The Routledge Handbook Of Spanish Phonology 1

Sonia Colina Editor download

https://ebookbell.com/product/the-routledge-handbook-of-spanish-
phonology-1-sonia-colina-editor-23269500

Explore and download more ebooks at ebookbell.com

Here are some recommended products that we believe you will be
interested in. You can click the link to download.

The Routledge Handbook Of Spanish History Andrew Dowling

https://ebookbell.com/product/the-routledge-handbook-of-spanish-
history-andrew-dowling-52605590

The Routledge Handbook Of Spanish Pragmatics Foundations And

Interfaces 1 Dale A Koike Editor

https://ebookbell.com/product/the-routledge-handbook-of-spanish-
pragmatics-foundations-and-interfaces-1-dale-a-koike-editor-23673822

The Routledge Handbook Of Spanish Morphology Antonio Fbregas

https://ebookbell.com/product/the-routledge-handbook-of-spanish-
morphology-antonio-fbregas-29392084

The Routledge Handbook Of Spanish As A Heritage Language Kim Potowski

https://ebookbell.com/product/the-routledge-handbook-of-spanish-as-a-
heritage-language-kim-potowski-33122602
The Routledge Handbook Of Spanish Translation Studies 1st Roberto A
Valden

https://ebookbell.com/product/the-routledge-handbook-of-spanish-
translation-studies-1st-roberto-a-valden-10424788

The Routledge Handbook Of Spanish Language Teaching Metodologas

Contextos Y Recursos Para La Enseanza Del Espaol L2 1st Edition Javier
Muozbasols Editor

https://ebookbell.com/product/the-routledge-handbook-of-spanish-
language-teaching-metodologas-contextos-y-recursos-para-la-enseanza-
del-espaol-l2-1st-edition-javier-muozbasols-editor-10992382

The Routledge Handbook Of Spanish In The Global City Andrew Lynch

Editor

https://ebookbell.com/product/the-routledge-handbook-of-spanish-in-
the-global-city-andrew-lynch-editor-12200158

Lexicografa Hispnica The Routledge Handbook Of Spanish Lexicography

Sergi Tornerpaz Battanerirene Renau Paz Battaner Irene Renau

https://ebookbell.com/product/lexicografa-hispnica-the-routledge-
handbook-of-spanish-lexicography-sergi-tornerpaz-battanerirene-renau-
paz-battaner-irene-renau-59365154

The Routledge Handbook Of Variationist Approaches To Spanish Taylor

Francis Group

https://ebookbell.com/product/the-routledge-handbook-of-variationist-
approaches-to-spanish-taylor-francis-group-42923492
 The Routledge Handbook
of Spanish Phonology

The Routledge Handbook of Spanish Phonology brings together leading experts in Spanish
phonology to provide a state-of-the-art survey of the field.
The five sections present current research on the phonological structure of Spanish including
the most prominent segmental processes, suprasegmental features, the ways Spanish phonology
interacts with other modules of grammar, the acquisition of Spanish phonology by first and
second language learners, and an analysis of phonological variation and sound change.
This volume provides comprehensive and detailed coverage of Spanish phonology. It addresses
major burning questions and pressing issues that have arisen in the study of Spanish phonology,
and is an essential reading resource for graduate students and researchers in the field.

Sonia Colina is a professor of Hispanic linguistics at the University of Arizona, Tucson,

Arizona, USA.

Fernando Martínez-Gil is an associate professor of Hispanic linguistics at The Ohio State

University, Columbus, Ohio, USA.
Routledge Spanish Language Handbooks
Series Editors: Manel Lacorte, The University of Maryland, USA,
and Javier Muñoz-Basols, The University of Oxford, UK

Routledge Spanish Language Handbooks provide comprehensive and state-of-the-art over-

views of topics in Hispanic Linguistics, Hispanic Applied Linguistics and Spanish Language
Teaching. Editors are well-known experts in the field. Each volume contains specially-commis-
sioned chapters written by leading international scholars.
Each Handbook includes substantial pieces of research that analyse recent developments in
the discipline, both from a theoretical and an applied perspective. Their user-friendly format
allows the reader to acquire a panoramic perspective of selected topics in the fields of Spanish
language and linguistics.
Published in English or in Spanish, the Handbooks are an indispensable reference tool for
undergraduate and postgraduate students, teachers, university lecturers, professional researchers,
and university libraries worldwide.They are also valuable teaching resources to accompany text-
books, research publications, or as self-study material. Proposals for the series will be welcomed
by the Series Editors.

The Routledge Handbook of Spanish Language Teaching

Metodologías, contextos y recursos para la enseñanza del español L2
Edited by Javier Muñoz-Basols, Elisa Gironzetti and Manel Lacorte

The Routledge Handbook of Spanish Translation Studies

Edited by Roberto A.Valdeón and África Vidal

The Routledge Handbook of Spanish in the Global City

Edited by Andrew Lynch

The Routledge Handbook of Spanish Phonology

Edited by Sonia Colina and Fernando Martínez-Gil

For more information about this series please visit: www.routledge.com/Routledge-Spanish-

Language-Handbooks/book-series/RSLH
The Routledge Handbook
of Spanish Phonology

Edited by Sonia Colina and

Fernando Martínez-Gil

SERIES EDITORS

MANEL LACORTE &

JAVIER MUÑOZ-BASOLS
First published 2020
by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Routledge
52 Vanderbilt Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2020 selection and editorial matter, Sonia Colina and Fernando Martínez-Gil;
individual chapters, the contributors
The right of Sonia Colina and Fernando Martínez-Gil to be identified as the authors of
the editorial material, and of the authors for their individual chapters, has been asserted
in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or reproduced or utilised in
any form or by any electronic, mechanical, or other means, now known or hereafter
invented, including photocopying and recording, or in any information storage or
retrieval system, without permission in writing from the publishers.
Trademark notice: Product or corporate names may be trademarks or registered trademarks,
and are used only for identification and explanation without intent to infringe.
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
Names: Colina, Sonia, editor. | Martínez-Gil, Fernando, editor.
Title: The Routledge handbook of Spanish phonology/edited by Sonia Colina,
Fernando Martínez-Gil.
Description: London; New York, NY: Routledge, 2020. | Series: Routledge Spanish
language handbooks | Includes bibliographical references and index.
Identifiers: LCCN 2019035214 (print) | LCCN 2019035215 (ebook)
Subjects: LCSH: Spanish language—Phonology.
Classification: LCC PC4131. R68 2020 (print) | LCC PC4131 (ebook) |
DDC 461—dc23
LC record available at https://lccn.loc.gov/2019035214
LC ebook record available at https://lccn.loc.gov/2019035215
ISBN: 978-0-415-78569-3 (hbk)
ISBN: 978-1-315-22811-2 (ebk)
Typeset in Bembo
by Apex CoVantage LLC
Este libro va dedicado a la memoria de mis padres (Fernando Martínez-Gil)
A mis padres (Sonia Colina)
 Contents

List of figuresx
List of tables xii
Contributorsxiii
Introductionxvii
Acknowledgmentsxxvi

PART I
Segmental phonology 1

1 Phonemic contrast and neutralization 3

Francesc Torres-Tamarit

2 Spirantization and the phonology of Spanish voiced obstruents 34

Fernando Martínez-Gil

3 Consonant assimilation 84
Carolina González

4 Vowel harmony 100

Jesús Jiménez and Maria-Rosa Lloret

PART II
Suprasegmental phonology 129

5 Phonotactic constraints on syllable structure 131

Sonia Colina

6 Glides and high vowels in Spanish 145

Ellen M. Kaisse

7 Syllable merger 162

José Ignacio Hualde

vii
Contents

8 Spanish verb and non-verb stress 181

Iggy Roca

9 Current issues and challenges in Spanish intonational research 222

Pilar Prieto and Paolo Roseano

PART III
Interfaces235

10 Spanish rhotics and the phonetics-phonology interface 237

Travis G. Bradley

11 Laboratory Phonology methods in Spanish phonology 259

Rebeka Campos-Astorkiza

12 Phonological encoding and the phonology of Spanish: the role

of the syllable 274
Miquel Simonet

13 Allomorphic variation 288

María Ohannesian

14 Plural formation and gender allomorphy 307

Caroline R.Wiltshire

15 Spanish phonology at the syntax interface 325

Eulàlia Bonet

16 Segmental, syllabic, and phrasal stratum domains of Dominican

Surprise-[s] hypercorrection 341
Rafael A. Núñez Cedeño, Lucía Badiola, and Ariane Sande-Piñeiro

PART IV
Spanish phonology and acquisition 375

17 Spanish phonology and phonological development: first steps in the

acquisition of focus 377
Conxita Lleó

18 Guiding principles for advancing experimental research on the second

language acquisition of Spanish phonology and phonetics 398
Holly J. Nibert

viii
 Contents

19 Language contact in Patagonia: durational control in the acquisition

of Spanish and Afrikaans phonology 416
Nicholas Henriksen, Lorenzo García-Amaya, Andries W. Coetzee,
and Daan Wissing

20 The phonological system of adult heritage speakers of Spanish in the

United States 439
Rajiv Rao

PART V
Variation and change in Spanish phonology 453

21 Spanish phonological variation 455

John M. Lipski

22 Sociolinguistic variation 470

Manuel Díaz-Campos and Valentyna Filimonova

23 A survey of Spanish diachronic phonology 484

Gary K. Baker and D. Eric Holt

24 The historical emergence of Spanish palatal consonants 502

André Zampaulo

Index521

ix
 Figures

1.1 Contrast neutralization and allophonic alternation 4

1.2 Feature geometry for consonants 11
1.3 Feature geometry for vowels 13
5.1 N-bar model of syllable structure 131
5.2 Moraic model of syllable structure 131
5.3 Sonority scale for some varieties of Chilean Spanish 136
8.1 First Hasse diagram 196
8.2 Second Hasse diagram 200
9.1 Spectrogram and intonational pitch contour of the interrogative sentence
¿Venden mandarinas? ‘Do you-pl. sell tangerines?’ (left panel) and the
exclamative sentence ¡Venden mandarinas! ‘They sell tangerines!’ (right panel) 223
9.2 Spectrogram and intonational pitch contour of the sentences La madre de
Ana y María ‘The mother of Anna and María’ (left panel) and La madre de
Ana, y María ‘Anna’s mother, and María’ (right panel) 223
9.3 Spectrogram, pitch contour, and phonological and phonetic intonational
transcription of the intonation of two yes-no questions in Cantabrian Spanish 227
9.4 Spectrogram and intonational pitch contour of the three main nuclear
configurations of yes-no questions in Spanish exemplified by the utterance
¿Marina? ‘Marina?’ 228
10.1 Gradient continuum of weakening in NCPS rhotics 239
10.2 Intergestural coupling graphs for NCPS rhotics in four syllable contexts 244
10.3 TaDA (Task Dynamics Application) model 245
10.4 Morphological and prosodic word structure of a “phrasal minimal pair” 250
10.5 Gestural scores for /abɾo/ (15b) in careful (left) versus casual (right) speech 251
10.6 Gestural scores for /aɾd̪ e~ard̪ e/ (16b,c) in careful (left) versus casual
(right) speech 252
16.1 Example of resyllabification in PDS: más efectividad360
16.2 Example of non-resyllabification in PDS: clases altas360
16.3 Example of resyllabification in PDS: es una emergencia361
16.4 Example of non-resyllabification of Surprise-[s] in PDS: uste[h] anteriormente362
16.5 Example of non-resyllabification in PDS: muys importante363
17.1 Evolution of the prosodic structures occurring in Spanish child language
for the word mariposa: in (1a) the child has only one foot available in his
prosodic inventory; in (1b) the foot is preceded by an unfooted syllable, and
in (1c) both feet of the word are already available 382

x
 Figures

17.2 Waveform, spectrogram and F0 trace for the German wh-question Was ist
denn los? ‘What’s going on?,’ produced by the bilingual child Manuel (3;0) 385
17.3 Waveform, spectrogram and F0 trace for the Spanish wh-question ¿Dónde
está la morsa? ‘Where is the walrus?,’ uttered by a Spanish adult native speaker 386
19.1 Waveform, spectrogram, and illustration of segmentation criteria for the
sentence La casa de la profesora no parece pequeña ‘The professor’s house does
not seem small,’ produced by a Spanish monolingual speaker (female, age 53) 419
19.2 Control-group comparisons for (a) rPVI-C and (b) nPVI-V 426
19.3 Mean rPVI-C data based on speaker group in (a) Afrikaans and (b) Spanish 427
19.4 Mean nPVI-V data based on speaker group in (a) Afrikaans and (b) Spanish 428
19.5 (a) Mean vowel duration (in milliseconds) for control and bilingual
Afrikaans speakers for short /ɑ/ and long /ɑː/. (b): Ratios (long /ɑː/ to
short /ɑ/) for individual speakers 429
19.6 (a) Mean vowel duration (in milliseconds) for control and bilingual
Afrikaans speakers for unstressed and stressed vowels. (b) Ratios (stressed-to-
unstressed vowels) for individual speakers 429
19.7 (a) Mean vowel duration (in milliseconds) for control and bilingual
Afrikaans speakers in sentence-initial vs. sentence-final syllables. (b) Ratios
(final-to-initial syllable vowels) for individual speakers 430
22.1 Retention of the syllable-final /ɾ/ in Caracas Spanish by age and
socioeconomic level 479
22.2 Retention of intervocalic /ɾ/ in alternation between para and pa in Caracas
Spanish by age and socioeconomic level 480

xi
 Tables

10.1 Acoustic characteristics of rhotic components 240

10.2 Lingual gestures for Spanish voiced coronal consonants 241
16.1 Contexts in which /s/ appears in coda position in Spanish 346
16.2 Consonantal sequences divided by consonant group 347
16.3 Frequencies of /s/ by consonant group according to the Diccionario de uso
del español actual347
16.4 Chi-square test by consonant group in the Diccionario de uso del español actual348
16.5 Frequency of each sequence divided by consonant group in the Léxico
disponible de la República Dominicana348
16.6 Chi-square test by consonant group in the Léxico disponible de la República
Dominicana349
16.7 Examples of the distribution of Surprise-[s] extracted from the interviews 352
16.8 Processes in the postlexical stratum 367
17.1 Prosodic hierarchy 381
17.2 Truncation in Spanish child language at the stage at which each
phonological word comprises one single trochaic foot 383
17.3 Ages of the monolingual and bilingual children at the time of data recording 388
17.4 Utterances produced by three monolingual German children and analyzed
auditorily and acoustically with PRAAT 390
17.5 Utterances produced by three monolingual Spanish children and analyzed
auditorily and acoustically with PRAAT 390
17.6 Utterances produced by three bilingual German-Spanish children and
analyzed auditorily and acoustically with PRAAT 391
19.1 Demographic information for the eight Afrikaans-Spanish bilinguals 425
22.1 Type and token frequency of labial and velar consonants in syllable coda 475
22.2 Characterization of phonological phenomena subject to lexical vs.
postlexical restrictions 476
22.3 Contextual factors when the semantically rich morphemes {s} and {n}
present redundant information 478

xii
 Contributors

Lucía Badiola is a PhD candidate at the University of Illinois at Chicago. Her research focuses
on bilingualism, code-switching, and morphosyntax.

Gary K. Baker, University of Georgia, is Senior Academic Professional in the Department of

Romance Languages and the Linguistics Department. His scholarly interests include the pho-
netics-phonology interface, Spanish dialectology, and the nature of sound change, both syn-
chronic and diachronic.

Eulàlia Bonet is Associate Professor and member of the Centre de Lingüística Teòrica at the
Universitat Autònoma de Barcelona. She obtained her PhD in 1991 from MIT. Her work
focuses on the interactions between phonology, morphology, and syntax.

Travis G. Bradley is Professor of Spanish Linguistics in the Department of Spanish and Portu-
guese at the University of California, Davis. He has published in diverse international journals,
such as Linguistics: An Interdisciplinary Journal of the Language Sciences, Probus, and Estudios de foné-
tica experimental. He has contributed to various edited volumes, including Fonética y fonología
descriptivas de la lengua española, Optimality-Theoretic Studies in Spanish Phonology, and Laboratory
Approaches to Spanish Phonology. Along with Rafael A. Núñez Cedeño and Sonia Colina, he
coedited Fonología generativa contemporánea de la lengua española (2nd edition).

Rebeka Campos-Astorkiza is Associate Professor at the Ohio State University. She works on
phonetics and phonology from a theoretical and experimental perspective. Her work illustrates
how phonetic data can help us develop theoretical models to explain sound patterns. In addition,
she also has projects related to second language acquisition and sociophonetics.

Andries W. Coetzee is Professor of Linguistics at the University of Michigan, and also holds
a position as extraordinary professor at the North-West University, South Africa. His research
focuses on phonetic and phonological variation, with a particular focus on languages of South-
ern Africa.

Sonia Colina is Professor of Spanish in the Department of Spanish and Portuguese at the
University of Arizona. She has published numerous chapters and refereed articles on Spanish
phonology, specializing in syllable structure. Professor Colina is the author of Spanish Phonology
(2009) and coeditor of Fonología generativa contemporánea de la lengua española (2014), Optimality-
Theoretic Studies in Spanish Phonology (2006), and Romance Linguistics 2009: Selected Proceedings of
the 39th LSRL.

xiii
Contributors

Manuel Díaz-Campos is Professor of Hispanic Linguistics at Indiana University. He conducts

research in the areas of phonological and morphosyntactic variation as well as acquisition of
phonology by native and nonnative speakers of Spanish. His publications include The Handbook
of Hispanic Sociolinguistics (2011), Introducción a la sociolingüística hispánica (2014), and Introducción y
aplicaciones contextualizadas a la lingüística hispánica (2017).

Valentyna Filimonova is a doctoral candidate and an associate instructor in general and His-
panic linguistics at Indiana University. Her dissertation is focused on the variation in production
and perception of social deixis and politeness in Mexican Spanish.

Lorenzo García-Amaya is Assistant Professor at the University of Michigan. His research focuses
on the psycholinguistic processes underlying language acquisition, especially in second-language
language immersion contexts such as study abroad. He is particularly interested in the relation-
ship among cognition, oral fluency, and complexity development.

Carolina González (PhD, University of Southern California) is Associate Professor in the

Department of Modern Languages and Linguistics at Florida State University. She is interested
in phonology; its interface with phonetics, morphology and syntax; and its acquisition.

Nicholas Henriksen is Associate Professor at the University of Michigan, holding appointments

in Romance languages and literatures and in linguistics. His research explores themes pertaining
to the phonetics-phonology interface of the Spanish sound system, especially in Manchegan and
Andalusian Spanish, spoken in the central and southern regions of Spain.

D. Eric Holt is Associate Professor of Spanish and Linguistics in the Department of Languages,
Literatures, and Cultures and the Linguistics Program at the University of South Carolina.
His scholarly interests include Hispanic linguistics, historical phonology, dialectology, and
phonological theory, as well as second language phonology, especially of connected speech
phenomena.

José Ignacio Hualde is Professor in the Department of Spanish and Portuguese and the Depart-
ment of Linguistics at the University of Illinois at Urbana-Champaign. He specializes in syn-
chronic and diachronic phonology with a focus on Spanish, other Romance languages, and
Basque.

Jesús Jiménez holds a PhD in philology from the Universitat de València (1997). He’s a full
Professor in the Department of Catalan Philology at the Universitat de València. His current
research is on phonetics, phonology, and linguistic variation in Catalan and other Romance
languages.

Ellen M. Kaisse is Professor Emerita of Linguistics at the University of Washington, where she
has taught phonology since 1976. She has coedited the journal Phonology since 1988. Her spe-
cializations include the interaction of phonology with morphology and syntax and the phonol-
ogy of Argentinian Spanish, Modern Greek, and Turkish.

John M. Lipski is Professor of Spanish and Linguistics at the Pennsylvania State University.
His research interests include Spanish phonology, Spanish and Portuguese language variation,

xiv
 Contributors

bilingual code-switching, creole languages, and the contribution of the African diaspora to
Spanish and Portuguese. Among his recent publications are Afro-Bolivian Spanish (2008), Varieties
of Spanish in the United States (2008), and El habla de los Congos de Panamá (2011).

Conxita Lleó holds two PhD degrees, one from the University of Washington (Seattle) on gen-
eral linguistics and one from the University of Barcelona on Romance languages. She has taught
at several universities, and from 1985 until retirement she was Professor for Spanish and Catalan
Linguistics at the Romance Languages Department of the University of Hamburg (Germany).
Her research, supported by the German Science Foundation (DFG), focuses on bilingualism,
child language, phonological acquisition, and sound change.

Maria-Rosa Lloret (PhD in linguistics, Indiana University, 1988) is a Full Professor in the
Department of Catalan Philology and General Linguistics at the Universitat de Barcelona. Her
current research is on phonology, morphology, and linguistic variation in Catalan and other
Romance languages.

Fernando Martínez-Gil (PhD, University of Southern California) is Associate Professor of

Hispanic Linguistics at the Ohio State University. He has coedited several volumes on His-
pano-Romance phonology, including Issues in the Phonology and Morphology of the Major Iberian
Languages (with Alfonso Morales-Front, 1997), Advances in Hispanic Linguistics: Papers from the 2nd
Hispanic Linguistics Symposium (with Javier Gutiérrez-Rexach, 1999), and Optimality-Theoretical
Studies in Spanish Phonology (with Sonia Colina, 2006). His main research interests are Spanish
and Galician phonology, the phonology-morphology interface, and the historical evolution of
the phonology and morphology of the Hispano-Romance languages.

Holly J. Nibert is Associate Professor of Hispanic Linguistics and Director of the Spanish and
Portuguese language programs at the Ohio State University, Columbus. Her research interests
include Spanish phonology and phonetics, the acquisition of Spanish L2 pronunciation, and
the principles and practices of L2 classroom instruction, instructor training, and language pro-
gram direction. She is the coauthor of two Spanish language textbooks (¡Arriba! and Día a día)
and has published in the Journal of Phonetics, Probus, and the Journal of American and Comparative
Cultures.

Rafael A. Núñez Cedeño is Professor Emeritus at the University of Illinois at Chicago, special-
izing in phonology, morphology, and dialectology. He has published extensively in linguistic
journals, anthologies, encyclopedias, and newspapers and has further (co)authored eight books
on Spanish and general Romance linguistics. He is a coeditor of Probus: International Journal of
Romance Linguistics.

María Ohannesian is Associate Professor in the Department of Spanish Language and Litera-
ture of the Autonomous University of Barcelona. Her research area centers on phonological and
morphological processes, especially in relation to Romance languages.

Pilar Prieto is an ICREA (Catalan Institute for Research and Advanced Studies) Research Pro-
fessor at the Department of Translation and Language Sciences at the Universitat Pompeu Fabra
in Barcelona. Her main research interests revolve around prosody, facial and manual gestures, and
language acquisition.

xv
Contributors

Rajiv Rao is Associate Professor in the Department of Spanish and Portuguese at the University
of Wisconsin–Madison. He uses a laboratory phonology approach to examine heritage Spanish,
Afro-Hispanic varieties, and prosody’s interface with pragmatics and syntax.

Iggy Roca is Emeritus Professor of Linguistics at the University of Essex in England, from which
he holds MA and PhD degrees, also in linguistics, and where he has been engaged in teach-
ing from his 1969 arrival until his recent early retirement. His additional commitment to both
Spanish and phonology research has led to a number of publications in the areas of gender and
of stress.

Paolo Roseano is a linguist working at the University of Barcelona (Phonetics Laboratory and
Department of Modern Languages and Literatures and of English Studies) and at the Univer-
sity of South Africa (Department of Linguistics and Modern Languages). His research interests
include the prosody of Romance languages, language contact, and morphology.

Ariane Sande-Piñeiro teaches Spanish and Culture at the University of Limerick (Ireland). She
received her PhD in Hispanic Linguistics from the University of Illinois at Chicago. She special-
izes on bilingualism, code-switching, and syntax.

Miquel Simonet is Associate Professor of Spanish Linguistics at the University of Arizona. He is

a laboratory phonologist who specializes in the Ibero-Romance languages. One of his principal
areas of expertise is bilingualism; in particular, the phonetics, phonology, and psycholinguistics
of bilingualism and second language learning. The goal of such research is to understand the
nature of phonological representations in the bilingual mind and the patterns that character-
ize the phonetic behavior (pronunciation, perception, lexical access) of bilinguals. A secondary
interest of Miquel Simonet is diachronic phonology, in particular, the phonetic bases of regular
sound change.

Francesc Torres-Tamarit is a CNRS/Paris 8 researcher at the laboratory Structures Formelles

du Langage since 2015. His research focuses on metrical and prosodic phonology, phonological
opacity, parallel and serial Optimality Theory, and the phonology of Romance languages.

Caroline R. Wiltshire is Associate Professor of Linguistics at the University of Florida. Her

focus of research is theoretical phonology, especially syllable phonology, and its interactions with
morphology, phonetics, and second language acquisition, in languages including Spanish, Tamil,
Arabic, and World Englishes.

Daan Wissing is Professor Emeritus and research associate at the North-West University, South
Africa. His research interests include the phonetics and phonology of Afrikaans, South-African
English, and other languages of South Africa.

André Zampaulo is Associate Professor of Spanish and Portuguese Linguistics at California

State University, Fullerton. His research focuses on Romance linguistics, particularly Ibero-
Romance phonetics, phonology, dialectology, and historical linguistics. He is the author of
Palatal Sound Change in the Romance Languages: Diachronic and Synchronic Perspectives (Oxford
University Press) and Os sons do português: Uma introdução prática à fonética e à pronúncia da língua
portuguesa (Routledge).

xvi
 Introduction

In this introduction, we describe the content, structure, and organization of this volume; provide
a summary of each of the sections and their corresponding chapters; and indicate the contribu-
tion of each author to the advancement of current research in Spanish phonology; we conclude
with a brief section of acknowledgments of anyone who has helped us in bringing the volume
to fruition.
Phonology, as one of the traditional subdisciplines of modern linguistics, has been at the
forefront of linguistic research since the beginning of the 20th century.The phonology of Span-
ish, a major world language, has likewise attracted the attention of numerous linguists, first
guided by descriptive purposes and later by the generative linguistics goal of explaining native
speaker competence. Building on almost two centuries of phonological research, this volume
aims to cover the most relevant topics in Spanish phonology today. Recent decades have seen an
increase in work in various areas of experimental linguistics related to phonology, such as pho-
netics, laboratory phonology, sociophonetics, and language acquisition. With increased research,
the need arises for solid, accessible presentations of up-to-date research and state-of-the-art
reviews of pertinent phonological topics. Guided by that need as well as the intrinsic value of
presenting our current knowledge of Spanish phonology, this volume offers an overview of the
current understanding of crucial topics for researchers and students involved in phonological
and phonologically related research who want to obtain a panoramic view of the issues. While
the chapters are not tied to any specific theoretical framework, they are written with the rigor
and theoretical sophistication expected of major leaders in the field.
It is an unquestionable fact that the study of Spanish phonological patterns has experienced
considerable advances in the past four decades or so, impelled initially by the emergence of new
theoretical models that expanded on the initial generative research emanating from seminal
studies such as those of Foley (1965), Harris (1969), and Cressey (1978), as well as numerous
articles published by these and other authors during the 1970s. The emergence of new models
of phonological analysis during the 1980s, including the autosegmental, syllabic, lexical, metri-
cal, and intonational frameworks, as well as the irruption since the mid-1990s of Optimality
Theory, undoubtedly the most influential phonological theory of the past decades, has not only
produced a wealth of research that provides valuable new insights into the Spanish phonological
system, thus significantly advancing our understanding of Spanish phonology, but also fostered
a sizable expansion of the empirical coverage to sets of data that previously either had not been
adequately explained or simply had not been taken into account.
Likewise, novel phonological theories that arose in the past few decades have been success-
fully applied to problems in a number of areas within the study of the Spanish sound structure,
many of which had not been satisfactorily accounted for nor comprehensively explored in

xvii
Introduction

earlier periods, encompassing topics in the synchronic phonological analysis of segmental phe-
nomena, the interaction of phonology with other areas of the grammar (phonetics, morphol-
ogy, syntax), the study of phonological variation and historical change, and the development of
phonological competence in both first and second language acquisition.
The idea for this Handbook grew from an effort to address what we consider to be the burn-
ing questions and pressing issues that have arisen in the study of Spanish phonology during
the past three decades. Spanish is spoken by close to 500 million people as a native language,
and about 100 million as a second language. Indeed, Spanish is currently the language with the
second-largest number of native speakers in the world, only after Mandarin. It is widely rec-
ognized as a vehicle for artistic creativity, cultural diversity, and scientific research. And yet no
single comprehensive up-to-date description of the Spanish phonological system from a theo-
retical (nondescriptive) perspective is currently available, even though such an account would
seem indispensable for research and graduate coursework on the language’s sound patterns in
institutions of higher education, and also for the task of achieving a rigorous and complete
characterization of its dialectal and stylistic diversity. An updated and in-depth description of
Spanish phonology that encompasses allophonic and allomorphic variation in both standard
and dialectal varieties, analyzed according to the latest theoretical frameworks, has yet to be
produced, in spite of the considerable amount of research brought about by recent develop-
ments in phonological theory, as well as the availability of new technologies for the study of
the speech sounds. This Handbook aims at rectifying this particular shortcoming by presenting a
comprehensive and detailed coverage of Spanish phonology that cannot be attained in volumes
addressing all areas of linguistics.
The objective of the Handbook of Spanish Phonology is to provide a comprehensive survey
of current research on a wide range of topics on the phonological structure of Spanish: (a) its
most prominent segmental processes; (b) its suprasegmental features, including syllable structure
and syllabification, word accentual prominence, and intonation; (c) the ways Spanish phonol-
ogy interacts with other modules of the grammar, namely, phonetics, morphology, and syntax;
(d) the acquisition of Spanish phonology by first and second language learners; and (e) the anal-
ysis of phonological variation and sound change. The Handbook offers a state-of-the-art survey
in the field of Spanish phonology, while at the same time making an essential contribution to
two related but distinct fields, namely, Spanish linguistics and phonological theory.
The Handbook will be useful for a variety of audiences, ranging from researchers in Spanish
phonology and Hispanic linguists to those engaged in research in phonology and linguistics in
general. It is also suitable as core reading for advanced graduate-level phonology courses and
seminars in Hispanic linguistics programs and as a reference book for researchers in linguistics
and related fields with interest and/or specialization in phonology, phonological theory, and
linguistics. Furthermore, given the current multidisciplinary interest in Spanish, The Handbook of
Spanish Phonology is appropriate for reading lists in courses in the areas of cognitive science, first
and second language acquisition, education, bilingualism, and speech and hearing science. After
this introductory chapter, the book is structured around five major areas of research in Spanish
phonology: segmental phonology, suprasegmental phonology, and closely related subareas, such
as the interface (phonetics, morphology, and syntax), variation (dialectal and historical), and
phonological acquisition of Spanish as a first language, as a second language, and as a heritage
language. Each chapter deals with case studies of particular topics in Spanish phonology, and
contains an introductory summary on how the theoretical approaches to the topic(s) under
study have evolved during the past three decades and up to the present day. References at the
end of each chapter guide the reader interested in further reading and research.

xviii
 Introduction

Part I: “Segmental Phonology”

In Chapter 1, Francesc Torres-Tamarit offers a comprehensive overview of the phonemic con-
trasts operative in Spanish. The theoretical framework employed is Optimality Theoretic, and
the data are from north-central Peninsular Spanish. The empirical focus is on the following
phonological processes: (i) nonassimilatory and assimilatory neutralization of place contrasts
in coda nasal consonants, (ii) coda depalatalization, (iii) the distribution and neutralization of
rhotic consonants, and (iv) the range of neutralization processes that target stop consonants in
coda position. The analyses—the most informing formal analyses available in the literature—
are framed within the two main approaches to phonemic contrast and neutralization devel-
oped within Optimality Theory: structure-based approaches based on licensing and positional
faithfulness, which concentrate on determining in which environments phonemic contrasts are
preserved or neutralized, and cue-based approaches such as Dispersion Theory, which focuses
on the role of contrast in shaping sound inventories. A case of phonological opacity involving
neutralization processes that target stop consonants in coda position is also analyzed in Optimal-
ity Theory with Candidate Chains, a derivational version of Optimality Theory. In Chapter 2,
Fernando Martínez-Gil addresses the topic of voiced obstruents, the stop ~ spirant allophonic
distribution, and the ways these segments interact with other phonological processes, in both
standard and dialectal varieties of Spanish. The adequate formal description of the surface reali-
zation of voiced obstruents as spirants in certain contexts, a process widely known as spirantiza-
tion, has been intensively debated from almost every conceivable theoretical perspective during
the past few decades.Two fundamental issues have been at the heart of the debate: (a) the nature
of the underlying segments and (b) the formal mechanism(s) that account for the surface dis-
tribution. This chapter reviews previous approaches to spirantization and proposes a constraint-
based analysis of the stop ~ spirant surface distribution of Spanish voiced obstruents within
the framework of Optimality Theory in an attempt to provide a principled solution to the
problems that plagued former approaches to the topic, and at the same time extends its empiri-
cal coverage to several sparsely studied facts of dialectal variation. In this approach, the stop ~
spirant distribution follows from the interaction of faithfulness and markedness constraints, as is
standard in Optimality Theory, which accounts for the core of the stop ~ spirant distribution in
a straightforward manner, independently of whether underlying stops or spirants are assumed. In
Chapter 3, Carolina González presents a comprehensive overview of consonant assimilation in
Spanish. Four main types are discussed and exemplified with data from diverse Spanish dialects:
specifically, place assimilation, voice assimilation, manner assimilation, and gemination (total
assimilation). She discusses several empirical issues related to the study of consonant assimilation
in Spanish, the most important being the extent to which some of the documented assimila-
tory phenomena in this language are categorical vs. gradient, or optional vs. obligatory. This
impinges on how best to analyze variable assimilation theoretically. It is suggested that future
studies should investigate the possible connection between apparently unrelated assimilatory
phenomena, such as interdentalization and gemination, in order to increase our understanding
of consonant assimilation in Spanish. After a brief overview of recent theoretical frameworks
used in the phonological analysis of assimilation, the chapter presents a case study of palataliza-
tion in Chilean Spanish that connects to some of the empirical and theoretical issues raised
previously. In Chapter 4, Jesús Jiménez and Maria-Rosa Lloret address the topic of vowel har-
mony, the phonological phenomenon that requires vowels in certain morphological or prosodic
domains to agree in specific phonological features. The authors present a thorough revision of
vowel harmony in southeastern Peninsular Spanish, where the loss of some final consonants is

xix
Introduction

compensated by opening the preceding vowel, and the lax character of this vowel extends to
the preceding syllables. Although the opening of the rightmost vowel is usually transmitted to
the left irrespective of the morphological filiation of the deleted consonant (as in Granada and
Murcia), vowel harmony may be limited to lax vowels related to the loss of consonants belong-
ing to certain inflectional suffixes (as in Jaén). The behavior of high vowels is another source of
variation: they can either fully participate in the process (as in Jaén) or act as neutral vowels (as
in Granada and Murcia). Regarding the harmonic domain, vowel harmony obligatorily targets
the stressed syllable—the most prominent in the word—but it can also affect the stressed and
the posttonic syllables—the main foot—or all the syllables in the word, sometimes permitting
gapped configurations (as in Granada). After formalizing the different harmonic patterns attested
in southern Peninsular Spanish within a prominence-based licensing approach in Optimality
Theory, the vowel harmony typology is extended to other Iberian Romance varieties displaying
vowel harmony (Cantabrian Spanish, Asturian, and Valencian Catalan).

Part II: “Suprasegmental Phonology”

In Chapter 5, Sonia Colina discusses phonotactic constraints—restrictions over sequences of
segments—on Spanish syllable structure, organized around two major types: those driven by
sonority, a factor that plays an important role in determining the sequencing of segments within
syllabic constituents; and those that refer to specific segments and their featural makeup, often
reflecting the affinity or lack of it between a specific feature and a syllabic position. The phono-
tactics of segments and features are discussed in connection with the phonotactics of the onset,
nucleus, and coda, simple and complex, noting that some phonological processes in Spanish
(e.g., nasal assimilation, obstruent coda neutralization) are ultimately a consequence of seg-
mental and featural phonotactics. In addition, this chapter summarizes recent work pertinent
to dialectal variation in phonotactics, in connection with complex onsets in some varieties of
Chilean Spanish and with prevocalic postconsonantal glides in Sonoran Mexican Spanish. Argu-
ments and preliminary experimental results are introduced regarding the proposal that glides
could be part of a complex onset in this variety (instead of the nucleus, as in most varieties of
Spanish). In Chapter 6, Ellen M. Kaisse reviews the phonetic and phonological characteristics
of glides, as well as their syllabification. Phonetic characteristics of glides that distinguish them
from vowels can include short duration, absence of a steady state, and gradual formant transi-
tions. Phonologically, glides in Spanish are allophones of underlying vowels and are realized
as obstruents in syllable onset position: [ɟ, ʝ, ɡʷ, ɣʷ]. When a high vowel is adjacent to a more
sonorous vowel, the lower vowel forms the peak of the syllable and the higher vowel is real-
ized nonsyllabically. Several arguments are summarized that indicate that a prevocalic glide is
syllabified in the nucleus when a consonant precedes it, but that the glide fills the onset and
becomes an obstruent when no consonant precedes. Kaisse also argues for the lexical specifica-
tion of exceptional hiatus, where a high vowel is realized as a separate peak even though it abuts
another vowel. In Chapter 7, José Ignacio Hualde reviews the reduction of lexically heterosyl-
labic vowel sequences in Spanish. At the lexical level, word-internal vowel sequences may be
syllabified as either tautosyllabic (diphthong) or heterosyllabic (hiatus). Postlexically, hiatus, both
within words and across word boundaries, may be avoided by merger of heterosyllabic vowels in
a single syllable. However, not all sequences undergo merger or may do so with equal readiness.
Syllable count in poetry provides useful evidence regarding which specific types of sequences
may be reduced to a single syllable. Hualde examines the factors that have been claimed to affect
the likelihood of merger and the phonetic realization of merged sequences. In Chapter 8, Iggy
Roca provides an account of both verb and nonverb primary stress in Spanish. The location of

xx
 Introduction

word stress is often not uniform, both in Spanish and in other languages. In nouns and the like,
Spanish stress invariably occurs within the confines of the “stem,” and the possible additional
“desinence” is hence ignored: cf. sˈában]a ‘bedsheet’, savˈan]a ‘savannah’, with “]” separating the
stem from the -a desinence, itself systematically stressless: cf. *saban]ˈá, with the -a desinence
illegitimately (“*”) stressed. In verbs, stress in turn is usually borne by certain given morphemes,
e.g., the theme vowel in past tenses (e.g., cant-á-ba-mos ‘we used to sing’) or the TAM (Tense
Aspect Mood) morpheme in future-oriented ones (cant-a-rˈe-mos). In impersonal forms (infini-
tive, gerund, and past participle), stress also falls on the theme vowel (e.g., cant-ˈa-r, cant-ˈa-ndo,
cant-ˈa-do), diphthongized in the 2nd and 3rd conjugations (corr-ie-ndo ‘running’, par-ie-ndo ‘giv-
ing birth’) and with the past participle vowel raising to i (corr-i-do), all of this in contrast with the
infinitive corr-e-r.This chapter concludes with a detailed account on the relation of stress and the
syllabification of vowel–high vowel and high vowel–vowel clusters into either one or two syl-
lables. In Chapter 9, Pilar Prieto and Paolo Roseano summarize the current state of the research
about Spanish intonation and aim at pointing out the most relevant issues and challenges in this
research field. The first section of the chapter contains an overview of the linguistic functions
of intonation in Spanish, as well as an explanation of some basic concepts. The second section
comprises a summary of the most important advances in Spanish intonation research in the past
few decades.The third section presents a panorama of the main research issues that are currently
being investigated and will probably gain momentum over the next few years, among which
are the adoption of a universally accepted coding system for transcribing Spanish intonation, as
well as the proper integration of intonation into other areas of linguistic inquiry such as syntax
and pragmatics.

Part III: “Interfaces”

In Chapter 10, Travis G. Bradley examines the patterning of rhotics in north-central Peninsular
Spanish within the context of contemporary research on the phonetics-phonology interface.
He reviews experimental studies that have provided detailed information about the production
and perception of these consonants, which have found that the phonetic realization of rhotics is
much more continuous and gradient than suggested by the discrete and categorical symbols of
broad transcription.Within this context, Bradley presents a novel analysis of Spanish rhotics that
formalizes a modular interface between an optimization-based phonology and an implementa-
tion-based phonetics. He argues that both operate on articulatory gestures that can be coupled,
or coordinated with each other in time, as represented in an intergestural coupling graph.When
optimal coupling graphs are phonetically implemented, lawful changes in the overlap and mag-
nitude of gestures give rise to gradient and continuous variation in the articulatory and acous-
tic output. Furthermore, combining perceptual distinctiveness with gestural coupling makes
possible a simpler analysis of the phonological distribution of Spanish rhotics, as compared
with two alternative theoretical accounts. The proposed analysis distinguishes between phonol-
ogy and phonetics as distinct but representationally linked modules. In Chapter 11, Rebeka
Campos-Astorkiza explores how laboratory phonology methods have been applied to Spanish
phonology by utilizing experimental and instrumental approaches, which allow researchers to
advance theoretical models based on empirical findings, as well as the use of systematic data
analysis, which usually goes hand in hand with an expanded view of phonology compared to
traditional or generative phonology. Laboratory studies have been crucial in the development
of phonological theories that explicitly address and model the phonetics-phonology interface.
Given the wide range of laboratory studies dealing with Spanish phonology, this chapter dis-
cusses four areas that have attracted considerable attention, including studies that are hypothesis

xxi
Introduction

based and aim to contribute to theories of phonology, and at the same time challenges previous
assumptions regarding Spanish sound patterns; it also introduces diverse methodologies focusing
on different types of data, namely, acoustic and articulatory data, including stop weakening, nasal
place assimilation and velarization, /s/ weakening, and the realization of vowel sequences. In
Chapter 12, Miquel Simonet reviews the experimental phonological encoding literature about
the Spanish syllable. During speech comprehension, listeners engage in phonological encod-
ing, a phenomenon that comprises processes such as assigning gradient acoustic information to
discrete phonemic categories (categorization), detecting possible word onsets in a continuous
signal (segmentation), and locating words in the lexicon (lexical activation, competition, and
recognition). The encoding evidence reviewed suggests that listeners whose native language
is Spanish deploy their knowledge of their language’s syllabification patterns in order to seg-
ment the speech chain into processable chunks, and to activate (and deactivate) word entries in
the mental lexicon, and that Spanish speakers’ knowledge of the phonology of their language
includes the syllable. Most aspects of Spanish speakers’ phonological knowledge, such as assimila-
tion or lenition rules, remain to be investigated from the perspective of phonological encoding.
In Chapter 13, María Ohannesian takes on a survey of allomorphic variation in Spanish. After
a brief general overview of the topic, this chapter presents an Optimality Theoretic analysis of
allomorphy in Spanish focusing on a well-known case involving morphophonology, namely, the
diphthongization of the Vulgar Latin open mid vowels in tonic position and its overapplication
in specific derived contexts such as evaluative derivation.The chapter also addresses the suitabil-
ity of Optimality Theory for dealing with phonological opacity phenomena, and the different
attempts to solve this problem, whether from parallel or from stratal perspectives. This chapter
frames allomorphic variation in Spanish within the Stratal Optimality Theory model, paying
especial attention to the difference between external (or optimizing) and internal (or nonopti-
mizing) allomorphy. While allomorph selection is subject to the Emergence of the Unmarked
effects in the external allomorphy (that is, allomorph choice improves unmarked structures),
in the case of internal allomorphy the selection is arbitrary and must be lexically specified. In
Chapter 14, Caroline R. Wiltshire reviews various accounts of plural and gender allomorphy
with the goal of trying to distinguish between what must be stored as lexical information and
what can be predicted in these morphophonological phenomena. She argues for a limited role
for epenthesis in plural formation, in order to streamline the classes required for allomorphy and
reassign some information from stored allomorphy to predictable phonology. In Chapter 15,
Eulàlia Bonet describes the segmental and suprasegmental phonological phenomena that are
found at the phrasal and sentential levels, and describes and discusses the behavior of the definite
article, which involves the interaction of phonology and syntax with morphology. She examines
the interactions between syntax and phonology and approaches to this interface, and investigates
questions such as (i) whether syntax is blind to phonological information or is sensitive to it,
an issue that is exemplified with prosodically motivated syntactic movement; and (ii) whether
phonology has access to syntactic information. Most approaches to the syntax-phonology inter-
face are indirect reference approaches (i.e., syntactic structure is not accessed directly by the
phonology but is mapped onto prosodic constituents of different sizes). In Optimality Theoretic
approaches, specific constraints relate syntactic structure and prosodic structure, capturing the
observed lack of isomorphism between the two types of structures. This point is illustrated
through the discussion of phrasing in Spanish, in which constraints on maximal and minimal
prosodic size interact with constraints that align syntactic and prosodic structure. Bonet’s contri-
bution also briefly discusses direct reference approaches. Finally, in Chapter 16, Rafael A. Núñez
Cedeño, Lucía Badiola, and Ariane Sande-Piñeiro analyze the widely studied topic of unex-
pected insertion of [s]/[h], which they label Surprise-[s], in popular Dominican Spanish. Among

xxii
 Introduction

other claims, it has been regarded as hypercorrection of the deletion of rhyme /s/; it has also
been held to be subject to constraints on syllable position; and it has been thought to obey voic-
ing restrictions across a word boundary and phrase-finally. This chapter seeks to gain a deeper
understanding of Surprise-[s] by proposing two different hypotheses regarding its appearance.
First, the chapter attempts to demonstrate that its distribution is not due to any phonological
restrictions but, rather, to the phonotactic and lexical frequencies of some consonantal sequences
in Spanish. And, second, the authors show that whereas underlying /s/ resyllabifies, Surprise-[s]
fails to do so. They propose that there exists an abstract, syntactic disjuncture expressed in terms
of a metrical grid, preventing Surprise-[s] from resyllabifying.

Part IV: “Spanish Phonology and Acquisition”

In Chapter 17, Conxita Lleó describes the acquisition of L1 Spanish segments, vowels, and
consonants, and the acquisition of L1 Spanish prosody, namely, syllables, metrical feet, prosodic
words, and phonological phrases. She also presents a novel study of the acquisition of intonation
by three monolingual German, three monolingual Spanish children, and three German-Spanish
bilingual children. Child utterances produced at two time points were analyzed within the
Autosegmental-Metrical model of intonation. Lleó finds that the bilingual children, heritage
speakers of Spanish, with German as the dominant language, acquired the intonation of each
language in a way that can be considered autonomous, but there was also influence of one
language onto the other, with transfer of a German pitch accent onto Spanish. In Chapter 18,
Holly J. Nibert reviews the advances in experimental research on second language acquisition
of Spanish phonology and phonetics. In recent decades, researchers in Spanish phonology and
phonetics have turned their attention to the acquisition of the language’s sound system by adult
second language (L2) learners, resulting in a better understanding of the L2 acquisition of Span-
ish phonology and phonetics. However, it remains difficult to articulate a broader understanding
of the facts of Spanish and their implications, due in part to a need for further studies but also
to confounding theoretical and methodological variation across studies that complicates their
connection. The main goal of Nibert’s chapter is twofold: first, to outline five interrelated guid-
ing principles regarding future experimental research goals and methodologies; and, second, to
review a selection of model studies that illustrate these principles, with an eye toward facilitat-
ing an explicit connection of findings as well as their practical application to L2 pronuncia-
tion instruction, with a main focus on vowels and, to a lesser extent, prosody. The chapter thus
is intended as a contribution to our current understanding of the L2 acquisition of Spanish
phonology and phonetics, as well as making a positive impact on the ongoing development of
the field. In Chapter 19, Nicholas Henriksen, Lorenzo García-Amaya, Andries W. Coetzee, and
Daan Wissing survey a case of language contact in Patagonia, with focus on durational control
in the acquisition of Spanish and Afrikaans phonology. Its two main goals are, first, to investigate
the effect of long-term bilingualism on patterns of temporal organization of speech in a unique
language-contact situation in Patagonia, Argentina, and, second, to examine specific aspects of
the phonetic and phonological grammars that contribute to cross-language transfer effects. The
analysis draws from a corpus of read speech obtained from three speaker groups: L1 Afrikaans/
L2 Spanish bilingual speakers who live in Patagonia, Argentina; speakers of L1 Spanish, also from
Patagonia, Argentina; and speakers of L1 Afrikaans from South Africa. It is shown that Afrikaans-
Spanish bilinguals display native-like patterns in consonant variability in both languages, but not
in vowel variability. Specifically, the bilingual speakers show native-like patterns in the vowel
variability of their L2 Spanish, but not of their L1 Afrikaans. It is also shown that differences
between the Afrikaans control and bilingual speakers stem from differences in how each speaker

xxiii
Introduction

group controls local segmental duration in their speech, namely, via phonemic vowel contrasts,
stress-induced vowel reduction, and final lengthening. The authors argue that this finding of
L2-to-L1 influence for vowels (but not consonants) derives from a combination of two forces:
the presence of phonetic and phonological processes affecting vowels in L1 Afrikaans, and the
relative complexity of phonotactic patterns in Afrikaans compared to Spanish. Altogether, this
chapter’s findings speak to the malleability of temporal organization patterns in bilingual gram-
mars, especially in situations of close long-term contact where the L2 becomes the dominant
language. In Chapter 20, Rajiv Rao examines the phonological system of adult heritage speak-
ers of Spanish in the United States. He provides an overview of the relatively young history of
research on the phonological system of heritage Spanish that has been conducted within a range
of US-based contexts (i.e., specifically on Spanish-English bilinguals). At the segmental level,
Rao focuses on vowels, mainly in terms of reduction processes, but also considering perception
versus production and vowel sequences, and on consonants, discussing a series of linguistic and
extralinguistic factors that influence voiced and voiceless stops, laterals and rhotics, and frica-
tives. At the less-studied suprasegmental level, Rao presents work on statement and question
intonation, speech rhythm, and stress. Finally, he concludes by aiming to inform and inspire
ways of advancing the field through a series of points for researchers to consider critically (e.g.,
data elicitation procedures, methods of drawing interspeaker comparisons) and, more important
for this volume, ideas for bridging the gap between acoustic analysis and phonological theory.

Part V: “Variation and Change in Spanish Phonology”

In Chapter 21, John M. Lipski examines phonological variation in the Spanish phonological
system, (vs. the more commonly studied subphonemic regional and sociophonetic variation).
The majority of instances examined involve consonants, and range from the retention/loss
of phonemes in all contexts (e.g., /θ/ and /ʎ/) to partial neutralization in specific positions
(e.g., between /l/ and /ɾ/, /ɾ/ and /r/, and /ɾ/ and /d/). Phonological variability in the vowel
system includes distinctive laxing in eastern Andalusian dialects, and partial neutralization of
mid-high vowel contrasts in Spanish dialects in contact with Arabic and Quechua. Ongoing
phonetic variation is also leading to emergent oppositions that may coalesce in some dialects
(e.g., intervocalic voiced stops, phonological glottal stops, distinctive vowel length). In Chap-
ter 22, Manuel Díaz-Campos and Valentyna Filimonova define and illustrate the concept of a
sociophonological variable as a function of diatopic, diastratic, and diaphasic aspects of variation
across Spanish-speaking communities. They are guided by the leading views on phonetic varia-
tion, sociophonological variables, and approaches to the study of sociophonological phenomena.
The term sociophonological variation is used to emphasize the implications that a phenomenon
could have for the sound system of a language, ranging from stigma to prestige and to language
change more broadly. The perspective of the quantitative variationist method is used to define
the sociolinguistic variable and apply it to the study of such Spanish-language phenomena as
lateralization, elision, and gliding of liquid consonants; hiatus resolution; elision of the syllable-
final /s/; velarization in consonant clusters; and intervocalic /ɾ/ deletion, among other variables.
While by no means an exhaustive account, the chapter attempts to present not only a variety
of relevant phenomena situated in the contemporary context but also a wide geographical and
theoretical scope of sociolinguistic study of sound variation. In Chapter 23, Gary K. Baker and
D. Eric Holt offer an overview of theoretical concepts and approaches to the study of Span-
ish diachronic phonology and explore a number of the principal changes that characterize
the phonological evolution of Spanish from Latin. A variety of contemporary approaches are
surveyed, both rule based and constraint based, including treatments that rely on nonlinear and

xxiv
 Introduction

geometric representations of features, as well as various levels of prosodic and metrical constitu-
ency, and that adopt multiple functional considerations based in both perception and produc-
tion. A major focus of the chapter is the extent to which phonological treatments increasingly
exploit our growing understanding of the subtle articulatory and acoustic/perceptual details of
human speech, as in listener-based accounts, in which perceptual biases may lead to reanalysis of
underlying forms. Also emphasized is the considerable role of structural and systemic influences:
syllable structure, for example, and the ways it interacts with the phonetic component, drives
much phonological change in Hispano-Romance and allows for a more perspicuous under-
standing of the mechanisms underlying some change. The chapter concludes with an extended
sketch of a gesture-based analysis of the origins of Spanish palatal sonorants /ʎ, ɲ/, in which
the authors claim that an increasing degree of intergestural timing and overlap was a response
to syllable-based dynamics. Finally, in Chapter 24, André Zampaulo explores the emergence
of palatal consonants in the history of the Romance languages, and suggests that it represents
a case of phonological innovation, since Latin displayed only labial, dento-alveolar, and velar
consonants. Because of their variability, complexity, and, in many respects, unique diachronic
paths, Spanish palatals constitute a challenging case study worth the attention of Romance and
general linguists alike. This chapter reviews the origins of such segments and presents a formal
account that builds on the insights of previous proposals. Specifically, it provides a constraint-
based analysis that focuses on the initiation of a change event during the interaction between a
speaker and a listener-turned-speaker in spoken communication. A sound change is formalized
as the difference in constraint ranking between the faithful realization of the speaker’s input
and the listener-turned-speaker’s reanalysis (as input) of one of the unfaithful realizations of
the original speaker’s input. The word palatal in the present analysis is used as a cover term for
a range of sounds whose passive place of articulation includes not only the palate itself but also
the postalveolar region. As such, the emergence of sonorants [ʎ, ɲ] and obstruents [ɟ, ʝ, ʒ, ʃ, tʃ] in
the history of Spanish is discussed.

xxv
 Acknowledgments

The Routledge Handbook of Spanish Phonology has been made possible through the participation
of countless individuals. We express our gratitude to all of them here, to those mentioned in the
following and to those we will inevitably forget to include, despite our best efforts.
First of all, we would like to thank all the authors for agreeing to be part of this project; for
following through with their contributions, without which this volume would not be possible;
and for also providing assistance as internal reviewers of the chapters.
Our gratitude also goes to the anonymous reviewers of the manuscript and the reviewers of
the chapters who offered their time and expertise to further improve the quality of the con-
tributions: Lourdes Aguilar, Javier Arias Navarro, Meghan Armstrong, Karen Baertsch, Jessica
Barlow, Ricardo Bermúdez-Otero, Laura Colantoni, Antonio Fábregas, Stephen Fafulas, Elena
Feliú, Sarah Finley, Juana Gil, Elizabeth Goodin-Mayeda, César Gutiérrez, Juan Manuel Hernán-
dez Campoy, Gillian Lord, Violeta Martínez-Paricio, Isabel Molina Martos, Francisco Moreno
Fernández, Erin O’Rourke, Joaquín Romero, Mario Saltarelli, and Kenneth Wireback.
In addition, we would like to convey our sincere appreciation to the Routledge editorial
team and in particular to Samantha Vale Noya for her endless patience and invaluable assistance,
and the University of Arizona for the sabbatical that allowed Sonia Colina to complete this
project.
Finally, we would like to thank our families for their unfailing support.
Sonia Colina and Fernando Martínez-Gil

xxvi
 Part I
Segmental phonology
 1
Phonemic contrast and
neutralization
Francesc Torres-Tamarit

1. Introduction: the notion of contrast and neutralization

The phonology of a language is best characterized as a set of phonologically contrastive pairs or
natural classes of sounds, that is, sounds that appear in opposition with each other. Contrastive
sounds in a particular language, traditionally referred to as phonemes, can usually occupy the same
structural position within a word and automatically trigger a change in meaning. For instance,
consider the examples in (1), in which only the first vowel in each word changes. Each of these
vowels occupies the nucleus position of the stressed syllable and changes the meaning of each
word. These data serve to illustrate one feature of Spanish, namely that it has a system of five
contrastive, or phonemic, vowels. Words like these that are distinguishable from each other by
only one contrastive sound are referred to as minimal pairs (e.g. piso/peso).

(1) Phonemic vowels in Spanish

p[ˈi]so ‘flat’ p[ˈe]so ‘weight’ p[ˈa]so ‘step’ p[ˈo]so ‘deposit’ p[ˈu]so ‘he put’
However, a language like Belarusian (Slavic), which also has the five contrastive vowels /a, e, i,
o, u/, neutralizes the contrast between the mid vowels /e, o/ when these occur in unstressed
position; both /e/ and /o/, which are contrastive in stressed position (e.g. [ˈnoɣi] ‘legs’; [ˈreki]
‘rivers’), neutralize to [a] in unstressed position (e.g. [naˈɣa] ‘leg’; [raˈka] ‘river’) (Crosswhite
2004), a phonological process known as vowel reduction. Neutralization, or contrast reduction,
occurs when a phonological contrast, like the one holding between the two mid vowels /e, o/
in Belarusian, is suspended in certain structural environments, such as unstressed position. Some
Romance languages also exhibit this phenomenon of vowel reduction, like most eastern Catalan
varieties, for instance, in which /a, e, ɛ/ in stressed position alternate with a schwa in unstressed
position, [ə], and /o, ɔ/ alternate with [u] (Mascaró 1976; Wheeler 2005). Contrast neutraliza-
tion can be schematically represented as in Figure 1.1a, in which two phonologically contrastive
sounds /α/ and /β/ neutralize, for instance, to [α] in certain restricted environments, here repre-
sented by the context Ci. Neutralization should not be confounded with allophonic alternations
(Figure 1.1b), in which a single contrastive sound maps onto more than one context-dependent
phonetic realization; allophones stand in complementary distribution. Allophonic alternations

3
Francesc Torres-Tamarit

can be seen in Spanish between the contrastive voiced stops /b, d, ɡ/ and their corresponding
approximant allophones [β, ð, ɣ] (e.g. [ˈbota] ‘boot’, but [la ˈβota] ‘the boot’) (see the chapter
“Spirantization and the Phonology of Spanish Voiced Obstruents”).

a. Contrast neutralization b. Allophonic alternation

/ / / / / /

[ ] [ ] [ ]

Context Ci Context Ci Context Cj

Figure 1.1 Contrast neutralization and allophonic alternation

In the schematic representation in Figure 1.1a, the neutralized sound in context Ci, [α], is
identical to one of the contrastive sounds, /α/. However, the neutralized segment can also be
distinct from both contrastive sounds.Yu (2011) establishes a simple typology of contrast reduc-
tion that distinguishes between (i) structure-preserving reduction, (ii) structure-building reduction, and
(iii) free variation. In structure-preserving reduction, the neutralized sound is identical to one of
the two contrastive sounds, as in the schematic representation. In structure-building reduction,
the resulting sound after contrast reduction is not identical to either of the two constrastive
sounds (like vowel reduction in Belarusian or Catalan). Finally, a single form can vary between
two or more sounds, in which case contrast reduction is subject to free variation.
Neutralization can be further divided into assimilatory versus nonassimilatory contrast
reduction. In the former, neutralization is the result of assimilating a specific set of features,
or feature values, depending on the theory of phonological representations, to a target from a
neighboring segment. In the latter, contrast is lost in specific positions without acquiring any
feature or feature value from a neighboring segment, most of the time owing to the absence of
such a neighboring segment.
In phonological theory, two major approaches to phonemic contrast and neutralization still
coexist today: structure-based approaches and cue-based approaches (see Yu 2011; Hall 2011 for more
details). On the one hand, structure-based approaches focus on identifying the prosodic or
structural positions that disfavor the preservation of phonological contrasts. This approach has
been extensively explored within Optimality Theory (henceforth OT; for a comprehensive
introduction to Optimality Theory, see McCarthy 2008) under the rubric of positional faithful-
ness (Beckman 1998; Lombardi 1999, 2001). In Optimality Theory, contrast reduction is always
the result of ranking a markedness constraint against an illicit phonological structure above a
faithfulness constraint that requires preservation of such a phonological contrast. If markedness
outranks faithfulness, contrast reduction is pervasive in the language and phonological contrast is
never visible. For example, a language with no voiced stops, like Hawaiian (Schütz 1994), ranks
the markedness constraint *VoicedObstruent, against voiced obstruents, above the faithfulness
constraint Ident-[voice], which is violated when the feature specification for voice in the out-
put is different from the voice specification in the input. Spanish has both voiceless and voiced
stops. However, it only has voiceless fricatives and affricates (there are no [z] or [d͡ ʒ]), except

4
 Phonemic contrast and neutralization

/ʝ/ (see section 2.3 for more details). In this sense, Spanish also exhibits an unmarked system in
its series of fricatives (a less stringent constraint against voiced continuant obstruents outranks
Ident-[voice]). In order to restrict contrast reduction to specific positions, a positional faithful-
ness constraint that incorporates the prosodic or structural position in which the phonological
contrast is maintained must outrank the relevant markedness constraint against the offending
segment. In languages with obstruent final devoicing like Catalan, for instance, in which voiced
obstruents devoice in word-final coda position, but in which obstruents maintain their voicing
specification in onset position, the positional faithfulness constraint Ident(Onset)-[voice] dom-
inates *VoicedObstruent, which in turn dominates the more general, context-free faithfulness
constraint Ident-[voice].This ranking is exemplified in (2) for Catalan obstruent final devoicing.
The variety of Spanish spoken by native speakers of Catalan also exhibits obstruent final devoic-
ing, as do some South American varieties of Spanish.The ranking of *VoicedObstruent above
Ident-[voice] is responsible for selecting the output candidate with a final voiceless obstruent.
When the same obstruent is syllabified in onset position owing to the presence of an inflectional
suffix, the positional faithfulness constraint Ident(Onset)-[voice] is responsible for blocking
devoicing. Notice that the underlying voiced stop spirantizes in onset position when followed
by a vowel, resulting in [ð], like in Spanish. In this chapter most neutralization effects will be
accounted for using positional faithfulness.

(2) Positional faithfulness analysis of final obstruent devoicing in Catalan

/nəbod/ ‘nephew’ Ident(Onset)-[voice] *VoicedObstruent Ident-[voice]
 a. [nəˈβot] * *
b. [nəˈβod] **!
/nəbod-ə/ ‘nice’ Ident(Onset)-[voice] *VoicedObstruent Ident-[voice]
 a. [nəˈβoðə] **
b. [nəˈβotə] *! * *

On the other hand, there also exist phonetically based OT approaches to contrast neu-
tralization. These accounts rely on the notion of perceptual cues, like licensing by cue or the
P[erceptual]-map (Steriade 1994, 1997, 2001, 2008; Hayes 1999). In cue-based approaches,
contrast neutralization is claimed to be the effect of poor perceptual cues linked to specific
environments; if in a certain position the perceptibility of a contrast is diminished, then the pho-
nological contrast is cancelled. Licensing by cue, like positional faithfulness, rests on the notion
of licensing: contrasts are licensed in some environments, but not in others. However, in cue-
based approaches to neutralization, the positions are identified not on formal structural distinc-
tions, but instead on phonetic grounds such as those supplying the auditory cues for perceiving
the contrast. Different phonetic cues relate with different environments, which can be arranged
according to a scale of perceptibility that ranges from the best environment for the perception
of the contrast to the worst one. From this phonetic knowledge, a fixed hierarchy of universally
ranked constraints is projected that prohibits the relevant phonological contrast in those envi-
ronments. For instance, the perception of a voicing contrast in obstruents is more difficult in
final position than in presonorant position. This is so because the latter context provides more
perceptibility cues than the former, and a larger number of cues correlates with higher distinc-
tiveness. To formalize this fact, the markedness constraint *Voice/__#, which militates against
voiced obstruents in word-final position, is always ranked higher than *Voice/__[+son], which
prohibits voiced obstruents before sonorants.Then, placing the faithfulness constraint Preserve-
[voice], against devoicing and parallel to Ident-[voice], in different positions with respect to the
fixed hierarchy of markedness constraints produces different patterns of voicing neutralization.

5
Francesc Torres-Tamarit

At this point it is necessary to make explicit what determines the status of contrastiveness in
OT analyses of phonological contrast and neutralization.This is constraint interaction. If a faith-
fulness constraint that requires maintenance of a feature outranks a markedness constraint against
such a feature, the language exploits that feature contrastively. Therefore, phonemic inventories
in OT are just by-products of constraint ranking. However, the learning procedure known as
lexicon optimization, which guides language learners in encoding morphemic underlying forms
in the absence of alternations and therefore of independent evidence, allows them to come up
with a set of contrastive segments (see Prince and Smolensky 1993: 225 for more details). From
now on the terms phoneme and contrastive sound or segment will be used indistinguishably.
Although contrastive segments are in principle seen as epiphenomenal in OT, there is one
specific theory of segment inventories that is implemented in OT terms, Dispersion Theory
(Flemming 1995, 2002). Dispersion Theory seeks to explain the shape and size of segment
inventories, both contrastive or underlying and surface inventories, by means of three types of
constraints: MaximizeContrast constraints, which force specific numbers of surface contrasts;
MinimalDistance constraints, which call for robust contrasts; and MinimizeEffort constraints,
which disfavor marked articulatory gestures. Later on a Dispersion Theoretic analysis of the
distribution of rhotic consonants in Spanish will be presented as developed by Bradley (2006).
The remainder of this chapter is as follows. In section 2 the inventory of contrastive and
surface segments in Spanish is presented, as well as the basic feature geometric representations
attributed to Spanish segments that will be tacitly assumed in subsequent analyses. Section 3
deals with the main set of neutralization processes: (i) nonassimilatory and assimilatory neutrali-
zation of place contrasts in coda nasal consonants, (ii) coda depalatalization, (iii) the distribution
and neutralization of rhotic consonants, and (iv) the range of neutralization processes that target
stop consonants in coda position.The analyses of these processes are couched within the formal-
ism of OT, and most of them are based on previous literature. Analyses reported from previous
literature are sometimes slightly modified in order to achieve an understanding of neutralization
in Spanish that is as comprehensive and uniform as possible. Section 4 concludes the chapter. All
the data presented in this chapter are accompanied by their source reference so the reader can
easily find more detailed information.

2. The contrastive segment inventory of Spanish

2.1. Vowels
As already noted, Spanish has five contrastive vowels (see Campos-Astorkiza 2012; Martínez-Gil
2012 for other recent accounts of Spanish phonemes). These are illustrated in (3) as classified in
Hualde (2005).

(3) The vocalic phonemes of Spanish (Hualde 2005)

Front Central Back
High i u
Mid e o
Low a
← Nonround Round
Spanish has two pairs of front and back vowels, two high vowels (/i, u/) and two mid vowels
(/e, o/), respectively, and one low central vowel (/a/). The vocalic system of Spanish is simple
in comparison with other Romance languages. First, there is no contrast between close and

6
 Phonemic contrast and neutralization

open mid vowels (/ɛ, ɔ/), as in Catalan (Mascaró 1976;Wheeler 2005), Standard Italian (Krämer
2009), or Portuguese (Mateus and d’Andrade 2000) (French, Occitan, and Rhaeto-Romance
also have rounded front vowels; see Bonet and Torres-Tamarit to appear). Navarro Tomás (1932)
describes a context-dependent allophonic distribution among close and high mid vowels for
Spanish, but more recent findings demonstrate a mismatch between acoustic and articulatory
data: Martínez Celdrán and Fernández Planas (2007) have shown that no systematic openness
can be supported from their acoustic data; although their articulatory data show a systematic
effect of openness precisely in those contexts pointed out by Navarro Tomás (1932), there is little
if any evidence for positing an allophonic distribution between close and open mid vowels given
that any such distinction is completely absent from the acoustic signal.
Second, Spanish has no phonological process of stress-dependent vowel reduction as exists
in Catalan or Standard Italian, in which the surface inventory of vowels is reduced in unstressed
position. Nadeu (2013) has investigated the effects of stress and pitch accent in the acoustic reali-
zation of Spanish vowels and has concluded that there is no phonologized use of vowel quality
to signal stress across speakers.
Third, neither nasalization nor duration is contrastive in Spanish vowels. All contrastive vowels
in Spanish are oral and short. However, according to Hualde (2005: 123), nasalization is an allo-
phonic feature in some Caribbean and Andalusian dialects when a vowel is followed by a nasal
consonant that can occasionally delete (e.g. [pãŋ] ~ [pã] ‘bread’). Given the gradual and variable
character of such context-dependent nasalization, we can conclude that there is no allophonic
alternation between oral and nasal vowels in Spanish, at least from a phonological perspective.

2.2. Glides
Spanish has two approximant glides: the palatal high glide [j] and the labiovelar high glide [w].
Both glides can form rising and falling diphthongs in combination with all nonhigh vowels, in
both stressed and unstressed position. Combinations of two distinct high vowels always result in
a rising diphthong ([ju, wi]), except in some parts of northern Spain (Navarro Tomás 1932: 65),
and the combinations *ji, *ij, *wu, and *uw are impossible. In (4) the possible combinations of
rising and falling diphthongs are exemplified. Spanish also has the triphthongs [jaj, jej, waj, wej]
in words like desprec[ˈjaj]s ‘you-pl disregard’, limp[ˈjej]s ‘you-pl clean-subj’, averig[ˈwaj]s ‘you-pl
find out’, and b[ˈwej] ‘ox’ (Navarro Tomás 1932: 68).

(4) Rising and falling diphthongs in Spanish

rising falling rising falling
j+V V+j w+V V+w
*ji *ij ˈrwiðo ruido *iw
‘noise’
ˈpjeðɾa piedra ˈpejne peine ˈpweɾta puerta ˈfewðo feudo
‘stone’ ‘comb’ ‘door’  ‘feudal
grant’
ˈraβja rabia ˈbajle baile aˈxwaɾ ajuar ˈkawsa causa
‘anger’ ‘dance’ ‘dowry’ ‘cause’
ˈlaβjo labio esˈtojko estoico ˈaɾðwo arduo ˈsowl soul
‘lip’ ‘stoic’ ‘arduous’ ‘soul’
ˈbjuða viuda *uj *wu *uw
‘widow’

7
Francesc Torres-Tamarit

Some words containing rising diphthongs accept pronunciations with a hiatus (e.g. [bi.ˈa.xe]
~ [ˈbja.xe] ‘trip’, [su.ˈa.βe] ~ [ˈswa.βe] ‘soft’, [ˈkɾu.el] ~ [ˈkɾwel] ‘cruel’) (Navarro Tomás 1932:
67). According to Harris (1983), prevocalic glides are not part of the onset and are syllabified in
the nucleus. This claim is based on the observation that Spanish rhymes (nucleus + coda) allow
up to three segments (e.g. p[eɾs]picaz ‘keen-sighted’). If prevocalic glides were syllabified in a
complex onset, one would expect the existence of syllables like *[pweɾs], for instance, which
are in fact unattested (cf. [ˈpweɾ.ta], [peɾs.pi.ˈkaθ]) (see the chapter “Phonotactic Constraints on
Syllable Structure”).
With respect to the contrastive status of glides, should glides be considered phonemic, or
should they be derived from high vowels? A first observation is that there exist (near-)minimal
pairs in which high vowels contrast with glides in both rising and falling diphthongs, as illus-
trated in (5) with data from Hualde (2004).

(5) (Near-)minimal pairs containing high vowels and glides (Hualde 2004)
du.ˈe.to ‘duet’ but ˈdwe.lo ‘duel’
pi.ˈe ‘I chirped’ but ˈpje ‘foot’
re.i.ˈɾe ‘I will laugh’ but rej.ˈne ‘I ruled’
Despite these (near-)minimal pairs, most phonological analyses derive glides from underlying
high vowels whenever the context for gliding is met, that is, if there is a non-stress-bearing high
vowel adjacent to another vowel (cf. [ma.ˈɾi.a] ‘Mary’ versus [ma.ˈɾja.no] ‘Marian’). What needs
to be explained is the blocking of the gliding process. Exceptional hiatuses in Spanish like those
in the left column in (5) are in fact conditioned by both prosodic and morphological factors
(Colina 1999; Harris and Kaisse 1999; Hualde and Prieto 2002; Hualde and Chitoran 2003;
Hualde 2004; Cabré and Prieto 2007; Cabré and Ohannesian 2009). First, some exceptional hia-
tuses containing a high vowel are due to paradigmatic pressure effects in which the high vowel
is stressed in other morphologically related forms ([pi.ˈe] but [ˈpi.o] ‘I chirp’). Second, blocking
of gliding can also be due to the presence of a morphological boundary ([bi-ˈe.njo] ‘biennium’).
Third, there is a prosodic initiality condition according to which high vowels preceding other
vowels at the left edge of stems are less prone to glide. The further it is from the left word edge,
and also the further from the stressed vowel, the higher the probability that a high vowel will
glide ([di.ˈa.lo.ɣo] ‘dialogue’ but [di.a.ˈlo.ɣo] ~ [dja.ˈlo.ɣo] ‘I dialogue’, [dja.lo.ˈɣe] ‘I dialogued’).
After stress, there are no hiatuses in Spanish ([is.ˈto.ɾja] ‘history’; cf. *[is.ˈto.ɾi.a]) (see the chapter
“Glides and High Vowels in Spanish” for more details).

2.3. Consonants
The set of phonemic consonants in north-central Peninsular Spanish is given in (6). Spanish
exhibits a set of three pairs of stops, voiced and voiceless, distributed along three places of articula-
tion: bilabial /p, b/, dental /t, d/, and velar /k, ɡ/. There are three contrastive nasals: bilabial /m/,
alveolar /n/, and palatal /ɲ/, and two rhotic consonants: a trill /r/ and a tap /ɾ/ (see section 3.3
for more details on the distribution of rhotics). Five contrastive fricatives are found (/f, θ, s, ʝ, x/).
All fricatives are voiceless except for /ʝ/, which has replaced the palatal lateral /ʎ/ in yeísta dialects
of Spanish; the palatal lateral is conserved in parts of Spain, the Andean region, Paraguay (Hualde
2005: 8), and the variety of Spanish spoken by some native speakers of Catalan, in which the
words vaya and valla create a minimal pair ([ˈbaʝa] ‘go-sg.pres.subj’ versus [ˈbaʎa] ‘fence’), which
are otherwise merged in yeísta dialects in favor of [ˈbaʝa].The only lateral found in all Spanish dia-
lects is the alveolar one, /l/.The voiceless interdental fricative /θ/ is present only in north-central

8
 Phonemic contrast and neutralization

Peninsular Spanish. In all other Spanish dialects, referred to as seseo dialects, casa [ˈkasa] ‘house’ and
caza [ˈkaθa] ‘hunting’ do not create a minimal pair (both words are pronounced [ˈkasa]). There is
yet one more phonemic system, referred to as ceceo, originally from parts of Andalusia and now
present in some Central American areas like El Salvador, Honduras, and Nicaragua, in which there
is no phonemic /s/ and words like seta ‘mushroom’ and ceta ‘zed’ are pronunced equally as [ˈθ̪eta],
a predorso-dental fricative (Hualde 2005: 156). The voiceless velar fricative /x/ likewise receives
different phonetic interpretations depending on the dialect: uvular [χ], characteristic of Castil-
ian Spanish; velar [x], characteristic of Mexico, Peru, and Argentina, and used in this chapter for
north-central Peninsular Spanish for ease of exposition; palatal fricative [ç], found in Chile before
the front vowels /i, e/; and [h] in Caribbean Spanish, Central America, Colombia, Andalusia, and
the Canary Islands (Hualde 2005: 156). Therefore, in the latter dialects /h/ can be considered a
phoneme of Spanish, although it is also an allophone of /s/ due to the process of /s/ aspiration in
coda position (see the chapter “Vowel Harmony” for more details on /s/ aspiration). Spanish has
only one voiceless postalveolar affricate, /t͡ ʃ/. In areas of Chile this affricate is realized as an alveolar
affricate [t͡ s], and in the Canary Islands it can be realized as a voiceless or voiced palatal stop [c ~ ɟ]
([muˈcaco] ~ [muˈɟaɟo] ‘young man’) (Hualde 2005: 152). In parts of Andalusia, northern Mexico
(Sonora and Chihuahua), Panama, and parts of Chile, there is no affricate whatsoever, and /t͡ ʃ/ has
been replaced by the voiceless postalveolar fricative /ʃ/ (Hualde 2005: 152).

(6) The consonantal phonemes of Spanish

bilabial labiodental dental alveolar postalveolar palatal velar
stop
p b t d k g
nasal ɲ
m n
trill r
flap ɾ
fricative θ s ʝ x
f
affricate ͡tʃ
lateral
l

The status of the voiced palatal fricative /ʝ/ as a phoneme is not without controversy (see
Hualde 2004, 2005; Rost Bagudanch 2014; and Lloret and Martínez-Paricio to appear for more
details). In Castilian Spanish, /ʝ/ can receive a set of phonetic realizations that range from the
palatal glide [j] to different degrees of constriction depending on the structural context: a frica-
tive [ʝ], a stop [ɟ], and even an affricate [ɟ͡ʝ] (e.g. yema [ˈjema] ~ [ˈʝema] ~ [ˈɟema] ~ [ˈɟ͡ʝema]
‘yolk’) (see Aguilar 1997; Martínez Celdrán and Fernández Planas 2001). Hualde (2004) points
out that it is more appropriate to see these realizations as a gradient along a continuum of degree
of constriction, and that a four-way categorization is just the result of imposing a symbol from
the International Phonetic Alphabet on each realization. There are two possible phonological
interpretations of these facts. On the one hand, if [ʝ] can be seen as the strengthening of [j] in
onset position, [ʝ] could be derived from the phonemic vowel /i/, because the glide can be
derived contextually from the vowel, as stated in the previous subsection. Under this view, all
cases of [ʝ], including those without a morpheme boundary, would derive from /i/ (e.g. calle

9
Francesc Torres-Tamarit

[ˈkaʝe] < /kaie/ ‘street’, mayo [ˈmaʝo] < /maio/ ‘May’). Such an interpretation finds evidence
in morphophonological alternations like rey [ˈrej] ‘king’, with a glide, and [ˈreʝes] ‘kings’ (or
[ˈreʒes] or [ˈreʃes] in Argentinian Spanish). If the /i/ is syllabified in coda position, it is realized
as a glide. In onset position, the /i/ undergoes fortition (see Harris and Kaisse 1999 for more
evidence in favor of a single underlying representation /i/). However, one could argue that the
near-minimal pairs in (7) are sufficient evidence to posit a distinct /ʝ/ phoneme.

(7) Near-minimal pairs between [j] and [ʝ] (Hualde 2004)

de.ˈsjeɾ.to ‘desert’ but des.ˈʝe.lo ‘thawing’
a.ˈβjeɾ.to ‘open’ but ab.ˈʝek.to ‘abject’
bo.ˈnja.to ‘yam’ but ˈkon̠ .ʝe.βa ‘it implies’
A closer look at these data reveals that the contrast between [j] and [ʝ] can be simply
derived from a difference in syllabification, which correlates transparently with morphologi-
cal structure: all the words in the right column contain a prefix. An OT alignment constraint
requiring the left edge of the stem to coincide with the left edge of a syllable would force the
stem to be syllabified independently from the prefix, and then a strengthening process would
derive the fricative realization of /i/ when followed by another vowel. However, there is a very
interesting contrast in some dialects of Spanish that seems to support a second interpretation
of the facts whereby there is a near-phonemic distinction between two contrasts, each of them
along its own continuum: the high vowel/glide contrast, on the one hand, and the glide/frica-
tive/obstruent contrast, on the other hand, most probably due to an effect of orthography or
fetichismo de la letra (Rosenblat 1963). This is most clear in Argentinian Spanish, in which the
palatal consonant under scrutiny has undergone a process of strengthening toward obstruents,
either [ʒ] or [ʃ] (e.g. yema [ˈʒema] or [ˈʃema]). However, orthographic (h)i is pronounced [j]
or [ʝ] (e.g. yerba [ˈʒeɾβa] ‘yerba mate’ but hierba [ˈjeɾβa] ~ [ˈʝeɾβa] ‘grass’). Castilian Spanish also
shows this intricate word-dependent contrast between two categories ranging along the degree
of constriction, as put forward in Hualde (2004, 2005) and illustrated in (8), in which different
words show a different cutoff point along the continuum ranging from the approximant glide
to the voiced palatal stop.

(8) Different cutoff points along the constriction scale (Hualde 2004)
hiato ‘hiatus’ i.ˈa.to ~ ˈja.to
hiena ‘hyena’ ˈje.na ~ ˈʝe.na
yema ‘yolk’ (ˈje.ma ~) ˈʝe.ma ~ ˈɟe.ma
The situation depicted in (8) is difficult to account for by any model of phonology based
exclusively on categorical representations and a mere interpretative phonetics component. This
state of affairs could just be a synchronic static picture of a distribution that is characteristic of
situations of in-progress language change. In such a scenario, two distinct classes of words start
behaving differently due to orthography and perhaps lexical frequency. One can imagine that
this unstable situation may develop into a robust phonological system with two contrastive
sounds over time, one vocalic (/i/) and the other consonantal (/ʝ/).

2.4. Featural specifications

Most of the phonological analyses of neutralization processes presented in section 3 rely on the
basic feature geometry in Figure 1.2, based on Lloret and Martínez-Paricio (to appear), which

10
 Phonemic contrast and neutralization

is adapted from Uffmann (2011) (see also Núñez Cedeño 2014 for an account of phonological
features in Spanish using feature geometrical representations). The assumed feature geometry is
enough to account for the processes of Spanish that we are interested in.

Root [±consonantal, ±sonorant]

[±continuant]
Laryngeal [±lateral]

[±voice] [±nasal]
[spread glottis]
Place

[dorsal]
[coronal]

[labial] [±anterior]

[±distributed] (depends from [+anterior])

Figure 1.2 Feature geometry for consonants

Source: Adapted from Lloret and Martínez-Paricio (to appear).

In a feature geometrical tree, features are hierarchically organized in different tiers. We

assume all phonological features to be binary except for the major place features [labial],
[coronal], and [dorsal], as is common in the phonological literature (for theories of feature
underspecification, see Archangeli 2011), and [spread glottis]. The segmental root, linked
to a skeletal or timing position that mediates between melody and syllabic structure, is
equivalent to the features [±consonantal, ±sonorant]. These two features are equated with
the root because they never assimilate or dissimilate. They distinguish two major classes of
segments: consonants, which are [+consonantal], and vowels, which are [‒consonantal]. The
feature [±sonorant] distinguishes between obstruents (stops and fricatives), which are [‒
sonorant], and sonorants (the remaining consonants and all vowels), which are [+sonorant].
Three terminal manner features are immediately dominated by the root: [±continuant],
which distinguishes fricatives, which are [+continuant], from stops, which are [‒continuant]
(the affricate /t͡ ʃ/ is both [+continuant] and [‒continuant]); [±lateral], which characterizes
laterals, which are [+lateral], versus rhotics, which are [‒lateral] (the two rhotics can be
distinguished by the feature [±trill], or [±tense], but we will argue in favor of considering
trills as [‒continuant] and flaps as [+continuant]); and [±nasal], which differentiates between
nasals and nonnasal consonants. The root node also dominates two more nodes, Laryngeal
and Place, each of which dominate their own respective terminal features. The Laryngeal
node dominates [±voice], which distinguishes voiced from voiceless segments, and [spread
glottis], which is characteristic of voiceless fricatives (see Lloret and Martínez-Paricio to
appear for a more detailed discussion of [spread glottis]). We assume that only /s/ has the

11
Francesc Torres-Tamarit

feature [spread glottis] in Spanish, which we conceive as a monovalent feature. The Place
node directly dominates three monovalent features, [labial], [coronal], and [dorsal]. Only
the feature [coronal] dominates two terminal features, [±anterior] and [±distributed]. Seg-
ments which are (inter)dental and alveolar are [+anterior], and postalveolar (or prepala-
tal) segments are [‒anterior]. Within [+anterior] segments, [+distributed] characterizes /s/,
and [‒distributed] characterizes /θ/. All velar consonants are [dorsal], and the palatal nasal
is characterized as being both [coronal] and [dorsal], as in Piñeros (2011), among others.
A table with the featural specifications for all consonants is illustrated in (9a); “x” stands for
the presence of a monovalent place feature.
According to Unified Feature Theory (Clements and Hume 1995), both consonants and
vowels share the same features (Figure 1.3) (we present the featural specification of vowels
although no process involving them will concern us). The Place node is characterized in this
theory as C-place, which dominates a Vocalic node. This node dominates in turn two more
nodes:V-place and Aperture. It is enough for a five-vowel system like Spanish that V-place domi-
nates only two features, [coronal] and [dorsal]. Aperture dominates the feature [open]. For con-
sonants, the C-place node directly dominates the three terminal place features not dominated
by the Vocalic node (see Uffmann 2011 for more details on feature geometry and the Unified
Feature Theory). The front vowels (/i, e/) are thus [coronal], and the back vowels (/u, o/) are
[dorsal].The aperture feature [open] distinguishes between high vowels and mid vowels, the lat-
ter being specified for V-place and for [open]. The low vowel (/a/) is only specified for [open]
and is neither [coronal] nor [dorsal]. This characterization of vowels resembles that of Element
Theory (Backley 2011). Featural specifications for vowels are illustrated in (9b).

(9) Summary of featural specifications for contrastive segments in Spanish

a. Consonants
Root Laryngeal Place Manner
[cons] [son] [voi] [sp gl] [lab] [cor] [dor] [nas] [lat] [cont]
[ant] [distr]
p + ‒ ‒ x ‒ ‒ ‒
b + ‒ + x ‒ ‒ ‒
t + ‒ ‒ x + ‒ ‒ ‒ ‒
d + ‒ + x + ‒ ‒ ‒ ‒
k + ‒ ‒ x ‒ ‒ ‒
ɡ + ‒ + x ‒ ‒ ‒
f + ‒ ‒ x ‒ ‒ +
θ + ‒ ‒ x + ‒ ‒ ‒ +
f + ‒ ‒ x x + + ‒ ‒ +
ʝ + ‒ + x x ‒ ‒ +
x + ‒ ‒ x ‒ ‒ +
͡tʃ + ‒ ‒ x ‒ ‒ ‒ ‒+
m + + + x + ‒ ‒
n + + + x + + + ‒ ‒
ɲ + + + x x + ‒ ‒
ɾ + + + x + + ‒ ‒ +
r + + + x + + ‒ ‒ ‒
l + + + x + ‒ + +

12
 Phonemic contrast and neutralization

b. Vowels
V-place Aperture
[cor] [dor] [open]
i x
e x x
a x
o x x
u x

C-place

[labial] Vocalic

[coronal] V-place Aperture

[dorsal] [open]

[coronal]

[dorsal]
Figure 1.3 Feature geometry for vowels
Source: Adapted from Uffmann (2011).

3. Neutralization processes in Spanish

3.1. Nasal place assimilation and neutralization

In this section we describe and formalize one of the major neutralization processes in Spanish,
nasal place assimilation (see also the chapter “Consonant Assimilation”).We have so far established
that Spanish makes use of a triple place contrast within nasals. In syllable onset position, the bila-
bial, alveolar, and palatal nasals contrast (10). Recall that phonemic contrast in onset position is
guaranteed by the undominated position of the Ident(Onset) family of faithfulness constraints.

(10) Nasal contrast in onset position

[ˈkana] ‘gray hair’
[ˈkama] ‘bed’
[ˈkaɲa] ‘cane’
However, the three-way place contrast is neutralized in coda position, both medially and
word-finally. As exemplified in (11), in medial position all nasals assimilate to the place of articu-
lation of the following consonant. Nasals acquire bilabial, labiodental, dental, alveolar, postalveo-
lar, and velar place specifications depending on the place specification of the following onset
consonant. (For other structure-building neutralization processes, like voice assimilation, see the
chapter “Consonant Assimilation”.)

13
Francesc Torres-Tamarit

(11) Nasal place assimilation (Hualde 2005: 104)

[ˈkampo] ‘field’
[ˈeɱfasis] ‘emphasis’
[ˈtanto] ‘so much’
[ˈansja] ‘anxiety’
[ˈaɲt͡ ʃo] ‘wide’
[ˈaŋka] ‘haunch’

We follow Piñeros (2006) and Colina (2009) in making use of the constraint Coda-Condi-
tion, defined in (13) to exclusively target nasal consonants. This markedness constraint incor-
porates the idea that coda consonants cannot license their own place features. However, in a
place-sharing configuration in which the consonant in coda position shares the same place
specification as a following onset, Coda-Condition is satisfied. In autosegmental terms, place
assimilation is represented as in (12). The segment [ŋ] stands for a nasal consonant that has
acquired its velar place specification from a following velar stop via an autosegmental opera-
tion of spreading-cum-delinking. Laterals also acquire the place of articulation of the following
consonant (e.g. interdental lateral as in e[l̟ θ]irco ‘the circus’; dental lateral as in e[l d]ía ‘the day’;
alveolar lateral as in e[l r]ío ‘the river’; palatal lateral as in e[ʎ ɟ]eso ‘the plaster’) except when the
following consonant is labial, because [labial] and [+lateral] are incompatible features, or velar
(e.g. e[l β]arco ‘the ship’; e[l ɣ]ato ‘the cat’). Therefore, Coda-Condition targets both nasals and
laterals; the constraints *[labial]&[+lateral], against labial nasals, and *[ʟ], against velar laterals,
outrank Coda-Condition. The faithfulness constraint Ident-[place] is defined in (14), violated
in output forms exhibiting nasal place assimilation.

(12) Nasal place assimilation

ŋ g

Place Place

[cor] [dor]
(13) Coda-Condition[nasal] (based on McCarthy 2008)
Assign one violation mark for every token of a place feature that is associated with a nasal
consonant in the syllable coda.
(14) Ident[place]
Assign one violation mark for every input-output discrepancy between place specifications.

Consider an input form like /son ɡatos/ ‘they are cats’ in (15). Candidate (15b) fatally violates
Coda-Condition[nasal] because the coda is linked to its own place specification, in this case
[coronal]. The winning candidate (15a), in which the [dorsal] place specification of the follow-
ing onset spreads, violates low-ranked Ident-[place]. In the following tableaux, a W indicates
that a winner-favoring constraint dominates a loser-favoring constraint, indicated by L.

(15) Tableau for /son ɡatos/ → [ˈsoŋ ˈɡatos]

/son ɡatos/ Coda-Cond[nas] Id-[pl]
 a. [ˈsoŋ ˈɡatos] *
b. [ˈson ˈɡatos] *W L

14
 Phonemic contrast and neutralization

Word-finally before a pause, nasals also neutralize. In most dialects of Spanish, we only find
[n] in this position. Sometimes this is even reflected in the orthography, like in the word ron
[ˈron] from English ‘rum’, or esmoquin [esˈmokin] from English ‘smoking’. Sometimes it is not,
like in referéndum ‘referendum’, which is pronounced [refeˈɾendun] anyway, or álbum ‘album’,
pronounced [ˈalβun] (cf. álbumes [ˈalβumes], but also [ˈalβunes] and less commonly [ˈalβuns];
see Lloret and Mascaró 2006; Bermúdez-Otero 2006). In other Spanish dialects, however, the
point of articulation of word-final nasals is velar, instead of coronal (e.g. [ˈroŋ], [refeˈɾenduŋ],
[ˈalβuŋ]). We follow some analysts in interpreting these nasals as debuccalized (see Trigo 1988;
Bakovic 2000). More alternations between [n] and [m] are exemplified in (16), taken from Llo-
ret and Mascaró (2006). No alternations between [n] and [ŋ] exist because velar nasals are not
contrastive in Spanish.

(16) [n] ~ [m] alternations

isla[n] ‘Islam’ islámico ‘Islamic’
íte[n] ‘item’ itemización ‘itemization’
oh[n] ‘ohm’ óhmico ‘ohmic’
tóte[n] ‘totem’ totemismo ‘totemism’
Abrahá[n] ‘Abraham’ abrahámico ‘Abrahamic’
Vietna[n] ‘Vietnam’ vietnamita ‘Vietnamese’
Amsterda[n] ‘Amsterdam’ amsterdamés ‘Amsterdamese’
Surina[n] ‘Surinam’ surinamés ‘Surinamese’
Fro[n] ‘Fromm’ frommiano ‘Frommian’
Neutralization in absolute final position can be attributed to two context-free markedness
constraints against noncoronal place features, *[dorsal] and *[labial] (17). The strategy adopted
to satisfy these constraints is not achieved by means of spreading, because there is no consonant
from which any place feature could spread. Instead, the input place feature, if it is not [coronal],
maps onto [coronal]. The fact that [coronal] is inserted and not, for instance, [dorsal] derives
from the same universally ranked constraint hierarchy that motivates centralization (neutraliza-
tion toward [coronal]) (17). This fixed hierarchy derives from the relative markedness of place
specifications (Lombardi 1999, 2001); the feature [coronal] is less marked than [labial] and
[dorsal].

(17) Universally ranked place markedness constraints (Lombardi 2001)

*[dorsal], *[labial] >> *[coronal]
For velarizing dialects, the high ranking of Coda-Condition[nasal] is enough to derive neu-
tralization. We claim that word-final nasals before a pause in velarizing dialects do not acquire a
[dorsal] feature, which would be ruled out by the universal constraint ranking in (17), but rather
are debuccalized; that is, they have no place (but see Ramsammy 2013 for an interpretation of
final velar nasals as nasals fully specified for place). Placeless nasals surface phonetically as velar
nasals because of the necessary lowering of the velum. Placeless segments violate the markedness
constraint Have-Place.
Tableaux (18) and (19) illustrate the ranking that is necessary to account for nonvelarizing
and velarizing dialects, respectively. The candidates that include [ŋ] have undergone debuccali-
zation, but not insertion of [dorsal] (usually debuccalized nasals are transcribed as [ɴ], the IPA
symbol for uvular nasals); candidates with a [dorsal] specification are omitted. In nonvelarizing

15
Francesc Torres-Tamarit

dialects (18), the ranking of *[labial] above *[coronal] and Ident-[place] discards the faith-
ful candidate (c) with a final labial nasal. Candidate (18b), with debuccalization, is discarded
because it violates the undominated constraint Have-Place, which is ranked above Coda-
Condition[nasal], *[coronal], and Ident-[place] (using Max-[place] would be more appropri-
ate since place features are monovalent, but we use Ident-[place] instead for ease of exposition).
In velarizing dialects (19), however, Have-Place is dominated, and debuccalization can apply
without violating any other constraint; recall that in the absence of a place feature, *[labial] and
*[coronal] are vacuously satisfied, as well as Ident-[place].

(18) Tableau for /album/ → [ˈalβun]

/album/ *[lab] Have-Pl Coda-Cond[nas] *[cor] Id-[pl]
 a. [ˈalβun] * * *
b. [ˈalβuŋ] *W L L L
c. [ˈalβum] *W * L L

(19) Tableau for /album/ → [ˈalβuŋ]

/album/ Coda-Cond[nas] *[lab] Id-[pl] *[cor] Have-Pl
 a. [ˈalβuŋ] *
b. [ˈalβun] *W *W *W L
c. [ˈalβum] *W *W L

The two OT grammars are represented as a Hasse diagram in (20); lines stand for domination
relations, and dashed lines indicate disjunctive rankings (see McCarthy 2008 for more details on
constraint rankings in OT and how to represent them).

(20) Hasse diagrams for nasal place assimilation and nasal place neutralization
a. Nonvelarizing dialects b. Velarizing dialects

*[lab] HAVE-PL CODA-COND[nas] *[lab]

*[cor] CODA-COND[nas] ID-[pl] *[cor]

ID-[pl] HAVE-PL

3.2. Coda depalatalization

Besides neutralization of /m/ as in álbu[n] (cf. álbumes), this process of place centralization, as
we could call it, also affects the palatal nasal /ɲ/ and the palatal lateral /ʎ/, which surface as [n]
and [l] in coda position, respectively. The [ʎ] ~ [l] alternation exists only in lleísta dialects, those
that have contrastive /ʎ/ instead of the palatal fricative /ʝ/. For yeísta dialects, the alternation
between [ʝ] and [l] could be better accounted for by means of phonologically conditioned allo-
morph selection. The data in (21) are taken from Lloret and Mascaró (2006). Transcriptions are
orthographic; ñ stands for [ɲ], and ll for [ʎ].

16
 Phonemic contrast and neutralization

(21) Nasal and lateral depalatalization

Word-final Nonfinal in Nonfinal in
inflection inflection and
derivation
a. [n] [n] [ɲ]
desdén ‘disdain’ desdenes ‘disdains’ desdeñar ‘to disdain’
desdeñoso ‘disdainful’
Don ‘sir’ doña ‘madam’
champán ‘champagne’ champanes ‘champagnes’ champañería ‘champagne
shop’
champañera ‘champagne
cooler’
b. [l] [l] [ʎ]
Él ‘he’ ella(s) ‘she/they-f’
ello(s) ‘it/they-m’
aquel ‘that’ aquella(s) ‘that/those-f’
aquello(s) ‘that/those-m’
doncel ‘squire’ donceles ‘squires’ doncella(s) ‘maid(s)’
doncellez ‘maidenhood’
clavel ‘carnation’ claveles ‘carnations’ clavellina ‘small
carnation’
piel ‘skin’ pieles ‘skins’ pellejo ‘skin (colloquial)’
c. Assimilatory Sabadellense ‘from Sabadell’
environment
beldad ‘beauty’ bello ‘beautiful-m’
Lloret and Mascaró (2006) take the behavior of borrowings as clear evidence of the produc-
tivity of the centralization process. Centralization from /m/ to [n] and depalatalization from
Catalan source forms, mostly proper nouns, are illustrated in (22). Orthographic ny in Catalan
represents the palatal nasal [ɲ].

(22) Centralization in Spanish

/m/ → [n] /ɲ/ → [n] /ʎ/ → [l]
Prim champán Sabadell
Grimm seny Maragall
réquiem Montseny Coll
eslálom Capmany Urgell
módem Fortuny Llull
referéndum Montmany Collell
médium Jubany Bofill
Boom Ferreny Tusell
tedèum Sispony Creixell
fòrum Montgrony Moll
As has been shown for nasals in the previous subsection, we claim that lateral depalatalization
is also due to a context-free markedness constraint on the [dorsal] place specifications typical of
palatal segments, which are also specified for [coronal] (see 9a). We repeat the universal ranking

17
Francesc Torres-Tamarit

of place markedness constraints in (23). Not only labial nasals are prohibited in coda position,
but also palatal nasals and laterals.

(23) Universally ranked place markedness constraints

*[dorsal], *[labial] >> *[coronal]
Coda depalatalization is accounted for by ranking *[dorsal] above Ident-[place], as illustrated in
(24). The positional faithfulness constraint Ident(Onset)-[place] is responsible for choosing the
candidate with the palatal lateral when syllabified in onset position.

(24) Positional faithfulness analysis of coda depalatalization

donceʎ Id(Ons)-[pl] *[dorsal] Id-[pl]
 a. doncel *
b. donceʎ *W L
donceʎ+a Id(Ons)-[pl] *[dorsal] Id-[pl]
 a. donceʎa *
b. doncela *W L *W

In order to explain the overapplication of depalatalization in onset position observed in

the plural inflected forms in (21) (also observed in diminutive forms), Lloret and Mascaró
(2006) propose an output-output faithfulness constraint Identity-OBaseOAffixedForm-[place] that
prevents affixed forms from mapping faithfully onto the [dorsal] specification of nasals and
laterals although syllabified in onset position (25). This kind of output-output faithfulness con-
straint in OT expresses paradigmatic pressure effects (see Benua 2000). See Lloret and Mascaró
(2006) for a more detailed definition and discussion of what constitutes a base for output-output
faithfulness constraints, and Bessett and Colina (2017) for a recent account of depalatalization
in Spanish that extends Lloret and Mascaró’s (2006) analysis of depalatalization by integrating
diachronic data and loanword adaptation.

(25) Analysis of overapplication of depalatalization in inflected forms

donceʎ+es Id-OO-[pl] Id(Ons)-[pl] *[dorsal] Id-[pl]
base: doncel
 a. donceles * *
b. donceʎes *W L *W L

The final grammar for nasal and lateral coda depalatalization is illustrated in (26) as a Hasse
diagram.

(26) Hasse diagram for coda depalatalization

Id-OO-[pl]

Id(Ons)-[pl]

*[dorsal]

Id-[pl]
Another process of neutralization in Spanish that involves the alveolar lateral is liquid neutrali-
zation, by which the contrast between /l/ and /ɾ/ is neutralized in favor of [ɾ] in word-medial

18
 Phonemic contrast and neutralization

coda position (both caldo ‘broth’ and cardo ‘thistle’ can be realized as [ˈkaɾðo]), and in favor of [l]
in word-final coda position in some dialects of Spanish (e.g. amo[l] ‘love’ but amo[ɾ]es ‘loves’).We
refer the reader to Martínez-Paricio (2008) for an OT analysis of these phenomena based on
sonority distinctions among liquids (see also the references therein).

3.3. The distribution and neutralization of rhotic consonants

The distribution of the trill and the flap in Spanish is puzzling as this contrast is only found
intervocalically. Stem-initially and after an homorganic heterosyllabic consonant, only the trill
is possible; in coda position, the flap is typically realized ([r] is possible under emphatic speech);
as the second member of a complex onset, only the flap is possible. Some data appear in (27),
mostly taken from Bradley (2006) and Colina (2009) (see also Roca 2005; Colina 2010).

(27) Distribution of rhotics

a. Stem-initial position and after [r]osa ‘rose’
homorganic and heterosyllabic
consonants (only [r])
pre[.r]equisito ‘prerequisite’
mata[.r]atas ‘rat poison’
con [r]osa ‘with Rose’
la [r]osa ‘the rose’
hon[r]a ‘honor’
al[r]ededor ‘around’
b. Coda position (only [ɾ]) ve[ɾ]; ve[.ɾ] osas (cf. ve ‘to see’; ‘to see bears-f’
[r]osas) (cf. ‘(s)he sees roses’)
ma[ɾ], ma[ɾ]es (cf. *ma[r]es) ‘sea’, ‘seas’
c. Intervocalic position ca[r]o ‘car’
ca[ɾ]o ‘expensive’
pe[r]o ‘dog’
pe[ɾ]o ‘but’
d. Second member of a complex g[ɾ]ano (cf. *g[r]ano) ‘grain’
onset
b[ɾ]azo (cf. *b[r]azo) ‘arm’

Colina (2009) makes the following assumptions regarding the trill: (i) the trill is a single ele-
ment (not a geminate), (ii) it is [‒continuant], and (iii) it is less sonorous than the flap (for a detailed
explanation of the notion of sonority, see the chapter “Phonotactic Constraints on Syllable Struc-
ture”). To these assumptions, we add the following: (i) flaps, as opposed to trills, are [+continuant],
and (ii) continuancy is the feature that distinguishes trills from flaps in Spanish (Colina 2010 uses
the feature [tense], but others assume an underlying flap that undergoes strengthening (Harris
1983; Bonet and Mascaró 1997 for Catalan; Roca 2005)).The phonotactic generalization accord-
ing to which trills are banned after heterosyllabic glides (28) seems to support a [+continuant]
characterization of flaps; the only exceptions to this generalization are two words borrowed from
Basque (aurragado [awraˈɣaðo] ‘tilled in a bad way’ and aurresku [awˈresku] ‘traditional Basque
dance’).This generalization may be due to [+continuant] spreading, as in voiced stop spirantization
(see the chapter “Spirantization and the Phonology of Spanish Voiced Obstruents”). For a word
like Israel, pronounced as [i(r).ra.ˈel], in which the rhotic is preceded by a sibilant, things are a bit
more complex: why does the rhotic following the sibilant surface as a trill and not a flap, if both the

19
Francesc Torres-Tamarit

sibilant and the flap are [+continuant] segments? One could argue for assimilation of the sibilant
to the rhotic (/sɾ/ → [ɾ.ɾ]) with concomitant strengthening of two adjacent flaps (ɾ.ɾ → [(r).r]). An
additional argument in favor of considering trills as [‒continuant] and flaps as [+continuant] is that
only trills are possible in word-initial position, in line with the onset preference for less sonorous
segments ([‒continuant] segments are less sonorous than [+continuant] segments) (see also Colina
2009). Also, in most varieties of Spanish, obstruents in coda position must be [+continuant]. Flaps,
but not trills, are precisely preferred in the coda position. In any case, which feature distinguishes
flaps from trills is a complicated matter that is still under debate.

(28) After heterosyllabic glides, no trill

a[j.ɾ]e ‘air’
vie[j.ɾ]a ‘scallop’
bo[j.ɾ]a ‘fog’
centa[w.ɾ]o ‘centaur’
An analysis of the distribution of rhotics in Spanish within Dispersion Theory has been
set out in Bradley (2006). In Dispersion Theory phonological contrast is directly evaluated
by the grammar, which includes constraints that promote maximizing the perceptual dis-
tinctiveness of a contrast. For rhotics, this type of minimal distance constraint is Spaceduration,
defined in (29a), according to which rhotic duration contrasts should be as robust as they are
in intervocalic position; it is assumed that trills are intrinsically longer than flaps, and that this
duration is better perceived in intervocalic position. The constraint Spaceduration is therefore
violated by any contrast in all contexts except for intervocalic contexts. Identduration (29b) is
the faithfulness constraint that penalizes input-output mismatches in terms of duration. The
next relevant constraint is *σ[ɾ, which penalizes flaps in onset position, a prominent position
that requires stronger segments, a constraint based on sonority considerations (see Clements
1990). Finally, the two context-free markedness constraints *r and *ɾ are also relevant for
the analysis.

(29) Constraints in Bradley (2006)

a. Spaceduration
Potential minimal pairs differing in rhotic duration differ at least as much as rhotics do
between vowels.
b. Identduration
Corresponding input and output rhotics are identical in duration.
c. *σ[ɾ
A rhotic in syllable-initial position is [r].
d. *r
e. *ɾ
The contrast between trills and flaps in intervocalic position derives from ranking Identduration
above *σ[ɾ, *r, and *ɾ, as illustrated in (30); the contrast in the input must be preserved in the out-
put. In intervocalic position, Spaceduration is never violated because the intervocalic context is the
one that best enhances the perceptual contrast of duration between the two rhotics. Candidates
(30b) and (30c) are ruled out because the constrast between the trill and the flap in intervocalic
position has been neutralized.

20
 Phonemic contrast and neutralization

(30) Rhotic-flap contrast in intervocalic position

VrV1 – VɾV2 Spaceduration Identduration *σ[ɾ *r *ɾ
 a.VrV1 – VɾV2 * * *
b.VrV1,2 *W L * L
c.VɾV1,2 *W * L *

However, in non-intervocalic contexts, like word-initial position or post-consonantically,

the cues for perceiving the duration contrast between the two rhotics is diminished, and
Spaceduration, which dominates Identduration, blocks the maintenance of the contrast, as shown in
(31). The reason why the trill is preferred over the flap is the activity of the strengthening con-
straint *σ[ɾ, which acts to ban flaps from the onset position in those contexts in which contrast
cannot be maintained for perceptibility reasons.

(31) Neutralization in non-intervocalic position

rV1 – ɾV2 Spaceduration Identduration *σ[ɾ *r *ɾ
 a. rV1,2 * *
b. ɾV1,2 * *W L *W
c. rV1 – ɾV2 *W L * *W

One attractive aspect of Bradley’s (2006) analysis is that it does not need any specific con-
straint banning trills in coda position (cf. Colina’s 2009 *[r]/Coda) in order to account for neu-
tralization toward the flap in coda position. The specific ranking of the low-ranked markedness
constraints, with *r above *ɾ, is sufficient, as exemplified in (32). The same holds for complex
onsets, in which the second element is always the flap.

(32) Neutralization in coda position

Vr1 – Vɾ2 Spaceduration Identduration *σ[ɾ *r *ɾ
 a. ɾV1,2 * *
b. rV1,2 * *W L
c.Vr1 – Vɾ2 *W L *W *

The Hasse diagram of Bradley’s (2006) analysis of the distribution of rhotics in Spanish is
shown in (33).

(33) Hasse diagram for the distribution of rhotics (Bradley 2006)

Spaceduration

Identduration

*σ[ɾ

*r

*ɾ

21
Francesc Torres-Tamarit

3.4. Coda stop neutralization processes

In this section we focus on the phonology of stops in coda position in north-central Peninsular
Spanish, which exhibits intricate patterns of neutralization and preservation of contrast that are
usually subject to variation.

3.4.1. Word-final coda stops

If our analysis is limited to native words of Spanish, only a subset of consonants are allowed in
word-final position, all of them being coronal: /ɾ, l, n, d, s, θ/; as noted earlier, /s/ in coda posi-
tion is debuccalized to [h] in aspirating varieties of Spanish (Andalusia and Latin America; see
the chapter “Vowel Harmony” for more details on the process of /s/ aspiration). Some examples
appear in (34).

(34) Possible word-final consonants in Spanish

so[ɾ] ‘sister (nun)’
so[l] ‘sun’
so[n] ‘they are’
se[ð] ‘thirst’
to[s] ‘cough’
pe[θ] ‘fish’
In north-central Peninsular Spanish, word-final /d/ is realized as an approximant [ð], espe-
cially when it is followed by another word, either vowel-initial or consonant-initial, or a pro-
nominal clitic. Some examples taken from Navarro Tomás (1932) appear in (35).

(35) Word-final /d/ in north-central Peninsular Spanish

juventu[ð] estudiosa ‘studious youth’
liberta[ð] absoluta ‘absolute freedom’
eda[ð] media ‘Middle Ages’
eda[ð] dorada ‘Golden Age’
llama[ð]lo ‘call-pl him!’
escribi[ð]nos ‘write-pl us!’
The specific realization of word-final /d/ is also subject to stylistic variation, and Navarro
Tomás (1932) reports cases of weakening and partial devoicing in absolute final position, which
he transcribes as [ð̥ ] (36).

(36) Weakened and partially devoiced word-final /d/ before pause

liberta[ð̥ ] ‘freedom’
huéspe[ð̥ ] ‘host’
bonda[ð̥ ] ‘goodness’
virtu[ð̥ ] ‘virtue’
veni[ð̥ ] ‘come!-pl’
espera[ð̥ ] ‘wait!-pl’
trae[ð̥ ] ‘bring!-pl’
calla[ð̥ ] ‘be-pl quiet!’

22
 Phonemic contrast and neutralization

In colloquial speech, Navarro Tomás (1932) also reports cases of complete deletion of final
/d/ (e.g. virtu[∅], verda[∅] ‘truth’, juventu[∅], liberta[∅], uste[∅] ‘you-formal’, Madri[∅] ‘Madrid’)
that is not found in monosyllables (e.g. se[ð] ‘thirst’, re[ð] ‘net’, vi[ð] ‘vine’). Finally, in some areas
in Castile and Madrid, word-final /d/ is not only spirantized but also completely devoiced (e.g.
virtu[θ], verda[θ], juventu[θ], uste[θ]); this same phenomenon of spirantization and final devoicing
also targets /ɡ/, realized as [x], in some Castilian varieties, e.g. bulldo[x] ‘bulldog’. The devoiced
realization [θ] is dispreferred before vowel-initial words (e.g. se[ð] enorme ‘enormous thirst’).
Intermediate realizations like [ð̥ ], with partial devoicing, are also possible according to Hualde
(2005). Word-final /d/ devoices and fails to undergo spirantization in the Spanish spoken by
native speakers of Catalan, and in some parts of Latin America in careful speech, as reported by
Hualde (2005) (e.g. salu[t] ‘health’, Madri[t]).
All other consonants in word-final position, that is, all obstruents except for /d/ (/p, t, k, b,
ɡ, f, ͡tʃ, x, (ʝ)/) and noncoronal sonorants (/m, ɲ, (ʎ)/), are prohibited in such a position. In fact,
most roots ending in these consonants take an inflectional ending /-e/ if they are patrimonial
words or established loans (e.g. nube ‘cloud’, toque ‘touch’, jefe ‘boss’, eje ‘axis’; see Harris 1999;
Lloret and Mascaró 2006).
The situation is different when we look at borrowings, where all voiceless obstruents (the
stops /p, t, k/, the affricate /t͡ ʃ/, and the fricatives /f, x/) are possible; a nonborrowed word like
reloj ‘watch’ is usually pronounced as [reˈlo], only as [reˈlox], in more affected speech or specific
idiolects. All the data in (37) and (38) are taken from Navarro Tomás (1932), Hualde (2005), and
Lloret and Mascaró (2006), or are my own.

(37) Word-final voiceless stops in Spanish

/p/ /t/ /k/
[ˈket͡ ʃup] ‘ketchup’ [ˈaβitat] ‘habitat’ [buˈtik] ‘boutique’
[ˈkɾep] ‘crêpe’ [aɾˈɣot] ‘slang’ [anoˈɾak] ‘anorak’
[esˈkajp] ‘Skype’ [ˈswit] ‘suite’ [aˈðok] ‘ad hoc’
[esˈtop] ‘stop’ [entɾeˈkot] ‘entrecôte’ [koˈɲak] ‘cognac’
[ˈwasap] ‘WhatsApp’ [akˈθesit] ‘additional prize’ [esˈtok] ‘stock’
[ˈθenit] ‘zenith’ [tikˈtak] ‘tick tock’
[ˈimput] ‘input’ [ˈblok] ‘writing pad’
[maˈmut] ‘mammoth’ [biˈβak] ‘bivouac’
[roˈbot] ‘robot’ [ˈfɾak] ‘tailcoat’

(38) Word-final voiceless affricates and fricatives in Spanish

/t͡ ʃ/ /f/ /x/
[ˈkit͡ ʃ] ‘kitsch’ [ˈrif] ‘Rif ’ [ˈsix] ‘Sikh’
[ˈmat͡ ʃ] ‘match’ [ˈnif] ‘NIF (fiscal ID number)’ [ˈbox] ‘boxwood’
[ˈθaɾeβit͡ ʃ] ‘czarevitch’ [ˈt͡ ʃef] ‘chef ’
[ˈkambɾit͡ ʃ] ‘Cambridge’ [naˈif] ‘naïve’
[ˈpuf] ‘pouf ’
[ˈrozβif] ‘roast beef ’
[ˈpaf] ‘pub’
Regarding the realization of the word-final voiced stops /b, ɡ/ in borrowings, as happens
with word-final /d/ in native words, the most common realization is an approximant (39). Final

23
Exploring the Variety of Random
Documents with Different Content
 ANNE BOLEYN'S WINDOW, DEAN'S CLOISTERS
Henry VI, the founder of Eton College in 1440, was born at Windsor,
and buried in the south aisle of St George's Chapel, but not until
some years after his death.
Edward IV rebuilt St. George's Chapel, or began the building which
was completed under Henry VIII and Edward VI. On the north side
of the Chapel he built the Dean's and Canons' houses, and those of
the petty Canons. Edward and his queen were buried near the altar
under a tomb of such splendour that it was plundered in 1642.
Henry VIII began the royal tomb-house at the east end of St.
George's as a sepulchral chapel for himself. Later he granted it to
Wolsey, who caused a black marble sarcophagus to be made,
bordered and canopied with costly bronze work. The Cardinal never
lay under it. It was stripped, and the ornaments sold, by Parliament
soldiers a century later. The sarcophagus itself was afterwards used
to cover the body of Nelson at St. Paul's. In the choir of the Chapel
of St. George lie Jane Seymour and Henry VIII, who built the
gateway bearing his name, under which the public enter the lower
ward.
When Henry VIII's queen, Anne Boleyn, was crowned in June, 1533,
Henry Howard, Earl of Surrey, and then about sixteen years old,
carried the fourth sword. This young man, who was to be atrociously
executed at the age of thirty by the same Henry, lived at Windsor for
some time as the companion of the king's bastard son, the Duke of
Richmond, and while confined there some years later he wrote a
poem which gives perhaps the most beautiful picture connected with
Windsor. I will belittle the rest of this little book by here quoting the
poem in full: "Prisoned in Windsor he recounteth his pleasure there
passed":
So cruel prison how could betide, alas,
As proud Windsor? Where I, in lust and joy,
With a King's son, my childish years did pass,
In greater feast than Priam's sons of Troy.
Where each sweet place returns a taste full sour,
The large green courts, where we were wont to hove,
With eyes cast up into the Maiden's tower,
And easy sighs, such as folk draw in love.
The stately seats, the ladies bright of hue,
The dances short, long tales of great delight;
With words and looks that tigers could but rue;
Where each of us did plead the other's right.
The palme-play, where, despoiled for the game,
With dazzled eyes oft we by gleams of love
Have miss'd the ball, and got sight of our dame,
To bait her eyes, which kept the leads above.
The gravel'd ground, with sleeves tied on the helm,
On foaming horse, with swords and friendly hearts;
With chere, as though one should another whelm,
Where we have fought, and chased oft with darts.
With silver drops the mead yet spread for ruth,
In active games of nimbleness and strength,
Where we did strain, trained with swarms of youth,
Our tender limbs, that yet shot up in length.
The secret groves, which oft we made resound
Of pleasant plaint, and of our ladies' praise;
Recording oft what grace each one had found,
What hope of speed, what dread of long delays.
The wild forest, the clothed holts with green;
With reins availed, and swiftly-breathed horse,
With cry of hounds, and merry blasts between,
Where we did chase the fearful hart of force.
The wide walls eke, that harbour'd us each night:
Wherewith, alas! reviveth in my breast
The sweet accord: such sleeps as yet delight;
The pleasant dreams, the quiet bed of rest;
The secret thoughts, imparted with such trust;
The wanton talk, the divers change of play;
The friendship sworn, each promise kept so just,
Wherewith we past the winter night away.
And with this thought the blood forsakes the face;
The tears berain my cheeks of deadly hue:
The which, as soon as sobbing sighs, alas!
Up-supped have, thus I my plaint renew:
"O place of bliss! renewer of my woes!
Give me account, where is my noble fere?
Whom in thy walls thou dost each night enclose;
To other lief; but unto me most dear."
Echo, alas! that doth my sorrow rue,
Returns thereto a hollow sound of plaint.
Thus I alone, where all my freedom grew,
In prison pine, with bondage and restraint:
 And with remembrance of the greater grief,
To banish the less, I find my chief relief.
The critics, I believe, regard this poem as a conventional poetical
exaggeration of some unimportant or wholly imaginary event in
Surrey's life, because he was then married, and because the lady
who is conjectured to have been the subject of his "Description and
Praise of Geraldine" was then only twelve years old.
Queen Elizabeth built the North Terrace of the Castle in 1576, a
gallery to the west of it now used as a library, and an octagon
banqueting hall, at the east end, which Charles I pulled down to
substitute a gateway and drawbridge leading into the Home Park. He
also demolished the fountain of Queen Mary Tudor in the Upper
Ward. He thought, but in vain, to build another banqueting hall, and
to construct a fountain, where Hercules was to have been seen
strangling Antæus, so as to make it appear that "by squeezing of
him the water came out of his mouth". Charles often held his Court
at Windsor, and was at the Castle in January when the Civil War was
at hand; there was a garrison of forty officers and four hundred
horse, and wagons of ammunition were arriving. But in October,
1642, appeared a pamphlet, entitled "Exceeding true and happy
news from the Castle of Windsor declaring how several troops of
Dragoons have taken possession of the said Castle to keep it for the
use of the King and Parliament". "For King and Parliament" was a
euphemism. Windsor was esteemed one of the strongest places in
the kingdom, and could the Cavaliers have retained and fortified it,
they might have descended upon London. And so "several well-
affected Gentlemen and valiant Religious Commanders have gone to
raise several troops of Dragooners and Volunteers, some of which
are already arrived at Windsor, and have taken possession of the
Castle". The intruders took the chapel plate of St. George's and
coined it into money for the Parliament; they despoiled Wolsey's
tomb; and they carried off Edward IV's embroidered surcoat of
crimson velvet, wrought with gold and pearls and decorated with
rubies, which had hung over his tomb since the opulent funeral of
1483.

NORTH TERRACE AND WINCHESTER TOWER

Prince Rupert attacked the Castle in the same year, 1642, but
without success, and in the winter and spring following Essex made
it his headquarters and a prison for Royalists, while Rupert flickered
here and there about Oxford. At the end of the war Windsor was the
strange foil to that notable prayer meeting of the Army officers held
some time early in 1648. The Army was uneasy in its relations with
people and Parliament; it had cause to fear a revival of royalism; and
some officers had thought of laying down their arms, because what
they had done, and were willing to do, for the nation was not
acceptable to it. Therefore they spent two days together in prayer at
Windsor Castle, enquiring when it last was that they could say with
confidence: "The presence of the Lord was among us". On the third
day the "gracious hand of the Lord" showed them how they had
come to their present trouble and uncertainty. It was through their
treating with the king and his party, this of course being prompted
by their own "conceited wisdom, fear, and want of faith". Thus they
were led to loathe their iniquities. They wept for shame of their
unbelief and trust in the wisdom of this world, and they arrived at a
humble confidence and "a very clear and joint resolution, That it was
our duty, if ever the Lord brought us back again in peace, to call
Charles Stuart, that man of blood, to an account for that blood he
had shed, and mischief he had done to his utmost, against the
Lord's Cause and People in these poor Nations". These are the words
written in 1659 by Adjutant Allen, who was at the prayer meeting.
In less than a year, on Christmas Eve, 1648, there was "terrible and
bloody news from Windsor". The king was brought from Hurst Castle
by Colonel Harrison and ten troops of horse. At the passing of the
king the people of Windsor cried: "God bless your majesty and send
you long to reign"; and after he entered the Castle the Royalists of
the town drank a carouse to their dread sovereign, but were "taken
off from that ceremonial and cant-like action" by several files of
musketeers, not before several had been wounded and three killed.
Charles did not return to Windsor again until he was dead. His body
was borne thither without pomp or noise. When the attendant lords
—the Duke of Richmond, the Marquis of Hertford, and the Earls of
Southampton and Lyndsy—requested that the body might be buried
according to the form of the Common Prayer Book, the Governor
"expressly, positively, and roughly refused to consent to it, and said
it was not lawful; that the Common Prayer Book was put down...."
As the coffin was brought to St. George's Chapel, snow fell and gave
the black pall the "colour of innocency". Such were the dismal
mutations of the Chapel, that the lords scarce knew where they
were. "A fellow of the town" showed them the vault of Henry VIII
and Jane Seymour, and there they laid him. At a later date the vault
was opened to receive a nameless child of Queen Anne's. There in
the vault just before the altar, John Evelyn in 1654 found "our
blessed martyr, King Charles".
Cromwell occasionally lived at Windsor. Charles II used it as his
summer lodging, and Nell Gwynn had a house, called Burford House,
close to the Castle. The king was at St. George's Feast in 1663 with
Lady Castlemaine as well as the queen. Pepys heard that the Duke
of Monmouth danced with the queen, his hat in his hand, and that
"the king came in and kissed him, and made him put on his hat,
which everybody took notice of". Pepys spent a cheerful, carnal day
in the Castle and at Eton on February 26, 1665, admiring the Chapel
and the banners and the singing and "the most romantique castle
that is in the world", and "giving a great deal of money to this and
that man and woman". When Evelyn saw the Castle in August, 1670,
Prince Rupert was Constable, and "had begun to trim up the keep or
high round tower, and handsomely adorned his hall with furniture of
arms, which was very singular, ... so disposing the bandoleers,
holsters and drums, as to represent festoons, and that without any
confusion, trophy-like. From the hall we went into his bed-chamber,
and ample rooms hung with tapestry, curious and effeminate
pictures so extremely different from the other, which presented
nothing but war and horror." The king was hunting the stag, walking
in the Park, and planting it with rows of trees. The Castle was
"exceedingly ragged and ruinous", and about to be repaired. Wren
Italianised the façade, and the Castle was to some extent rebuilt and
altogether remodelled into something which later critics considered
monotonous and commonplace. The interior was decorated by the
carvings of Gibbons and Antonio Verrio's inert and luscious paintings
of "Judith and Holofernes", "Leda and the Swan", and the like, which
Evelyn, who saw the frescoes of St. George's Hall in 1683, admired
for their "full and flowing, antique and heroical" style. Gibbons also
made the copper statue of the king on horseback, which was newly
set up in July, 1680, on its pedestal of white marble, where it still
stands. The outer ditches of the Castle were filled in. Terraces were
formed on the south and east, and the north terrace was enlarged.
The Devil's Tower was given to the Maids of Honour. Charles meant
to face the mound of the Round Tower with red brick, but was
prevented.
 NELL GWYN'S HOUSE AND HENRY VIII GATEWAY
James II turned the Tomb House into a chapel for the practice of his
own religion, and its ceiling was decorated by Verrio. At Windsor he
received the Papal Nuncio. On the king's downfall a revolutionary
crowd therefore attacked the Chapel, destroyed the windows, and
left the interior in ruin. Further alterations planned by William III
were not carried out. Queen Anne's work was in the Park. In her
reign Swift was often at Windsor with his friend Harley, the Secretary
of State, supping with Prior and Arbuthnot, playing twelvepenny
picquet and winning seven shillings at it, putting his thumb out of
joint by boxing Patrick's ear for carelessness, riding out in the forest
on "the finest day in the world"—October 4, 1711—"a noble caravan
of us", Maids of Honour, the Duke and Duchess of Shrewsbury,
Arbuthnot, and others, some driving, some riding, and Swift on
horseback in a coat of light camlet, faced with red velvet, and silver
buttons. On September 1, 1711, Swift came to Windsor with a
basket of fruit for his friend Lewis from Lord Peterborough's garden
at Parson's Green. "I durst not eat any fruit, but one fig," he writes
to Stella, and asks, "Does Stella never eat any? What, no apricots at
Donnybrook? Nothing but claret and ombre? I envy people
maunching and maunching peaches and grapes, and I not daring to
eat a bit. My head is pretty well, only a sudden turn any time makes
me giddy for a moment, and sometimes it feels very stuffed; but if it
grows no worse, I can bear it very well. I take all opportunities of
walking; and we have a delicious park here just joining to the castle,
and an avenue in the great park very wide, and two miles long, set
with a double row of elms on each side. Were you ever at Windsor?
I was once a great while ago; but had quite forgotten it."
The two first Georges neglected the Castle—though the second
placed there Windsor's last prisoner, the Maréchal de Belleisle—to
such an extent that George III had to build the "Queen's Lodge" for
himself and his family. This building is immortalized by the king's
remark to Fanny Burney, a Maid of Honour to his Queen Charlotte,
"Was there ever such stuff as much of Shakespeare? ... only of
course one must not say so"; but it is now pulled down. St. George's
Chapel had been completely neglected, probably because there was
no need of it, for many years before George III began to renovate
and repave it in 1787. He removed the tracery and glass of the east
window, in order to exalt a new picture of the Resurrection in
painted glass. The walls were then stained to harmonize with the
heavy colouring of this picture, and finally the Chapel was darkened
by the blocking up of the clerestory, to destroy the painful contrast
between the sunlit walls and the glass. George III also restored the
ruined Tomb House and dug a royal vault beneath it. Under George
IV the monument to the Princess Charlotte of Wales was placed in
the chantry at the west end of the north aisle. Two years before,
Wyatville (né Wyatt) began to rebuild the Castle. The incongruous
buildings and external additions of the Restoration were removed.
The entrance gateway to the upper ward, with its two towers of
Lancaster and York, was made. The height of the Round Tower was
increased by thirty-nine feet and a flag turret added. The old houses
under the Curfew, Garter, and Salisbury towers were cleaned away.
Thus in four years the exterior of the Castle, shaken more free of the
little town clambering and clustering about it, was brought to its
present state, which may or may not have blasted the hopes of the
author of this epigram:
Let restless George who can leave nothing quiet,
Change if he will the good old name of Wyatt:
But let us hope that their united skill
May not make Windsor Castle Wyatville.
The Tomb House was converted by Queen Victoria and Sir Gilbert
Scott into the Albert Memorial Chapel, on the death of the Prince
Consort in 1861. As a memorial to the same prince, the mullions
were restored to the east window of St. George's Chapel in 1863.
Queen Victoria lived much at Windsor, and in her time the interior of
the Castle attained its present height of costliness and domesticity.
The majesty without and the splendour within answer fully to the
expectations usually founded upon a reading of history and a sober
loyalty to the Crown.
ETON COLLEGE FROM WINDSOR
 WINDSOR FOREST AND PARK
In the Conqueror's time and before, it must have been hard to say
what was Windsor Forest, or what was not, on the south side of the
middle course of the Thames. After choosing the mound of Windsor
for a castle, William enlarged the Forest so that it included a great
part of Berkshire, as far west as Hungerford; some of
Buckinghamshire, which is on the north side; parts of Middlesex,
Oxfordshire, and Hampshire, and in Surrey both banks of the Wey as
far as Guildford. The Forest and the river surrounded and isolated
the Castle on every side. The Forest was named after Windsor from
early times, but was also sometimes called Oakingham or
Wokingham Forest. All this wild virgin country of heath, swamp,
tangled wood, and high land held many deer for the king's hunting,
and fattened many swine. Partly by the number of swine feeding in
it the value of a forest was estimated; and the right to send swine
among the acorns of Windsor was retained or acquired by many of
the dwellers at the edge of the Forest or within it, from the boor to
the nuns of Ankerwyke near Datchet.
There were many portions of cultivated land running into the forest
or islanded in its midst. Some even of the woods inside the borders,
such as Clewer, Bray, Hurley, Bisham, and Finchhampstead,
remained under separate ownership, with their own woodwards,
though open to admit the king's game. The oaks of the Forest are
often mentioned in early records, together with alders, birches,
beech, and ash. There are oaks at Cranbourne, and a beech at
Smith's Lawn, which are conjectured to have been seedlings at the
Conquest, perhaps earlier. Gifts of timber for building were
frequently made to religious houses in the neighbourhood and to
private men. Six oaks were sent to the Tower in 1276, wherewith to
burn lime for the masonry; and the builders of Windsor Castle in
William I's, Henry II's, Henry III's, and Edward III's time must have
drawn abundantly from the oaks in the clay of the lower lands. The
game in the Forest was of many kinds. The red deer was the noblest
in appearance, in speed, and in esteem. Fox, otter, badger, wild cat,
and hare were also hunted. There were wild cattle as late as 1277,
for in that year the Constable of Windsor was ordered to capture and
sell them. Among the Forest offences were the carrying away of
boughs and felling of trees, the pasturing of sheep, the taking of
does with a noose, hunting with greyhounds, hawking at pheasants
and partridges. The poachers included labourers, husbandmen,
gentlemen, and a rector. A tenth part of the venison, under Henry I,
Henry II, and Richard I, was granted to God and St. Mary of
Abingdon.
In the reign of Edward I the Chief Forester was under the orders of
the Constable of Windsor. Under Edward III the Constable was also
Parker of the Great Park, which had gradually been fenced in out of
the larger and vaguer extent of the Forest itself. Yet another
enclosure was made in 1467 by Edward IV, namely two hundred
acres close to Windsor, which were the origin of the "Home Park",
once called the "Little Park". There Henry VII and Philip of Castile
killed deer "with their own hands, with their crossbows"; even so
early was it a notable thing for a sovereign to do what many a man
does without thinking about it. Henry VIII loved the chase, and
hunted in Windsor Forest all day, from morning until nightfall. He
also shot, hawked, fished, and played tennis, and having killed the
deer, watched the men who ate quantities of venison for a wager.
Elizabeth hunted at Windsor, attended by half a hundred ladies on
hackneys, and once, in 1602, shot a great fat stag, and sent it to
Archbishop Parker as a gift. It is supposed to have been in her
childhood, in her father's reign, that the events which led to the
story of Herne the hunter took place:
There is an old tale goes that Herne the hunter,
Sometime a keeper here in Windsor forest,
Doth all the winter-time, at still midnight,
Walk round about an oak, with great ragg'd horns;
 And there he blasts the tree, and takes the cattle,
And makes milch-kine yield blood, and shakes a chain
In a most hideous and dreadful manner.
So speaks Mistress Page in opening her plans for the discomfiture of
Falstaff. It is said that a yeoman hanged himself on a tree for fear of
the king after hunting in the Forest without leave. The tree was
cursed, and a ghostly stag haunted the place and butted at the tree
and breathed smoke and fire as it tore the roots. There was also a
story that Herne was a keeper and went mad after being gored by a
stag. He tied a pair of antlers upon his head, ran naked through the
Forest, and hanged himself on the tree, near Shakespeare's Oak in
the Home Park, which was called Herne's Oak for centuries, and was
blown down in 1863, or, according to another opinion, cut down by
George III. Queen Victoria planted the oak which marks the site of
the legendary tree.

VIRGINIA WATER
In Elizabeth's reign the first systematic planting was begun by Lord
Burleigh. Thirteen acres near Cranbourne Tower were sown with
oaks which were never pollarded, like most other trees in the Forest,
to provide browsing for the deer. This planting in 1580 was to supply
the navy, especially in case the Spaniards should destroy the oaks of
the Forest of Dean, as they had planned to do. Since that date a
more or less contemporary record of successive plantings has been
made, and where the planter has been a royal or distinguished
personage, his or her name is attached to the recording plate.
James I hunted in the Forest, closed the Little Park against the
public, and turned out some wild pigs, of which a few are still left. In
his time the circumference of the unenclosed Forest on the Berkshire
side of the river measured seventy-seven miles and a half, and here
ran the red deer. The Home Park of two hundred and eighty acres
held two hundred and forty fallow deer, and the Great Park of three
thousand six hundred and fifty acres held eighteen hundred. Charles
I also hunted there, and at the beginning of the Civil War deer were
lawlessly killed and the pales of the Park destroyed. Bulstrode
Whitelocke was Constable of the Castle and Keeper of the Forest
under the Commonwealth, but could not keep down the poaching.
Charles II and William III planted the Long Walk.
Queen Anne hunted in a chaise, and Swift, in 1711, says that she
was hunting until four in the afternoon, and covered more than forty
miles. She planted with oaks the ride known by her name. Sarah,
Duchess of Marlborough, was Ranger for many years, and has to be
gratefully remembered for protecting the trees against Walpole when
he was in need of money. Though the first two Georges did nothing
for the Castle except by neglecting it, in their reigns there were
several plantings of trees. The avenue of lime trees east of
Cumberland Lodge was made under George I. Under George II were
formed some of the plantations round about the heathery Smith's
Lawn at the south end of the Park: these were the first to be shaped
according to the lines of the ground, and not circular or in
parallelograms as before; and it is said that some of this work was
given, for lack of anything else, to soldiers raised against the
Rebellion of 1745. In the time of George II thirteen-hundred red
deer ran in the Forest. By 1806 they numbered only three hundred,
though as late as 1813 the Forest was fifty-six miles and a half in
circumference, and included Wokingham and a great part of Bagshot
Heath. The Forest was still unenclosed, but squatters had been
steadily enlarging their pieces of land by carrying forward their
ditches at the time of scouring them, while parishes within the
boundaries had raised money by allowing persons to enclose and
acquire portions of the common land. In 1817 awards were given,
settling the claims of various occupants, and the Forest, or every
tract of it which retained that title, was enclosed and the deer driven
into the Great Park. This is now eighteen hundred acres in extent,
and holds a thousand fallow deer and a hundred red deer,
Cranbourne Park holding a small herd of white deer.
Though crossed by public footpaths and roads, it is at most times
and places clear that the Park is the front garden of Windsor Castle.
There is even a sense of privacy unintentionally disturbed at spots
here and there where the family grief or rejoicing of royalty has been
celebrated by planting a tree—as when Queen Victoria planted an
oak to mark the place where the Prince Consort finished his last
day's shooting, November 23, 1861. Yet the Park is about six miles
in length from the Castle southward to Virginia Water, and at most
points from two to three miles wide. Considering this extent, it has
no great effect of space. This is due to the lack of any great quality
of art or nature in the Park. Its outline has no natural wholeness,
and the boundaries, marked by fences and walls and several lodges,
are not easily forgotten. The eighteen hundred acres have little
grace of undulation or natural variety; and they are made up of a
number of separate but not integral parts, so that it is not one but
many. Curiosity, admiration, respect, and surprise follow one another
too rapidly for any but the first and last to be satisfied. There are a
thousand excellent or notable things—some due to chance and
antiquity, some to deliberation and design—but the Park as a whole
has no supremacy over others of the same or even less extent. I
have no sooner admired the exquisite giant birches, or the craggy
vast oaks, or the perfectly formed younger ones, than I come to
lines of rhododendrons, the symbols of very modern riches, or to
lines of venerable stately trees which are not satisfying except on
the rare occasions when they overhang some human stateliness or
splendour. The Park was grand and stern under Plantagenets or
Tudors, when the poet could say of it—
No Forest, of them all, so fit as she doth stand,
When Princes, for their sports, her pleasures will
command,
No Wood-nymph as herself such troops hath ever
seen,
Nor can such quarries boast as have in Windsor been;
it was sweet and gallant under Stuarts and early Hanoverians. But
the charm is faded and the grandeur confounded, and the Park
should either be artistically treated as a whole, or allowed a century
of nature and wise neglect, if these qualities are to return in a
measure worthy of its repute and history.

TRANSCRIBER'S NOTES:
Obvious typographical errors have been corrected.
*** END OF THE PROJECT GUTENBERG EBOOK WINDSOR CASTLE
***

Updated editions will replace the previous one—the old editions will
be renamed.

Creating the works from print editions not protected by U.S.

copyright law means that no one owns a United States copyright in
these works, so the Foundation (and you!) can copy and distribute it
in the United States without permission and without paying
copyright royalties. Special rules, set forth in the General Terms of
Use part of this license, apply to copying and distributing Project
Gutenberg™ electronic works to protect the PROJECT GUTENBERG™
concept and trademark. Project Gutenberg is a registered trademark,
and may not be used if you charge for an eBook, except by following
the terms of the trademark license, including paying royalties for use
of the Project Gutenberg trademark. If you do not charge anything
for copies of this eBook, complying with the trademark license is
very easy. You may use this eBook for nearly any purpose such as
creation of derivative works, reports, performances and research.
Project Gutenberg eBooks may be modified and printed and given
away—you may do practically ANYTHING in the United States with
eBooks not protected by U.S. copyright law. Redistribution is subject
to the trademark license, especially commercial redistribution.

START: FULL LICENSE

THE FULL PROJECT GUTENBERG LICENSE
 PLEASE READ THIS BEFORE YOU DISTRIBUTE OR USE THIS WORK

To protect the Project Gutenberg™ mission of promoting the free

distribution of electronic works, by using or distributing this work (or
any other work associated in any way with the phrase “Project
Gutenberg”), you agree to comply with all the terms of the Full
Project Gutenberg™ License available with this file or online at
www.gutenberg.org/license.

Section 1. General Terms of Use and

Redistributing Project Gutenberg™
electronic works
1.A. By reading or using any part of this Project Gutenberg™
electronic work, you indicate that you have read, understand, agree
to and accept all the terms of this license and intellectual property
(trademark/copyright) agreement. If you do not agree to abide by all
the terms of this agreement, you must cease using and return or
destroy all copies of Project Gutenberg™ electronic works in your
possession. If you paid a fee for obtaining a copy of or access to a
Project Gutenberg™ electronic work and you do not agree to be
bound by the terms of this agreement, you may obtain a refund
from the person or entity to whom you paid the fee as set forth in
paragraph 1.E.8.

1.B. “Project Gutenberg” is a registered trademark. It may only be

used on or associated in any way with an electronic work by people
who agree to be bound by the terms of this agreement. There are a
few things that you can do with most Project Gutenberg™ electronic
works even without complying with the full terms of this agreement.
See paragraph 1.C below. There are a lot of things you can do with
Project Gutenberg™ electronic works if you follow the terms of this
agreement and help preserve free future access to Project
Gutenberg™ electronic works. See paragraph 1.E below.
1.C. The Project Gutenberg Literary Archive Foundation (“the
Foundation” or PGLAF), owns a compilation copyright in the
collection of Project Gutenberg™ electronic works. Nearly all the
individual works in the collection are in the public domain in the
United States. If an individual work is unprotected by copyright law
in the United States and you are located in the United States, we do
not claim a right to prevent you from copying, distributing,
performing, displaying or creating derivative works based on the
work as long as all references to Project Gutenberg are removed. Of
course, we hope that you will support the Project Gutenberg™
mission of promoting free access to electronic works by freely
sharing Project Gutenberg™ works in compliance with the terms of
this agreement for keeping the Project Gutenberg™ name associated
with the work. You can easily comply with the terms of this
agreement by keeping this work in the same format with its attached
full Project Gutenberg™ License when you share it without charge
with others.

1.D. The copyright laws of the place where you are located also
govern what you can do with this work. Copyright laws in most
countries are in a constant state of change. If you are outside the
United States, check the laws of your country in addition to the
terms of this agreement before downloading, copying, displaying,
performing, distributing or creating derivative works based on this
work or any other Project Gutenberg™ work. The Foundation makes
no representations concerning the copyright status of any work in
any country other than the United States.

1.E. Unless you have removed all references to Project Gutenberg:

1.E.1. The following sentence, with active links to, or other

immediate access to, the full Project Gutenberg™ License must
appear prominently whenever any copy of a Project Gutenberg™
work (any work on which the phrase “Project Gutenberg” appears,
or with which the phrase “Project Gutenberg” is associated) is
accessed, displayed, performed, viewed, copied or distributed:
 This eBook is for the use of anyone anywhere in the United
States and most other parts of the world at no cost and with
almost no restrictions whatsoever. You may copy it, give it away
or re-use it under the terms of the Project Gutenberg License
included with this eBook or online at www.gutenberg.org. If you
are not located in the United States, you will have to check the
laws of the country where you are located before using this
eBook.

1.E.2. If an individual Project Gutenberg™ electronic work is derived

from texts not protected by U.S. copyright law (does not contain a
notice indicating that it is posted with permission of the copyright
holder), the work can be copied and distributed to anyone in the
United States without paying any fees or charges. If you are
redistributing or providing access to a work with the phrase “Project
Gutenberg” associated with or appearing on the work, you must
comply either with the requirements of paragraphs 1.E.1 through
1.E.7 or obtain permission for the use of the work and the Project
Gutenberg™ trademark as set forth in paragraphs 1.E.8 or 1.E.9.

1.E.3. If an individual Project Gutenberg™ electronic work is posted

with the permission of the copyright holder, your use and distribution
must comply with both paragraphs 1.E.1 through 1.E.7 and any
additional terms imposed by the copyright holder. Additional terms
will be linked to the Project Gutenberg™ License for all works posted
with the permission of the copyright holder found at the beginning
of this work.

1.E.4. Do not unlink or detach or remove the full Project

Gutenberg™ License terms from this work, or any files containing a
part of this work or any other work associated with Project
Gutenberg™.

1.E.5. Do not copy, display, perform, distribute or redistribute this

electronic work, or any part of this electronic work, without
prominently displaying the sentence set forth in paragraph 1.E.1
Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.

More than just a book-buying platform, we strive to be a bridge

connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.

Join us on a journey of knowledge exploration, passion nurturing, and

personal growth every day!

ebookbell.com