Date: Wed, 22 Jan 2003 16:04:28 -0500
From: Marc Picard <picard@vax2.concordia.ca>
Subject: Romance Phonology and Variation
Wiltshire, Caroline R., and Joaquim Camps, eds. (2002) Romance Phonology
and Variation. John Benjamins, x+238pp, hardback ISBN 1-58811-079-6,
$72.00.
Marc Picard, Concordia University
This book is a collection of selected papers from the Thirtieth Annual
Meeting of the Linguistic Symposium on Romance Languages (LSRL 30) which
took place in February, 2000 at the University of Florida in
Gainesville. The thirteen articles are preceded by a section entitled
"Romance phonology and variation" in which the editors (henceforth W&C)
present an overview of the articles. Although some of these papers may
combine issues of the two major themes, W&C introduce them in terms of
either phonology or variation. The phonology-based papers can be roughly
categorized according to whether they deal with either
markedness/correspondence, typology or representations, while the
variation-based articles are either dialectal or contact-oriented in
nature.
The first paper is Barbara E. Bullock's "Constraining the vagaries of
glide distribution in varieties of French" wherein she examines
"language-external evidence that points to how speakers of French
actually treat surface glide-vowel (GV) sequences in linguistic
performance" (p. 11). She concludes on the basis of data drawn from
child language, language games and dialectal variation that "speakers
actually treat surface GV strings as belonging, as a unit, to the
nucleus of the syllable" (p. 17), thus calling into question previous
analyses of French glide distribution in which it was assumed that some
or all GV sequences are derived from VV ones.
"On the relationship between comprehension and production data in
codeswitching" by Paola E. Dussias deals with a phenomenon known as 'the
functional element effect' whereby "there is a systematic favoritism for
switches that involve certain grammatical categories over others during
codeswitched speech" such that "whereas functional elements tend to
appear in one language, their complements appear in the other language"
(p. 27). Having shown in a previous study that Spanish-English
bilinguals took significantly longer to read sentences like:
(1) La maestra no sabia que the boy had left
where the functional head and its complement appeared in the same
language, than they took to read sentences like:
(2) La maestra no sabia that the boy had left
where these elements appeared in different languages, the author sets
out to investigate whether these comprehension preferences can be
replicated in production data. Her results indicate that the two types
of preferences do not always match, and she suggests possible
explanations based on certain linguistic, psycholinguistic and discourse
principles.
In "Focus, word order variation and intonation in Spanish and English:
an OT account", Rodrigo Gutiérrez-Bravo compares languages like Spanish,
which have what is known as focus-related word order variation, and
English, which does not, as shown in the following sentences:
(3) Ayer compró el periódico Juan
(lit. "Yesterday bought the newspaper Juan")
(4) John bought the newspaper yesterday
Rejecting "recent analyses [which] connect this word order variation in
Spanish . . . with some structural condition that requires a focused
constituent to occupy a specific syntactic position" (p. 39), he argues
that given the fact that the focused subject in both the Spanish VOS
structure and the English SVO structure are assigned sentential stress,
the variation can be accounted for in an Optimality Theoretical model
through the differential ranking of a prosodic constraint - the Nuclear
Stress Rule (NSR) - relative to syntactic constraints on focus and
subject position.
"Morphological complexity and Spanish object clitic variation" by David
Heap examines the two different types of third-person object-pronoun
paradigms that exist in various social and geographical dialects of
Spanish. These two types of pronominal paradigms, which he refers to as
the etymological (or case-based) system and the referential system, make
differential use of morphological contrasts mostly based on the features
[DATIVE], [FEMININE] and [PLURAL]. For example, in the first system we
find structures like:
(5) Lo conoci "I met him"
(6) Le di un regalo "I gave her a present"
which correspond in the second system to:
(7) Le conoci "I met him"
(8) La di un regalo "I gave her a present"
Given that an analysis which considers the pertinent features to be
binary overpredicts the number of contrasts which are in reality highly
constrained, the author proposes "a Feature Geometry account which
allows for the attested range of variation in pronoun paradigms without
opening the door to unconstrained variation" (p. 55).
In "Catalan phonology: cluster simplification and nasal place
assimilation", Dylan Herrick examines the complex interaction of these
two processes "from the perspective of a parallel non-serial version of
Optimality Theory" (p. 69). The problem is that cluster simplification
(CL) and nasal place assimilation (NPA) in Catalan yield opaque surface
forms, as shown in the following example:
(9) /tin+k bint botas+s/ > [tiN bim bot@s] "I have twenty wineskins"
where the first two words undergo CL but only the second undergoes NPA
since the velar nasal [N] fails to assimilate to the following labial.
By resorting to Correspondence Theory and output-output constraints, the
author proposes to "account for the apparent opacity . . . without the
need for additional theoretic machinery such as cyclicity, multiple
levels, or underspecification" (p. 69)
The central issue addressed by D. Eric Holt in "The articulator group
and liquid geometry: implications for Spanish phonology present and
past" is one that has preoccupied a number of generative phonologists
over the years, viz., whether liquids are continuants or
non-continuants. His goal is to "propose a novel approach to the
understanding of the ambivalent status of the feature [±continuant] of
/l/, whose value is not universally accepted" (p. 85). To this end, he
reanalyzes the synchronic process of Spirantization in Spanish, zeroing
in on its oft-debated post-liquid application before voiced labial and
dorsal stops (/lb lg/ > [lB lG]) but not before coronals (/ld/ > [ld]),
and he also discusses various historical changes involving /l/ in that
language.
The purpose of José Ignacio Hualde's article "Intonation in Spanish and
the other Ibero-Romance languages: overview and status quaestionis" is
to "examine some specific issues in the intonation of Spanish and its
close relatives that have arisen in recent work and are currently
controversial" (p. 101). The author focuses on three issues in the
context of the Autosegmental-Metrical model of intonational analysis as
applied mainly to Spanish and also peripherally to Portuguese and
Catalan. These are (1) the phonological analysis of rising pitch-accents
which are defined as tones associated with stressed syllables, (2) the
nature and phonological characterization of final declarative contours,
and (3) intonational phrasing where variance can be attributed to
pragmatic considerations such that, for example, a sentence like:
(10) Mariano me dio la moneda de oro "Mariano gave me the gold coin"
can have a different intonational contour depending on whether the
context is a response to something like "what did Mariano give you?" as
opposed to "what did Mariano do?".
In "'Partial' Spanish: strategies of pidginization and simplification
(from Lingua Franca to 'Gringo Lingo')", John M. Lipski examines the
kind of widespread Spanish foreigner talk which stereotypically employs
the infinitive as default verb and *mi* as subject pronoun, and omits
definite and indefinite articles, e.g., *mi ver soldado* for *yo veo a
el soldado* 'I see the soldier'. Given that "today no known
second-language learners of Spanish speak in this fashion" (p. 118), he
sets out to examine real examples of reduced language in Afro-Iberian,
'Moorish' Spanish, Anglo (mostly American), Philippine, Chinese and
Amerindian pidgin Spanish, Basque Spanish, and Spanish child language in
order to find answers to the two following questions: "[w]hat . . . is
the relationship between imagined and real 'foreigner' Spanish, and how
has a reasonably cohesive model of such 'almost-Spanish' remained in the
Spanish collective unconscious for so long?" (p. 118).
The gist of D. Gary Miller's "The death of French in medieval England"
consists of the presentation of "[s]everal new pieces of evidence in
support of [the] thesis . . . that Anglo-French (AF) was dead by ca.
1400" (p. 145). The article comprises sections on the expansion of
French in England, the resistance of English to this linguistic
onslaught, the lexical and morphological signs of the imminent death of
AF,and the characteristic features of late AF that point to convergence
with English as found both in literature and in the records of the
London Grocers' Company. Essentially, it is the sharp decline of French
calques in English along with the significant increase of French
suffixes in English hybrids that lead the author to conclude that by ca.
1400, AF was in the throes of "a typical language death situation, in
which the dying language employs extreme measures of convergence as an
attempted survival strategy" (p. 157)
As the title "Discourse context and polysemy: Spanish *casi*" clearly
indicates, Scott A. Schwenter's paper has implications for both Hispanic
linguistics and semantics/pragmatics. In the former case, the central
issue revolves around an innovative use of *casi* 'almost', notably in
Spain's Valencian Community. Thus, beside the normal use of this adverb
in sentences such as:
(11) ¡Casi no me lo dices! 'Now you tell me!" (lit. 'You almost don't
tell me!')
speakers of this dialect can say with the same meaning:
(12) ¡Casi me lo dices! 'Now you tell me!" (lit. 'You almost tell me!')
What the author shows is that this 'inverted' *casi*, as he terms it, is
neither ironic nor possible in a non-temporal setting, and so should be
analyzed as a distinct polysemy. In sum, he "offers up a case study
which illustrates how to distinguish between contextually-determined
interpretations of a lexical item, on the one hand, and conventionalized
senses - polysemies - of the same lexical item, on the other" (p. 161).
Given previous claims to the effect that reduplication in French nouns
and adjectives is too unpredictable and unproductive to be amenable to a
unified analysis, Mary Ellen's Scullen's intent in "New insights into
French reduplication" is to "present new data on 'invented'
reduplications in the domain of French baby-talk which strongly suggest
that French reduplication is indeed productive and that it can be
analyzed straightforwardly with a constraint based approach such as
Optimality Theory" (p. 177). In essence, what she succeeds in showing by
asking native speakers to invent reduplicated forms is that there is a
productive process that is basically right-edged and consonant-initial,
e.g., *toto* < *auto* 'car', and that preserves a final coda consonant
while deleting it in the reduplicated syllable, e.g., *bébête* < *bête*
'silly, foolish'. Although the language does contain a number of
exceptions, such as *pipi* < *pisser* 'to piss' and *fifi* < *fille*
'girl', the vast majority of these are clearly lexicalized since such
configurations are almost never produced in invented forms.
In Optimality Theory, it has been proposed that certain constraints must
work in tandem and that these conjoined constraints are violated if and
only if both conjuncts are violated. In "Local conjunction in Italian
and French phonology", Bernard Tranel and Francesca Del Gobbo present
arguments on three fronts in favor of the conjoined constraint
{ONSET&NOCODA}. First, they claim that "this constraint makes sense of
the acquisition pattern of Dutch syllable structure" (p. 191) which has
the peculiarity of being (1) CV, (2) CV, CVC, (3) CV, CVC, V, i.e., of
having V syllables without concomitant VC syllables. Second, they argue
that Local Conjunction can be "shown to play a key role in explaining
the suppletive distribution of the masculine plural definite article in
Italian" (p. 191) whereby *gli* occurs before vowels, geminates and
obstruent + obstruent clusters, and *i* everywhere else. Third, they
propose that {ONSET&NOCODA} can account for "the exceptional behavior of
*h-aspiré* words with respect to optional schwa deletion in French" (p.
191), e.g. *le héros* /l@ero/ 'the hero' (where /@/ is a mid-low front
rounded vowel) as opposed to *l'étau* /leto/ 'the vise'.
Having previously "submitted a number of observations which strongly
suggest that mora count has a major role to play in an analysis of main
stress in non-verbs in Brazilian Portuguese" (p. 219), notwithstanding
the widespread belief among contemporary phonologists that "languages
without a distinctive quantity opposition cannot have a weight sensitive
stress rule" (p. 221), W. Leo Wetzels sets out to discover in "On the
relation between quantity-sensitive stress and distinctive vowel length:
the history of a principle and its relevance for Romance" whether
Trubetzkoy is really the source, as is often claimed, of the argument
that contrastive vowel length is a prerequisite to weight-sensitive
stress. His general conclusion is that "it does not make much sense to
refer to Trubetzkoy to dismiss the possibility of a weight-sensitive
stress rule for Romance" (p. 232) since his main concern was to find a
way to represent distinctive vowel length rather than weight.
COMMENTS
Anyone who has patiently read through the foregoing descriptions may
find, as I did, that there is something incongruous about this
collection. Indeed, it might well have been titled "Romance Phonology et
al." for the thread that connects the articles W&C have included under
the theme of 'variation' with those that are phonologically oriented is
a mighty fine one. In fact, the same might also be said of the
'variation' papers themselves which constitute a pretty disparate bunch,
dealing as they do with codeswitching, morphology, pidginization,
language death and pragmatics. Still, this does not in any way detract
from the quality of the individual articles themselves which are, on the
whole, informative, interesting and well written. Most linguists, even
those who are not steeped in Romance phonology, should find something of
interest in this volume.
|
