish, Norwegian, and Faroese, the
presence of an adjective in a definite DP triggers an additional determiner (as
in 2 below in Swedish), which can encode some aspect of definiteness like
inclusivity or specificity. Danish and Icelandic on the other hand do not
feature this double definite structure (e.g. in Danish a preadjectival article
is introduced, but the suffix is omitted). Use of adjectival inflection in
Scandinavian languages also results in additional semantic effects related to
definiteness.
(2) a. film-en b. den rolig-a film-en
film-def def funny-weak.infl film-def
The author argues that each component of definiteness marking can identify a
distinct semantic feature in the representation. In those languages with double
definiteness marking, use of the suffixed definite article when an adjective is
present is shown to correlate with specificity (or identifiability). The
preadjectival article signals the introduction of a new modified definite entity
into the discourse. Adjectival inflection individuates the members in the A+N
denotation. The author proposes the same general structure of the DP for all
Scandinavian languages, arguing that the difference between them is in whether
the specificity and novel discourse entity features head their own phrases or
are combined.
The sixth chapter, “A semantic approach to noun phrase structure and the
definite -- indefinite distinction in Germanic and Romance” by Ulla
Stroh-Wollin, argues for a semantically-motivated model of noun phrase structure
and shows how structural differences account for cross-linguistic variation in
the use of definite and indefinite noun phrases. The author argues in favor of a
notion of the definite -- indefinite distinction which is (purely pragmatically)
based on how the restriction of the set of referents in derived, which takes
place in the D-domain. The Germanic languages accomplish this by moving a
lexically contentful noun to D, however when an adjective is present a different
strategy must be used; for example, the double definiteness languages merge a
free definite article in D, formally specifying the set of referents, while
Icelandic moves the entire dP to Spec,DP. Romance languages on the other hand
are argued to have a null affix in d which attracts all nouns, and imposes a
restriction on the set of referents.
The seventh chapter, “Definite determiners in two English-based creoles:
Specificity or definiteness?” by Ekaterina Bobyleva, investigates whether the
distribution of definiteness marking in Jamaican and Sranan (two English-based
creoles) can be accounted for using the same constraints underlying determiner
use in the substrate language Gbo. This hypothesis had been proposed to explain
why these (and other creoles) do not mark definiteness categorically as their
European lexifiers do. In Gbo, overt determiners only occur with specific NPs --
perhaps Jamaican and Sranan determiner use follows this pattern rather than the
English one. The author uses corpus evidence to show, however, that in both
creoles overt definite determiners generally follow a definiteness- rather than
specificity-based pattern. Bare definites in both languages are most likely to
occur in prepositional phrases and with uniquely referring name-like NPs.
English is suggested as the source of the overt definiteness markers in both
Sranan (the determiners ‘da’ and ‘den’ from ‘that’ and ‘them’) and Jamaican
(‘di’ from ‘the’). The author suggests that the omission of the determiner in
prepositional phrases and for uniquely determined referents comes from universal
pragmatic principles; in the former case definite marking is less likely because
the NP typically has low discourse prominence, in the latter case the identity
of the referent is self-evidence and therefore overt marking unnecessary.
The eighth chapter, “Form-function mismatches in (formally) definite English
noun phrases” by Christopher Lucas, marks the transition from Part 1
(“Variation”) to Part 2 (“Change”), and proposes an explanation for why some
morphosyntactically definite nouns do not appear to behave as semantically
definite. The first case addressed is so-called activity-denoting weak
definites, which function semantically parallel to bare NPs in English (Carlson
et al. 2006). After providing data to show that weak definites behave not like
typical definites or indefinites but rather like bare NPs -- with an
activity-denoting interpretation, i.e. the weak definite in “Go to the pub” is
parallel to the bare NP in “Go to church” -- the author presents a diachronic
source for the difference in article use. In particular, he shows that the
activity-denoting use of some bare NPs began before the definite article was
obligatory in English (Old English, Middle English period), and once the
definite article did become obligatory, these lexical items (at least in this
context) were shielded from adopting it. Other lexical items did not evolve
their activity-denoting weak interpretation until during the Late Middle English
period, and thus the definite article appears obligatorily. The author further
argues that although new lexical items can come to be used in this way, whether
they enter the class of bare NPs or the class of weak definites is determined
through analogy (i.e. ‘the cell’ with ‘the telephone’, but ‘synagogue’ with
‘church’). The chapter then turns to relational weak indefinites, e.g. “The bank
of the river”, arguing that the interpretation of relational nouns -- which is
necessary coupled with another category, here “the river”, and cannot exist
independently -- is more compatible with definiteness than indefiniteness,
explaining why these (mostly) occur with definite articles.
The ninth chapter, “The emergence of the definite article in English” by Paola
Crisma, picks up on the question of when the definite article began to be used
regularly in English using corpus data, and proposes a source for it. The author
begins by introducing the Old English morpheme ‘se’, which was seemingly used
both as a demonstrative and a definite article. She then argues in favor of
Greenberg’s (1978) distinction between demonstrative and definite article,
namely that a demonstrative has become a definite article when it is obligatory
and generally indicates identifiability. By this reasoning a language has a
definite article if a noun phrase can only be interpreted as semantically
definite if it is overtly marked with it, and an overt D must be expressed even
when semantic definiteness is expressed by other means (at least for arguments).
Corpus data reported show that in Old English, like other languages with
definite articles, even intrinsically definite proper names appear almost
categorically with ‘se’ in certain contexts (when an adjective is present, as in
Italian and German, and the DP is an argument). A representative sample of the
data also reveals that, bare NPs aside, almost all subject and object noun
phrases interpreted as definite co-occur with ‘se’, suggested that indeed the
definite article system was in place by the Old English period. Timing and
surface similarity lead the author to suggest Celtic influence as the cause of
this development.
The tenth chapter, “On the syntax of Romanian definite phrases” by Alexandra
Cornilescu & Alexandru Nicolae, examine variation in the position of the
definite article in Old Romanian. In Modern Romanian the definite article
‘-(u)l’ is a suffix always attached to the first noun or adjective in the DP. By
contrast in Old Romanian it could occur either on the first noun or adjective OR
on a lower noun, allowing other constituents (e.g. adjectives) to precede the
definite noun. The authors argue that the difference between Modern and Old
Romanian in terms of the position of the definite article is the result of the
fact that Old Romanian allowed the definiteness feature of D to be valued by a
c-commanded nominal phrase which need not be in local relationship to it (thus
the nominal phrase to which it is attached need not be the first c-commanded by
D). In Modern Romanian, however, local Agree is the only available option. They
further suggest that the possibility of multiple instances of definiteness
marking in Old Romanian, e.g. on both a prenominal adjective and a noun, might
have triggered the more restrictive conditions in Modern Romanian. The authors
propose that the existence of a lower article in Romanian may support the
hypothesis that the diachronic source for the article what the Latin
POST-nominal ‘ille’ rather than a prenominal demonstrative. Noting a strong
correlation between the use of the lower article and the presence of genitive
complement phrase, the authors suggest that lower article may have been used in
order to facilitate the use of the more economical (simpler) type of genitive
construction. This bare genitive could only be used when a definite noun was
immediately preceding, exactly the result if the suffix was attached to a lower
noun rather than a higher adjective.
The eleventh chapter, “Coexisting structures and competing functions in genitive
word order” by Elisabetta Magni, examines variation in genitive ordering in
Latin and English and investigates how alternative orders arose and the contexts
in which they are favored. In English, the ‘of’-genitive and ‘s’-genitive have
been in competition since Old English, with the ‘of’-genitive initially rare
before gaining ground and then being again supplanted by the ‘s’-genitive in
Present Day English. Although which genitive is used is subject to probabilistic
variation, ‘of’-genitive use is correlated with inanimacy of the possessor while
with animate possessors the ‘s’-genitive is preferred. This difference in
function is potentially related to the sources for the two constructions; the
‘of’-genitive comes from a prepositional construction used to express spatial
movement from an inanimate source, the ‘s’-genitive possible evolved from a
construction with an animate pronominal modifier ‘his’ (e.g. “the bishop of Rome
his laws”). In Latin, the author also illustrates a functional load distinction
for the two genitive orders -- genitive-noun order appears to be used when the
two parts of the phrase are to be interpreted as a unit, while noun-genitive
order indicates some contrastive or new information. The author suggests that
genitive-noun constructions may have grammaticalized during a period of OV
syntax, while the use of noun-genitive constructions may preserve the order from
semantically similar noun-adjective phrases.
The twelfth chapter, “Anaphoric adjectives becoming determiners: A corpus-based
account” by Freek Van de Velde, argues that certain anaphoric adjectives in
Dutch can function as determiners and shows that this use is increasing in
frequency. The author uses several criteria proposed to hold (for the most part)
of determiners -- mutual exclusivity with other determiners, obligatoriness (at
least when a noun phrase is used as an argument), linear ordering at the
periphery with respect to numerals and adjectives, and exclusion from predicate
position. The adjectives in question are anaphoric adjectives like ‘voornoemd’
(‘aforementioned’), ‘bedoeld’ (‘intended’), ‘gezegd’ (‘said’), each of which can
seemingly occur in constructions without a proper determiner as in (3).
(3) Voornoemde heer bezigde onheuse taal
Aforementioned man used inappropriate language
In these cases, the anaphoric adjectives do not co-occur with any other
determiners, follow inflectional patterns of determiners rather than adjectives,
precede all numerals, and cannot appear in predicate position. Although they are
sometimes preceded by an article, the author argues that this is not unusual for
elements that can function as determiners but can also fill another slot (e.g.
English adnominal elements like ‘many’). Using corpus data, the author
illustrates the trend away from using the article with these newly developing
determiners throughout the period of Late Modern Dutch (the corpus spans from
1841-1990).
The thirteenth chapter “From N to D: Charting the time course of the internal
rise of French n-words” by Viviane Déprez, investigates the structural position
of French n-words and proposes a set of diachronic steps connecting changes in
position with changes in modificational and quantificational properties. The
author begins by showing that contemporary French n-words like ‘personne’
(‘nobody’) and ‘rien’ (‘nothing’) behave like Ds rather than nominals, occurring
in argument positions with no determiner, and not triggering adjectival
agreement (although ‘personne’ for example can also still function as a noun
meaning ‘person’ when it does occur with a determiner). Further the
modificational properties of n-words are similar to those of existential
quantifiers and not other nominal expressions, for example n-words and
existential quantifiers can be modified by ‘d’autre’ (‘else’), while nominals
cannot (“quelq’un d’autre” (“someone else”), “personne d’autre” (“no one else”),
but not “*une personne d’autre”). The author then traces the evolution of these
n-words from nominals structurally placed within the NP layer, to elements which
are number and gender invariant but have climbed only to NumP, and finally to
quantificational elements high up in the DP. The latter two stages are
distinguished in that only in the quantificational stage can they be modified by
‘d’autre’.
EVALUATION
Overall, the articles that comprise this book present very interesting novel
data and analysis regarding noun phrase variation and change in Romance and
Germanic (although Germanic may be better represented than Romance in this
particular collection). The use of diachronic and corpus data to address
synchronic questions is in general well done and thought provoking. For example,
Van de Velde (chapter 12) uses data from written Dutch corpora spanning 150
years to argue that the determiner category in Dutch is in some sense not pure,
but admits certain anaphoric adjectives. Interestingly, the data also suggest
that despite DECREASING in frequency over time nevertheless these elements are
continuing their path of grammaticalization. Deprez (chapter 13) also makes
critical use of diachronic data in proposing a series of steps that French
negative words like 'personne' ('nobody') have taken as they changed category
from N to D. Bobyleva (chapter 7) is able to argue forcefully against a
prevailing hypothesis that constraints on definiteness marking in Jamaican and
Sranancan Creoles are derived from the substrate language Gbo. Rather the corpus
data she investigates shows clearly that while the latter is specficity-based,
the former is definiteness-based.
All the chapters basically assume a generative framework (e.g. Cirillo, Wood &
Vikner, Deprez), but in fact a number of them also address the
functional/typological (e.g. Bobyleva, Magni) and variationalist (e.g. Van de
Velde) perspective. While a number of the articles leave quite a bit of work to
be done, this book would certainly be a useful read for any linguist interested
in the syntax and morphosyntax of the noun phrase. That being said, a
significant number of the chapters (5, 6, 7, 8, 9,10 and 12, 13 to some extent)
address definiteness and definiteness marking, leaving less room for other
topics of interest. For example, although the introductory chapter discusses
some intriguing facts concerning word order--similarities and difference among
the Romance and Germanic languages, and various changes undergone--the topic is
somewhat neglected in this volume.
REFERENCES
Bennis, H., Corver, N. & den Dikken, M. (1998). Predication in nominal phrases.
Journal of Comparative Germanic Linguistics 1: 85-117.
Carlson, G., Sussman, R., Klein N., & Tanenhaus, M. (2006). Weak definite NP’s.
In Proceedings of NELS 36, Vol.1, C. Davis et al. (eds), 179-198. Amherst, MA: GLSA.
Giusti, G. (1990). Floating quantifiers, scrambling and configurationality.
Linguistic Inquiry 21: 633-641.
Sportiche, D. (1988). A theory of floating quantifiers and its corollaries for
constituent structure. Linguistic Inquiry 19: 425-449.
ABOUT THE REVIEWER
|
