Publishing Partner: Cambridge University Press CUP Extra Publisher Login
amazon logo
More Info

New from Oxford University Press!


Style, Mediation, and Change

Edited by Janus Mortensen, Nikolas Coupland, and Jacob Thogersen

Style, Mediation, and Change "Offers a coherent view of style as a unifying concept for the sociolinguistics of talking media."

New from Cambridge University Press!


Intonation and Prosodic Structure

By Caroline Féry

Intonation and Prosodic Structure "provides a state-of-the-art survey of intonation and prosodic structure."

Review of  Segmental and Prosodic Issues in Romance Phonology

Reviewer: Benjamin Schmeiser
Book Title: Segmental and Prosodic Issues in Romance Phonology
Book Author: Pilar Prieto Joan Mascaró Maria-Josep Solé
Publisher: John Benjamins
Linguistic Field(s): Phonetics
Issue Number: 19.3341

Discuss this Review
Help on Posting
EDITORS: Prieto, Pilar; Mascaró, Joan; Solé, Maria-Josep
TITLE: Segmental and Prosodic Issues in Romance Phonology
SERIES TITLE: Current Issues in Linguistic Theory 282
PUBLISHER: John Benjamins Publishing Company
YEAR: 2007

Benjamin Schmeiser, Department of Languages, Literatures, and Cultures, Illinois
State University

The book under review is part of the John Benjamins series entitled, 'Current
Issues in Linguistic Theory', and focuses on the phonetics and phonology of
Romance languages, of which varieties of Catalan, French, Italian, Portuguese,
and Spanish are researched. It is a collection of selected papers from the
conference, Phonetics and Phonology in Iberia (PaPI) 2005, which took place at
the Universitat Autònoma de Barcelona in June 2005 in Barcelona, Spain. The book
is divided into three sections: Segments and Processes, Prosodic Structure, and
Acquisition of Segmental Contrasts and Prosody, each section comprising four
articles. The introduction notes the conference's emphasis on the tradition of
'laboratory phonology' and the importance of empirical results, both of which
form the foundation of this volume.

The first section of the book is entitled 'Segments and Processes'. The first
article, by Nguyen, Wauquier-Gravelines, Lancia, and Tuller, discusses the
detection of liaison consonants in French. The authors are particularly
interested in what way liaison consonants are processed and represented
differently, if at all, from fixed consonants. Their findings are that liaison
consonants are more difficult to detect than fixed consonants (second word (W2)
initial consonants) by listeners. In their study, subjects tended to miss a
liaison target more often at the end of a short function word (determiner /
preposition), as opposed to an adjective or adverb. The authors surmise that the
exemplar-based approach to the analysis is more advantageous than the
auto-segmental one because ''the detection of liaison is to some extent dependent
upon the strength of the connections between the words at the juncture of which
liaison is realized'' (21). For example, if we compare a determiner + noun (e.g.
son hôtel 'his hotel') sequence to one that is adjective + noun (un lointain ami
'a distant friend'), the authors suggest that the former is more likely to form
a single processing unit, given the higher probability of co-occurrence of the
two words. In short, the authors do not see the difficulty of detecting liaison
consonants over W2-initial consonants to be attributable to acoustic
differences, but rather it is due to the way that the liaison is ''represented in
the speaker-listener's grammar'' (21).

The second article, by Daniel Recasens, considers how the speaker realizes
strings of phonetic segments in articulatory terms. The author tests the
predictive power of the degree of articulatory constraint (DAC) model by testing
the size of V-to-C anticipatory and carryover effects for a set of consonants
that favor different patterns of C-to-V coarticulatory direction. In the DAC
model, the higher the segment is constrained, the more resistant it is to
coarticulation and the more it exerts coarticulation effects on adjacent
segments. The author tests the coarticulatory direction, that is, either
anticipatory or carryover, using the flanking vowels [i] and [a] for both
Catalan and Spanish. The data suggest that the tendency in Catalan and Spanish
is toward vocalic carryover instead of anticipation, with only the Catalan dark
(velarized) [l] exhibiting anticipatory effects. The study confirms that the
predictive powers of the DAC are quite strong, which only strengthens our
understanding of processing phonetic segments in articulatory terms.

The third article, by Maria-Josep Solé, discusses how frication may be
endangered by concurrent and coarticulatory nasality. The author looks at the
stability of fricatives when combined with nasality both within a segment and in
contact with a nasal as an adjacent segment. The hypothesis that is posited is
that the very physical and auditory principles that condition how features
interact within segments are the same ones that condition the interaction of
features in adjacent segments. The author offers many examples of diachronic
change in which the fricative is either weakened or elided, epenthesis occurs,
or the nasal is elided. The author then offers empirical data on English
fricative + nasal sequences in both fast and slow speech. The results show that
there can be anticipatory velopharyngeal opening during the realization of the
fricative gesture, which causes instability of the fricative; in addition, a
reduction in the oral pressure, coupled with a reduction of the intensity of the
high-frequency noise may result in a listener's misperception of the cluster.
For the author, these aerodynamic and acoustic consequences indeed reveal why
languages do not evidence nasal fricatives.

The fourth and final article of the first section, by Francisco Torreira, treats
pre- and post-aspirated stops in Andalusian Spanish. The proposed hypothesis was
that Andalusian stops preceded by aspirated /s/ are consistently postaspirated.
The author tested data from both a reading task and spontaneous speech. From the
reading task, in terms of VOT duration findings were statistically significant
for all three voiceless stops in Andalusian Spanish; in the data from the
Northern Peninsula speakers, no significant differences were found, except for
those of the voiceless dental stop, /t/. The author notes that in the previous
work of Gerfen (2002), stop closure duration is considered an acoustic cue of
/s/ aspiration. Therefore, the author proceeds to test this phenomenon as well.
For stop closure duration, results were significant for the Andalusian speakers.
In the case of spontaneous speech, results again support the notion that
postaspiration of Andalusian voiceless stops occurs when there is a preceding
underlying or etymological /s/, typically realized as a period of aspiration.
Finally, given variability in the durations of aspirated /s/ and in the VOT
values, the author chooses to analyze the results in Articulatory Phonology,
offering a gestural score to illustrate the precise articulatory conditions that
produce aspiration in Andalusian Spanish.

The second section of the book is entitled 'Prosodic Structure'. The first
article, by Lluïsa Astruc-Aguilera and Francis Nolan, focuses on the variation
in the intonation of extra-sentential elements (ESEs) in English and Catalan in
three experiments. The first experiment was a comparative study of Southern
British English and Central Catalan in which the speakers read a sample. The
results suggest that ESEs tend to be prosodically independent, except for
appositions and vocatives. In addition, initial ESEs behave differently in
prosodic terms from their non-initial counterparts. For the second and third
experiments, the authors considered right-dislocated phrases in Catalan. Their
goal for these two experiments was to assess whether the right-dislocated phrase
receives a true pitch accent or not. Data results from the second experiment
point to a preference for accentual cues rather than phrasing ones and results
from the third experiment suggest that dislocations tend to form separate tonal
units; they did not find support for the hypothesis that right-dislocations are
accented. The authors conclude the article by stating their view of the role of
ESEs; that is, the role of ESEs is primarily semantic, and this role is signaled
in prosodic terms via tonal and/or junctural cues.

In the second article, by Laura Colantoni and Jeffrey Steele, the authors
discuss consonant + tap cluster simplification asymmetries in Argentine and
Chilean Spanish and European and Quebec French. In this study consisting of a
reading-task experiment, the authors submitted three hypotheses. First, the
authors predicted a higher degree of simplification in stop-rhotic clusters than
in stop-lateral clusters. The findings support their hypothesis as intrusive
vowel rate of occurrence was quite high for both varieties before a rhotic and
extremely low before a lateral in Spanish and around 50% for both varieties
before a rhotic and again extremely low before a lateral in French. In their
second hypothesis, the authors predict that an intrusive vowel will be the
preferred outcome if the rhotic is similar to the stop; if the rhotic is less
similar, however, assimilation will likely occur. The results support their
hypothesis in that Argentine and Chilean Spanish exhibit a high percentage of
vowel intrusion; in French, the rhotic is a dorsal fricative and thus less
similar to the tap, hence a higher rate of voicing assimilation found is not
surprising. Their third hypothesis states that ''a compensatory lengthening
effect should be observed in the second member of the cluster, provided that it
be [+continuant]'' (114). In Spanish, the results indicate that there is no
compensatory lengthening effect in the Argentine data, but there is in the
Chilean data; in French, data from both varieties suggest a compensatory
lengthening effect. Lastly, the findings are discussed in an OT framework in
which the authors propose five constraints, three based on markedness
constraints and two based on faithfulness ones.

In the third article, by Sónia Frota, MariaPaola D'Imperio, Gorka Elordieta,
Pilar Prieto, and Marina Vigário, the authors investigate the phonetics and
phonology of intonational phrasing in the four Romance languages of Catalan,
Spanish, European Portuguese (two varieties), and Italian. In phonological
terms, they studied nuclear accents and contours. With regard to nuclear
accents, of the languages/varieties examined, two groups formed; those with only
rising accents, or at least with a strong tendency toward rising accents
(Spanish and Catalan) and those with both rising and falling accents (Italian
and both varieties of European Portuguese); the same language division is found
for nuclear contours, in that the Spanish and Catalan data showed very little
connection between L+H* and sustained pitch and L*+H and continuation rise; that
withstanding, the Italian and Portuguese data did in fact show a strong
connection between these categories. In terms of phonetics, the authors analyze
the dominant boundary cue found in all languages studied, namely the H(igh)
boundary tone (HBT). Data results suggest nuclear pitch accent choice plays a
major role on HBT scaling in all of the languages/varieties studied; the impact
of phrase length and scaling correlation between the first peak and HBT again
offers the pattern found in the phonological analysis of a clear division of
Spanish and Catalan in one group and Italian and the two varieties of European
Portuguese in the other.

The fourth and final article of the second section, by Marta Ortega-Llebaria and
Pilar Prieto, examines stress contrast in deaccented syllables in Spanish. The
authors analyzed four correlates of stress, namely syllable duration, vowel
quality, overall intensity, and spectral balance under the scope of four
conditions, stressed and unstressed syllables in both accented and unaccented
environments. By doing so, the authors sought to determine how stress contrast
is maintained in the presence or absence of a pitch accent. Data results reveal
that the presence of an accent does not necessarily trigger lengthening on the
stressed syllable, which prompts the authors to conclude that duration is a
secondary acoustic marker in determining an accented vs. unaccented syllable.
Moreover, the authors found that the presence of an accent does not affect
formant frequency values, with accented and unaccented vowels having similar
vowel qualities. Data results also suggest that syllable duration and vowel
quality are primary cues for stress and are secondary cues in accent. In
addition, the authors also contest the commonly-held view that intensity plays
no role as a cue to stress; rather, they find intensity does play a role in
determining stress, with their findings exhibiting differences in intensity
levels at higher regions of the spectrum. Another novel claim made by the
authors is that, for stress, the phonetic cues for accented contexts differ from
those for unaccented ones; for accented contexts, pitch is a strong phonetic
cue, along with duration and intensity; in unaccented contexts, duration,
intensity and vowel quality are cues, but not pitch.

The third section of the book, entitled ''Acquisition of Segmental Contrasts and
Prosody'', is comprised of four studies that discuss the acquisition of segmental
contrasts and prosody. The first article, by M. João Freitas, considers the
early acquisition of unstressed vowels in European Portuguese. The central goal
of the paper is to investigate children's early sensitivity to the phonological
processes that affect vowel inventory. In European Portuguese, the following
vowel neutralizations occur: /a/ reduces to a low central vowel; mid front
unrounded vowels reduce to a high central untrounded vowel; back central vowels
reduce to [u]. Data results show that the subjects chose syllable deletion over
using an unstressed vowel in the initial stage of acquisition. During this
initial stage, however, Freitas did find that the subjects began to employ a low
central vowel before a high central vowel. The author points out that this is
not surprising in that a process from /a/ to a low central vowel requires only a
change in Height features, which is easier than one that involves multiple
processes (i.e. mid front vowels reducing to a high central vowel). The children
later used repair vowels instead of deletion, often choosing [i] for the high
central unrounded vowel and dorsals (including the high central unrounded vowel)
for the low central vowel. In short, the study shows that Portuguese children
are indeed quite aware of phonological processes at an early age and start
acquiring the vowel reduction process quite early on in phonological
development. The study also suggests that children will employ vowel reduction
first in more simple processes, followed by more complex ones.

In the second article, by Ferran Pons and Laura Bosch, the authors discuss the
perception of lexical stress patterns by Spanish and Catalan infants. The
authors consider the influence that stress cues can have on early word
segmentation by conducting two experiments. In the first experiment, the authors
sought to explore six-month old infants' preference for iambic or trochaic lists
of words. They tested this by using a slightly modified version of Jusczyk and
et al's (1993) Head-Turn Preference Procedure (HPP); the stimuli was produced by
an adult female speaker who read two-syllable CVCV nonsense words (half
trochaic; half iambic); as with previous studies, no stress pattern preference
was evidenced for the six-month old infants. In the second experiment, infants
around nine months old were tested, and this time their data differed from
previous work in that the infants again did not exhibit a stress pattern
preference. In short, the study suggests that children acquiring Catalan and
Spanish, both at six and nine months of age, are attuned to different cues than
their stress-timed counterparts, English, Dutch, and German. The authors are
currently testing a hypothesis that considers syllable weight in lexical stress
assignment in infants to further understand how children of syllable-timed
languages acquire lexical stress patterns.

The third article, by Geoffrey Stewart Morrison, considers how logistic
regression analysis can be applied to both first- and second-language perception
data. The article serves as a tutorial for both first language (L1) and second
language (L2) studies. The author reminds us that logistic regression has been
frequently used in L1 research, but that is not the case for L2 research. The
aim of the article is to aid L2 researchers in utilizing this form of analysis.
As a whole, Morrison goes to great lengths to present a very difficult concept
in a clear fashion. Utilizing actual data from two studies, he offers different
examples to aid comprehension, beginning with one stimulus dimension, binomial
responses and followed by multiple stimulus dimensions, multinomial response
categories. In the next section, he shifts to the interpretation of logistic
regression coefficients in the form of graphical representations, offering many
types of figures for data analysis. In sum, the article teaches us how we can
use linear regression to quantify how listeners use acoustic cues, how to create
illustrations of their use of acoustic cues and finally, as a statistical model
to determine significant differences in perception of stimuli.

The fourth and final article of the section, by Laurence White and Sven L.
Mattys, explores rhythmic typology in L1 and L2 environments. The authors here
are concerned with the difference between stress-timed languages (e.g. English)
and syllable-timed languages (e.g. Spanish) and break their study down into
three parts: the first is a cross-linguistic study comprised of eight groups of
speakers, with the first four groups composed of L1 speakers and the second four
groups comprised of L2 speakers. The second study regarded British accents and
was comprised of five groups, Standard Southern British English (SSBE) and four
groups of regional British accents. A third and final experiment considered the
English-speaking subjects' perception of speech as native and non-native. For
all three experiments, seven metrics were used. In terms of rhythmic
differences, data from the first study suggested three metrics, listed in (1),
most clearly discriminated the stress-timed languages from the syllable-timed ones:

(1) a. %V – Sum of vocalic interval duration divided by the total duration of
vocalic and
consonantal intervals
b. VarcoV – Standard deviation of vocalic interval divided by the total
duration of vocalic
and consonantal intervals
c. nPVI-V – Normalized Pairwise Variability Index for vocalic intervals

In addition, data from the second experiment regarding British accents also
exhibited the above metrics of (1a) and (1b) as the most discriminant metrics of
rhythmic differences. Finally, with regard to the third experiment on perception
of English speech as native or non-native, (1b) proved the most effective
metric, though (1a) and (1c) also exhibited correlations with accent ratings.

In the 'Introduction', the editors discuss the importance of 'laboratory
phonology' to Romance linguistics noting that it ''...sets out to answer a wide
array of research questions through the use of experimental methods. In other
words, experimental methodology previously associated with phonetic studies is
applied to the realm of phonology with the goal of exploring the crucial
correspondence between empirical data and theoretical claims'' (vii). This volume
quite successfully attains this goal in that the articles test theoretical
claims for Romance phonetics and phonology by using empirical data to
corroborate or refute their hypotheses. The result is a series of articles that
offers both a rich cross-linguistic element and novel, original research that
combine to greatly advance our current knowledge of Romance phonetics and

In terms of a rich cross-linguistic element, the vast majority of the articles
not only include empirical data of a particular language or variety of Romance,
but also compare and contrast languages and varieties both within Romance and
with another language family, namely Germanic. For example, Recasens analyzes
data from both Catalan and Spanish subjects; Torreira compares data from
different varieties of Spanish; Astruc-Aguilera and Nolan consider data from
both Catalan and English; Colantoni and Steele offer data from varieties of both
French and Spanish; Frota, D'Imperio, Elordieta, Prieto, and Vigário test data
from Catalan, Italian, Spanish and two varieties of Portuguese; Ortega-Llebaria
and Prieto compare their results on Spanish to those of American English and
Dutch; Pons and Bosch study data from Catalan and Spanish; in his helpful
article on statistics, Morrison considers data from both English and Spanish;
and finally, White and Mattys analyze data from Dutch, English, French, and
Spanish. Additionally, the volume's dedication of one entire section to studies
in both L1 and L2 acquisition adds insight into how segmental contrasts and
prosody are acquired in different languages; of particular interest in this
section are the two studies (Pons and Bosch; White and Mattys) which discuss
differences in acquisition between stress-timed languages (English, German, and
Dutch) and syllable-timed ones (Spanish and Catalan).

In terms of the novel, original research produced in the volume, many authors
cause us to reconsider previously-maintained views and/or assumptions about the
given topic. For example, in discussing the well-known processes of /s/
aspiration before voiceless stops and postaspiration, Torreira notes that his is
the only known study to actually use instrumental data that demonstrates the
phenomenon. Moreover, Ortega-Llebaria and Prieto contest the commonly-held view
that intensity plays no role as a cue to stress; rather, they find intensity
does play a role in determining stress, with their findings exhibiting
differences in intensity levels at higher regions of the spectrum. Another novel
claim made by Ortega-Llebaria and Prieto is that, for stress, the phonetic cues
for accented contexts differ from those for unaccented ones. More specifically,
for accented contexts, pitch is a strong phonetic cue, along with duration and
intensity; in unaccented contexts, duration, intensity and vowel quality are
cues, but not pitch. Other authors in this volume greatly expand our
understanding of current theoretical models. Nguyen, Wauquier-Gravelines,
Lancia, and Tuller convincingly argue for an exemplar-based approach over an
auto-segmental one in liaison consonant analysis in French. Recasens's study
offered evidence in support of the predictive power of the DAC model, which
sheds light on inter-gestural timing and thus, how we produce speech. Finally,
Colantoni and Steele offered new constraints in Optimality Theory to rank
different outputs for cluster simplification.

Overall, I highly recommend this volume, given the reasons mentioned above.
Seeing that the conference, in the editors' words, ''proved a great success in
bringing together scholars from all around the world, all of them involved in
researching contemporary issues in phonetics, phonology, and related areas...''
(vii), it really is not a surprise to see such quality work and intellectual
rigor. Perhaps it is due to this success, though, that I do think one minor
fault of the volume is the book's length. Though twelve articles that comprise
over 250 pages would hardly count as 'too short', a greater number of articles
could have been incorporated, especially, but not bound to, ones that discussed
issues of Romance intonation. That withstanding, this volume is well-edited,
contains innovative research in Romance linguistics, and merits a careful read.

Gerfen, Chip (2002) ''Andalusian codas'' _Probus_ 14: 247-277.

Jusczyk, Peter W., Cutler, Anne, and Nancy Redanz (1993) ''Preference for the
predominant stress pattern of English words'' _Child Development_ 64: 675-687.

Benjamin Schmeiser is an Assistant Professor in the Department of Languages,
Literatures, and Cultures at Illinois State University in Normal, Illinois. His
main interests are in Spanish phonetics and phonology. His recent work, which
has concentrated on Spanish intrusive vowel analysis and on gemination and
nasals in Pali, uses the gestural model as its framework.